Demographics

There is a strong consensus across all demographic groups on the broad view that the present system does not serve the interests of the whole public as much as it should, and that the majority public should influence the government more than it does. But while different demographic groups endorse these views equally, not all groups feel equally able to influence the actions of elected officials. This sense of political efficacy and government responsiveness are important, because as the political scientist C. V. Hamilton notes, "People participate where, when, and how they think it matters." Hence, while groups’ attitudes about the desirability of the publics influence on government decisionmaking only vary from positive to extremely positive, their sense of political efficacy and government responsiveness varies widely with an individual’s education level, income, age, political identification, race and region. There is less agreement among groups, however, on the means by which the public’s preferences might be ascertained.

For the analyses presented below, we have used the political efficacy and government responsiveness scales developed and used by the National Election Study since the 1950s. The political efficacy index is developed from three questions: "People like me don’t have any say about what the government does"; "I don’t think public officials care much what people like me think"; and "Sometimes politics and government seem so complicated that a person like me can’t really understand what’s going on." Many studies have found an important link between a sense of political efficacy and whether people vote. The government responsiveness index is composed of two questions: "Over the years, how much attention do you feel the government pays to what the people think when it decides what to do?", and "How much do you feel that having elections makes the government pay attention to what the people think?" Both indexes are scaled from 0 to 100, with low scores representing low efficacy and low responsiveness respectively.

Finally, the reader should bear in mind that when results are not presented it is because there were no reliable differences in the way groups responded.

Education

Throughout the forty-year history of the National Election Study, higher levels of education have been associated with a stronger sense of political efficacy. The findings in this poll are no exception to this rule. Respondents who earned an advanced degree had a mean political efficacy score of 58 (of 100). Political efficacy declines steadily with decreasing levels of education, to a score of 26 for those who did not finish high school. This is most clearly illustrated by the responses to the statement, "Public officials don’t care much what people like me think." A majority of those with an advanced degree (55%) disagreed with this statement, while a strong majority of those who did not finish high school (64%) agreed. Responses for those with moderate levels of education fell between these two points. Similarly, Americans with higher levels of education perceive the government as more responsive. While respondents with an advanced degree had an average score of 49 (out of 100) on the NES government responsiveness scale, those without a high school degree average a score of 34. When asked how much elections made the government pay attention to what the people think, an overwhelming number of those holding an advanced degree (87%) said either "a good deal" (44%) or "some" (43%), while only 55% of those who did not finish high school said "a good deal" (25%) or "some" (30%) and (42%) said "not much."

Although there were differences in perceptions of how responsive the government is, there were almost no reliable differences among people of various education levels on what they thought the role of the public should be in government. The sole effect was that as the level of education increased, larger majorities rejected the arguments against government officials attending to the preferences of the majority of the public when making decisions. For example, 78% of those with some college or more found unconvincing the argument that the judgments of officials, "who are well informed and are able to think through issues thoughtfully and objectively … should count for more than the views of the public." Seventy-five percent of high school graduates found it unconvincing, while 53% of those without a high school degree found it unconvincing. There were no differences in the levels at which people of differing educational achievement found the three ‘pro’ arguments for the influence of majority public opinion convincing.

Household Income

Higher levels of income are associated with a steady increase in a sense of political efficacy, even when the effect of education is taken into account. Those who report a household income of $100,000 or more per year have an average political efficacy score of 56 (out of 100), compared to a score of 30 for those making $15,000 or less. This is most clearly illustrated by the responses to the statement, "People like me don’t have any say about what the government does." A strong majority (63%) of those with incomes of $25,000 or less agreed with this statement; in contrast, those with incomes of more than $70,000 were split (47% agree and 47% disagree).

Consistent with the public’s perception of how Washington works, those with higher levels of income perceived the government as somewhat more responsive. While respondents with incomes of $100,000 or higher had an average score of 48 (out of 100) on the NES government responsiveness scale, those with incomes of $15,000 per year or less averaged a score of 32. When asked how much attention they felt the government pays to what the people think, a majority (53%) of those with incomes of more than $70,000 said either a good deal (10%) or some (43%), while a strong majority (66%) of those whose incomes were $15,000 or less said the government did not pay much attention to what the people thought. While a strong majority in all income groups felt the people in government did not really understand them all that well, this was strongest for people at the lowest income level (71%) and declined with income to 63% of those with incomes of more than $70,000.

The number of respondents disapproving of the impeachment decision declined steadily with rising household income. This was true even when party identification was held constant. An overwhelming majority (72%) of those with incomes of $15,000 or less disapproved of the decision to impeach: this declined steadily with rising income, with those with household incomes between $45,000 and $70,000 splitting on the issue and a majority (58%) of those making more than $70,000 approving of the decision. A similar pattern was found on the question of whether President Clinton should be removed from office, but support for this action was still less than half at the top income level.

Likely Voters

Likely voters were those who indicated that they were currently registered to vote (83.5%), said they had voted in the last presidential election (68.9%) and gave at least one positive answer to the three questions that comprise the National Election Study’s (NES) external political efficacy scale. Likely voters comprised 48.5% of the sample (the turnout in the 1996 presidential election was 49.4%).

While a majority of those not likely to vote feel that government does not pay much attention to what the people think (59%), only a plurality of likely voters (48%) felt not much attention was paid to the people when the government decided what to do. Similarly, likely voters were less pessimistic about how much people in government understand Americans with 58% saying "not that well" compared to 68% of those not likely to vote. Likely voters also show a stronger sense that the government is responsive (mean score: 45 of 100) than do those not likely to vote (mean score: 33).

A larger portion (59%) of likely voters felt that members of Congress should seek integrative solutions when the member’s opinion on an issue diverges from the course favored by his or her constituents, compared to those not likely to vote, of whom 52% chose the integrative solution. The option chosen least by likely voters was that the member should do what he or she thinks is best (15%).

More likely voters felt that members of Congress should pay attention to the views of the majority over views of people who write letters and call—59% among likely voters as compared to 49% among the rest.

Although likely voters thought Congress should pay attention to polls they did so by a smaller margin than those unlikely to vote. While 70% of those not likely to vote thought Congress should pay attention to polls showing their approval ratings, 52% of likely voters felt this way. Since approval ratings reflect the opinions of those who do not vote as well as those who do, arguably they have relatively greater value for those who do not vote. Therefore, it is not surprising that those unlikely to vote would find such polls more attractive than likely voters. Using the same reasoning, however, we can conclude that for both groups it is difficult for them to make their voice heard on particular issues, so it makes sense that this gap is narrower on the question of Congress paying attention to polls on specific issues. Among likely voters 63% felt Congress should pay attention to polls on issues and 72% of those unlikely to vote felt that way.

Forty-four percent of likely voters thought that having one of their Senators vote to remove Clinton would make them at least a bit less likely to vote for that Senator in the next election. However, 33% felt this would make them more likely to vote for that Senator, resulting in an eleven point gap in favor of Senators voting against removal. For the remainder of the sample, 47% thought they would be inclined to vote against that Senator in the next election, while 33% reported being more inclined to vote for a Senator who voted for removal—a 13-point gap. A fairly strong majority (60%) of likely voters felt Clinton should not be removed from office; this was slightly lower than for those unlikely to vote, of whom 67% felt the Senate should not remove Clinton from office. This finding was replicated for the effect of the impeachment vote by the House of Representatives. Here, likely voters continued to show a greater tendency to feel like voting Democratic (38%) than Republican (28%) as a result of the impeachment process.

Party Identification

Americans who identify themselves as independents show much more disaffection with the political system than do Democrats and Republicans. They have a lower sense of political efficacy (mean score = 35 of 100) than do Republicans (mean score 46) or Democrats (mean score =39). Moreover, this has been the case throughout the 46-year history of the National Election Study. Independents also feel that the government is less responsive than do Republicans and Democrats. Independents rated the government as less responsive (37 out of 100) than did either Republicans (41) or Democrats (45). A strong majority of independents (61%) felt that the government did not pay much attention to what people think, while 53% of Republicans and just a plurality (46%) of Democrats felt this way. A very strong majority of independents (69%) also felt that the government did not understand most Americans, as did majorities of both Republicans (57%) and Democrats (55%). This may be related to the even stronger feeling among independents that the government is "pretty much run by a few big interests" (82%) compared to 72% of Republicans and 70% of Democrats. Similarly, independents feel their own member of Congress is less influenced by the majority’s views (4.9 out of 10) than do Republicans (5.7) or Democrats (5.6).

While still quite substantial, Republicans have the smallest gap (3.1 points out of 10 possible) between the extent they want the views of the majority of the public to influence elected officials and the extent to which they believe the majority’s views do influence officials. For Democrats this influence gap averages 3.8 points, and for independents 4.0 points. Fewer Republicans saw growth in the influence of the wealthy (65%) than did Democrats (70%) or independents (72%). However, there were no reliable differences between the groups on whether corporate influence had grown (68%).

Although there were no clear differences in the degree to which people of differing political affiliations wanted elected officials to attend to the views of the majority of the public, there were differences in their attitudes towards one method of ascertaining those views; public opinion polling. A strong majority of Democrats (75%) and independents (71%) thought that members of Congress should use polls to get a sense of the public’s views on issues. Only half of Republicans (50%), however, thought members should read polls on the issues. Similarly, 51% of Republicans thought members of Congress should pay attention to polls showing their approval ratings, while 75% of Democrats thought they should, and 59% of independents approved of this approach. This effect may be related to the confidence each group has in their ability to judge polls. While a very strong majority of Democrats (73%) said they had "a fair amount" or more confidence that they could judge whether a poll was done in a fair and scientific manner, and a majority (54%) of independents felt this way, only 38% of Republicans said they had "a fair amount" or more confidence to judge polls.

As already seen in an abundance of polling data on impeachment from other organizations, party identification was strongly associated with supporting or opposing Clinton’s impeachment or removal, with Republicans more likely to support and Democrats more likely to oppose.

Ideology

Approximately one third (35%) identified themselves as conservative, 29% considered themselves moderates, 10% identified themselves as liberals, and 21% could not say where they fit in this continuum. While a strong majority of all those asked felt that if the leaders of the nation followed the views of the public more closely the nation would be better off, an overwhelming number (90%) of those who considered themselves liberals felt this way, compared to 77% of moderates and 79% of conservatives. Similarly, when asked about the influence of corporations on the government, strong majorities from each of the three ideological positions felt corporations had gained influence over the last ten years. This was strongest, however, among liberals, 80% of whom felt this way, while 72% of moderates and 67% of conservatives felt this way.

The strongest differences came not on the issue of whether government officials should listen to the public, but on how they should do so. While a majority of both liberals (60%) and moderates (70%) thought that Congress should pay attention to polls that showed approval ratings, only a plurality of conservatives (49%) agreed with this point of view. While less striking, a similar pattern emerged when asked if Congress should listen to polls about the public’s views on issues. Seventy-six percent of liberals and 67% of moderates, but just 60% of conservatives thought that members of Congress should read up on polls when deciding how to vote on an issue.

Those who considered themselves moderates and liberals were consistently more positive about the decisions that a representative sample of 500 Americans would make when compared to Congress. A majority of moderates (60%) and liberals (57%) felt that such a group would be more likely to reach a consensus when deliberating than Congress would, while a plurality of conservatives (49%) felt this would be the case. A similar pattern, though less pronounced, was present when respondents were asked about the decisions such a group would make compared to the wishes of the majority. A strong majority of moderates (74%) and liberals (70%) thought that the decisions of such a group would be more like the wishes of the majority of the public than Congress’ decisions. A smaller number, though still a majority, of conservatives (62%) felt this way.

Conservatives were the group most likely to say they followed public affairs more closely than average (73%), compared to 63% of liberals and 50% of moderates. A majority (57%) of conservatives also said they followed what was going on in government and public affairs most of the time, compared to 44% of liberals and 45% of moderates.

An overwhelming majority (77%) of liberals disapproved of the decision to impeach President Clinton, as did a strong majority of moderates (64%). However, a strong majority of conservatives (63%) approved of the decision. More liberals (80%) and moderates (77%) were opposed to removing Clinton from office than were opposed to impeaching him. But support for removal among conservatives, although still a majority (54%) was lower than their support for impeachment .

The Attentive Public

Those who say they follow public affairs most of the time are also those with the highest sense of political efficacy (mean score = 41.6 as compared to 31.8 among the rest). Even among this group though, a majority said they did not have any say about what the government does (53%), that public officials do not care what people like them think (60%), and a plurality (50%) agree that politics and government sometimes seem so complicated that a person can’t really understand what’s going on. There were no differences in how this group felt about the role of the public in government decision making or their attitudes toward the possible methods of determining public opinion.

Gender

There were no reliable gender differences on attitudes toward the role of the public in governance or the nature of public opinion. Women did, however, differ significantly from men on the issue of the impeachment process. Sixty percent of women disapproved of the House of Representatives voting to impeach President Clinton, while a smaller majority of men (54%) disapproved. The gap was wider when asked whether "the Senate should or should not remove President Clinton from office." A very strong majority (69%) of women said Clinton should not be removed, compared to 58% of men who opposed removal. This ‘gender gap’ was most apparent when respondents were asked "if one or both of your Senators votes to remove President Clinton from office" would this make you more likely or less likely to vote for that Senator. While there was a tendency for men on the whole to be less likely to vote for the Senator (net loss 3.7%), among women the net loss of people less inclined to vote against minus those more inclined to vote for such a Senator was 20.4 percentage points.

Region

While there were no differences between regions on whether the government understood "most Americans," the majority feeling that the government does not understand people "like themselves" was largest in the South. Seventy-five percent of those in the South felt that way as compared to 64% in the Northeast, 62% in the West, and 57% in the Midwest. As in previous surveys by the National Election Study (1952 through 1996), southerners had a slightly lower sense of political efficacy compared to non-southerners. This was most apparent in their stronger tendency to agree that "people like me don’t have any say about what the government does." Sixty percent of southerners agreed with this statement contrasted with 54% of those not in the South.

Race

Although the size of this study does not allow us to distinguish between different non-white or Hispanic groups, minorities as a whole tended to show slightly lower levels of political efficacy (mean efficacy score 36.5) than white Americans (mean efficacy score 40.5). This pattern is consistent with findings that show higher levels of political efficacy for white Americans compared to African-Americans since the 1950s in the National Election Study.

An overwhelming majority (81%) of members of minority groups disapproved of the vote to impeach Clinton, compared to just a majority of white Americans (53%). Similarly, while large majorities of both groups felt Clinton should not be removed from office, an overwhelming 87% of minorities felt this way, in contrast to 61% of whites. This leads to a large gap in the net loss of votes for a Senator who voted for removal (i.e. the difference between those who said they would be less likely and those who said they would be more likely to vote for the Senator). Among minorities there was a 49-percentage-point net loss of votes, compared to a 14 percentage point net loss of votes for whites.

Age

Older Americans have a lower sense of political efficacy (mean score: 30) than those 65 and under (mean score: 41). For example, 69% of those over 65 agree that "officials don’t care much what people like me think," compared to 57% of those 18 to 65. This has been the case throughout the history of the scale at NES. There was no difference among age groups, however, in the extent to which they considered the government responsive.

Older Americans appeared to feel less strongly that majority opinion should influence government decision making. They rated the level of influence the views of the majority should have on the decisions of elected officials as lower (7.9 out of 10) than did those 18 to 65 (8.5). While majorities of all age groups found the pro arguments convincing, these majorities were smaller for Americans over 65 than for younger Americans. Overwhelming majorities of those 18 to 65 found the argument convincing that since this is a democracy, the government should be guided by the will of the majority, while 63% of those over 65 supported this argument. Seventy-seven percent of 18-to-65 year olds found the argument that nobody knows what’s best for the people better than the people convincing, while a smaller majority (56%) of those over 65 found it convincing. Along the same line, fewer of those in the oldest group found the ‘con’ arguments unconvincing than did other age groups. Only a plurality (45%) of those over 65 found the argument that "there is really no reason to believe that the public is any better than the people in the government" unconvincing, while an overwhelming majority of those 18 to 65 (75%) rejected this argument. Consistent with this, a much smaller majority of those over 65 (57%) think that the way voters in a representative’s district feel about an issue should be more important than the representative’s judgement compared with overwhelming support for considering the way voters feel among younger groups (79%).

Younger Americans were the most positive about the use of polls. Seventy-eight percent of those under 29 thought Congress should pay attention to polls to get a sense of the public’s views on issues; 71% of those 30 to 40 years of age thought they should, as did 59% of those 46 to 65 and 61% of those over 65. A strong majority (70%) of those 18 to 29 thought members of Congress should pay attention to approval ratings, as did 64% of those 30 to 40 and 60% of those 46 to 65, but only a plurality (45%) of those over 65 thought members should attend to approval ratings. This mirrors the confidence these age groups have in their ability to tell whether polls are done fairly and scientifically. A strong majority of the youngest group (68%) said they had a fair amount or more of confidence in their ability to judge; and 54% of those 30 to 65 felt this way; while less than a plurality (47%) of the oldest group had this much confidence in their ability. Similarly, a strong majority (71%) those under 65 thought that a group of 500 citizens, given the necessary information, would come closer to the wishes of the majority of Americans than would Congress, but only half of those over 65 felt this way. A strong majority (72%) of those 18 to 45 and 65% of those 46 to 65 thought the decisions of such a group would be better than Congress’s, while among those over 65, 53% thought this was true.