The Nature and Origins of Mass Opinion (Cambridge Studies in Public Opinion and Political Psychology)

Citation preview

96

N A T U R E AND ORIGINS OF MASS O P I N I O N

a public that has no fixed attitude t o w a r d w h a t it wants d o n e , but simply a r a n g e of only partially consistent c o n s i d e r a t i o n s , s o m e o n e has got to play the role of crystallizing issues in a way that can lead to a c t i o n . Consider, in this c o n n e c t i o n , D o u g l a s A r n o l d ' s recent b o o k , The Logic of Congressional Action (1990). A r n o l d shows that C o n g r e s s often takes blatantly contradictory actions. It favors a balanced budget w h e n this is the issue, and favors increased spending w h e n asked to vote on p a r t i c u l a r s p e n d i n g provisions. L i k e w i s e , it favors " t a x i n c e n t i v e s " w h e n asked to vote on t h e m one item at a t i m e , and also tax simplification w h e n asked to vote on tax reform. T h e key to these and n u m e r o u s other inconsistencies in C o n g r e s s ' s r e c o r d , according to Arn o l d , is h o w issue entrepreneurs frame t h e given issue, thereby linking it to o n e rather than a n o t h e r potential distribution of public o p i n i o n . S u c c e s s in c o n g r e s sional politics turns on the ability to get o n e ' s colleagues and o t h e r elites, e s pecially in t h e p r e s s , to think about o n e ' s issue in a way that will p r o d u c e majority s u p p o r t for it rather t h a n , as might be the c a s e u n d e r a different issue frame, majority o p p o s i t i o n . A similar a r g u m e n t might be m a d e for leadership in g e n e r a l . Political leaders are s e l d o m t h e passive i n s t r u m e n t s of majority o p i n i o n . Nor, as it s e e m s to m e , do they often a t t e m p t openly to c h a l l e n g e public o p i n i o n . But they do regularly attempt to play on the contradictory ideas that are always present in p e o p l e ' s m i n d s , e l e v a t i n g the salience of s o m e and h a r n e s s i n g t h e m to new initiatives while d o w n p l a y i n g or ignoring other ideas - all of w h i c h is just a n o t h e r way of talking about issue framing. I s u g g e s t , t h e n , that my a c c o u n t of an ambivalent public is not only m o r e faithful to the wealth of m i c r o e v i d e n c e e x a m i n e d e a r l i e r in this chapter, but also m o r e faithful to the c o m p l e x role of public opinion in d e m o c r a t i c politics. All of the analysis in this and the last c h a p t e r involves the form and nature of attitudes r a t h e r than the content of p e o p l e ' s opinion s t a t e m e n t s and the process by which they are formed. T h e next five chapters are c o n c e r n e d with d r a w i n g out the implications of the m o d e l in these latter areas.

The mainstream and polarization effects

W i t h the national inflation rate a p p r o a c h i n g the then-startling level of 7 percent, President N i x o n went on television in late s u m m e r 1971 to a n n o u n c e a surprise decision to i m p o s e w a g e and p r i c e controls on the e c o n o m y . A l t h o u g h such controls w e r e a major d e p a r t u r e from administration policy, t h e decision w a s imm e d i a t e l y hailed by c o m m e n t a t o r s across the political s p e c t r u m as a n e c e s s a r y step in t h e battle against inflation. By g o o d luck, there exist excellent d a t a on the effect of N i x o n ' s s p e e c h on public attitudes. A C o l u m b i a University survey of political activists h a p p e n e d to be in t h e field at t h e t i m e of N i x o n ' s a n n o u n c e m e n t , and G a l l u p surveys on price controls b r a c k e t e d the s p e e c h . T h e C o l u m b i a study found, first of all, that the s p e e c h had little effect on D e m o c r a t i c activists, w h o tended to favor w a g e and price controls even before N i x o n s p o k e . But t h e effect of t h e speech on R e p u b lican activists w a s d r a m a t i c . Virtually o v e r n i g h t , s u p p o r t for controls a m o n g R e p u b l i c a n activists shot up from 37 percent to 82 percent, a rise of s o m e 45 p e r c e n t a g e points ( B a r t o n , 1 9 7 4 - 5 ) . T h e G a l l u p s u r v e y s , m e a n w h i l e , showed that t h e public as a w h o l e b e c a m e about 10 p e r c e n t a g e points m o r e favorable toward price controls in the w e e k s following the N i x o n s p e e c h . T h i s c a s e suggests that a p o p u l a r president b a c k e d by a unified W a s h i n g t o n c o m m u n i t y can have a powerful effect on public o p i n i o n , especially that part of the public that is most attentive to politics. T h i s is the first of a series of c h a p t e r s that a i m s at a c c o u n t i n g for the effects of such elite c o m m u n i c a t i o n s on m a s s attitudes. In this c h a p t e r we e x a m i n e two simple ideal typical s i t u a t i o n s , o n e type in which elites achieve a c o n s e n s u s or near c o n s e n s u s on a value or policy, so that virtually all c o m m u n i c a t i o n s take t h e s a m e side of t h e given issue, and a n o t h e r type in which elites disagree along p a r t i s a n or ideological lines, so that there is a roughly even flow of c o m m u n i c a t i o n s on b o t h sides of t h e issue. T h e c a s e of w a g e and price controls is an e x a m p l e of the first type of situation, and the nearly unified s u p p o r t of A m e r i c a n elites for the w a r in V i e t n a m in 1964 is another. T h e sharply ideological division of elites o v e r V i e t n a m in t h e late 1960s is an e x a m p l e of t h e s e c o n d . T h e R A S m o d e l , as we shall see, leads us to expect that these t w o types of situations will have regular and predictable effects on public attitudes.

98

N A T U R E AND ORIGINS OF MASS O P I N I O N

L a t e r c h a p t e r s will e x a m i n e m o r e c o m p l i c a t e d c a s e s , ones in which the pattern of elite m e s s a g e s switches from mainly consensual to mainly conflictual, and others in which elites are d i v i d e d , but in w h i c h the relative intensity of c o m m u n i c a t i o n s c h a n g e s o v e r t i m e . Such c h a n g e s in the flow of elite c o m m u n i c a tions p r o d u c e quite interesting and nonintuitive patterns of c h a n g e in m a s s attitude r e p o r t s , as will b e c o m e apparent.

MAINSTREAM

EFFECT

W h a t , we m a y now a s k , would be t h e theoretically expected effect on public opinion if elites across the political s p e c t r u m were to achieve a c o n s e n s u s in s u p port of a p a r t i c u l a r " m a i n s t r e a m " policy? Or, to ask the same question in the l a n g u a g e of the m o d e l : W h a t w o u l d be t h e e x p e c t e d effect on public opinion if virtually all the persuasive m e s s a g e s c a r r i e d in political media on a p a r t i c u l a r policy were favorable to that policy, and if there were no cueing m e s s a g e s to alert p e o p l e that the policy w a s inconsistent with their values?

Mainstream

and

polarization

effects

99

eign policies (see also S i g e l m a n and Conover, 1981). T h e i r explanation for this paralleled K e y ' s a r g u m e n t . Information m e a s u r e d " o n e ' s a t t a c h m e n t t o the m a i n s t r e a m and the resultant exposure t o influences such a s the m a s s m e d i a " (1966: p. 189). M c C l o s k y and Brill's (1983) a r g u m e n t that e d u c a t i o n p r o m o t e s the " s o c i a l l e a r n i n g " of libertarian ideals, and M u e l l e r ' s (1973) claim that better e d u c a t e d p e r s o n s were m o r e likely to s u p p o r t the V i e t n a m W a r because they were better " f o l l o w e r s " of official policy likewise appeal to the notion that exposure to " m a i n s t r e a m " values tends to e n h a n c e s u p p o r t for t h e m . M o r e recently, the t e n d e n c y of better e d u c a t e d p e r s o n s to be m o r e o p p o s e d to the quarantining of A I D S v i c t i m s ( S n i d e r m a n et a l . , 1991) appears to reflect the internalization of a m e d i c a l c o n s e n s u s that such action is u n n e c e s s a r y to prevent the spread of the disease (Colby and C o o k , 1 9 9 1 ) . '

Noting that e d u c a t i o n w a s associated with greater support for racial equality, private health i n s u r a n c e , and tolerance of n o n c o n f o r m i s t s , Key explained that " f o r m a l e d u c a t i o n may serve to indoctrinate p e o p l e into the more-or-less official political values of the c u l t u r e " (p. 3 4 0 ) .

In a c o m p a r a t i v e study of the United S t a t e s and Britain, C a i n , Ferejohn, and Fiorina (1987) turned up a finding that nicely illustrates the " i n d o c t r i n a t i n g eff e c t " that e x p o s u r e to a p a r t i c u l a r elite culture often p r o d u c e s . Citizens in both c o u n t r i e s w e r e asked w h e t h e r elected representatives should " s u p p o r t the position their p a r t i e s take w h e n s o m e t h i n g c o m e s up for a v o t e , or should they m a k e up their o w n m i n d s r e g a r d l e s s of h o w their p a r t i e s want t h e m to v o t e . " In B r i t a i n , w h e r e P a r l i a m e n t d e p e n d s on a high degree of p a r t y discipline, collegee d u c a t e d p e r s o n s were m o r e likely than those with only high school e d u c a t i o n to say that representatives should h e w the p a r t y line. But in the United S t a t e s , with its a n t i p a r t y and individualist political tradition, c o l l e g e - e d u c a t e d persons were m o r e likely to say that representatives should vote their o w n opinions. T h u s , the better e d u c a t e d in each c o u n t r y are the m o r e faithful adherents of their c o u n t r y ' s r e s p e c t i v e national traditions. If the m a i n s t r e a m a r g u m e n t is c o r r e c t , c o r r e l a t i o n s b e t w e e n a w a r e n e s s and s u p p o r t for a policy should be strongest w h e n elite c o n s e n s u s is strongest and less strong w h e n elite c o n s e n s u s is less strong or nonexistent (D25). M u c h p u b lished e v i d e n c e (to be s u p p l e m e n t e d later in this chapter) s u p p o r t s this e x p e c tation. For e x a m p l e , M u e l l e r notes that the correlation b e t w e e n e d u c a t i o n and s u p p o r t for the V i e t n a m W a r w a s strong early in the war, w h e n most elites supp o r t e d it, and weak in the late p h a s e s of the war, w h e n p a r t y and ideological elites b e c a m e deeply divided. In a s y s t e m a t i c test of this hypothesis in the d o m a i n of civil l i b e r t i e s , M c C l o s k y and Brill ( 1 9 8 3 : p. 4 2 1 ) classified m o r e than 100 civil liberties items according to the d e g r e e of support for the libertarian option in relevant S u p r e m e C o u r t decisions and in the attitudes of s o m e 2 , 0 0 0 elites they had s u r v e y e d . T h e y found that for items on which the C o u r t and other elites had strongly e n d o r s e d the civil liberties position, m e m b e r s of the general public w h o had attended college w e r e , on a v e r a g e , 24 p e r c e n t a g e points more libertarian than w e r e those with less than a high school e d u c a t i o n . Yet e d u c a t i o n had a progressively w e a k e r effect in inducing support for libertarian policies as elite s u p p o r t for t h e m d e c l i n e d , until finally, e d u c a t i o n had a slightly n e g a t i v e

W r i t i n g a few years later, G a m s o n and Modigliani (1966) noted a substantial correlation b e t w e e n political information and support for the g o v e r n m e n t ' s for-

1 This is my interpretation of the education effect reported by Sniderman, Brody, and Tetlock (1991, chap. 4). For a further discussion of this point, see Chapter 12.

A x i o m Al s u g g e s t s , first of all, that t h e greater a c i t i z e n ' s level of political a w a r e n e s s , the greater the likelihood of reception of persuasive m e s s a g e s on this hypothetical m a i n s t r e a m issue. If all of the cueing m e s s a g e s on this policy were favorable, no o n e w o u l d have any basis via A2 for resisting it. F r o m this we can d e d u c e that the greater a p e r s o n ' s level of political a w a r e n e s s , the greater the n u m b e r of m a i n s t r e a m m e s s a g e s t h e person would internalize in the form of considerations and h e n c e , all else e q u a l , the greater the p e r s o n ' s level of expressed support for the m a i n s t r e a m policy (D25). R e s e a r c h e r s w o r k i n g on a variety of substantive p r o b l e m s have r e p o r t e d s u p p o r t for this implication of the m o d e l . In fact, t h o u g h using different v o c a b u l a r i e s , several have m a d e roughly the s a m e a r g u m e n t as h e r e . For exa m p l e , in Public Opinion and American Democracy (1961), V. O. Key, Jr., w r o t e that a p e r s o n ' s level of formal e d u c a t i o n may be an indicator of the extent to which the person has been influenced by s o c i e t y ' s traditional or " o f f i c i a l " values. Key w r o t e : Probably a major consequence of education for opinion consists in the bearing of education on the kinds of influences to which a person is subjected throughout his life. The more extended the educational experience, the more probable it is that a person will be exposed to the discussions of issues as they arise. When, as so often occurs, the current discussion is heavily loaded on one side, it might be expected that this educationally conditioned exposure would have some bearing on the direction of opinion. (1961: p. 341)

100

N A T U R E AND ORIGINS OF MASS O P I N I O N

association with s u p p o r t for civil liberties on those (few) items on w h i c h the pre-Rehnquist C o u r t and most elites took an antilibertarian position (for exa m p l e , civil d i s o b e d i e n c e ) . It is w i d e l y supposed that political awareness - w h e t h e r m e a s u r e d by k n o w l e d g e , p a r t i c i p a t i o n , or e d u c a t i o n - e n g e n d e r s resistance to elite influence rather than, as a s s u m e d in the m a i n s t r e a m m o d e l , susceptibility to it. As will b e c o m e clear in C h a p t e r s 7 - 1 1 , this supposition has s o m e validity. Political a w a r e n e s s does a p p e a r to e n g e n d e r resistance to the political c o m m u n i c a t i o n s of g o v e r n i n g authorities. But awareness does so less by e n g e n d e r i n g resistance per se than by increasing t h e p e r s o n ' s sensitivity to the c o m m u n i c a t i o n s of countervailing elites, especially the ideological o p p o n e n t s of the r e g i m e . T h u s , for e x a m p l e , it will turn out in C h a p t e r 9 that a major source of opposition to t h e V i e t n a m W a r was the e x p o s u r e of politically a w a r e citizens to a n t i w a r c o m m u n i c a t i o n s that were t o o faint to be picked up by the less a w a r e . T h e notion that politically aware persons resist all forms of political persuasion is highly dubious. O n e other c o m m e n t . T h e r e are in every society ideas on w h i c h virtually eve r y o n e agrees. In such c a s e s , the idea is unlikely to b e c o m e the object of studies of public o p i n i o n , except perhaps in studies of c u l t u r e . S u c h " m o t h e r h o o d iss u e s " in t h e United States m i g h t include m a i n t e n a n c e of free e l e c t i o n s , taxs u p p o r t e d public s c h o o l s , and state-organized a t t e m p t s to repulse an invading e n e m y . T h e m a i n s t r e a m model is less useful for policies of this type than for policies on which there is p o p u l a r reluctance to go along with an elite c o n s e n s u s , such as tolerance of disliked g r o u p s , or support for w a r w h e n the nation is not immediately threatened. 2

Mainstream

and

polarization

effects

101

to see the ideological implications of the m e s s a g e s they r e c e i v e , they will be likely to reject c o n s e r v a t i v e a r g u m e n t s on this issue; these c u e i n g m e s s a g e s will not, however, i m p e d e their internalization of liberal m e s s a g e s . Less politically aware liberals, by contrast, will be exposed to few persuasive m e s s a g e s , a n d , owing to their low reception of cueing m e s s a g e s and the lower accessibility of these cues in m e m o r y , will be less selective about the persuasive m e s s a g e s they internalize. In c o n s e q u e n c e of this d y n a m i c , the most a w a r e liberals will fill their h e a d s , so to speak, with a large n u m b e r of c o n s i d e r a t i o n s that a r e , on b a l a n c e , favorable to the liberal side of the issue. Less aware liberals, for their p a r t , will fill their h e a d s with a smaller n u m b e r of c o n s i d e r a t i o n s , and these c o n s i d e r a t i o n s will not consistently favor the liberal side of the issue. T h e s a m e a r g u m e n t , mutatis mutandis, applies to c o n s e r v a t i v e s . Highly a w a r e c o n s e r v a t i v e s should fill their heads with mostly c o n s e r v a t i v e c o n s i d e r a t i o n s , while less a w a r e c o n s e r v a t i v e s should fill their heads with a smaller n u m b e r of c o n s i d e r a t i o n s that are less consistently c o n s e r v a t i v e . O u r e x p e c t a t i o n , t h e n , is that for cases in which there is a roughly even flow of o p p o s i n g p a r t i s a n m e s s a g e s , the ratio of ideologically consistent c o n siderations to ideologically inconsistent o n e s should increase as political awareness increases. Figure 4 . 1 h a s already c o n f i r m e d this e x p e c t a t i o n . As s h o w n there, the r a t i o of consistent c o n s i d e r a t i o n s to total c o n s i d e r a t i o n s increases from about .5 a m o n g the least informed persons to about .80 a m o n g the m o s t informed. Two of the slopes in Figure 4 . 1 are statistically significant at the .01 level and the third is significant at the .10 l e v e l . O n e m a y expect that an increasing r a t i o of ideologically consistent to inconsistent c o n s i d e r a t i o n s should translate into differences in p e o p l e ' s attitude statem e n t s : M o r e a w a r e liberals will be m o r e likely to call to mind c o n s i d e r a t i o n s favorable to the liberal position and h e n c e will be m o r e likely to s u p p o r t it. Less aware liberals will be less likely to be able to recall c o n s i d e r a t i o n s of any k i n d , which will lead to higher no-opinion r a t e s , and less likely to endorse the liberal position w h e n they do offer an o p i n i o n . 3

THE POLARIZATION EFFECT T h e r e a r e , of c o u r s e , many cases in which political elites heatedly d i s a g r e e , so that no " m a i n s t r e a m " exists. In cases of this t y p e , the R A S m o d e l leads us to expect quite different patterns of m a s s attitudes. To see why, let us a s s u m e a situation in which elites are roughly evenly divided on a p a r t i s a n issue, with o n e p a r t i s a n c a m p s p o n s o r i n g persuasive m e s sages favoring t h e liberal position and the other s p o n s o r i n g m e s s a g e s in support of the c o n s e r v a t i v e position. We further a s s u m e that each c a m p sponsors cueing m e s s a g e s indicating why the given policy is or is not consistent with liberal (or c o n s e r v a t i v e ) values. Finally, let us a s s u m e that all of these m e s s a g e s are equally intense in that a person at a given level of political a w a r e n e s s w o u l d be equally likely to e n c o u n t e r and take in any one of t h e m . Within the general p u b l i c , increases in awareness will lead to increased reception of persuasive m e s s a g e s favoring both the liberal position and the c o n servative position (from A l ) and also increased reception of c u e i n g m e s s a g e s c o n c e r n i n g t h e issue. Let us focus first on how this affects liberals. Since p o litically a w a r e liberals will be likely to possess cueing m e s s a g e s that e n a b l e t h e m 2 See also Chong, McClosky, and Zaller, 1984.

4

T h e logic of this a r g u m e n t again applies equally to c o n s e r v a t i v e s . T h a t is, increases in awareness m a k e m a s s c o n s e r v a t i v e s increasingly likely to m a k e conservative attitude s t a t e m e n t s w h e n asked about the issue. T h u s , in the c a s e of an evenly divided p a r t i s a n elite and a balanced flow of partisan c o m m u n i c a t i o n s , the effect of political a w a r e n e s s is to p r o m o t e the 3 The relationships depicted in Figure 4.1, however, apply to the sample as a whole; closer inspection of the data reveals that the expected relationships hold only for liberals, where they hold very strongly. For conservatives, there appears to be little change in the ratio of consistentto-inconsistent considerations as awareness increases. The reason for this complication appears to be that the assumed conditions for the test have not been met, namely, a roughly even division of elite support for the opposing policy alternatives. For none of the three options is the division of mass opinion close to 50-50, as it ought to be in the case of an equal elite division. See Chapter 8 for further tests of the effect of awareness and ideology on the internalization of considerations. 4 See Krosnick and Milburn, 1990, for a review of the evidence on the effects of political awareness on no-opinion rates.

102

N A T U R E AND O R I G I N S OF MASS O P I N I O N

polarization of attitude r e p o r t s as m o r e aware liberals gravitate m o r e reliably to the liberal position and m o r e aware c o n s e r v a t i v e s gravitate m o r e reliably to the c o n s e r v a t i v e position (D26).

Mainstream

support for

the

polarization

effect

M u c h empirical e v i d e n c e s u p p o r t s the e x p e c t a t i o n of an awareness-induced p o larization of liberals and c o n s e r v a t i v e s on p a r t i s a n issues. T h e polarizing effect of political a w a r e n e s s on p a r t i s a n (as against m a i n s t r e a m ) issues w a s first noted by G e o r g e B e l k n a p and A n g u s C a m p b e l l ( 1 9 5 1 - 2 ) and w a s incorporated into the M i c h i g a n s c h o o l ' s classic, The American Voter (Campbell et a l . , 1960: p p . 186, 2 0 7 ) . U s i n g different theoretical v o c a b u l a r i e s , G a m s o n and Modigliani (1966) and C h o n g , M c C l o s k y , and Zaller (1984) have noted the same effect. T h e y exa m i n e public attitudes toward n u m e r o u s issues on which elites d i s a g r e e , issues r a n g i n g from foreign policy to civil liberties to welfare to race to e c o n o m i c policy. In each c a s e , increases in political a w a r e n e s s were associated with a s h a r p e r p o l a r i z a t i o n of attitudes b e t w e e n liberals (or D e m o c r a t s ) , on o n e s i d e , and conservatives (or R e p u b l i c a n s ) , on the o t h e r . 5

T h e d a t a in Figure 6.1 illustrate both the m a i n s t r e a m and p o l a r i z a t i o n effects of political a w a r e n e s s . W h e n , in 1964, A m e r i c a n elites nearly all s u p p o r t e d the V i e t n a m War, increases in a w a r e n e s s led nonelite liberals and c o n s e r v a t i v e s to b e c o m e m o r e s u p p o r t i v e of the " m a i n s t r e a m " war policy. Yet w h e n , in 1970, A m e r i c a n elites had b e c o m e d e e p l y divided about the war, increases in awareness are a s s o c i a t e d with greater p o l a r i z a t i o n of the attitudes of m a s s liberals and c o n s e r v a t i v e s . T h e Persian G u l f W a r affords a n o t h e r o p p o r t u n i t y to o b s e r v e both the mainstream and the p o l a r i z a t i o n effect. F r o m the Iraqi invasion of K u w a i t in A u g u s t 1990 through t h e fall 1990 congressional e l e c t i o n , there w a s only light criticism of President B u s h ' s h a n d l i n g of the crisis a n d , in particular, virtually no a r t i c ulate opposition to the policy of sending U . S . forces to the region. T h u s , as J. W. A p p l e w r o t e on the eve of the e l e c t i o n , 6

[A] midterm election campaign has taken place with war threatening in the Persian Gulf, and . . . the major foreign policy issue confronting the nation has generated almost no debate among the candidates about what the U.S. should do. Instead, President Bush has traded insults with Saddam Hussein of Iraq, and the Democrats have barely mentioned the subject. (New York Times, 6 November 1991, p. A l ) 5 In Gamson and Modigliani, these findings are the basis for a "cognitive consistency" model of opinion formation; in McClosky et al., they are the basis for a "contested norms" model of opinionation. Yet in both cases, the empirical regularity being explained, as well as the operational constructs in the models, are the same as in the Belknap and Campbell polarization model. 6 To validate these claims concerning elite consensus and division, I asked a research assistant to classify cover stories on Vietnam in Newsweek and Time. In 1964 prowar cover stories outnumbered antiwar ones by a margin of approximately 3 to 1; in 1970, the ratio was close 1 to 1. (See also Hallin, 1986.)

polarization

103

effects

1970

1964 100

100

Conservatives

Conservatives

80 Empirical

and

Percent 60 supporting war 40

Liberals Liberals

20 0 Political awareness Cons. N

42 25

53 32

71 65

102 33

86 51

40 12

53 19

120 30

114 41

153 73

Figure 6 . 1 . An illustration of the mainstream and polarization models. Liberals are defined as persons who rated liberals fifteen or more points higher than conservatives on separate 100-point feeling thermometers; conservatives are persons who exhibited the reverse pattern. Persons supporting the war are those who said either that the United States should "keep troops in Vietnam but try to end the fighting," which was the position of both the Johnson and Nixon administrations, or that the U.S. should take a stronger stand on the war. The awareness measure is described in the Measures Appendix. Source: 1964 and 1970 CPS surveys. Two days after the e l e c t i o n , however, Bush a n n o u n c e d a d e c i s i o n to send several h u n d r e d t h o u s a n d additional troops to the gulf. T h i s decision s p a r k e d strong congressional c r i t i c i s m , leading to congressional h e a r i n g s in which administration policy w a s harshly criticized and later to a congressional vote on a w a r policy resolution. As in the V i e t n a m c a s e , D e m o c r a t s w e r e the most salient critics of the a d m i n i s t r a t i o n ' s h a w k i s h policies and R e p u b l i c a n s were t h e most salient defenders. In v i e w of t h i s , we should expect, in the period before C o n g r e s s reacted critically to B u s h ' s troop a n n o u n c e m e n t , to find e v i d e n c e of the m a i n s t r e a m effect; after criticism b e g a n , we should expect to o b s e r v e the p o l a r i z a t i o n p a t t e r n . By g o o d luck, the 1990 National E l e c t i o n Study went into t h e field on the day after the election and w a s able to c o m p l e t e m o r e than 2 5 0 i n t e r v i e w s before congressional c r i t i c i s m of B u s h ' s military buildup b e g a n . It also c a r r i e d a question a s k i n g w h e t h e r " w e did the right thing in sending U . S . military forces to the Persian Gulf, or should we have stayed o u t ? ' ' R e s u l t s , which are d e r i v e d from a m a x i m u m likelihood logistic regression that controls for political a w a r e n e s s , p a r t y a t t a c h m e n t , gender, r a c e , and Jewish ethnicity are s h o w n in Figure 6 . 2 . ( T h e coefficients on which the figure is b a s e d are in Table 6 . 3 of the a p p e n d i x to this chapter.) As e x p e c t e d , the d a t a 7

7 Democrats and Republicans in the figure are constructed as persons with a score of ± 1.3 on the party variable, where party ranges from -2 (strong Republican) to +2 (strong Democrat). Awareness scores in the figure run from —1.8 SD to +2.57 SD.

104

N A T U R E AND ORIGINS OF MASS O P I N I O N

Before congressional criticism 100 T 75

Percent who say it was "right thing" 50 to send troops to Gulf 25

Mainstream

and

polarization

effects

105

After congressional criticism 100y

Republicans

Republicans Democrats

Political awareness

i Figure 6.2. Partisans become more polarized over time on "right to send troops." Estimates are derived from coefficients in Table 6.3. Source: 1990 NES survey.

betray little e v i d e n c e of p a r t i s a n p o l a r i z a t i o n in the period prior to t h e c o n g r e s sional c r i t i c i s m of B u s h ' s policies, but clear p o l a r i z a t i o n afterward. Public division d o e s not a p p e a r to be as s h a r p as in the V i e t n a m c a s e , but this is probably b e c a u s e elite polarization on Persian Gulf policy did not a p p r o a c h that of the V i e t n a m p e r i o d in t e r m s of either d u r a t i o n or intensity. Before c o n t i n u i n g the analysis of opinion on Persian G u l f policy, I must discuss a m e t h o d o l o g i c a l issue. In c r e a t i n g Figure 6 . 2 from the coefficients in a logistic regression m o d e l , I had to m a k e certain c o d i n g decisions. For e x a m p l e , to show the effects of political a w a r e n e s s , I m a n i p u l a t e d scores from roughly the 1st percentile on political a w a r e n e s s to the 98th percentile. Since I need to m a k e many similar decisions about how to c r e a t e g r a p h s from coefficients in the next several c h a p t e r s , I want to s t a n d a r d i z e my p r o c e d u r e s in an intelligible set of c o n v e n t i o n s . A s u m m a r y of these c o n v e n t i o n s , w h i c h will be used for the rest of the b o o k , is given in the a c c o m p a n y i n g b o x . T h e 1990 N E S survey c a r r i e d o n e other q u e s t i o n w h i c h is useful for g a u g i n g public opinion on the gulf crisis. It reads: Which of the following do you think we should do now in the Persian Gulf:

Conventions of graphical analysis For graphs showing the relationship between political awareness, political predispositions, and a political attitude or attitude change, the following conventions will apply in the remainder of the book: Basic design. In all cases, political awareness will be treated as the principal independent variable and plotted against the jr-axis. The dependent variable, usually the probability of a political attitude or attitude change, the will be plotted against the y-axis, as in Figure 6.2. The effect of differences in political dispositions (such as, being a Democrat rather than a Republican) will be shown by separate lines within the graphs, as in Figure 6.2. Range of political awareness. Except as noted, graphs depict the simulated effect of moving from about the 1st percentile to about the 98th percentile on political awareness. This range leaves about 1 percent of the cases outside each endpoint, though, of course, lumpiness in the data makes it impossible to achieve this range in every case. Because different awareness scales have different skews, the range of political awareness scores will not always correspond to a particular z-score range, such as ±2 SD. The particular z-score ranges used in the simulations will be provided in footnotes. Range of simulated attitude scores. With one clearly noted exception, graphs showing probabilities or proportions will use a scale of 0 to 1.0. When means are used, graphs will reflect the range of mean values in the data. Thus, in the case of means, the ranges can vary from figure to figure. However, unless explicitly noted, identical scales will be used in figures that are being compared to one another. Range of predispositional variables. Throughout the analysis, party attachment is coded from -2 (strong Republican) to -1 (weak or independent Republican) to +1 (weak or independent Democrat) to +2 strong Democrat, with all others assigned to the score of zero. In graphs that depict the effect of being a Republican or Democrat, partisans are simulated by scores of either —1.3 or + 1.3, as appropriate. The effects of other predispositional measures (such as equalitarianism, hawkishness) are simulated differently in different cases, depending on how many measures are available for use in a given model. For example, if only one measure is used in a model, the range may be ±2 SD for that variable; if three measures are used, their joint effect - that is, the effect of identical movements on all three variables - will be depicted over a smaller range. The exact values are provided in each case. The aim will be to approximate the raw data, insofar as the raw data can be directly observed.

Pull out U.S. forces entirely. Try harder to find a diplomatic solution. Tighten the economic embargo. Take tougher military action. All but the first of these o p t i o n s imply support for the basic United States policy of military involvement in the Persian Gulf. S i n c e , with the possible exception of the congressional Black C a u c u s , virtually all of B u s h ' s elite critics accepted this policy, we should expect to find that, a m o n g the p u b l i c , political

a w a r e n e s s is associated with g r e a t e r s u p p o r t for keeping U . S . forces in the G u l f - w h i c h is to say, greater levels of rejection of the " p u l l o u t " option in favor of o n e of t h e o t h e r three r e s p o n s e a l t e r n a t i v e s . T h i s e x p e c t a t i o n is c o n firmed in Figure 6 . 3 (see Table 6 . 3 , the c h a p t e r a p p e n d i x , for coefficients). E v e n within the g r o u p most resistant to using military action against Iraq black D e m o c r a t i c w o m e n - rejection of the pullout option rises from about 54 percent in the lowest a w a r e n e s s c a t e g o r y to about 92 percent in the highest

106

Mainstream

N A T U R E AND ORIGINS OF MASS O P I N I O N

100 T

50

polarization

When Bush more conciliatory

Modal Republicans

75 Percent who favor option other than pulling out U.S. troops

and

effects

107

When Bush more threatening

100

Black Democratic women

Percent who ^ favor tougher military action 50 against Iraq

Modal Democrats

25

25 0 Political awareness Political awareness Figure 6.3. Support for keeping U.S. troops in the Persian Gulf. Estimates are derived from coefficients in Table 6.3. Source: 1990 NES survey. 8

c a t e g o r y . T h e trends in Figure 6 . 3 were about the s a m e throughout the period of the s u r v e y . It is difficult to be certain w h a t to expect from the three response o p t i o n s m o r e d i p l o m a c y , a tighter e m b a r g o , and t o u g h e r military action - that I have c o u n t e d as i m p l y i n g s u p p o r t for military involvement in t h e Gulf. Certainly, D e m o c r a t s en m a s s e wouid be e x p e c t e d to reject the c h o i c e of " t o u g h e r military a c t i o n , " since t h e D e m o c r a t i c p a r t y in C o n g r e s s w a s clearly identified with o p position to this idea. T h e p r o b l e m is that it is not c l e a r that R e p u b l i c a n s should be e x p e c t e d to e m b r a c e it. For B u s h ' s public position, especially in the early m o n t h s of the c r i s i s , w a s that an e m b a r g o , in c o m b i n a t i o n with skilled diplomacy, w o u l d m a k e it possible to avoid the use of force. In m i d - D e c e m b e r , however, the Bush administration rejected a proposed J a n u a r y 12th m e e t i n g in B a g h d a d on the g r o u n d s that it w a s t o o near the United N a t i o n s deadline for Iraqi w i t h d r a w a l from K u w a i t to be useful for averting military action. By that point, therefore, it w a s clear at least that willingness to use force w a s a key feature of Bush policy. O n e m i g h t therefore expect that at about that t i m e o p p o s i n g partisan g r o u p s in t h e public b e c a m e increasingly polarized over the q u e s t i o n of military force. 9

Figure 6 . 4 a p p e a r s to s u p p o r t this e x p e c t a t i o n . H i g h l y aware D e m o c r a t s and R e p u b l i c a n s were apparently m o r e polarized over the use of force after D e c e m b e r 15 than they had been before. D e s p i t e this, however, the increase in p o 8 In separate regressions for blacks and whites, political awareness is associated with rejection of the pull-out option at least as strongly among blacks as among whites. (In a simple linear regression of this question [scored 0-1] on political awareness, the intercept and slope for blacks are .55 and .047, respectively; for whites, the intercept and slope are .78 and .022, with all terms highly statistically significant; the range of political awareness is 0 through 13.) See Chapter 9 for additional discussion of the effects of elite opinion leadership on Afro-Americans. 9 Insofar as there was a time trend, it was toward less party polarization, but the trend did not approach either statistical or substantive significance.

Figure 6.4. Partisans become more polarized over time on use of military force. Estimates are derived from coefficients in Table 6.3. Source: 1990 NES survey. larization does not achieve statistical significance and must therefore be taken as equivocal s u p p o r t for my e x p e c t a t i o n (coefficients s h o w n in Table 6 . 3 ) . T h e key point here is that exposure to public affairs, as m e a s u r e d by tests of political a w a r e n e s s , has i m p o r t a n t effects on m a s s a t t i t u d e s , but that these effects differ across policies and across t i m e , d e p e n d i n g on the positions taken by political elites and reflected in the m a s s m e d i a . A w a r e n e s s is associated with s u p p o r t for those a s p e c t s of g o v e r n m e n t policy that have the consensual s u p p o r t of political and m e d i a e l i t e s , but is a s s o c i a t e d with higher levels of polarization over policies on which elites are d i v i d e d . 1 0

To d e m o n s t r a t e this point m o r e rigorously, I selected items from the 1972— 7 4 - 7 6 N E S survey that s e e m e d o n their face t o exemplify m a i n s t r e a m policies and p a r t i s a n policies of the early 1 9 7 0 s . " Table 6.1 c o n t a i n s a list of these items. S e l e c t i o n of t h e items w a s b a s e d on my j u d g m e n t of the positions of liberal and c o n s e r v a t i v e elites, political p a r t y elites, and the m a s s m e d i a at the t i m e of t h e N E S study. To confirm these j u d g m e n t s , I a s k e d a research assistant to read t h e platforms of the D e m o c r a t i c and Republican p a r t i e s in 1972, and to rate each p a r t y on each issue. T h e research assistant w a s u n a w a r e of my e x p e c tations and d i d the r a t i n g s on the basis of instructions that were conveyed in w r i t i n g . I expected that both p a r t i e s w o u l d explicitly e n d o r s e the policies I had identified as m a i n s t r e a m policies, and w o u l d take sharply o p p o s i n g positions on policies I had identified as p a r t i s a n policies. T h e s e e x p e c t a t i o n s were largely 1 2

10 In showing that the public became more polarized in its attitudes toward Vietnam and Persian Gulf policy, I have been, in effect, examining mass opinion change. The actual patterns of change occurring in these cases are, however, considerably more complicated than I have been able to demonstrate in this initial treatment. For further examination of opinion change on Vietnam, see Chapter 9; for a treatment of opinion change on Gulf war policy along the lines sketched in Chapter 7, see Zaller (1992). 11 I used the panel data because this test was part of a study to test the comparative ability of political knowledge, education, political interest, media exposure, and political participation to specify relationships that a good measure of political awareness should specify. It turned out that political knowledge outperformed all of the alternative measures (see Zaller, 1990). 12 The written instructions are available from the author upon request.

108

N A T U R E AND ORIGINS OF MASS O P I N I O N

Mainstream

and

polarization

effects

109

T h e m o d e l used to e s t i m a t e the effect of a w a r e n e s s on each of these policy Table issues

6.1.

Question

stems for

opinions

on

mainstream

and partisan

items w a s as follows: Prob(Lib. Response) = Prob(Opinionation) x Prob(Lib. I Opinionation)

Mainstream issues This country would be better off if we just stayed home and did not concern ourselves with problems in other parts of the world. Do you think that mainland China should be a member of the United Nations, or do you think it should not? Should farmers and businessmen be allowed to do business with Communist countries or should they be forbidden to do business with Communist countries? Should the government support the right of black people to go to any hotel or restaurant they can afford, or should it stay out of this matter? Recently there has been a lot of talk about women's rights. Some people feel that women should have an equal role with men in running business, industry, and government. Others feel that women's place is in the home. Partisan issues There is much discussion of the best way to deal with racial problems. Some people think achieving racial integration of schools is so important that it justifies busing children to schools out of their neighborhoods. Others think letting children go to their neighborhood schools is so important that they oppose busing. Some people feel the government in Washington should see to it that every person has a job and a good standard of living. Others think the government should just let each person get ahead on their own. Do you think we did the right thing in getting into the fighting in Vietnam or should we have stayed out? Source: 1972 NES survey.

c o n f i r m e d . T h e o n e exception involved an item about w h e t h e r the federal gove r n m e n t should g u a r a n t e e blacks t h e right to equal treatment in hotels and restaurants. A provision on equal a c c o m m o d a t i o n s w a s a key p a r t of 1964 Civil Rights Act, w h i c h , S e n a t o r B a r r y G o l d w a t e r n o t w i t h s t a n d i n g , passed the C o n gress with majority s u p p o r t from both the D e m o c r a t i c and Republican p a r t i e s . T h e D e m o c r a t i c platform, as I had e x p e c t e d , explicitly e n d o r s e d this policy, but the R e p u b l i c a n Party, although professing general s u p p o r t for equal r i g h t s , m a d e no direct reference to it. I c o n t i n u e , in light of the b i p a r t i s a n history of the Civil Rights Act and the fact that even S o u t h e r n opposition to it had collapsed by 1972, to c o n s i d e r equal a c c o m m o d a t i o n s in hotels and restaurants a m a i n s t r e a m government policy. 13

13 In addition to the items in Table 6.1,1 asked my research assistant to rate an item on whether the government should act against inflation. There was strong endorsement of this principle by both parties, but a ceiling effect on mass support for the policy prevented a test on the effect of political awareness on support for this idea.

T h a t i s , t h e probability of a liberal r e s p o n s e is the probability of offering any opinion at all, t i m e s t h e probability of m a k i n g a liberal r e s p o n s e , given that an opinion has b e e n offered. T h e t w o p a r t s of the m o d e l have been estim a t e d separately. T h e probability of a liberal o p i n i o n , given that an opinion s t a t e m e n t has been m a d e , h a s b e e n m o d e l e d as a logit function of a w a r e n e s s , ideological selfd e s i g n a t i o n , p a r t y identification, and standard d e m o g r a p h i c variables ( r a c e , a g e , i n c o m e , and r e s i d e n c e in a S o u t h e r n state). In addition to these v a r i a b l e s , the initial specification of each e q u a t i o n contained an interaction t e r m for A w a r e n e s s x Ideology and A w a r e n e s s X Party. T h i s e q u a t i o n w a s e s t i m a t e d separately for each of the five m a i n s t r e a m and three p a r t i s a n issues. To m a x i m i z e c o m p a r a b i l i t y of results across different item f o r m a t s , each item w a s coded to a three-point scale r u n n i n g from 0 to 0 . 5 to 1 . 0 . 14

15

T h e e x p e c t a t i o n from the m o d e l is that for p a r t i s a n policies, the t w o A w a r e ness x Values interaction t e r m s will be s t r o n g , but that for m a i n s t r e a m policies these interactions will be a n e m i c . T h e s e c o n d e x p e c t a t i o n is that awareness will have an i m p o r t a n t positive impact on s u p p o r t for m a i n s t r e a m policies. T h e first of these e x p e c t a t i o n s is largely c o n f i r m e d . T h e Ideology x A w a r e ness t e r m gets coefficients that are large for t h e three p a r t i s a n issues and trivial for t h e five m a i n s t r e a m issues, exactly as e x p e c t e d . T h e P a r t y x A w a r e n e s s t e r m b e h a v e s erratically, but its coefficients are either statistically insignificant or t o o small to have m u c h i m p a c t , thus leaving the ideology interaction t e r m to d o m i n a t e the results. Let us l o o k first at results for t h e three p a r t i s a n issues. T h e coefficients for the three p a r t i s a n issues are s h o w n in the left-hand side of Table 6 . 2 , and a graphical analysis of these coefficients is s h o w n in the top half 14 The question asked respondents to place themselves on a seven-point scale that ran from "extremely liberal" to "liberal" to "slightly liberal" to "moderate, middle of the road" to "extremely conservative." The question was asked in all three waves of the survey, and in the test reported below, responses over all three waves were averaged. People who gave no opinion in one year were assigned their average for the other two years; people who gave a response in only one year were assigned their response from that year. People who gave a no-opinion response all three times were assigned to the sample average. This way of including respondents with missing data would be expected to produce differences in item reliabilities across different respondents, but since this difference is constant across all dependent variables, and since the key hypothesis involves differences in the effect of ideology across different items, it would not be expected to produce biased results. Omitting respondents with any missing data would, on the other hand, undermine ability to detect the effect of awareness on support for mainstream policies, since the people omitted would be mainly less-informed persons. 15 When the original item was an agree/disagree item, "in between" responses were coded to .5 and other responses were coded zero or one. The jobs and women's rights items were originally seven-point scales; 4 was coded to .5 and the other points were coded to zero or one. Busing was also originally a seven-point scale, but it was so skewed in the antibusing direction that it was necessary to transform it; the far conservative position, which contained 68 percent of all respondents, was coded to zero, the next most conservative position was coded to .5, and the remaining five scale points were coded to one.

3 £

3 I- H

w -n

3

3

§

3 EFT

>

O

X

3: e EN P O 3 M «< J -P A, »= S. S

~

g-OSS' " 3

3

C. ET O £ FT O -I 3* PO FT

FT RT FT 2 D. 00 A, O CO 3 P A. FT ^ PO O 3

E

o

0

a

3 3. 3 2. 3 3 " ET FT

a a.

P

^ .

1 3*/ ~ ~ FT R«. E T E T _ 3 C5 / FT O EE. FT O 5 S P1 FT § RT RT B. 3 „ 3 3 I" 3 5 3 FT |g £ FT EFT 3> _. x Q. ~ FT F» o 2, CD P 8 g. o_ C o RT —• ES SR. © a 23 P

a |

o

c a m p a i 8 n

-

f , o w

p , o t s

voters to recall the i n c u m b e n t ' s n a m e . But, they m a i n t a i n , i n c u m b e n c y advantage d o e s not require recall ability; it requires only that the voter be able to reco g n i z e the i n c u m b e n t ' s n a m e w h e n e n c o u n t e r i n g it in the voting b o o t h . H e n c e the rise of the personal vote m a y rest on little m o r e than an increase in t h e ability of voters to r e c o g n i z e i n c u m b e n t s ' n a m e s . T h e n u b of this a r g u m e n t is that i n c u m b e n c y a d v a n t a g e , which is large at the level of the ballot b o x , d e p e n d s on cognitive u n d e r p i n n i n g s that are exceedingly slight. A n d this is exactly the pattern I have found: T h e effect of seniority w h i c h , I reiterate, is m e r e l y a proxy for t h e things that i n c u m b e n t s do to build a personal vote - is large at the level of vote defections in Figure 10.2 but small at the level of c a n d i d a t e c o n s i d e r a t i o n s in Figure 1 0 . 5 . 1 5

Political awareness , n f

245

8ive

estimates

for

mainly a m o n g o u t p a r t i s a n s scoring m i d d l e to low on political a w a r e n e s s , which is roughly the s a m e g r o u p that w a s most affected by seniority in t h e e a r l i e r analysis of v o t e defections. On t h e right of Figure 10.5, we see that an increase in c a m p a i g n intensity from low to high p r o d u c e s higher than b a s e l i n e s u p p o r t for the i n c u m b e n t a m o n g l e s ; a w a r e o u t p a r t i s a n s , and lower than baseline s u p p o r t a m o n g t h e highly a w a r e o u t p a r t i s a n s . T h e s e results also m a t c h the effects of c a m p a i g n intensity on o u t p a r t i s a n defections, as s h o w n in F i g u r e 1 0 . 2 . 1 4

G i v e n the differences in t h e m e t h o d s used to p r o d u c e Figures 10.2 a n d 10.5 t h e former involving the votes of 2 7 7 o u t p a r t i s a n s , the latter the net likes and dislikes of the t w o c a n d i d a t e s by s o m e 1,500 respondents - t h e s u b s t a n t i v e corr e s p o n d e n c e b e t w e e n t h e t w o sets of e s t i m a t e s s e e m s quite impressive. A n d yet, as can be seen, the m a g n i t u d e of the effect of seniority on c o n s i d e r a t i o n s s e e m s q u i t e small in relation to the m a g n i t u d e of t h e effect of defections, as s h o w n earlier. W h a t is g o i n g o n ? An explanation is s u g g e s t e d by an important analysis by M a n n and Wolfinger (1980). T h e y n o t e that t h e rise of a personal v o t e for i n c u m b e n t H o u s e m e m b e r s h a s not been a s s o c i a t e d with any d e t e c t a b l e increase over t i m e in the ability of 14 In these and subsequent simulations of candidate considerations, the effect of political awareness is simulated by simultaneously manipulating scores on the three awareness measures: knowledge, interest, and voting. Knowledge runs from - 1 . 8 SD to 1.5 SD, with the asymmetry reflecting the skew in the measure. In Figure 10.2, interest was included in the Awareness scale; in the present analysis, it is entered as a separate variable in the reception function and, for purposes of simulation, is manipulated over the range of 1.0 to 2.6 on the original three-point scale, with an adjustment for purging. Vote turnout is manipulated over a range of 0 to 1, with an adjustment for purging. The loquacity variable, the sum of party likes and dislikes, was set to 2.0 in all cases, thereby eliminating its correlation with political awareness.

My c o n c l u s i o n , therefore, is that the m o d e s t m a g n i t u d e of the seniority effect in Figure 10.5 is p r o b a b l y an accurate indication that the effect really is small, but not so small that it c a n n o t m a k e a big difference in the voting b o o t h w h e n the i n c u m b e n t is c o m p e t i n g against a c h a l l e n g e r w h o is all but invisible. Before leaving Figure 1 0 . 5 , I want to return to the q u e s t i o n of w h y exactly high c a m p a i g n intensity has the cross-cutting effects that it does. S i n c e Figure 10.5 has been p r o d u c e d by c o m p u t e r s i m u l a t i o n , it is possible to say exactly what h a s g e n e r a t e d t h e cross-cutting effects. T h e c a u s e of these cross-cutting effects, as I suggested earlier, is differential information flow. S c o r e s on net c a n d i d a t e c o n s i d e r a t i o n s in t h e baseline race have been produced by a d o m i n a n t but not very intense p r o i n c u m b e n t c o m m u nication flow, and a c o u n t e r v a l e n t p r o c h a l l e n g e r m e s s a g e that is extremely faint. As c a m p a i g n intensity i n c r e a s e s , the d o m i n a n t p r o i n c u m b e n t m e s s a g e gains sufficient p o w e r to reach low-awareness voters w h o could not pick it up in a b a s e line r a c e . H e n c e s u p p o r t for the i n c u m b e n t a m o n g t h e least a w a r e o u t p a r t i s a n s increases a s c a m p a i g n intensity increases. T h e c o u n t e r v a l e n t c h a l l e n g e r m e s s a g e also gains in intensity, but only e n o u g h to reach the most politically aware p e o p l e , w h o , if they are o u t p a r t i s a n s , then b e c o m e less favorable toward the incumbent. T h e s e points are m a d e c l e a r in Table 1 0 . 5 , w h i c h shows the simulated n u m b e r of d o m i n a n t c o n s i d e r a t i o n s ( i n c u m b e n t likes plus c h a l l e n g e r dislikes) and c o u n tervalent c o n s i d e r a t i o n s ( i n c u m b e n t dislikes plus c h a l l e n g e r likes) in both a lowintensity and a high-intensity r a c e . T h e effects are shown separately for a highawareness o u t p a r t i s a n voter and a l o w - a w a r e n e s s o n e . In the first row of the t a b l e , which depicts a low-intensity r a c e , neither highawareness nor l o w - a w a r e n e s s o u t p a r t i s a n s volunteer many c o n s i d e r a t i o n s , and most of w h a t they do express involves information from the d o m i n a n t c a m p a i g n . T h e c h a l l e n g e r ' s c a m p a i g n , t h u s , is almost invisible even to highly aware outpartisans. 15 The effects are small in the sense that the net of positive and negative considerations that seniority can explain is small; the coefficients responsible for the separate positive and negative effects, as shown in Table 10.4, are not small.

246 Table among

N A T U R E AND ORIGINS OF MASS O P I N I O N

10.5. Effect outpartisans

of campaign

intensity

High awareness outpartisan Proincumbent Prochallenger considerations considerations

on formation

Information flow and electoral choice

of considerations

Low awareness outpartisan Proincumbent considerations

Prochallenger considerations

Low-intensity campaign

0.24

0.07

0.16

0.01

High-intensity campaign

1.09

1.43

0.92

0.26

+1.36

+0.76

Gain from campaign

+0.85

Net gain/loss to incumbent

-0.41

+0.25 +0.51

Note: Cell entries are simulated mean number of considerations in each category for modal outpartisan voters, where simulations are based on coefficients in Table 10.4. Source: 1978 NES survey.

In a high-intensity r a c e , as s h o w n in the s e c o n d row of the t a b l e , highly a w a r e voters acquire additional c o n s i d e r a t i o n s , but m o r e c o u n t e r v a l e n t c o n s i d e r a t i o n s than d o m i n a n t ones. ( T h e s e respondents were heavily exposed to b o t h c a m p a i g n s , but, as o u t p a r t i s a n s , they accept m a i n l y countervalent c o n s i d e r a t i o n s . ) As a result, their net evaluation of the i n c u m b e n t falls below the b a s e l i n e . M e a n w h i l e , the effect of an intense c a m p a i g n on less a w a r e o u t p a r t i s a n s is quite different. T h e y a r e , as we have seen in Table 10.4, predisposed to accept c o u n t e r v a l e n t i n f o r m a t i o n , but most of w h a t reaches t h e m is i n f o r m a t i o n from t h e d o m i n a n t c a m p a i g n . H e n c e , in contrast to h i g h - a w a r e n e s s o u t p a r t i s a n s , they acquire m o r e d o m i n a n t c o n s i d e r a t i o n s than c o u n t e r v a l e n t o n e s , w h i c h leads t h e m to raise their net evaluations of the i n c u m b e n t above baseline levels. T h e s e d a t a on t h e effects of differential information flow offer a striking parallel to the n o t a b l e c a s e of c h a n g e s in public attitudes toward the V i e t n a m W a r b e t w e e n 1964 and 1966, w h e n less aware liberals b e c a m e m o r e s u p p o r t i v e of the w a r while m o r e a w a r e liberals b e c a m e less so. T h e only difference is t h a t , in the c a s e of H o u s e e l e c t i o n s , we have been able to capture the effects of a two-sided information flow b o t h at the level of s u m m a r y s t a t e m e n t s of preference (that is, vote c h o i c e s , as in Figure 10.2) and at the m o r e basic level of the c o n s i d e r a t i o n s u n d e r l y i n g s u m m a r y decisions.

Summary

on

House

elections

T h e analysis of H o u s e e l e c t i o n s h a s given us o u r best o p p o r t u n i t y to pick a p a r t the d y n a m i c s of resistance to persuasion by a d o m i n a n t political m e s s a g e , and especially the role of political a w a r e n e s s in such resistance. W h a t have we learned?

247

C o n v e r s e , in his study of information flow in p a r t i s a n e l e c t i o n s , proposed o n e m e c h a n i s m by w h i c h a w a r e n e s s induces resistance to p e r s u a s i o n . M o r e aware p e o p l e , he a r g u e d , d e v e l o p larger stores of l o n g - t e r m p a r t i s a n i n f o r m a t i o n , and this internalized ballast e n a b l e s t h e m to withstand the c a m p a i g n m e s s a g e s they encounter. T h e H o u s e d a t a produced s u p p o r t for this view. P e o p l e w h o are generally attentive to politics do have larger stores of preexisting p a r t i s a n conside r a t i o n s (Table 10.2), and preexisting c o n s i d e r a t i o n s are a s s o c i a t e d with resistance to i n c u m b e n t - d o m i n a t e d c a m p a i g n s (Tables 10.1 and 1 0 . 3 , and Figure 10.1). This type of resistance is w h a t I have t e r m e d inertial resistance. But political a w a r e n e s s r e m a i n s associated with resistance even after controlling for its intervening effect as a proxy for inertial c o n s i d e r a t i o n s . O n e reason for this is that a w a r e n e s s also e n h a n c e s partisan resistance to political c o m m u nications at the point of e n c o u n t e r i n g and d e c i d i n g w h e t h e r to accept t h e m , as s h o w n in Table 10.4. Yet t h e effects of a w a r e n e s s - i n d u c e d resistance to persuasion w e r e neither large nor consistent. T h e y played a role in resistance to i n c u m b e n t - d o m i n a t e d c a m p a i g n s , but not a large o n e . T h i s b r i n g s us to the third form of resistance to d o m i n a n t political c a m p a i g n s , countervalent resistance. C o u n t e r v a l e n t resistance involves sensitivity to sources of information o t h e r than the d o m i n a n t c a m p a i g n , w h i c h , in the present c a s e , m e a n s sensitivity to the prochallenger and a n t i i n c u m b e n t information from the c h a l l e n g e r c a m p a i g n . T h e major finding in this area h a s been t h a t , as s h o w n most clearly in Figure 10.4, highly a w a r e p e r s o n s , but not most o t h e r p e o p l e , receive significant a m o u n t s of information from the countervailing challenger c a m p a i g n . As a result, highly aware o u t p a r t i s a n s , but not m o s t o t h e r s , develop r e a s o n s for o p p o s ing the i n c u m b e n t and s u p p o r t i n g the challenger. T h i s , in t u r n , e n a b l e s highly aware o u t p a r t i s a n s to resist the d o m i n a n t i n c u m b e n t c a m p a i g n in the e l e m e n t a r y sense that, a l t h o u g h they m a y internalize s o m e p r o i n c u m b e n t m e s s a g e s , they d o not e n d up s u p p o r t i n g the i n c u m b e n t . It is w o r t h e m p h a s i z i n g that w h a t d r i v e s the c o u n t e r v a l e n t resistance of t h e politically aware is not resistance per s e , but t h e ability to pick up low-intensity c o m m u n i c a t i o n s from the political e n v i r o n m e n t . C o u n t e r v a l e n t resistance a p pears to be the m o s t i m p o r t a n t source of resistance to d o m i n a n t political campaigns. It might be objected that w h a t I am calling c o u n t e r v a l e n t resistance to a d o m inant m e s s a g e is not really resistance at all, but merely susceptibility to alternative sources of p e r s u a s i o n . This objection m a y perhaps be r e a s o n a b l e in cases in w h i c h the c o u n t e r v a l e n t m e s s a g e is nearly as intense as the d o m i n a n t o n e . But it should not be forgotten that c o u n t e r v a l e n t resistance involves, most fundamentally, being e x p o s e d to t w o streams of influence and picking out the o n e that is m o r e congenial w i t h o n e ' s values r a t h e r than simply the o n e that is louder. As long as this sort of a u t o n o m o u s choice is being m a d e , it m a k e s sense to say that o n e m e s s a g e has been c h o s e n and the o t h e r resisted.

248

Information

N A T U R E AND ORIGINS OF MASS O P I N I O N HOUSE E L E C T I O N S IN COMPARATIVE PERSPECTIVE

M o s t analysis of electoral c h o i c e , as I indicated in C h a p t e r 1, focuses on a single type of e l e c t i o n . B e c a u s e t h e present analysis of H o u s e elections has been cast in t e r m s of the general R A S m o d e l , however, there is every reason to try to extend it to m o r e c a s e s . T h i s is what I u n d e r t a k e to do in this section. O n e w o u l d not, of c o u r s e , expect identical patterns of electoral c h o i c e in w i d e l y dissimilar c a s e s . O n e w o u l d , however, expect the basic processes of inf o r m a t i o n diffusion and influence to be t h e s a m e . If s o , voting b e h a v i o r across types of e l e c t i o n s should a p p e a r fundamentally similar, o n c e s y s t e m a t i c differe n c e s in t h e flow of information are taken account of. Since H o u s e e l e c t i o n s provide less information to voters than any other type of national e l e c t i o n , and in this sense m a r k a sort of limiting c a s e , the o b v i o u s question b e c o m e s : W h a t h a p p e n s to electoral c h o i c e as c a m p a i g n intensity increases, thereby providing voters with larger a m o u n t s of c a n d i d a t e i n f o r m a t i o n ? To a n s w e r this q u e s t i o n , I brought together d a t a on voter defection to the inc u m b e n t from three other types of races: low- and high-intensity S e n a t e c a m p a i g n s , and the 1984 presidential r a c e . T h e s e d a t a are shown in Figure 10.6. Let me briefly d e s c r i b e the p r o c e d u r e s used to g e n e r a t e these d a t a . House elections. Defection patterns in low- and high-intensity H o u s e races are the s a m e as in t h e r i g h t - h a n d panel of Figure 10.2. T h e s e p a t t e r n s , as m a y be r e c a l l e d , depict the effects of c a m p a i g n intensity, net of the effect of seniority, which is held at its m i n i m u m value.

flow

and

electoral

Outpartisan defections to incumbent in contested 1990 Senate elections

Outpartisan defections to incumbent in contested 1978 House elections

Lowintensity

choice

249

Democratic defections to Reagan in 1984 presidential election

Low-intensity races

Political awareness Figure 10.6. Defections to incumbent party in U.S. national elections. House estimates are based on coefficients in Table 10.1; Senate estimates are based on coefficients in Table 10.6; presidential data are based on a polynomial regression. Source: National Election Study surveys. Presidential elections.

E s t i m a t i o n of defection rates to the i n c u m b e n t ' s p a r t y in

1984 w a s sufficiently straightforward that it could be a c c o m p l i s h e d by a simple p o l y n o m i a l regression.

Senate elections. Defection patterns in S e n a t e e l e c t i o n s have been e s t i m a t e d from t h e 1990 w a v e of the N E S Senate election study. T h e s e d a t a involve the twenty-five races in w h i c h an i n c u m b e n t sought reelection under c h a l l e n g e from the opposition party. T h e m o d e l and p r o c e d u r e s used in e s t i m a t i n g defection p a t t e r n s are exactly the s a m e as in the c a s e of H o u s e elections. T h e defection p a t t e r n s shown in the figure are e s t i m a t e s for m o d a l p a r t i s a n s , as d e s c r i b e d in A p p e n d i x A to this chapter. As in the c a s e of the H o u s e d a t a , low-intensity races have been defined by scores on the three c a m p a i g n - i n t e n s i t y v a r i a b l e s and corr e s p o n d to roughly the 10th percentile of overall c a m p a i g n intensity; highintensity races have been defined by scores at about the 90th p e r c e n t i l e . A l s o as in the case of H o u s e e l e c t i o n s , the i n c u m b e n t is a s s u m e d to have j u s t c o m pleted a first t e r m . Details of variable c o n s t r u c t i o n are found in A p p e n d i x A. 16

16 To control for differences in state population, I divided campaign spending by the number of congressional districts in the state. Low-intensity races were ones in which the total of the incumbent's spending, including party and other sources, was $1.5 million per district, the total of the challenger's spending was $500,000 per district, and the aggregate media intensity score was near the bottom of its range at 0.05. High-intensity races were ones in which incumbent and challenger spending were $10 million, and media intensity was .75.'

In e x a m i n i n g Figure 10.6, it is i m m e d i a t e l y a p p a r e n t that the n o n m o n o t o n i c i t y often associated with t h e r e c e p t i o n - a c c e p t a n c e model shows up strongly in H o u s e elections but is essentially absent from the other c a s e s . As explained in C h a p t e r 8, this is neither a p r o b l e m nor even a surprise for the R A S m o d e l , since n o n m o n o t o n i c i t y is only o n e of the possible o u t c o m e s of the r e c e p t i o n a c c e p t a n c e process. If t h e c h a n g e - i n d u c i n g m e s s a g e is sufficiently intense to reach the least politically aware stratum of the e l e c t o r a t e , politically u n a w a r e persons can be e x p e c t e d to be more susceptible to influence than a n y o n e else. T h u s , D r y e r (1971) found that the n o n m o n o t o n i c i t y that C o n v e r s e discovered in t h e 1952 presidential e l e c t i o n did not show up in any s u b s e q u e n t presidential contest. Instead, there has been a m o n o t o n i c a l l y n e g a t i v e r e l a t i o n s h i p b e t w e e n awareness and probability of vote defection. " T h e flow of s h o r t - t e r m s t i m u l i , " as D r y e r explained, " h a s effectively p e n e t r a t e d all s e g m e n t s of the e l e c t o r a t e , " thereby ironing out t h e n o n m o n o t o n i c i t y and p r o d u c i n g attitude c h a n g e even in the least politically active stratum of the e l e c t o r a t e (Dryer, 1971: p. 5 3 3 ) . W h a t Figure 10.6 adds to D r y e r ' s account is that even S e n a t e c a m p a i g n s now achieve sufficient intensity to penetrate to t h e least aware stratum of the e l e c t o r a t e , but that most H o u s e elections do not. T h e r e is, however, a p r o m i n e n t feature of Figure 10.6 that cannot be readily a c c o m m o d a t e d within a o n e - m e s s a g e f r a m e w o r k . It is the fact that peak levels

250

N A T U R E AND ORIGINS OF MASS O P I N I O N

Information

of defection o c c u r in H o u s e elections and in low-intensity S e n a t e e l e c t i o n s , which have relatively low levels of c a m p a i g n intensity, while the lowest defection rates occur in a presidential e l e c t i o n , which h a s the highest i n t e n s i t y . Nothing in my typology of attitude c h a n g e , as presented in Figure 8 . 2 , or any o t h e r m a n i p u l a t i o n of the logic of the o n e - m e s s a g e m o d e l , can explain this o c c u r r e n c e . T h e cross-cutting effects of c a m p a i g n intensity in the c a s e of H o u s e e l e c t i o n s likewise defy the logic of a o n e - m e s s a g e m o d e l . In order to explain the patterns of vote defection in Figure 10.6, w h i c h should be u n d e r s t o o d most generally as patterns of attitude c h a n g e , it is n e c e s s a r y to take account of both sides of the c o m m u n i c a t i o n flow to which citizens have been e x p o s e d . 17

W i t h respect to H o u s e e l e c t i o n s , this task has already been a c c o m p l i s h e d by my discussion of the effect of differential information flow on the f o r m a t i o n of c a n d i d a t e c o n s i d e r a t i o n s . Figure 10.7 gives us the e v i d e n c e n e c e s s a r y to extend this discussion to S e n a t e and presidential e l e c t i o n s . Figure 10.7 displays s u m m a r y m e a s u r e s of c a n d i d a t e evaluations for all three types of e l e c t i o n s , b a s e d on administration of the s a m e likes/dislikes q u e s t i o n s in each c a s e . T h e d a t a on H o u s e e l e c t i o n s in the figure are derived from the coefficients presented in Table 10.4; the d a t a for S e n a t e e l e c t i o n s have been simulated by the s a m e p r o c e d u r e s as used for H o u s e elections (see A p p e n d i x A to this c h a p t e r ) ; and the d a t a for the 1984 presidential election are simply plots of the raw m e a n s for E^emocrats and R e p u b l i c a n s by levels of political awareness. T h e i m m e d i a t e l y evident pattern in these d a t a is that m o r e intense c a m p a i g n s , both across types of e l e c t i o n s and within t h e m , lead inpartisans and o u t p a r t i s a n s to d e v e l o p overall c a n d i d a t e evaluations - m e a s u r e d as the net of the four cand i d a t e c o n s i d e r a t i o n m e a s u r e s — that are m o r e polarized along p a r t i s a n lines. G i v e n that, as we saw in Table 10.3, c a n d i d a t e c o n s i d e r a t i o n s powerfully affect vote d e c i s i o n s , the effect of this polarization of c o n s i d e r a t i o n s along p a r t y lines is to g e n e r a t e m o r e party-line voting in the more intense races and h e n c e lower defection rates. Since it has been p r o d u c e d by s i m u l a t i o n , the greater polarization of net cand i d a t e c o n s i d e r a t i o n s in high-intensity races is fully e x p l a i n a b l e . Two m a i n factors are at w o r k : M o r e intense races entail, in the construction of these figures as well as in actual p r a c t i c e , both more evenly balanced c o m m u n i c a t i o n flows and a larger overall v o l u m e of c o m m u n i c a t i o n . It is no m y s t e r y that a balanced c o m m u n i c a t i o n flow c o n t r i b u t e s to p a r t y p o l a r i z a t i o n , since it gives e a c h side the informational basis of a partisan evaluation of the c a n d i d a t e s , but sheer volu m e of c o m m u n i c a t i o n is also extremely i m p o r t a n t . Let us see why. We have already seen that D e m o c r a t s tend to accept m o r e of the proD e m o c r a t i c m e s s a g e s they e n c o u n t e r than they do of the p r o - R e p u b l i c a n o n e s , 17 The 1984 presidential election, of course, produced higher levels of defection to the inparty than in most presidential elections, but is otherwise typical of other cases. Even among House elections, higher-intensity races produce lower overall levels of defection than low-intensity races (mean 41 percent vs. 59 percent, t = 2.87, p < .01).

Net like/dislike scores in contested 1978 House elections

flow

and

electoral

Net like/dislike scores in contested 1970 Senate elections

251

choice

Net likeldislike scores in contested 1984 Presidential elections

Pro( incumbent

Republicans

ProChallenger Political awareness KEY 1. Members of inparty, high intensity race 2. Members of inparty, low intensity race 3. Members of outparty, low intensity race 4. Members of outparty, high intensity race Figure 10.7. Candidate evaluations in U.S. national elections. House estimates are based on Equation 10.1 and the coefficients in Table 10.4; Senate estimates are based on coefficients in Table 10.7; presidential data depict raw data. Sources: National Election Study surveys. and that R e p u b l i c a n s exhibit the reverse tendency. Let us a s s u m e , for the sake of a r g u m e n t , that t h e a c c e p t a n c e rate is 67 percent for congenial m e s s a g e s , but 33 percent for uncongenial ones. T h i s a m o u n t s to p a r t i s a n bias in information processing that is roughly in line with the rates implied for median respondents by the coefficients in Table 10.4. If, t h e n , a D e m o c r a t is e x p o s e d to three proD e m o c r a t i c m e s s a g e s and three pro-Republican o n e s , his expected net score will be +1 in the D e m o c r a t i c direction (he will accept t w o of the first type and one of the s e c o n d , for a net score of + 1 ) . But if he is exposed to six m e s s a g e s of each t y p e , his expected net score will increase to +2 (he will accept four congenial m e s s a g e s and two u n c o n g e n i a l o n e s ) . In g e n e r a l , the larger the n u m b e r of m e s sages on which this or any consistent D e m o c r a t i c bias o p e r a t e s , the larger the expected net score in the D e m o c r a t i c d i r e c t i o n . T h e same p r o c e s s , of c o u r s e , drives R e p u b l i c a n s to a higher net score in t h e Republican d i r e c t i o n .

252

Information flow and electoral choice

N A T U R E AND ORIGINS OF MASS O P I N I O N

T h i s a r g u m e n t explains why, all else e q u a l , there is more p a r t y p o l a r i z a t i o n in high-intensity (high-information) races. Since highly aware p a r t i s a n s ingest m o r e c a m p a i g n information than less aware p a r t i s a n s , it also explains why highly aware D e m o c r a t s and R e p u b l i c a n s are m o r e polarized in their net cons i d e r a t i o n s scores than are less a w a r e p a r t i s a n s of each stripe. It is w o r t h e m p h a s i z i n g , t h e n , that the m a i n reason for the greater p a r t i s a n p o l a r i z a t i o n of m o r e a w a r e voters - especially in presidential e l e c t i o n s , w h e r e the flow of c a n d i d a t e information is essentially evenly balanced - is not that the m o r e aware voters are m o r e selective in d e c i d i n g which c o m m u n i c a t i o n s to accept (though there m a y be s o m e t e n d e n c y for t h e m to be). It is, rather, that m o r e a w a r e p a r t i s a n s , o w i n g to the effect of a w a r e n e s s on r e c e p t i o n , ingest m u c h m o r e information than less aware p a r t i s a n s , and processing this greater volume of information with (almost) the same partisan bias as e v e r y o n e else leads t h e m to form net e v a l u a t i o n s that are m o r e highly p o l a r i z e d . T h i s a r g u m e n t explains why there is m o r e party-line voting in e l e c t i o n s in w h i c h information flow is m o r e intense, and why, within both presidential and S e n a t e e l e c t i o n s , there tends to be m o r e p a r t y loyalty a m o n g m o r e a w a r e voters. In H o u s e e l e c t i o n s there is a n o n m o n o t o n i c relationship between a w a r e n e s s and defection, but the reason for the exceptional p a t t e r n is clear. T h e a m o u n t of inf o r m a t i o n about the c a n d i d a t e s reaching less a w a r e voters in H o u s e e l e c t i o n s is lower than in any o t h e r t y p e of r a c e , an a m o u n t that is fairly close to n i l . Voters w h o are undisturbed by any new information do not defect to t h e o p p o s i t i o n . A modified form of C o n v e r s e ' s original information flow a r g u m e n t w o u l d , in light of all t h i s , explain t h e cross-election p a t t e r n of defections in Figure 10.6 as follows: Voters at the lowest awareness levels and in the lowest intensity e l e c tions r e m a i n loyal to their p a r t y b e c a u s e , as C o n v e r s e originally a r g u e d , they get almost no new i n f o r m a t i o n . As the information reaching voters rises from nearly nil to s o m e , p a r t i s a n instability shoots to its highest levels; these are the cases of m i d d l e - a w a r e n e s s voters in low-intensity H o u s e r a c e s , and l o w - a w a r e n e s s voters in low-intensity S e n a t e races. T h e reason for the m a r k e d instability of these voters is that most of the information they get is p r o i n c u m b e n t i n f o r m a t i o n , w h i c h they have little inertial capacity to resist. As c a m p a i g n information reaching outp a r t i s a n voters c o n t i n u e s to r i s e , instability falls from its peak levels, reaching a limit of almost no p a r t i s a n defection a m o n g highly aware voters in presidential e l e c t i o n s . Inertial resistance is only a small p a r t of the explanation for the d e cline in instability from peak levels. T h e m a i n reason is that more attentive voters are receiving higher volumes of more balanced c a n d i d a t e i n f o r m a t i o n , all of w h i c h c o n d u c e toward greater party p o l a r i z a t i o n of c a n d i d a t e e v a l u a t i o n s and h e n c e higher levels of p a r t i s a n loyalty. 1 8

T h i s analysis leads to an o b s e r v a t i o n of s o m e theoretical significance: Alt h o u g h resistance to d o m i n a n t political c a m p a i g n s d e p e n d s heavily on factors that are internal to individuals - notably stored c o n s i d e r a t i o n s , political values and a t t a c h m e n t s , and political awareness - intraindividual predispositions to18 The simulated mean number of remarks across both candidates for voters in the lowest awareness category in a low-intensity, low-seniority House race is 0.13.

253

w a r d resistance c a n n o t , by t h e m s e l v e s , explain very m u c h . In o r d e r to have real effect, they must be n o u r i s h e d by exposure to a countervalent information flow. W h e n people are e x p o s e d to two c o m p e t i n g sets of electoral i n f o r m a t i o n , they are generally able to c h o o s e a m o n g t h e m on the basis of their p a r t i s a n s h i p and values even w h e n they do not score especially well on tests of political awareness. But w h e n individuals are exposed to a o n e - s i d e d c o m m u n i c a t i o n flow, as in low-key H o u s e and S e n a t e e l e c t i o n s , their capacity for critical resistance appears quite limited. T h e conclusion I d r a w from this is that the most i m p o r t a n t source of resistance to d o m i n a n t c a m p a i g n s - certainly in e l e c t i o n s a n d , as e v i d e n c e from t h e mainstream model i n d i c a t e s , perhaps in other contexts as well - is c o u n t e r v a l e n t information c a r r i e d within the overall stream of political i n f o r m a t i o n . Before I c o n c l u d e this a r g u m e n t , let us e x a m i n e one final c a s e of attitude form a t i o n and c h a n g e .

T H E DYNAMICS OF P R E S I D E N T I A L PRIMARIES O n e difficulty in studying attitude c h a n g e in t h e context of p a r t i s a n e l e c t i o n s is that m o s t vote decisions are so strongly m o o r e d to stable p a r t y identifications that there is little o p p o r t u n i t y to o b s e r v e c h a n g e . In e x a m i n i n g rates of defection of voters in H o u s e e l e c t i o n s , for e x a m p l e , it w a s n e c e s s a r y to set aside most voters on the g r o u n d s that, as p a r t i s a n s of the inparty, they p r o d u c e d almost no cases of defection from p a r t y v o t i n g . T h i s difficulty obviously d o e s not arise in presidential p r i m a r y e l e c t i o n s . Indeed, the p r o b l e m is more nearly the opposite: Preferences s o m e t i m e s shift so rapidly that it is impossible to get a g o o d fix on t h e m . It is therefore interesting to see w h e t h e r the u n d e r s t a n d i n g of electoral d y n a m i c s that w o r k s so well in the relatively stable context of p a r t i s a n e l e c t i o n s applies to a m o r e free-wheeling e l e c t i o n as w e l l .

Background In t h e first w e e k s of the 1984 D e m o c r a t i c n o m i n a t i o n contest, it a p p e a r e d that a serious race m i g h t never d e v e l o p . J o h n G l e n n w a s m o u n t i n g an inept and faltering c a m p a i g n , and this left Walter M o n d a l e almost u n o p p o s e d in a D e m o cratic pack c o n s i s t i n g , except for h i m , of u n k n o w n s . Shortly before the N e w H a m p s h i r e p r i m a r y , t h e New York Times released a national poll s h o w i n g that M o n d a l e w a s further a h e a d than any c a n d i d a t e had ever been at that point in the n o m i n a t i o n contest. But big things w e r e stirring in Iowa. In t h e w e e k prior to that s t a t e ' s c a u c u s e s , G a r y H a r t w a s m o v i n g up fast, and this b e g a n to be reflected in national n e w s weeklies. If Time or Newsweek included a picture of any c a n d i d a t e other than M o n d a l e or G l e n n , it w a s likely to be H a r t . H a r t went on to place well in Iowa, finishing s e c o n d t o M o n d a l e . T h e n , m o r e unexpectedly, H a r t t r o u n c e d M o n d a l e in N e w H a m p s h i r e a n d , on the m a s s i v e t i d e of publicity that followed, s e e m e d on t h e verge of k n o c k i n g M o n d a l e out of the r a c e . H a r t ' s p i c t u r e , of c o u r s e ,

254

N A T U R E AND ORIGINS OF MASS O P I N I O N

graced the covers of the n e w s w e e k l i e s a n d , o v e r n i g h t , his n a m e b e c a m e a h o u s e hold w o r d . But a l t h o u g h performing very strongly in the S u p e r Tuesday p r i m a ries that followed t w o w e e k s after N e w H a m p s h i r e , H a r t c o u l d n ' t quite force M o n d a l e from the r a c e . Instead, he suddenly found himself the butt of j o k e s about his n a m e and his a g e , and trying to a n s w e r the question " W h e r e ' s the b e e f ? " T h r o u g h late M a r c h , April, and May, H a r t and M o n d a l e e n g a g e d in trench warfare, with M o n d a l e finally prevailing. T h e s e e v e n t s , as c o m p l e x a stimulus to m a s s opinion as any e n c o u n t e r e d in this b o o k , provide a final opportunity to o b s e r v e the effects of information flow on the evolution of the p u b l i c ' s political preferences.

Data

and

model

D u r i n g the p r i m a r y p e r i o d , the N E S c o n d u c t e d a c o n t i n u o u s survey of attitudes toward the c o n t e n d i n g c a n d i d a t e s . A l t h o u g h these surveys contacted only about forty-five D e m o c r a t s a w e e k , they provide an extremely v a l u a b l e , if slightly fuzzy, series of snapshots of the r a c e . In analyzing these d a t a , I have broken t h e p r i m a r y season into p e r i o d s that c o i n c i d e with the major s w i n g s in public attitudes. T h e first, c o v e r i n g J a n u a r y and early February, is the t ime w h e n M o n d a l e s e e m e d to have the race locked up and w h e n H a r t w a s a mefe blip in the national polls. T h e s e c o n d , covering, the last t w o w e e k s of F e b r u a r y w a s the period w h e n M o n d a l e w a s widely c o n s i d e r e d unstoppable but H a r t w a s beginning to build m o m e n t u m in the Iowa and N e w H a m p s h i r e c a m p a i g n s . T h e third is the t h r e e - w e e k period from i m m e d i a t e l y after N e w H a m p s h i r e to just before the Illinois p r i m a r y , the race in w h i c h H a r t suffered his first serious defeat; this w a s the period in which H a r t enjoyed his a m a z ing surge in the polls, but also the period in which he began to e n c o u n t e r the first questions about his c h a r a c t e r and his supposedly " n e w i s s u e s . " T h e fourth period r u n s from late M a r c h through the end of the p r i m a r y s e a s o n , a period of many m i n o r ups and d o w n s from which M o n d a l e eventually e m e r g e d the victor. Figure 10.8 provides a b r e a k d o w n of m e a n s u p p o r t for each c a n d i d a t e a m o n g D e m o c r a t s in each t i m e p e r i o d . Even from visual inspection, these d a t a tell an o b v i o u s story of differential information flow - that is, a story of the differential p e n e t r a t i o n of c a n d i d a t e m e s s a g e s of differing intensities. In the p e r i o d before the Iowa p r i m a r i e s , higher levels of political awareness are associated with steadily greater s u p p o r t for M o n d a l e , the c a n d i d a t e w h o , at that point, had no debilitating w a r t s and w a s , according to the m e d i a , the D e m o c r a t with the best c h a n c e of winning t h e n o m i n a t i o n (Brady and J o h n s t o n , 1987; Tables 7 and 8). G a r y H a r t had minimal s u p p o r t , but w h a t little he had w a s c o n c e n t r a t e d a m o n g the most politically aware m e m b e r s of the D e m o c r a t i c party, w h o were the only people sufficiently attentive to politics to have b e c o m e aware of such a " s e c o n d tier" candidate. 1 9

19 Thus, as Sam Popkin has noted, the politically aware are often "leading indicators" of future change (oral remarks. Southern California Running Dog Seminar,,San Diego, February 1988).

256

N A T U R E AND ORIGINS OF MASS O P I N I O N

T h e finding that the most aware D e m o c r a t s w e r e most susceptible to influe n c e by the d o m i n a n t M o n d a l e c a m p a i g n of this period m a y s e e m s u r p r i s i n g . But it s h o u l d n ' t . If, as has been a r g u e d , a w a r e n e s s is associated with resistance to persuasion b e c a u s e it proxies for r e c e p t i o n of a l t e r n a t i v e information sources - notably, the stored m a s s of previously formed c o n s i d e r a t i o n s , cueing m e s s a g e s , and c o u n t e r v a l e n t information flows - there is no reason to expect a w a r e n e s s - i n d u c e d resistance effects in this p e r i o d . D e m o c r a t s had no m e m o r y of bad e x p e r i e n c e s with M o n d a l e ; national elites largely s u p p o r t e d his candidacy, or at any rate p r o v i d e d no opinion leadership against h i m ; and the H a r t c a m p a i g n w a s t o o w e a k to be an effective source of c o u n t e r v a l e n t i n f o r m a t i o n . In such c i r c u m s t a n c e s , there w o u l d be no informational basis for resistance to persuasion and no c u e i n g m e s s a g e s that w o u l d i m p e d e the internalization of proM o n d a l e m e s s a g e s , so that influence should d e p e n d wholly on levels of r e c e p tion of the d o m i n a n t c a m p a i g n m e s s a g e . I n the t w o - w e e k period preceding the N e w H a m p s h i r e p r i m a r y , H a r t b e g a n his r i s e . But at this e a r l y point, the esoteric H a r t m e s s a g e w a s still able to reach only t h e most a w a r e q u a r t i l e of the D e m o c r a t i c rank-and-file. T h e r e w a s no resistance to this m e s s a g e , b e c a u s e H a r t , like M o n d a l e , w a s getting one-sidedly positive c o v e r a g e in the press and b e c a u s e D e m o c r a t s had no prior information about H a r t that w o u l d give t h e m a basis for resistance. M e a n w h i l e , the M o n d a l e m e s s a g e , which w a s still more intense than that of H a r t before the N e w H a m p shire p r i m a r y , w a s m a k i n g c o n v e r t s a m o n g t h e r a n k s of the m o d e r a t e l y a w a r e , w h o had not been paying sufficient attention to receive the M o n d a l e m e s s a g e in the p r e c a m p a i g n p e r i o d and were now j u s t tuning in. But the M o n d a l e m e s s a g e could m a k e n o m o r e c o n v e r t s a m o n g the m o s t a w a r e D e m o c r a t s b e c a u s e , within this s e g m e n t of t h e D e m o c r a t i c c o m m u n i t y , it w a s in direct c o m p e t i t i o n with the Hart campaign. So in this p e r i o d , there w a s still no e v i d e n c e of resistance to p e r s u a s i o n . T h e most a w a r e were clearly the most i n f l u e n c e a b l e , and by a very w i d e m a r g i n . ( T h e s h a r p n o n m o n o t o n i c i t y in net s u p p o r t for M o n d a l e in the s e c o n d period w a s not e v i d e n c e of resistance to c h a n g e ; it w a s e v i d e n c e that the m o s t a w a r e D e m o c r a t s were d e s e r t i n g M o n d a l e in o r d e r to follow H a r t , and that less a w a r e o n e s had not yet gotten the n e w s about H a r t ' s rising star.) Trends were m a r k e d l y different in the third t i m e p e r i o d . A l t h o u g h H a r t w a s enjoying his p o s t - N e w H a m p s h i r e publicity s u r g e , he was unable to gain any m o r e s u p p o r t a m o n g highly aware D e m o c r a t s . N o r did M o n d a l e lose any support in this g r o u p . T h u s , highly a w a r e persons had b e g u n to exhibit resistance to the d o m i n a n t c a m p a i g n m e s s a g e . T h e r e a s o n , presumably, w a s that they had now acquired e n o u g h information about the t w o c a n d i d a t e s that they could no longer be blown about by every n e w turn in the c a m p a i g n . M e a n w h i l e , H a r t ' s p o s t - N e w H a m p s h i r e publicity binge w a s strong e n o u g h finally t o reach m o d erately aware D e m o c r a t s , w h o rapidly d e s e r t e d M o n d a l e and s w u n g to H a r t in large n u m b e r s . T h e largest attitude s w i n g s in this period were clearly c o n c e n trated a m o n g p e r s o n s of m o d e r a t e political a w a r e n e s s . T h e m a g n i t u d e of this

257

Information flow and electoral choice Proportion switching to Hart after Iowa

Proportion switching to Hart after New Hampshire

Proportion switching to Mondale in late campaign period

Political awareness Figure 10.9. Estimated rates of preference change in Mondale vs. Hart contest. Note: Estimates are derived from coefficients in Table 10.8 and model in Appendix B to Chapter 10. Estimates involve Democrats only. p o p u l a r surge - to w h i c h t h e most aware and least aware D e m o c r a t s c o n t r i b u t e d little - w a s what n e a r l y d r o v e M o n d a l e from the r a c e . In the final p e r i o d , m o d e r a t e l y a w a r e D e m o c r a t s , following t h e n e w s of M o n d a y ' s increasing success i n t h e later p r i m a r i e s , s w u n g back t o M o n d a l e w h i l e the most politically a w a r e D e m o c r a t s c o n t i n u e d largely to hold their g r o u n d . A g a i n , t h e n , political a w a r e n e s s w a s associated with resistance to c h a n g e , pres u m a b l y b e c a u s e highly a w a r e persons h a d , by this t i m e , been able to firmly anchor their evaluations in an inertial m a s s of stored information about the t w o c a n d i d a t e s . T h e least a w a r e D e m o c r a t s were also fairly s t a b l e , with most of t h e m u n a b l e even to form a preference in the p r i m a r y contest. T h i s left m o d e r ately a w a r e persons m o s t susceptible to t h e c u r r e n t s of the t i m e . 2 0

Unfortunately, there are no data at the level of c a n d i d a t e likes and dislikes with which to c h e c k this account of the m i c r o d y n a m i c s of attitude c h a n g e . It is, however, possible to fit the r e c e p t i o n - a c c e p t a n c e m o d e l to the d a t a s h o w n in Figure 10.8 in order to see w h e t h e r the story I h a v e j u s t told c a n , in a statistical s e n s e , hold together. T h i s m o d e l i n g exercise is relegated to A p p e n d i x B of this chapter. T h e only point that needs to be m a d e h e r e is that the m o d e l fits the d a t a extremely well and c o n f i r m s my basic a c c o u n t in every i m p o r t a n t respect. For e x a m p l e , Figure 10.9 gives the r a t e s of attitude c h a n g e toward H a r t and then toward M o n d a l e in the final p h a s e s of t h e c a m p a i g n , as e s t i m a t e d from coefficients in the m o d e l . Summary

on

primary

elections

T h e s e findings, especially those from the e a r l y p h a s e s of the c a m p a i g n , underscore a critical theoretical point: A w a r e n e s s - i n d u c e d resistance to d o m i n a n t political c a m p a i g n s is not a u t o m a t i c . It d e p e n d s on access to a l t e r n a t i v e c o m 20 From different perspectives, Bartels (1988) and Brady and Johnston (1987) also stress that voter learning takes place over the course of the primary campaign.

256

T h e finding that the m o s t a w a r e D e m o c r a t s were most s u s c e p t i b l e to influe n c e by the d o m i n a n t M o n d a l e c a m p a i g n of this period may s e e m s u r p r i s i n g . But it s h o u l d n ' t . If, as has been a r g u e d , a w a r e n e s s is associated with resistance to persuasion b e c a u s e it proxies for r e c e p t i o n of a l t e r n a t i v e information sources - notably, the stored m a s s of previously formed c o n s i d e r a t i o n s , c u e i n g m e s s a g e s , and c o u n t e r v a l e n t information flows - there is no reason to expect a w a r e n e s s - i n d u c e d resistance effects in this p e r i o d . D e m o c r a t s had no m e m o r y of bad e x p e r i e n c e s with M o n d a l e ; national elites largely s u p p o r t e d his candidacy, or at any rate p r o v i d e d no opinion l e a d e r s h i p against h i m ; and the H a r t c a m p a i g n w a s too w e a k to be an effective source of c o u n t e r v a l e n t i n f o r m a t i o n . In such c i r c u m s t a n c e s , there w o u l d be no informational basis for resistance to persuasion and no c u e i n g m e s s a g e s that w o u l d i m p e d e the internalization of proM o n d a l e m e s s a g e s , so that influence should d e p e n d wholly on levels of r e c e p tion of the d o m i n a n t c a m p a i g n m e s s a g e . I n the t w o - w e e k period preceding the N e w H a m p s h i r e p r i m a r y , H a r t began his r i s e . But at this e a r l y point, the esoteric H a r t m e s s a g e was still able to reach only t h e most a w a r e q u a r t i l e of the D e m o c r a t i c rank-and-file. T h e r e w a s no resistance to this m e s s a g e , b e c a u s e H a r t , like M o n d a l e , w a s getting one-sidedly positive c o v e r a g e in the press and b e c a u s e D e m o c r a t s had no prior information about H a r t that w o u l d give t h e m a basis for resistance. M e a n w h i l e , the M o n d a l e m e s s a g e , which w a s still r i o r e intense than that of H a r t before the N e w H a m p shire p r i m a r y , w a s m a k i n g c o n v e r t s a m o n g the r a n k s o f the m o d e r a t e l y a w a r e , w h o had not been p a y i n g sufficient attention to receive the M o n d a l e m e s s a g e in the p r e c a m p a i g n p e r i o d and were now j u s t tuning in. But the M o n d a l e m e s s a g e could m a k e n o m o r e c o n v e r t s a m o n g the m o s t a w a r e D e m o c r a t s b e c a u s e , within this s e g m e n t of the D e m o c r a t i c c o m m u n i t y , it w a s in direct c o m p e t i t i o n with the Hart campaign. So in this p e r i o d , there w a s still no e v i d e n c e of resistance to p e r s u a s i o n . T h e most a w a r e were clearly the most i n f l u e n c e a b l e , and by a very w i d e m a r g i n . ( T h e s h a r p n o n m o n o t o n i c i t y in net s u p p o r t for M o n d a l e in the s e c o n d p e r i o d w a s not e v i d e n c e of resistance to c h a n g e ; it w a s e v i d e n c e that the m o s t a w a r e D e m o c r a t s were d e s e r t i n g M o n d a l e in o r d e r to follow H a r t , and that less a w a r e o n e s had not yet gotten the n e w s about H a r t ' s rising star.) Trends were m a r k e d l y different in the third t i m e p e r i o d . A l t h o u g h H a r t w a s enjoying his p o s t - N e w H a m p s h i r e publicity s u r g e , he w a s u n a b l e to gain any m o r e support a m o n g highly aware D e m o c r a t s . Nor did M o n d a l e lose any s u p p o r t in this g r o u p . T h u s , highly a w a r e p e r s o n s had b e g u n to exhibit resistance to the d o m i n a n t c a m p a i g n m e s s a g e . T h e r e a s o n , presumably, w a s that they had now acquired e n o u g h information about the t w o c a n d i d a t e s that they could no longer be blown about by every new turn in the c a m p a i g n . M e a n w h i l e , H a r t ' s p o s t - N e w H a m p s h i r e publicity b i n g e w a s strong e n o u g h finally t o reach m o d erately a w a r e D e m o c r a t s , w h o rapidly d e s e r t e d M o n d a l e and s w u n g t o H a r t i n large n u m b e r s . T h e largest attitude s w i n g s in this period were c l e a r l y c o n c e n trated a m o n g p e r s o n s of m o d e r a t e political a w a r e n e s s . T h e m a g n i t u d e of this

257

Information flow and electoral choice

N A T U R E AND ORIGINS OF MASS O P I N I O N

Proportion switching to Hart after Iowa

Proportion switching to Hart after New Hampshire

Proportion switching to Mondale in late campaign period

1.0 .8 .6 .4 .2 0

]

/

' Political awareness

Figure 10.9. Estimated rates of preference change in Mondale vs. Hart contest. Note: Estimates are derived from coefficients in Table 10.8 and model in Appendix B to Chapter 10. Estimates involve Democrats only. p o p u l a r surge - to w h i c h t h e most aware and least aware D e m o c r a t s c o n t r i b u t e d little - w a s what n e a r l y d r o v e M o n d a l e from the r a c e . In the final p e r i o d , m o d e r a t e l y a w a r e D e m o c r a t s , following t h e n e w s of M o n d a y ' s increasing success i n t h e later p r i m a r i e s , s w u n g back t o M o n d a l e w h i l e t h e m o s t politically a w a r e D e m o c r a t s c o n t i n u e d largely to h o l d their g r o u n d . A g a i n , t h e n , political a w a r e n e s s w a s associated with resistance t o c h a n g e , pres u m a b l y b e c a u s e highly a w a r e persons h a d , by this t i m e , been able to firmly a n c h o r their evaluations in an inertial m a s s of stored information about the t w o c a n d i d a t e s . T h e least a w a r e D e m o c r a t s w e r e also fairly s t a b l e , with most of t h e m u n a b l e even to form a preference in the p r i m a r y contest. T h i s left m o d e r ately aware p e r s o n s m o s t susceptible t o t h e c u r r e n t s o f the t i m e . Unfortunately, there are no d a t a at t h e level of c a n d i d a t e likes and dislikes with which to c h e c k this account of t h e m i c r o d y n a m i c s of attitude c h a n g e . It is, however, possible to fit t h e r e c e p t i o n - a c c e p t a n c e m o d e l to the d a t a s h o w n in Figure 10.8 in order to see w h e t h e r the story I have j u s t told c a n , in a statistical s e n s e , h o l d together. T h i s m o d e l i n g exercise is relegated to A p p e n d i x B of this chapter. T h e only point that needs to be m a d e here is that t h e m o d e l fits t h e d a t a extremely well and c o n f i r m s my basic account in every i m p o r t a n t respect. For e x a m p l e , Figure 10.9 gives the rates of attitude c h a n g e t o w a r d H a r t and then toward M o n d a l e in the final p h a s e s of the c a m p a i g n , as e s t i m a t e d from coefficients in t h e m o d e l . 2 0

Summary

on

primary

elections

T h e s e findings, especially those from the e a r l y p h a s e s of the c a m p a i g n , underscore a critical theoretical point: A w a r e n e s s - i n d u c e d resistance to d o m i n a n t political c a m p a i g n s is not a u t o m a t i c . It d e p e n d s on access to a l t e r n a t i v e c o m 20 From different perspectives, Bartels (1988) and Brady and Johnston (1987) also stress that voter learning takes place over the course of the primary campaign.

258

N A T U R E AND O R I G I N S OF MASS O P I N I O N

Information

m u n i c a t i o n s - either in the form of stored c o n s i d e r a t i o n s and information from past c a m p a i g n s or in t h e form of c u r r e n t reception of c o u n t e r v a l e n t c o m m u n i c a t i o n s or cues. H e n c e w h e n , as in the early stages of the 1984 p r i m a r i e s , the m o s t a w a r e persons have no informational basis for r e s i s t a n c e , they exhibit no hint of resistance. To the c o n t r a r y , they are m o r e reactive to c a m p a i g n s of the m o m e n t than any o t h e r s e g m e n t of the p u b l i c . O n l y in the later stages of the c a m p a i g n , by w h i c h t i m e there w a s a two-sided information flow and s o m e d e v e l o p m e n t of inertial c o n s i d e r a t i o n s , do we find that the most a w a r e D e m o c r a t s are most resistant to c h a n g e .

CONCLUDING

REMARKS

T h i s c h a p t e r has been only incidentally c o n c e r n e d with e l e c t i o n s . Its real aim has been t o s h o w h o w p e o p l e ' s s u m m a r y political d e c i s i o n s , w h e t h e r vote choices or r e s p o n s e s to c l o s e d - e n d e d survey i t e m s , are b a s e d on c o n s i d e r a t i o n s that form from the c o m p e t i n g influences to which they have been e x p o s e d . T h e a d v a n t a g e of this c h a p t e r over p r e c e d i n g ones is that it has h a d , for the first t i m e , a b u n d a n t , m e a s u r a b l e variability in t h e intensities of t h e o p p o s i n g m e s s a g e s to w h i c h individuals have been e x p o s e d , along with m e a s u r e s of the c o n s i d e r a t i o n s that u n d e r l i e s u m m a r y e x p r e s s i o n s of political preferences. This has m a d e it possible to directly o b s e r v e the effects of inertial resistance (in the form of party likes and dislikes) a n d , m o r e i m p o r t a n t in the context of e l e c t i o n s , the effects of c o u n t e r v a l e n t resistance (in the form of prochallenger and antiinc u m b e n t c o m m u n i c a t i o n f l o w s ) . But the theoretical m a c h i n e r y u n d e r l y i n g these o b s e r v a t i o n s has b e e n exactly the s a m e as in e a r l i e r analyses of attitude formation and c h a n g e .

APPENDIX A: D E F E C T I O N S IN S E N A T E E L E C T I O N S T h e m o d e l used in e s t i m a t i n g defection to the i n c u m b e n t in S e n a t e e l e c t i o n s w a s the s a m e as for H o u s e defections. T h e variables used in the m o d e l are as follows: Political awareness. A l t h o u g h the S e n a t e election study contained an unusually weak selection of a w a r e n e s s i t e m s , I built the best m e a s u r e I c o u l d , as d e s c r i b e d in the M e a s u r e s A p p e n d i x . Campaign intensity.

T h e 1990 S e n a t e study a s k e d all respondents:

How many stories did you read, see or hear regarding the campaign in this state for the U.S. Senate? Would you say that you read, saw, or heard a good many, just one or two, or none? As described in the c a s e of the H o u s e d a t a , I c o n v e r t e d responses to this question to a state-level m e a s u r e of the intensity of m e d i a c o v e r a g e . T h e variable r u n s from 0 to 0 . 7 4 scale units, as d e n o m i n a t e d by the original two-unit scale. 21

21 There may be some heteroskedasticity in measurement of the aggregate media variable, but it is likely to be small since N's of the state-level means are approximately the same.

flow

and

electoral

choice

259

Table 10.6. Coefficients predicting vote for Senate incumbent among outpartisans Reception function Intercept Awareness (standardized) Incumbent spending (in $10 thousands per cong. district) Challenger spending (in $10 thousands per cong. district) Media coverage (range 0 - .74) Seniority (logged) Seniority X Challenger spending Acceptance function Intercept Awareness (standardized) Partisan strength (range 1-2) Policy distance (see text) Challenger spending N

0.19 0.40 (.39) 0.015 (.008) -.012 (.008) -2.65 (1.42) 1.14 (1.41) -0.0035 (.0031) 4.05 -1.31 (.50) -1.18 (.57) -0.70 (.31) 0.001 (.001) 450

Note: Model is Equation 10.1, estimated by maximum likelihood. Standard errors are in parentheses. Dependent variable is whether outpartisan voted for the incumbent. Source: 1990 NES Senate survey.

Political values. T h e S e n a t e study c a r r i e d the traditional party identification m e a s u r e , scored in t h e usual way. It w a s also possible to m e a s u r e ideological or policy distance from the i n c u m b e n t . T h e d i s t a n c e m e a s u r e w a s c o n s t r u c t e d by s u m m i n g i n c u m b e n t s ' scores on the A m e r i c a n s for D e m o c r a t i c A c t i o n and A m e r i c a n C o n s e r v a t i v e U n i o n r a t i n g scales from the year prior to the e l e c t i o n , standardizing the s c o r e s , and subtracting from a standardized m e a s u r e of res p o n d e n t s ' ideological self-description. T h e v a r i a b l e has been c o d e d so that, if the D e m o c r a t i c c a n d i d a t e is to the left of the Republican c a n d i d a t e , higher scores indicate greater policy affinity to the challenger. T h e coefficients for the S e n a t e defection m o d e l are s h o w n in Table 10.6. O n e point of interest in t h e s e coefficients is t h a t , a l t h o u g h the seniority effects in S e n a t e e l e c t i o n s a r e s m a l l e r than in H o u s e e l e c t i o n s (as s h o w n in Table 10.1)

260

Information flow and electoral choice

N A T U R E AND ORIGINS OF MASS O P I N I O N

T a b l e 1 0 . 7 . Coefficients for diffusion of likes and dislikes in contested 1990 Senate races Incumbent likes Reception function Intercept -1.60 Awareness 1.00 (standardized) (.18) Voted in 1978 1.20 (range 0-1) (.24) Incumbent spending 5.11 (log $10,000s per cong. dist.) (2.23) Challenger spending -3.80 (2.02) News coverage -0.89 (range 0 - .75) (.53) Seniority 1.02 (log of years in office) (.36) Prior elective office (0 or 1) Acceptance function Intercept Party attachment (range -2 to +2) Issue distance from Incumbent Ideological identification (standardized) Domestic spending attitudes (standardized)

Incumbent dislikes

Challenger likes

Challenge dislikes

-2.96 0.82 (.14) 0.35 (.15) 5.37 (1.57)

-3.10 0.62 (.12) 0.71 (.18) 2.91 (1.51) -2.73 (1.37) 2.51 (.52)

-2.20 0.94 (.16) 0.94 (.23) -1.77 (2.20) 1.18 (1.99) 5.45 (.89)

-4.99 (1-45) 1.70 (.42) 0.03 (.20)

—

—

-0.50 -0.21 (.03) -0.12 (.05)

-1.32 0.30 (.06) 0.38 (.09)

-

-

-0.06 (.04)

0.06 (.06)

—

—

0.12 (.07)

0.27 (.09)

-0.79 0.43 (.09)

-1.32 -0.12 (.04)

_ 0.03 (.07) 0.21 (.08)

_ -0.29 (.05) -0.13 (.05)

Note: Dependent variables are sums of all remarks of given type, recoded to 0 - 1 range. Up to five remarks were counted in conection with each probe. The model is Equation 10.1. Approximate standard errors are in parentheses. A' of cases, which involve all states in which an incumbent sought reelection in a contested race in 1990, is 1866. Source: 1990 NES Senate survey.

261

I might add that the effects of the predispositions v a r i a b l e s , w h i c h are not displayed in Figure 10.6 b e c a u s e it shows trends only for modal o u t p a r t i s a n s , are substantial. A m o n g voters ideologically closest to the i n c u m b e n t , defection patterns actually increase with political a w a r e n e s s ; a m o n g those most ideologically distant, the decline in defection associated with a w a r e n e s s is steeper than shown in Figure 10.6. T h u s , t h e pattern of defection for different types of p a r t i s a n s s o m e w h a t r e s e m b l e s the voter defection c u r v e s in Figure 1 0 . 1 , except for a m o r e intense m e s s a g e . My e s t i m a t e s of net c a n d i d a t e c o n s i d e r a t i o n scores for S e n a t e e l e c t i o n s in Figure 10.7 were p r o d u c e d by the same general m e t h o d used in c r e a t i n g the H o u s e e s t i m a t e s , as r e p o r t e d in Table 10.4. In the reception function, I used political a w a r e n e s s , voter t u r n o u t (purged of the effects of c a m p a i g n intensity), logged s p e n d i n g by the i n c u m b e n t , logged s p e n d i n g by the challenger, logged seniority, and m e d i a c o v e r a g e ; in t h e m o d e l s for c h a l l e n g e r likes and dislikes, I also used a v a r i a b l e m e a s u r i n g the previous electoral e x p e r i e n c e of the challenger. In the a c c e p t a n c e function, I used party a t t a c h m e n t and policy d i s t a n c e . I specified the distance v a r i a b l e in t h e a c c e p t a n c e e q u a t i o n as the absolute value of policy distance (note, I a s s u m e that l e f t - r i g h t policy distance affects vote c h o i c e , as d e scribed a m o m e n t a g o , but that absolute policy difference affects liking and disliking, so that, for e x a m p l e , a c o n s e r v a t i v e voter to the right of a c o n s e r v a t i v e S e n a t o r will not be led to vote against the i n c u m b e n t but may be led to dislike h i m or h e r ) . T h e survey also c a r r i e d a series of q u e s t i o n s on budget priorities. I c o m b i n e d these items into a scale, s t a n d a r d i z e d the scale, and adjusted the scoring so that low scores indicate affinity with t h e i n c u m b e n t . To m e a s u r e issue distance from the challenger, I used r e s p o n d e n t s ' ideological self-identification, s t a n d a r d i z i n g and c o d i n g it to fit the p a r t y affiliation of the challenger. N o t e that I did not use political a w a r e n e s s in the a c c e p t a n c e function of the e s t i m a t e s of likes and dislikes for the S e n a t e c a n d i d a t e s . This is b e c a u s e , w h e n it w a s included, it b e h a v e d oddly, picking up large positive coefficients (indic a t i n g less resistance a m o n g the highly a w a r e ) , while causing the a w a r e n e s s variable in the r e c e p t i o n function also to pick up n e g a t i v e coefficients (indicating less reception a m o n g the m o r e a w a r e ) . I do not understand why E q u a t i o n 10.1 b e h a v e d nonsensically in this c a s e after it had w o r k e d well in so m a n y othe r s , but elimination of a w a r e n e s s from t h e a c c e p t a n c e function restored sense to the e s t i m a t e s , as s h o w n in Table 1 0 . 7 . 2 2

and do not a p p r o a c h statistical significance, they follow the s a m e p a t t e r n as in the H o u s e case: a positive m a i n effect for seniority and a n e g a t i v e interaction with c h a l l e n g e r s p e n d i n g .

T h e values used to c r e a t e the idealized o u t p a r t i s a n s in Figure 10.7 w e r e the s a m e as those used in the H o u s e analysis of likes/dislikes: p a r t y d i s t a n c e scores of 1.3 and — 1 . 3 , absolute policy distance scores of 2 and 0, budget attitude

T h e x-axis for S e n a t e e l e c t i o n s in Figure 10.6 m a n i p u l a t e s political awareness over a r a n g e of — 1.9 SD to + 1 . 8 6 , w h i c h , as u s u a l , represents 98 percent of the total r a n g e of t h e s c a l e . T h e idealized o u t p a r t i s a n s used in the analysis in Figure 10.6 are p e o p l e having a p a r t y identification score of 1.3 and a policy distance score o f + 2 .

22 There was a similar tendency in estimates of the House likes/dislikes data when the loquacity measure (a count of all party likes and dislikes, regardless of direction) was left out of the House model. The loquacity measure, which is, of course, positively correlated with awareness, was not available in the Senate data. A possible explanation for this difficulty is that the model, which contains some of the same variables in both a reception and an acceptance function, is not identified if the true value of the awareness coefficient in the acceptance function is zero.

262

N A T U R E AND ORIGINS OF MASS O P I N I O N

Information

scores of 1 and — 1, and ideological distances from the challenger of 1 and — 1 for o u t p a r t i s a n s and i n p a r t i s a n s , respectively.

Table 10.8. model

To m o d e l trends in c a n d i d a t e support in the 1984 D e m o c r a t i c p r i m a r i e s , we specify an initial s u p p o r t e q u a t i o n for H a r t and M o n d a l e , as follows: //, = P ,

(10.4o)

Mi = P

(10.4ft)

H e r e H and M, are levels of support for H a r t and M o n d a l e at t i m e 1, and P , and P are functions that specify this s u p p o r t , " H a r t c h a n g e " and " M o n d a l e c h a n g e , " in relation to levels of political a w a r e n e s s . Each s u b s e q u e n t t i m e period is then conceived as a " c a m p a i g n " that either adds to M o n d a l e ' s s u p p o r t by subtracting from H a r t ' s or vice versa. x

electoral

Reception function

PRIMARY C O N T E S T

MCl

and

263

choice

Coefficients for presidential primaries

A P P E N D I X B : MODELING T H E M O N D A L E - H A R T

HC

flow

Acceptance function

Haiti intercept

-4.92 (2.12)

a

Mondale i intercepts

1.17 (.64)

-0.02 (.27)

Hart2 intercept

-2.59 (.50)

a

Hart3 intercepts

0.32 (.40)

2.44 (1.46)

Mondale4 intercepts

-0.44 (.41)

2.41 (1.62)

Awareness (standardized)

1.17 (0.80)

-2.87* (1.43)

HC

MC]

D u r i n g the s e c o n d t i m e p e r i o d , a certain p r o p o r t i o n of D e m o c r a t s c o n v e r t to H a r t , as given by P , which is also a function of political a w a r e n e s s . T h e n , following E q u a t i o n 7 . 7 , w e may w r i t e HC

H = Hi + P c * (1 - Hi) 2

H

(10.4c)

2

Mi = Mi - P

* Mi

HCi

(10Ad)

T h a t is, H a r t s u p p o r t at t i m e 2 is equal to initial H a r t s u p p o r t plus the proportion of initial n o n s u p p o r t e r s w h o c o n v e r t to H a r t . M o n d a l e s u p p o r t at t i m e 2 is equal to initial M o n d a l e s u p p o r t m i n u s the p r o p o r t i o n of initial s u p p o r t lost to H a r t . T h e r e m a i n i n g e q u a t i o n s are then Hy = H

+ P

2

M H M

3

= M

4

= /Y

4

= M

* (1 - H )

HCy

(\0Ae)

2

a

H

=

4

[[P

HCL

+

2

- P *M

(10.4/)

+

3

- P *H

(10.4g)

~

PMC,

(10Ah)

+

PHC,

HCi

2

MCi

3

+ P

3

* (1

MC)

- M) 3

w h e r e P and P are the p r o p o r t i o n s c o n v e r t i n g to H a r t in period 3 and to M o n d a l e in period 4. T h e s e P t e r m s m o d e l the effect of the d o m i n a n t m e s s a g e of each p e r i o d , and each is a form of the usual r e c e p t i o n - a c c e p t a n c e p r o c e s s , as in E q u a t i o n 7.5 in the text. E q u a t i o n s \0Aa-h represent a series of nested e q u a t i o n s , such that t e r m s from each earlier e q u a t i o n appear in later ones as well. After a p p r o p r i a t e substitutions, for e x a m p l e , the four H a r t e q u a t i o n s b e c o m e HC}

Note: The data for this model are the 40 aggregate means in Figure 10.8. Standard errors appear in parentheses. Acceptance function eliminated to conserve degrees of freedom; see text. * Coefficient applies to acceptance function only in third and fourth time periods. Source: 1984 NES primary election survey.

MC

PHC,

PHC,

*

*

(1

*

[[PHC,

* (1

"

—

(1

-

[PHC, + [PHC,

P ,)] HC

+

PHC,

PHC, *

(1

*

(1

-

-

PHC,)])]

PHC,)]

+ PHC, * (1

~~

PHC, I ) ] ) ]

w h e r e each of the P t e r m s represents a r e c e p t i o n - a c c e p t a n c e process having the form of E q u a t i o n 7 . 5 . T h e four M o n d a l e e q u a t i o n s have a c o m p a r a b l e form. HC

c

Hi H

2

H3

= = =

PHC, [P

+

HCL

[[P , HC

+

PHC, +

PHC,

*

*

(1

-

PHC,)]

PHC,

*

(1

-

(1

"

[PHC,

PHC,)] +

PHC

E a c h of t h e five P t e r m s in these e q u a t i o n s has four p a r a m e t e r s - an intercept for the r e c e p t i o n and a c c e p t a n c e functions, and a coefficient on attentiveness in both functions - w h i c h add up to a total of twenty p a r a m e t e r s to be simultaneously estimated. C

In e s t i m a t i n g this c u m b e r s o m e m o d e l , I used the a g g r e g a t e d a t a s h o w n in Figure 10.8 - s o m e 40 d a t a points (4 p e r i o d s x 2 c a n d i d a t e s x 5 awareness g r o u p s ) . To p r e s e r v e d e g r e e s of freedom, I simplified the m o d e l by constraining or e l i m i n a t i n g as m a n y p a r a m e t e r s as p o s s i b l e . First, I constrained a w a r e n e s s to have the s a m e effect in t h e reception functions of all P t e r m s , leaving the intercept t e r m s , a , to pick up differences in t h e intensities of t h e various c

*

(1

-

P ,)])] HC

0

264

N A T U R E AND ORIGINS OF MASS O P I N I O N

m e s s a g e s . This is a strong but not u n r e a s o n a b l e constraint: After controlling for m e s s a g e intensity via intercept shifts, a w a r e n e s s ought to have the s a m e strength of relationship with r e c e p t i o n in each c a m p a i g n in each t i m e p e r i o d . S e c o n d , since visual e x a m i n a t i o n of the d a t a and initial e s t i m a t e s indicated that the acc e p t a n c e function m a d e no contribution to the e s t i m a t i o n of E q u a t i o n s 1 0 . 4 a - d , it w a s e l i m i n a t e d , w h i c h saved several p a r a m e t e r s . T h e a w a r e n e s s variable w a s eliminated as an a c c e p t a n c e factor in E q u a t i o n 10.46 on the s a m e g r o u n d s . Finally, I constrained a w a r e n e s s to have the s a m e effect in the a c c e p t a n c e functions of E q u a t i o n s \0.4e-h; separate a w a r e n e s s coefficients w o u l d have been preferable h e r e , but n o n e of the coefficients could have a p p r o a c h e d statistical significance if e s t i m a t e d separately. T h e imposition of these constraints left ten coefficients for t h e forty d a t a points. E q u a t i o n s 10.4a-/* were s i m u l t a n e o u s l y e s t i m a t e d with a n o n l i n e a r regression p r o g r a m .

11

Evaluating the model and looking toward future research

23

T h e coefficient e s t i m a t e s are s h o w n in Table 10.8. B e c a u s e most of the t e r m s represent intercept shifts in the reception or a c c e p t a n c e functions and h e n c e involve only five or ten d a t a points, it is not w o r r i s o m e that s o m e fall short of statistical significance in this highly interactive m o d e l . W h a t is i m p o r t a n t is that coefficients d e n o t i n g t h e effects of a w a r e n e s s on reception and on a c c e p t a n c e both achieve s o m e d e g r e e of statistical s i g n i f i c a n c e . As a further indication of g o o d n e s s of fit, it m a y be noted that the c o r r e l a t i o n b e t w e e n the y - e s t i m a t e s from t h e model and the r a w d a t a is .92 ( r = . 8 4 ) . Given the s a m p l i n g e r r o r in the r a w d a t a and t h e constraints on the m o d e l , this would s e e m to be as m u c h a c c u r a c y as could be h o p e d for. 24

2

23 Coefficient estimates for b tended toward indicating no resistance. 24 Because the model has been estimated from data aggregates derived from different cell sizes, significance tests should be regarded as being only roughly suggestive. 0

" I f t h e public had an opinion and there w a s no pollster a r o u n d to m e a s u r e it, w o u l d public opinion e x i s t ? " Like the old c o n u n d r u m about the tree falling in the forest with no o n e to h e a r it, this q u e s t i o n is not c o m p l e t e l y v a c u o u s . T h e a n s w e r d e p e n d s on w h a t o n e m e a n s by public o p i n i o n . If by public opinion one m e a n s the h o p e s , fears, feelings, and reactions to events of o r d i n a r y citizens as they go about their p r i v a t e lives, then certainly there is public opinion w h e t h e r or not there is a pollster to m e a s u r e it. But if by public opinion o n e m e a n s ord i n a r y citizens w a l k i n g a r o u n d saying t o t h e m s e l v e s things like " I strongly a p prove o f the way G e o r g e Bush i s doing his j o b a s p r e s i d e n t " o r " I think w e should take a stronger s t a n d , even if it m e a n s invading N o r t h V i e t n a m , " then most of w h a t gets m e a s u r e d as public opinion d o e s not exist except in the presence of a pollster. T h e R A S model has been about both k i n d s o f public o p i n i o n , the " c o n s i d e r a t i o n s " that p e o p l e form in r e s p o n s e to t h e flow of political c o m m u n i c a t i o n s , and t h e process by w h i c h they translate typically disorganized c o n s i d e r a t i o n s into the survey r e s p o n s e s that virtually e v e r y o n e now takes as constituting p u b lic o p i n i o n . To t h e extent o n e evaluates t h e R A S m o d e l on the conventional criteria of empirical breadth and theoretical parsimony, it l o o k s quite g o o d . Its four a x i o m s can be used to explain a w i d e variety of p h e n o m e n a , including s o m e c o n c e r n i n g attitude c h a n g e , that w o u l d be difficult to explain except by essentially similar ideas. 1

T h e most i m p o r t a n t feature of t h e m o d e l is its m a r r i a g e of a t h e o r y of inform a t i o n diffusion t h r o u g h a p o p u l a t i o n that is differentially attentive to politics (axioms Al and A 2 ) , with a t h e o r y of h o w p e o p l e transform this information into survey responses ( A 3 and A 4 ) . It is the union of these t w o p r o c e s s e s that gives the m o d e l its breadth of c o v e r a g e , and that c r e a t e s the potential for further extensions and a p p l i c a t i o n s of the m o d e l . However, the model h a s t h e vice of its virtues: Breadth and p a r s i m o n y have been achieved at the e x p e n s e of significant simplification and o m i s s i o n . H e n c e , 1 An exception is Ginsberg, 1986. For an analysis of differing conceptions of public opinion, see Price, 1992.

268

N A T U R E AND ORIGINS OF MASS O P I N I O N

W H O LEADS W H O M ?

The

model

and future

research

269

c r e a t e s p r o b l e m s for my a s s u m p t i o n that the c o u r t w a s p r e d o m i n a n t l y leading r a t h e r than following m a s s opinion in the B r o w n c a s e .

A c c o r d i n g to an oft told t a l e , a n i n e t e e n t h - c e n t u r y revolutionary w a s sipping w i n e in a cafe o n e afternoon w h e n a m o b suddenly rushed past in t h e street. " M o n D i e u ! " he c r i e d , " t h e p e o p l e are on the m o v e , I am their leader, I must follow t h e m . " He then raced out into the street and to the head of t h e c r o w d . A l t h o u g h the story is perhaps a p o c r y p h a l , the p r o b l e m it raises is real. O n e regularly o b s e r v e s " l e a d e r s " at the head of m a s s m o v e m e n t s , but o n e cannot, on this e v i d e n c e a l o n e , be certain that elites are actually leading. T h e y m i g h t , of c o u r s e , be l e a d i n g , but they might also be following. W i t h o n e e x c e p t i o n , this is not a p r o b l e m I have been able to deal with syst e m a t i c a l l y in this study. I have taken the intensities of o p p o s i n g political c o m m u n i c a t i o n s to be e x o g e n o u s variables and have limited my analysis to the s u p p o s e d effects of these c o m m u n i c a t i o n s on public o p i n i o n . My w o r k i n g a s s u m p t i o n , t h u s , h a s been that elite c o m m u n i c a t i o n s shape m a s s opinion r a t h e r than vice versa. 2

I justify this w o r k i n g a s s u m p t i o n on two g r o u n d s . First, in several of the cases e x a m i n e d , it is c l e a r from the nature of the p a r t i c u l a r cases that the elite stimulus to m a s s attitude c h a n g e really w a s either mainly or wholly e x o g e n o u s . It is p e r h a p s most c l e a r in the c a s e of the I r a n - C o n t r a scandal, w h i c h p r o d u c e d c h a n g e in public attitudes toward presidential p e r f o r m a n c e and Central A m e r i c a policy, as discussed in C h a p t e r s 7 and 8. At t h e t i m e the scandal b r o k e , President R e a g a n ' s popularity was high and he had no serious p r o b l e m s on the h o r i z o n . T h e congressional attacks on R e a g a n ' s role in the I r a n - C o n t r a e v e n t s , t h u s , must have o c c u r r e d in spite of his initial standing with t h e public rather than b e c a u s e his p o p u l a r i t y w a s already low. Similar a r g u m e n t s can be m a d e for four o t h e r c a s e s :

2. T h e n u c l e a r freeze m o v e m e n t of the early 1980s involved a great deal of g r a s s r o o t s public activity. A l m o s t certainly, m a n y politicians got on t h e freeze b a n d w a g o n only b e c a u s e they perceived it to be an extremely p o p u l a r issue. T h e i m p e t u s for the freeze m o v e m e n t , however, can be clearly traced to p e o p l e w h o w o u l d have to be c o n s i d e r e d elites, n a m e l y m e m b e r s of the professional a r m s control c o m m u n i t y w h o had been losing internal d e b a t e s and calculated that politicians w o u l d not pay any attention to t h e m until they mobilized public s u p p o r t for their cause ( P r i n g l e , 1982). T h e impetus to the m a s s m o v e m e n t that led many politicians to c l i m b on t h e freeze b a n d w a g o n w a s t h u s itself the product of other elites w h o were quite self-consciously a t t e m p t i n g to lead o p i n i o n . 3. T h e bullish n e w s r e p o r t s that led to o p t i m i s t i c a s s e s s m e n t s of the U . S . e c o n o m y in late 1982 a m o n g s o m e m e m b e r s of the public w e r e no d o u b t s o m e c o m bination of hopeful p r o g n o s t i c a t i o n s by President R e a g a n and the b u s i n e s s c o m m u n i t y , and technical e c o n o m i c r e p o r t s . In neither c a s e , and especially the latter, could o n e say that these r e p o r t s were a r e s p o n s e to what t h e public already believed.

1. It is h a r d to a r g u e that the S u p r e m e C o u r t ' s decision in Brown v. Board of Education w a s a r e s p o n s e to p o p u l a r pressure for an end to S o u t h e r n s c h o o l segr e g a t i o n . Inside a c c o u n t s indicate that, on the contrary, the C o u r t perceived the public as largely hostile to federal intervention on this issue. T h e C o u r t d i d indeed have a p r o b l e m with public o p i n i o n , but the p r o b l e m w a s not h o w far it w o u l d be forced to go in o p p o s i n g d i s c r i m i n a t i o n but how far it could go without p r o v o k i n g a political b a c k l a s h (Kluger, 1975). To be sure, the N a t i o n a l Association for the A d v a n c e m e n t of Colored People w a s pressuring for a d e s e g r e g a t i o n r u l i n g , but it h a d , at that point, minimal political clout. T h u s , the B r o w n d e c i sion appears to have b e e n m o t i v a t e d p r i m a r i l y by the m o r a l c o n v i c t i o n s of the j u s t i c e s . H o w these c o n v i c t i o n s f o r m e d , and why they took the form they did in the m i d - 1 9 5 0 s , is o b v i o u s l y a d e e p and i m p o r t a n t q u e s t i o n , but it is not o n e that

4. T h e Bush a d m i n i s t r a t i o n ' s mobilization of public support for t h e Persian Gulf W a r is o n e of the m o s t striking cases of elite opinion leadership e x a m i n e d in this b o o k . W h e n Iraq invaded Kuwait in A u g u s t , 1990, only a small fraction of A m e r i c a n s w a s aware that K u w a i t existed ( L a n g , L a n g , and L a r s o n , 1991). Yet within two w e e k s , public s u p p o r t for the use of A m e r i c a n troops to prevent further Iraqi aggression w a s t o p p i n g 80 percent in t h e polls, and this s u p p o r t eventually c a r r i e d over into solid backing for a w a r to liberate K u w a i t (Mueller, 1992). W h a t m a k e s this mobilization of m a s s opinion impressive is that it w a s a c c o m p l i s h e d in the a b s e n c e of any c o m m u n i s t threat, which for 40 years had been the standard justification for the use of A m e r i c a n troops a b r o a d , and d e spite the fact t h a t , before t h e w a r b e g a n , m o s t p e o p l e e x p e c t e d it to be costly in A m e r i c a n lives. Since no polls were apparently taken in the period b e t w e e n the initial Iraqi invasion and President B u s h ' s d e c l a r a t i o n that the aggression would not be allowed to s t a n d , t h e r e is no e m p i r i c a l basis for ruling out the possibility that Bush w a s simply r e s p o n d i n g to a g r o u n d s w e l l of public d e m a n d for a firm A m e r i c a n r e s p o n s e . However, few informed o b s e r v e r s of these events are likely to take this possibility seriously. T h e Persian G u l f W a r w a s , by all a p p e a r a n c e s , a c a s e in w h i c h a politically skilled p r e s i d e n t , with t h e s u p p o r t of the press and m o s t other e l i t e s , handily s h a p e d m a s s o p i n i o n .

2 In the analysis of campaign effects in House elections, I controlled for incumbents' prior victory margins in order to find out whether the most intense campaigns were directed against incumbents who were already weak. The evidence indicated essentially no reason for concern in this particular case. See footnote 10 in Chapter 10.

T h e general point here is t h a t , however difficult it m a y be to resolve the direction of e l i t e - m a s s influence in the abstract, it is often possible to m a k e plausible j u d g m e n t s in p a r t i c u l a r c a s e s .

271

N A T U R E AND ORIGINS OF MASS O P I N I O N

The model and future research

S e c o n d , even w h e n politicians pursue a certain policy b e c a u s e of perceived public pressure to do s o , it is often the anticipation of that p r e s s u r e , rather than actual p r e s s u r e , that is critical. T h i s is a point that V. O. Key (1961) has e m p h a s i z e d . T h e public opinion to which elites try to be responsive is not the public opinion that is reflected in polls at the point of m a k i n g d e c i s i o n s , but t h e public opinion that an o p p o n e n t might be able to call into existence at the next e l e c t i o n . Politicians w h o slavishly follow existing opinion a r e , as K e y a r g u e s , likely to fare p o o r l y in the long r u n .

h o m e in A r k a n s a s . In fact, the e v i d e n c e indicates that he t o o k his a n t i w a r p o sition in spite of the best advice of his political staff (but b e c a u s e of advice he w a s getting from his foreign policy a d v i s o r y s t a f f ) . A n o t h e r early m a i n s t r e a m o p p o n e n t of t h e war, R e p r e s e n t a t i v e Tip O ' N e i l l of M a s s a c h u s e t t s , likewise relates that his e a r l y o p p o s i t i o n w a s a r e s p o n s e to w h a t military a u t h o r i t i e s privately told h i m about t h e futility of the war, and in spite of his belief that opposition to t h e w a r w a s political suicide ( O ' N e i l l , 1987: p. 195). T h e m e d i a w e r e , o f c o u r s e , another i m p o r t a n t source o f a n t i w a r c o m m u n i c a tions. To a very c o n s i d e r a b l e extent, however, m e d i a r e p o r t i n g reflected, as it always d o e s , w h a t the m e d i a ' s sources w e r e saying. For e x a m p l e , C h a r l e s Mohr, a c o r r e s p o n d e n t for Time and later for the New York Times, h a s written of t h e a n t i w a r r e p o r t i n g that o r i g i n a t e d a m o n g r e p o r t e r s in South V i e t n a m :

270

Let me take a s i m p l e c a s e of such a n t i c i p a t o r y influence, followed by a m o r e c o m p l i c a t e d o n e . President N i x o n ' s surprise a n n o u n c e m e n t o f w a g e and price controls in s u m m e r 1 9 7 1 , as described at the b e g i n n i n g of C h a p t e r 6, w a s surely an act calculated to o b v i a t e criticism of his record on handling the e c o n o m y in the u p c o m i n g presidential e l e c t i o n . In this sense N i x o n was r e s p o n d i n g to public o p i n i o n . Yet he w a s also leading it, since, as we saw, his s p e e c h had a definite impact in m o v i n g public opinion further in the direction of s u p p o r t for controls. N o w consider the infinitely more c o m p l i c a t e d c a s e of the V i e t n a m War. F r o m G e l b ' s account ( G e l b w i t h B e t t s , 1979), we k n o w that President J o h n s o n w a s well advised on the difficulties of trying to use A m e r i c a n military force to stabilize a n o n c o m m u n i s t g o v e r n m e n t in South V i e t n a m . J o h n s o n w a s also well a w a r e that there w a s no groundswell of p o p u l a r s u p p o r t for sending troops to V i e t n a m . Yet J o h n s o n f e i r e d that if he p e r m i t t e d South V i e t n a m to fall to the N o r t h V i e t n a m e s e , he w o u l d open himself to the c h a r g e of being " s o f t on c o m m u n i s m " and having " l o s t V i e t n a m " i n the s a m e way that President T r u m a n had supposedly " l o s t C h i n a . " Q u i t e likely, J o h n s o n feared, this w o u l d d o o m his G r e a t Society d o m e s t i c a g e n d a and set off a n e w round of M c C a r t h y i s t r e c r i m inations in which he w o u l d be the loser. In this situation, J o h n s o n led as well as followed the public into a w a r that neither he nor it w a s e a g e r to p r o s e c u t e . He w a s following public opinion in the sense that he a n t i c i p a t e d a harsh retrospective j u d g m e n t if he allowed a c o m munist takeover, but leading public opinion in the sense that, in o r d e r to avoid this j u d g m e n t , he had to mobilize public s u p p o r t for policies that w e r e by no m e a n s i m m e d i a t e l y popular. If, t h e n , J o h n s o n w a s , in a s e n s e , genuinely leading public opinion on the war, perhaps a case can be m a d e that a n t i w a r politicians were the o n e s w h o were following public opinion m o r e than leading it. T h o u g h m o r e p l a u s i b l e , t h i s , t o o , is a hard case to m a k e . For it w a s at the high tide of p o p u l a r support for the w a r in 1966 that m a i n s t r e a m politicians began to o p p o s e the war. Further, the earliest m a i n s t r e a m o p p o n e n t s of t h e war were from c o n s t i t u e n c i e s that could not be d e scribed as hotbeds of a n t i w a r feeling. Take the c a s e of S e n a t o r J. W i l l i a m Fulbright of A r k a n s a s , the earliest i m p o r t a n t political figure to turn against the war. Fulbright must have m a d e a calculation that an a n t i w a r position w o u l d not lead to i n s u r m o u n t a b l e p r o b l e m s with his constituents. Yet it is i m p l a u s i b l e to think that Fulbright w a s in any m o r e direct way responding to a n t i w a r pressures back

3

The debate was not essentially, as some seem to believe, a quarrel between the press and the U.S. officials in Vietnam. It was, rather, a quarrel between factions within the U.S. Mission. For the most part, field advisers closest to the action and to the Vietnamese took the pessimistic view. Some of the more senior officials in Saigon, who were reporting to Washington on the progress of the programs they themselves were administering, were publicly and persistently optimistic. The reporters quickly became aware of this dispute, because brilliant young field officers, as exemplified by the late John Paul Vann, increasingly turned to the journalists. The reporters did not invent the somber information that sometimes appeared in their stories. (1983: p. 56) Since these sources w e r e mostly in the military, the C I A , and t h e State D e p a r t m e n t (Hallin, 1986; M o h r 1983; H a l b e r s t a m , 1979), it seems unlikely that their leaks of a n t i w a r i n f o r m a t i o n w e r e in any simple sense a response to c h a n g e s in 4

public o p i n i o n . Perhaps the m o s t i m p o r t a n t exception to the claim that press r e p o r t i n g tends to reflect press sources o c c u r r e d d u r i n g t h e Tet offensive in early 1968, w h e n the press depicted w h a t w a s apparently a decisive A m e r i c a n military victory a s , instead, a serious defeat ( B r a e s t r u p , 1979). T h i s apparent m i s r e p o r t i n g of the n e w s in an a n t i w a r d i r e c t i o n d o e s not, however, a p p e a r to have been a r e s p o n s e to public opinion in t h e U n i t e d S t a t e s , which still largely s u p p o r t e d the war at the t i m e of Tet. Rather, r e p o r t i n g a p p e a r s to have reflected the press c o r p s ' surprise that t h e Viet C o n g were able to m o u n t an apparently credible n a t i o n w i d e military offensive after A m e r i c a n officials had been saying that the Viet C o n g were all but defeated. (It is c o n v e n t i o n a l w i s d o m in A m e r i c a n electoral politics that any c o m p e t i t o r w h o d o e s " b e t t e r than e x p e c t e d " by the press finds h i m - or herself t h e beneficiary of an excited o v e r r e a c t i o n by the press; see Polsby, 1983.) N o n e of this is to deny a significant d e g r e e of elite responsiveness to m a s s o p i n i o n , p a r t i c u l a r l y in the later p h a s e s of t h e V i e t n a m War. Most likely, t h o u g h it is difficult to i m a g i n e h o w anyone could rsally be sure, m a r g i n a l r e d u c t i o n s in public s u p p o r t for the w a r m a d e it i n c r e m e n t a l l y safer for those politicians 3 Halberstam, 1972: p. 420. 4 For an account of the disdain with which the foreign policy elite regard public opinion, see Cohen, 1973.

272

The

N A T U R E A N D ORIGINS OF MASS O P I N I O N

predisposed against t h e w a r to go public with their a n t i w a r sentiments. P r e s u m ably a similar interaction o c c u r r e d b e t w e e n press and public, t h o u g h it is again i m p o r t a n t to note that the p r e s s , especially the elite p r e s s , w a s c a r r y i n g significant a m o u n t s of a n t i w a r information in 1965 and 1966, well before there w a s any indication that d e c l i n i n g public s u p p o r t for t h e war was m a k i n g it safe to do so. E a c h increment of additional elite opposition m a d e safe by c h a n g e s in public opinion may then have p r o d u c e d an additional increment of c h a n g e in m a s s o p i n i o n , which m a y have further e m b o l d e n e d t h e potential elite o p p o s i t i o n , and so forth. T h e late entry of R o b e r t K e n n e d y as an a n t i w a r c a n d i d a t e in t h e 1968 D e m o c r a t i c presidential p r i m a r i e s m a y be an e x a m p l e of a politician w h o acted only after p r e l i m i n a r y s o u n d i n g s of public opinion - in the form of E u g e n e M c C a r t h y ' s success in t h e N e w H a m p s h i r e p r i m a r y - signaled that an a n t i w a r p o sition w a s politically p r o m i s i n g .

model

and future

research

p r o c e e d s via an e l i t e - d o m i n a t e d m e d i a or by h u m b l e w o r d - o f - m o u t h ,

273 the

r e c e p t i o n - a c c e p t a n c e m o d e l can capture the joint effects of both m o d e s of 5

S u c h e l i t e - m a s s d y n a m i c s are by no m e a n s inconsistent with an a s s u m p t i o n of elite opinion l e a d e r s h i p . O n e a r g u m e n t m i g h t be that w e a k e n i n g public s u p port for the V i e t n a m W a r c r e a t e d discretion for elites to m o v e in a similar direction. T h e elites w h o t o o k a d v a n t a g e of this discretion w o u l d have been those predisposed by ideology or analysis to do so. T h u s , elites act a u t o n o m o u s l y to shape o p i n i o n , but only after they calculate that it is safe to do s o . An a l t e r n a t i v e a r g u m e n t for elite leadership would posit a differentially a u t o n o m o u s elite, s o m e p a r t s o f w h i c h , p r i m a r i l y ncluding u p w a r d l y m o b i l e politicians, are s o m e w h a t responsive to m a s s o p i n i o n , and other p a r t s of w h i c h , including m o s t foreign policy e x p e r t s , military officers, and g o v e r n m e n t officials critical of the war, are largely independent of public opinion. T h e m o r e a u t o n o m o u s s e g m e n t s of the elite, by this a r g u m e n t , use the media to g e n e r a t e public pressure to w h i c h other, m o r e responsive e l i t e s , the politicians, then a c c e d e . A weak form of t h e argum e n t for elite opinion leadership - and o n e that s e e m s to me m u c h t o o weak w o u l d be that a n t i w a r elites were simply r e s p o n d i n g to m a s s o p i n i o n , but that their responses t o o k t h e form of a n t i w a r m e s s a g e s , which diffused t h r o u g h t h e m e d i a in a m a n n e r specified by the r e c e p t i o n - a c c e p t a n c e p r o c e s s , thereby prod u c i n g further m a s s c h a n g e , which in turn further pressured elites t o w a r d c h a n g e . By this w e a k a r g u m e n t , it would be a n o n a u t o n o m o u s elite that w a s influencing m a s s o p i n i o n .

diffusion. T h e allusion to g e r m s , incidentally, is not fanciful. T h e r e c e p t i o n - a c c e p t a n c e m o d e l , as formalized in this b o o k , is a g e n e r i c diffusion m o d e l . G i v e n a p p r o priate m e a s u r e s of e x p o s u r e and resistance, it m i g h t be used to d e s c r i b e the diffusion of new fashions, religious beliefs, or even diseases. W i t h i n an adult p o p u l a t i o n , for e x a m p l e , susceptibility to a socially transmitted d i s e a s e might be n o n m o n o t o n i c with respect to age: Young adults m i g h t be high on the social interactions associated with exposure to the d i s e a s e but also healthy e n o u g h to resist it; the old might be v u l n e r a b l e but socially reclusive; and so the m i d d l e - a g e d might be hardest hit. T h e incidence of A I D S m a y present an e x a m p l e of this sort of interaction, since r a w d a t a published by t h e U . S . C e n t e r s for Disease Control suggests that the incidence of A I D S w a s , at least for a t i m e , n o n m o n o t o n i c with respect to age ( " F e d e r a l P l a n n i n g , " 1987). W h a t all this indicates is that a close fit b e t w e e n the r e c e p t i o n - a c c e p t a n c e m o d e l and d a t a on m a s s attitude c h a n g e d o e s not, by itself, indicate anything about the extent of elite influence. T h e model is simply a m e a n s for specifying the process of diffusion of elite influence, w h e r e the r e a s o n s for believing that it is elite influence, r a t h e r than s o m e t h i n g e l s e , that is diffusing must be s u p plied independently of the m o d e l . (See the discussion on p p . 4 3 - 4 . ) T h e r e is, as I have b e e n arguing (and argued e a r l i e r in C h a p t e r 2 ) , an a m p l e supply of such reasons. A n d yet, a fully a d e q u a t e account of elite opinion leadership is not o n e that sees a public that responds to elite initiatives in a c o m pletely m e c h a n i c a l fashion, t h o u g h a positive r e s p o n s e by the public to wellcrafted initiatives can usually be c o u n t e d u p o n . Rather, it is an account in w h i c h elites - always h a v i n g s o m e ideas that are a u t o n o m o u s l y their o w n , always p o tentially split a m o n g t h e m s e l v e s along p a r t i s a n lines and m a n e u v e r i n g for partisan a d v a n t a g e , and a l w a y s looking over their shoulders to see w h a t the public is thinking and might think in the future - a t t e m p t to lead and to follow at the s a m e t i m e . Or, to put the point s o m e w h a t differently, the q u e s t i o n is not w h e t h e r elites lead or follow, but how much and which elites lead r a t h e r than follow m a s s o p i n i o n , and under what circumstances they do so. A full analysis

It is i m p o r t a n t to r e c o g n i z e that, strictly s p e a k i n g , o n e cannot use the results of my m o d e l i n g to s u p p o r t even a weak a r g u m e n t for elite influence. T h e reason is that the R A S m o d e l , as f o r m u l a t e d , is entirely agnostic about the sources of the political c o m m u n i c a t i o n s that m o v e public o p i n i o n . T h u s , it is consistent with the model that e l i t e s , including even t h e m e d i a , have no influence w h a t soever on public o p i n i o n , and that all political c o m m u n i c a t i o n s diffuse t h r o u g h the p o p u l a t i o n by personal c o n t a c t , like g e r m s . So long as political a w a r e n e s s is positively associated w i t h reception of political c o m m u n i c a t i o n s and negatively associated with a c c e p t a n c e - as it would be in either a g e r m or elite-driven scen a r i o - the model applies. T h i s m a y be part of the reason the m o d e l p e r f o r m s so well: B e c a u s e the diffusion process has t h e s a m e formal p r o p e r t i e s w h e t h e r it

5 It should be noted, however, that the classic studies of personal influence provide no mechanism for explaining how interpersonal communications could bring about systematic changes in the distribution of mass opinion independently of elite influence. Rather, these studies have emphasized the role of personal discussion either as a conservatizing force that buttresses the individual against external influence, as in the Berelson, Lazarsfeld, and McPhee (1954) study titled Voting, or as a mediating force, as in the Katz and Lazarsfeld (1955) study titled Personal Influence, which shows how local opinion leaders pick up ideas in the media and pass them on to friends. In the first case, no attitude change occurs, while in the second, opinion leaders may be said merely to aid in the diffusion of elite-generated information. With respect to the second point, MacKuen and Brown (1987) conclude that political discussion is basically another "information channel" (p. 485): "[T]he social circle is not an active independent force in politics but instead shapes the information that comes from outside. Thus it is proper to think of the social environment not as a source of influence but as an intervening mechanism in a larger communication system" (p. 483).

274

N A T U R E AND ORIGINS OF MASS O P I N I O N

of these p r o b l e m s is well beyond the scope of this b o o k . My only aim in this brief discussion h a s b e e n , first, to m a k e c l e a r w h a t exactly my formal m o d e l s of attitude diffusion do and do not prove about elite opinion l e a d e r s h i p , and seco n d , to m a k e it a p p e a r plausible that, in a broad r a n g e of c i r c u m s t a n c e s , the a m o u n t of a u t o n o m o u s leadership exercised by elites is substantial. (I return to this p r o b l e m in C h a p t e r 12.)

CRITICAL REVIEW OF BASIC AXIOMS Critical evaluation of the R A S model m a y be o r g a n i z e d a r o u n d its defining axi o m s , beginning with A l , the R e c e p t i o n A x i o m . T h e a s s e r t i o n i n A l i s that greater attention to an issue is associated with greater reception of m e s s a g e s c o n c e r n i n g that issue. T h i s claim would s e e m initially u n e x c e p t i o n a b l e . Yet w h a t exactly is it that people receive? A l m o s t certainly, different people can be exposed to t h e s a m e m e s s a g e and yet receive quite different m e s s a g e s , or even no intelligible m e s s a g e , d e p e n d i n g on their prior k n o w l e d g e about the iss u e . T h e r e are at least t w o ways in which this can h a p p e n . First, as noted in C h a p t e r 8, p e o p l e m a y better c o m p r e h e n d stories on subjects about which they have m o r e initial familiarity. T h u s , for e x a m p l e , m o r e p e o p l e l e a r n e d about the s e n t e n c i n g of the colorful M a r i n e , Oliver N o r t h , for his c o n v i c t i o n in t h e I r a n C o n t r a controversy than learned the reason for the resignation of the c o m p a r a tively bland H o u s e S p e a k e r Jim W r i g h t , even t h o u g h the latter story played out o v e r a period of several m o n t h s and received far more extensive c o v e r a g e . (In fact, o n e poll s u g g e s t e d that almost as m a n y p e o p l e k n e w the n a m e of Oliver N o r t h ' s striking secretary, F a w n H a l l , as k n e w why W r i g h t resigned the speake r s h i p . ) S e c o n d , and m o r e speculatively, individuals w h o are e x p o s e d to the s a m e m e s s a g e may, if they take notice of it, perceive it differently. For e x a m p l e , the air force m a y stage an e l a b o r a t e spectacle for the maiden flight of the B 2 , its n e w flying w i n g b o m b e r , to show that the plane really w o r k s . S o m e m e m b e r s of the public m a y receive this as a p r o - B 2 m e s s a g e , but others may " s e e " in the flight " a n o b v i o u s l y useless b a t m o b i l e . " In b o t h of these c a s e s , differences in reception (given equal attentiveness to the s a m e m e s s a g e ) will d e p e n d on p e o p l e ' s previously existing ideas (or s c h e m a t a ) w h i c h m a y differ both in content and degree of d e v e l o p m e n t across individuals (Fiske and Kinder, 1981). T h e R A S m o d e l , as presently constituted, m a k e s no allowance for these w a y s in w h i c h prior opinion m a y affect reception or p a r t i s a n perception of the c o m m u nications one e n c o u n t e r s . C o n s i d e r next the Resistance A x i o m , A 2 , w h i c h claims that p e o p l e can resist persuasion only to the extent that they have acquired an a p p r o p r i a t e c u e i n g m e s s a g e , w h e r e c u e i n g m e s s a g e s are a s s u m e d to have an origin that is, at least in p r i n c i p l e , s e p a r a t e from the persuasive m e s s a g e itself. T h i s specification of the Resistance A x i o m is intended to follow C o n v e r s e ' s (1964) notion that p e o p l e need to acquire contextual k n o w l e d g e of " w h a t g o e s with w h a t " in order to d e v e l o p conventionally l e f t - r i g h t belief s y s t e m s . W i t h

The

model

and future

research

275

the inference (from A 1 ) that m o r e aware persons are m o r e likely to possess c u e ing i n f o r m a t i o n , this a x i o m played a crucial role in explaining attitude c h a n g e . All else e q u a l , m o r e a w a r e persons are m o r e likely to resist m e s s a g e s , given rec e p t i o n of t h e m , that are inconsistent with their basic values - that is, to exhibit p a r t i s a n resistance. But how can we be certain that it is possession of relevant contextual inform a t i o n , rather than s o m e t h i n g e l s e , that explains the greater critical resistance of m o r e aware p e r s o n s ? As the reader may have n o t i c e d , I have been u n a b l e to p r o d u c e any direct e v i d e n c e on this point, that is, any d a t a that credibly m e a s u r e possession or n o n p o s s e s s i o n of a p a r t i c u l a r cueing m e s s a g e and link it to t h e a c c e p t a n c e or n o n a c c e p t a n c e of a new c o n s i d e r a t i o n . I h a v e , however, pointed to abundant e v i d e n c e of the i m p o r t a n c e of contextual information about m e s s a g e sources in m a s s c o m m u n i c a t i o n s - e v i d e n c e that inf o r m a t i o n or cues about t h e person providing a m e s s a g e greatly affects h o w p e o ple respond to it. So my e m p h a s i s on c u e i n g m e s s a g e s rests on a solid foundation of past r e s e a r c h . I h a v e , moreover, been able to p r o d u c e a fair a m o u n t of indirect e v i d e n c e of the i m p o r t a n c e of c u e i n g i n f o r m a t i o n . An i m p o r t a n t piece of this e v i d e n c e c o m e s from the V i e t n a m War. W h e n a m a i n s t r e a m elite c o n s e n s u s existed, exposure to the elite c o n s e n s u s , as m e a s u r e d by i n d i v i d u a l s ' levels of political a w a r e n e s s , w a s associated with greater s u p p o r t for the war, with no resistance a m o n g the m o s t politically a w a r e . W i t h the app e a r a n c e of elite ideological d i s a g r e e m e n t s in 1966, however, politically sophisticated liberals b e g a n to resist prowar m e s s a g e s and to accept a n t i w a r ones. T h u s , elite cues functioned to activate ideological predispositions a m o n g the p o litically a w a r e . T h e e x a m i n a t i o n of the m a i n s t r e a m and p o l a r i z a t i o n m o d e l s in C h a p t e r 6 provided several other cases in which political a w a r e n e s s either w a s or w a s not a s sociated with political p o l a r i z a t i o n , d e p e n d i n g on the configuration of elite c u e s . Perhaps the most striking of these c a s e s involved attitudes on r e c o g n i t i o n of C o m m u n i s t C h i n a in 1972. Political a w a r e n e s s w a s not a s s o c i a t e d with resistance to this policy, even t h o u g h it is e a s y to imagine that s o m e politically sophisticated c o n s e r v a t i v e s w o u l d have eagerly resisted it if a p p r o p r i a t e ideological cues had been g i v e n . 6

T h u s , most of my e v i d e n c e for the i m p o r t a n c e of cueing m e s s a g e s derives from e x a m i n a t i o n of the attitudes of the politically a w a r e , w h o are extremely responsive to w h a t ideologically congenial elites urge t h e m to b e l i e v e . In one type of c a s e , however, the entire m a s s public - and not j u s t t h e most politically attentive s e g m e n t - relies heavily on p a r t i s a n cues. This is the c a s e of contested e l e c t i o n s , w h e r e all citizens are about equally p a r t i s a n in their t e n d e n c y to accept t h e information supplied by their o w n p a r t y ' s c a n d i d a t e and to reject that of the opposition p a r t y ' s c a n d i d a t e . 6 Congressional Quarterly reports "a rising chorus of praise" for establishing relations with China, but also some criticism. Among the critics were John Ashbrook, an aspirant for the Republican presidential nomination, and columnist William F. Buckley (March 4, 1972: 472-3).

276

N A T U R E AND ORIGINS OF MASS O P I N I O N

It is t h u s clear that a w a r e n e s s is associated with resistance to persuasion in s o m e cases but not others. If o n e rejects my explanation - that cueing m e s s a g e s are subtle in s o m e cases (typical p a r t i s a n issues, such as j o b g u a r a n t e e s ) , c o m pletely o b v i o u s in others (partisan e l e c t i o n s ) , and nonexistent in still others ( m a i n s t r e a m issues), so as to m a k e highly a w a r e persons m o r e c a p a b l e of partisan resistance in the first type of situation but not in the second or third - then it b e c o m e s n e c e s s a r y to p r o p o s e s o m e other explanation for why a w a r e n e s s has this pattern of resistance effects. S i n c e , for the m o m e n t , no a l t e r n a t i v e explanation is a p p a r e n t , it is r e a s o n a b l e to accept the idea that, in m a n y but not all situations, awareness brings with it esoteric k n o w l e d g e of " w h a t g o e s with w h a t , " and that this contextual k n o w l e d g e is an i m p o r t a n t e l e m e n t of resistance to p e r s u a s i o n . A m o r e serious c r i t i c i s m of the Resistance A x i o m can be m a d e on conceptual g r o u n d s . Let us a s s u m e , as I suggested earlier, that s o m e people see in the m a i d e n flight of t h e B2 a comical b a t m o b i l e - and see t h i s , I should a d d , precisely because they have e a r l i e r received a n d , as d o v e s , accepted stories about p o o r p e r f o r m a n c e by h i g h - t e c h n o l o g y w e a p o n s - while others see in t h e flight a magnificent flying m a c h i n e . People w h o s e p r e c o n c e p t i o n s so influence w h a t they perceive do not then need to e n g a g e in a s e p a r a t e evaluation to see w h e t h e r the m e s s a g e should be accepted as a p r o - B 2 c o n s i d e r a t i o n . Perception and evalu a t i o n constitute for t h e m a single, s c h e m a - d r i v e n s t e p . H e n c e , it m a y reasonably be objected that t h e R A S m o d e l , in s h a r p l y distinguishing the reception step from the a c c e p t a n c e s t e p , is creating a purely theoretical d i s t i n c t i o n . T h e a l t e r n a t i v e w o u l d be to design a theory in which the internalization of prior, ideologically cued c o m m u n i c a t i o n s affects both the likelihood of receiving a m e s s a g e and the form in which the m e s s a g e will be perceived. T h u s , if s o m e o n e hears from a trusted source that the B2 is a costly and useless c o l o s s u s , the person may be m o r e likely to notice future n e w s about the B2 and to interpret that n e w s in w a y s consistent with this initial information bias. T h e r e is m u c h research on mental o r g a n i z a t i o n in social p s y c h o l o g y indicating that perception is t h e o r y - l a d e n in this way. It w o u l d therefore be valuable to graft a theory of perception o n t o the R A S m o d e l . T h e lack of such a t h e o r y is, in my opinion, one of the greatest deficiencies of the present m o d e l . T h e Accessibility A x i o m , A 3 , is perhaps the most defensible of the four axioms. T h e notion that ideas that have been used recently or frequently are m o r e readily recalled from m e m o r y is extremely well s u p p o r t e d both by e x p e r i m e n t a l p s y c h o l o g y (as in W y e r and Srull, 1989) and by research on political attitudes (Iyengar and Kinder, 1987; Iyengar, 1991), and the R A S m o d e l simply capitalizes on this e v i d e n c e . T h e r e must, however, be more to accessibility than recency of a c t i v a t i o n . For e x a m p l e , the studies of Wilson et al. (for reviews see W i l s o n , Kraft, and D u n n , 1989; Wilson et a l . , 1989) and Millar and Tesser (1986) clearly indicate that p e o ple call to mind different c o n s i d e r a t i o n s w h e n a s k e d to a r t i c u l a t e the r e a s o n s for their attitudes than w h e n asked to think about their feelings on a subject. T h i s

The

model

and future

research

211

w o u l d s e e m to indicate, unsurprisingly, that m e m o r y search d e p e n d s on subtle features of the task in w h o s e service it is being p e r f o r m e d . It s e e m s likely that m e m o r y search is also affected by such m a t t e r s as m o o d , social c o n t e x t , and other recent activities, even if unrelated to the task at h a n d . I have finessed these potential c o m p l i c a t i o n s by a s s u m i n g a large d e g r e e of simple r a n d o m n e s s in t h e m e m o r y search process - or, as I have expressed it, r a n d o m n e s s in " t h e ideas that happen to be at the top of o n e ' s head at a given m o m e n t . " A l t h o u g h this s e e m s to me a c c e p t a b l e as an initial strategy, future m o d e l builders might profitably pay more attention to how m e m o r y search o c curs. T h e expected payoff w o u l d be the ability to explain more of the instability over t i m e that is associated with p e o p l e ' s survey responses. A n o t h e r s h o r t c o m i n g of the Accessibility A x i o m is that, since it p e r m i t s the activation of o n e idea to increase the accessibility of related or similar i d e a s , it is natural to w o n d e r w h a t exactly it is that d e t e r m i n e s w h e n ideas are related or similar. T h i s is s o m e t h i n g that the R A S m o d e l d o e s not currently a d d r e s s but m i g h t , with greater attention to the r e p r e s e n t a t i o n of ideas in the m i n d , eventually be able to address. It might well turn out t h a t , for e x a m p l e , w h i c h ideas are related to one a n o t h e r is less a function of logic or linguistic similarity than of elite cues and other features of the external e n v i r o n m e n t . But since, n o t w i t h s t a n d i n g these l i m i t a t i o n s , A3 s e e m s to a c c o m p l i s h its role in t h e m o d e l quite adequately, there w o u l d s e e m less need to t a m p e r with it than with s o m e of the o t h e r a x i o m s . T h e R e s p o n s e A x i o m , w h i c h claims that individuals a n s w e r survey q u e s t i o n s by averaging across the c o n s i d e r a t i o n s that are most i m m e d i a t e l y accessible in m e m o r y , is the most c o n t e s t a b l e of the four a x i o m s . O n e i m p o r t a n t s h o r t c o m i n g is that it provides no m e a n s of taking into account either the " s t r e n g t h " or the " e x t r e m i t y " of the attitude statements p e o p l e are willing to m a k e . T h a t i s , it allows people to favor o n e or the other side of an issue, but not to take positions that are m o r e or less s t r o n g , or m o r e or less e x t r e m e . T h e r e is nothing inherently w r o n g with this b i n a r y a p p r o a c h . M u c h of the t i m e w h e n public opinion enters into elite p o l i t i c s , it enters as a single number, the p e r c e n t a g e w h o a p p r o v e of the p r e s i d e n t ' s h a n d l i n g of his j o b , or w h o s u p port b u s i n g to achieve racial i n t e g r a t i o n , or w h o want an i m m e d i a t e pullout from V i e t n a m . E l e c t i o n s , as c o n d u c t e d in the United S t a t e s , are likewise agg r e g a t i o n s o f 0 - 1 choices. N o n e t h e l e s s , it w o u l d be valuable to take account of attitude strength and extremity within the R A S m o d e l . O n e a p p r o a c h w o u l d b e t o a s s u m e that the extremity or strength of an individual's attitude r e p o r t v a r i e s with t h e net directional thrust of h e r c o n s i d e r a t i o n s and t h e total n u m b e r of salient conside r a t i o n s . " A p r e p o n d e r a n c e o f mutually consistent c o n s i d e r a t i o n s " w o u l d thus be t h e c a u s e of e x t r e m e or strong attitude r e p o r t s . I am aware of no d a t a for testing this idea in c o n n e c t i o n with political issues, but it is readily tested in c o n n e c t i o n with e v a l u a t i o n s of politicians. In the 1984 N E S election study, for e x a m p l e , r e s p o n d e n t s w e r e asked their likes and dislikes

278

N A T U R E AND ORIGINS OF MASS O P I N I O N

Table and

11.1. underlying

The relationship considerations

0=25. Average Net of likes and dislikes Average Sum of likes and dislikes N

between

evaluational

The

extremity

Rating of Reagan on 100-point scale 26-35 3fr43 46-55 56-65 66-75 75-100

-2.75

-2.09

-1.78

-0.91

-0.52

0.39

1.90

3.02

2.97

2.68

1.76

2.69

2.98

3.14

287

123

146

234

245

419

757

Source: 1984 NES survey.

about Ronald R e a g a n and then i m m e d i a t e l y afterward were asked to evaluate h i m on a 100-point feeling t h e r m o m e t e r . R e s p o n s e s to the former q u e s t i o n can be taken as a m e a s u r e of the n u m b e r and thrust of c o n s i d e r a t i o n s relating to R e a g a n that were i m m e d i a t e l y salient in the r e s p o n d e n t ' s m i n d , while r e s p o n s e s to the latter can be used to m e a s u r e attitude extremity. Table 11.1 presents the relevant data. As can be seen in the top p a n e l , there is a strong positive r e l a t i o n s h i p between net R e a g a n likes/dislikes and the overall w a r m t h of feelings toward h i m . As the b o t t o m panel further s h o w s , the relationship b e t w e e n total n u m b e r of Reagan c o n s i d e r a t i o n s and feelings scores is strongly n o n m o n o t o n i c . T h u s , p e o p l e with the largest average n u m b e r of R e a g a n likes and dislikes evaluated h i m most extremely. This little analysis raises the possibility that attitude extremity m i g h t be a relatively simple function of the n u m b e r and direction of accessible c o n s i d e r a t i o n s . A n o t h e r difficulty with my simple top-of-the-head response rule is that it is o p e n l y at odds with a c o n s i d e r a b l e research literature which argues that, instead of m a k i n g " m e m o r y - b a s e d " j u d g m e n t s from accessible i d e a s , as in A 4 , p e o p l e ' s j u d g m e n t s are n o r m a l l y m a d e " o n - l i n e " at the point of initial p r o c e s s i n g of i n c o m i n g information ( H a s t i e and P a r k , 1986; Lichtenstein and S r u l l , 1987; L o d g e , M c G r a w , and S t r o h , 1989; M c G r a w and Pinney, 1990). W h e n a s k e d to express an attitude, p e o p l e simply retrieve their current evaluation and r e p o r t it, rather t h a n , as I have c l a i m e d , construct their attitude reports on the basis of the ideas most i m m e d i a t e l y salient to t h e m . T h e on-line m o d e l is obviously a plausible o n e , but not necessarily equally plausible for every type of p r o b l e m . It is n o t a b l e , first of all, that m o s t of the e v i d e n c e H a s t i e and P a r k cite involves nonpolitical subjects, especially personality e v a l u a t i o n . As they point out, it is extremely difficult to prevent p e o p l e from forming integrated evaluations of other personalities as information c o m e s in. A p p a r e n t l y this is s o m e t h i n g that people do naturally and effortlessly. It is by no m e a n s o b v i o u s , however, that the s a m e is true of political c o n t r o versies. In fact, t h e o p p o s i t e may be nearer t h e truth. Indirect e v i d e n c e from a variety of sources indicates that it is difficult u n d e r the best of c i r c u m s t a n c e s to

model

and future

research

279

induce citizens to form c o h e r e n t impressions about political subjects ( s e e , for e x a m p l e , L a n e , 1962; C o n v e r s e , 1964; H o c h s c h i l d , 1981: ch. 9 ) . P a r k and H a s t i e do specify special c o n d i t i o n s in which p e o p l e m a k e evaluations on the basis of ideas i m m e d i a t e l y available in m e m o r y , namely, w h e n p e o ple e n c o u n t e r information that is so u n i m p o r t a n t to t h e m that they d o n ' t b o t h e r to k e e p u p d a t e d e v a l u a t i o n s but are then u n e x p e c t e d l y asked to state an o p i n i o n . W h e n this h a p p e n s , p e o p l e are forced to rely on ideas recalled from m e m o r y as the basis of attitude r e p o r t s . I suggest that these c o n d i t i o n s are t h e n o r m in p o litical information p r o c e s s i n g . In fact, the issues on which survey r e s p o n d e n t s are invited to offer o p i n i o n s are simply t o o n u m e r o u s , t o o m u l t i d i m e n s i o n a l , a n d , taken as a g r o u p , t o o o b s c u r e for it to be feasible for t h e m to e n g a g e in on-line processing of relevant information (see M c G r a w and Pinney, 1990). T h e strongest e v i d e n c e that the on-line m o d e l is inappropriate in the d o m a i n of political attitudes is that p e o p l e ' s attitude r e p o r t s exhibit so m u c h purely chance variation over t i m e and are systematically affected by t h i n g s , such as q u e s t i o n order, that o u g h t , by the logic of the on-line m o d e l , to be irrelevant. E v e n attitude r e p o r t s c o n c e r n i n g w e l l - k n o w n politicians, as an e x p e r i m e n t by W i l s o n , Kraft, and D u n n (1989) h a s s h o w n , a p p e a r to d e p e n d on the ideas at t h e top of the head at t h e m o m e n t of r e s p o n s e rather than any d e e p e r true attitude. It t h u s a p p e a r s that t h e on-line m o d e l suffers from t h e s a m e w e a k n e s s e s as the c o n v e n t i o n a l " t r u e a t t i t u d e " m o d e l s discussed earlier. In fact, the on-line m o d e l is essentially j u s t a variant of the old notion that survey r e s p o n s e s represent revelations of preexisting states of o p i n i o n , a notion that, as I have argued at l e n g t h , fails to w o r k very well. H a v i n g now, I b e l i e v e , strongly defended t h e R e s p o n s e A x i o m from criticism m o t i v a t e d by the on-line m o d e l , I w o u l d like to b a c k off a bit. For a l t h o u g h the m e m o r y - b a s e d j u d g m e n t p r o c e s s e m b o d i e d in A4 is defensible, it should be reg a r d e d only as a r o u g h first a p p r o x i m a t i o n of w h a t must actually occur. It seems to me likely that p e o p l e do e n g a g e in w h a t might be called bounded on-line processing of political i n f o r m a t i o n . T h a t is, they m a y m a k e on-line e v a l u a t i o n s of p a r t i c u l a r subjects but then fail to integrate e a c h newly u p d a t e d evaluation into a fresh global e v a l u a t i o n . So if, for e x a m p l e , p e o p l e e n c o u n t e r information about t h i r d - g e n e r a t i o n welfare families, they m a k e an on-line u p d a t e of their j u d g m e n t of the value of social spending for the alleviation of poverty. If, s o m e t i m e later, they e n c o u n t e r information about the n e e d s of h o m e l e s s p e r s o n s , they u p d a t e their j u d g m e n t on t h e u n m e t welfare n e e d s of the n a t i o n . But w h a t they do not d o , unless called to do so by an unusual event such as a public opinion survey, is to m a k e an u p d a t e of their global attitude toward the welfare s y s t e m . T h u s , if asked on a survey to m a k e a global statement on the p r o p e r level of welfare s p e n d i n g , they will have no u p - t o - d a t e global evaluation to retrieve but will instead have to m a k e an on-the-spot j u d g m e n t from w h a t e v e r ( u p d a t e d ) cons i d e r a t i o n s , w h e t h e r involving h o m e l e s s p e r s o n s o r s o m e t h i n g e l s e , c o m e most readily to m i n d . By this a c c o u n t , A4 is not so m u c h w r o n g as underspecified: M e m o r y - b a s e d d e c i s i o n m a k i n g o c c u r s , but s o , at a n o t h e r level, d o e s on-line processing.

280

N A T U R E AND ORIGINS OF MASS O P I N I O N

T h i s limitation points to a more general w e a k n e s s of the entire R A S framew o r k , namely its failure to p r o v i d e any m e c h a n i s m for integration of information that h a s been acquired. By its a x i o m s , p e o p l e screen information at the point of first e n c o u n t e r i n g it, but o n c e internalized, each bit of information b e c o m e s just a n o t h e r c o n s i d e r a t i o n in a mental " b i n " full of such atomized c o g n i t i o n s . T h i s is obviously a d r a s t i c simplification. A l t h o u g h total d i s c o n n e c t e d n e s s may, as Luskin (1987) skillfully a r g u e s , occur within certain s e g m e n t s of t h e p u b l i c , it cannot be the w h o l e story. S o m e p e o p l e , and probably m o s t , surely do build up c o m p l e x l y differentiated c o g n i t i o n s that cannot be a d e q u a t e l y captured by my simple notion of a c o n s i d e r a t i o n . C o n t r a r y to the R A S m o d e l , these m e n tal structures u n d o u b t e d l y g r o w in size and subtlety as new information is enc o u n t e r e d and integrated, t h u s perhaps increasing their chronic accessibility and h e n c e relative " i m p o r t a n c e " as d e t e r m i n a n t s of attitude r e p o r t s . I n t e g r a t e d inf o r m a t i o n structures might also play a major role in both the perception and critical scrutiny of i n c o m i n g c o m m u n i c a t i o n s . A n o t h e r s h o r t c o m i n g in this s a m e vein is that the m o d e l m a k e s no provision for m u l t i p l e reception of the s a m e m e s s a g e . D o e s a m e s s a g e that has b e e n accepted t w o or m o r e t i m e s then b e c o m e t w o or m o r e c o n s i d e r a t i o n s ? Or, m o r e in the spirit of the R A S m o d e l , might re-reception of an idea increase its salience and h e n c e its likelihood of use? Or finally, d o e s re-reception m a k e it m o r e likely that an idea will be integrated within s o m e larger, m o r e c o m p l e x mental structure? Probably the answer to all of these q u e s t i o n s , each of which c h a l l e n g e s the basic structure of the R A S m o d e l , is yes. T h e reason that I have left so m u c h that I believe to be true out of the R A S m o d e l is, quite simply, that there has been no pressing need to include it. T h e m a c h i n e r y of the current m o d e l has been able to explain a large part of the varia n c e in the existing survey e v i d e n c e that s e e m s presently a m e n a b l e to s y s t e m a t i c e x p l a n a t i o n , and I have been loathe to m a k e the R A S model any m o r e e l a b o r a t e than n e c e s s a r y to do this, v In the longer r u n , however, greater e l a b o r a t i o n of the model will b e c o m e app r o p r i a t e . Perhaps a fifth a x i o m could specify that the effect of t h o u g h t about a subject - perhaps even t h o u g h t that is independent of any elite influence (see Tesser, 1978) - is to build up integrated c o g n i t i v e structures, p r e s u m a b l y schem a t a . As these structures grow, they might c o l o n i z e or neutralize o p p o s i n g , less developed c o n s i d e r a t i o n s , thereby reducing r e s p o n s e instability. T h e y m i g h t also influence the interpretation and processing of new c o m m u n i c a t i o n s , thereby preventing the future f o r m a t i o n of o p p o s i n g c o n s i d e r a t i o n s . It w o u l d be easy, however, to take the notion of integrated c o g n i t i v e structures t o o far. O n e s h o u l d , in particular, never take this idea so far as to o v e r l o o k the substantial extent to w h i c h most p e o p l e ' s political ideas are not integrated into coherent mental s t r u c t u r e s , the extent to which o p p o s i n g ideas and feelings regularly coexist c o m f o r t a b l y within the s a m e b r a i n , and the large e l e m e n t of c h a n c e in the process by which one rather than a n o t h e r idea c o m e s to the top of the head and exerts m o m e n t a r y control over a b e h a v i o r or attitude s t a t e m e n t .

The

model

and future

research

281

W h a t would be most d e s i r a b l e , t h e n , is not the R A S model I have p r o p o s e d , but a R e c e i v e - A c c e p t - / « / e ^ r a r e - S a m p l e M o d e l , in which the o u t c o m e of initial information processing affects r e c e p t i o n , a c c e p t a n c e , and integration of subseq u e n t information. I d o u b t , however, that theoretical e l a b o r a t i o n of the R A S model along these lines will b e c o m e fruitful to analysts of public opinion until we learn how to m e a s u r e more than the simple " l i k e s " and " d i s l i k e s " that have been featured in my analysis. T h i s , in t u r n , will require s o m e way of m e a s u r i n g , a m o n g o t h e r t h i n g s , the v o l u m e and c o m p l e x i t y of discrete c o g n i t i o n s , and w h e r e , in a stream of verbal statements from a survey r e s p o n d e n t , t h o u g h t s o r g a n i z e d by o n e c o g nitive structure leave off and those associated with a n o t h e r begin. T h e s e sorts of i m p r o v e m e n t s in m e a s u r e m e n t capacity in m a s s surveys are not, so far as I am a w a r e , i m m e d i a t e l y in p r o s p e c t . S o , a l t h o u g h I readily a c k n o w l e d g e p h e n o m e n a that the a x i o m s of t h e R A S f r a m e w o r k cannot presently a c c o m m o d a t e , and can m a k e fairly specific suggestions about how the f r a m e w o r k might be p a r s i m o n i o u s l y e l a b o r a t e d in o r d e r to better a c c o m m o d a t e t h e m , I do not see any reason for opinion researchers to m a k e these e l a b o r a t i o n s until the d a t a n e c e s s a r y for their testing b e c o m e available.

MODELS OF T H E R E C E P T I O N - A C C E P T A N C E PROCESS If t h e most general a c h i e v e m e n t of the R A S model is its integration of a w i d e r a n g e of empirical regularities within a c o m p a c t m o d e l , its most p a r t i c u l a r success centers on its treatment of attitude c h a n g e . T h e core idea here has been that reception of persuasive c o m m u n i c a t i o n s increases with attentiveness to politics, and that capacity for resistance to uncongenial c o m m u n i c a t i o n s also increases with attentiveness. T h i s dual c l a i m , which o r i g i n a t e d in w o r k by Philip C o n verse and William M c G u i r e in the 1960s, h a s both m o t i v a t e d the discovery of s o m e interesting patterns of attitude c h a n g e and provided a basis for organizing t h e m , as in the typology proposed in C h a p t e r 8. E v e n if, as will d o u b t l e s s occur, future research turns up cases of attitude c h a n g e that are not i m m e d i a t e l y explicable within this typology, researchers should be reluctant to discard either the r e c e p t i o n - a c c e p t a n c e d y n a m i c or the role of political attentiveness in m e d i a t i n g it. As I have a t t e m p t e d to show, individual differences in propensity to receive political c o m m u n i c a t i o n s are so great and so c o n s e q u e n t i a l u n d e r a w i d e r a n g e of conditions that it is virtually never safe to ignore t h e m . T h e effects of attentiveness on a c c e p t a n c e or n o n a c c e p t a n c e of persuasive m e s s a g e s , given r e c e p t i o n , are less universal - in particular, they s e e m to w a s h out in high-intensity p a r t i s a n e l e c t i o n s - but they a r e n o n e t h e l e s s extremely i m p o r t a n t in many types of s i t u a t i o n s , and therefore also d a n g e r o u s to ignore. But despite the general i m p o r t a n c e of the r e c e p t i o n - a c c e p t a n c e d y n a m i c , there are a n u m b e r of q u e s t i o n s that may be raised about my d e v e l o p m e n t of

t

282

N A T U R E AND ORIGINS OF MASS O P I N I O N

p a r t i c u l a r r e c e p t i o n - a c c e p t a n c e models. I g r o u p these questions u n d e r t w o rub r i c s : general theoretical issues and m e t h o d o l o g i c a l issues.

Theoretical

The model and future research Table 11.2. Two measures of opinion change on U.S. policy in Central America

issues

In my o p e n i n g discussion of attitude c h a n g e in C h a p t e r 7, I said that attitude c h a n g e cannot be u n d e r s t o o d within the R A S m o d e l as a conversion experie n c e - the r e p l a c e m e n t of o n e crystallized belief by another. It must instead be u n d e r s t o o d as a c h a n g e in the balance of positive and n e g a t i v e c o n s i d e r a t i o n s relating to a given issue. O w i n g to the lack of a p p r o p r i a t e d a t a , I have been u n a b l e to do full j u s t i c e to this theoretical p o s i t i o n . T h a t is, I have b e e n unable to show how c h a n g e s over t i m e in the b a l a n c e of c o n s i d e r a t i o n s in p e o p l e ' s h e a d s lead to g r a d u a t e d c h a n g e s in c l o s e d - e n d e d attitude r e p o r t s , such as a m o v e m e n t from " s t r o n g a g r e e m e n t " with a p a r t i c u l a r policy to " n o t so s t r o n g " a g r e e m e n t . Rather, all of my analyses have dealt with the d i c h o t o m o u s c a s e : Either people c h a n g e their initial position to the o p p o s i t e position or they r e m a i n unchanged. D i c h o t o m o u s scoring of attitude c h a n g e m e a s u r e s d o e s have technical advantages: T h e 0 - 1 scoring m e t r i c , which distinguishes those w h o s u p p o r t a given position from all o t h e r s , including those w h o say " D o n ' t k n o w , " is the only o n e - d i m e n s i o n a l metric hat avoids treating " D o n ' t k n o w " responses a s missing d a t a . Since people w h o respond " D o n ' t k n o w " are less politically a w a r e than c o u n t e r p a r t s w h o d o a n s w e r survey q u e s t i o n s , their loss a s missing d a t a w o u l d introduce significant distortion into e s t i m a t e s of t h e effect of a w a r e n e s s on opinion f o r m a t i o n and c h a n g e . For e x a m p l e , most of the attitude c h a n g e in the early stages of the V i e t n a m W a r consisted of the conversion of respondents without -opinions into s u p p o r t e r s of the war. N o t w i t h s t a n d i n g all t h i s , attitude scales that m e a s u r e strength or extremity of attitude reports are c o n v e y i n g real information about p e o p l e ' s opinions. C h a n g e s in the d e g r e e of strength or extremity of an attitude, even if they do not involve crossing t h e threshold to s u p p o r t a policy that w a s previously o p p o s e d , likewise convey i n f o r m a t i o n . T h e p r o b l e m is h o w to use it. T w 6 possibilities are s u g g e s t e d in Table 11.2. T h e issue is U . S . involvement in Central A m e r i c a , on w h i c h attitudes were m e a s u r e d on a seven-point scale at t w o t i m e points. In the top half of Table 1 1 . 2 , 1 simply t o o k the m e a n difference o v e r t i m e i n scores o n the seven-point scale, o m i t t i n g " D o n ' t k n o w " respondents. As can be seen, these patterns of c h a n g e are radically different from those found in the e a r l i e r analysis. However, there is an o b v i o u s difficulty with this m e t h o d of m e a s u r i n g attitude c h a n g e : M a n y highly aware h a w k s were at points 5 or 6 on the seven-point scale at the first interview, so that little p r o i n v o l v e m e n t c h a n g e w a s possible for t h e m ; s o m e were even at point 7, w h e r e no m o v e m e n t to the right w a s p o s s i b l e . In g e n e r a l , the greater a r e s p o n d e n t ' s propensity to c h a n g e in the direction of a particular m e s s a g e , the m o r e limited the r o o m for c h a n g e w a s likely to b e . M e a n -

283

Level of political awareness Low High Method one: Attitude change as difference of means Hawks 1986 mean 1987 mean

3.65 4.15

4.23 4.59

5.26 5.60

0.61

0.50

0.36

(18)

(40)

(44)

0.34 (15)

3.47 4.07

2.76 3.95

3.53 4.05

3.71 3.71

0.60 (30)

1.19 (37)

0.52 (40)

0.00 (14)

Doves 1986 mean 1987 mean

3.87 4.07

2.97 3.16

2.60 3.00

2.27 3.19

Mean difference

0.20

0.19

0.40

0.92

(15)

(3D

(50)

(26)

3.67 4.28

Mean difference N Centrists 1986 mean 1987 mean Mean difference v

a

Method two: Attitude change as proportionate

change

Hawks 0.18* (17)

0.03 (37)

0.09 (41)

0.18 (13)

-0.01

0.23 (37)

0.14 (37)

-0.19 (14)

0.03 (30)

0.05

0.16

(49)

(26)

Centrists (29)

Doves -0.09 (14)

a

Cell entry is mean score on 7-point scale. * Cell entry is proportionate change between surveys, as described in text. Source: 1986 and 1987 NES surveys.

w h i l e , highly aware d o v e s were c o n c e n t r a t e d at points 1 and 2, w h e r e there w a s m u c h r o o m for r i g h t w a r d m o v e m e n t , including m o v e m e n t that did not require t h e m to cross the threshold into actual s u p p o r t for greater involvement in Central America. As a way of taking into account these floor and ceiling factors, I calculated c h a n g e scores for each r e s p o n d e n t , as follows,

284

The

N A T U R E AND ORIGINS OF MASS O P I N I O N

model

and future

285

research

A n o t h e r i m p o r t a n t q u e s t i o n about t h e r e c e p t i o n - a c c e p t a n c e p r o c e s s i s w h y ex-

Time2 — T i m e i Change = — — 7 — Time]

actly critical resistance to persuasion often increases with levels of political

T h i s m e a s u r e , calculated from panel d a t a , scores each r e s p o n d e n t ' s c h a n g e as a p r o p o r t i o n of his o p p o r t u n i t y for c h a n g e . S i n c e division by z e r o is i m p o s s i b l e , this a p p r o a c h requires the e l i m i n a t i o n of r e s p o n d e n t s w h o were at point 7 at the first i n t e r v i e w , a plausible r e q u i r e m e n t in v i e w of t h e impossibility of further r i g h t w a r d m o v e m e n t by such persons. T h e results of this maneuver, as s h o w n in t h e lower half of Table 11.2, are s o m e w h a t closer to the results of my earlier a n a l y s i s , which also implicitly controls for floors a n d c e i l i n g s . However, except for taking into account ceiling effects, this set of results has no m o r e theoretical w a r r a n t than the first. For like the s i m p l e subtraction a p p r o a c h , it fails to link probability of switching sides on an issue with g r a d u a t e d c h a n g e in the m i x of c o n s i d e r a t i o n s in p e o p l e ' s h e a d s . 7

To see w h a t linkage m i g h t look like, c o n s i d e r an issue in which c o m m u n i c a t i o n s so strongly favor o n e side - say, t h e liberal side - that even c o n s e r v a t i v e s take the liberal side of a d i c h o t o m o u s i t e m . But such c o n s e r v a t i v e s u p p o r t w o u l d p r e s u m a b l y b e a m b i v a l e n t . H e n c e any intensification o f c o n s e r v a t i v e c o m m u n i c a t i o n s w o u l d tip the b a l a n c e of c o n s i d e r a t i o n s for many c o n s e r v a t i v e s , p r o d u c i n g heavy switching to the c o n s e r v a t i v e side of the d i c h o t o m y . L i b e r a l s , b e c a u s e initially less a m b i v a l e n t , would be less likely to cross the t h r e s h o l d . T h u s , even if both g r o u p s formed exactly t h e s a m e n e w c o n s i d e r a t i o n s - and in this sense exhibited equal c h a n g e - c o n s e r v a t i v e s w o u l d be m o r e apt to switch s i d e s , with the most a w a r e most likely to do so o w i n g to heavier e x p o s u r e to n e w ideas and ( m o s t likely) g r e a t e r initial a m b i v a l e n c e resulting from p r i o r r e c e p t i o n of c o u n t e r v a l e n t c o m m u n i c a t i o n . This sort of a r g u m e n t c o u l d , I believe, m a k e a useful s t a r t i n g point for future research in this a r e a . 8

a w a r e n e s s . I h a v e offered a t h r e e - p a r t e x p l a n a t i o n for this r e s i s t a n c e , c e n t e r i n g on the m e c h a n i s m s of inertial r e s i s t a n c e , c o u n t e r v a l e n t resistance, and p a r t i s a n resistance. Of t h e t h r e e , c o u n t e r v a l e n t resistance - t h e c l a i m that better informed p e o p l e are m o r e resistant to d o m i n a n t persuasive m e s s a g e s b e c a u s e they are m o r e likely to b e c o m e a w a r e of low-intensity c o u n t e r v a l e n t information flows - s e e m s to h a v e t h e strongest e m p i r i c a l support. It is h a r d to imagine how o n e m i g h t explain the p e c u l i a r p a t t e r n s of c h a n g e that arose d u r i n g the V i e t n a m W a r and in voting in c o n g r e s s i o n a l e l e c t i o n s w i t h o u t reference to the effects of c o u n t e r v a l e n t communications. P a r t i s a n r e s i s t a n c e , t h e c l a i m that p e o p l e reject i d e a s that they a r e a b l e t o r e c o g n i z e as being inconsistent with their v a l u e s , likewise s e e m s to be on a firm f o u n d a t i o n , as j u s t i n d i c a t e d . T h i s b r i n g s us to inertial r e s i s t a n c e , w h i c h is the idea that p e o p l e h a v i n g a large store of initial c o n s i d e r a t i o n s on an issue a r e m o r e resistant to s u b s e q u e n t c h a n g e . T h i s resistance m e c h a n i s m , a l t h o u g h d a t i n g t o C o n v e r s e ' s (1962) introduction of the r e c e p t i o n - a c c e p t a n c e d y n a m i c , is o p e n to q u e s t i o n . T h e probl e m s are both theoretical and e m p i r i c a l . On t h e theoretical s i d e , s u p p o s e that C and Z), are c o u n t e r v a l e n t and d o m i n a n t c o n s i d e r a t i o n s in a p e r s o n ' s m i n d at an initial point in t i m e , and that C and D are c o u n t e r v a l e n t and d o m i n a n t cons i d e r a t i o n s acquired o v e r a s u b s e q u e n t t i m e p e r i o d . T h e e x p e c t e d r e s p o n s e c h a n g e o v e r t i m e is t h e n given by: {

2

2

• , , , C h a n g e in r e s p o n s e probability = nu

(Ci + C )

h

2

C, 7 The pattern in Table 11.2 would be perfect, except for high-awareness doves, who show much larger than expected change. I was naturally curious about this problem and so investigated it. Virtually all of the difficulty is with two individuals (in a cell of 26) who shifted from 1 on the seven-point Central America scale in 1986 to 7 in 1987, a maximal change in the wrong direction. Immediately following this question in the 1987 survey was a question about whether aid to the Contra guerrillas should be increased, decreased, or kept the same. Both of the problematic doves responded by volunteering that aid to the Contras should be entirely cut off! Apparently, then, their stated support for "much greater U.S. involvement in Central America" is misleading. One possibility is that they wanted the United States to be more involved in development projects or humanitarian relief. More likely, however, is that they simply made an error in using the sevenpoint scale. This would have been easy since the 1987 survey was conducted by telephone. In this mode, interviewers describe a seven-point scale to respondents rather than, as in a regular NES survey, give respondents a show-card with a labeled seven-point scale. I monitored many of the 1987 telephone interviewers and heard one case in which I was certain that a respondent picked the wrong number to describe her self-described position on Central America and another in which I felt it was likely that an error had been made. The SRC interviewing staff is generally reluctant to use seven-point scales over the telephone, and this case may illustrate why. 8 Because it is difficult to anticipate issues on which attitudes will change in the real world, it might be necessary to pursue the relationship between considerations and susceptibility to change in a laboratory setting. A type of opinion that is particularly suitable for use in examining attitude change in a natural setting is presidential approval, which can be counted upon to exhibit relatively large amounts of variation over time. s

,

(11.1)

C, + D ,

T h e claim of inertial resistance is that the larger C , , the greater the p e r s o n ' s resistance to the effects of the d o m i n a n t m e s s a g e . If s o , the derivative of E q u a tion 7 . 1 w i t h respect t o C , o u g h t t o b e p o s i t i v e , w h i c h w o u l d indicate that h i g h e r initial n u m b e r s of c o u n t e r v a l e n t c o n s i d e r a t i o n s a r e , all else e q u a l , associated with higher levels of c o n t i n u e d s u p p o r t for the c o u n t e r v a l e n t position. However, this d e r i v a t i v e turns out not to be p o s i t i v e ; it is a c o m p l i c a t e d function that indicates no c l e a r r e l a t i o n s h i p b e t w e e n C

{

and resistance to a d o m i n a n t

9

message. Despite t h i s , the notion of inertial resistance r e m a i n s an intuitively plausible o n e . T h e r e w a s , moreover, s o m e c l e a r e m p i r i c a l s u p p o r t for it in the analysis of v o t i n g in H o u s e e l e c t i o n s . O n e m i g h t , in a d d i t i o n , c o u n t t h e steadfast loyalty of 9 Numerical analysis of the derivative with plausible values for the key terms indicates that C, is, as expected, often associated with resistance to a dominant message, but never very strongly. Derivatives of Equation 7.1 testing for the effects of C and D , which are associated with countervalent and partisan resistance, do have the expected signs. 2

2

N A T U R E AND ORIGINS OF MASS O P I N I O N

The model and future research

sophisticated older p a r t i s a n s to established ideas c o n c e r n i n g K o r e a and Vietn a m , as discussed in C h a p t e r 8, as e v i d e n c e for inertial resistance. T h e idea here w o u l d be that old p a r t i s a n s have longer e x p e r i e n c e with these ideas and h e n c e a greater store of inertial c o n s i d e r a t i o n s m a k i n g t h e m resistant to c h a n g e . But there are also s o m e e m p i r i c a l p r o b l e m s . A p a r t i c u l a r l y troubling o n e involved t h e p u b l i c ' s r e s p o n s e t o President N i x o n ' s initiatives toward the c o m m u n i s t g o v e r n m e n t of m a i n l a n d C h i n a , as discussed in C h a p t e r 6. T h i s w a s an issue that had exercised c o n s e r v a t i v e s for t w o d e c a d e s , and so I e x p e c t e d that sophisticated c o n s e r v a t i v e s , especially older o n e s , would exhibit s o m e inertial resistance to a d m i s s i o n of " R e d C h i n a , " as the c o u n t r y w a s then called, to the U n i t e d N a t i o n s . But I s e a r c h e d diligently for e v i d e n c e of such inertial resist a n c e - w h i c h should have taken the form of a n o n m o n o t o n i c r e l a t i o n s h i p b e tween awareness and s u p p o r t for recognition of C h i n a - and could find n o n e . T h e relationship to t h e n e w m a i n s t r e a m n o r m w a s , as reported in C h a p t e r 6, as monotonically positive a m o n g c o n s e r v a t i v e s a s a m o n g other types o f p e o p l e .

insufficient to implant a n e w idea in a n y o n e ' s h e a d , but a d e q u a t e to a w a k e n and

286

A n o t h e r d i s a p p o i n t m e n t , from the point of view of inertial resistance, w a s the b r e a t h t a k i n g l y swift m o v e m e n t of Republican activists, individuals w h o w o u l d u n d o u b t e d l y score at the top of m e a s u r e s of b o t h p a r t i s a n s h i p and political a w a r e n e s s , toward w a g e and price controls o n c e N i x o n a n n o u n c e d his s u p p o r t of t h e m in 1 9 7 1 , as also discussed in C h a p t e r 6. T h i s policy w a s as antithetical to traditional Republican c o n s e r v a t i s m as any that o n e can i m a g i n e , but it p r o v o k e d scarcely any resistance. T h e s e t w o c a s e s , in c o m b i n a t i o n with the m o d e r a t e m a g n i t u d e s of t h e effect of inertial resistance in congressional e l e c t i o n s , feed the impression that inertial resistance is less i m p o r t a n t than either p a r t i s a n or c o u n t e r v a l e n t resistance as a source of resistance to d o m i n a n t political m e s s a g e s . T h e notion of inertial resistance should certainly be kept alive, but it should also be treated with r e s e r v e until a reformulated t h e o r y or better d a t a can p r o v i d e it with a stronger w a r r a n t . O n e possibility is that .inertial r e s i s t a n c e , like the other t w o forms of resist a n c e , d e p e n d s on n o u r i s h m e n t from c u r r e n t c o m m u n i c a t i o n s . I suggest this bec a u s e in the two c a s e s in w h i c h there w a s s o m e t h i n g like inertial resistance to a n e w idea - the c o n t r a s t i n g cases of sophisticated o l d e r c o n s e r v a t i v e s resisting t h e K o r e a n W a r and of sophisticated older liberals resisting a n t i - V i e t n a m W a r ideas - there w a s a m p l e reinforcement for preexisting attitudes in elite disc o u r s e . In the c a s e in w h i c h , as it s e e m e d to m e , inertial resistance should have been present but apparently w a s not - N i x o n ' s reversal of U . S . policy toward c o m m u n i s t C h i n a - there w a s virtually no elite opposition to N i x o n ' s n e w policy and also no resistance a m o n g sophisticated c o n s e r v a t i v e s . H e n c e w h a t I am calling inertial resistance m a y w o r k s o m e t h i n g like this: P e o p l e with large stores of preexisting c o n s i d e r a t i o n s with respect to an issue m a y exhibit unusual resistance to new ideas but only so long as elite discourse gives s o m e indication of the c o n t i n u e d relevance of their old feelings. A b s e n t m i n i m a l elite v a l i d a t i o n , old feelings m a y be s w a m p e d by ideas of the m o m e n t , as in N i x o n ' s policy reversal on C h i n a . Elite v a l i d a t i o n , on this view, m i g h t be

287

invigorate preexisting ideas. It is useful to recall at this point that t w o of my three resistance m e c h a n i s m s , p a r t i s a n and c o u n t e r v a l e n t resistance, d e p e n d on n o u r i s h m e n t from elite disc o u r s e (cueing m e s s a g e s in the first c a s e , c o u n t e r v a l e n t c o m m u n i c a t i o n s in the s e c o n d ) . N o w it turns out that even inertial r e s i s t a n c e , already the least important of t h e three m e c h a n i s m s , m a y also d e p e n d on elite c o m m u n i c a t i o n . T h e implication of all this is painfully clear: T h e capacity of citizens for a u t o n o m o u s resistance to d o m i n a n t elite c o m m u n i c a t i o n s is very distinctly limited. I t u r n , finally, to a large but almost invisible simplification in the receptiona c c e p t a n c e m o d e l , n a m e l y the implicit a s s u m p t i o n that individuals never think for t h e m s e l v e s , but instead rely exclusively on the reception of c o m m u n i c a t i o n s reaching t h e m from t h e external e n v i r o n m e n t . Technically, this presupposition m a y be stated as follows: Prob(Attitude change) = Prob(Reception) x Prob(Acceptance I Reception) + P r o b ( A c c e p t a n c e I No reception) w h e r e P r o b ( A c c e p t a n c e I No reception) is c o n s t r a i n e d always to be z e r o . It is difficult to say h o w large a simplification is involved h e r e . But before discussing that q u e s t i o n , I want to put t h e m o d e l ' s constraint c o n c e r n i n g indep e n d e n t thought in p e r s p e c t i v e by noting w h a t t h e m o d e l does allow. It d o e s allow p e o p l e to v a r y in their political p r e d i s p o s i t i o n s , so that citizens m a y be m o r e or less strongly predisposed toward equality, hawkish military policies, tolerance of dissent, and so forth. W h e t h e r these predispositions are p r o d u c t s of socialization, adult e x p e r i e n c e , self-interest, inborn personality, or r e a s o n e d c h o i c e , they represent c r i t i c a l , internally m o t i v a t e d influences on opinion a n d , as s u c h , afford citizens an important d e g r e e of a u t o n o m y from elite inform a t i o n flows. To put the m a t t e r m o r e strongly: T h e m o d e l d o e s not c l a i m , as a reader of an early draft of this b o o k c o m p l a i n e d , that individuals are the passive receptacles of w h a t e v e r elites want t h e m to believe; the c l a i m , rather, is that citizens pick and c h o o s e , on the basis of their predispositions, from the m e n u of elitesponsored a l t e r n a t i v e s to which they have b e e n e x p o s e d . T h i s is still a strong c l a i m , but a less o u t r a g e o u s o n e . T h e a r g u m e n t is that most c i t i z e n s , to the extent they are attentive to politics, align t h e m s e l v e s with leaders or g r o u p s that share their predispositions and have m o r e information t h a n they d o . In this way, they avoid the necessity - and perhaps t h e practical impossibility - of really thinking about issues. (The n o r m a t i v e and research implications of this claim are c o n s i d e r e d in C h a p t e r 12.) Let me play out this logic as it applies to a c o n c r e t e c a s e that h a s been extensively e x a m i n e d in this b o o k , the V i e t n a m War. T h i s is an issue that is often said to have b e e n agonizingly difficult for m a n y p e r s o n s , especially e d u c a t e d liberals. T h e s e were p e o p l e w h o t e n d e d to be strong supporters of the w a r in its

288

289

N A T U R E AND ORIGINS OF MASS O P I N I O N

The model and future research

initial phase but found the w a r increasingly difficult to s u p p o r t and eventually t u r n e d , a m i d m u c h d i s c o m f o r t , o v e r w h e l m i n g l y against it. T h e m o d e l , of c o u r s e , explains both the initial s u p p o r t and s u b s e q u e n t o p p o sition of e d u c a t e d liberals as a response to the c u e s of liberal elites. If, as w a s the c a s e , it proved difficult for many e d u c a t e d liberals to turn against t h e w a r as liberal elite opinion leaders b e c a m e increasingly antiwar, this might simply be a c a s e of a m b i v a l e n c e b e t w e e n o p p o s i n g c o n s i d e r a t i o n s , w h e r e the a m b i v a l e n c e has been brought about by " i n e r t i a l r e s i s t a n c e " to the new a n t i w a r m e s s a g e . S u c h inertial resistance w a s , as the model w o u l d expect, m u c h stronger a m o n g o l d e r liberals, w h o , in contrast to y o u n g e r o n e s , had built up large stores of antic o m m u n i s m and o t h e r C o l d W a r - t y p e c o n s i d e r a t i o n s . A l t h o u g h few e d u c a t e d liberals may have consciously experienced t h e m s e l v e s as suffering t h e effects of inertial resistance to a new elite n o r m , it does not s e e m to me that t h e m o d e l d o e s an o b v i o u s injustice in claiming that this is w h a t q c c u r r e d .

However, the d y n a m i c s of a " p u b l i c opinion b a s e d on intergroup f e e l i n g s " need not fundamentally differ from the d y n a m i c s of a public opinion b a s e d on elite d i s c o u r s e , as p r o p o s e d in this b o o k . P e o p l e w h o pay little attention to elite d i s c o u r s e would still need to acquire information about the w o r l d and to evaluate it in light of s o m e u n d e r s t a n d i n g of w h a t the information m e a n s ; they w o u l d also still formulate attitude r e p o r t s on the basis of top-of-the-head information (or feelings), with all that follows from this. O n l y t h e sources of the leadership c u e s and information flows n e e d be different: T h e y m i g h t originate in informal or n e i g h b o r h o o d subcultures r a t h e r than in a m a i n s t r e a m elite, and might diffuse by w o r d of m o u t h rather t h a n via t h e m e d i a . S u c h c o m m u n i c a t i o n flows m a y be as likely to contradict as to c o r r o b o r a t e information from the elite c o m m u n i c a t i o n c h a n n e l s , as an e x a m p l e from the former Soviet U n i o n will shortly illustrate, but they n e e d not contradict any principles of the r e c e p t i o n - a c c e p t a n c e m o d e l . As I o b s e r v e d earlier, c o m m u n i c a t i o n s transmitted by face-to-face interaction in the m a n n e r of g e r m s m a y obey the principles of t h e m o d e l to the same extent as elite c o m m u n i c a t i o n s that diffuse through the m a s s m e d i a .

I am less confident that the a r g u m e n t of this b o o k can capture the processes by which the attitude r e p o r t s of less attentive p e o p l e are formed. T h e reason is that, being less politically a t t e n t i v e , they are less influenced by the flow of p o litical c o m m u n i c a t i o n s that is at the h e a r t of the b o o k ' s a r g u m e n t . H o w less attentive p e r s o n s respond to political issues in the absence of attention to or information about many of t h e m is a difficult q u e s t i o n . P a r t of the answer, as we have s e e n , is that their attitude r e p o r t s tend to be u n r e l i a b l e , in that they lack t e m p o r a l stability and are p o o r l y correlated with their political predispositions. A n o t h e r p a r t of the a n s w e r is that politically inattentive p e o p l e m a y often give " m i s t a k e n " attitude r e p o r t s , in t h e sense that they r e s p o n d on the basis of an incorrect g u e s s about the m e a n i n g of the q u e s t i o n s they have been a s k e d . I w a s able to report o n e clear c a s e of this in C h a p t e r 5, w h e r e it w a s s h o w n that inattentive persons r e s p o n d e d to a sparely w o r d e d q u e s t i o n about U . S . s p e n d i n g on aid to the C o n t r a rebels as if they were being asked about g o v e r n m e n t spending in g e n e r a l . T h e result w a s that social welfare liberals were most s u p p o r t i v e of aid to" the C o n t r a s and social welfare c o n s e r v a t i v e s m o s t o p p o s e d . This w r o n g - w a y c o r r e l a t i o n with ideology d i s a p p e a r e d , however, w h e n a m o r e e l a b o r a t e q u e s t i o n c o n v e y e d the information that the C o n t r a s w e r e o p p o nents of c o m m u n i s m . T h e s e o b s e r v a t i o n s a c c e n t u a t e the n e g a t i v e , in that they stress w h a t is missing from the attitude r e p o r t s of the politically inattentive. T h i s is not i n a p p r o p r i a t e , but there must be a positive s i d e , as well, to indicate what does occur. It s e e m s likely that the s a m e predispositions that a n i m a t e the attitude r e p o r t s of the politically a w a r e are present and influential a m o n g the less a w a r e . It is j u s t that they m a y o p e r a t e m o r e directly on opinion - or at least o p e r a t e indep e n d e n t l y of any m e d i a t i o n via the information and leadership c u e s c a r r i e d in elite political discourse. T h u s , for e x a m p l e , S n i d e r m a n , Brody, and Tetlock (1991) report t h a t , a m o n g p o o r l y e d u c a t e d p e r s o n s , feelings or affect toward v a r i o u s political and social g r o u p s are a potent influence on attitude r e p o r t s , w h e r e a s , a m o n g better e d u c a t e d p e r s o n s , indicators of ideology are t h e d o m i nant influence.

But might not the ideas that o r i g i n a t e a m o n g the public and diffuse by word of m o u t h be different from those that o r i g i n a t e a m o n g elites? I n d e e d , they m i g h t . Taxpayer revolts and b a c k l a s h e s of v a r i o u s t y p e s , including w h i t e backlashes against federal efforts to p r o m o t e racial equality, are probably o n e type of attitude that tends to o r i g i n a t e m o r e often a m o n g the public than a m o n g elites t h o u g h , even h e r e , there a r e cases in w h i c h politicians stir up b a c k l a s h e s and voter revolts b e c a u s e they expect to profit politically by t h e m . All this brings us b a c k to the q u e s t i o n discussed earlier: H o w often do elites a u t o n o m o u s l y lead opinion and how often do they simply go along with a public that o p e r a t e s on its o w n internal d y n a m i c ? A n d as in my e a r l i e r d i s c u s s i o n , I have no general a n s w e r to this q u e s t i o n . I can only o b s e r v e that in the p a r t i c u l a r cases e x a m i n e d in this b o o k , there is reason to believe that elites p r e d o m i n a t e l y led rather than followed.

Methodological

issues

T h e central effort of the latter half of this b o o k h a s been to show h o w public opinion is formed from an interaction b e t w e e n aggregate-level variation in the intensity of political c o m m u n i c a t i o n s and individual-level differences in political a w a r e n e s s and political values. Additional issues a r i s e in the statistical m o d eling of this interaction. I have taken two general a p p r o a c h e s , o n e that a s s u m e s the diffusion of a single m e s s a g e through the p u b l i c , and a n o t h e r that a s s u m e s t w o c o m p e t i n g m e s sages. T h e first a p p r o a c h , t h o u g h b a s e d on a simplification, yields surprisingly a c c u r a t e predictions about w h a t should be e x p e c t e d u n d e r a r a n g e of p a r t i c u l a r c o n d i t i o n s . But, as C h a p t e r s 9 and 10 have m a d e clear, there are i m p o r t a n t features of m a s s opinion that cannot be explained by a o n e - m e s s a g e m o d e l . In particular, the cross-cutting m o v e m e n t s of liberal opinion early in t h e V i e t n a m War, and t h e inverse association b e t w e e n c a m p a i g n intensity and aggregate-level

290

N A T U R E A N D ORIGINS OF MASS O P I N I O N

defection to the o p p o s i t i o n party in e l e c t i o n s , require the explicit m o d e l i n g of a t w o - m e s s a g e information flow. I suspect that s o m e readers will e x p e r i e n c e t h e progression of my a r g u m e n t from o n e - m e s s a g e to t w o - m e s s a g e m o d e l s with a degree of frustration. T w o m e s s a g e m o d e l s are statistically c u m b e r s o m e , i m p o s e d a t a r e q u i r e m e n t s that can rarely be m e t , and fail to yield c l e a r predictions about w h a t will h a p p e n in a given situation. A n d if the predictions of the o n e - m e s s a g e m o d e l - which are not only c l e a r but aesthetically attractive - have only limited applicability, these readers will w o n d e r w h a t , after all, a r e c e p t i o n - a c c e p t a n c e m o d e l is really w o r t h . I think, however, that any d i s p a r a g e m e n t of t w o - m e s s a g e variants of the r e c e p t i o n - a c c e p t a n c e m o d e l is u n w a r r a n t e d . For although t w o - m e s s a g e m o d e l s c a n n o t readily be used deductively, they can certainly be used, as in t h e V i e t n a m c a s e and the presidential p r i m a r i e s c a s e , as t h e basis for r i g o r o u s statistical m o d e l i n g . Moreover, t w o - m e s s a g e m o d e l s p e r f o r m t h e i m p o r t a n t service of explaining why the s i m p l e r o n e - m e s s a g e m o d e l breaks d o w n in the cases in w h i c h it d o e s break d o w n , and of d o i n g so in a way that vindicates the essential principles of the r e c e p t i o n - a c c e p t a n c e process. If criticism is d u e the r e c e p t i o n - a c c e p t a n c e m o d e l s I have d e v e l o p e d , it is that there is t o o little c o n n e c t i o n b e t w e e n the o n e - m e s s a g e and t w o - m e s s a g e versions of t h e m . It w o u l d be preferable to have a single statistical m o d e l that w o u l d , by m e a n s of explicit simplifying a s s u m p t i o n s , reduce to various s i m p l e r forms. But I h a v e been unable to formulate such a m a s t e r r e c e p t i o n - a c c e p t a n c e m o d e l . A n o t h e r c o n c e r n is statistical robustness. S o m e of my results, as has b e e n a p p a r e n t , are quite fragile. T h e fragility manifests itself in large s t a n d a r d e r r o r s for coefficients and the need to constrain p a r a m e t e r s in m o d e l s in o r d e r to obtain conventionally significant e s t i m a t e s . T h i s statistical imprecision h a s s o m e t i m e s involved the two m o s t theoretically i m p o r t a n t effects in the R A S m o d e l , the effects of a w a r e n e s s on r e c e p t i o n and on resistance to p e r s u a s i o n , given r e c e p t i o n . T w o r e m a r k s are in o r d e r h e r e . First, c o n t e m p o r a r y statistical theory, t h o u g h not m o s t public opinion r e s e a r c h , tends to d o w n p l a y the i m p o r t a n c e of statistical significance as e s t i m a t e d from a single test of a m o d e l on a single d a t a s e t . It instead e m p h a s i z e s the stability of results o b t a i n e d across related p r o b l e m s and d a t a s e t s , and e m p h a s i z e s , as w e l l , the substantive m a g n i t u d e of p a r t i c u l a r effects and the theoretical p a t t e r n into which they fit ( A c h e n , 1983). E v a l u a t e d in light of these criteria, the results reported in C h a p t e r s 7 through 10 of this b o o k are strong. Political a w a r e n e s s , in particular, h a s proven itself extremely important, p r o d u c i n g l a r g e , theoretically intelligible and occasionally nonintuitive effects. W h e n the m a g n i t u d e s of these a w a r e n e s s effects have v a r i e d , the variation h a s often been in conformity with theory (as in t h e lack of an awareness effect on resistance in certain situations). T h e c o l l e c t i v e weight of this e v i d e n c e , including the b a c k g r o u n d t h e o r y and research on w h i c h the R A S m o d e l is founded, m a k e s it s e e m bootless to w o r r y about t h e s t a n d a r d errors of p a r t i c u l a r coefficients.

The

model

and future

291

research

A n d yet, the information conveyed by the frequently large standard e r r o r s and loose statistical fits is real. E v e n with the constraints I have i m p o s e d on the R A S m o d e l , it has often been u n a b l e to obtain statistically precise fits. T h e r e appear to be three sources of the p r o b l e m . O n e is t h e highly interactive nature of my o n e - m e s s a g e and t w o - m e s s a g e m o d e l s . T h e o r y requires political a w a r e n e s s to be entered into these m o d e l s several different times. T h u s , in the o n e - m e s s a g e m o d e l in E q u a t i o n 7 . 7 , a w a r e n e s s must be entered in the reception functions of both the baseline and c h a n g e m o d e l s , and in the a c c e p t a n c e functions of both m o d e l s . In the t w o - m e s s a g e m o d e l proposed in C h a p t e r 9, awareness a p p e a r s in the r e c e p t i o n function for both the liberal and c o n s e r v a t i v e m e s s a g e s , in the a c c e p t a n c e function for both m e s s a g e s , and in the recall function. T h e predispositional variables are entered t w i c e . T h e net effect of all t h i s , even after s o m e constraints have been i m p o s e d , is to create a large a m o u n t of multicollinearity, and thereby a loose statistical fit to the data. A n o t h e r source of statistical imprecision is that m u c h of what the m o d e l is trying to capture involves the b e h a v i o r of small subgroups of the overall s a m p l e . In particular, the m o d e l ' s e s t i m a t e of the effect of a w a r e n e s s on resistance d e p e n d s heavily on the handful of persons w h o are both heavily exposed to a m e s sage and predisposed to resist it. S u c h p e o p l e m a y constitute about half of the upper 10 percent of the s a m p l e on political a w a r e n e s s , which is to say, about five percent of the s a m p l e - a s a m p l e that, in several of my cases of attitude c h a n g e , w a s quite small to begin w i t h . Multicollinearity in the p r e s e n c e of small c a s e n u m b e r s in critical s u b g r o u p s can k n o c k the statistical precision out of any m o d e l , no m a t t e r how well crafted it might b e . In c o n s e q u e n c e of all t h i s , the difference b e t w e e n a fit in w h i c h political awareness affects resistance to persuasion and o n e in which it has no such effect m a y be small. This is a p p a r e n t in Figure 1 1 . 1 , which shows w h a t small differe n c e is m a d e in e s t i m a t e d patterns of opinion c h a n g e on U . S . involvement in Central A m e r i c a if a w a r e n e s s is deleted from the a c c e p t a n c e f u n c t i o n . (Note that I have altered the scale of the y-axis in Figure 11.1 from the usual 0 - 1 r a n g e in o r d e r to m a k e it easier to see what is h a p p e n i n g . ) Little wonder, in light of this figure, that the effect of a w a r e n e s s on resistance is difficult to reliably detect. 10

E v e n the effect of a w a r e n e s s on r e c e p t i o n , which appears by my e s t i m a t e s to be a highly reliable effect, m a y be difficult to detect in p a r t i c u l a r c a s e s . L o o k b a c k , for e x a m p l e , at the patterns of voter defections to the R e p u b l i c a n presidential c a n d i d a t e , as s h o w n in Figure 10.6. G r e a t e r a w a r e n e s s is associated with greater resistance to the presidential c a m p a i g n , with no hint that political awareness m a y have any positive effect on r e c e p t i o n . T h e r e is similarly little e v i d e n c e , either visual or statistical, of strong reception effects from political a w a r e n e s s apparent in the d a t a on vote defection in S e n a t e e l e c t i o n s , as also s h o w n in Figure 10.6. Yet t h e m a s s of d a t a presented in the s e c o n d half of this b o o k , as well 10 In estimating the model on which Figure 11.1 is based, I fixed the coefficients in the baseline function to be the same as in Table 7.4, thereby allowing only the coefficients in the change function to freely vary.

292

NATURE

AND

ORIGINS

OF

MASS

The

OPINION

.50-,

Political awareness

Figure 11.1. When awareness-induced resistance is set to zero. Solid black lines show estimated probabilities of opinion change from standard model; gray lines show estimated change when political awareness is omitted from the acceptance function. Source: 1986 and 1987 NES surveys. as p a r t i c u l a r d a t a involving these e l e c t i o n s (as in Figure 10.7), m a k e it c l e a r that a w a r e n e s s is associated with large individual-level differences in r e c e p t i o n in these cases as in t h e o t h e r s we have e x a m i n e d . In t h e face of these difficulties - multicollinearity, the d e p e n d e n c e of results on relatively small s u b g r o u p s of the overall s a m p l e , and effects that m a y be either subtle or invisible in p a r t i c u l a r contexts - w h a t can be d o n e ? It is i m p o r t a n t , first of all, to be clear that n o n e of these difficulties is a difficulty of the R A S m o d e l itself. It is not, that is, a difficulty of the m o d e l that political a w a r e n e s s h a s b o t h reception and a c c e p t a n c e effects, and that these effects may, u n d e r certain c o n d i t i o n s , interact in w a y s that p r o d u c e gentle nonm o n o t o n i c curves that are h a r d to distinguish from negatively sloped straight lines or almost flat lines. T h e difficulty, rather, is in finding w a y s to test the m o d e l that give its e x p e c t e d effects a fair c h a n c e to show up in typically noisy survey data. 1

T h e p r e c a u t i o n s n e c e s s a r y to conduct reliable tests of statistical m o d e l s are generally w e l l - k n o w n , but let me add a few r e m a r k s relating to p e c u l i a r i t i e s of the R A S m o d e l . Reliable measures. I n a s m u c h as results often d e p e n d on the b e h a v i o r of relatively small s u b g r o u p s at the extremes of high and low a w a r e n e s s , it is essential to have m e a s u r e s of political awareness that can reliably distinguish those subgroups. As explained in C h a p t e r 2 and in the M e a s u r e s A p p e n d i x , simple tests of political k n o w l e d g e s e e m to discharge this m e a s u r e m e n t function better than alt e r n a t i v e m e a s u r e s . In e v e r y c a s e in w h i c h I had an o p p o r t u n i t y to m a k e c o m p a r i s o n s , other types of m e a s u r e s did w o r s e in the sense that they showed w e a k e r effects than m e a s u r e s of political k n o w l e d g e were able to detect. T h u s ,

model

and future

research

293

if I had used variables o t h e r than political k n o w l e d g e in my tests of the r e c e p t i o n - a c c e p t a n c e m o d e l , I would have obtained significantly w e a k e r results than I did - and in s o m e c a s e s , nothing of interest at all. It is an interesting oddity that different disciplines prefer different types of items to m e a s u r e w h a t I have been calling political awareness. Political scientists are partial to interest in politics, especially self-report m e a s u r e s of interest in election c a m p a i g n s . " C o m m u n i c a t i o n specialists prefer self-reports of m e d i a exposure (though see P r i c e and Zaller, 1990). Sociologists like to use e d u c a t i o n , which they see as a m e a s u r e of social stratification. O n l y p s y c h o l o g i s t s , for reasons no less r o o t e d in disciplinary prejudice, m a k e w i d e s p r e a d use of political k n o w l e d g e , w h i c h they take as a m e a s u r e of " e x p e r t i s e , " on which they have a large literature. It is, as a l w a y s , also i m p o r t a n t to use statistically reliable m e a s u r e s . T h r e e - or four-item k n o w l e d g e s c a l e s , in which all of the items have roughly m i d d l i n g levels of difficulty, will n o r m a l l y fail to provide g o o d d i s c r i m i n a t i o n in the critical extremes of high and low political awareness. If a researcher can use only three or four i t e m s , they should at least be k n o w l e d g e tests that v a r y in difficulty from easy to h a r d . 1 2

T h e R A S m o d e l is no less sensitive to reliable m e a s u r e m e n t of political predispositions. For e x a m p l e , my findings on attitude c h a n g e on U . S . involvement in Central A m e r i c a are entirely d e p e n d e n t on the availability of an eight-item militarism scale, w h i c h , by pure luck, h a p p e n e d to be u n d e r g o i n g testing in a study that spanned the I r a n - C o n t r a controversy. Survey size. It s e e m s likely that the t w o - w a v e survey of 3 6 0 p e r s o n s , as used in the e s t i m a t i o n of attitude c h a n g e on Central A m e r i c a , is near the m i n i m u m n u m b e r of cases n e c e s s a r y to adequately test the r e c e p t i o n - a c c e p t a n c e m o d e l . I say this b e c a u s e the a m o u n t and p a t t e r n of attitude c h a n g e , as apparent in the raw d a t a , c o n f o r m well to theoretical e x p e c t a t i o n s , and the m e a s u r e s n e c e s s a r y to capture this c h a n g e w e r e about as g o o d as they could reasonably have b e e n , including, as they d i d , b o t h a 22-item a w a r e n e s s scale with an alpha reliability of .87 and the P e f f l e y - H u r w i t z militarism i t e m s to which I just referred. D e s p i t e all t h i s , the R A S m o d e l p r o d u c e d coefficient e s t i m a t e s that, a l t h o u g h having g o o d m a g n i t u d e s , w e r e barely able to achieve c o n v e n t i o n a l levels of statistical significance, even after constraining the m o d e l . It will, of c o u r s e , be quite difficult to achieve large s a m p l e sizes in surveys designed to capture attitude c h a n g e , since it is rarely possible to anticipate in a d v a n c e w h e n attitude c h a n g e will occur and w h a t q u e s t i o n s will be n e c e s s a r y to c a p t u r e it. U See pp. 235-6 in Chapter 10 for a discussion of the particular pitfalls of this measure. 12 Knowledge items that pose easy or hard tests tend, because of their limited variance, to produce misleadingly low item-to-total correlations (see Zaller, 1985, for examples). Thus, researchers wanting to build powerful measures of awareness need either to ignore the item-to-total correlations of their easy and hard items, or use suitable nonlinear and nonstandardized measures of statistical association (for example, item-to-total logit or probit coefficients).

294

N A T U R E AND ORIGINS OF MASS O P I N I O N

Variance in information flow. L a r g e datasets and reliable m e a s u r e s , despite their i m p o r t a n c e , m a y bring only m i n o r gains in testing of the R A S m o d e l . W h a t is most essential is g o o d v a r i a n c e on the principal independent variable in the m o d e l , namely, the aggregate-level flow of c h a n g e - i n d u c i n g persuasive c o m m u nications. To get this v a r i a n c e , it is n e c e s s a r y to m o v e away from s i m p l e cases like those e x a m i n e d in C h a p t e r s 7 and 8, w h e r e t h e w h o l e public h a s b e e n exp o s e d to t h e s a m e flow of persuasive c o m m u n i c a t i o n s between a single pair of survey o b s e r v a t i o n s , and to try instead to locate cases in which there is usable aggregate-level v a r i a t i o n in information flow. C h a p t e r 9, which analyzed attit u d e c h a n g e on the V i e t n a m W a r over a six-year p e r i o d , and C h a p t e r 10, which e x a m i n e d the effect of 77 different H o u s e c a m p a i g n s on voter a t t i t u d e s , showed w h a t can be l e a r n e d from such c a s e s . T h e fine analyses of E r b r i n g , G o l d e n b e r g , and Miller (1980), M a c K u e n (1984), and B a r t e l s (1988) p r o v i d e further e x a m ples of the testing of sophisticated individual-level m o d e l s of attitude f o r m a t i o n and c h a n g e by capitalizing on v a r i a n c e in the flow of political i n f o r m a t i o n . Modeling technology. N o t h i n g in the four a x i o m s of the R A S m o d e l requires that it be o p e r a t i o n a l i z e d and tested by m e a n s of t h e c o m p l i c a t e d nonlinear m o d els I have used in this b o o k . But given the inevitable limitations of the d a t a available for testing t h e m o d e l , it is essential to squeeze as m u c h m e a n i n g as possible from t h e available e v i d e n c e . At least in the c a s e of the R A S m o d e l , this will m e a n frequent resort to c o m p l i c a t e d n o n l i n e a r m o d e l s . My analysis of attitude c h a n g e would have been m u c h less successful if it had relied on s t a n d a r d linear regression m o d e l s , including interactive variants of these m o d e l s , instead of taking t h e path it h a s . I certainly do not c l a i m that the o p e r a t i o n a l forms into which I have cast the R A S m o d e l are ideal. But they do have t h e virtue of taking central a c c o u n t of the c o m p l e x interaction that exists a m o n g political a w a r e n e s s , political predisposit i o n s , and t h e flow of political i n f o r m a t i o n , and s o m e type of nonlinear m o d e l that can efficiently do this is essential for testing the R A S m o d e l . M o d e l i n g t e c h n o l o g y is t h e o n e factor in testing the R A S m o d e l that is fully u n d e r the control of opinion researchers. T h e y should not fail to m a k e t h e most of it. A final m e t h o d o l o g i c a l issue is my m e a s u r e m e n t of aggregate-level information flow. A l t h o u g h t h e flow of elite c o m m u n i c a t i o n s in the m e d i a is a p r i m e indep e n d e n t variable in my a n a l y s i s , I have m a d e no attempt to m e a s u r e it precisely. T h i s is true even in my analysis of attitudes t o w a r d V i e t n a m , w h e r e I c o m m i s sioned counts of stories for and against the w a r in several n e w s m a g a z i n e s . T h e difficulty is t h a t , as explained a few p a g e s a g o , m e s s a g e intensity involves m o r e than just the n u m b e r and salience of stories that are c a r r i e d in the m e d i a . It also involves characteristics of a m a s s a u d i e n c e that d o e s or d o e s not find a story interesting. Ia the long r u n , it will be desirable to distinguish sharply the s e p a r a t e c o n t r i b u t i o n s of m e d i a attention and m a s s receptivity to the

The model and future research

295

p e n e t r a t i n g power of a given m e s s a g e , which is h o w I defined m e s s a g e intensity. P r o g r e s s on this p r o b l e m will involve m o r e than better s t o r y - c o u n t i n g techniques. It will likely involve t h e union of a m o r e refined t h e o r y of the " m e s s a g e r e c e i v e r ' ' with a closely allied t h e o r y of m e d i a c o n t e n t analysis. In t h e m e a n t i m e , r e s e a r c h e r s c o m p a r i n g p a r t i c u l a r cases of m a s s attitude c h a n g e will have to d e p e n d on their judgments of the intensities of the c o m m u nication flows that have b r o u g h t about c h a n g e , even w h e n they have m a d e precise c o u n t s of m e d i a r e p o r t a g e . A n d to this extent, their tests of the m o d e l will be s o m e w h a t ad h o c . H a v i n g now d o n e my best to evaluate my f o r m u l a t i o n and testing of the R A S m o d e l , I wish to c o n s i d e r possible extensions.

PARSIMONY AS A V A L U E T h e centrifugal urge in a c a d e m i c r e s e a r c h , by w h i c h I m e a n the t e n d e n c y to d e velop n e w c o n c e p t s and n e w theoretical m a c h i n e r y for each n e w substantive p r o b l e m , is very strong in t h e political b e h a v i o r field. It a p p e a r s , moreover, that the m o s t gifted r e s e a r c h e r s e x p e r i e n c e t h e urge most strongly. T h e result, as I l a m e n t e d in the o p e n i n g of this b o o k , is that opinion research has b e c o m e deeply and needlessly f r a g m e n t e d , almost c o m p l e t e l y lacking in theoretical c o h e r e n c e as each n e w survey q u e s t i o n e m e r g e s as the object of its o w n specialized research literature. T h i s b o o k has been w r i t t e n in self-conscious opposition to this centrifugal u r g e . T h e a i m h a s been to identify a c o m p a c t set of c o r e ideas and to tie as m a n y p h e n o m e n a as possible to this c o r e . A l t h o u g h this effort might be unflatteringly c h a r a c t e r i z e d as i m p e r i a l i s t i c , the a n i m a t i n g spirit has been that of p a r s i m o n i o u s u n d e r s t a n d i n g . For, all else e q u a l , the m o r e c o n c e p t s and d i s t i n c t i o n s used to account for a given set of p h e n o m e n a , t h e less well those p h e n o m e n a are understood. In the spirit of p a r s i m o n y rather than i m p e r i a l i s m , t h e n , I w o u l d like to suggest s o m e additional p h e n o m e n a that currently exist as s e p a r a t e research topics but t h a t , I b e l i e v e , can be fruitfully analyzed within the f r a m e w o r k of the R A S m o d e l . T h e issues are evaluation of presidential character, trust in g o v e r n m e n t , and p o p u l a r s u p p o r t for a u t h o r i t a r i a n r e g i m e s .

Evaluation

of presidential

character

In a classic of the voting literature, S t o k e s (1966) argued that fluctuations in the p u b l i c ' s a s s e s s m e n t s of the personal qualities of presidential c a n d i d a t e s are the m o s t i m p o r t a n t factor in explaining interelection s w i n g s in the v o t e . R e c e n t research has tried to specify m o r e precisely the nature of these c h a r a c t e r assessm e n t s . Two streams of analysis have e m e r g e d . O n e is c o n c e r n e d with e m o t i o n a l reactions to the c a n d i d a t e s - feelings of fear, h o p e , p r i d e , and so forth. T h e o t h e r focuses on the p u b l i c ' s e v a l u a t i o n s of presidential character, especially the

296

N A T U R E AND ORIGINS OF MASS O P I N I O N

extent to which t h e c a n d i d a t e s s e e m to possess the traits of personal integrity and competence. Both streams of research have tended to focus on the factor structure of p o p ular reactions to presidential c h a r a c t e r - that is, the d i m e n s i o n s of j u d g m e n t that u n d e r l i e p e o p l e ' s a s s e s s m e n t s ( s e e , for e x a m p l e , Kinder, 1986). It has not d e voted m u c h attention to how, and whether, p e o p l e form their a s s e s s m e n t s of presidential character. T h e implicit a s s u m p t i o n in this area of opinion r e s e a r c h , as in most of the public opinion field, has been that all citizens pay e n o u g h attention to politics to d e v e l o p structured reactions to the n a t i o n ' s leading political figures. T h e a p p r o a c h t a k e n h e r e , in contrast, is to c o n s i d e r " i n f o r m a t i o n " about a p r e s i d e n t ' s c h a r a c t e r a " c a m p a i g n m e s s a g e " like other persuasive m e s s a g e s , and to use the r e c e p t i o n - a c c e p t a n c e model to c a p t u r e the diffusion p a t t e r n s of this m e s s a g e . T h e idea is that people w h o pay m o r e attention to politics receive m o r e information relating to presidential c h a r a c t e r and that they r e s p o n d to this information on t h e basis of their general levels of political a w a r e n e s s , their i d e o logical v a l u e s , a n d their p a r t i s a n s h i p . Less attentive citizens e n c o u n t e r less evidence of presidential c h a r a c t e r and are less able to m a k e p a r t i s a n and ideological evaluations of the e v i d e n c e they do c o m e across. S u c h an a p p r o a c h s e e m s at least initially p l a u s i b l e . Presidential c a n d i d a t e s always try in a general w a y to project sterling personal qualities and to k e e p their w a r t s h i d d e n from view. Occasionally, their efforts m a y b e c o m e c o n s p i c u o u s , as in J i m m y C a r t e r ' s a t t e m p t to d e m o n s t r a t e t r u s t w o r t h i n e s s in the 1976 c a m p a i g n or R o n a l d R e a g a n ' s d e t e r m i n a t i o n to d e m o n s t r a t e levelheadedness in t h e 1980 contest. But h o w e v e r subtle or o b v i o u s these c a m p a i g n m e s s a g e s , citizens m a y still vary in both their attentiveness to and their disposition to accept t h e m . Let u s , t h e n , turn to analysis of e v a l u a t i o n s of Reagan in the 1984 election c a m p a i g n . R e a g a n w a s by this point as well k n o w n as most presidential candidates are likely to get,. W i t h a strong e c o n o m y and a record u n b l e m i s h e d by personal s c a n d a l , o n e w o u l d expect h i m to have at least m o d e r a t e l y high r a t i n g s on c o m p e t e n c e and, given his ability to project sincerity, perhaps even h i g h e r ratings on personal integrity. To m e a s u r e a s s e s s m e n t s of R e a g a n on t h e s e t w o dim e n s i o n s , which fall on the two major d i m e n s i o n s on w h i c h A m e r i c a n s are said to assess presidential c h a r a c t e r (Kinder, 1986), I rely on the following items from the N E S battery: Now I'd like to know your impressions of Ronald Reagan. I am going to read a list of words and phrases people use to describe political figures. After each one I would like you to tell me how much the word or phrase fits your impression of Ronald Reagan. Moral Knowledgeable A c c o r d i n g to p o p u l a r lore, o n e of R e a g a n ' s principal a c h i e v e m e n t s as president w a s to m a k e A m e r i c a feel g o o d about itself again. T h e idea that u n d e r Reagan the United S t a t e s w a s " s t a n d i n g t a l l " in w o r l d affairs w a s a c o n t i n u i n g

The

model

and future

research

297

refrain of his years in office. At the s a m e t i m e , m a n y liberals confessed to a p prehension; there w a s s o m e t h i n g about R e a g a n that s e e m e d t o scare t h e m . T h e N E S study of the 1984 e l e c t i o n included q u e s t i o n s m e a s u r i n g v o t e r s ' e m o t i o n a l reactions to R e a g a n on b o t h these d i m e n s i o n s : Now we would like to know something about the feelings you have toward the candidates for President. I am going to name a candidate, and I want you to tell me whether something about the person, or something he has done, has made you have certain feelings like anger or pride. Think about Ronald Reagan. Now, has Reagan - because of the kind of person he is, or because of something he has done - ever made your feel: Angry Proud To g a u g e public r e s p o n s e s to these items I used the basic r e c e p t i o n - a c c e p t a n c e model as e m b o d i e d in E q u a t i o n 7 . 5 . T h e resulting e s t i m a t e s , s h o w n in Table 1 1 . 3 , are depicted graphically in Figure 1 1 . 2 . ( T h e r a w d a t a closely r e s e m b l e the e s t i m a t e s in Figure 11.2 and so are not s h o w n . ) T h e s e results are n o t a b l e in several respects. T h e first is that A m e r i c a n s are highly p a r t i s a n in their a s s e s s m e n t s of presidential character. T h i s , of c o u r s e , is no longer n e w s at this point ( C a m p b e l l et a l . , 1960; C o n v e r s e and D e P e u x , 1962). W h a t is n e w s , at least to the scholarly literature of this subject, is how strongly individual differences in political a w a r e n e s s interact with p a r t i s a n s h i p and ideology in the a s s e s s m e n t of presidential character. A m o n g less a w a r e citizens, p a r t i s a n s h i p has less effect; a m o n g the highly informed, t h e effect of partisanship is m o r e p r o n o u n c e d . T h e third salient feature of these results is how differently t h e public reacted to R e a g a n ' s knowledgeability than to his other c h a r a c t e r i s t i c s . By these d a t a , R e a g a n w a s less successful in projecting c o m p e t e n c e t h a n in proj e c t i n g morality, instilling p r i d e , and, in certain g r o u p s , p r o v o k i n g fear. Highly aware Republican c o n s e r v a t i v e s , the g r o u p that most strongly s u p p o r t s R e a g a n on other m a t t e r s , failed to rally behind h i m on this aspect of presidential perf o r m a n c e . Rather, it is m o d e r a t e l y a w a r e R e p u b l i c a n c o n s e r v a t i v e s w h o were most likely to assert that R e a g a n w a s extremely k n o w l e d g e a b l e about politics. T h u s we h a v e , in this c a s e but no o t h e r s , a strong n o n m o n o t o n i c relationship a m o n g R e p u b l i c a n s b e t w e e n political a w a r e n e s s and belief in R e a g a n ' s good qualities. 1 3

T h i s pattern i s , I b e l i e v e , best explained in t e r m s of the differential intensities of a d o m i n a n t and c o u n t e r v a i l i n g m e s s a g e . T h e d o m i n a n t m e s s a g e is the c a r e fully crafted i m a g e of presidential m e d i a advisors w h o , by skilled scripting of R e a g a n ' s public a p p e a r a n c e s , effectively project t h e idea that R e a g a n is the c o m m a n d i n g leader of t h e national g o v e r n m e n t . T h i s is the m e s s a g e that most p e o p l e e n c o u n t e r e d m o s t of the t i m e they received information about R e a g a n in 13 Although Figure 11.2 is based on Reagan evaluations in the postelection NES survey, highly similar patterns appear in data in the preelection survey. In particular, awareness has a resistance effect for knowledge but not for the other three ratings.

r

298

The

N A T U R E AND ORIGINS OF MASS O P I N I O N

Table

11.3.

Coefficients for presidential Traits Knowledgeable

Reception function Intercept Awareness

Moral

2.94 1.52 (.36)

0.90 0.96 (.35)

0.48 -1.25 (.24)

0.32 -0.17 (.17)

1.27 -0.21 (.13)

-0.72 0.20 (.21)

Party (range -2 to +2)

-0.56 (.11)

-0.52 (.08)

-0.88 (.11)

0.66 (.12)

Ideology self-id (range - 3 t o +3)

-0.37 (.10)

-0.20 (.07)

-0.34 (.07)

0.30 (.09)

Race (1 if black, else 0)

0.11 (.35) 1921

-1.09 (.28) 1923

-1.47 (.26) 1932

0.05 (.28) 1987

Acceptance Intercept Awareness

N

299

research

Percent who see Reagan as "extremely knowledgeable"

Conservative strong Republicans

Emotional reaction. Pride Fear

2.95 1.59 (.75)

and future

Percent who see Reagan as "extremely moral"

evaluation

1.18 0.74 (.28)

model

Conservative strong Republicans

function

Note: Model is equation 7.5, estimated by maximum likelihood. Trait items coded 1 for "extremely," 0 otherwise; emotional reaction items coded 1 ibi "yes," 0 otherwise. Other political variables coded liberal high. Dependent variables are from postelection survey. " N o o p i n i o n " respondents on ideology (V369 and V370) were coded to middle position. Standard errors appear in parentheses. Source: 1984 NES survey.

the m e d i a . But there w a s also a countervailing m e s s a g e which a p p e a r e d , t h o u g h m u c h less prominently, i n " o p - e d " and " n e w s a n a l y s i s " p i e c e s , and w h i c h d e picted R e a g a n as u n i n f o r m e d and uninterested in the details of g o v e r n a n c e . T h i s m e s s a g e , b e c a u s e less i n t e n s e , reached only t h e most politically a w a r e s e g m e n t s of the p u b l i c . But within this g r o u p of p e o p l e , it c o u n t e r a c t e d the effects of the d o m i n a n t p r o - R e a g a n m e s s a g e , thereby p r o d u c i n g the n o n m o n o t o n i c i t y that is a p p a r e n t in Figure 11.2. Note that political a w a r e n e s s has an i m p o r t a n t effect in the a c c e p t a n c e function of t h e knowledgeability m o d e l in Table 1 1 . 3 , but in n o n e of the o t h e r m o d els. T h i s is consistent with my a r g u m e n t , d e v e l o p e d in C h a p t e r 7, that one reason that awareness is often associated with resistance to a d o m i n a n t m e s s a g e is that it is a proxy for exposure to c o u n t e r v a l e n t c o m m u n i c a t i o n s that are insufficiently intense to reach the w h o l e public. Unfortunately, there is no easy way to m a k e direct m e a s u r e m e n t s of t h e intensity of the actual p r o - R e a g a n and a n t i - R e a g a n m e s s a g e s on each of t h e characteristics e x a m i n e d in Figure 11.2, and h e n c e no way to c h e c k the validity of this a r g u m e n t b e y o n d an appeal to the r e a d e r ' s intuition. My a r g u m e n t does

Democrats Political awareness Percent who say Reagan has made them feel "proud" IOOT

Percent who say Reagan has made them feel "afraid"

Conservative strong Republicans

Liberal strong Democrats.

Republicans

Political awareness Figure 11.2. Partisan and ideological reactions to candidate Reagan in 1984. Estimates are derived from coefficients in Table 11.3. Awareness runs from —1.97 SD to + 2 . 0 2 SD. Modal partisans have scores of ± 1 . 3 on the party measure and ± 1 on a 7-point measure of ideological self-location. Strong partisans have scores of ±2 on each of these measures. Source: 1984 NES survey. nonetheless provide a new and, I h o p e , interesting slant on the p u b l i c ' s j u d g ment of presidential character. T h i s a r g u m e n t , quite simply, is that the d y n a m i c s of public opinion on this m a t t e r are m u c h the s a m e as on o t h e r s : Political c o m m u n i c a t i o n s e m a n a t e from the elite centers of society, in this c a s e political c a m p a i g n s , m e d i a a d v i s o r s , a n d the p r e s i d e n t ' s record in office; the m a s s r e s p o n s e to these m e s s a g e s d e p e n d s on individual differences in value o r i e n t a t i o n s and attention to politics, and on the relative intensities of the v a r i o u s m e s s a g e s .

TRUST IN G O V E R N M E N T In t h e m i d - 1 9 6 0 s , as the United States w a s about to enter the w a r in V i e t n a m , p o p u l a r trust in g o v e r n m e n t appeared impressively high ( L a n e , 1965). But t h e n ,

300

N A T U R E AND ORIGINS OF MASS O P I N I O N

The

be g i n n i n g about 1965, c o n f i d e n c e b e g a n to e r o d e , a trend that soon affected attitudes not only toward g o v e r n m e n t but t o w a r d virtually the full r a n g e of A m e r ican institutions, including even the clergy (Miller, 1974; Citrin, 1974; U p s e t and Schneider, 1982). T h e reason for the d e c l i n e has never been entirely clear, but it has been widely a r g u e d that the t u m u l t u o u s events of the 1960s, especially the u n p o p u l a r war in V i e t n a m , the protest against it, and t h e black urban r i o t s , were at the root of the phenomenon. S h o u l d the w a v e of c y n i c i s m that swept over the c o u n t r y in this p e r i o d conform to the d y n a m i c s of the r e c e p t i o n - a c c e p t a n c e m o d e l ? If one c o n c e i v e s of t h e decline in trust as a sort of m o o d c h a n g e , a h u m o r reflecting the generally bad t i m e s on w h i c h the c o u n t r y had fallen, the a n s w e r w o u l d probably be n o . T h e r e c e p t i o n - a c c e p t a n c e m o d e l is d e s i g n e d to capture the diffusion of p a r t i c u l a r ideas and m e s s a g e s , not the spread of a m o r p h o u s m o o d s and h u m o r s . But if, on the o t h e r h a n d , o n e c o n c e i v e s of t h e decline in trust as the p u b l i c ' s r e s p o n s e to attacks on national integrity by alienated protesters and leftist intellectuals, then the a n s w e r w o u l d b e yes. T h e m e s s a g e o f " l i b e r a l a l i e n a t i o n " should follow the s a m e diffusion path as any o t h e r type of m e s s a g e . S i n c e I could never d e c i d e w h i c h of these c o n c e p t u a l i z a t i o n s w a s m o r e valid, and since I feel o b l i g a t e d to report both the successes of the m o d e l and its failu r e s , I w a s reluctant to subject the m o d e l to a test in this d o m a i n . Recently, h o w ever, Jack C i t r i n , w h o specializes in analysis of trust in g o v e r n m e n t , c o n c l u d e d that t h e r e c e p t i o n - a c c e p t a n c e m o d e l might plausibly be applied to the p r o b l e m , and asked for my a d v i c e in building political a w a r e n e s s scales. I sent h i m copies of these scales for all N E S surveys and waited for the results of his analysis. S o m e of his results are s h o w n in Figure 1 1 . 3 . As can be s e e n , t h e increase in g o v e r n m e n t a l distrust o v e r t h e p e r i o d of the V i e t n a m W a r appears to have followed the p a t t e r n a s s o c i a t e d with the diffusion of liberal m e s s a g e s . A m o n g libe r a l s , there is a r o u g h l y linear and positive r e l a t i o n s h i p between a w a r e n e s s and c h a n g e in the direction of a liberal m e s s a g e ; a m o n g c o n s e r v a t i v e s , this relationship is s h a r p l y n o n m o n o t o n i c . (Centrists, not s h o w n , follow an indistinct m i d d l e p a t t e r n . ) Citrin, in e x a m i n i n g patterns of c h a n g e in trust in g o v e r n m e n t over several o t h e r t i m e p e r i o d s , found little additional s u p p o r t for t h e receptiona c c e p t a n c e m o d e l . O n e reason for t h i s , no d o u b t , is that fairly large a m o u n t s of attitude c h a n g e are n e c e s s a r y before t h e patterns of c h a n g e begin to stand out against b a c k g r o u n d levels of m e a s u r e m e n t error, and most c h a n g e s in trust levels since 1972 have b e e n of m o d e s t m a g n i t u d e . Another, m o r e substantive reason m a y be that n o n e of the c h a n g e - i n d u c i n g information flows had a c l e a r ideological v a l e n c e . It h a s n o t , in other w o r d s , a l w a y s been clear w h e t h e r liberals or c o n s e r v a t i v e s or p e r h a p s even centrists should be most susceptible to t h e presumed change-inducing message. 1 4

14 Citrin's results are part of a book manuscript in progress: "Governing the Disenchanted," Department of Political Science, University of California, Berkeley.

model

and future

301

research

1.00Liberals 0.75-

Proportionate reduction in trust in government,

u

-

: > u

"

o 25-

Conservatives

1964-72

0.00 ^ -0.25-

Political awareness Figure 11.3. Decline in trust in government. Data are from 1964 and 1972 CPS surveys, as provided by Jack Citrin, University of California, Berkeley. Trust is measured by a two-item index, whether government can be counted to do what is right, and whether government is run by a few big interests. C i t r i n ' s investigation of t h e p e r f o r m a n c e of the r e c e p t i o n - a c c e p t a n c e m o d e l in t h e d o m a i n of g o v e r n m e n t trust u n d e r s c o r e s a key theoretical point: T h e model applies only to cases in which o n e can d e m o n s t r a t e or plausibly a s s u m e the existence of p a r t i c u l a r c h a n g e - i n d u c i n g m e s s a g e s having distinct ideological c o l o r a t i o n s . W h e n o n e cannot d o t h i s , the r e c e p t i o n - a c c e p t a n c e m o d e l d o e s not perform, and should not be expected to p e r f o r m , in the m a n n e r suggested by my analysis.

POPULAR SUPPORT FOR A U T H O R I T A R I A N REGIMES T h e R A S model o r i g i n a t e d in studies of A m e r i c a n public opinion and A m e r i c a n political psychology. T h e r e is, however, no reason to believe that it applies only in the United States. It is an entirely general m o d e l a n d , as such, o u g h t to apply in any situation in which public opinion forms and c h a n g e s in r e s p o n s e to the diffusion of political c o m m u n i c a t i o n s . In an effort to show t h i s , I c o l l a b o r a t e d with B a r b a r a G e d d e s , a specialist in c o m p a r a t i v e politics, in an application of the m o d e l to a less developed n a t i o n . 1 5

T h e p r o b l e m that G e d d e s and I e x a m i n e d w a s p o p u l a r support for authoritarian r e g i m e s . E v e r y a u t h o r i t a r i a n g o v e r n m e n t a t t e m p t s to control the flow of n e w s and information to t h e public - but with w h a t effect? T h e R A S m o d e l , with its e m p h a s i s on the diffusion of elite c o m m u n i c a t i o n , is ideally suited to a n s w e r i n g this q u e s t i o n . P r e s u m a b l y the most politically aware m e m b e r s of the public receive the largest a m o u n t s of g o v e r n m e n t p r o p a g a n d a , and are also most 15 The joint paper was proposed by Geddes, who, upon becoming familiar with the receptionacceptance model, asserted that if it was as general as I claimed, it ought to be capable of explaining patterns of popular support for, and resistance to, the authoritarian regime then in power in Brazil. This was the beginning of an extremely fruitful collaboration.

302

N A T U R E AND ORIGINS OF MASS O P I N I O N

c a p a b l e of resisting it. H e n c e we should expect that p e o p l e in the broad m i d d l e r a n g e s of political a w a r e n e s s are most likely to exhibit support for the policies of authoritarian regimes. W i t h o u t going into technical detail that is readily available e l s e w h e r e ( G e d d e s and Zaller, 1989), this is exactly what we found. U s i n g opinion d a t a c o l l e c t e d in Brazil at the height of its a u t h o r i t a r i a n period in the early 1970s, G e d d e s and I show that s u p p o r t for the g o v e r n m e n t ' s a u t h o r i t a r i a n policies w a s greatest a m o n g citizens exhibiting m o d e r a t e levels of political awareness. S u p p o r t for a u t h o r i t a r i a n policies rises steadily with increases in political awareness (which is m e a s u r e d , as in this b o o k , by simple tests of factual k n o w l e d g e about politics) until about the 9 0 t h percentile on awareness. At that point, the trend flattens and then reverses as s u p p o r t for g o v e r n m e n t policies begins to d e c l i n e . T h i s n o n m o n o t o n i c p a t t e r n w a s not obtained for all policy i t e m s , however. For policies in w h i c h the a u t h o r i t a r i a n g o v e r n m e n t simply c a r r i e d forward t h e mains t r e a m policies of t h e previous d e m o c r a t i c r e g i m e , there w a s no n o n m o n o t o n i c ity: T h e most a w a r e p e r s o n s w e r e most likely to support the g o v e r n m e n t ' s m a i n s t r e a m line. H e n c e in Brazil as in the United S t a t e s , m a i n s t r e a m c o m m u n i c a t i o n s e n c o u n t e r e d n o resistance. N o n m o n o t o n i c patterns were o b t a i n e d only in policy a r e a s , such as c e n s o r s h i p of the p r e s s , in w h i c h the military g o v e r n m e n t d e p a r t e d from policies of the previous r e g i m e . T h e Brazilian d a t a were not strong e n o u g h to show why exactly m o r e aware p e r s o n s w e r e less willing to accept a u t h o r i t a r i a n policies a n d , in particular, w h e t h e r their unwillingness w a s r o o t e d in inertial, c o u n t e r v a l e n t , or p a r t i s a n resistance. T h e Brazilian survey d i d , however, c o n t a i n an abbreviated version of the classic F - s c a l e , and this e n a b l e d us to extend t h e m o d e l in a n o t h e r d i r e c t i o n . T h e F - s c a l e , as readers m a y recall, w a s p r o p o s e d by the authors of The Authoritarian Personality ( A d o r n o et a l . , 1950) as a m e a s u r e of predispositions tow a r d fascist ideologies. T h e items in the scale involve such matters as o b e d i e n c e to p a r e n t s , respect for e l d e r s , and c o n t e m p t for w e a k n e s s . A l t h o u g h n o n e of the i t e m s contains manifestly political c o n t e n t , they are nonetheless intended to m e a s u r e deeply r o o t e d personality predispositions toward a u t h o r i t a r i a n political policies and ideologies. Unfortunately, the effectiveness of this scale has never been fully tested in a setting in which the m a s s e s of o r d i n a r y citizens w e r e being heavily e x p o s e d to a d o m i n a n t a u t h o r i t a r i a n m e s s a g e . T h e Brazilian dataset therefore afforded a v a l u a b l e o p p o r t u n i t y to see w h e t h e r the scale p e r f o r m s as e x p e c t e d . As can be seen in F i g u r e 11.4, both the F - s c a l e and political a w a r e n e s s are associated with s u p p o r t for a u t h o r i t a r i a n p o l i c i e s , but not in a straightforward w a y . Persons w h o pay little attention to politics do not reliably s u p p o r t the policies of the military g o v e r n m e n t even w h e n their personality d i s p o s i t i o n s , as 1 6

16 Both relationships are highly statistically significant in a fully specified model (p < .01) (see Geddes and Zaller, 1989: table 3).

The

model

and future

research

303

High on F-scale

.60

40 Net score on index of support for authoritarian policies . 20

Low on F-scale

Political awareness Figure 11.4. The F-scale and support for the authoritarian regime in Brazil. Dependent variable is an 8-item scale indicating support for authoritarian policies, such as press censorship and suppression of unions. The F-scale consists of five items, purged of the effects of education, sex, race, and residential location. General procedures used in constructing the figure are the same as reported in figure 3 of Geddes and Zaller (1989). Authoritarians and nonauthoritarians in the figure have been constructed from scores of approximately ± 1 . 2 5 SDs on the F-scale; other background variables have been set to their modal values. Source: 1972-73 survey by researchers at the Instituto Universitario de Pesquisas do Rio de Janeiro and the University of Michigan.

m e a s u r e d by the F - s c a l e , indicate they ought to do s o . This is p r e s u m a b l y b e c a u s e they pay t o o little attention to politics to k n o w w h a t policies the authoritarian g o v e r n m e n t stands for. (Note that persons scoring low on a w a r e n e s s in Brazil w e r e , by U . S . s t a n d a r d s , extremely uninformed about politics; m a n y were illiterates w h o did not even k n o w the n a m e of the c o u n t r y ' s president.) M o d e r ately a w a r e persons exhibit m o r e support for a u t h o r i t a r i a n policies, but not m u c h m o r e e v i d e n c e of ideological a c u m e n . Like the p a r t i s a n defectors in H o u s e e l e c t i o n s , they are simply responding uncritically to the d o m i n a n t political m e s s a g e in the e n v i r o n m e n t . O n l y a m o n g highly aware persons is there an important relationship b e t w e e n a u t h o r i t a r i a n predispositions and s u p p o r t for authoritarian policies. This is b e c a u s e , in Brazil as in the United S t a t e s , highly aware persons are the only o n e s c a p a b l e of r e s p o n d i n g in a consistently i d e o logical fashion to the political c o m m u n i c a t i o n s they encounter. T h e s e results represent an i m p o r t a n t qualification of the t h e o r y of the authoritarian personality. T h e y s u g g e s t , first of all, that exposure to a u t h o r i t a r i a n prop a g a n d a may be at least as i m p o r t a n t in a c c o u n t i n g for s u p p o r t for a u t h o r i t a r i a n r e g i m e s as are a u t h o r i t a r i a n personality o r i e n t a t i o n s . T h e y further suggest that personality predispositions are just that: predispositions. W h e t h e r , or to w h a t extent, a u t h o r i t a r i a n predispositions manifest t h e m s e l v e s in the form of s u p p o r t for a u t h o r i t a r i a n policies d e p e n d s , as a l w a y s , on a critical interaction with political awareness.

304

NATURE

AND

ORIGINS

OF

MASS

The

OPINION

Dispositions toward a u t h o r i t a r i a n i s m are not, of c o u r s e , the only factor influencing s u p p o r t for a u t h o r i t a r i a n policies. E d u c a t i o n is also i m p o r t a n t . Better e d u c a t e d Brazilians, p e r h a p s b e c a u s e greater literacy increased their reception of c o u n t e r v a l e n t c o m m u n i c a t i o n s that w e r e largely u n c e n s o r e d by t h e military, were significantly m o r e resistant to a u t h o r i t a r i a n policies, given r e c e p t i o n of t h e m , than were p o o r l y e d u c a t e d ones. T h e result w a s a pattern m u c h like that s h o w n in Figure 11.4, except with e d u c a t i o n substituted for scores on the F-scale. O n e surprise in o u r analysis involved chronological a g e . We e x p e c t e d older, politically a w a r e B r a z i l i a n s , b e c a u s e of their e x p e r i e n c e with the d e m o c r a t i c n o r m s of the previous r e g i m e , to be m o r e resistant than their y o u n g c o u n t e r p a r t s to the policies of the a u t h o r i t a r i a n g o v e r n m e n t . A g e , however, turned out to have no d i s c e r n i b l e effect, p e r h a p s b e c a u s e it w a s a s s o c i a t e d with t w o mutually canceling effects, c o m m i t m e n t to the d e m o c r a t i c n o r m s of the previous r e g i m e and a preference for the o r d e r and stability of a u t h o r i t a r i a n g o v e r n m e n t . O n e should not place t o o m u c h e m p h a s i s o n the p a r t i c u l a r patterns o f r e g i m e s u p p o r t found in Brazil. T h e r e c e p t i o n - a c c e p t a n c e model would lead o n e to expect significant v a r i a t i o n in these p a t t e r n s from o n e country to another. T h e s e e x p e c t a t i o n s are set out in Figure 1 1 . 5 , a typology c l a i m i n g that variations in r e g i m e s u p p o r t d e p e n d on how vigorously g o v e r n m e n t s seek to indoctrinate their citizens, and h o w m u c h access citizens have to c o u n t e r v a l e n t communication. In c o u n t r i e s in which the g o v e r n m e n t m a k e s energetic efforts to indoctrinate its citizens - t h e g o v e r n m e n t s of Nazi G e r m a n y and the now defunct Soviet U n i o n are e x a m p l e s - even the least politically informed m e m b e r s of society m a y (in c o m p a r i s o n w i t h t h e least informed Brazilians) exhibit m o d e r a t e l y high levels of s u p p o r t for r e g i m e n o r m s ( c o m p a r e the cases in the left-hand c o l u m n of Figure 11.5 to those in t h e right-hand c o l u m n ) . A s e c o n d source of variation in p o p u l a r s u p p o r t for a u t h o r i t a r i a n r e g i m e s is the access of citizens in different c o u n t r i e s to c o u n t e r v a l e n t sources of values. M a n y citizens in a u t h o r i t a r i a n Brazil retained access to b o o k s and other small circulation publications that c a r r i e d a l t e r n a t i v e values; s o m e also r e m e m b e r e d the d e m o c r a t i c n o r m s of the previous r e g i m e . In other a u t h o r i t a r i a n c o u n t r i e s , h o w e v e r - N o r t h K o r e a c o m e s to m i n d - scarcely any s e g m e n t of the p o p u l a t i o n h a s either personal m e m o r y of, or access t o , sources of a n t i r e g i m e values. In c o n s e q u e n c e , the d e c l i n e in supp o r t for r e g i m e policies that we found a m o n g the most aware Brazilians should show up m o r e weakly, or p e r h a p s not at all, a m o n g the most aware citizens of c o u n t r i e s such as N o r t h K o r e a ( c o m p a r e cases in the top row of Figure 11.5 to those in the b o t t o m r o w ) .

model

and future

305

research

REGIME EFFORT TO MOBILIZE OPINION

High effort Nazi Germany ?

Low effort Authoritarian

Brazil

Political awareness North Korea?

United States (mainstream issues only)

Citizens have no access to alternative political values

Political awareness

1 7

17 In developing this typology, we assume that authoritarian regimes espouse authoritarian policies. The Gorbachev government in the Soviet Union, however, was a partial exception. Its glasnost and pereslroika policies must be viewed, in the context of Soviet history, as nonauthoritarian. One would therefore expect patterns of support for these policies to differ from those forecast for other cases.

Figure 11.5. A typology of support for government policies.

It is difficult to think of m o d e r n a u t h o r i t a r i a n c o u n t r i e s that fall clearly into the lower right cell of Figure 11.5. However, o n e might reasonably c o n s i d e r the United States an e x a m p l e of a n o n m o b i l i z i n g r e g i m e w h o s e elites have achieved high levels of v o l u n t a r y a g r e e m e n t on certain n o r m s relating to c a p i t a l i s m , d e m o c r a c y , a n d , at t i m e s , foreign policy. T h e analysis of C h a p t e r 6 indicates that w h e n such elite c o n s e n s u s exists, A m e r i c a n public opinion indeed c o n f o r m s to the p a t t e r n in the lower right cell. (See Figure 6 . 5 . ) In p r o p o s i n g this typology, G e d d e s and I had in m i n d cases in which t h e gove r n m e n t m a i n t a i n s at least t h e a p p e a r a n c e of c o m p e t e n c e and effectiveness. Certainly this w a s true of the Brazilian r e g i m e at the t i m e of o u r survey. For cases in w h i c h g o v e r n m e n t s are m a r k e d l y less effective (or less lucky with the e c o n o m y ) , o n e w o u l d expect m o r e resistance t o g o v e r n m e n t policies. T h e m o d e l readily a c c o m m o d a t e s the effects of such h e i g h t e n e d resistance. R e t u r n i n g to Figures 8.2 and 8 . 3 , we see that as resistance i n c r e a s e s , p a t t e r n s of policy s u p port b e c o m e n o n m o n o t o n i c and perhaps n e g a t i v e . H e n c e for the c a s e of an ineffective g o v e r n m e n t , we w o u l d expect, ceteris p a r i b u s , a gently n o n m o n o t o n i c

306

The

N A T U R E AND ORIGINS OF MASS O P I N I O N

p a t t e r n of support a m o n g the m o d e r a t e to poorly e d u c a t e d , and a n e a r l y flat or n e g a t i v e relationship with awareness a m o n g the better e d u c a t e d . (It is straightforward to r e d r a w Figure 11.5 to illustrate such c a s e s . ) T h e r e c e p t i o n - a c c e p t a n c e m o d e l thus specifies the w a y s in which a variety of factors - a n a t i o n ' s p r i o r e x p e r i e n c e with a l t e r n a t i v e i d e o l o g i e s , the intensity of the r e g i m e ' s efforts at public m o b i l i z a t i o n , the access of citizens to a l t e r n a t i v e sources of i d e a s , and the r e g i m e ' s p e r f o r m a n c e - interact to affect overall patt e r n s of r e g i m e s u p p o r t .

model

and future

research

307

result, according to a R a d i o Free E u r o p e survey of Soviet citizens traveling in the w e s t , w a s that b e t w e e n 1984 and 1986

1 8

A s m a t t e r i n g of e v i d e n c e from w h a t used to be k n o w n as E a s t e r n bloc c o m munist c o u n t r i e s in the p r e - G o r b a c h e v e r a provides additional s u p p o r t for the R A S m o d e l . A 1984 study by Polish researchers at the University of W a r s a w found that s u p p o r t for m a r t i a l law and o p p o s i t i o n to the union Solidarity were greatest a m o n g p e o p l e at low-to-middle levels of e d u c a t i o n . In 1985, w h e n the g o v e r n m e n t had b e g u n to p r o m o t e a return to n o r m a l c y , many p o o r l y e d u c a t e d r e s p o n d e n t s had lapsed into no opinion, so that s u p p o r t for martial law w a s then greatest a m o n g p e o p l e a t m i d d l e levels o f e d u c a t i o n ( R y s z k a , 1987: p . 2 5 3 ) . T h e s e findings are readily intelligible if, as s e e m s r e a s o n a b l e , o n e is willing to a s s u m e that e d u c a t i o n is positively associated with reception of p r o r e g i m e c o m m u n i c a t i o n s and negatively associated with a disposition toward uncritical acc e p t a n c e of these c o m m u n i c a t i o n s . Similar, t h o u g h s o m e w h a t sketchier e v i d e n c e is available from the f o r m e r S o viet U n i o n . Philip R o e d e r ( 1 9 8 5 , 1989) reviews studies s h o w i n g that, a l t h o u g h o p p o s i t i o n to the Soviet r e g i m e was greatest within the best e d u c a t e d s e g m e n t s of society, e d u c a t i o n w a s positively a s s o c i a t e d with individual susceptibility to indoctrination p r o g r a m s in factory settings. In o t h e r w o r d s , s u p p o r t for the reg i m e s e e m e d to increase with e d u c a t i o n - except at the very highest levels of e d u c a t i o n , w h e r e s u p p o r t declined s o m e w h a t . On the basis of this and o t h e r evi d e n c e , R o e d e r suggests that " t h e relationship b e t w e e n dissent and social m o bilization is curvilinear, declining with early social mobilization u n d e r Party t u t e l a g e , but rising as-education and urbanization b e c o m e still m o r e a d v a n c e d . " T h e inverse of R o e d e r ' s dissent c u r v e is, of c o u r s e , a pattern in w h i c h support for t h e r e g i m e increases as e d u c a t i o n i n c r e a s e s , but then declines at t h e highest levels of e d u c a t i o n . 1 9

By far the most p r o v o c a t i v e e v i d e n c e from a former E a s t e r n block nation involves c h a n g i n g p a t t e r n s of s u p p o r t for the Afghan W a r a m o n g the Soviet p u b l i c . It seems that after an initial p e r i o d in the e a r l y 1980s in which the Soviet m e d i a gave relatively little publicity to the war, the g o v e r n m e n t b e c a m e conc e r n e d about signs of public d i s q u i e t u d e . C o n s e q u e n t l y , it m a d e a " d e c i s i o n to d e v e l o p ' i n f o r m a t i o n ' - p a r t i c u l a r l y on television - about the w a r . " T h e 2 0

18 It is interesting to note that when the Brazilian economy faltered, support for the regime collapsed first in just those quarters in which our analysis would lead one to expect it: among the most highly educated and politically involved segments of the population (Alves, 1984; Lamounier, 1980). 19 I thank Stanislaw Gebethner for bringing these data to my attention. 20 Foreign Opinions (June 1987), Centre d'Analyse et de Prevision, Ministere des Affairs Etrangers. Number 4, Paris. I thank Chris Achen for bringing these data to my attention.

approval increased among those who had so far proved hesitant, in particular young people, uneducated, [and] central Asian citizens. On the other hand, opposition has developed in key sectors such as Moscow, Leningrad, among graduates and party members. (Roeder, 1985) In a d d i t i o n , no-opinion r a t e s , which had been high in 1984, fell d u r i n g this p e r i o d of intensified p r o g o v e r n m e n t c o m m u n i c a t i o n s , and fell most of all in provincial areas. In S i b e r i a , t h e most peripheral region of the Soviet U n i o n , " D o n ' t k n o w " r a t e s declined from 5 9 % t o 3 1 % b e t w e e n 1984 and 1986, while s u p p o r t for the w a r rose from 2 4 % t o 5 1 % and o p p o s i t i o n rates r e m a i n e d u n c h a n g e d . But in M o s c o w and L e n i n g r a d , the least peripheral r e g i o n s , the biggest c h a n g e w a s an increase in o p p o s i t i o n from 2 0 % to 3 5 % . T h e s e Soviet results present a striking parallel to c h a n g i n g patterns of s u p p o r t for t h e V i e t n a m W a r in the United States b e t w e e n 1964 and 1966. It will be recalled t h a t , in this p e r i o d , s u p p o r t for the w a r rose in the least aware s t r a t u m in r e s p o n s e to an intensified p r o w a r c a m p a i g n by t h e g o v e r n m e n t , but that, at t h e s a m e t i m e , it fell a m o n g the mostly highly a w a r e d o v e s , w h o were the only o n e s likely b o t h to receive and to accept the a n t i w a r m e s s a g e . O n e w o n d e r s about t h e source of the a n t i - A f g h a n W a r m e s s a g e in t h e Soviet U n i o n . O u r only e v i d e n c e about it is that a n t i w a r Soviets were most likely to m e n t i o n " w o r d of m o u t h " as their m a i n source of information about the war. T h i s no doubt represents a g e n u i n e and i m p o r t a n t difference from the United States. W h e r e the m e d i a are controlled by the g o v e r n m e n t , c o u n t e r v a l e n t c o m m u n i c a t i o n s must diffuse through informal c h a n n e l s . But aside from t h i s , the effects of d o m i n a n t and c o u n t e r v a l e n t m e s s a g e s a r e , as far as one can tell from the available d a t a , surprisingly similar in these two cases of u n p o p u l a r foreign w a r s . T h e r e is, however, an i m p o r t a n t caveat to all this. M u c h e v i d e n c e suggests that p o p u l a r support that is built up by m e a n s of one-sided c o m m u n i c a t i o n s can collapse virtually overnight w h e n t w o - s i d e d c o m m u n i c a t i o n s are allowed. A d e m o n s t r a t i o n of this o c c u r r e d in a u t h o r i t a r i a n Brazil shortly after o u r d a t a were c o l l e c t e d . In a highly controlled " e l e c t i o n c a m p a i g n " in which only g o v e r n m e n t approved c a n d i d a t e s w e r e allowed access t o t h e m a s s m e d i a , the g o v e r n ment c a n d i d a t e s were so far ahead of the opposition in the public opinion polls that the g o v e r n m e n t felt it safe to permit the o p p o s i t i o n c a n d i d a t e s limited access to the m e d i a a few w e e k s before t h e 1974 e l e c t i o n . W h e n it did s o , s u p p o r t for t h e official c a n d i d a t e s p l u m m e t e d . M a n y found t h e m s e l v e s in suddenly close r a c e s , and a n u m b e r were actually defeated ( L a m o u n i e r , 1980). T h a t m a s s opinion m a y prove highly c h a n g e a b l e is by no m e a n s e v i d e n c e of its insignificance. W e a k l y held or superficial o p i n i o n s that c o n d u c e toward a c q u i e s c e n c e in a u t h o r i t a r i a n i s m or are the basis for choices b e t w e e n radically different c a n d i d a t e s in e l e c t i o n s can be i m m e n s e l y c o n s e q u e n t i a l in spite of their superficiality. M a s s opinion m a y b e , in m a n y c a s e s , an inherently superficial

308

N A T U R E AND ORIGINS OF MASS O P I N I O N

force, but it is directly attached to the levers of g o v e r n m e n t , and this can m a k e it extremely powerful. It should also be noted t h a t , in s o m e a u t h o r i t a r i a n c o u n t r i e s , o p p o s i t i o n a c tivists d e v e l o p unofficial n e t w o r k s to challenge r e g i m e d o m i n a t i o n of political ideas. T h i s h a p p e n e d in t h e Philippines d u r i n g the M a r c o s r e g i m e and recently in S o u t h K o r e a . M o r e strikingly, as events in R o m a n i a showed in 1989, it is possible for a citizenry to stage a s p o n t a n e o u s rebellion even in the a b s e n c e of such c o m m u n i c a t i o n s n e t w o r k s . It is obviously important to understand how such events occur. T h e d y n a m i c s of the R A S m o d e l , however, offer no c l e a r insight to such p r o b l e m s . 2 1

2 2

CONCLUDING

REMARKS

T h e c o m m o n f r a m e w o r k developed in this b o o k involves the interactive effects of four types of variables: at the a g g r e g a t e level, variation in the intensity of political c o m m u n i c a t i o n s ; at the individual-level, variation in c i t i z e n s ' attention to politics and in their political predispositions (especially values); a n d , again at t h e a g g r e g a t e level, in t h e accidental factors that make considerations momentarily salient to p e o p l e , such as n e w s reports or q u e s t i o n n a i r e c o n s t r u c t i o n . T h e s e variables interact in a four-axiom m o d e l that specifies how individuals acquire political c o n s i d e r a t i o n s from the political e n v i r o n m e n t and use t h e m to formulate a n s w e r s to q u e s t i o n s about their political preferences. T h e general m o d e l s e e m s to w o r k a p p r o x i m a t e l y equally well in a w i d e range of political c o n t e x t s , from e l e c t i o n s to s u p p o r t for foreign wars to evaluation of presidential character. T h e four-axiom R A S m o d e l actually reduces to t w o main ideas. T h e first is that individuals do not possess " t r u e a t t i t u d e s , " in the usual technical sense of the t e r m , on most political issues, but a series of c o n s i d e r a t i o n s that a r e typically rather p o o r l y integrated. T h e c r u d e m o d e l I have proposed d o e s not begin to do j u s t i c e to the process by which c o m p e t i n g c o n s i d e r a t i o n s are f o r m e d , interact with one another, and stochastically d e t e r m i n e p e o p l e ' s responses to particular survey questions. But the core claim of the R A S model that people do not typically have " j u s t o n e a t t i t u d e " on issues is, as I have sought to show, essentially c o r r e c t and d e s e r v e s , in o n e form or another, a central place in o u r und e r s t a n d i n g of the nature of m a s s o p i n i o n . T h e o t h e r m a i n idea in the R A S model is that an interaction between political a w a r e n e s s and political predispositions is fundamental to the process by which citizens use information from the political e n v i r o n m e n t to form opinions. T h i s interaction w a s almost c o m p l e t e l y neglected in studies of political b e h a v i o r in the 1960s and 1970s but s e e m s to be getting m o r e attention from attitude re21 In the latter case, the government, although authoritarian, allowed the mass media a fair amount of latitude for criticizing its policies. 22 A report on National Public Radio, however, contended that the Romanian revolution began in peripheral areas of the country because that was where access to alternative communications, in the form of news reports from neighboring countries, was greatest.

The

model

and future

research

309

searchers - t h o u g h not yet from many voting b e h a v i o r specialists - in recent years (see E r b r i n g , G o l d e n b e r g , and Miller, 1980; M a c K u e n , 1984; Franklin and K o s a k i , 1989; K r o s n i c k and Kinder, 1990; Stoker, 1990; Jacoby, 1991; Pollock, L i s l e , and Vittes, 1991; S n i d e r m a n , Brody, and Tetlock, 1991; F r a n k l i n , 1992; and H u r w i t z and Peffley, in p r e s s ) . But even s o , s t a n d a r d p r o c e d u r e in the m a jority of political b e h a v i o r studies is still to a s s u m e that citizens react to the stimuli they e n c o u n t e r - w h e t h e r political c a m p a i g n s , presidential p e r f o r m a n c e , the state of the e c o n o m y , issues of war, p e a c e , or d o m e s t i c policy - solely on the basis of their values and interests. T h e idea that, o w i n g to differences in p o litical a w a r e n e s s , citizens m a y vary substantially in their ability to act on their values and interests is typically not r e c o g n i z e d . T h e m o d e l I have p r o p o s e d is certainly not w i t h o u t w e a k n e s s e s . In view of t h e m , especially the inability of the m o d e l in its present form to a c c o m m o d a t e either the integration of discrete information into larger mental s t r u c t u r e s , or the p r o c e s s by which political perception is influenced by past e x p e r i e n c e , the m o d e l must be c o n s i d e r e d a p p r o x i m a t e and p r o v i s i o n a l . It i s , I should also n o t e , very m u c h an o p e n q u e s t i o n w h e t h e r the p a r t i c u l a r o p e r a t i o n a l m o d e l s by w h i c h I have sought to a c c o m m o d a t e the interactive effects of political a w a r e n e s s and political p r e d i s p o s i t i o n s , especially as e m b o d i e d in E q u a t i o n s 7.7 and 9 . 8 , a r e ideal. I certainly do not c l a i m that they a r e . My c l a i m is only that r e s e a r c h e r s need to find s o m e way of taking these interactions into central account in their o w n m o d e l b u i l d i n g , b e c a u s e these interactions are both very c o m m o n and substantively i m p o r t a n t . D e s p i t e its v a r i o u s l i m i t a t i o n s , the general R A S m o d e l h a s t h e virtue of pulling t o g e t h e r within o n e theoretical s y s t e m a w i d e variety of empirical p h e n o m e n a , m a n y of w h i c h a r e analyzed by specialists w h o rarely or never c o m m u n i c a t e with o n e another. T h e specialized division of labor is a strength of scientific activity, but w h e n the vast majority of w o r k c o n c e n t r a t e s on issues of topical i m p o r t a n c e - even great topical i m p o r t a n c e , such as racial a t t i t u d e s , or s u p p o r t for U . S . intervention in foreign w a r s , or e l e c t i o n s - with little regard for anything besides t h e inherent i m p o r t a n c e of the topic, specialization has great costs. My feeling, as I have indicated, is that these costs have b e c o m e t o o great, and that it is t i m e to start i n c u r r i n g a few costs on the o t h e r side of the ledger. T h e R A S m o d e l , in a c c e p t i n g the costs of n o n s p e c i a l i z a t i o n , h a s been able to m a k e c l e a r t h e g r e a t , presently largely u n t a p p e d potential for synthesis within the public opinion field. T h e R A S model also s w i m s d e t e r m i n e d l y against the current o f m u c h political b e h a v i o r r e s e a r c h , w h i c h cheerfully e m p h a s i z e s the diversity of i n d i v i d u a l s ' r e s p o n s e s to politics. In so d o i n g , the m o d e l reflects my conviction that the a p p r o p r i a t e scientific r e s p o n s e to the diversity of nature is not to rejoice, but to r e d o u b l e t h e effort to find strong m o d e l s and broad generalizations that will fruitfully simplify t h e realities that we e x p e r i e n c e in daily life.

Elite

12

Epilogue: The question of elite domination of public opinion

The voice of the people is but an echo. The output of an echo chamber bears an inevitable and invariable relation to the input. As candidates and parties clamor for attention and vie for popular support, the people's verdict can be no more than a selective reflection from the alternatives and outlooks presented to them (p. 2). -V. O. Key, Jr., The Responsible Electorate In t h e 1930s and 1940s, m a n y o b s e r v e r s feared that the rise of the m o d e r n m a s s m e d i a w o u l d b r i n g a n e w e r a of totalitarian d o m i n a t i o n . M a s s circulation n e w s p a p e r s , t h e newly invented r a d i o , and m o t i o n pictures s e e m e d ideal tools for playing upon the fears of the n e w m a s s s o c i e t i e s , and the great t h o u g h t e m p o r a r y success of Hitler in G e r m a n y , Mussolini in Italy, and Stalin in the Soviet U n i o n s e e m e d to confirm e v e r y o n e ' s worst fears. G e o r g e O r w e l l ' s famous novel 1984 is p e r h a p s the best-known expression of this foreboding over the d a r k potential of the m a s s m e d i a , but many social scientists shared O r w e l l ' s a p p r e h e n s i o n . As a result, a t t e m p t s to m e a s u r e t h e effects of the m a s s m e d i a on public opinion were a staple of early opinion r e s e a r c h . T h i s early research t u r n e d out to be r e a s s u r i n g , however. C o m p a r e d to w h a t m a n y feared the m e d i a m i g h t be able to a c c o m p l i s h , surveys found m e d i a effects to be relatively small (Klapper, 1960). T h e m e d i a most often served to reinforce and activate existing opinion rather than to c r e a t e it, and m u c h of the impact the m e d i a did have w a s m e d i a t e d by c o m m u n i t y leaders in a kind of " t w o - s t e p f l o w " (Lazarsfeld, B e r e l s o n , and G a u d e t , 1944; Berelson, Lazarsfeld, and M c P h e e , 1954). If m e d i a effects w e r e , as almost all research i n d i c a t e d , " m i n i m a l , " then the d a n g e r from political elites w h o might exploit the m e d i a to m a n i p u l a t e m a s s opinion must be m i n i m a l as well - or so it s e e m e d to the majority of m a i n s t r e a m c o m m u n i c a t i o n researchers w h o , n o t w i t h s t a n d i n g the i m p o r t u n i n g s of a handful of identifiably leftist a c a d e m i c s (for e x a m p l e , M i l i b a n d , 1969), were content to ignore the possibility of elite d o m i n a t i o n of m a s s opinion by m e a n s of the mass media. Recently, however, t h e s t a n d a r d belief in " m i n i m a l " effects has c o m e u n d e r severe attack. U s i n g w i d e l y different m e t h o d o l o g i e s but converging on the s a m e c o n c l u s i o n , several research g r o u p s have found e v i d e n c e of substantial m e d i a

domination

of public

opinion

311

effects (Patterson and M c C l u r e , 1974; Patterson, 1980; Iyengar, Peters, and Kinder, 1982; Iyengar and Kinder, 1987; B a r t e l s , 1988; F a n , 1988; Brody, 1991; P a g e , S h a p i r o , and D e m p s e y , 1987; Page and S h a p i r o , in p r e s s ) . No single n e w s story or broadcast may have great effect, but the c u m u l a t i v e effect of many stories over a period of m o n t h s or years may n o n e t h e l e s s be large (Iyengar, 1991). T h u s , although the old m o d e l of the m e d i a as a " h y p o d e r m i c n e e d l e " that could inject ideas into the b o d y politic on c o m m a n d h a s not been revived, m a i n s t r e a m c o m m u n i c a t i o n research has now developed a healthy respect for w h a t the m e dia, and t h e politicians w h o use it, can a c c o m p l i s h . If t h e theory of m i n i m a l m e d i a effects h a s lost its a c a d e m i c w a r r a n t , then so has c o m p l a i s a n c e over the potential for elite m a n i p u l a t i o n of m a s s opinion through use of the m e d i a . In this concluding s e c t i o n , I therefore turn to this issue. My discussion will be heavily structured by the c a t e g o r i e s of my earlier, empirical analyses.

DEFINING

ELITE

DOMINATION

T h e a r g u m e n t of this b o o k is, on first i n s p e c t i o n , scarcely e n c o u r a g i n g with respect to d o m i n a t i o n of m a s s opinion by elites. M a n y c i t i z e n s , as w a s a r g u e d , pay too little attention to public affairs to be able to respond critically to t h e political c o m m u n i c a t i o n s they e n c o u n t e r ; rather, they are blown about by w h a t e v e r current of information m a n a g e s to develop the greatest intensity. T h e m i n o r i t y of citizens w h o are highly attentive to public affairs are scarcely m o r e critical: They respond to new issues mainly on the basis of t h e p a r t i s a n s h i p and ideology of the elite sources of t h e m e s s a g e s . If m a n y citizens are largely uncritical in their response to political c o m m u nications as c a r r i e d in the m a s s m e d i a , and if most of the rest respond m e c h a n ically on the basis of p a r t i s a n c u e s , how can o n e deny t h e existence of a substantial degree of elite d o m i n a t i o n of public o p i n i o n ? It all d e p e n d s on how o n e defines elite d o m i n a t i o n . If one takes it to m e a n any situation in which the public c h a n g e s its opinion in t h e direction of the " i n f o r m a t i o n " and leadership c u e s supplied to it by elites, indeed, there is not m u c h to a r g u e about. Not only t h e present study, but several others provide a b u n d a n t e v i d e n c e of this sort of elite d o m i n a t i o n (Iyengar and Kinder, 1987; P a g e , Shap i r o , and D e m p s e y , 1987; F a n , 1988). Yet the matter cannot be d e c i d e d so easily. Of course the public r e s p o n d s to elite-supplied information and leadership cues. H o w could it be o t h e r w i s e in a w o r l d in which events are a m b i g u o u s and in w h i c h the public m u s t regularly have opinions about m a t t e r s that a r e , to use Llippmann's p h r a s e a g a i n , " o u t of reach, out of sight, out of m i n d " ( 1 9 2 2 , 1946: p. 21)? P a g e and Shapiro (in p r e s s ) , recognizing an u n a v o i d a b l e d e p e n d e n c y of public opinion on elite d i s c o u r s e , frame the p r o b l e m in t e r m s of the "quality of information and interpretation [that] is conveyed to the p u b l i c . " T h e y c o n t i n u e ,

312

N A T U R E AND ORIGINS OF MASS O P I N I O N

To the extent that the public receives useful interpretations and correct and helpful information - information and interpretations that help it arrive at the policy choices it would make if fully informed - the policy preferences it expresses can be considered "authentic." . . . Individuals or institutions that influence public opinion by providing correct, helpful political information can be said to educate the public. On the other hand, to the extent that the public is given erroneous interpretations or false, misleading, or biased information, people may make mistaken evaluations of policy alternatives and may express support for policies harmful to their own interests and to values they cherish. An extreme result of such mistaken evaluations could be the systematic "false consciousness" or "hegemony" of which some Marxists and other theorists speak. . . . Those who influence public opinion by providing incorrect, biased, or selective information may be said to mislead the public. If they do so consciously, and deliberately, by means of lies, falsehoods, deception, or concealment, they manipulate public opinion, (ch. 9; emphasis in original) T h e difficulty in this way of framing the p r o b l e m , as Page and Shapiro a c k n o w l e d g e , is that it requires independent k n o w l e d g e of (or a s s u m p t i o n s about) which interpretations and information are c o r r e c t , and such independent k n o w l e d g e is largely unavailable. Despite this, however, Page and S h a p i r o att e m p t to identify cases of elite m a n i p u l a t i o n . For e x a m p l e , they w r i t e that President Reagan misled or manipulated the public on a variety of foreign policy matters. Calling the Soviet Union an "evil empire," with leaders willing to " l i e , cheat, and steal" for their ends, he made exaggerated charges that the Soviets had broken the SALT arms control treaties, and he portrayed the U.S. as advocating arms control while he in fact resisted reaching agreement, (ch. 9)

Elite

domination

of public

opinion

313

W i t h this in m i n d , I define elite d o m i n a t i o n as a situation in which elites induce citizens to hold opinions that they would not hold if aware of the best available information and analysis. T h i s c o n c e p t i o n is still p r o b l e m a t i c in that it d e p e n d s on an a s s e s s m e n t of what the public w o u l d believe if it w e r e fully informed. But this difficulty is not, I believe, an i n s u r m o u n t a b l e o n e , as the following p a r a b l e will suggest.

T H E PARABLE OF P U R P L E L A N D O n c e there w a s a c o u n t r y that w a s inhabited by t w o kinds of p e o p l e , blues and reds. Blues and reds shared m a n y v a l u e s , but they evaluated public policies differently. Blue people preferred s h o r t , round policies expressed in strong c o l o r s , w h e r e a s the reds preferred tall, rectangular policies a r t i c u l a t e d in pastel colors. In c o n s e q u e n c e of their ideological differences, w h i c h might or m i g h t not have been rooted in differences of material interest, reds and blues w e r e in constant political d i s a g r e e m e n t . But both sides valued reason and e v i d e n c e , and so each c o m m i s s i o n e d e x p e r t s to advise t h e m . Of c o u r s e , blues hired blue persons as experts and reds hired red persons as e x p e r t s , but they c h a r g e d their expert advisors to a r g u e , d i s c u s s , and d e b a t e with o n e a n o t h e r in an effort to a c h i e v e , if at all p o s s i b l e , the best resolutions to policy p r o b l e m s . To e n c o u r a g e experts to get the best a n s w e r s to policy d i l e m m a s , they offered very large prizes - consisting of s t a t u s , research s u p p o r t , a n d , in a handful of c a s e s , public recognition - to those e x p e r t s w h o were able to m a k e c o n v i n c i n g a r g u m e n t s to other experts.

T h o u g h not wishing to defend either the R e a g a n a d m i n i s t r a t i o n ' s policies or its use of rhetoric in these instances, I would also be reluctant to cite t h e m as cases of m a n i p u l a t i o n of public opinion. W h o , after all, can say that the Soviet U n i o n w a s not, in s o m e s e n s e , an evil e m p i r e , or that it w a s truly the United S t a t e s r a t h e r than the Soviet g o v e r n m e n t that w a s d r a g g i n g its feet on a r m s control? J u d g m e n t s on such m a t t e r s are inherently political - which often m e a n s ideological - and it is a m i s t a k e , in my view, to u n d e r t a k e an evaluation of e l i t e - m a s s relations on the a s s u m p t i o n that o n e ' s own j u d g m e n t s c a n , in gene r a l , rise above p a r t i s a n s h i p .

Like all free c o u n t r i e s , P u r p l e Land had professional politicians and political activists to take the lead in public affairs. But the politicians and activists of P u r p l e Land were p r a g m a t i c and p e o p l e - o r i e n t e d sorts w h o rarely c a m e u p with ideas on their o w n . Rather, they looked to e x p e r t s of their o w n c o l o r a t i o n for i d e a s , and w h e n a congenial expert g r o u p p r o p o s e d s o m e t h i n g new, t h e politicians and activists d i d n ' t ask many h a r d q u e s t i o n s . T h e i r m a i n c o n c e r n s were the readiness of t h e public to receive the idea, a d v a n t a g e o u s framing of t h e idea in partisan d e b a t e , and other matters of effective m a r k e t i n g , l l e n c e , the politicians rarely ventured beyond t h e p a r a m e t e r s of expert d i s c o u r s e .

Yet Page and S h a p i r o ' s notion of " i n f o r m a t i o n and interpretations that help [the public] a r r i v e at the policy choices it w o u l d m a k e if fully i n f o r m e d " is, I b e l i e v e , a conceptually useful o n e . For o n e t h i n g , " f u l l y i n f o r m e d " j u d g m e n t s , if they refer to all available information r a t h e r than to all possible i n f o r m a t i o n , need not always b e correct j u d g m e n t s . O n e can m a k e o n e ' s best decision o n the basis of available information and still be dead w r o n g . Further, p e o p l e w h o are " f u l l y i n f o r m e d " may n o n e t h e l e s s d i s a g r e e , a s e x p e r i e n c e regularly s h o w s . O n these t w o c o u n t s , t h e n , there is latitude for o p p o s i n g groups to disagree radically w i t h o u t each of t h e m risking a c h a r g e from the other that it is seeking to m a nipulate - or in my t e r m s , to d o m i n a t e - public o p i n i o n .

N e i t h e r red nor blue citizens were especially interested in politics. T h e y preferred to devote their t i m e to their j o b s , their families, and to b a s e b a l l , the national p a s t i m e . So they d i d n ' t take t h e trouble to follow political d e b a t e very carefully; rather, they c o m m i s s i o n e d c o m m u n i c a t i o n specialists to k e e p t h e m inf o r m e d , in general and easily c o m p r e h e n s i b l e t e r m s , of what each political group thought. Citizens were so apolitical that few paid attention to which e x p e r t s or politicians e n d o r s e d which p a r t i c u l a r policies, but those w h o did notice w o u l d m e chanically adopt the o p i n i o n s of their o w n type of elite, as r e p o r t e d by c o m m u n i c a t i o n specialists in the press. T h e r e m a i n d e r simply spouted w h a t e v e r

314

N A T U R E AND ORIGINS OF MASS O P I N I O N

idea w a s at the top of their h e a d s , without attaching much significance to w h a t they said. T h e one thing no citizen ever did w a s to think for himself or herself. All simply selected from the m e n u of elite-supplied options. If elite d o m i n a t i o n consists of elites inducing the public to hold attitudes that it would not hold if fully informed, it may be said that the citizens of P u r p l e L a n d entirely avoided elite d o m i n a t i o n . W h e n , despite differences in o u t l o o k , blue and red experts agreed with one a n o t h e r and got most politicians and citizens to go along with t h e m , citizens could feel assured that, even if they d e voted their w h o l e lives to investigating the given policy p r o b l e m , they w o u l d not reach conclusions m u c h different from the o n e s a d v o c a t e d by the e x p e r t s . For the expert c o m m u n i t y included persons having the s a m e values as the c o m m u nity at large, and reached its conclusions after extensive analysis of the best available information. E v e n in cases of elite d i s a g r e e m e n t , in which each type of citizen m e c h a n i cally followed the advice of his or her o w n type of politician or e x p e r t , there w a s no elite d o m i n a t i o n . For citizens could still be confident that, the m o r e closely they looked into a subject, the m o r e likely they would be to reach the s a m e conclusion reached by the expert s u b c o m m u n i t y sharing their o w n values. This p a r a b l e shows that it is possible at least to imagine c o n d i t i o n s in w h i c h the d e p e n d e n c e of m a s s op-nion on the information and analyses c a r r i e d in elite disc o u r s e is great, and y e : in which elite d o m i n a t i o n of public o p i n i o n , by a plausible construction of the t e r m , is unlikely to occur. T h e s e conditions are 1. predispositional differences among the experts paralleling those within the general public, such that experts are motivated to examine issues from all viewpoints; 2. institutional incentives for experts to develop effective solutions to pressing problems; 3. a press that, whatever else it also does, provides ample coverage of all expert viewpoints, where the term " e x p e r t " is broadly construed to include anyone having specialized knowledge of a problematic subject; 4. politicians and activists that keep within the parameters of expert opinion; 5. a citizenry that is capable, in cases of elite disagreement, of aligning itself with the elite faction that shares its own predispositions. A l t h o u g h one may be able to imagine better or stronger safeguards against elite d o m i n a t i o n , the o n e s proposed here w o u l d be r e a s o n a b l y effective, and they have the virtue of being researchable by s t a n d a r d empirical t e c h n i q u e s . T h e researcher need have no special or suprapolitical insight into the " c o r r e c t n e s s " of the leadership p r o v i d e d by elites. It is only n e c e s s a r y to e x a m i n e the processes by which leadership cues are g e n e r a t e d and diffused. T h e r e m a i n d e r of t h e c h a p t e r will use these c o n d i t i o n s as the basis for e x a m ining the degree of elite d o m i n a t i o n that exists in the United States. T h e a i m will not be to settle the q u e s t i o n , which is obviously impossible in the few p a g e s that r e m a i n in this study, but to show the kinds of issues that need to be discussed and the kinds of additional e v i d e n c e n e c e s s a r y to reach a c o n v i n c i n g c o n c l u s i o n .

Elite

domination

of public

opinion

315

T H E P O L I T I C A L C O M M U N I C A T I O N S SYSTEM O F T H E U N I T E D STATES The

press

Scholars interested in the quality of information and analyses available to the public have tended to focus on the press as the m o s t p r o x i m a t e source of that i n f o r m a t i o n . H e n c e , a great deal of information about it is available. O n e of the clearest findings to e m e r g e from the scholarly literature on the press is that reporters will regard as n e w s w o r t h y that which their " l e g i t i m a t e " sources say is n e w s w o r t h y ( C o h e n , 1963). As L e o n Sigal (1973) w r i t e s , Even when the journalist is in a position to observe an event directly, he remains reluctant to offer interpretations of his own, preferring instead to rely on his news sources. For the reporter, in short, most news is not what has happened, but what someone says has happened, (p. 69) As o n e journalist put it: " W e d o n ' t deal in facts, but in attributed o p i n i o n s " (cited in G a n s , 1980: p. 130). O n e reason for the well-established reliance of reporters on their sources is that j o u r n a l i s t s have a tendency to " g o n a t i v e , " identifying with the sources and w a n t i n g to p r o m o t e the s o u r c e ' s w o r l d view. As Russell B a k e r has w r i t t e n , The State Department reporter quickly learns to talk like a fuddy-duddy and to look grave, important, and inscrutable. The Pentagon man always looks like he has just come in from maneuvers. The Capital Hill reporter . . . affects the hooded expression of a man privy to many deals. Like the politicians he covers, he tends to garrulity, coarse jokes, and bourbon and to hate reform, (cited in Sigal, 1973: p. 49) A m o r e fundamental reason that r e p o r t e r s rely so heavily on sources is that they have no real c h o i c e . R e p o r t e r s have neither the t i m e nor the training nor, in most c a s e s , the inclination t o d o p r i m a r y investigations. A s David H a l b e r s t a m has written in The Powers That Be (1979) of the j o u r n a l i s t s w h o cover foreign affairs, they had come to journalism through the traditional routes, they had written the requisite police stories and chased fire engines and they had done all that a bit better than their peers, moving ahead in their profession, and they had finally come to Washington. If after their arrival in Washington they wrote stories about foreign policy, they did not dare inject their own viewpoints, of which they had none, or their own expertise, of which they also had none. Rather they relied almost exclusively on what some American or possibly British official told them at a briefing or at lunch. The closer journalists came to great issues, the more vulnerable they felt. (pp. 517-18) Journalists have r e p u t a t i o n s as s w a s h b u c k l i n g c h a r a c t e r s w h o are never afraid to say w h a t they think is t r u e . It would be m o r e accurate to say that j o u r n a l i s t s are never afraid to say w h a t other people think is t r u e . Yet in a w o r l d in which there are e n o r m o u s pressures - s o m e political, s o m e e c o n o m i c , and s o m e a r i s ing from t h e prejudices of the populace - to suppress w h a t is t r u e , an intense c o m m i t m e n t to publicizing the v i e w s of a w i d e r a n g e of sources can be an extremely valuable service.

316

Elite domination of public opinion

N A T U R E AND ORIGINS OF MASS O P I N I O N

317

secular understanding of this subject, r e g a r d e d homosexuality as a d i s e a s e , press c o v e r a g e of the issue w a s d o m i n a t e d by a n t i h o m o s e x u a l stereotypes. T h u s , if o n e l o o k s up h o m o s e x u a l i t y in the New York Times Index of 1950, o n e finds a request to see the perversion and scandal listings. But w h e n the A P A d e c l a r e d , by a poll of its m e m b e r s h i p in 1974, that h o m o s e x u a l i t y w a s no longer to be c o n s i d e r e d a d i s e a s e , the press began to e m p l o y a " c i v i l r i g h t s ' frame of refere n c e alongside the old " v i c e " frame, thus offering the public an a l t e r n a t i v e way of c o n c e p t u a l i z i n g the issue of homosexuality.

40

Count of 30 news stories using civil rights frame on ABC, 20 CBS, and NBC news programs 10

69 70 71 72 73 74 75 76 77 78 79 Year Figure 12.1. Number of stories using civil rights frame on network television news. Source: Vanderbilt Television News Archives. It i s , however, a m i s t a k e to think of the relationship between the press and its sources in m e c h a n i s t i c t e r m s . T h e sources responsible for a given story m a y not be q u o t e d in it, c o n s u l t e d for it, or even personally familiar to the reporter. T h i s is b e c a u s e , most generally, w h a t the press a t t e m p t s to do is to cover the n e w s from " a l l l e g i t i m a t e points of v i e w , " w h e r e legitimacy is conferred not only by g o v e r n m e n t a u t h o r i t i e s and interest g r o u p leaders, but also by subject m a t t e r specialists w h o have b e e n , in effect, accredited by m a i n s t r e a m institutions such as the g o v e r n m e n t , universities, r e s e a r c h institutes, or major b o o k publishers. H e n c e , if there is a significant fraction of m a i n s t r e a m expert opinion that h o l d s a p a r t i c u l a r v i e w , t h e press will l o o k for w a y s to use that v i e w as the basis of n e w s r e p o r t s . W h e n , for e x a m p l e , C B S television n e w s c o r r e s p o n d e n t M o r e l y Safer filed his famous story of a U . S . m a r i n e using his cigarette lighter to b u r n a V i e t n a m e s e p e a s a n t ' s t h a t c h e d roof hut ( H a l b e r s t a m , 1979), a story that w a s obviously a n t i w a r in its i m p l i c a t i o n s , it is doubtful that Safer c h e c k e d with any p a r t i c u l a r source on how to frame t h e story. However, he w a s n o n e theless reflecting a view of t h e w a r that w a s held by a significant fraction of foreign policy e x p e r t s . In a s s e r t i n g such a broad d e p e n d e n c e of m e d i a r e p o r t i n g on expert o p i n i o n , I am going beyond the existing literature on the p r e s s , w h i c h h a s formulated press d e p e n d e n c y in t e r m s of d e p e n d e n c e on p a r t i c u l a r s o u r c e s , most often g o v e r n m e n t officials dealing with a foreign policy issue. However, there is no reason to believe that a press w h i c h is congenitally d e p e n d e n t on g o v e r n m e n t sources for a type of story that is d o m i n a t e d by the g o v e r n m e n t will shed that d e p e n d e n c e in other types of situations. T h e b r o a d e r d e p e n d e n c e m a y be more difficult to d o c u m e n t , but it is likely to be present nonetheless. C o n s i d e r the treatment of homosexuality in the m e d i a . In the period in which the A m e r i c a n Psychiatric Association ( A P A ) , as the most authoritative source of

T h e c h a n g e in n e t w o r k television n e w s c o v e r a g e of homosexuality, in p a r t i c ular, o c c u r r e d quite suddenly at the t i m e of the A P A v o t e . In the five years prior to the A P A vote in 1974, TV n e t w o r k n e w s devoted 14 m i n u t e s to stories that, implicitly or explicitly, referred to h o m o s e x u a l s as a minority g r o u p seeking its rights. In the next five y e a r s , the n e t w o r k s devoted 135 m i n u t e s to stories of this k i n d , a nearly tenfold i n c r e a s e . T h e s a m e d a t a , a r r a y e d in t e r m s of story c o u n t s , are shown in Figure 1 2 . 1 . 1

My e x a m i n a t i o n of this n e w civil rights c o v e r a g e , which w a s generally but not always s y m p a t h e t i c to g a y s , indicates that most fell into one of three c a t e g o r i e s : 2

1. Coverage of a lawsuit by Air Force Lieutenant Leonard Matlovich, who was attempting to reverse the decision of the military to dismiss him for being homosexual. Over several days, all three networks devoted significant attention to this story. 2. Coverage of state and local referenda on gay rights. Following the APA vote, the media regarded these referenda, beginning with one in Dade Country, Florida, as having national political significance. 3. Human interest stories on gay and lesbian organizations, especially those forming on college campuses. These stories examined the loneliness of homosexual students, and their efforts to form support organizations, hold social events such as gay dances, and oppose discriminatory rules. It m i g h t be objected that n o n e of these stories had any direct c o n n e c t i o n with the A P A v o t e , and that w h a t really c h a n g e d w a s t h e behavior of h o m o s e x u a l s , w h o were b e c o m i n g increasingly militant and increasingly effective in d e m a n d ing their rights. C e r t a i n l y the increased militancy and visibility of h o m o s e x u a l s c o n t r i b u t e d to the c h a n g e in press c o v e r a g e . But it w a s hardly the w h o l e e x p l a n a t i o n . For exa m p l e , nothing in t h e gay rights m o v e m e n t c o m p e l l e d the spate of largely s y m pathetic stories about c a m p u s gay rights o r g a n i z a t i o n s . In one of these stories, N B C r e p o r t e r Betty Rollins noted: The question is, do homosexual organizations encourage homosexuality? Psychiatrists we spoke to think not. They point out that in this society, no one wants to be a homosexual who isn't one and that if a kid goes as far as joining a homosexual club, he is a homosexual. All the club does is to make him feel less alone and less terrible about the sexual feelings he has and can't help having. 3

1 Calculated from the Television News Archive of Vanderbilt University. 2 Some of news under the civil rights rubric included persons arguing against civil rights for gays, as in the case of the well-publicized opposition of singer Anita Bryant to a Dade County, Florida ordinance that outlawed discrimination on the basis of sexual preference. 3 NBC Evening News, April 13, 1973.

N A T U R E AND ORIGINS OF MASS O P I N I O N

Elite domination of public opinion

T h e explicit reference to expert opinion in this story is not, however, typical of the new coverage of homosexuality. To the extent that expert opinion w a s important in most stories, it w a s as deep b a c k g r o u n d . For w h a t c h a n g e d in the 1970s w a s that the press w a s no longer seeing h o m o s e x u a l s as social d e v i a n t s , but as o r d i n a r y citizens suffering the effects of h o m o p h o b i c prejudice. Given this c h a n g e of view, w h i c h as I show later o n , clearly originated a m o n g psychiatric e x p e r t s , m u c h else followed more or less automatically. C o n s i d e r the contrast b e t w e e n the last story dealing with gay rights on C B S n e w s p r i o r to the A P A v o t e , and the first such story following t h e v o t e . In the earlier story, gays had w o n a c o u r t victory p e r m i t t i n g t h e m to d e m o n s t r a t e at the D e m o c r a t i c national c o n v e n t i o n in M i a m i in 1972. Walter C r o n k i t e , in a slot often used for light n e w s at the end of his p r o g r a m , reported the victory this way:

c a m e out in opposition to the m e a s u r e : California g o v e r n o r J e r r y B r o w n , former g o v e r n o r and presidential aspirant R o n a l d R e a g a n , and President J i m m y Carter. It is h a r d to imagine such b e h a v i o r on the p a r t of a m b i t i o u s politicians in an e r a w h e n e x p e r t s were in u n a n i m o u s a g r e e m e n t that h o m o s e x u a l i t y w a s a form of m e n t a l illness and in w h i c h the press routinely c a t e g o r i z e d n e w s about h o m o sexuality u n d e r the r u b r i c of p e r v e r s i o n .

318

Miami Beach had laws against female impersonation. Today a federal judge in Miami struck them down, saying they were too vague, and besides, discriminated against men. The suit was brought by the Civil Liberties Union on complaint of gay activists who plan to demonstrate at the Democratic convention next month. They'll apparently give the fashion reporters something to contemplate. And that's the way it i s . 4

T h u s , gays were still fair target for mild ridicule on the national n e w s in 1972. T h e next gay rights stor> on C B S o c c u r r e d shortly after the A P A vote and involved the a t t e m p t s of gay activists in N e w York City to win passage of an antid i s c r i m i n a t i o n law, an effort that had been u n d e r way for several years. C r o n k i t e introduced the story as follows: Part of the new morality of the 1960s and 70s is a new attitude toward homosexuals. The homosexual men and women have organized to fight for acceptance and respectability. They have succeeded in winning equal rights under the law in many communities, but in the nation's biggest city, the fight goes o n . s

C o v e r a g e of the M a t l o v i c h c a s e represents an equally significant c h a n g e in press behavior. M a t l o v i c h w a s far from being the first h o m o s e x u a l to protest being d i s m i s s e d from a g o v e r n m e n t j o b b e c a u s e of his sexual o r i e n t a t i o n . He w a s , however, the first to get e x t e n s i v e , respectful c o v e r a g e on the evening TV n e w s for doing s o . T h a t this coverage c a m e very soon after the A P A ' s d e c l a r a t i o n on h o m o s e x u a l i t y s e e m s m o r e than c o i n c i d e n c e . T h e point, t h e n , is that h o m o s e x u a l s had been seeking equal rights for s o m e years w h e n , in the a f t e r m a t h of the A P A v o t e , t h e m a i n s t r e a m press began to take a different view of their efforts. T h e c h a n g e in media r e p o r t i n g c a n n o t be linked in a m e c h a n i c a l way to the A P A vote but is nonetheless best u n d e r s t o o d as reflecting the p r e s s ' s general sensitivity to expert authority. It w a s not, incidentally, only the p r e s s , but also m a n y m a i n s t r e a m politicians w h o s e b e h a v i o r c h a n g e d in the aftermath of the A P A vote. T h u s , w h e n voters in California w e r e asked in 1978 to d e c i d e on a.ballot initiative that w o u l d m a k e it easier to fire h o m o s e x u a l s c h o o l t e a c h e r s , three nationally prominent politicians 4 June 22, 1972.

5 May 6, 1974.

319

T h e r e is, in s u m , reason to believe that the press is to a c o n s i d e r a b l e d e g r e e d e p e n d e n t on subject m a t t e r specialists, including g o v e r n m e n t officials a m o n g m a n y o t h e r s , in framing and r e p o r t i n g the n e w s . I have a t t e m p t e d to s u m m a r i z e the nature of this d e p e n d e n c e by a s s e r t i n g that t h e press u n d e r t a k e s , in g e n e r a l , to cover t h e n e w s from all legitimate points of view, w h e r e legitimacy is determ i n e d by w h a t m a i n s t r e a m e x p e r t s take s e r i o u s l y . 6

Experts T h e a r g u m e n t of the p r e c e d i n g s e c t i o n , if c o r r e c t , places the real responsibility for selecting the ideas that reach the public with the p r e s s ' s sources. As V. O. Key, Jr. (1961), c o n c l u d e d in this r e g a r d , " T h e picture of the press collectively as the w i e l d e r of great p o w e r on its own initiative d o e s not fit the f a c t s " ( p . 394). The managers of the mass media, he continues, should no more be held accountable for the materials that flow through their channels than should the managers of transportation concerns be blamed for the quality of the printed matter they transport from place to place. The tone and quality of the content of the media tend to be mightily influenced, if not fixed, by those who manufacture news, (p. 395) T h i s in turn suggests that if we are interested in t h e quality of the information reaching the public, we must understand how it is manufactured, w h i c h is to say, we must understand t h e politics of expert c o m m u n i t i e s as they relate to the g e n e r a t i o n and diffusion of k n o w l e d g e c l a i m s , policy r e c o m m e n d a t i o n s , and general frames of reference. We especially need to k n o w the extent to which these c o m m u n i t i e s are ideologically d i v e r s e , o p e n to c h a l l e n g e , and institutionally s u p p o r t i v e of s e r i o u s - m i n d e d a t t e m p t s to discover feasible solutions to pressing p r o b l e m s . T h e r e has b e e n , as far as I can tell, little scholarly attention to these q u e s tions. T h e r e a r e , to be s u r e , n u m e r o u s accounts of h o w p a r t i c u l a r g r o u p s of individuals have a t t e m p t e d to resolve p a r t i c u l a r policy d i l e m m a s . But a t t e m p t s to generalize from these a c c o u n t s , or to a n s w e r s y s t e m a t i c a l l y questions of the type just p o s e d , a p p e a r to be a l m o s t entirely lacking in the literature of political science and c o m m u n i c a t i o n studies. Let m e , t h e n , offer an anecdotal glimpse of the w o r k i n g s of the s u b c o m m u nity of persons specializing in questions of n u c l e a r strategy. It involves types of personal interactions t h a t , as far as I have been able to tell, are o u t s i d e existing 6 This argument is similar to Bennett's (1990) indexing hypothesis, except that it refers to a broader range of sources than Bennett considers to be influential.

320

Elite

N A T U R E AND ORIGINS OF MASS O P I N I O N

studies of c o m m u n i c a t i o n , but that a r e , I b e l i e v e , of the utmost i m p o r t a n c e for u n d e r s t a n d i n g the forces that u l t i m a t e l y shape public o p i n i o n . A few years a g o , it w a s p r o p o s e d by a g r o u p of scientists that any significant use of n u c l e a r w e a p o n s by either the United States or the former Soviet Union w o u l d result in an e n v i r o n m e n t a l disaster of m o n u m e n t a l p r o p o r t i o n s . T h e idea w a s that the explosion of large n u m b e r s of n u c l e a r b o m b s , possibly only a few d o z e n , w o u l d throw up e n o r m o u s a m o u n t s of dust, d e b r i s , and s m o k e into the a t m o s p h e r e , b l o c k i n g out t h e sun and c r e a t i n g a d e c a d e s long " n u c l e a r w i n t e r " of near-total d a r k n e s s and freezing t e m p e r a t u r e s . As a result, the use of n u c l e a r w e a p o n s , even in self-defense, w o u l d be suicidal not only for o n e ' s n a t i o n , but for o n e ' s race and for all of t h e e a r t h ' s creatures (Schell, 1982). T h e idea of n u c l e a r w i n t e r w a s , as would be e x p e c t e d , most p o p u l a r a m o n g libe r a l s , w h o used it as an a r g u m e n t for restricting p r o d u c t i o n of n u c l e a r w e a p o n s . At the t i m e w h e n these ideas were being d e b a t e d in the m e d i a , I w a s put in c h a r g e of organizing a speaker p r o g r a m at my university on t h e subject of the morality of n u c l e a r war. O n e of the people I invited w a s a well-known conservative e x p e r t w h o had w r i t t e n extensively on n u c l e a r war, including strategies for fighting n u c l e a r w a r s on a limited but, a c c o r d i n g to the theory of n u c l e a r winter, potentially c a t a s t r o p h i c scale. W h e n I a s k e d a liberal nuclear w a r policy expert to introduce this p e r s o n and to serve as m o d e r a t o r for the e n s u i n g disc u s s i o n , he flatly and firmly refused. T h e ideas of the c o n s e r v a t i v e war e x p e r t w e r e d e e p l y i m m o r a l , th? liberal expert said, and it w o u l d hardly be possible for h i m to attend the talk, let alone serve as m o d e r a t o r for it. However, after m u c h c o a x i n g - including my insistence that, as moderator, he would be in a better position to o p p o s e the baleful influence of t h e c o n s e r v a t i v e expert - he agreed to s e r v e as moderator. On t h e day of t h e talk, I w a s chatting with the c o n s e r v a t i v e expert w h e n the liberal d r o p p e d in. After s t a n d a r d introductions were m a d e , the liberal c a l m l y a s k e d his fellow e x p e r t , " W e l l , w h a t are you w o r k i n g o n these d a y s ? " 7

" N u c l e a r w i n t e r , " replied the c o n s e r v a t i v e . " T h a t ' s i n t e r e s t i n g , " said the liberal. " W h a t are you f i n d i n g ? " " T h a t it's basically a n o n s t a r t e r , " replied the c o n s e r v a t i v e . - • I stepped back and looked a r o u n d for shelter, fearing a nuclear explosion w a s about to occur. But t h e w o r r y w a s unfounded. " Y e a h , t h a t ' s pretty m u c h the way I see it, t o o , " said the liberal, w h o w a s a physicist by training. " I ' v e d o n e an awful lot of calculations and I c a n ' t c o m e up with any that m a k e n u c l e a r w i n t e r s e e m like a real possibility." W i t h that, t h e t w o m e n launched into a discussion of exactly w h a t t h e incine r a t i o n rates of various substances w e r e , h o w m u c h of what kinds of c o m b u s tible m a t e r i a l s are found in typical c i t i e s , and other technical p a r a m e t e r s of the process by w h i c h n u c l e a r w i n t e r m i g h t or might not be created. T h e y a p p e a r e d to agree on almost e v e r y t h i n g . 7 Quotations are approximate.

domination

of public

opinion

321

S h o r t l y after these e v e n t s , discussion of n u c l e a r winter d i s a p p e a r e d from the p o p u l a r m e d i a and the issue w a s seemingly forgotten. T h i s a n e c d o t e , w h i c h is straight out of P u r p l e L a n d , suggests several important points. First, despite a g r e e m e n t on n u c l e a r winter, there is clear e v i d e n c e of ideological diversity of a sort that doubtless manifests itself in other situations. S e c o n d , strong ideological differences are not an insuperable b a r r i e r to rational discussion and to the a c h i e v e m e n t of a g r e e m e n t on p a r t i c u l a r issues. T h e key to b r i d g i n g ideological differences appears to be the existence of a b o d y of conventional scientific k n o w l e d g e , in this c a s e the laws of physics and chemistry, w h i c h both experts a c c e p t e d . T h i s k n o w l e d g e is apparently sufficiently well d e veloped and routinized that it can lead its users to accept conclusions they are p r e d i s p o s e d against. T h i r d , the press apparently lost interest in the issue after relevant experts on both sides concluded that it w a s a nonstarter, and did so d e spite a c o n s u m e r m a r k e t for stories on n u c l e a r w i n t e r that w a s p r e s u m a b l y as strong as it had ever b e e n . H o w often these things are true of other issues and expert c o m m u n i t i e s is impossible to say. C e r t a i n l y o n e can readily i m a g i n e expert c o m m u n i t i e s that enforce ideological h o m o g e n e i t y on their m e m b e r s , or t h a t , if they do permit diversity, are so d e e p l y polarized that no rational discussion can occur, or that have no c o n v e n t i o n a l k n o w l e d g e that is sufficiently strong to c o m p e l a c c e p t a n c e a m o n g differently p r e d i s p o s e d users. T h e s e issues, I m a i n t a i n , are i m p o r t a n t ones. T h e y go to the h e a r t of t h e functioning of the political c o m m u n i c a t i o n s s y s t e m in the United S t a t e s , affecting not only public opinion but g o v e r n m e n t a c t i o n . Yet, as I e m p h a s i z e , they s e e m to be outside the s c o p e of current investigation. By way of m a k i n g a p r e l i m i n a r y b u t , I h o p e , suggestive e x a m i n a t i o n of these issues, let me return to t h e issue of homosexuality. I have already sketched an a r g u m e n t about the effect of expert opinion on m e d i a treatment of this issue. I will n o w l o o k briefly at h o w expert opinion on this issue d e v e l o p e d . At m i d - t w e n t i e t h century, there had been a lengthy history of h o m o s e x u a l rights g r o u p s seeking a c c e p t a n c e by society. But the g r o u p s had been small, short-lived, and ineffective. A l t h o u g h s o m e of t h e early rights g r o u p s had believed that " t h e p e r i o d of passing over the m a t t e r [of h o m o s e x u a l i t y ] and disr e g a r d i n g it is past, for g o o d , " Lauritsen and T h o r s t a d (1974) o b s e r v e d that 8

The early homosexual rights movement was cut short, in the 1930s, after 70 years of existence, having achieved virtually no lasting breakthroughs. The small, isolated groups that came into existence here and there during the next two decades could not even be called holding actions, since there was little left to hold onto. (p. 71) It s e e m s that h o m o s e x u a l s faced t w o p r o b l e m s . As a minority g r o u p almost universally considered d e v i a n t , they were not able to m a k e h e a d w a y without having at least a few allies inside the s y s t e m , and they had n o n e . A l s o , far from w a n t i n g to challenge m a i n s t r e a m s o c i e t y ' s c h a r a c t e r i z a t i o n of their plight, many 8 Cited in Lauritsen and Thorstad, 1974: p. 22.

322

N A T U R E AND ORIGINS OF MASS O P I N I O N

h o m o s e x u a l s accepted and internalized it. As a h o m o s e x u a l rights activist w r o t e in t h e 1960s: The homosexual, whether born or bred (and the psychiatric argument is still raging) has been conditioned to think of himself as . . . something to be despised. The minute he discovers he's "different" he avidly reads anything he can on the subject. And what does he find? More ammunition for his self-contempt. He's told by psychiatric "authorities" that he's " s i c k . " So he begins to tell himself not that " T h e psychiatrists say I'm sick," but " I ' m sick." He programs himself into perpetual feelings of unworthiness." (emphasis in the original) 9

In the late 1940s, however, a handful of professional psychiatrists and psychologists u n d e r t o o k research that placed h o m o s e x u a l i t y in a new light. T h e r e h a d , to be s u r e , been p r i o r research and t h e o r i z i n g on the subject, but t h e n e w research w a s , in a small but i m p o r t a n t way, m o r e r i g o r o u s than e a r l i e r efforts: R a t h e r than study h o m o s e x u a l s w h o were either i m p r i s o n e d for s o m e c r i m e , often a sex c r i m e , or actively seeking psychiatric h e l p , the new studies found w a y s of e x a m i n i n g h o m o s e x u a l s w h o were leading o r d i n a r y lives (Bayer, 1981). T h e results were a surprise to almost e v e r y o n e . Alfred Kinsey, in his i n t e r v i e w - b a s e d studies of m a l e sexuality, t u r n e d up evidence that, he said, indicated that a large fraction of m a l e s w o u l d have a h o m o s e x u a l e x p e r i e n c e ending in o r g a s m at s o m e point in their lives. As he further argued: The opinion that homosexual activity in itself provides evidence of a psychopathic personality is materially challenged by these incidence and frequency data. Of the 40 or 50 percent of the male population which has had a homosexual experience, certainly a high proportion would not be considered psychopathic personalities on the basis of anything else in their histories. (1948: 659) A n o t h e r researcher, E v e l y n H o o k e r o f U C L A , administered personality a s s e s s m e n t tests to m a t c h e d s a m p l e s of h o m o s e x u a l and n o n h o m o s e x u a l m e n in the early 1950s and found no differences in their overall levels of personal adj u s t m e n t . O t h e r researchers followed suit, and soon there w a s a c o n s i d e r a b l e b o d y of psychiatric e v i d e n c e that raised serious q u e s t i o n about w h e t h e r h o m o sexuality w a s a form of mental illness. It is n o t a b l e that Hooker, w h o s e research proved the most influential, used s t a n d a r d social scientific research techniques. T h a t i s , she administered s t a n d a r d tests of personal adjustment and interpreted their results in c o n v e n t i o n a l l y e s tablished ways. T h e s e tests, R o r s c h a c h inkblot tests interpreted by j u d g e s blind to which had been provided by h o m o s e x u a l s , m a y not inspire great c o n f i d e n c e a m o n g n o n p s y c h i a t r i s t s , but, with respect to the issue of homosexuality, they constituted an application of neutral scientific k n o w l e d g e . No d o u b t b e c a u s e of t h i s , the initial response of a significant fraction of the psychiatric c o m m u n i t y w a s i m m e d i a t e l y favorable. H e r w o r k w a s q u i c k l y accepted for publication in an i m p o r t a n t j o u r n a l , and H o o k e r herself went on to a distinguished career. S h e w a s even asked by t h e N i x o n administration to head a 1969 federal task force on homosexuality, and by the International Encyclopedia 9 Leo Martello, cited in Marotta, 1981: p. 103.

Elite

domination

of public

opinion

323

of the Social Sciences to w r i t e its section on homosexuality. T h u s , far from p e nalizing H o o k e r for her i d e a s , the psychiatric and political c o m m u n i t i e s rew a r d e d her. H o o k e r ' s research and follow-up studies gave homosexuality, for p e r h a p s the first t i m e in A m e r i c a n history, a solid foundation of s y m p a t h e t i c opinion in a m a i n s t r e a m institution. T h e n e w view w a s by no m e a n s universally held within p s y c h i a t r y : Traditional p s y c h o t h e r a p i s t s c o n t i n u e d to a r g u e that h o m o s e x u a l i t y w a s a d i s e a s e and to treat it as s u c h . But p s y c h i a t r y w a s no longer m o n o l i t h i c in its view of homosexuality, and if my general view of efforts by the m e d i a to r e p o r t t h e n e w s from " a l l legitimate points of v i e w " is c o r r e c t , this w a s e n o u g h to b r i n g about an i m p o r t a n t degree of c h a n g e in public discussion of t h e issue. T h e earliest and most attentive a u d i e n c e for this n e w research consisted of h o m o s e x u a l s . T h e i r r e s p o n s e w a s not only interesting - for it suggests that the authority of e x p e r t i s e is accepted by g r o u p s o t h e r than t h e m e d i a - but part of the story of h o w further c h a n g e o c c u r r e d within t h e psychiatric c o m m u nity. T h e initial r e s p o n s e , a c c o r d i n g to Bayer ( 1 9 8 1 ) , w a s o n e of keen but cautious interest. H o m o s e x u a l s w a n t e d to k n o w m o r e about the n e w research but d i d n ' t leap to any c o n c l u s i o n s about it. T h e i r m e e t i n g s and p u b l i c a t i o n s gave roughly equal play to the n e w research and to representatives of the traditional s c h o o l . " W h e r e medical m e n , p s y c h o a n a l y s t s , and social scientists fail to a g r e e , " o n e contributor to a h o m o s e x u a l publication w r o t e , " l a y m e n can only c o w e r in s i l e n c e . " Yet it w a s probably inevitable that h o m o s e x u a l s , having b e e n offered a choice b e t w e e n seeing t h e m s e l v e s as mentally ill and seeing t h e m s e l v e s as n o r m a l and healthy, would eventually g r a v i t a t e toward the latter view. As this o c c u r r e d , the n e w e r psychiatric view of h o m o s e x u a l i t y t o o k on great i m p o r t a n c e in efforts by h o m o s e x u a l s both to o r g a n i z e t h e m s e l v e s and to instigate social c h a n g e . F r a n k K a m e n y , t h e l e a d e r of a g r o u p of W a s h i n g t o n , D . C . , g a y s , w a s a key figure i n both respects. " T h e entire h o m o p h i l e m o v e m e n t , " h e said, " i s going to rise or fall upon the q u e s t i o n of w h e t h e r h o m o s e x u a l i t y is a s i c k n e s s . " " K a m e n y personally rejected the authority of p s y c h i a t r y over the status of h o m o sexuality but n o n e t h e l e s s invoked scientific values and t h e new p s y c h i a t r i c e v i d e n c e to p e r s u a d e m e m b e r s of his o r g a n i z a t i o n to proclaim publicly that their sexual preferences w e r e n o r m a l and healthy. However, it w a s not until 1965 that K a m e n y w a s able to win approval for such a s t a t e m e n t , and even t h e n , it w a s carefully hedged: " i n t h e a b s e n c e of valid e v i d e n c e to the c o n t r a r y , " read the resolution of the W a s h i n g t o n D . C . M a t t a c h i n e Society, " h o m o s e x u a l i t y is not a sickness . . . but is merely a p r e f e r e n c e . " In a still later p h a s e of its r e s p o n s e to the new psychiatric e v i d e n c e , h o m o sexuals b e g a n to publicly reject psychiatric authority. E v e n as this o c c u r r e d , however, friendly psychiatric experts r e m a i n e d w e l c o m e allies. T h e s e e x p e r t s , Bayer (1981) c o m m e n t s , 1 0

12

10 Cited in Bayer, 1981: p. 74. 11 Cited ibid.: p. 82. 12 Cited in D'Emillio, 1983: p. 164.

324

N A T U R E A N D ORIGINS OF MASS O P I N I O N

provided homosexuals with evidence and with a vocabulary of criticism that were invaluable in the effort to tear the mantle of authority from those who claimed that science itself had discovered the psychopathology inherent in homosexuality. The role of the expert at homophile meetings shifted from that of providing homosexuals with insight into the etiology of their sexual preferences to that of providing insight into the illegitimate power of psychiatry, (p. 86) It is interesting to note that the behavior of h o m o s e x u a l s in this period conforms to the basic n o t i o n s of the m a i n s t r e a m and p o l a r i z a t i o n m o d e l s . W h e n h o m o s e x u a l i t y w a s c o n s i d e r e d by virtually all specialists and the press to be a d i s e a s e , h o m o s e x u a l s largely accepted this m a i n s t r e a m n o r m by staying, as the p h r a s e g o e s , " i n the c l o s e t . " This orientation w a s obviously sustained in p a r t by realistic fears of w h a t w o u l d h a p p e n if they were discovered; but it is c l e a r that in m a n y c a s e s , the m a i n s t r e a m n o r m against h o m o s e x u a l i t y w a s to a significant d e g r e e internalized. But t h e n , offered by s o m e psychiatric authorities a c h o i c e of c o n s i d e r i n g t h e m s e l v e s sick or merely to have an a l t e r n a t i v e sexual preference, h o m o s e x u a l s naturally allied t h e m s e l v e s with the friendlier view. P a r t of t h e effort to o v e r c o m e s o c i e t y ' s prejudice against h o m o s e x u a l i t y w a s a y e a r s ' long series of confrontations b e t w e e n militant gays and the A m e r i c a n Psychiatric A s s o c i a t i o n , w h i c h c o n t i n u e d in its official publications to classify h o m o s e x u a l i t y as d i s e a s e . T h u s w h e n the A P A met in W a s h i n g t o n in 1970, F r a n k K a m e n y helped to lead the protest: The planned disruption occurred on May 3, when gay and anti-war activists stormed the prestigious Convocation oi Fellows. During the ensuing uproar, Kameny grabbed a microphone and denounced the right of psychiatrists to discuss the question of homosexuality. Borrowing from the anti-war movement, he declared, "Psychiatry is the enemy incarnate. Psychiatry has waged a relentless war of extermination against us. You may take this as a declaration of war against you." Fist-shaking psychiatrists, infuriated by the invaders, compared their tactics to that of Nazi storm troopers. (Bayer, 1981: p. 105) S u c h d i s r u p t i o n s no d o u b t impelled the leaders of the A P A to think m o r e c a r e fully about the issue of h o m o s e x u a l i t y than they w o u l d o t h e r w i s e have d o n e . As a result, w h e n the trustees of the A P A , and later a majority of the A P A m e m b e r s h i p in a referendum, voted to delete h o m o s e x u a l i t y from its official list of m e n t a l d i s o r d e r s , both those w h o w o n the vote and those w h o lost felt.that the action had involved a large political c o m p o n e n t - as obviously it h a d . But equally obviously, there w a s m o r e to the politics of this expert c o m m u n i t y than 1960s-style pressure tactics. As B a y e r o b s e r v e d , s o m e of the trustees voting for the c h a n g e w e r e genuinely c o n v i n c e d that homosexuality w a s not n e c essarily a d i s e a s e , while others " f e l t privately that homosexuality w a s indeed a d i s o r d e r . . . [but] n e v e r t h e l e s s a c k n o w l e d g e d that the evidence required to s u b stantiate their position w a s l a c k i n g " (p. 136). In short, s t a n d a r d ' p s y c h i a t r i c evi d e n c e , w h a t e v e r its intrinsic m e r i t , appeared to c o u n t . If this s e e m s a naive v i e w , consider the significance of the psychiatric evid e n c e in coldly political t e r m s . O n c e the findings of H o o k e r and others b e g a n to a c c u m u l a t e , the psychiatric profession faced a g e n u i n e d i l e m m a , o n e that had eventually to be faced w h e t h e r gay militants d e m a n d e d it or not. E i t h e r psychiatrists could r e m o v e h o m o s e x u a l i t y from the list of mental disorders on the

Elite

domination

of public

opinion

325

g r o u n d s that the s t a n d a r d a s s e s s m e n t t e c h n i q u e s had failed to uncover e v i d e n c e of p s y c h i c p r o b l e m s a m o n g h o m o s e x u a l s , or they could devalue the s t a n d a r d a s sessment t e c h n i q u e s - the basis for p s y c h i a t r y ' s authority in o t h e r d o m a i n s - on the g r o u n d s that they had failed to uncover the p a t h o l o g y inherent in h o m o s e x uality. T h i s m u s t , in the e n d , have been an easy c h o i c e . T h e reader should n o t i c e that n o w h e r e in this account have I evaluated the c o r r e c t n e s s of the view of h o m o s e x u a l i t y taken by psychiatrists. T h i s is b e c a u s e the c o r r e c t n e s s of elite v i e w s is, from the p e r s p e c t i v e I am p r o p o s i n g , irrelevant. W h a t matters is, in effect, procedural c o r r e c t n e s s : w h e t h e r the relevant expert c o m m u n i t i e s are ideologically diverse and w h e t h e r they have institutionalized incentive structures that impel m e m b e r s to m a k e their best efforts to resolve pressing p r o b l e m s . W i t h respect to these q u e s t i o n s , the psychiatric profession appears to d e s e r v e a favorable e v a l u a t i o n . A researcher of no initial standing in her field, an untenured research a s s o c i a t e at what w a s then a m i n o r state university, w a s able to o p e n the first i m p o r t a n t crack in a previously i m p r e g n a b l e p o w e r s t r u c t u r e , and to do so in the m i d - 1 9 5 0 s , a t i m e not obviously propitious for such an undertaking. T h i s is an impressive a c h i e v e m e n t not only for the individual, but for the institutional a r r a n g e m e n t s that m a d e it possible. T h e most i m p o r t a n t of these institutional a r r a n g e m e n t s , I have s u g g e s t e d , w a s the existence within t h e psychiatric c o m m u n i t y of s t a n d a r d personality a s s e s s m e n t t e c h n i q u e s , w h i c h I have referred to as c o n v e n t i o n a l scientific k n o w l e d g e . T h e last few p a g e s have used a case study in o r d e r to e x a m i n e the politics of an expert c o m m u n i t y . I do not b e l i e v e , however, that the u n d e r s t a n d i n g of the d y n a m i c s of expert c o m m u n i t i e s should be too heavily d e p e n d e n t on c a s e studies. Rather, an a t t e m p t should be m a d e to d e v e l o p a general account of h o w expert c o m m u n i t i e s g e n e r a t e ideas and h o w these ideas diffuse. T h e key issues in such a general a c c o u n t , I b e l i e v e , should be possible bias in recruitment to the expert c o m m u n i t i e s , professional incentive structures within the c o m m u n i t i e s , and the d e v e l o p m e n t of a b o d y of c o n v e n t i o n a l k n o w l e d g e strong e n o u g h to c o m p e l a g r e e m e n t a m o n g differently predisposed users. T h e r e exists s o m e solid literature on the first q u e s t i o n . It indicates that persons of different ideological o r i e n t a t i o n s are attracted to different a c a d e m i c disciplines (Lipset and L a d d , 1970) a n d , o n e a s s u m e s , to different o c c u p a t i o n a l g r o u p i n g s as well. W h e t h e r this results in ideological h o m o g e n e i t y within g r o u p s , or pressures that m a k e it difficult for the minority to do its w o r k , is an i m p o r t a n t but presently u n a n s w e r e d q u e s t i o n . To the extent that r e c r u i t m e n t to expert c o m m u n i t i e s , w h e t h e r by m e a n s of v o l u n t a r y self-selection or p u r p o s i v e enforcement of ideological s t a n d a r d s , limits t h e types of persons w h o can bec o m e e x p e r t s , the r a n g e of elite discourse is restricted and the specter of elite d o m i n a t i o n is raised. W i t h respect to the s e c o n d q u e s t i o n , the existing literature is, if a n y t h i n g , even m o r e lacking. F r o m my previous d i s c u s s i o n s , it nonetheless s e e m s plausib l e to maintain that at least s o m e e x p e r t s , professionals, and o t h e r types of

326

N A T U R E AND ORIGINS OF MASS O P I N I O N

subject m a t t e r specialists do often have incentive structures that strongly e m p h a s i z e the a d v a n c e m e n t of k n o w l e d g e for the solution of pressing societal problems. A c o n s p i c u o u s e x a m p l e of t h e m may be found in the medical research c o m m u n i t y , within which s t a t u s , research s u p p o r t , and salary all d e p e n d on the practical or theoretical value of the research o n e has d o n e , as j u d g e d by peers in the field. Similar incentive structures are in place in most a c a d e m i c disciplines or at least the individuals w h o serve on r e c r u i t m e n t and p r o m o t i o n c o m m i t t e e s within universities devote a great deal of energy to trying to m a k e sure that they a r e . T h u s , one need not posit experts and professionals w h o are altruistic in ord e r to believe that many of t h e m are genuinely interested in the discovery of k n o w l e d g e of g o o d public policy. O n e need only posit incentive structures that e n c o u r a g e t h e m to have such interests. Such incentive structures obviously cannot e l i m i n a t e bias or g u a r a n t e e correct c o n c l u s i o n s . But they may n o n e t h e l e s s serve as effective i n d u c e m e n t s to try to reach these goals. T h e incentive structures prevalent in expert c o m m u n i t i e s , especially if internalized, may provide a strong influence on a p e r s o n ' s behavior. I o n c e s p o k e to a sociologist doing research in association with medical scientists w h o , in the early stages of the A I D S e p i d e m i c , were trying to isolate the A I D S virus and d e t e r m i n e its m o d e of t r a n s m i s s i o n . M a n y of these r e s e a r c h e r s , the sociologist told m e , w e r e either gay or s y m p a t h e t i c to g a y s , and so had a strong desire to find that A I D S w a s not as readily transmissible as many viruses. For a short initial p e r i o d , he said, trfss p r e c o n c e p t i o n s e e m e d to affect research p l a n s . But very quickly, he said, all r e s e a r c h e r s , gay and n o n g a y alike, began to focus on w i n n i n g " t h e big p r i z e " - that is, being the first to discover the true nature of A I D S , r e g a r d l e s s of their personal interests or w i s h e s c o n c e r n i n g its n a t u r e . It is, I reiterate, an open e m p i r i c a l question h o w strongly such incentives are felt across w h a t r a n g e of expert and professional o c c u p a t i o n g r o u p s to which r e p o r t e r s and politicians go for information - but an empirical q u e s t i o n well worth investigating. My notion of professional and expert c o m m u n i t i e s obviously d r a w s heavily on t h e idealized c o n c e p t i o n of a scientific c o m m u n i t y (see M e r t o n , 1982; H a g s t r o m , 1965). T h i s should n o t , however, o b s c u r e the critically i m p o r t a n t fact that, for m u c h of the subject m a t t e r that is relevant to politics, the personal and political predispositions of experts and professionals can and regularly do influe n c e t h e conclusions they reach (or at least try to r e a c h ) . I am aware of no research on this point, but I d o u b t that many will d i s p u t e it. Certainly within any subfield of social s c i e n c e , all of us k n o w studies w h i c h reflect the ideological t e n d e n c i e s of researchers w h o c a n , on other g r o u n d s , be identified as m o r e or less liberal or c o n s e r v a t i v e . It seems quite likely that this is true of o t h e r professional and expert c o m m u n i t i e s as well. T h u s , to return to the A I D S e x a m p l e , it s e e m s likely that, if there were s o m e medical researchers w h o w a n t e d , for reason of their predispositions or perceived interests, to be able to show that A I D S could not be casually t r a n s m i t t e d , there were o t h e r researchers w h o s e p r e d i s p o sitions or perceived interests m o t i v a t e d the o p p o s i t e c o n c l u s i o n .

Elite

domination

of public

327

opinion

I see no conflict in c o m p a r i n g professional and expert c o m m u n i t i e s to scientific c o m m u n i t i e s , while also giving very w i d e scope to the o p e r a t i o n of ideologically m o t i v a t e d p r e c o n c e p t i o n s . A s m y U C L A colleague Richard S k l a r has o b s e r v e d , social science t h e o r i e s are often c o n c e i v e d in ideological sin rather than scientific virtue. But, a l t h o u g h people may be m o t i v a t e d by various prec o n c e p t i o n s , they m u s t , in order to maintain high standing in their specialized c o m m u n i t i e s , justify their eventual conclusions in t e r m s of a r g u m e n t s that eve r y o n e can accept, including persons w h o do not share their predispositions or p r e c o n c e p t i o n s . Moreover, a c a s e can be m a d e that ideological b i a s , so long as it is o p e r a t i n g within pluralistic professional and expert c o m m u n i t i e s , has salutary effects. It w o u l d tend to assure that every potentially controversial issue is scrutinized by persons having a r a n g e of m o t i v a t e d p r e c o n c e p t i o n s . To the extent that such differently m o t i v a t e d persons can agree on a c o m m o n answer, as very often they d o , t h e rest of society w h o are not specialists can have s o m e confidence that the a n s w e r p r o p o s e d by the e x p e r t s is sound - or if not s o u n d , the best that anyone can do for the m o m e n t . A n d to the extent that differently m o t i v a t e d p e r s o n s cannot a g r e e , they w o u l d , as sources for reporters and staff advisors to politicians, air their d i s a g r e e m e n t s in p u b l i c , thereby a l e r t i n g o r d i n a r y citizens to t h e existence of uncertainty. T h i s brings u s back t o o u r m a i n c o n c e r n about elite d o m i n a t i o n . W h e n , d e spite their divergent p r e d i s p o s i t i o n s , all relevant specialists agree on a policy, any source w h o m j o u r n a l i s t s consult will say roughly the s a m e t h i n g , with the result that society will have " e l i t e c o n s e n s u s " and a " m a i n s t r e a m n o r m " that will be m o s t strongly s u p p o r t e d by the most politically attentive m e m b e r s of society. But w h e n predispositions induce relevant specialists to d i s a g r e e , j o u r n a l ists will publicize the d i s a g r e e m e n t , often in starkly ideological t e r m s that invoke i m a g e s of g o o d and evil. Politicians and publicists, w h o maintain lines of c o m m u n i c a t i o n s to like-minded specialists, will also d i s a g r e e . T h e result will be a polarization of the general public along lines that m i r r o r the elite ideological conflict, with the most attentive m e m b e r s of the public most ideologically p o l a r i z e d . 13

Ideology, in this view, is a mechanism by which ordinary citizens make contact with specialists who are knowledgeable on controversial issues and who share the citizens' predispositions. As such, ideology can m a k e a valuable contribution to d e m o c r a t i c politics in a society in w h i c h people are expected not only to have opinions about a range of impossibly difficult issues, but to use those opinions as the basis for c h o o s i n g leaders and h o l d i n g t h e m a c c o u n t a b l e . Note that ideology, as t h e t e r m is being used h e r e , is not, as it is often taken by opinion researchers to b e , the unified product of an individual creative gen i u s , such as Karl M a r x or E d m u n d B u r k e or John L o c k e . Rather, it is an 13 Congressional committee and personal staffs, policy institutes and think tanks, and overlapping parts of the federal bureaucracy often have distinct political colorations, thereby tying politicians to members of specialist communities who share their values.

328

N A T U R E A N D ORIGINS OF MASS O P I N I O N

a g g l o m e r a t i o n of v i e w s of different specialists sharing a c o m m o n predispositional b e n t . T h u s , for e x a m p l e , w h a t passes for c o n s e r v a t i v e e c o n o m i c s o r c o n s e r v a t i v e foreign policy will be w h a t most c o n s e r v a t i v e e c o n o m i s t s and conservative foreign policy e x p e r t s agree needs to be d o n e . W h e n the e x p e r t s c h a n g e their v i e w s , as liberal experts c h a n g e d their opinion about V i e t n a m , m a s s belief s y s t e m s will, with the sorts of lags I have d o c u m e n t e d in earlier c h a p t e r s , tend to follow the elite view. T h e r e is no p a r t i c u l a r need for m e m b e r s of t h e general public to k n o w technical details of the expert d e b a t e on these iss u e s , or to k n o w why a given set of policies is c o n s e r v a t i v e or liberal, in order to take positions on t h e m that are consistent with their ideological predispositions; they need only be able to r e c o g n i z e which elites share their predispositions and take cues from t h e m . 1 4

I have now sketched an idealized s y s t e m of public information in which political ideas and perspectives d e v e l o p a m o n g various kinds of policy specialists and diffuse d o w n w a r d to the public via politicians and the m a s s m e d i a . T h i s idealized s y s t e m is obviously not t h e w h o l e p i c t u r e , for it neglects the extremely imp o r t a n t role of political institutions, p a r t i s a n c o m p e t i t i o n , and political c e n s o r s h i p in regulating t h e flow of information through society. But even s o , my account of this s y s t e m m a k e s two useful points. It sets o u t , first of all, a set of idealized c o n d i t i o n s t h a t , if m e t , w o u l d greatly mitigate the n o r m a t i v e l y o b j e c t i o n a b l e effects of a heavy d e p e n d e n c e of m a s s opinion on elite d i s c o u r s e . It further shows that, as an empirical matter, s o m e of these conditions - expert c o m m u n i t i e s o p e r a t i n g at least s o m e of the t i m e within salutary incentive structures, and a press that is attentive to these c o m m u n i t i e s - are at least partially met in the current information system of the United States. I turn now to s o m e brief r e m a r k s about the effect of political factors on the o p e r a t i o n of my idealized information s y s t e m .

POLITICS AND I N F O R M A T I O N In leaving until now any discussion of the effects of " p o l i t i c s " on t h e flow of i n f o r m a t i o n , I imply that the g e n e r a t i o n of information is p r i m a r y and that political factors that regulate its d i s s e m i n a t i o n , including the activities of g o v e r n m e n t officials, o t h e r politicians, and interest g r o u p s , a m o n g o t h e r s , are secondary. T h e reason for this e m p h a s i s is, quite simply, that political information prec e d e s political action. All political actors, including the g o v e r n m e n t , must form s o m e view of w h a t is h a p p e n i n g in the w o r l d before u n d e r t a k i n g any a c t i o n , including action to c e n s o r or m a n i p u l a t e the flow of information to the p u b l i c , and this view is invariably a product of various kinds of policy and information 14 This common bent presumably involves non-issue specific predispositions toward order, stability, and social control, among other things (for some empirically grounded conjectures, see McClosky and Zaller, 1985, ch. 7; Costantini and Craik, 1980).

Elite

domination

of public

opinion

329

specialists. Moreover, since m u c h of the information s y s t e m exists o u t s i d e of the g o v e r n m e n t , the ability of politicians to suppress or control information is, t h o u g h i m p o r t a n t , limited. For any given issue, there are expert c o m m u n i t i e s in p l a c e , each at least potentially divided along ideological lines, and each with established lines of c o m m u n i c a t i o n to g o v e r n m e n t decision m a k e r s , political o p positions a n d , via the p r e s s , to the public. T h i s p e r s p e c t i v e on the flow of information in the United States is u n o r t h o d o x . M o s t a c c o u n t s , w h i c h are based on studies of the " s o u r c e s " that give inf o r m a t i o n to t h e p r e s s , e m p h a s i z e the role of g o v e r n m e n t in the g e n e r a t i o n of information and its role in m a n i p u l a t i n g or c e n s o r i n g that information ( C o h e n , 1963; Sigal, 1973, Hallin, 1986; B e n n e t t , 1990). T h e s e a c c o u n t s , to the extent that they aspire to a n s w e r general q u e s t i o n s about the sources of the information that reaches the p u b l i c , focus t o o n a r r o w l y on w h a t is c o m p a r a t i v e l y easy to o b s e r v e and study. G o v e r n m e n t officials m a y be the p r o x i m a t e source of m u c h of the information that reaches the p u b l i c , but they are not necessarily the creators of the i n f o r m a t i o n , nor can they always control h o w information is framed by the press or w h a t the press is interested in, nor are they equally i m p o r t a n t on all types of issues. An illuminating c a s e of h o w m e m b e r s of the n o n g o v e r n m e n t a l information s y s t e m can exert influence independently of g o v e r n m e n t w a s described by Washington Post r e p o r t e r D a v i d B r o d e r s o m e twenty years after it o c c u r r e d . E a r l y in the V i e t n a m War, w h e n U . S . i n v o l v e m e n t still had the s u p p o r t of almost all m a i n s t r e a m politicians and w h e n the g o v e r n m e n t w a s providing little a n t i w a r i n f o r m a t i o n , S e n a t o r F r a n k C h u r c h , a y o u n g S e n a t o r from I d a h o with a strong interest in foreign policy, a r r a n g e d a " d i n n e r s e m i n a r " for s o m e r e p o r t e r s on the subject of V i e t n a m . As B r o d e r relates: Church's guest at the dinner was Hans J. Morgenthau of the University of Chicago, an authority on foreign policy. The two men tried their best to make a largely skeptical group of reporters reexamine the prevailing assumptions about Vietnam. The struggle taking place there, they asserted, was not aggression by proxy from China or Moscow but an indigenous revolution, led by a man Ho Chi Minn - who appeared to be the only authentic Vietnamese leader on the scene. . . . If Americans . . . went down the path the French had followed, Church and Morgenthau said, then we would pay a terrible price. I went home thoroughly unconvinced that night, but had many occasions in the next dozen years to recollect the warning. (Broder, 1984) T h e details of the discussion, even t h o u g h anticipating the line a n t i w a r liberals w o u l d eventually t a k e , are u n i m p o r t a n t . W h a t is i m p o r t a n t is C h u r c h ' s or p e r h a p s M o r g a n t h a u ' s - apparently self-conscious effort to influence t h e rep o r t i n g of leading j o u r n a l i s t s by exposing t h e m to a c a d e m i c v i e w p o i n t s . Informal discussions like t h e s e , w h i c h are by no m e a n s rare in j o u r n a l i s m , are o n e route by w h i c h t h e ideas of leading policy specialists reach and inform the w o r k of practicing j o u r n a l i s t s . B o o k s are probably at least as i m p o r t a n t . Bright y o u n g r e p o r t e r s a r r i v i n g on a beat about which they k n o w little routinely seek out the leading b o o k s on their n e w subjects and use t h e m for intellectual

330

N A T U R E AND ORIGINS OF MASS O P I N I O N

o r i e n t a t i o n . In the c a s e of V i e t n a m , these b o o k s were the b o o k s of B e r n a r d Fall, a historian w h o w a s apparently read by most of the journalists in V i e t n a m ( H a l b e r s t a m , 1979) and w h o s e t h e m e w a s that the V i e t n a m W a r should be und e r s t o o d in light of a n t i i m p e r i a l i s m and V i e t n a m e s e nationalism rather t h a n , as m o s t U . S . g o v e r n m e n t p o l i c y m a k e r s t h o u g h t , as a struggle b e t w e e n d e m o c r a c y and c o m m u n i s m . If, after absorbing such i d e a s , reporters found it difficult to be as staunchly a n t i c o m m u n i s t as many of their official sources, if they actively sought out dissenting voices within the g o v e r n m e n t , if by the q u e s t i o n s they asked at press conferences they constantly and impatiently pressured high officials to give e v i d e n c e of " r e s u l t s , " and if they often took the o p p o r t u n i t y to play up e m b a r r a s s m e n t s to U . S . policy (such as c o r r u p t i o n within the S o u t h Vietn a m e s e g o v e r n m e n t or mistreatment of V i e t n a m e s e peasants by the m i l i t a r y ) , at least part of the e x p l a n a t i o n should be laid to the influence of t h e extrag o v e r n m e n t a l experts w h o created the intellectual framework that helped m o t i vate such actions. 1 5

In stressing the n o n g o v e r n m e n t a l origins of m u c h official information, I by no m e a n s deny that g o v e r n m e n t regularly uses w h a t influence it can m u s t e r to color or to c e n s o r w h a t reaches the p u b l i c . It is also certainly true that " p o l i t i c s " the t e n d e n c y of p a r t i s a n elites to disagree with one another no m a t t e r what exp e r t s m i g h t say, t h e t e n d e n c y of politicians to p a n d e r to w h a t e v e r interests can help t h e m stay in office, and the t e n d e n c y of g o v e r n m e n t officials to suppress and m a n i p u l a t e the flow of information - can easily o v e r r i d e the influence of policy specialists and a c a d e m i c experts in the short r u n , w h e r e the short r u n m a y be sufficiently long to c o m m i t the nation to an ill-advised course of a c t i o n . But t h e c o r r u p t i n g influence of p a r t i s a n politics should not be e x a g g e r a t e d . For e x a m p l e , fear of the A I D S virus and of the h o m o s e x u a l s and d r u g users w h o were initially its principal c a r r i e r s w a s surely one of the most potentially explosive backlash issues of the 1980s. Yet a backlash political issue never e m e r g e d . A national strategy of public e d u c a t i o n and intensive scientific research c a m e to be accepted by both political p a r t i e s and virtually the entire p o litical elite as the best way of dealing with the p r o b l e m . Such d i s a g r e e m e n t as o c c u r r e d centered almost exclusively on speed of the research r e s p o n s a and levels of public funding rather than on the strategy itself. T h u s A I D S has r e m a i n e d a m e d i c a l and public health issue - that is, an e x p e r t - d o m i n a t e d one - despite having great potential as a p a r t i s a n political controversy (Colby and C o o k , 1991). Yet can anyone d o u b t that, if medical e x p e r t s had disagreed on this fund a m e n t a l strategy - if, for e x a m p l e , a significant n u m b e r had favored a quara n t i n e policy - the issue w o u l d have b e c o m e instantly and bitterly p a r t i s a n and ideological? A similar a r g u m e n t can be m a d e with respect to r a c e . No issue in A m e r i c a n society is m o r e deeply political than r a c e , and yet, a variety of e x p e r t s , most 15 It is not, I imagine, uncommon for politicians to look to expert opinion for orientation when they take on a new subject. Thus, Fall's books were reputed also to have been a major influence on Senator William Fulbright, an early critic of the Vietnam War.

Elite

domination

of public

opinion

331

notably psychologists c o n c e r n e d with h u m a n intelligence and with the harmful effects of racial s e g r e g a t i o n , have been active and seemingly influential on this issue t h r o u g h o u t the twentieth century. Early in the century, the thrust of expert opinion w a s p r e d o m i n a n t l y racist, as discussed in C h a p t e r 2. However, d u r i n g the 1920s, a period that w a s , incidentally, widely r e g a r d e d as a t i m e of cultural c o n s e r v a t i s m in the c o u n t r y as a w h o l e , expert opinion on g r o u p differences in intelligence u n d e r w e n t a virtual revolution. By 1930 a new scholarly c o n s e n s u s , m i n i m i z i n g the existence and especially the significance of any possible g r o u p differences in intelligence, had b e c o m e d o m i n a n t . It is at least a r g u a b l e that this n e w expert c o n s e n s u s , w h i c h soon c a m e to be reflected in the opinions of edu c a t e d w h i t e p e r s o n s , is the reason that the w h i t e power structure of the country, after c e n t u r i e s of suppressing the efforts of blacks to gain equality, b e g a n at last to see their point ( M y r d a l , 1944). T h e r e a r e , without q u e s t i o n , m y r i a d ways in which political a u t h o r i t i e s and o t h e r interests can short-circuit the idealized s y s t e m I have sketched and distort the flow of information to the public. E x p e r t c o m m u n i t i e s are t h e m s e l v e s prey to m y r i a d internally caused dysfunctions. T h e existing literature on the press and m a s s c o m m u n i c a t i o n s h a s , to s o m e extent, focused on the b r e a k d o w n s . An unintended c o n s e q u e n c e of this e m p h a s i s is that we s o m e t i m e s get a better idea of how t h e s y s t e m fails to w o r k than of how it w o r k s at all. In view of t h i s , my aim in this c h a p t e r h a s been to direct attention toward a b r o a d e r range of factors than is n o r m a l l y e x a m i n e d , especially the existence of policy specialists w h o are relatively independent of political authority and other interests, w h o play a major role in raising and framing issues, and w h o are linked to public o p i n i o n , via the press and politicians, by shared ideological predispositions.

CONCLUDING

REMARKS

In any c o m p l e x society, as in P u r p l e L a n d , there must be a division of labor in the c o n d u c t of politically relevant d e b a t e . This necessity suggests an objection to my general a r g u m e n t . If, as I have i m p l i e d , only specialists are c o m p e t e n t to c o n d u c t political d e b a t e , why bring the public into it at all? W h y not let gove r n m e n t policy reflect the p r e p o n d e r a n t weight of expert o p i n i o n ? T h e r e are several a n s w e r s . O n e is that, quite simply, it is the p e o p l e ' s right to settle any d e b a t e they feel m o v e d to settle. A n o t h e r is t h a t , as J. S. Mill and others have a r g u e d , political p a r t i c i p a t i o n is a value in itself, ennobling and selffulfilling to all w h o p a r t a k e of it. I, however, w o u l d not find either of these arg u m e n t s very attractive if I believed that the p u b l i c , in exercising its u n d o u b t e d rights and seeking its self-fulfillment, regularly m a d e decisions that I r e g a r d e d as morally abhorrent or technically stupid. I therefore offer a third reason for involving the public in political d e b a t e : T h a t g o v e r n m e n t , in the a b s e n c e of c h e c k s , invariably g o e s astray and b e c o m e s o v e r b e a r i n g o r w o r s e . A few specialists do not constitute a viable c h e c k on political authority. It is too e a s y to i g n o r e , j a i l , kill, or o t h e r w i s e silence t h e m . T h e i r only h o p e of being

332

N A T U R E AND ORIGINS OF MASS O P I N I O N

effective is to publicize and politicize w h a t e v e r d i s a g r e e m e n t s a r i s e , so that the m a s s e s of o r d i n a r y citizens can weigh into the d e b a t e . I do not see any o b v i o u s reason to believe that the w h o l e citizenry will b e , in g e n e r a l , wiser t h a n specialists, so long as the specialists are broadly recruited and given freedom to o p e r a t e within a salutary s y s t e m of incentives. It is t h e collective ability of citizens in a d e m o c r a c y to pressure leaders in useful d i r e c t i o n s and, w h e n n e c e s sary, to r e m o v e the l e a d e r s , m o r e than the collective w i s d o m of the p e o p l e , that seems to me critical. All of this a s s u m e s , of c o u r s e , that an independent public opinion can actually c h e c k political authorities. G i v e n the c o n t i n u e d existence of c o m p e t i t i o n a m o n g partisan elites for public s u p p o r t in the current A m e r i c a n political s y s t e m , I do not see this a s s u m p t i o n as p a r t i c u l a r l y p r o b l e m a t i c . T h e real p r o b l e m is g u a r a n t e e i n g the existence of an equally vigorous c o m p e t i t i o n a m o n g o p p o s ing ideas.

Measures appendix

POLITICAL

KNOWLEDGE

S i n c e political a w a r e n e s s is the key independent variable in this study, I have extensively investigated how it may be most effectively m e a s u r e d (Zaller, 1985, 1986, 1990; P r i c e and Zaller, 1990). A l t h o u g h s o m e of this w o r k has involved c o n c e p t u a l clarification, m o s t of it has been m u n d a n e l y e m p i r i c a l , as I have painstakingly tested a l t e r n a t i v e m e a s u r e s across a variety of datasets and issues. T h e routine empirical w o r k , however, has paid a useful dividend: It h a s s h o w n that t h e effects o n e attributes to political a w a r e n e s s can d e p e n d greatly on how one g o e s about m e a s u r i n g it.

How

to

measure

political

awareness

T h e surveys of the National E l e c t i o n Studies ( N E S ) , w h i c h provide the d a t a for almost all of the n e w analyses r e p o r t e d in this b o o k , contain n u m e r o u s m e a s u r e s that w o u l d seem suitable as m e a s u r e s of political a w a r e n e s s . T h e s e include level of political p a r t i c i p a t i o n (such as e n g a g i n g in political discussions with friends, giving m o n e y to c a n d i d a t e s ) , level of political interest, level of m e d i a u s e , educational a t t a i n m e n t , and neutral factual k n o w l e d g e about politics. T h e r e is no a g r e e m e n t in t h e existing scholarly literature about which of these m e a s u r e s is b e s t . E v e n C o n v e r s e has given m i x e d signals on this q u e s t i o n . In 1

1 Some scholars use a single-item control, most commonly education or information, or, less commonly, interest (e.g., Dean and Moran, 1977; Sniderman, Brody, and Tetlock, 1991; Judd and Milburn, 1980); sometimes they use up to five or six controls either simultaneously (Achen, 1975) or combined into an index (Erikson, 1979); and sometimes they introduce a series of single-item controls in bivariate form (Fiorina, 1981). In general, researchers seem simply to use whatever measure of awareness is handy and permits them to get on with their main analysis. These measures, all compounds of the same family of variables, are given a variety of names: political interest and involvement (Converse, 1964, 1980); cognitive ability (Stimson, 1975); political sophistication (Chong, McClosky, and Zaller, 1984; Erikson, 1979; Neuman, 1986; Sidanius, 1988; Luskin, 1990); political expertise (Fiske and Kinder, 1981; Fiske, Lau, and Smith, 1990); ideological sophistication (Knight, 1985); and indicators of rational civic man (Fiorina, 1981; Key with Cummings, 1966). Judd, Krosnick, and Milburn (1981), having been criticized by Converse (1980) for relying on education as their measure of awareness, show that their results hold whether they use education, interest, or political activity; Luskin (1987) and Fiorina (1981) also test several different measures across a limited range of criterion variables.

334

Measures

appendix

his classic 1964 study of m a s s belief s y s t e m s , he m a d e clear that there is a cluster of variables - i n f o r m a t i o n , activity, s o p h i s t i c a t i o n , e d u c a t i o n , status as a m e m b e r of the political elite, and political interest - that are associated with constrained belief s y s t e m s ; of t h e s e , the latter t w o , and especially the last, seem most central to his a r g u m e n t . Yet in a paper written a r o u n d the same t i m e , C o n verse (1962) e m p l o y s m e d i a u s a g e , a variable given little play in his belief syst e m s paper, as the m e a s u r e of the extent to which individuals are both exposed to p a r t i s a n election c a m p a i g n s and e n d o w e d with the capacity to resist their appeals. In his 1975 a r t i c l e in the Handbook of Political Science, C o n v e r s e further r e p o r t s that " a f t e r a long and rather u n r e w a r d i n g s e a r c h " he found that a m e a sure of " p a r t i s a n political i n v o l v e m e n t " functions as the most i m p o r t a n t determ i n a n t of attitude stability, t h o u g h not necessarily of other p r o p e r t i e s of welld e v e l o p e d belief s y s t e m s (p. 104). He reiterates this position in 1980 in c o m m e n t i n g on a paper that had used e d u c a t i o n as its control v a r i a b l e , here referring to t h e key variable m o r e simply as " p o l i t i c a l i n v o l v e m e n t . " E v e n in the absence of empirical evidence on c o m p a r a t i v e p e r f o r m a n c e of the v a r i o u s possible m e a s u r e s of a w a r e n e s s , o n e w o u l d find s o m e operational m e a sures m o r e attractive than others. For e x a m p l e , p e o p l e w h o d e v e l o p the abstract learning skills that are supposedly i m p a r t e d by formal schooling w o u l d s e e m likely, by that account, to be m o r e adept at the c o m p r e h e n s i o n of political ideas. Yet e d u c a t i o n is only m o d e r a t e l y correlated with political interest and m e d i a exp o s u r e . This suggests that m a n y e d u c a t e d p e o p l e , although having the cognitive skill n e c e s s a r y to d e v e l o p political a w a r e n e s s , n o n e t h e l e s s lack the interest or m o t i v a t i o n to k e e p abreast of political events ( L u s k i n , 1990). S t a n d a r d m e a s u r e s of m e d i a exposure w o u l d appear, at first g l a n c e , excellent c a n d i d a t e s for m e a s u r i n g a w a r e n e s s of political ideas. T h e y h a v e , however, important w e a k n e s s e s . O n e is that, as Price and I have argued (1990), it is p r o b ably n e c e s s a r y to distinguish between exposure to " l o w - b r o w " political m e d i a A c t i o n N e w s on TV, People m a g a z i n e , talk show r a d i o , local w e e k l y n e w s p a pers - and exposure to " h i g h - b r o w " m e d i a , such as the TV network n e w s , N a tional Public R a d i o , and the Wall Street Journal. O n l y the latter c a r r y the rich diet of national and international n e w s n e c e s s a r y to create political a w a r e n e s s . Yet the m e a s u r e s of m e d i a exposure on most existing surveys do not p e r m i t o n e to distinguish b e t w e e n low-brow and high-brow m e d i a . S e c o n d , even if s e p a r a t e m e a s u r e s were available, it would be difficult to k n o w how to calibrate t h e m ; intermittent attention to N a t i o n a l Public R a d i o m i g h t , for e x a m p l e , c o n t r i b u t e m o r e to political awareness than avid attention to local TV news. T h i r d , m e a sures of self-reported m e d i a exposure suffer from subjective differences in selfrating s t a n d a r d s , and also s o c i a l - d e s i r a b i l i t y - i n d u c e d e x a g g e r a t i o n . For e x a m p l e , P r i c e and I (1990) found that, according to survey self-reports, 40 percent of the A m e r i c a n public listens to N a t i o n a l Public R a d i o several t i m e s a w e e k , a level that is both implausibly high and a b o v e N P R ' s own internal estim a t e s by a factor of about 10. In a n o t h e r e x a m p l e , s o m e r e s p o n d e n t s , h a v i n g claimed to read the Wall Street Journal, responded to a follow-up that they nor-

Measures

appendix

335

mally read the p a p e r seven d a y s a w e e k . This sort of m i s r e p o r t i n g , even if it occurs fairly infrequently, can wreak havoc with m o d e l s that d e p e n d on t h e ability to d i s c r i m i n a t e accurately at the top levels of political awareness. A n d finally, m e d i a e x p o s u r e , even if well m e a s u r e d , is still only a m e a s u r e of e x p o s u r e to politics rather t h a n , as required by the m o d e l , propensity for actual reception of i n f o r m a t i o n . Political p a r t i c i p a t i o n also h a s serious w e a k n e s s e s as a m e a s u r e of a w a r e n e s s . For e x a m p l e , a city m a i n t e n a n c e w o r k e r w h o must c o n t r i b u t e w o r k or m o n e y to the p a r t y m a c h i n e in o r d e r to k e e p his j o b w o u l d be indistinguishable on most political p a r t i c i p a t i o n m e a s u r e s from an activist in an issue-oriented political c l u b , even t h o u g h their differences in political a w a r e n e s s (as defined here) could be very great. It i s , moreover, easily possible for a person to achieve very high levels of political a w a r e n e s s without ever giving m o n e y to c a n d i d a t e s , w o r k i n g for p a r t i e s , or o t h e r w i s e p a r t i c i p a t i n g in politics. T h i s b r i n g s us to neutral factual k n o w l e d g e about politics, a type of m e a s u r e that, to a greater extent than any of the o t h e r s , captures political l e a r n i n g that h a s actually o c c u r r e d - political ideas that the individual h a s e n c o u n t e r e d , u n d e r s t o o d , and stored in his h e a d . T h i s is exactly w h a t we want to be m e a s u r i n g . Factual k n o w l e d g e is preferable on other theoretical g r o u n d s . A l o n e a m o n g the possible m e a s u r e s of a w a r e n e s s , tests of political k n o w l e d g e are relatively i m m u n e to a social desirability r e s p o n s e set; that is, individuals cannot overstate their levels of information h o l d i n g b e c a u s e they perceive that it is socially d e sirable to a p p e a r politically a w a r e . Tests of factual k n o w l e d g e are also relatively i m m u n e to r e s p o n s e effects, such as B i s h o p et al.'s (1984) d e m o n s t r a t i o n that expressed levels of political interest are readily affected by the context in which interest q u e s t i o n s are a s k e d . Finally, unlike m e d i a exposure and political interest, tests of k n o w l e d g e do not require r e s p o n d e n t s to e s t i m a t e subjective behaviors or inner states; they either pass or fail p a r t i c u l a r tests. In an a t t e m p t to settle the m e a s u r e m e n t p r o b l e m empirically, I tested t h e performance of several a l t e r n a t i v e m e a s u r e s of political awareness in predicting a variety of relevant criterion v a r i a b l e s (Zaller, 1986, 1990). T h e a w a r e n e s s m e a sures were e d u c a t i o n , m e d i a e x p o s u r e , p a r t i c i p a t i o n in politics ( d o n a t i n g money, attending s p e e c h e s , e t c . ) , interest in politics, and political k n o w l e d g e . T h e criterion variables included levels of o b s e r v e d attitude stability, attitude consistency, and degree of c o r r e l a t i o n b e t w e e n vote choice and party attachm e n t . P r i c e and I (1990) also e x a m i n e d the c o m p a r a t i v e ability of political k n o w l e d g e , e d u c a t i o n , and m e d i a exposure to specify individual-level reception probabilities for p r o m i n e n t n e w s stories; we also tested the c o m p a r a t i v e perform a n c e of k n o w l e d g e and m e d i a use in a r e c e p t i o n - a c c e p t a n c e m o d e l of attitude c h a n g e . T h e results of these tests were u n a m b i g u o u s : In each individual test, neutral factual information w a s always a r e s p e c t a b l e c o m p e t i t o r and almost 2

2 Individuals can attempt to guess correct answers, but a guessing penalty can be introduced to correct for this. I have experimented with guessing penalties and found that they do not improve the performance of information scales.

336

Measures

appendix

Measures

always the leading performer. No other m e a s u r e gave an overall p e r f o r m a n c e that w a s nearly as g o o d . T h e s e results also agree with the s o m e w h a t n a r r o w e r r a n g e of tests c o n d u c t e d by Luskin (1987). Accordingly, neutral factual information is, on both theoretical and empirical g r o u n d s , the preferred m e a s u r e of political a w a r e n e s s and is used as such t h r o u g h o u t this b o o k . In t h e cases in which an insufficient n u m b e r of k n o w l e d g e i t e m s is available, they are s u p p l e m e n t e d by m e a s u r e s of interest and e d u c a t i o n , as d e s c r i b e d below.

A s u b s e q u e n t inquiry c o a u t h o r e d with Vincent P r i c e ( P r i c e and Zaller, 1990) reached a similar c o n c l u s i o n . F r o m necessity, this b o o k relies almost exclusively on g e n e r a l - p u r p o s e rather than domain-specific m e a s u r e s of political awareness. A l t h o u g h domain-specific m e a s u r e s w o u l d be preferable, the loss from not having t h e m is apparently slight.

Types Issues

in

the

measurement

of political

information

G r a n t i n g that political k n o w l e d g e is the best available m e a s u r e of political a w a r e n e s s , there remain several questions about how k n o w l e d g e itself should best be m e a s u r e d . O n e that has attracted interest is w h e t h e r information is best m e a s u r e d globally by m e a n s of general tests of political k n o w l e d g e , or w h e t h e r m e a s u r e s must be specific to the p a r t i c u l a r d o m a i n in which information effects are e x p e c t e d (Iyengar, 1990). F r o m a theoretical p e r s p e c t i v e , awareness clearly s h o u l d , to the extent p o s sible, be m e a s u r e d separately within each d o m a i n for w h i c h effects are exp e c t e d . This follows from a x i o m A l , which claims that reception d e p e n d s not on global a w a r e n e s s , but on attention to p a r t i c u l a r issues. Therefore, we should expect to find stronger a w a r e n e s s effects for domain-specific m e a s u r e s of inform a t i o n than for global m e a s u r e s . T h e e v i d e n c e s u p p o r t s this expectation (Iyengar, 1990; Zaller, 1986; M c G r a w and Pinney, 1990). T h e r e i s , however, a serious difficulty in p r o d u c i n g d o m a i n specific effects. T h e difficulty is that, as a practical matter, p e o p l e w h o pay attention to o n e facet of politics tend to pay attention to other facets as well. As a result, it is difficult to build domain-specific scales that exhibit a satisfactory d e g r e e of discriminant validity. Iyengar (1990) m a n a g e s to obtain such validity, but only at the cost of e l i m i n a t i n g many face valid items that were t o o highly c o r r e l a t e d with one a n o t h e r across d o m a i n s ; these exclusions significantly red u c e the reliability of his scales. U s i n g the s a m e d a t a s e t , I built d o m a i n - s p e c i f i c scales that included all face-valid items that were available (Zaller, 1986); w h e n I did s o , I found that domain-specific k n o w l e d g e scales c o n c e r n i n g e c o n o m i c s , foreign policy, and minority relations were intercorrelated in the r a n g e of .70 (without c o r r e c t i n g for scale unreliabilities). T h e s e domain-specific k n o w l e d g e scales outperformed a general k n o w l e d g e scale in predicting relevant criterion v a r i a b l e s (such as attitude stability), but never by m a r g i n s that a p p r o a c h e d statistical or substantive significance. I therefore c o n c l u d e d It is clear that the effects of political information on public opinion are, to some extent, domain specific. . . . At the same time, the superiority of domain specific measures of information is both modest and uneven. . . . Together, these results suggest that political information is a relatively general trait that can be effectively measured with a generalpurpose information scale. (Zaller, 1986)

337

appendix

of information

tests

T h e items available as m e a s u r e s of political information vary greatly in number, quality, and format over the twenty-four I C P S R d a t a s e t s that have been used in this b o o k . In a handful of c a s e s , notably the N E S studies of 1986 and 1988 and the Brazilian public opinion survey, there is a g o o d selection of items which had been consciously d e s i g n e d as tests of factual k n o w l e d g e . In a few other c a s e s , most seriously the 1978 N E S congressional study and the 1984 " r o l l i n g t h u n d e r " survey of presidential p r i m a r i e s and the 1990 N E S S e n a t e study, there were few or no such i t e m s . A n d in the majority of c a s e s , there w a s a handful of g o o d information t e s t s , but t o o few to p r o d u c e highly reliable a w a r e n e s s scales, by w h i c h I m e a n scales with alpha reliabilities of 0 . 8 5 or h i g h e r . H e n c e it w a s n e c e s s a r y to locate i t e m s that, t h o u g h not direct m e a s u r e s of political a w a r e n e s s , could be c o n v e r t e d to this u s e . In the course of this b o o k , I have m a d e extensive use of t h e following three types of information tests: 3

4

1. T h e ability to m a k e " c o r r e c t " c o m p a r a t i v e p l a c e m e n t s of g r o u p s and candidates on issue d i m e n s i o n s , which I call " l o c a t i o n t e s t s . " S i n c e 1968, C P S / N E S surveys have regularly asked respondents to place t h e m s e l v e s on an issue d i m e n s i o n and also to place certain political objects - such as t h e presidential n o m i n e e s , the p a r t i e s , and liberals and c o n s e r v a t i v e s - on these s a m e scales. T h u s , for e x a m p l e , a respondent might have been asked to place herself on a seven-point defense-spending scale ( w h e r e 1 is the most a n t i s p e n d i n g position and 7 is the most p r o - s p e n d i n g p o s i t i o n ) , and then asked to place 1988 presidential c a n d i d a t e s G e o r g e B u s h and Michael D u k a k i s on the s a m e scale. It is very difficult to say exactly w h e r e B u s h ' s " t r u e " position on this scale m i g h t b e , but it can be confidently asserted that Bush should be located to the pro3 The 1984 survey contains measures of citizen awareness of various presidential contenders, but these are not well suited to my purposes. The difficulty is that such measures, as Bartels (1988) has shown, register higher levels of awareness over the course of the primary campaign as people learned more about the candidates. This is a difficulty because my theory holds that people who are high on habitual awareness will react differently to the fresh information they encounter than will people who are low on habitual awareness. Hence it is necessary to measure people's habitual awareness, and measures that are reactive to the current campaign cannot do this. 4 A three- or four-item scale of the type used in many opinion studies will normally have a reliability in the range of .65. Such a scale can easily fail to detect mild nonmonotonicity of the type shown in Figure 8.2 or subtle features of attitude change curves, as depicted in Figure 8.5. For the purposes of this book, much lengthier scales are needed.

338

Measures

appendix

s p e n d i n g side of D u k a k i s . H e n c e a respondent w h o locates Bush to the right of D u k a k i s on this scale can be c o u n t e d as having " c o r r e c t l y " a n s w e r e d a test of political k n o w l e d g e . T h e analyst must use j u d g m e n t in c o n v e r t i n g location items into k n o w l e d g e tests since, for s o m e i t e m s , it is impossible to say that one c a n d i d a t e or party differs from the other. For e x a m p l e , it is not at all o b v i o u s w h e t h e r J i m m y C a r t e r or Gerald Ford w a s farther to the left on w o m e n ' s r i g h t s , which m a k e s this issue unsuitable for use in m e a s u r i n g political k n o w l e d g e . But in m a n y c a s e s , c o r r e c t relative p l a c e m e n t s are o b v i o u s to any informed o b s e r v e r of American politics. 5

2. Willingness to rate c e r t a i n political figures on " f e e l i n g t h e r m o m e t e r s , " which may b e referred t o a s " r e c o g n i t i o n i t e m s . " N E S surveys regularly ask r e s p o n d e n t s to rate v a r i o u s political figures - including p r e s i d e n t s , s e n a t o r s , and political n o t a b l e s (such as Ralph Nader) - on 100-point " f e e l i n g t h e r m o m e t e r s . " R e s p o n d e n t s a r e , however, offered the o p t i o n of saying they are unfamiliar with the given political figure, and t h o s e taking this option may be c o u n t e d as having failed a k n o w l e d g e test. It is c o m m o n for researchers to use these items as k n o w l e d g e tests ( B a r t e l s , 1988; M a n n and Wolfinger, 1980). N o n e t h e l e s s , b e c a u s e this type of k n o w l e d g e test is highly susceptible to a social desirability r e s p o n s e set, I use it only sparingly and only w h e n better a l t e r n a t i v e s are unavailable. 3 . M a n y C P S / N E S surveys require i n t e r v i e w e r s , upon c o m p l e t i n g e a c h interv i e w , to rate each r e s p o n d e n t ' s apparent level of political information on a fivepoint scale. I have extensively analyzed the p e r f o r m a n c e of these rating scales and c o n c l u d e d that, at least in surveys involving face-to-face i n t e r v i e w s and c o n s i d e r a b l e political c o n t e n t , they perform extremely well (Zaller, 1985). In particular, a single five-point interviewer rating scale performs about as well as a scale c o n s t r u c t e d from 10 to 15 direct k n o w l e d g e tests, w h e r e the m e a s u r e of p e r f o r m a n c e is the ability to predict relevant c r i t e r i a . 6

A fear in relying upon such interviewer ratings is that they might be s y s t e m atically biased in favor of higher-status p e r s o n s , notably whites and m a l e s . H o w ever, I c h e c k e d carefully for e v i d e n c e of such bias and w a s able to find n o n e (Zaller, 1985). 5 Prior to 1970, CPS surveys sometimes directly asked respondents which party was more likely to favor a certain policy (e.g., Medicare for the elderly). 1 found that, in all years in which this type of item was asked, it performed poorly as a knowledge test in the sense that item reliabilities were relatively low. I suspect this is because these items are one-shot items, whereas the comparative location items depend on three responses (self-rating, and ratings of each of two objects). The multiple-step measures provide more opportunities to screen out guessers. In general, I discarded items with low reliability, no matter their face validity, unless the low reliability was due to a strong skew. 6 This qualification is important. One would not expect interviewer ratings of respondent awareness to work unless the interviewer had ample opportunity to observe the respondent deal with political matters.

Measures

appendix

339

M o s t of the a w a r e n e s s scales used in this b o o k have been c o n s t r u c t e d from these three types of items - especially the c o m p a r a t i v e location tests - and from m o r e s t a n d a r d , direct tests of k n o w l e d g e (such a s , W h i c h party controls t h e H o u s e of R e p r e s e n t a t i v e s ? W h a t is the t e r m of office of a U . S . senator?). It is natural to w o n d e r w h e t h e r " c a t c h - a l l " scales of the type I use perform as well as scales c o n s t r u c t e d from m o r e c o n v e n t i o n a l k n o w l e d g e tests. T h e 1985 N E S pilot study affords an o p p o r t u n i t y to investigate this question. It c a r r i e s s o m e twenty-seven direct k n o w l e d g e tests, that is, questions w h i c h directly asked r e s p o n d e n t s to supply the a n s w e r to a factual question about politics. W h e n a scale c o n s t r u c t e d from these twenty-seven items w a s pitted against one of my catch-all scales, there w a s essentially no difference in their ability to predict three criteria: attitude stability, attitude consistency, and issue salience (Zaller, 1 9 8 6 ) . 7

Coding

conventions

In c o n v e r t i n g a variety of i t e m s to tests of political i n f o r m a t i o n , I h a v e , with exceptions explicitly noted below, adhered to t h e following coding c o n v e n t i o n s : 1. Each discrete test of information has been given equal weight. In cases in which a single item could capture multiple bits of information, each bit was counted separately. In 1966, for example, respondents were asked to name as many Supreme Court justices as they could; I awarded one point for each justice they could name (up to four names). For another example, the interviewer rating scale has five points; I awarded respondents one point for each level above the lowest level. 2. In keeping with the previous point, I created simple additive scales, except in a few cases as noted. 3. Persons who failed to give a correct answer because they responded "Don't know" were counted as having given an incorrect response. This includes cases in which respondents were not asked the comparative location of candidates on an issue scale because they had no opinion on the issue itself. 4. Persons with missing data were not eliminated unless more than two-thirds of their responses were missing. Instead, these persons were assigned the average score for the items for which data were available. 8

9

All of the awareness scales I u s e , except the 1951 s c a l e , the 1978 H o u s e e l e c tions s c a l e , and the 1990 S e n a t e e l e c t i o n s s c a l e , have alpha reliabilities in the 7 The catch-all scale in this case was based on items culled from the 1984 NES study. It had twentytwo items and an alpha reliability of .87. The twenty-seven-item scale had a reliability of .89. 8 I experimented with use of principal components analysis to extract the most important common factor from a set of items. Although this made little difference for the results I obtained, I was reluctant to use the technique because it gave small weights to items which, although strongly skewed, had high discriminating power. Technically, the problem is that principal components assume a linear relationship between items and the underlying awareness factor, and this assumption is egregiously false for skewed items, as shown in Zaller, 1985. I regarded skewed items as strong items, because they isolated respondents at the extremes of the awareness continuum; but principal components regarded them as weak items, because they appeared to explain little variance in the underlying construct. 9 In the 1968 survey, respondents were asked to rate candidates even on issues on which they had no opinion. It turned out that scarcely any people who had no opinion on an issue were willing to estimate the positions of others on that issue.

340

Measures

appendix

r a n g e from about .80 to . 9 0 , with most having a l p h a reliabilities a r o u n d . 8 5 . T h e three exceptions have reliabilities in the r a n g e of . 6 0 to . 7 0 . S i n c e , however, the a s s u m p t i o n s u n d e r l y i n g the use of the alpha s t a t i s t i c , especially the a s s u m p t i o n of equal item difficulty, a r e not met in the d a t a I u s e , and since other m e a s u r e s of reliability are not readily available, there s e e m s no point in providing reliabilities for each scale. O t h e r detailed information about the m e a s u r e s of a w a r e ness I have used is given below.

Scale

appendix

341

(V109); which party is c o n s e r v a t i v e ( V 3 6 1 ) ; congressional c a n d i d a t e s ' n a m e s and i n c u m b e n c y status (three i t e m s , V 3 8 6 , V 3 8 7 ) ; majority in C o n g r e s s ( V 3 0 5 , V 3 0 6 ) ; up to two points for expression of interest in foreign affairs ( V 1 0 2 ) . A l s o t w o interviewer rating s c a l e s , V 2 6 9 on preelection and V531 on p o s t e l e c t i o n , w h i c h were averaged to yield a m a x i m u m of four points; and two tests of ability to m a k e proper c o m p a r a t i v e p l a c e m e n t s of politicians on issue s c a l e s , as follows: J o h n s o n and Wallace o n V i e t n a m ( V 4 6 6 and V 4 6 9 ) ; H u m p h r e y and W a l lace on urban unrest (V461 and V 4 6 3 ) . T h e u n s t a n d a r d i z e d m e a s u r e h a s a m e a n of 11.39 and a s t a n d a r d deviation of 4 . 4 1 .

construction

Variable n u m b e r s refer to the relevant I C P S R c o d e b o o k . (The S P S S x c o d e for building these scales is available via bitnet to any scholar wishing to use it.) 1951 CPS terviewer (col. 6 5 ) , mentioned

Measures

"minor election study." A w a r e n e s s is m e a s u r e d by a three-point inrating of the r e s p o n d e n t ' s apparent understanding of w o r l d affairs plus a one-point m e a s u r e indicating w h e t h e r the person specifically D e a n Acheson in the course of the i n t e r v i e w (col. 6 7 ) .

1956 CPS election study. T h e r e were an insufficient n u m b e r of items to build an information scale from the 1956 study alone. H e n c e I used twelve tests scattered across all three waves of the 1 9 5 6 - 5 8 - 6 0 p a n e l , as follows: V 1 4 2 , V 1 4 5 , V 3 1 7 , V 3 1 8 , V 3 9 4 (two i t e m s ) , V 3 9 5 , V 6 0 8 , V 6 0 9 , V 6 1 1 , V 6 1 2 , V 6 1 3 , V 6 1 5 , V798.' T h e u n s t a n d a r d i z e d m e a s u r e h a s a m e a n of 7 . 8 and a standard deviation of 3 . 0 8 .

1970 CPS Awareness scale. An eighteen-point s c a l e , as follows: Attention to V i e t n a m ( V 4 0 ) ; which p a r t y is m o r e c o n s e r v a t i v e ( V 1 7 7 ) ; congressional candidate ( V 2 0 3 ) ; p e r c e n t a g e of tax dollar to defense ( V 2 5 6 ) ; w h o can c h a n g e law ( V 2 5 8 ) ; n u m b e r of allowable t e r m s as president ( V 2 5 9 ) ; t e r m of U . S . senator ( V 2 6 0 ) ; t e r m of congressional representative ( V 2 6 1 ) ; interviewer rating of res p o n d e n t ' s information (up to four p o i n t s , V 3 9 6 ) . T h e r e were four c o m p a r a t i v e location tests, as follows: D e m o c r a t s and R e p u b l i c a n s on V i e t n a m ( V 9 3 and V 9 4 ) ; M u s k i e and W a l l a c e o n aid t o m i n o r i t i e s ( V I 1 0 and V I 1 1 ) ; D e m o c r a t s and Nixon on inflation ( V I 1 4 and V I 1 6 ) ; and M u s k i e and W a l l a c e on criminal rights ( V 1 2 4 and V 1 2 5 ) ; also t w o 1-item location tests, Wallace on urban unrest and V i e t n a m ( V 9 0 , V 9 7 ) . T h e u n s t a n d a r d i z e d m e a s u r e has a m e a n of 9 . 5 4 and a s t a n d a r d deviation of 4 . 4 4 . 1972-74-76 NES panel awareness scale.

1964 CPS election study. A sixteen-point s c a l e , as follows: L . B . J . ' s and G o l d w a t e r ' s h o m e state and religion ( V 2 9 4 to V 2 9 7 ) ; which party is m o r e c o n s e r vative ( V 3 0 2 ) ; majority in C o n g r e s s ( V 3 0 5 , V 3 0 6 ) ; congressional c a n d i d a t e s ' n a m e s , i n c u m b e n c y status (three i t e m s , V 3 0 9 , V 3 1 2 ) ; which party favors g o v e r n m e n t utilities ( V 3 4 6 ) ; C h i n a ' s form of g o v e r n m e n t and U . N . status (two i t e m s , V 3 5 2 ) ; form of C u b a ' s g o v e r n m e n t ( V 3 5 4 ) ; k n o w l e d g e of 1964 Civil Rights Act ( V 4 0 6 , V 4 0 7 ) . T h e unstandardized m e a s u r e has a m e a n of 10.4 and a s t a n d a r d deviation of 4 . 1 3 . - .

T h i s scale is described in Zaller,

1990.

1966 CPS election study. A fifteen-point scale, as follows: congressional cand i d a t e s ' n a m e s and i n c u m b e n c y status (three i t e m s , V 9 0 , V 9 2 ) ; majority in C o n g r e s s ( V 1 0 0 , V 1 0 1 ) ; n a m e s of U . S . S u p r e m e C o u r t j u s t i c e s (up to four p o i n t s , V 1 6 7 ) ; interviewer rating of r e s p o n d e n t ' s information level (up to four p o i n t s , V 2 4 2 ) . Since this scale w a s used to m e a s u r e exposure to n e w s about V i e t n a m , I a w a r d e d up to t w o points for expression of interest in foreign affairs ( V 6 5 ) . T h e unstandardized m e a s u r e has a m e a n of 7 . 4 8 and a s t a n d a r d d e v i a t i o n of 4 . 7 .

1978 NES election awareness scale. An unusually diverse scale, as follows: M e a n i n g of liberalism ( V 4 1 0 ) , m e a n i n g of c o n s e r v a t i v i s m ( V 4 1 3 ) , party control of H o u s e of R e p r e s e n t a t i v e s before and after e l e c t i o n ( V 5 0 0 , V 5 0 1 ) . T h r e e recognition i t e m s , R o n a l d R e a g a n ( V 1 3 9 ) , J e r r y B r o w n ( V 1 4 1 ) , and Ralph N a d e r ( V 1 4 4 ) . Four c o m p a r a t i v e location items: D e m o c r a t s and R e p u b l i c a n s o n j o b g u a r a n t e e s ( V 3 5 9 , V 3 6 0 ) , o n rights o f the accused ( V 3 7 5 , V 3 7 6 ) , o n g o v e r n ment medical insurance ( V 3 8 3 , V 3 8 4 ) , and o n liberalism and c o n s e r v a t i v i s m ( V 4 0 1 , V 4 0 2 ) . Political interest, V 4 3 ; and interviewer rated i n f o r m a t i o n , V 6 3 6 . B e c a u s e I w a s unsure w h e t h e r these i t e m s d e s e r v e d equal w e i g h t i n g , I c o m b i n e d t h e m into a scale by m e a n s of weights o b t a i n e d from a principal c o m p o n e n t s analysis. B e c a u s e the political interest variable is correlated with the intensity of p a r t i c u l a r H o u s e r a c e s , as m e a s u r e d by the media-intensity variable (see text of C h a p t e r 10), I p u r g e d it of this influence before a d d i n g it to the a w a r e n e s s scale. T h e u n s t a n d a r d i z e d m e a s u r e has a m e a n of 7 . 4 and a s t a n d a r d deviation of 3 . 3 . (In the analysis of likes/dislikes in C h a p t e r 10, the interest item is used as a s e p arate variable.)

1968 CPS election study. A seventeen-point s c a l e , as follows: C h i n a ' s form of g o v e r n m e n t and U . N . status (two i t e m s , V 1 0 7 ) ; form o f C u b a ' s g o v e r n m e n t

1980 NES election awareness scale. A twenty-five-point s c a l e , as follows: ten c o m p a r a t i v e location items: C a r t e r and R e a g a n o n ideology ( V 2 6 8 and V 2 6 9 ) ;

342

Measures

appendix

K e n n e d y and Connally on ideology ( V 2 7 0 and V 2 7 1 ) ; two p a r t i e s on ideology ( V 2 7 8 and V 2 7 9 ) ; C a r t e r and Reagan on defense spending ( V 2 8 2 and V 2 8 3 ) ; p a r t i e s on g o v e r n m e n t services ( V 2 9 6 and V 2 9 7 ) ; p a r t i e s on j o b s and inflation trade-off ( V 3 0 6 and V 3 0 7 ) ; C a r t e r and Reagan on a b o r t i o n (V312 and V 3 1 3 ) ; C a r t e r and Reagan on c o o p e r a t i o n with Russia ( V 1 0 7 9 and V 1 0 8 0 ) ; p a r t i e s on j o b g u a r a n t e e s ( V I 1 2 1 and V I 1 2 2 ) ; Reagan and K e n n e d y on aid to m i n o r i t i e s ( V 1 0 6 4 and V 1 0 6 6 ) . T h e s e additional items: h e a r d poll results ( V 2 1 9 ) , expresses opinion on tax cut proposal ( V 3 2 2 ) , n a m e and p a r t y of congressional c a n d i d a t e s (four i t e m s , V 8 2 6 , V 8 2 9 ) , n a m e s of S e n a t e candidates (four i t e m s , w h e r e relevant, V 8 3 7 and V 8 4 0 ; respondents were not penalized for living in a state without a Senate r a c e ) , p a r t y that s u p p o r t s stronger federal g o v e r n m e n t ( V I 1 3 1 ) . T h e r e were t w o interviewer rating s c a l e s , V 7 2 6 o n p r e e l e c t i o n and V I 1 8 6 on p o s t e l e c t i o n , which were averaged to yield a m a x i m u m of four points. T h e u n s t a n d a r d i z e d m e a s u r e has a m e a n of 9 . 6 and a standard deviation of 5 . 7 . 1982 NES election awareness scale. A twenty-five-point scale, as follows: T h r e e r e c o g n i t i o n items: J o h n A n d e r s o n , B u s h , and M o n d a l e ( V 1 2 5 , V 1 3 4 , V 1 3 5 ) . Five c o m p a r a t i v e location items: D e m o c r a t s and R e p u b l i c a n s on ideolo g y scale ( V 4 0 4 , V 4 0 5 ) , o n defense spending scale ( V 4 0 9 , V 4 1 0 ) , o n j o b guarantees ( V 4 2 7 , V 4 2 8 ) , and o n g o v e r n m e n t services ( V 4 4 5 , V 4 4 6 ) ; liberals and c o n s e r v a t i v e s on aid to rr.inorities ( V 4 2 2 and V 4 2 3 ) . A l s o t w o single objectlocation tests: Ronald R e i g a n located right of c e n t e r on ideology ( V 3 9 4 ) and defense spending ( V 4 0 8 ) . N a m e and p a r t y of congressional c a n d i d a t e s (four i t e m s , V 1 0 2 , V 1 0 6 ) , n a m e and party of Senate c a n d i d a t e s (four i t e m s , w h e r e relevant, V I 1 5 , V I 1 8 ) . Party control o f H o u s e and S e n a t e , party that elected most n e w m e m b e r s to H o u s e ( V 5 2 3 to V 5 2 5 ) . I n t e r v i e w e r rating of r e s p o n d e n t information (up to four p o i n t s , V 7 6 8 ) . T h e u n s t a n d a r d i z e d m e a s u r e has a m e a n of 11.7 and a standard deviation of 6 . 3 . 1984 NES Continuous Monitoring Study awareness scale. A twenty-four-point s c a l e , a s follows: V 1 3 6 (three i t e m s ) , V 4 0 7 (four i t e m s ) , V 6 1 4 , V 6 2 2 , V 6 2 6 , 805 ( t w o i t e m s ) , V 8 7 6 . Eleven recognition i t e m s , V 2 2 0 , V 2 2 2 t o V 2 3 1 . B e c a u s e I w a s unsure w h e t h e r these items d e s e r v e d equal w e i g h t i n g , I c o m bined t h e m into a scale by m e a n s of weights o b t a i n e d from a principal c o m p o nents analysis. 1984 NES election awareness scale. A twenty-four-point scale, as follows: T h e r e were eleven c o m p a r a t i v e location i t e m s , several of which required respondents to place objects t w o or three points apart: R e a g a n and M o n d a l e on ideolo g y ( V 3 7 1 and V 3 7 2 , at least two points a p a r t ) , g o v e r n m e n t services ( V 3 7 6 and V 3 7 7 , at least two points a p a r t ) , aid to m i n o r i t i e s ( V 3 8 3 and V 3 8 4 ) , U . S . inv o l v e m e n t in Central A m e r i c a ( V 3 8 9 and V 3 9 0 ) , defense spending ( V 3 9 6 and V 3 9 7 ) , c o o p e r a t i o n with Russia ( V 4 0 9 and V 4 1 0 ) , j o b g u a r a n t e e s ( V 4 1 5 and V 4 1 6 ) , R e a g a n and M o n d a l e on ideology (distance of three points required for correct answer, V1017 and V1018); liberals and c o n s e r v a t i v e s on g o v e r n m e n t

Measures

appendix

343

services ( V 1 0 3 4 and V 1 0 3 5 , three points a p a r t ) , involvement in Central A m e r ica ( V 1 0 4 5 and V 1 0 4 6 , three points a p a r t ) , j o b g u a r a n t e e s ( V 1 0 5 4 and V 1 0 5 5 , three points a p a r t ) . In addition: n a m e and p a r t y of congressional c a n d i d a t e s (four i t e m s , V 7 4 1 , V 7 4 5 ) , which party i s more c o n s e r v a t i v e ( V 8 7 5 ) , which p a r t y h a s m o r e m e m b e r s in H o u s e and S e n a t e before and after election ( V 1 0 0 6 to V 1 0 0 9 ) , r e c o g n i t i o n of Bush and F e r r a r o ( V 2 9 2 and V 3 0 4 ) . T h e r e w e r e t w o interviewer rating scales, V 7 1 3 on preelection and V I 1 1 2 on p o s t e l e c t i o n , which w e r e averaged to yield a m a x i m u m of four points. T h e u n s t a n d a r d i z e d m e a s u r e has a m e a n of 9.8 and a standard deviation of 5 . 9 . 1986 NES election and 1987 pilot awareness scale. A nineteen-point s c a l e , as follows: ability to n a m e political offices held by B u s h , Weinberger, R e h n q u i s t , Volker, O ' N e i l l , D o l e ( V 6 3 5 t o V 6 4 2 ) , n a m e s o f H o u s e c a n d i d a t e s ( V 2 0 2 , V 2 0 6 ) , two r e c o g n i t i o n t e s t s , H a r t and J a c k s o n ( V 2 4 2 , V 2 4 4 ) . Four location items: D e m o c r a t s and R e p u b l i c a n s on ideology ( V 7 3 0 and V 7 3 1 ) , on defense spending ( V 7 4 9 and V 7 5 0 ) , on involvement in Central A m e r i c a (V811 and V 8 1 2 ) , and on g o v e r n m e n t services (V831 and V 8 3 2 ) ; c o m p a r a t i v e r a n k i n g s on ideology, defense, and services had to be at least t w o points apart to get credit for a c o r r e c t answer. An additional point w a s a w a r d e d for placing Reagan on the c o n s e r v a t i v e side of the ideology scale, which is a n o n c o m p a r a t i v e location test ( V 7 2 3 ) . T h e u n s t a n d a r d i z e d m e a s u r e , as used in the analysis of attitude c h a n g e on Central A m e r i c a , h a s a m e a n of 7 . 9 and a standard deviation of 4 . 2 . 1989 NES pilot awareness scale. A twenty-two-point s c a l e , as follows: budget k n o w l e d g e ( V 8 3 5 ) , w h i c h p a r t y is m o r e c o n s e r v a t i v e ( V 5 6 5 6 ) , the political offices held by K e n n e d y , S c h u l t z , R e h n q u i s t , G o r b a c h e v , Thatcher, and Arafat ( V 5 8 2 7 to V 5 8 3 3 ) , p a r t y control of C o n g r e s s ( V 8 3 4 , V 8 3 5 ) , and n a m e s of congressional c a n d i d a t e s ( V 5 1 0 6 ) . Seven location items: c o m p a r a t i v e p l a c e m e n t s of the D e m o c r a t s and R e p u b l i c a n s on ideology ( V 4 2 1 and V 4 2 2 ) , g o v e r n m e n t services ( V 6 0 6 and V 6 0 7 ) , defense spending ( V 6 1 0 and V 6 1 1 ) , medical insurance ( V 6 2 0 and V 6 2 1 ) , j o b g u a r a n t e e s ( V 6 2 7 and V 6 2 8 ) , and c o o p e r a t i o n with R u s sia ( V 7 1 2 and V 7 1 3 ) ; in a d d i t i o n , a point w a s given for r a n k i n g D u k a k i s t w o or m o r e points to the left of Bush on ideology (V417 and V 4 1 8 ) . R e s p o n d e n t s could also get up to four points from the interviewer rating of respondent information ( V 1 2 4 1 ) . T h e u n s t a n d a r d i z e d m e a s u r e has a m e a n of 10.5 and a standard deviation of 5 . 7 . 1990 NES awareness scale. A fourteen-point scale consisting of tests of t h e ability to give the political office held by Q u a y l e , M i t c h e l l , R e h n q u i s t , Gorbachev, Thatcher, M a n d e l l a , and Foley ( V 8 3 5 to V 8 4 1 ) ; plus p a r t y control of C o n g r e s s ( V 8 4 2 , V 8 4 3 ) and interviewer rating of respondent k n o w l e d g e ( V 1 2 3 1 ) . T h e uns t a n d a r d i z e d m e a s u r e has a m e a n of 4 . 9 5 and a standard deviation of 2 . 7 4 . 1990 NES Senate study awareness scale. T h e 1990 S e n a t e study contained only five political k n o w l e d g e i t e m s - tests of r e s p o n d e n t ' s ability to locate G e o r g e

344

Measures

appendix

B u s h , the D e m o c r a t i c Party, and the Republican Party on a seven-point l i b e r a l c o n s e r v a t i v e scale ( V 5 1 7 , V 5 3 3 , V 5 3 4 ) , and the n a m e s of the c a n d i d a t e s in H o u s e e l e c t i o n s ( V 4 1 , V 4 5 ) . I c o m b i n e d these with tests of willingness to evaluate B u s h ' s j o b p e r f o r m a n c e (V36) and to state an opinion on aid to m i n o r ities ( V 5 3 7 ) . This scale had a mean of 3 . 0 4 and a standard deviation of 1.53. T h e purged interest item (see 1978 a w a r e n e s s scale) and the scale were s t a n d a r d i z e d , c o m b i n e d , and restandardized, thereby giving equal weight to each c o m p o n e n t in the final awareness scale.

Measures

345

appendix

separately scaled by m e a n s of principal c o m p o n e n t s analysis. T h e items in t h e d o m e s t i c policy cluster w e r e V 3 2 , V 3 8 , V 5 3 , V 5 9 ; the items i n t h e foreign policy cluster were V 3 5 , V 5 0 , V 6 8 , V 7 1 . 1964,

1966,

1968,

1970 CPS election studies.

See a p p e n d i x to C h a p t e r 9.

1978 NES election study. T h e items used to m e a s u r e respondent attitudes are V 3 5 7 , V 3 6 5 , V 3 7 3 , V 3 9 9 , V 4 4 3 . See footnote 8 in C h a p t e r 10 for further information.

POLITICAL

PREDISPOSITIONS 1982 NES election study.

W i t h the exception of t h e 1951 and 1956 s t u d i e s , my m e a s u r e s of political a w a r e n e s s are quite strong and essentially c o m p a r a b l e in all years. M e a s u r e m e n t of political predispositions is m u c h m o r e uneven a n d , on the w h o l e , m u c h less a d e q u a t e . Except in t w o studies - notably the 1987 N E S pilot study, which contains t h e H u r w i t z - P e f f l e y (1988) m e a s u r e of foreign policy p r e d i s p o s i t i o n s , and the 1989 pilot study, which contains an excellent selection of C o n o v e r - F e l d m a n equality items - I w a s always dealing with m e a s u r e s that I felt were m o r e or less i n a d e q u a t e to the tasks to which I put t h e m . A l t h o u g h m a n y of the surveys contained reasonably g o o d value m e a s u r e s , they did not contain value m e a s u r e s of the particular dimensior on which public opinion w a s u n d e r g o i n g c h a n g e . C h a p t e r 2 provides the theoretical justification for opportunistically using w h a t ever m e a s u r e s are available. T h e least a d e q u a t e studies were the 1951 and 1956 studies. T h e only m e a s u r e of general p a r t i s a n o r i e n t a t i o n in the 1951 study w a s recalled vote in the 1948 e l e c t i o n . A l t h o u g h this m e a s u r e served as the basis for the original s t a t e m e n t of the M i c h i g a n t h e o r y of p a r t y identification ( B e l k n a p and C a m p b e l l , 1 9 5 1 - 2 ) , the m e a s u r e leaves m u c h to be desired. T h e 1956 study c o n t a i n s a n u m b e r of g o o d preference q u e s t i o n s , but n o n e that s e e m well suited for capturing the predispositions that underlie racial attitudes. Since 1984, N E S surveys have c a r r i e d a generally excellent selection of value m e a s u r e s .

Coding

conventions

In building value scales, I m a d e full use of w h a t e v e r information w a s c o n t a i n e d in an item (that is, I did not recode to a smaller n u m b e r of c a t e g o r i e s ) . T h e items were c o m b i n e d into scales by m e a n s of weights obtained from principal c o m ponents a n a l y s i s , unless all items had the same r a n g e , in which c a s e simple additive indices were c r e a t e d .

Scale

construction

1956 CPS election study. T h e r e were t w o m a i n sets of i t e m s , o n e c o n c e r n i n g d o m e s t i c politics and the o t h e r foreign politics. T h e items in each cluster were

For use in diffusion of e c o n o m i c n e w s , three e c o n o m i c

policy items: j o b g u a r a n t e e s ( V 4 2 5 ) , g o v e r n m e n t services ( V 4 4 3 ) , regulation of business ( V 4 5 9 ) . For use with n u c l e a r freeze issue, t w o defense spending items (V317 and V 4 0 7 ) . 1984 NES

election

study.

Four items:

liberal-conservative

self-identification

( V 3 6 9 ) , g o v e r n m e n t services ( V 3 7 5 ) , minority aid ( V 3 8 2 ) , and j o b g u a r a n t e e s (V414). 1987 NES pilot study.

I t e m s from H u r w i t z - P e f f l e y foreign affairs battery c o m -

bined into h a w k - d o v e s c a l e , a s follows: V 2 2 4 5 , V 5 2 3 3 , V 5 2 4 9 , V 5 2 5 1 , V 5 2 5 2 , V 5 2 5 3 , plus t w o additional items (Russia and Defense s p e n d i n g ) , V 7 4 2 and V 9 0 7 . T h e social welfare ideology scale consists of fourteen C o n o v e r - F e l d m a n items plus t w o m e a s u r e s of ideological self-designation, as follows: V 6 2 0 to V 6 2 2 , V 6 2 4 , V 6 2 6 , V 7 0 1 t o V 7 0 6 , V 2 1 7 6 , V 2 1 7 8 , V 2 1 7 9 , plus V 7 2 2 and V1010. T h e g o v e r n m e n t s p e n d i n g scale consisted o f V 7 1 6 - V 7 2 5 .