Home Up

Elections are fundamental to democratic politics. They are
supposed to be the chief means by which citizens control what their government does. Thus, they are the principle means by which popular sovereignty and majority rule are supposed to work. But do elections ensure that governments will do what their people want? If so, how?

In a small, participatory democracy such as a town meeting the answer is easy: people vote directly on what to do. In a large, complicated society such as the entire United States, however, it is not generally feasible to have everyone vote directly on policies. Representative democracy is the feasible democratic alternative. But how can the people control their representatives to ensure that they will work for what the people want rather than pursue their own ends or respond to special interests?

Democratic theorists suggest several ways in which two-party elections of representatives can or should lead to democratic control of government. The theory of responsible parties holds that parties turn majority preferences into policy.  The theory of responsible party government is based on the idea that elections provide voters with a choice. 

According to this theory, political parties stand for different policies, the voters choose between them, and the winning party carries out its mandate. More precisely, the responsible-party theory assumes that each of the two parties is cohesive and unified; that each takes a clear policy position that differs significantly from the other party's position; that citizens accurately perceive these positions and vote on the basis of them; and that the winning party, when it takes office, does exactly what it said it would do. If all these conditions are met, the party with the more popular policy positions will win and enact its program. This process enables citizens to better articulate their desires to the government and thus make government more responsive to them.

It is not clear, however, that responsible parties actually exist. Even if an election were to work exactly as the responsible-party theory dictates, this would not actually guarantee that the winning party would take policy positions that pleased the voters. It can only guarantee that the winner's stand is less unpopular than the loser's, which may not make it particularly popular at all.

Under the terms of the theory, crucial decisions about what the parties will stand for are still made by party leaders or interest groups instead of ordinary citizens. Moreover, the conditions under which responsible-party theories are supposed to work are not met in the United States. As we have seen, the Republican and Democratic parties are neither unified nor cohesive. 

Although the parties sometimes take clear stands, their policy positions are sometimes deliberately ambiguous. The parties may avoid unpopular or extreme positions, or take similar policy positions. Furthermore, voters do not vote solely on the issues, and parties do not always keep their promises.

Theories of electoral competition offer a different and less obvious means of democratic control of government. Once  again, the assumption is that the two political parties each take clear, unified  stands on the issues of the day, that citizens vote on the basis of the issues, and that the winning party does what it promises to do.

Electoral-competition theories have no expectation or desire, however, that the parties' stands will be sharply different from each other. Instead, the theory holds that both parties compete for votes by taking the most popular positions that they can. Both parties may well end up championing the same policies--those favored by the most voters. 

According to the theory, it does not matter which party wins; the winner enacts the platform that the voters want. Democracy is assured by the hidden hand of competition. The theory holds that parties will take positions near the median of public opinion, where exactly one half of the voters are more liberal and one half are more conservative. This  situation would be the ideal democratic outcome, because, by the logic of the theory, a majority of citizens would prefer policies at the median over policies anywhere else on the continuum.

The conditions assumed by electoral-competition theories are not likely to correspond exactly to what happens in the real world. Electoral competition ensures democratic control only if parties are unified, take clear stands, and are motivated solely by vote-seeking. The theory breaks down if parties are fragmented, care about policies without regard to majority preferences, or seek contributors' dollars rather than citizens' votes. The theory also assumes (perhaps incorrectly) that voters consider only issues in making their choices and that they know where the parties stand on the issues.

Finally, the theory of retrospective voting offers a mechanism for democratic control of government. This is the concept that citizens make voting decisions on the basis of retrospective (backward-looking) judgments about how well incumbent officials have done in the past, rewarding success with reelection and punishing failure by throwing the incumbents out. 

Voters don't bother to form preferences about complex issues and don't necessarily trust parties' promises. Because politicians are ambitious, they try to anticipate what the public wants and seek to accomplish it. This reward-and-punishment version of democratic control requires little of voters--no elaborate policy preferences, no study of campaign platforms--just judgments of how well or how badly things have been going. It allows time for deliberation, and it lets leaders try out experimental or temporarily unpopular policies, as long as the results work out well and please the public in time for the next election.

Nonetheless, reward and punishment may be a rather blunt instrument, getting rid of bad political leaders only after (not before) disasters happen, without guaranteeing that the next group of leaders will be any better. Moreover, the reward-and-punishment version focuses only on the most crucial issues and may leave room for unpopular policies on issues that are less visible. It may also encourage politicians to produce deceptively happy results around election day that fade later.

Each of the three theories only partly explains what happens in American elections. None of  them--alone or together--works well enough to guarantee perfectly democratic outcomes from U.S. elections. One flaw common to all of them is that these processes can bring about government responsiveness to all citizens only if all citizens vote. However, millions of Americans cannot or do not go to the polls. Their voices are not clearly heard and thus political equality is not achieved.

Another problem besetting all three theories is that money, organizational resources, and  activism--not just citizens' policy preferences and vote--influence the stands that parties take and the outcomes of elections. This means that money givers, activists, and the leaders of organized groups have more influence than ordinary citizens do. Thus, political equality is not realized.

Political Participation

 Political participation refers to political activity by individual citizens. It includes participation in demonstrations, boycotts, social movements, and conventional participation (writing letters, contacting officials, going to meetings, working on campaigns, and giving money). The most basic form of modern political participation, however, and the one that plays the most central part in theories of democratic control through elections, is the act of voting. 

In the early years of the United States, the right to vote was quite restricted. Most people could not vote at all. Slaves, native Americans, women, and those not owning property, for example, were not allowed to vote. People could vote for their state legislators and members of the House of Representatives but other offices in the national government were elected by means other than direct popular vote.

Furthermore, it was not always easy to get to the polls. The whole process was so restricted,  difficult, and confusing that only 11 percent of the Americans eligible to vote actually cast ballots in the first presidential election.

One of the most important developments in the political history of the United States, and an essential part of the struggle for democracy, has been the expansion of the right to vote. The extension of the franchise has been a lengthy process, lasting 200 years. The first barriers to fall were those concerning property and religion. By 1829, property and religious requirements had been dropped in all states except North Carolina and Virginia. That left universal white male suffrage, or ability to vote, firmly in place in the United States.

Expanding the suffrage to include blacks and women was more difficult and painful. Ironically, universal white male suffrage was often accompanied by withdrawal of voting rights from black freedmen. It took the Civil War to free the slaves, and the Fifteenth Amendment to the U.S. Constitution (1870) to formally extend the right to vote to all blacks. Even so, blacks were effectively disfranchised in the South by the end of the Nineteenth century and were generally kept from the polls until the civil rights movement and the Voting Rights Act of 1965.

Democracy was further expanded when women won the right to vote with the Nineteenth Amendment to the Constitution (1920), after a long political battle. Residents of the District of Columbia were allowed to vote for president (but not Congress) after 1961, and persons 18 to 20 years of age gained the franchise in 1971. The result of all these changes has been an enormous increase in the proportion of Americans who are legally eligible to vote.

Another trend is toward the direct election of government officials, overcoming the insulation of officeholders from the public. The development of a two-party system made a big difference. By the time of the Jefferson-Adams election of 1800, most state legislatures allowed popular voting for presidential electors.

That and the practice of electors pledging to vote for a particular presidential candidate, together with the rise of two political parties to focus voter choices, meant that the voters could choose their president more or less directly.

The more recent innovation of nominating presidential candidates in national conventions (by 1840) and of electing convention delegates directly in primaries probably also increased democratic control of government. The popular election of U.S. senators did not come until 1913, with the ratification of the Seventeenth Amendment to the Constitution.

Proportionally fewer people participate in politics in the U.S. today than did so during most of the nineteenth century. Since 1912, only about 55 to 65 percent of eligible Americans have voted in presidential elections. Still fewer Americans participate in other elections; 40 to 50 percent take part in off-year (non-presidential year) congressional elections and as few as 10 to 20 percent vote in minor local elections.

In recent years, the turnout rate has dropped to the low end of those ranges. Moreover, voting turnout in the United States is exceptionally low compared with other modern industrialized countries, where turnout rates of 80 percent are common. Most observers consider voter apathy a serious problem for democracy, particularly since those who vote tend to be different from those who do not. Nonvoters do not get an equal voice in political choices, therefore the principle of political equality is violated.

Why do so few Americans participate in elections? Scholars usually disagree on the answer to this question but several factors definitely seem to be at work. First, the U.S. has relatively restrictive voting registration laws. Only those who take the initiative to register before an election are permitted to vote in it. Sometimes registration is made difficult, with limited locations, limit d hours, and requirements that registration occur long before election day.

Political equality and popular sovereignty in the United States could be enhanced if voting were made easier. Some of the possible ways of increasing voting registration include the national use of postcard registration, same-day registration, the use of extended voting periods, or the broadened right of absentee voting.

Second, the nature of the U.S. two-party system limits voters' choices. Countries with proportional representation and multi-party systems have averaged an 83 percent turnout rate, whereas single-district, plurality vote countries (which usually have just two parties) have had a voter turnout rate that is closer to 70 percent. Also, in contrast to most European countries, the United States does not have a Workers' party to mobilize blue-collar workers and poor people.

Third, changes in eligibility rules have affected turnout rates. Turnout as a proportion of eligible voters dropped sharply just after women were enfranchised in 1920, because at first women were less likely to vote than men. Similarly, turnout percentages dropped a bit after persons 18 to 20 years of age won the vote in 1971, because young people do not participate as much as older people. But the enfranchisement of young voters, which reduced turnout, is partly balanced by rising education levels and by somewhat easier registration requirements.

The alienation and apathy about politics that many Americans  felt as a result of the Vietnam War, urban unrest, the Watergate and Iran-Contra scandals, and various economic troubles probably also contributed to recent declines in turnout. Furthermore, turnout may have been hurt by the failure of either of the major parties in the 1980s and 1990s to register poor black and Hispanic people.

Despite the low voter turnout levels in the United States, Americans are actually more likely than people in other countries to participate actively in campaigns. Exactly why Americans vote less but campaign more than citizens elsewhere is something of a mystery.

People with higher than average incomes and who are relatively well educated are more likely to be politically active than other persons. This difference in participation rates has important effects on the working of democracy, because it undermines political equality in that some kinds of people have more representation and more influence than others.

There is a class bias in both political participation and voting. The very young, the unemployed, and Hispanics all have especially low turnout rates. In the past, fewer black people voted than whites, but now the proportions are much more nearly equal. Blacks are at least equally likely to vote, and sometimes more so, than whites of similar background (same education and income levels). In general, however, statistical analyses indicate that the crucial factor in voter turnout is level of formal education.

How much difference do participation biases make? Some observers have argued that it doesn't matter if many people don't vote, because the preferences of nonvoters aren't much different from the preferences of voters. Other observers say that low voter turnout has a positive benefit for democracy, because less educated people are more easily swayed by political demagoguery.

However, we should not be too quick to accept these arguments. Even if the expressed preferences of nonvoters about politics are not distinctive from those of voters, their objective circumstances differ. Hispanics, the young, and people of low income would presumably benefit from particular kinds of government policies that are of less interest to other citizens. A political system that included and mobilized these people might produce policies more along the lines of European social welfare systems. Broader participation in U.S. elections would increase popular sovereignty and political equality, and thus would contribute to democracy.


The major party candidates for president are drawn from a rather small pool of potential candidates. For any given election, only a handful of candidates are serious possibilities. Almost all serious candidates for president have been middle-aged or elderly white men with strong educational backgrounds, fairly high incomes, and substantial experience as public figures--usually as government officials.

In recent years, the presidency has been practically monopolized by governors, vice-presidents, and U.S. senators. Serious candidates for president almost invariably represent mainstream American values and policy preferences and must be acceptable to the business community.

A senator or a governor who wants to run for president usually begins at least two or three years before the election by testing the waters, asking friends and financial backers if they will support a run, and observing how people react to a potential candidacy. The pre-candidate may commission a national poll to check for name recognition and image. He or she may assemble an informal organization to round up private endorsements, commitments, and financial contributions, perhaps setting up private PACs to gather money. If all goes well at this early stage, the presidential aspirant becomes more serious, assembling a group of close advisors, formulating strategy, raising large amounts of money, and putting together organizations in
key states.

Party nominees for president are chosen every four years at national party conventions, made up of state delegations from around the country. Since the 1970s, most of the delegates to the convention have been chosen in state primary elections, with direct voting by rank-and-file party members.

The popularly elected delegates are supplemented by super delegates, usually members of congress and local officials, who become convention delegates by appointment. Relatively few states now use caucus systems, in which active party members and officials choose delegates to state conventions, where, in turn, the national convention delegates are chosen. The smoke-filled rooms and deal making among party leaders that once characterized national conventions are mostly a thing of the past.

Candidates who establish momentum by winning early primaries and caucuses get press attention, financial contributions, and better standings in the polls, as voters and contributors decide they are viable candidates and must have some merit if people in other states have supported them.

Four main factors affect candidates' success in gathering delegates: the general attractiveness of the candidate, how the media and public opinion polls judge the candidate's viability, the candidate's organizational strength, and the amount of money a candidate has to spend (on television advertising, organization, and travel).

In recent years, most successful presidential nominees have won primary after primary, gathering in the supporters of losing candidates who bowed out, and have gone on to the convention with a substantial plurality of delegates. When there is a clear front-runner, the national convention generally becomes a coronation ceremony, in which delegates ratify the selection of the leading candidate, accept a ticket-balancing choice for the vice-presidency, and put on a colorful show for the media and the country.

Republican and Democratic delegates to the national conventions tend to be different from each other, reflecting the nature of their party coalitions. Delegates to both conventions are predominately white and financially well off, but the Democratic convention typically includes many more black, Hispanic, female, and working-class delegates.

In addition, convention delegates of the two parties tend to differ substantially on certain political issues, with Democratic delegates tending to be more liberal and Republican delegates more conservative than the average American citizen. Only occasionally nowadays is there a real contest at the national convention, in which delegates may have to decide between their professional minds and their purist hearts whether to nominate a likely winner or a candidate who stands for ideas that they hold dear.

The evidence from polls indicates that in virtually all recent conventions, the party nominee has been the candidate who was most popular with rank-and-file party identifiers in the nation as a whole. Nonetheless, the differences between the delegates of the two parties tend to push the nominees and the positions of the parties apart, as predicted by the responsible-party model rather than electoral-competition theories.

What does all this have to do with democratic control of government? The nomination process is fairly good at producing candidates who are personally attractive to voters and who take stands with wide popular appeal, as electoral- competition theories seek. Moreover, the sharp differences between Republican and Democratic convention delegates indicate that Republican and Democratic nominees tend to differ in certain systematic ways, in accordance with responsible-party theories.

The crucial role of party activists and contributors in selecting candidates means that nominees and their policy  stands are chosen partly to appeal to party elites and financial contributors, rather than to ordinary voters. Thus, neither party's nominee may represent what ordinary citizens want exactly, preventing a perfectly democratic outcome.

The path to re-nomination for incumbent presidents is different than the road taken by challengers. Although incumbent presidents must also enter and win primaries, they have the machinery of government working for them. They campaign on the job, taking credit for policy successes while discounting failures. Winning the nomination is usually easy for an incumbent president. 

What kinds of information do voters get in campaigns? Some of it concerns issues. In accordance with electoral- competition theories, both the Republican and Democratic candidates usually try to appeal to the average voter by taking similar, popular stands on policy, especially foreign policy. Nonetheless, as responsible-party theories suggest, Republican and Democratic candidates usually do differ systematically from each other on certain issues.

On all these issues, the Democratic candidate tends to take a more liberal stand than the Republican, just as Democratic party identifiers, activists, contributors, and convention delegates tend to be more liberal than their Republican counterparts. Presidential candidates usually do not say a great deal about specific issues. They tend to be vague in order to avoid offending voters who disagree.

Often candidates talk about past performance and future goals, with the "out" party blaming the "in" party for wars, recessions, and other calamities. Incumbent presidents, of course, can act in ways that accurately or inaccurately suggest successful performance. They can try to schedule recessions for off years, pumping up the economy in time for reelection (as in 1972 and 1984), or they can make dramatic foreign policy moves just before election day. This strategy does not always work, however, and the opposition sometimes fights back effectively.

Elections give voters a chance to learn about the real or alleged personal characteristics of the candidates. Even when the candidates are talking about something else, they can give an impression of either competence or incompetence. Candidates come across as warm or cold or can seem strong or weak. The sparse and ambiguous treatment of policy issues in campaigns, and the emphasis on past performance and personal competence, fit better with ideas about electoral reward and punishment than they do with the responsible parties or issue-oriented electoral competition models.

Evaluations of candidate personalities are indirectly related to democratic control of government. Obviously, it is useful for voters to pick presidents who possess competence, warmth, and strength, and citizens may be more skillful at judging people than at figuring out complicated policy issues. Nonetheless, voters can be fooled by dirty tricks or slick advertising that sells presidential candidates' personalities.

Moreover, the focus on personal imagery may distract attention from policy stands. If candidates who favor unpopular policies are elected on the basis of attractive personal images, democratic control of policymaking is weakened. By the purchase of advertising and the hiring of smart consultants, money may, in effect, overcome the popular will.

Presidential campaigns cost enormous amounts of money. Over the years, campaign costs have increased steadily and rapidly. Observers disagree about whether campaign expenditures are excessive. The $210.7 million in direct spending on all presidential candidates in 1988, for example, translates to less than $1 for each person in the country. However, the growth in campaign spending is real; it has far outstripped inflation or increases in the price of advertising.

Where does campaign money come from? Substantial amounts of money come from the federal treasury, paid by the taxpayers-- some $46 million for the general election in 1988. Taxpayers can check off a box on their tax form and thereby authorize a $1 contribution from public funds. The government uses this money to match small private contributions to candidates. This public money, however, accounted for only about 31 percent of campaign spending in 1988. The rest had to come from contributions by individuals and by PACs (political action committees) set up by businesses, labor unions, and other special interest groups.

Presidential candidates accepting federal money must follow certain rules. Individual contributors can give no more than $1,000 directly to any one candidate in one election cycle and $5,000 to any one PAC per year. But independent expenditures on behalf of a candidate or a party are unlimited, and people can contribute to as many different PACs as they want.

Total individual contributions to all candidates, PACs, and party committees are limited to $25,000 per year. PACs are limited to a $5,000 contribution per candidate per election, and a given PAC can receive no more than $5,000 from any one person. Parties are limited through a complicated formula in their contributions to, or on behalf of, candidates. Nonetheless, party-building activities at the state and local levels have no limit. Presidential candidates who receive public funding have to report all expenditures and contributions.

Much of the money comes from individuals and organizations that want something from government: tax benefits, regulatory relief, or military contracts. It is widely believed, though difficult to prove, that contributors often get something from their contributions. The point is not that politicians take outright bribes, but rather that cozy relationships develop between politicians and major money givers. Contributors gain access--a friendly hearing--from those whom they have helped to win office. Contributors' money increases the chances of victory for like-minded politicians who hold sympathetic views and can be counted on to do the right thing without any need for pressure.

Campaign contributors are different from average citizens in that they have special interests of their own. Many kinds of people are left out, with the result of political inequality. Those who are well organized or have a lot of money and are willing to spend it on politics have a better chance of influencing policy than do ordinary citizens. This is a major problem for the working of democracy in the United States.

Election Outcomes

  After the parties and candidates have presented their choices, the voters decide. As one might expect, it is not easy to explain why people vote the way they do. Scholars check what sorts of people (with what party affiliations, what opinions about issues, and what perceptions of candidates' personal qualities) tend to vote for one candidate rather than another. Even with the help of sophisticated methods, it is still difficult to be sure what causes people to behave in different ways. There is general agreement, however, that peoples' feelings about parties, candidates, and issues all have substantial effects on how people vote.

People's socioeconomic status, religion, and ethnic backgrounds are significantly related to how they vote. Since the 1930s, for example, black people, Jews, and lower- income citizens have tended to vote for Democrats, while most white, Anglo-Saxon, Protestants and wealthy persons have voted for Republicans.

This pattern usually holds true for both presidential and Congressional elections. To some extent, these social patterns reflect long-term attachments to political parties. As we discussed in previous chapters, most Americans say that they consider themselves Republicans or Democrats. Party loyalties vary among different groups of the population, often because of past or present differences between the parties on policy issues.

The relationship between party loyalties and certain policy preferences--especially those related to New Deal social welfare issue--means that party-line voting can accomplish some of the same things that issue voting is supposed to do in responsible-party theories. Party cues help people vote for candidates who are close to them on the issues. 

Because party loyalties are so important, some scholars speak of a normal vote, that is, a proportion of the votes that each party would win if only party and nothing else affected voting decisions. If nothing else were going on, the majority party, which since 1932 has been the Democratic Party, would win every election. In fact, this has been nearly true of congressional elections but not true of presidential elections. (Democrats have won only three presidential elections since 1952.)

Part of the explanation for this phenomenon has to do with the fact that party loyalties in general have been weakening. Even though strong party identifiers have continued to vote for their party, there have been fewer strong party identifiers since about 1968. Moreover, while the number of whites in the South who consider themselves Democrats declined sharply because of issues on civil rights, defense, and other issues, some continued for a while to call themselves Democrats but voted for Republican presidential candidates.

Finally, another reason why presidential-election outcomes have not simply reflected the party balance and why the Democrats have lost so often is that voters pay a lot of attention to their perceptions of the personal characteristics of candidates. They vote heavily for candidates who have experience, appear strong and decisive, and seem to display personal warmth. Democratic presidential candidates have not always measured up to these criteria. 

Voters also pay attention to issues, even beyond the party- cleavage issues that are reflected in party loyalties. Sometimes this means choosing between different policy proposals. More often, however, issue voting has meant retrospective voting, with voters making judgments about the past, especially on major questions about war or peace and the state of the economy. Voters tend to reward the incumbent party for what they see as good times and to punish it for what they consider bad times.

One way to see the regular, systematic effects of electoral reward and punishment is to look at the number of seats in Congress that the incumbent party wins in off-year elections, when there are no presidential candidates with unique personalities to complicate voting choices. Changes in the economy one year before an off-year election are strong predictors of how well the incumbent party will do. When personal incomes rise, voters are much more likely to vote for the incumbent than when their incomes go down.

When we ask whether elections make a difference, we are likely to look for dramatic cases in which big issues were at stake between the two parties and the election ushered in a new era of policymaking and party realignment. Such decisive elections, however, are quite rare. In most elections, it does not make a dramatic difference whether the Republicans or the Democrats win. 

Nonetheless, regardless whether the final choice is critical, elections do matter, in several different ways that bear upon questions of democratic control. In terms of the responsible-party governmental theory, for example, the fact that Republicans tend to be more conservative than Democrats on a number of economic and social issues provides voters with a measure of democratic control by enabling them to detect differences and make choices.

With the theory of electoral punishment, the control voters exercise is expressed by either reelecting successful incumbents or throwing failures out of office.

Finally, as electoral competition theories point out, elections force parties to compete by nominating centrist candidates and by taking similar issue stands that are close to what most Americans want. This, in fact, may be the chief way in which citizens' policy preferences affect what their government does.

Although elections help make the public's voice heard, they do not bring about perfect democracy. Two key reasons for this are the limited and biased participation of citizens, and the crucial role of money and activists in affecting election outcomes. Both of these reasons impair political equality by giving some sorts of people more political influence than others. The nature of our political parties, of course, is related to both of these reasons.