The Vanishing Voter: Why Are the Voting Booths So Empty?


National Civic Review, Winter 2002, volume 91, issue 4, pages 367-378


The period from 1960 to 2000 marks the longest ebb in voter turnout in the nation's history. Turnout was nearly 65 percent in 1960 but fell in each of the five succeeding presidential elections. It rose by one percentage point in 1984 but then fell by three points in 1988. Although analysts viewed that drop with alarm, the warning bells really sounded in 1996, when more Americans stayed home than went to the polls on Election Day.

Less publicized but no less dramatic has been the decline in voting in other elections. Turnout in the 2002 midterm elections was 39 percent. Several decades ago, these elections drew half the adult population to the polls. Many states and communities have recently experienced a record low voting rate. Turnout has also fallen in primary elections; in congressional primaries it was nearly 35 percent in the 1970s, but a mere 18 percent in 2002.

Fewer voters are not the only sign of the public's waning interest in political campaigns. In 1960, 60 percent of the nation's television households had their sets on and tuned to the October presidential debates. In 2000, less than 30 percent were tuned in. Few Americans today pay even token tribute to a presidential election. In 1974, Congress established a fund to underwrite candidates' campaigns, financed by a check-off box on the personal income tax return that allowed citizens to assign one dollar (later raised to three) of their tax liability to the fund. Initially, one in three taxpayers checked the box. By the late 1980s, only one in five marked it. Since then, the number has fallen to one in eight. (1)

There is a puzzling aspect to the decline. The percentage of college graduates in the population has tripled since 1960. Literacy tests, poll taxes, and lengthy residency requirements have been abolished. Registration has been simplified. Yet turnout has fallen.

What is going on here? Why are Americans less engaged by political campaigns today than even a few decades ago? Some analysts claim that participation follows a natural cycle and will therefore rise again, just as it did after downturns in the 1890s, 1920s, and 1940s. But this argument overlooks the persistence of the current trend and the special nature of those earlier downturns (which were caused by the introduction of registration requirements, women's suffrage, and war weariness, respectively).

Analysts agree that generational replacement is part of the explanation for the decline. The civic-minded generations molded by the Depression and the Second World War have been gradually replaced by the more private-minded generation X and generation Y, who lived through childhood and adolescence without experiencing a grave crisis that called them to action. Today's young adults are less politically interested and informed than any cohort of young people on record. The voting rate of adults under age 30 was 50 percent in 1972. It barely exceeded 30 percent in 2000.

Generational replacement, however, is not the sole cause of the downward trend. Somehow, the United States has managed to create some of the least inviting and least savory campaigns imaginable. This development is the subject of this article, which is based on my recently published book The Vanishing Voter. The book, in turn, is an outgrowth of the Vanishing Voter Project, which was conducted during the 2000 campaign by the Joan Shorenstein Center on the Press, Politics, and Public Policy at Harvard University's Kennedy School of Government, with the support of a grant from the Pew Charitable Trusts. Our research team conducted weekly national surveys to discover why Americans were following or ignoring the campaign. By the time the study was concluded, we had interviewed nearly one hundred thousand Americans--easily the largest study of campaign involvement ever conducted (for information on the project, including state-by-state comparison of involvement level, see the project Website: www.vanishingvoter.or g).

Competition, Anyone?

With party control of both the House and Senate at issue, the 2002 midterm election was a pundit's dream. However, the intense competition for control of Congress masked the fact that the vast majority of House races were uncompetitive. Only about three dozen of the 435 House seats were actually in play in 2002. In nearly twice that many districts, there was literally no serious competition: the weaker major party did not bother even to nominate a candidate. In several hundred other districts, the competition was so one-sided that the result was known even before the campaign began. As was the case in 2000, the victors won by an average margin of more than two to one.

House incumbents breezed to victory in 2002, just as they have in other recent election campaigns. Ninety-eight percent of the incumbents seeking another term in the House were reelected. U.S. House races are less competitive--and by a wide margin--than those of any other freely elected national legislative body in the world.

House incumbents have created a lock on the offices they hold. When the campaign finance laws were changed during the 1970s in reaction to Watergate, political action committees (PACs) suddenly sprouted, increasing in number from six hundred to four thousand within a decade. This new source of money turned out to be a bonanza for incumbents, since PACs are reluctant to oppose politicians who are already in power. Today upwards of 85 percent of PAC money ends up in the pockets of incumbents,2 who also operate year-around reelection campaigns at taxpayer expense. When members of Congress in the 1960s voted to greatly enlarge their personal staff, they argued that the additional personnel were needed to offset the executive branch's domination of policy information. However, an estimated 50 percent and more of congressional staff resources are devoted to public relations, constituency service, and other activities that serve primarily to keep House members in office.

Competition is the lifeblood of a democratic election, and when it dries up participation suffers. In many of the House districts in 2002, there was no campaign to speak of, and the news media provided little or no coverage. Voters in these districts were deprived of an opportunity to learn of the issues and the candidates and, on Election Day, to cast a meaningful vote. Analysts offer varying estimates of the effect of uncompetitive House races on turnout, but a 3-5 percent decrease is a reasonable figure.

Seated senators and governors find it harder to use their office in a way that ensures reelection, and these positions often attract challengers who are well heeled or well known. Close competition for these offices in a dozen states supplied excitement and interest to the 2002 election. The turnout average in these states was higher than it was in 1998. Yet competitive races have become an anomaly The trend in House races is matched by what has been taking place in state legislatures. As these bodies have become more professionalized with larger staffs and salaries, their members have been able to use the advantages of being in office to stay in office and to turn politics into a lifelong career. In 2002, there were--continuing a trend--a record number of uncontested state legislative seats.

Many voters are also effectively disenfranchised by the way in which presidential primaries are structured. Front loading of the nominating schedule--the placement of a large number of state contests near the front end of the process--has led presidential hopefuls to raise and spend tens of millions on these early contests in an effort to secure nomination with a decisive victory on Super Tuesday One effect is to make money the king of the nominating process. Not since John Connally in 1980 has the candidate who raised the most money before the first contests in Iowa and New Hampshire lost a nominating race.3 A second effect is to deprive millions of citizens of the opportunity to cast a meaningful vote. Bush and Gore's Super Tuesday victories in 2000 completely devalued the yet-to-be-held presidential primaries and caucuses in other states. Turnout in those states was a third lower than that in the early-contest states and would have been next to nothing if nominations for other offices were not being contes ted. Our Vanishing Voter surveys revealed that residents of the late-scheduled states were also much less likely to talk about the campaign and to follow news about it. They were also less informed about the candidates and issues.

In the 1970s, when the nominating schedule unfolded a state at a time until the final month or so, the races lasted longer, money was less influential, and residents of nearly all states had a chance to cast a meaningful vote. Turnout nationally was twice the level that it is now.

In the presidential general election, Americans' opportunity to be part of the action is determined by the Electoral College. Although this feature of our constitutional system has always distorted the process to some extent, the fact that today's campaigns are based on money rather than volunteers has exaggerated the effect. Unlike volunteers, who work within the communities where they live, money can be targeted and withheld at will. During the 2000 general election campaign, there were no ad buys and no candidate visits in Kansas, a lopsidedly Republican state. In neighboring Missouri, which was a battleground state, there were eighteen candidate visits and millions of dollars were spent on televised political advertising. "The process effectively takes half the country and says, 'you're just spectators,'" notes Kathleen Hall Jamieson, dean of the University of Pennsylvania's Annenberg School of Communication. (4)

In 2000, residents of battleground states had a voting rate that was 3 percentage points higher than that of residents of other states. In fact, although the overall voting rate in 2000 was somewhat higher than it had been in 1996, turnout actually fell in nine states, all of which were safely in the Bush or Gore column. Residents of these and the other noncompetitive states also talked less about the campaign and paid less attention to election news than did the residents of battleground states. (5)

Clear-Cut Parties, Anyone?

The issue of whether voters have a choice includes the clarity and significance of that choice. Here, too, the situation is less favorable than it once was.

There was a long period in American history when elections were waged on economic issues powerful enough to define the two major parties and divide the public. These issues stemmed from Americans' deepest hopes and fears and had the power to cement their loyalty to a party and draw them to the polls. That era ended with the triumph of Franklin Roosevelt's New Deal, which (along with Lyndon Johnson's Great Society) put in place government programs that greatly reduced the sources of economic resentment and insecurity that had fueled party conflict. A safety net for the economically vulnerable was in place, as were policy mechanisms for stabilizing the economy. An electoral majority that could be easily rallied by calls for economic redistribution no longer existed.

As the impact of economic issues on voting behavior weakened, a large set of less comprehensive issues emerged. Civil rights, street crime, school prayer, and welfare dependency were among the earliest of these issues, which were followed by abortion, the environment, education, global trade, and others. All were important, but they intersected with each other in confounding ways.

None had the reach or the endurance of the economic issues. As a result, the issues of one election were usually different from the issues that had dominated the previous election or would be at the forefront in the next one.

How could the political parties create cohesive and enduring coalitions out of this mix of issues? The short answer is that they could not do so. The issues were too crosscutting and too numerous for either party to combine them in a way that could easily satisfy a following. By the 1970s, self-described independents accounted for a third of the electorate. People also found it increasingly difficult to think and talk about the parties. Americans were better educated than they had been in the 1950s, but they had a harder time saying what the parties represented. In the 1950s, fewer than one in ten had nothing to say when asked in polls what they liked and disliked about the parties. By the 1970s, three in ten had nothing to say. (6)

Since then, political parties have not recovered their prominence. The two major parties are now relatively weak objects of loyalty and thought, and the decline in party loyalty and identification has diminished Americans' concern with election politics. Like any other emotional attachment, party loyalty heightens interest and commitment. For its part, party awareness reflects people's ability to recognize what is at stake in election politics and the options available to them. (7) "My mind has just gone blank," said a Florida resident in 2000 when asked in one of our surveys to describe the parties. (8)

Americans who today have a party loyalty and awareness of the parties have a voting rate more than twice that of those who call themselves independent and who cannot find words with which to describe the parties. This was true also in the 1950s; the difference today is that the percentage of citizens in the high-voting group is much smaller and the percentage in the low-voting group is much larger than in the 1950s. (9) The type of citizen who votes less often has been gradually replacing the type who votes more often.

The change in party politics helps to explain why, disproportionately, the decline in participation has been concentrated among Americans of low income. Although a class bias in turnout has been a persistent feature of U.S. elections, the gap has now widened to a chasm. The voting rate among those at the bottom of the income ladder is only half that of those at the top. During the era when the electorate was divided by basic economic issues, working-class Americans were at the center of political debate and party conflict. They now occupy the periphery of a political world in which money and middleclass concerns are ascendant. In 2000, low-income respondents were roughly 30 percent more likely than those in the middle or top income groups to say the election's outcome would have little or no impact on their lives. (10)

The change in party politics also helps to explain why candidates now have trouble crafting a message that voters find compelling. Candidates have never had so many communication weapons at their disposal, yet they have never found it so hard to frame their message. As Franklin Roosevelt's voice crackled into living rooms through the vacuum-tube radio, his pledge to "the forgotten man" had a persuasive power that today's media consultants would envy. Listeners didn't have to be told what FDR had in mind or to whom he was speaking. Campaign messages today are strikingly different in the range of issues they address, the contradictions they contain, the speed with which they turn over, and the small percentage of voters with whom they resonate. After their defeat in the 2002 midterm election, Democratic leaders were roundly criticized for failing to put out a message that captivated voters. However, Democratic politicians are neither stupid nor apolitical. If a simple and compelling message were readily availab le, they would have seized it. Such messages are today quite rare. If Republicans could not rely on their perennial "let's cut taxes" pitch--which is now closer to a fight song than a true governing philosophy--they would face the same problem.

A century ago, James Bryce worried that the growing complexity of American society threatened the parties' ability to forge and mobilize a cohesive majority Social complexity is now orders of magnitude greater and has clearly overtaken the parties. The consequences include a lower rate of electoral participation.

Uplifting Campaigns, Anyone?

Beginning in the late 1950s and early 1960s, control of an election campaign began to shift from the political parties to the candidates, largely because of television and refinement in techniques of mass persuasion. Americans were initially thrilled by the chance for a close-up look at the candidates and their campaigns. Theodore H. White's The Making of the President, 1960, topped the best-seller list.

However, Americans have come to dislike nearly everything about the modern campaign. The new style has brought out aspects of politics that were once largely out of sight. Ambition, manipulation, and deception have become as prominent as issues of policy and leadership. But politicking, like sausage making, is best viewed from a distance. An election is supposed to bring out the issues. It is not supposed to ruin one's appetite, but that's the best way to understand much of what Americans now see during a campaign and why they don't have much taste for it.

Negative campaigning has long been part of campaign politics but now dominates it. Candidates have discovered that it is easier in many situations to attract swing voters by tearing down one's opponent than by talking about one's own platform. Research indicates that negative advertising has more than tripled since the 1960s. Such ads now account for well over half the ads featured in most presidential and congressional races. (11)

Our surveys indicate that a cumulative effect of negative politics, campaign after campaign, is reduced interest in the election. Attack politics wears some people down to the point where they simply don't want to hear about the campaign. On an average day during the 2000 campaign, Americans who believed that negative messages are a defining feature of U.S. elections were less likely than those who had a different view of the prominence of such messages to discuss the campaign and pay attention to news about it. The difference was not large, but it occurred across the course of the campaign. Day in and day out, those who believed campaigns are akin to mud wrestling were less attentive to the campaign, even when level of education and income were controlled. (12)

Today's campaigns are also characterized by promises--endless promises. Unlike their predecessors in the age of party-centered politics, today's candidates are unable to campaign on broad statements of principle within the context of a reliable base of party loyalists. Today's candidates build their following by pledges of support to nearly every conceivable voting group. The changing nature of party platforms tells the story. Whereas the platform was once a declaration of broad goals and ideals, it has become a promissory note to special interests. The 1948 Democratic and Republican platforms were less than three thousand words in length. By the 1980s, they exceeded twenty thousand words. (13)

An effect of this relentless flow of campaign promises is a public wary of taking candidates at their word. In our Vanishing Voter surveys during the 2000 campaign, 81 percent agreed with the statement "most politicians will say almost anything to get themselves elected." Respondents who felt this way had a significantly lower voting rate than others; they were also less likely to talk about the campaign and follow news about it.

Modern-day politics also exalts personality, increasing the likelihood that personal blunders and failings will loom large in a campaign. Through the 1972 presidential election, personal controversy did not receive even half as much news coverage as did policy issues. Since 1972, it has received nearly equal time. (14) Even a short list indicates just how salient personal controversy has become: Gerald Ford's blundering statement on Eastern Europe, Jimmy Carter's "lust in my heart" Playboy interview, Geraldine Ferraro's tax returns, Gary Hart's affair with Donna Rice, Dan Quayle's assault on the fictional Murphy Brown, Bill Clinton's relationship with Gennifer Flowers, and Al Gore's Buddhist Temple appearance. The revelation in 2000 that Bush had been arrested a quarter century earlier for drunken driving dominated the headlines in the closing days of the campaign. The incident got more coverage on the evening newscasts in a few days than did all of Bush and Gore's foreign policy statements during the entire general election.

Although a startling revelation can perk up a campaign, citizens do not particularly like the prominence it attains. In our surveys, 62 percent agreed with the statement "political campaigns today seem more like theater or entertainment than something to be taken seriously" Although those who held this opinion were no less likely than others to vote on Election Day, they were less likely to discuss election politics and to attend to news about it.

The length of the modem campaign is also a turnoff for many Americans. Today's candidates are self-starters who depend on themselves rather than the party to win nomination and election. As a result, active campaigning now begins much earlier in the election year than it once did. In our 2000 election surveys, respondents repeatedly expressed displeasure with the campaign's length. The long campaign also numbed people to the point where many tuned it out. A week before the 2000 Republican national convention, only one in five American respondents knew it was only days away Not surprisingly, a large share of those who ended up watching the Republican convention did so only because they stumbled across it while channel surfing.

Americans' disenchantment with campaigns is not the major reason for their declining participation rate, but it is one of them. Unsavory campaigns also leave a sour taste even with those who do vote. In the concluding week of the 2000 campaign, when asked whether the campaign had been "rather depressing, that it hasn't been nearly as good as a campaign should be" or whether it had been "uplifting, that it made [you] feel better about elections," respondents in our survey said by more than two to one that the campaign had been depressing.

Good News, Anyone?

On the network evening newscasts during the 2000 general election, George W. Bush's coverage was 63 percent negative in tone and only 37 percent positive. Al Gore's coverage was no better. A good deal of Bush's coverage suggested that he was not too smart. There were nine such claims in the news for every contrary claim. Gore's coverage was dotted with suggestions he was not all that truthful. Such claims outpaced rebuttals by seventeen to one. (16)

Although the press is often accused of having a liberal bias, its real bias is a preference for the negative. The news was not always so downbeat. When John F. Kennedy and Richard Nixon sought the presidency in 1960, 75 percent of their coverage was favorable in tone and only 25 percent was unfavorable. By the 1980s campaign, however, election news coverage had reached a point where more than half of it was negative. Since then, no major-party presidential nominee has received on balance more positive news than negative news. (17)

This change is attributable in part to the poisonous effect of Vietnam and Watergate on the relationship between the journalist and the politician. A larger influence, however, has been the emergence of an interpretive style of reporting. In the 1960s, this style began to supplant the older descriptive style in which the journalist's main goal was straightforward reporting of the facts of events. Since the facts were often based on what newsmakers had said or done, they had considerable control over the coverage they received. Much of the "good press" that Kennedy and Nixon received in 1960 came from what they themselves said about their candidacy.

On the other hand, interpretive journalism thrusts the reporter into the role of analyst and judge. The journalist gives meaning to a news event by supplying the analytical context. The journalist is thus positioned to give shape to the news in a way that the descriptive style did not allow. The power of the journalist to construct the news is apparent from the extent to which journalistic voices now dominate the coverage. Whereas reporters were once the passive voice behind the news, they now get more time than the newsmakers they cover. On the nightly newscasts, the journalists covering Bush and Gore in 2000 spoke six minutes for every minute the candidates spoke. (18)

Interpretive reporting has unleashed the skepticism traditional in American journalism. The candidates' failings and disputes have been played up; their successes and overtures have been played down. The 1996 Republican nominating race is a case in point. Media analyst Robert Lichter examined the GOP hopefuls' television ads and stump speeches. Over half the ads (56 percent) were positive in tone and nearly two-thirds (66 percent) of the assertions in candidate speeches were positive statements about what they hoped to accomplish if elected. These dimensions of the Republican campaign were seldom mentioned in news reports. The candidates' negative ads and their attacks on opponents filled the news. "Forget about the issues," ABC'S Peter Jennings said of the Republican race, "there is enough mud being tossed around... to keep a health spa supplied for a lifetime." (19)

The tone of news coverage affects people's opinions about candidates. A study of the 1960-1992 campaigns found that a negative impression of presidential candidates increased step by step with the increase in negative coverage. (20) Gallup polls are another indicator of the effect of the increase in negative coverage. Between 1936 and 1968, Barry Goldwater was the only major-party presidential nominee whose public image at the end of the campaign was more negative than positive. Since 1968, in the era of interpretive journalism, a third of the presidential nominees have been perceived unfavorably and another third have had marginally favorable ratings.

Negative news is only one of the reasons Americans are dissatisfied with candidates and campaigns, but it is clearly among those reasons.

What Might Be Done to Reverse the Trend?

The developments that have diminished Americans' interest in election politics are deep rooted and unlikely to be reversed easily or soon. Campaign participation in all its forms, from voting to watching debates, could continue its downward slide.

There are some changes that would retard the trend. Election Day registration is one of them. In the 1950s, 90 percent of Americans lived in a state that closed its registration rolls two or more weeks in advance of the election. (21) The situation is not much different now; today, 87 percent live in a state that shuts down registration two or more weeks before Election Day. (22) In the six states that allow Election Day registration, turnout in 2000 was 1.5 percentage points higher than elsewhere. Although these states (which include Minnesota and Wisconsin) have a history of a high participation rate, all of them moved up in the turnout rankings after implementing same-day registration. Studies indicate that universal same-day registration could boost turnout by as much as 5 percentage points. (23)

Turnout would also increase if polling hours were extended. Amid the uproar over ballot irregularities in Florida in 2000, no commentator saw fit to ask why the polis in that state close at 7:00 P.M. local time. Florida is one of twenty-six states that shut down their polls before 8:00 P.M. Not surprisingly, turnout in these states is several percentage points below that of states where the polls are open until 8:00 P.M. or later. Limits on polling hours go back decades and have been a convenient way to discourage the participation of lower-income workers, who are stuck at their jobs during the day and do not get home in time to cast a ballot.

Turnout might also increase if Election Day were declared a national holiday, as the National Commission on Federal Election Reform has recommended. It is noteworthy that support for this reform, as well as for same-day registration and extended polling hours, is highest among young adults. They are particularly likely to forget to register on time and to require an extra boost to get them to the polls on Election Day They would benefit most from laws that make participation easier. (24)

Structural change by itself will not be enough to turn things around. When turnout dropped sharply in the 1920s, Arthur M. Schlesinger and Erik McKinley Eriksson wrote "no stone should be left unturned" in the effort to lure citizens back to the polls. (25) Today, the schools can do more to give students a decent civic education and help them register so that the first election upon graduation is a step toward lifelong participation. Other entities--the churches, news media, universities, nonprofits, unions, and corporations--must also use their power to assist people in exercising the vote, because if citizens cannot be encouraged to participate more fully the nation will face the far greater challenge of how to maintain self-government when people don't vote.

This article has been adapted with permission of the publisher from The Vanishing Voter by Thomas E. Patterson, Knopf, 2002.


(1.) Federal Elections Commission data.

(2.) Federal Elections Commission data.

(3.) Norrander, B. "Candidate Attrition During the Presidential Nominating System." Paper presented at the Joan Shorenstein Center on the Press, Politics, and Public Policy roundtable, John E Kennedy School of Government, Harvard University, Oct. 16, 2000.

(4.) Quoted in Marks, P. "The Forgotten State: Dearth of Ads Makes Race in Kansas a Snooze." New York Times, Oct. 27, 2000, p. A26.

(5.) Vanishing Voter Survey data.

(6.) National Election Studies data.

(7.) See Wattenberg, M. P. The Decline of American Political Parties, 1952-1996. Cambridge, Mass.: Harvard University Press, 1998.

(8.) Vanishing Voter Survey, Nov. 3-7, 2000.

(9.) National Election Studies data, 1952-2000.

(10.) The question was asked in several surveys. Across all of them, 46 percent of those in the bottom third, compared with 36 percent in the middle third and 32 percent in the upper third, said the elections outcome would have little or no effect on their lives.

(11.) See, for example, West, D. M. Air Wars: Television Advertising in Election Campaigns, 1952-1992. Washington, D.C.: Congressional Quarterly Press, 1993.

(12.) Vanishing Voter survey data.

(13.) Fishel, J. Presidents and Promises: From Campaign Pledge to Presidential Performance. Washington, D.C.: Congressional Quarterly Press, 1985.

(14.) Patterson, T. E. Out of Order. New York: Knopf, 1993, chapter four.

(15.) Patterson, T. E. The Vanishing Voter. New York: Knopf, 2002.

(16.) See Lichter, R. "A Plague on Both Your Parties: Substance and Fairness in TV Election News." Harvard International Journal of Press/Politics, 2001, 6(6), 16; Project for Excellence in Journalism data, Web download Mar. 7, 2002.

(17.) Patterson (1993), chapter one.

(18.) Lichter (2001).

(19.) "The Bad News Campaign." Media Monitor, 1996, 10(2), 3-6.

(20.) Patterson (1993), chapter one.

(21.) Rosenstone, S. J., and Hansen, J. M. Mobilization, Participation, and Democracy in America. New York: Macmillan, 1993.

(22.) Percentage calculated from Federal Election Commission data.

(23.) Fenster, M. J. "The Impact of Allowing Day of Registration Voting on Turnout in the United States from 1960 to 1992." American Politics Quarterly, 1994, 22, 74-87.

(24.) Patterson (2002), chapter six.

(25.) Schlesinger, A. M., and Eriksson, E. M. "The Vanishing Voter." New Republic, Oct. 15, 1924, p. 147. The rest of this paragraph and the article's final sentence are indebted to the final paragraph of their article, even to the choice of the phrasing and some of the words in the final sentence.