Cognitive Engagement and Citizen World Views* Samuel L. Popkin and Michael A. Dimock Department of Political Science University of California, San Diego La Jolla, CA 92093-0521 February, 1995 WORKING PAPER DRAFT AS OF 4/1/95 *Prepared for presentation at the PEGS Conference on Citizen Competence and the Design of Democratic Institutions, February 10-11, 1995, Washington, D.C. The purpose of this paper is to show that institutional knowledge structures the kinds of inferences which citizens make about the world. The way that citizens reason about politics, and the kinds of information they use to make political decisions are all affected by their understanding of political institutions. Persons with knowledge of political institutions have conceptual maps of the world that are less vague and uncertain and anxiety provoking. Without this knowledge people see economic and social change as more uncertain. Any discussion of citizen competence must recognize the importance of political knowledge in helping persons to evaluate politicians and policies and cope with social and economic change. The effects of limited knowledge of political institutions upon citizen world views are particularly strong in this country because Americans have so little knowledge about the way their own government works and about the institutions which govern their society. Building upon recent advances by John Zaller and Scott Keeter, we show that knowledge about government affects not just how well people respond to their leader's messages or how well they are able to identify and pursue their individual or collective interests, but how they react to international conflict, how they view the global economy and whether or not they vote. The lower the level of institutional knowledge about politics a person has, the more likely it is that that person will be fearful of immigration, international trade, and foreign involvement. The less a person knows about government, the more likely it is that the voter will judge representatives by their personal character and the less a voter knows, the less likely it is that he or she will vote. Recent work by Zaller and Keeter has revived interest in political knowledge of the civics textbook variety. Keeter, with Michael Delli Carpini, revived the moribund practice of evaluating the levels of basic knowledge among the citizenry and showing that civic literacy affects political participation, attitude stability, political tolerance and the ability to link preferences and choices. In short, Keeter and Delli Carpini argue that civic knowledge is necessary for good citizenship. Zaller's concern in his seminal work was the way that citizens respond to political debates and messages. While neither the levels nor the effects of civics knowledge were his ultimate concern he showed the importance of the concept of cognitive engagement as a determinant of political behavior and opinion formation. People who understand the basic facts of political institutions, who are cognitively engaged, are more attentive to and comprehending of communications from the political environment. A completely separate tradition of research following from Anthony Downs' work developed in response to the demonstrably low levels of information among the electorate. Downs' central insight is that citizens do not have much incentive to gather information about politics solely in order to improve their voting choice. They will rely on information shortcuts as substitutes for more complete information about parties, candidates and policies. Because citizens use shortcuts to obtain and evaluate information, they are able to store far more data about politics than measurements of their textbook knowledge would suggest. Shortcuts for obtaining information at low cost are numerous and include party identification, endorsements, and demographic characteristics. Given the limited information about government which most citizens possess, determining which issues will matter in any election is problematic. It is not the importance of a policy, nor even the extent to which parties or candidates differ on it, that determines when an issue will become central to voter decision making. What makes it central will be the information persons have about the issue and the manner in which their beliefs about how the world works connect an issue to their own life situation and to the office, candidate and party for which they are voting. Here, we concentrate on the role an understanding of political institutions plays when citizens decide whether to vote and evaluate candidates and policies. Following Zaller, we are calling knowledge of the institutions of government cognitive engagement to distinguish this type of engagement from affective or emotional engagement -- often labeled in reverse fashion as emotional alienation. Persons who do not know about institutions and how they work cannot incorporate these institutions in their thinking about news of the world. However, in contrast to Zaller's focus on political communication, our concern is with the ways that persons with different levels of institutional knowledge think about the world. This is a new and very different branch of the tree that is sprouting from Zaller's work. Because Zaller's concern is with response to elite messages, he concentrates upon the likelihood that a citizen will adopt a particular stance on an issue of political debate. He takes it literally as axiomatic -- the receptiveness axiom -- that cognitive engagement, knowledge about the basic structure of politics, is directly related to the probability of both receiving and comprehending a message. The less well informed either miss the message, uncritically accept it, or relate to the message according to partisan cues. The data that will be presented here show that people who do not know as much about the structure of political institutions, do not think like the more knowledgeable, only slower or a bit fuzzier. Neither do they uncritically accept what is placed before them. The less cognitively engaged have a different world view and view the world and politics differently. When people evaluate news and think about politics, their representations of the world are the foundation upon which they build. Their level of cognitive engagement has two clear implications for how they utilize information. First, we will show that the information affects the cues that will be salient and the kinds of information people will utilize to evaluate candidates and parties. People who process news with and without institutional familiarity follow stories differently. Citizens with low levels of institutional information are more likely to use assessments of personal character as substitutes for evaluations of a candidate's positions or party affiliation. Candidate- centered politics, and the emphasis on scandal, are both, in part, consequences of low levels of political information in the citizenry. They will also be less able to perceive differences between candidates and parties. Non-voting results from a lack of knowledge about what government is doing and where parties and candidates stand, not from a knowledgeable rejection of government or parties. Second, institutional information affects their levels of concern with new situations and social and economic change. The ability to include political institutions in thinking about the world is a central and key determinant of how people will view the world economically, socially, politically and militarily. The world is more out of control and frightening when citizens don't know about or don't understand institutions. They are, in general more afraid than those who understand something about political institutions. This tendency to experience the unfamiliar as uncertain and threatening also makes persons with low knowledge of political institutions more afraid of immigration and foreign investment in America and more concerned about foreign trade. They are also more likely to perceive danger in the world and to believe that force is necessary to resolve international conflicts. Thus, knowledge of institutions makes people feel more secure when they think about the way foreign people and events effect them. Ignorance, in short, is far from bliss. A world viewed without institutions is a world viewed fearfully, not with indifference. The citizen without some knowledge of the government institutions which seek to control conflict and change in the country and world is more anxious and uncertain about the future. Further, it is not the poor performance of political institutions as much as it is ignorance of the institutions that is the source of many current discontents. The Nature and Extent of Cognitive Engagement For more than 50 years, voting studies and public opinion polls have showed low levels of textbook political knowledge in the mass electorate. Indeed, the level of factual knowledge about the basic structure of government within the electorate, and the extent of information about specific legislation is so low, survey researchers are generally reluctant to ask too many factual questions for fear of embarrassing respondents, who might terminate the interview or become too flustered to answer other questions. To directly assess the changes in civics knowledge, which we call cognitive engagement, since the 1940s, Keeter and Delli Carpini conducted a national survey asking the same basic questions that were asked in the 1940s. They replicated questions testing knowledge of certain elementary facts, such as which party now controls the House, what the first ten amendments to the Constitution are called, the name of the vice-president, the definition of a presidential veto, and how much of a majority is required for the Senate and House to override a presidential veto. Overall, they found, "the level of public knowledge of some basic facts has remained remarkably stable." That the overall extent of cognitive engagement has remained stable despite increases in education emphasizes that specific knowledge of political institutions is conceptually different from education; moreover, as we shall see, there is a great deal of variance in cognitive engagement at every level of education. Fifty years ago, three-fourths of the electorate had not finished high school and only 10 percent had any college experience. Today three-fourths of the electorate has finished high school and nearly 40 percent has been to college. But despite this increase in education, voter turnout and factual knowledge about government and current political debates is at best only marginally higher. Clearly, knowledge of government and its institutions is not a simple function of education. People are no more likely than they were 50 years ago to know the name of their congressional representative. Despite all the publicity he received, Vice- President Dan Quayle was only marginally better known in 1989 than Richard Nixon was in 1952; 75 percent could name Quayle, while 69 percent had been able to name Nixon. Thanks to Zaller's efforts, the American National Election Studies now contain a large number of factual questions which can be used to develop scales of cognitive engagement. Table I shows the seven questions on that survey that we utilize throughout this paper as a cognitive engagement scale. Seven out of eight respondents could identify the position held by Dan Quayle but only one in four could identify the job of Tom Foley and only one in twelve could identify the job held by William Rehnquist. Between fifty and sixty percent could identify the majority party in the House and Senate, say who nominated judges or who decided whether a law was constitutional. These are easy straightforward questions, yet the average number of right answers among respondents was 3.5. Only 1 in six respondents got six or seven right answers and one out of three had two or less right answers. (Table 2) While the scores for respondents are correlated with education level, there is substantial variation at every level of education. Figure 1 shows the distribution of cognitive engagement scores by level of education. Only among college graduates is it the case that at least half of the respondents could answer five or more of the questions correctly. Knowledge of political institutions, though correlated with education, is not merely a surrogate for education. The type of institutional understanding that is the focus of our paper is distinct from education and has important influences on perceptions even when education is controlled for. [In all the multi- variate analyses of the effects of cognitive engagement, however, we will control for education, income, age and ideology to make even clearer the effect of institutional knowledge upon political worldviews.] Cognitive Engagement and Personal Character Reactions to the house banking scandal demonstrate how cognitive engagement affects whether persons evaluate their representatives in personal or political terms. The level of cognitive engagement was directly and dramatically related to both knowledge of the scandal and to attitudes towards the representatives involved. This scandal mattered most to people with the least knowledge about politics, and the least ability to judge candidates on other grounds. Personal character matters to principals when they must choose agents for jobs in which the agent will have a great deal of leeway. Personal character, as Aristotle noted, makes a speaker credible. "We believe good men more readily and fully than others; this is true generally whatever the question is, and absolutely true where exact certainty is impossible and opinions are divided." We care more about the character of a baby-sitter with whom we leave a baby than we do about the character of the person from whom we buy the child's clothing. Votes in Congress are based upon a hard-to-decipher mixture of compromises between ideal positions and practical realities. Watching candidates perform and make promises in situations replete with compromises and role playing, people will wonder whether a candidate's support for a cause was strategic or reflected a true commitment. Because of uncertainty, they will wonder whether the candidate is sincere about his or her concerns, whether there is congruence between avowed and actual feelings. Did the representative do his or her best in the smoke-filled rooms, and what will he or she do next time? Voters with institutional knowledge of politics can better sort through the posturing of candidates and use partisan and issue cues to evaluate incumbents. Voters less able to use these cues rely on estimates of personal character instead of on attitudes about parties and issues. In other words, reliance on personal character as a proxy for political character is related to uncertainty and uncertainty is related to a lack of understanding about politics. If this argument is correct, then the persons for whom the scandal mattered most should be those with the least information about government and therefore needed most to rely on personal cues about the candidates. The House Banking Scandal contributed to the largest turnover of representatives since 1948. After a six year period in which 90% of all representatives had been reelected, over 25% in 1992 either retired, were defeated in primaries or lost to an opponent in November. The Congressional bank, which allowed representatives to get interest free loans for short periods of time instead of bouncing checks, involved no taxpayer money; the interest-free loans came from other members of Congress. Yet, in contrast to the Savings and Loan bailout which cost hundreds of billions of people's tax dollars but which was virtually ignored by them, the bounced checks became an issue which captured public attention and talk shows for months. Why should the house bank became such a big issue? As Rush Limbaugh succinctly stated, it was an easily comprehended issue of personal character: " I mean, the public is angry as they can be about that and there's one good reason for it. This is easy to understand. This is something they can't do. This is the epitome of arrogance. . ." However, not everyone who understood the scandal thought it mattered. While more than 90% of those with higher levels of political information were aware of the banking scandal during 1992, the proportion aware of the scandal among the lowest third of the population in information about government was under 70%. (Figure 2) This is, of course, completely predictable. People with knowledge about government, whatever their level of exposure to media, are far more able to store and recall information about political events. While cognitively engaged persons were more aware of the scandal than those with low levels of cognitive knowledge about government, they were far less likely to think that the scandal mattered enough to warrant punishing bouncers. Among all persons who knew about the banking scandal, 45.1% took a hard-line position that check-bouncing incumbents "should be voted out of office." As Figure 3 so dramatically illustrates, persons in the lowest three categories of information about political institutions were nearly twice as likely to be hard-liners as were persons in the top three information categories. The banking scandal is a classic example of an information shortcut, using estimates of personal character to assess public character. That persons with less information about government took the hardest line on check-bouncers demonstrates how levels of engagement affect the types of cues and thinking that citizens adopt. Candidate-centered politics, and a focus on the personal character as opposed to the political character of candidates, is in part a function of low information about the structure of government. Less-engaged people used evaluations of personal character -- evidence of the arrogance and privilege referred to by Rush Limbaugh -- as a substitute for information about the political character of incumbents. The persons most concerned about personal character and integrity, who were the most outraged about the privilege and arrogance reflected in the scandal, were precisely those who were least able to infer the candidate's true commitments from his or her past votes. The political and theoretical implication of this analysis also suggests why so many voters were turned off by Watergate, or the confrontation between Anita Hill and Clarence Thomas at his Supreme Court confirmation hearings. For persons with low levels of cognitive knowledge of institutions, in particular, there are no white knights once the dirt hits the fan. That is, when two groups engage in a long series of charges and countercharges, most people lose track of the issues or principles behind the skirmishes; what might have started as good guys versus bad guys soon becomes nothing more than a mudslinging free-for-all in which everyone looks bad. To persons who understand the institutions of politics, a long set of exchanges between, say, Bill Clinton and Robert Dole, can be as clear as a sustained volley in tennis; to persons without any knowledge of institutions their exchange is hard to follow and becomes indistinguishable from a food fight or mud slinging. Cognitive Engagement and Turnout As we have shown, persons with different levels of engagement use different kinds of information shortcuts. However, it is not all obvious that these differences should carry over to turnout. If, for example, persons had enough faith in an individual leader or a political party, then turnout would be independent of cognitive engagement. Over the past thirty years every measure of trust in government has steadily declined. Whereas once nearly three quarters of the citizenry believed they could trust the government in Washington to "do what is right" most of the time, today only a quarter of the country believes this. On question after question there are similar dramatic declines in trust in government. People increasingly believe government is run by crooks; run for a few big interests; wastes a lot of money; isn't influenced by elections and doesn't pay much attention to what people think. Since turnout is also lower than thirty years ago, much of the popular, casual analysis concludes that the declines in voting are related to this deterioration of trust in government. While trust in government does bear some relation to turnout, there is a much stronger relation between turnout and cognitive engagement than between turnout and trust. The frequent arguments that non-voters would be mobilized by more attractive policies are arguments that link lack of trust in government to knowledge about government and rejection of government policies. This might be called high- information distrust. Some advocates of policies that would overcome distrust and cynicism are arguing for economic policies to mobilize the underclass; some for moral and cultural policies to make participation more attractive to religious citizens; and some for tax policies to make government more attractive to the middle class. In each case though, there is an implicit assumption that distrust and cynicism is knowledgeable rejection of government and this rejection is what has caused the decline in turnout. There is also another, low-information possibility, however, that the persons who do not participate really don't even know what government is doing. For all the attention that has been placed on whether or not citizens trust their government and believe it cares about them, evidence is scant that distrust is related to turnout or that the citizens supposedly turned-off by tweedle-dum tweedle-dee political parties actually know anything about the government and policies that are supposedly turning them off. To test whether it is lack of cognitive engagement in government (i.e., a lack of basic understanding of political institutions) or a lack of trust (i.e., cynicism) behind non- voting we conducted multi-variate probit analyses of turnout on the 1992 American National Election Study. These tests demonstrate that the effects of cognitive knowledge about the structure of government are far more significant and dramatic than the effects of trust in government. Figure 4 graphs the relationship between turnout and trust and between turnout and cognitive engagement. Trust in government is related to turnout but the effect is slight. Controlling for age, sex, education, income, and partisanship, the differences in turnout between the highest and lowest categories of trust are less than 15%. The differences in turnout between the highest and lowest levels of cognitive engagement, however, are over twice as large. Persons in the top third of the cognitive engagement scale are more than 20% more likely to vote than persons in the bottom third. People who do not understand the basic features of political institutions are less likely to grasp the connections between issues, candidates, parties and offices. Anthony Downs related the decision to vote to the size of the estimated party differential citizens perceived between the parties. He took for granted that all citizens would have an estimate of the size of the differential and would vote when the estimated difference in policy outcomes outweighed the cost of actually voting. In America, however, non- voting is most often because potential voters literally do not know whether the parties or candidates are tweedle-dum and tweedle-dee or night and day. we suggest that turnout is so dramatically related to cognitive engagement in the American electorate because in the American system of federalism, presidentialism and primaries fragments any ongoing ties to political parties or candidates. American voters are constantly bombarded with cues suggesting that any long-standing estimates of differences between parties are inapplicable which in turn pressures them to recreate their estimates of the distances between parties and candidates in each election. Persons with low knowledge about government are less able to do so. The lack of understanding about institutions is more central to non-voting than the lack of trust in government. Cognitive Engagement and the Global Economy With little knowledge or awareness of political institutions persons see free trade, foreign investment and immigration as more threatening than do those who have some knowledge about the political institutions which maintain sovereignty and regulate these exchanges. The world economy is more of an unrestrained free-for-all, and therefore more threatening, when people try to make sense of the world economy without any knowledge of the institutions that shape their involvement in it. Because so many questions about economics and security are connected with ideology, we control for ideology as well as education, age and income in the rest of our analysis. Whether a person is liberal or conservative, educated or uneducated, institutional knowledge affects the way that person views the world economically and militarily. The lower the level of cognitive engagement with political institutions, the greater the fear that immigrants will take jobs away and the greater the belief that immigrants are a burden. The 1992 ANES survey asked respondents two questions about job loss and immigration, how likely it was that the growing number of Hispanics would take jobs away from people already here and an identical question about Asians. The two sets of answers were nearly identical; 49% thought it extremely or very likely that Hispanics would take jobs away from persons already here and 50% thought it likely that the same would happen with Asian immigrants. For both Hispanic and Asian immigrants, persons in the lowest three information categories are more than twice as likely as persons in the top categories to think immigration leads to job loss; in the three lowest categories nearly 60% thought immigrants took jobs, in the highest three categories it was only half as many. Even when controlling for the other demographic variables in a multivariate logit model, this effect remains as strong. (Figure 5) These results are robust and cannot be explained away. Cognitive involvement with political institutions makes immigration seem significantly less negative. People who view the world in a manner which incorporates knowledge about institutions, no matter their age, education, ideology or income, are less threatened by immigration. Similar analyses are possible on many of the surveys conducted by The Times Mirror Center for The People & The Press. They also asked whether immigrants strengthen the country or are a tax burden, and also whether they opposed or supported free trade agreements like NAFTA and GATT; they also asked factual questions about government. The more people know about politics, controlling for education and ideology, the more accepting and appreciative they are of the ways that immigrants strengthen the country. The effect of cognitive engagement upon the support of free trade is just as strong and robust. Opposition to free trade declined from 45% in the lowest information category to 23% in the highest category. When the confounding effects of education, income and ideology are removed, the decline is just as steep. There is a similar pattern with respect to attitudes about foreign investment in the United States. The more people are cognitively engaged in politics the less likely they are to think foreign investment is a bad influence on the country. The less people know about political institutions, controlling for education and ideology, the more likely it is that they will be apprehensive about foreign people, foreign money and foreign trade. It is hard to avoid the inference here that persons without an awareness of national political institutions feel naked in the face of the new world economy, and believe their sovereignty is being abrogated. If so, then the data shows that free trade and the free exchange of people, money and ideas requires not the absence of political institutions -- a laissez-faire world -- but strong institutions that can assure citizens that they are not facing change and uncertainty alone. Isolationism Persons who are less cognitively engaged in politics are also more isolationist than persons with more knowledge about government. Involvement in general is less appealing to the less knowledgeable, and when we do get involved, they are less likely to find value in the involvement. One in four Americans agrees with the straightforward blanket isolationist sentiment that "This country would be better off if we just stayed home and did not concern ourselves with problems in other parts of the world." Among the lowest third of respondents on cognitive engagement the proportion is over 40% while among the most knowledgeable third of the public only 13% agree with such a bald statement. When the confounding effects of education, income, age and ideology are removed the effects of information are even more dramatic. As Figure 6 illustrates, isolationism plummets as knowledge increases. Figure 6 also shows the relation between cognitive engagement and opposition to giving economic assistance to developing democracies in Eastern Europe. Well over half of low-information Americans oppose loans, while less than one in three of the most knowledgeable third of the public opposes loans. Lending money is not, in other words, an alternative to involvement; it is opposed by the non-engaged just as much as any other form of involvement. In retrospective evaluations of the Gulf War, respondents with low levels of knowledge about political institutions were less willing to go along with prevailing sentiments that the U.S. did the right thing in the Gulf War or that the Gulf War was worth the costs. The direct effects of are very strong in both cases. Low information Americans, whether or not costs are an issue, were far more begrudging of the Gulf War. (Figure 7) Less engaged Americans, in other words, are neither patriotic flag wavers always proud that the US won the war, nor band wagoneers going along with the prevailing sentiments after the Gulf War. Since the retrospective questions were asked in 1992, it is possible that the survival of Sadam Hussein made the war look less successful to low-information persons. It is far more likely, given the pattern of responses, that the connections between the war and their own future or the country's future were just much less apparent to these respondents. Whatever the explanation, the less information about political institutions people have, the more isolationist they are. What makes these results so dramatic is that they hold when controlling for education, ideology, income and age. Furthermore, they are opinions that do not reflect a begrudging acceptance of the general tone of the media. Without institutional context, specific international actions -- be they be foreign aid or military involvement -- do not appear to make as much sense either in advance or after the fact. International Danger The persons who are less aware of political institutions are not complacent isolationists, willing to stay out of foreign involvements because all is well. They perceive a more dangerous world than persons with more institutional knowledge. They were less likely in 1992 to believe the Cold War was coming to an end and they were far more worried about future conventional and nuclear wars. Perceptions of the possibility of future nuclear or conventional wars were also related to cognitive engagement. These effects were particularly strong for worries about nuclear as opposed to conventional war. Two-thirds of the less knowledgeable but only one-third of the more knowledgeable respondents were very or somewhat worried about future nuclear war. (Figure 8) By 1992, two thirds of all Americans thought the cold war between the Soviet Union and the United States was coming to an end. While three-quarters of the most knowledgeable third of the country thought this was so, only half of the bottom third agreed. Even on somethng as basic to the entire last 40 years of international politics as relations with Russia, persons with less awareness and knowledge about their country's institutions were less likely to see the implications of the changes in Eastern Europe, the unification of Germany and the collapse of communism. Earlier we suggested that less engaged persons are not as able to distinguish principled debates from mud-slinging or food fights. Here, an analogoue possibility is consistent with the data. Persons with less institutional knowledge were, we suggest, less likely to believe the cold war was ending because they were less able to distinguish arguments over, say, the Baltic States from arguments during the cold war. And the increased concern over war may follow from a similar inability to distinguish between peripheral skirmishes in places like Somalia from skirmishes that have more serious implications. Conclusion Why should knowing who William Rehnquist is and Tom Foley was, or which party controlled congress, have affected attitudes about immigration, foreign investment, free trade or isolationism? While it is clear why such knowledge would be related to reliance on cues about the personal as opposed to political character of candidates, or to the ease with which citizens arrive at summary judgments about the stakes in elections, it is less obvious why political knowledge affects attitudes about the global economy and isolationism. We suggest this basic political information is related to anxiety and concern about change and things foreign because the knowledge is part of a map of the world that helps persons navigate and deal with change. Whatever their level of attentiveness to news, we suggest that persons with institutional knowledge sort and process news differently from those without this knowledge. A tourist familiar with the geography of New York can place reports of violent crimes in perspective -- knowing, for example, that the South Bronx is far from mid-town Manhattan. In a similar fashion, we suggest, a person with knowledge of political institutions might be better able to place reports of violence around the world in perspective, knowing the difference between a moral and humanitarian crisis like Rwanda, and a civil war in a nuclear state like North Korea or China. Further, we suggest persons whose thinking about the world includes awareness of and knowledge about institutions will not feel as vulnerable and apprehensive because they will understand they are not isolated and that they are not facing social change on their own. If this is correct much of the apprehension Americans feel about the global economy is not due to the poor performance of their political institutions but to their lack of awareness that these institutions are mediating the changes they are undergoing. Endnotes Michael X. Delli Carpini and Scott Keeter. "Political Knowledge of the U.S. Public: Results from a National Survey," paper given at the annual meeting of the American Association for Public Opinion Research, May, 1989; What Americans Know About Politics and Why It Matters, forthcoming, Yale University Press, 1995. John Zaller. The Nature and Origins of Mass Opinion. New York: Cambridge University Press, 1992. Anthony Downs. An Economic Theory of Democracy. New York: Harper and Row, 1957. Samuel L. Popkin, John Gorman, Jeffrey Smith and Charles Phillips. "Toward an Investment Theory of Voting Behavior: What Have You Done for Me Lately?" American Political Science Review Vol. LXX (3) (September 1976):779-805. Samuel L. Popkin The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. Chicago: University of Chicago Press, 1991. Zaller, op cit, p. 42. ibid., p. 45. See also p. 217. . ibid., p. 3. . ibid., p. 5. . Michael X. Delli Carpini and Scott Keeter, "Political Knowledge of the U.S. Public: Results from a National Survey," paper given at the annual meeting of the American Association for Public Opinion Research, May, 1989. . Since 1958 studies conducted by the University of Michigan Survey Research Center have asked respondents to name their congressman. The proportion of all adults who could do so has never been over 50 percent, and the proportion of adult voters who could do so has ranged from a low of 46 percent in 1980 to a high of 64 percent in 1968. These surveys are conducted immediately after elections; if a survey asks persons the name of their congressman before a campaign, the numbers are much lower. In October 1977 and January 1978, the CBS News/New York Times asked respondents the name of their congressman and less than one-third of the adults in the two polls, and only 49 percent of those who were college graduates, could do so. Since 1978 the name of each respondent's Congressional representative has been included in a list and the respondents have been asked if they recognize the name. The percentage recognizing the name of their representative (not necessarily as a representative but as someone whose name they recognize, is over 90 percent. Aristotle, The Rhetoric and Poetics of Aristotle. New York: Random House, 1954. p. 25. Lionel Trilling, Sincerity and Authenticity, (New York: Harcourt Brace Jovanovich, 1971), p. 4. Gary C. Jacobson and Michael A. Dimock "Checking Out: The Effects of Bank Overdrafts in the 1992 House Elections" American Journal of Political Science, 38:3 (August 1994) Mr. Limbaugh's comment ws made on the McNeil-Lehrer Newshour Friday March 13, 1992.a see, for examples, the comments by Curtis Gans of the Committee for the Study of the American Electorate as cited by Mark D. Uehling in "All American Apathy" American Demographics November, 1991 The two scales were normalized to run from zero to one so they could be graphed together. The probabilities were estimated in a multivariate logit model controlling for education, age, gender, income, and being a party identifier. The actual logit model is reported in Appendix I. Times Mirror Center for The People & The Press, "The New Political landscape," uly 1994. However, there were only three factual questions that could be used to measure cognitive engagement: "Who is the current vice-president?"; "Which party has a majority in the U.S. House of representatives?" and "Who is the president of Russia?" Popkin and Dimock, 1995