Dec 1 1992

Issues, Images & Impact

A FAIR survey of voters' knowledge

While there was no shortage of polls published during the election campaigns, most opinion surveys offered little more than a snapshot of attitudes or voting intentions. FAIR’s survey of the U.S. electorate during the 1992 presidential election campaign, by contrast, was concerned not only with what people thought, but with the knowledge that lay behind those attitudes, and where that information came from.

Democracy in the U.S. depends increasingly upon the news media. Surveys show that for most citizens, the media are the principal source of information about politics and candidates. Of the many news outlets available, the most important is network television, which is the place—to paraphrase ABC News—where most Americans get their news. Our study focused on the role played by the news media (and TV news in particular), asking the following questions:

  • Did the news media successfully communicate enough information for voters to understand the political issues?
  • How did the public’s knowledge of hard issues compare with more trivial knowledge, and how did this relate to media exposure?
  • Did the information people receive suggest any bias toward certain political positions? Did, for example, the media effectively communicate more “pro-Bush facts” or “pro-Clinton facts”? Did the information communicated by the news media make voters more inclined toward the left, the right or the center?

The study was based on a nationwide telephone survey, conducted Sept. 30 to Oct. 3, of 601 randomly selected Americans who said they would “probably” or “definitely” vote. These likely voters tended to have more years of education than the average American.

Millie, Murphy & the Issues

Why do people vote for one candidate rather than the other? This question is never as simple as it looks, and the many possible answers to it are enough to stimulate the thriving industry of political punditry. What we can say with some degree of certainty is that people make their decisions to vote based on what they know about the candidates. This led us, in this survey, to ask people not so much what they thought of the candidates, but what they knew about them. The answers often exhibited a stunning ignorance.

Like most other surveys taken at the beginning of October, out poll showed Clinton leading with 41 percent, Bush with 32 percent, Perot with 10 percent and 17 percent undecided.

Despite voters’ proclaimed desire to base their votes on the issues rather than on more trivial information about the candidates, most people, we discovered, knew very little about the former and a great deal about the latter. When asked facts about the candidates’ policies and backgrounds, the only questions that a majority knew the answer to were “Which candidate’s family has a dog called Millie?” and “Do you recall which TV character Dan Quayle criticized for setting a poor example of family values?” Eighty-six percent knew that Millie belonged to the Bush family, and 89 percent correctly identified Murphy Brown.

This compared with 19 percent who could name the Reagan/Bush cabinet member recently indicted for his role in the Iran-contra scandal, Caspar Weinberger. (Perhaps the only surprise, in this respect, was that only 23 percent could correctly recall the name of the woman alleged during the primaries to have had an affair with Clinton—Gennifer Flowers.)

Even an issue that was topical in the week the survey took place had a majority confused. At the beginning of the week, President Bush vetoed an attempt by Congress to impose sanctions on China for human rights abuses, a position Bush had long held. Despite this, when asked what position Bush had taken on China since the Tiananmen Square crackdown, only 44 percent knew that Bush was against sanctions, while roughly as many (43 percent) stated that Bush had actually imposed trade sanctions. In other words, people were as likely to link the President to a policy he steadfastly opposed as they were to link him with the one he actually pursued.

When it comes to Clinton’s record in Arkansas, perceptions of the candidate on various issues bore even less resemblance to reality. According to independent calculations, Arkansas state taxes are among the lowest in the nation. Despite this, when asked, “To your knowledge, how high were Arkansas state taxes while Clinton has been governor?” only 21 percent responded correctly that they were among the lowest in the nation,” while more—32 percent—reported that they were “among the highest in the nation.”

Similarly, the Green Index, compiled by an independent monitoring group, has put Arkansas near the bottom of all states in most areas of environmental policy. But when asked, “how has Governor Clinton’s record on the environment been rated by an independent monitoring group?” only 19 percent correctly said “among the worst in the nation.” (52 percent said “about average” and 7 percent “among the best.”)

Of the 21 factual questions about the candidates on the issues, the average percentage of correct responses was 32 percent, a figure that drops to 27 percent if the more trivial, less “issue-oriented” questions are excluded. Not surprisingly, those with a college education tended to score higher than those without. But while Clinton supporters were more likely to be drawn from less educated groups (his lead among those with no college education stretched to 25 percent), they were, overall, the best informed group of voters in this survey, followed by Perot supporters, with Bush supporters scoring the lowest.

These differences become even more striking when we look at the most knowledgeable 20 percent and the least knowledgeable 20 percent. Among the most knowledgeable group, Clinton led by almost 30 points (55 percent to 26 percent) over Bush (with Perot at 9 percent). Among the least knowledgeable, on the other hand, Bush had a commanding 44 percent to 26 percent lead (with Perot at 8 percent). Predictably, we found more undecided voters—22 percent—among the least knowledgeable, while only 10 percent of the most knowledgeable were undecided.

It was also notable that exposure to the main information source in our culture, television, id not increase knowledge. Heavy TV viewers knew slightly less than light TV viewers. Similarly, those who relied on TV as their main source of news scored slightly lower than those who relied on other sources, such as newspapers.

The Successes of the Republican Campaign

Many in the Bush campaign lamented (in public, at least) about Bush’s failure to get his message across—often blaming the “liberal media” for this failure. Our survey suggested that on a number of issues, the opposite was true. This related not simply to Clinton’s record on taxes, but other areas where Republican attacks appeared to have hit home.

Take, for example, a question about the candidates and the draft: “Of the four candidates for president and vice president [Bush, Clinton, Quayle, Gore], have any been accused of using family influence to avoid being sent to Vietnam?” This statement applied to both Bill Clinton and Dan Quayle, yet only 23 percent named both, while almost three times as many picked only Clinton (41 percent) as only Quayle (15 percent). To put it another way, we could say that in a competition between a Republican campaign answer, a Democratic campaign answer and the correct answer, the Republican answer won.

An issue that potentially undermined Vice President Quayle was the fact that, despite his criticisms of the “cultural” or “media elite,” he himself comes from a family that owns a newspaper chain. The good news for Quayle was that few people were aware of this—only 24 percent in our survey.

Several Democratic attacks on the Bush record, on the other hand, didn’t hit home. Democrats complained, for example, that despite Bush’s account of the “tax and spend” Congress, Bush proposed a bigger budget last year than the one Congress finally approved (Washington Post, 8/26/92). While this is true, most people believed the opposite to be the case. When asked, “Last year, which do you think was greater: the amount of money which President Bush proposed spending on the federal budget, or the amount Congress actually passed?” 66 percent said Congress, and only 22 percent correctly identified Bush. Even Clinton supporters chose the wrong answer by a margin of more than two to one.

Despite surveys suggesting that on the issues (such as family leave) that voters consider to be the most important “family” issues, most preferred Clinton’s proposals over Bush’s (Time, 8/31/92), Bush was still seen as more supportive of “family values” than his opponent. When asked which candidate was more supportive of “family values,” 45 percent chose Bush and only 28 percent said Clinton. One theme that Democratic strategists seemed more successful at getting across than the Republicans was tying their opponent to a symbol of elite privilege, the Ivy League. Fifty-seven percent identified Bush as a candidate who had attended an Ivy League university, while 44 percent named Clinton. In fact, both attended an Ivy League school (Yale), although only 14 percent chose the “both” answer.

Despite the apparent success of the GOP in getting much of its message across, the fact remains that people who stated a preference tended to choose Clinton over Bush. This suggests that the general level of antipathy toward Bush was so deep-rooted that voters could absorb Republican criticisms of Clinton as a “draft-dodging, tax and spend liberal”—and still vote for him.

The Republican claim that the news media promoted “pro-Clinton” facts more aggressively than “pro-Bush” facts was not supported by this survey. Far from evidence of a “liberal bias,” the data on the media’s impact on public attitudes suggests that the reverse is more plausible.

There was also a more profound problem highlighted by these results. The news media’s emphasis on reporting campaign rhetoric rather than the facts, their reluctance to focus on the record rather than on “claims” about the record, seemed to make it difficult for voters to distinguish between truth and propaganda. In the stress on claims and counter claims, the facts become elusive and, in the end, unimportant. It would appear that the news media, the message carriers in a modern democracy, have played a key role in the evolution of a political process in which most voters know little about what is really going on, The winner, in this impoverished political climate, was not democracy, but the manipulative world of public relations.

Clinton, the Imaginary Liberal

Perhaps the most conspicuous success of the Bush campaign was its portrayal of Clinton as a liberal. When asked about the candidates’ policy questions, voters consistently attributed positions to Clinton than were more liberal than he had endorsed, while his more conservative stances were mostly unknown.

So, for example, only 37 percent knew that he supported the death penalty, only 37 percent knew he supported “right to work” laws opposed by organized labor, and only 32 percent knew that Clinton supported cuts in capital gains taxes. While is was true that Clinton had proposed reducing military spending by more than Bush, 73 percent agreed with the statement that he had proposed a cut of 50 percent over the next five years—significantly more substantial than the 30 percent cut that Clinton had proposed.

Perhaps the most glaring misconception concerned the source of the Democrats’ campaign money. Independent assessments (from groups such as the Center for Responsive Politics) indicated that both the Republican and Democratic presidential candidates relied much more on corporate and business interests than on labor for campaign money. Yet when asked who contributes more to Clinton, people overwhelmingly said labor (69 percent) rather than business (16 percent). While it is true that organized labor was more supportive of Clinton than of Bush, this answer indicated an underestimation of Clinton’s ties to business—and an exaggerated sense of the importance of labor to Clinton’s campaign.

Just as it was attractive to GOP strategists to portray the election as a battle between liberal and conservative ideologies, it was easier for the media to emphasize the differences rather than the similarities between the two major candidates. Our survey suggests, however, that such a framework was misleading to the public.

What People Knew About the Issues

Some of the most striking gaps in voters’ knowledge came in more general questions about the society they live in. Perhaps the most extraordinary finding in the entire survey concerned people’s perceptions of how the federal government spends their money. While decisions about where to spend money will be some of the most important choices any president will make, it was clear that most voters had little idea where their federal taxes are spent.

We asked respondents what the federal government spent more on in 1992: foreign aid, the military of welfare. The most popular answer, given by 42 percent, was foreign aid. In fact, foreign aid consumes a tiny portion of the budget—just 1 percent, according to the Senate Budget Office. (Of the developed countries, the U.S. spends among the least on foreign aid, per capita.)

The second most popular answer, at 30 percent, was welfare, which consumes just 5 percent of the federal budget, while military was named by only 22 percent of our respondents—even though, at 21 percent of the budget, it is by far the largest of these three items, more than four times larger than welfare spending.

How can we explain such an extraordinary misconception? Since most people relied on television for such information, it seems that TV news had done a poor job in communicating basic facts about how tax dollars are spent. Indeed, while public perceptions bore no relation to actual spending, they may have been a rational response to the extensive coverage given to the “welfare problem” and the need for “welfare reform,” and to the oft-repeated and rarely challenged statement that the U.S. spends “too much” on foreign aid and not enough on folks at home. Reporting on military spending, on the other hand, has tended to focus on cutback sand the negative consequences on jobs. (Reporting on the cutbacks in the non-military budget that occurred throughout the Reagan/Bush years, by contrast, rarely focused on the consequent job losses.)

Part of the problem was that media took much of their agenda from the mainstream political candidates, so if major candidates were encouraging this distorted view, the media were more likely to report it than challenge it. It was not surprising, then, that these distortions were more vivid among heavy TV news viewers, particularly when it came to overestimating welfare and underestimating military spending.

The political consequences of these distortions were complex, although they would appear to favor centrist and right-wing candidates. Bush might have benefited from misconceptions about welfare spending, while Clinton may have benefited if Bush were linked with inflated perceptions of foreign aid. A candidate on the left who supported foreign aid and welfare spending, and proposed major military cutbacks, would have to fight pervasive spending myths. Since Clinton, while not the candidate of the left, drew support from that direction, it was not surprising that Clinton supporters were the least likely to be misinformed on this issue, particularly on welfare and military spending. (Among Clinton supporters, military was more likely than welfare to be seen as the biggest expense, while Bush and Perot supporters were twice as likely to believe welfare costs more than the military.)

Welfare spending was subject to a number of misconceptions. A majority in our survey (60 percent of those who ventured an opinion) overestimated the portion of welfare recipients who are black, while 80 percent overestimated the average number of children women on welfare have. The average number of women ascribed to women on welfare by our respondents was 3.3. The actual number is 1.9. (Only 2 percent of our respondents underestimated this figure.)

Again, these misconceptions were linked to TV news viewing: Heavy TV news viewers were more likely to overestimate the proportion of people on welfare who are black and the number of children “welfare mothers” have. Only 35 percent of people who named TV news as their primary source of information, for example, did not overestimate the number of blacks on welfare, compared to 51 percent of people who got their news from other sources.

Another myth that had taken firm hold of the electorate was the issue of the tax burden on the middle class. We asked, “Which income group pays the highest percentage of their income in state and local taxes: the richest 1 percent, the middle 20 percent or the poorest 20 percent?”

Although our respondents were extremely confident they knew the answer to this (only 1 percent chose “don’t know”), the vast majority, it turned out, did not. Only 11 percent gave the correct answer (the poorest 20 percent), while an overwhelming 85 percent incorrectly identified the middle 20 percent as the main victims of local tax inequities. This seemed to reflect the emphasis by politicians and journalists on the “beleaguered middle class”; the beleaguered poor tend to be ignored by both.

While voters seemed to be misinformed on some issues, on others they were simply confused. Reporters and politicians have become used to talking about the influence of “special interests” on politics, but who, exactly, are these “special interests”? When we asked this question of our respondents, we found that, even when prompted, almost a third (32 percent) could give no answer at all, a very high proportion of “don’t know” responses. TV news viewers were particularly baffled by the term: 36 percent of those who chose TV news as their primary source of information said “don’t know,” compared to 26 percent of those who chose other sources.

Respondents who were able to identify “special interests” gave an extraordinary array of answers. Some respondents said the term referred to groups with the financial power to influence politicians, like corporations; others said that it meant groups that had special needs, like minorities or the disabled; others cited groups that focused on particular issues, like environmentalists or activists in the abortion debate, or even the “National Organization for Women.” Given the various ideas the phrase evoked in the public—and the large number who had no idea what it meant—perhaps journalists would have done well to use a more precise term or explain what they meant when they said “special interests.”

Political Bias & the Media

While most scholarly research on the media emphasize its tilt toward mainstream or conservative views, the most well-know criticism has come from the political right, with conservatives to Spiro Agnew to Dan Quayle accusing the media of having a liberal bias. The “liberal bias” criticism surfaced again during the ’92 campaign, particularly during the Republican convention.

What did our survey reveal about the question of bias? First of all, it was clear that the Republicans had gotten their message about bias across: While a plurality (45 percent) felt the media are “pretty balanced,” the rest were four times more likely to see the media as liberal (34 percent) than conservative (8 percent). This did not, of course, prove the Republican case: On the contrary, it may simply suggest that the media actually give more space to criticism of themselves from the right then from the left. (Journalists would rather bee seen as balanced, but may prefer to be attacked from the right as over-zealous watchdogs of the establishment, rather than by the left as establishment lapdogs.)

In order to ask a question about media slant whose correct answer could be verified, we asked, “Overall, in recent presidential elections, would you say more newspapers have endorsed the ore liberal candidates or the more conservative candidates?” Most people said that more liberals were endorsed (57 percent) than conservatives (27 percent). But in reality, every conservative candidate from Nixon in 1968 to Bush in 1988 received between 66 percent and 80 percent of daily newspaper endorsements. The notion of “liberal bias” was thus assumed even when the evidence contradicted it.

The overall data in this survey also finds little to support the conservative claim of bias. As we have indicated, the Republicans were, on many issues, quite successful in getting their message across: When asked about using family influence to avoid going to Vietnam, people thought of Clinton and not Quayle; when asked who the big spenders are, they thought of Congress rather than Bush; when asked to evaluate Clinton’s tax record in Arkansas, they preferred Republican claims to the independent assessments that contradicted it.

Overall, this study, conducted only a month before election day, painted a picture of a fairly uninformed electorate who still knew little about the candidates and even less about the underlying issues. On the whole, those who watched more TV news did not know more about the issues, and in many cases appeared to know less about the realities of the political world. The questions asked about TV’s poor coverage in the wake of the 1988 campaign remain with us.