American Government › Section F · Public opinion › Topic F.4

The effects of public opinion

Polls are everywhere, but does government actually follow them? The answer differs for campaigns, presidents, Congress and the courts, and it depends on what kind of representation we think officials owe the people.

Should officials follow opinion?

Political theorists describe three models of representation. A delegate votes as constituents wish, so polling helps them learn what that is. A trustee uses independent judgement for the common good, even against local opinion; the Irish-born British politician Edmund Burke defended this view to Bristol's voters in 1774. A politico switches between the two, following opinion on issues voters care deeply about and using judgement on others. Most American officials behave as politicos. They also live in a permanent campaign, governing with the next election in mind, which helps explain why polling continues between elections.

Worked example: trustee or delegate?

A senator's state polls show 60 percent support for a tariff that most economists, and the senator, believe would raise prices. Analyse the choice. 1. Delegate view: voters' wishes should decide, so she votes yes. 2. Trustee view: her duty is to judge the public good, so she votes no and explains why. 3. Intensity: if supporters care little while opponents, such as importing businesses, care a great deal, the electoral risk of voting no is smaller. 4. Timing: a senator elected for six years has time to persuade voters before facing them again. 5. Likely outcome: as a politico she may vote no but campaign to change opinion. 6. Lesson: the same poll can justify different votes under different theories.

Polls and campaigns

Elections are where polls have their clearest effects. Candidates who poll well receive more news coverage, more donations and places on debate stages, since the parties have used polling thresholds to decide who may join primary debates. A bandwagon effect occurs when voters, donors and journalists drift towards the candidate who appears to be winning; an underdog effect is sympathy for the trailing candidate. Early success can snowball: a strong showing in Iowa or New Hampshire raises poll numbers, which attracts money, which pays for advertising. Weak early polling can end a campaign before most voters have paid any attention.

Horse-race coverage

Horse-race coverage reports a campaign like a race: who is ahead, who gained and who slipped, often with a new poll each day. It is popular because audiences want to know who will win, and polls give journalists fresh, measurable news. Critics argue that it crowds out reporting on candidates' records and policies, overstates small or random movements in single polls, and can feed bandwagon effects. Poll averages and aggregators, which combine many surveys, reduce the noise of any single poll, though they cannot remove errors shared by most polls. Media coverage is examined further in "The impact of the media".

Exit polls and early calls

Exit polls interview voters after they cast their ballots and help networks explain who voted for whom and why. Because they produce early estimates, they raised a worry that announcing results could affect people who had not yet voted. With early and mail voting now common, exit pollsters also survey people who voted before election day. Networks project winners from exit polls and counted votes together, and a projection can be wrong: on election night in 2000, networks called Florida for Al Gore, withdrew the call, then called it for George W. Bush and withdrew that too. The result was settled only after weeks of recounts and a Supreme Court ruling.

Worked example: the 1980 early call

Analyse the controversy over the 1980 presidential election. 1. Event: using exit polls, NBC projected Ronald Reagan's victory at about 8:15 p.m. Eastern time, and President Jimmy Carter conceded before polls had closed on the West Coast. 2. Claim: some Democrats argued that western supporters stayed home, costing the party close races further down the ballot. 3. Evidence: researchers found at most a small turnout effect, and Reagan's lead was large. 4. Response: Congress held hearings, and in 1985 the major networks pledged not to project a state's winner before most of its polls had closed. 5. Conclusion: even an uncertain effect on turnout was enough to change media practice, because it touched voters' confidence in fair elections.

Presidents and opinion

Franklin Roosevelt was the first president to commission private polls, and every recent White House has tracked opinion closely. High approval is a resource: a popular president finds Congress more willing to cooperate, and after a clear election win presidents often claim a mandate for their programme. Presidents also try to lead opinion by going public, appealing to citizens over the heads of Congress, a modern form of what Theodore Roosevelt called the bully pulpit. The evidence suggests presidents more often try to move the public than move towards it. Bill Clinton backed loans to Mexico in 1995, and George W. Bush sent extra troops to Iraq in 2007, despite public opposition.

Congress and dynamic representation

Members of Congress face voters often, House members every two years, so they have strong reasons to watch opinion. In an influential 1995 study, James Stimson, Michael MacKuen and Robert Erikson found evidence of dynamic representation: when the public's overall mood shifts in a liberal or conservative direction, policy tends to follow. They traced two routes: members in office adjusting their votes in anticipation of the next election, and elections replacing members with others closer to the new mood. Members also weigh the views of organised groups and donors, whose preferences are clearer and more intense than the general public's, which complicates any simple story of responsiveness.

The courts and public opinion

Federal judges are appointed for life precisely so that they can rule against majority opinion, for example to protect unpopular minorities. Yet research suggests the Supreme Court rarely stays far from sustained national opinion for long. The political scientist Robert Dahl argued in 1957 that because presidents appoint justices and the Senate confirms them, the Court usually reflects the governing coalition of its era. Other studies find modest effects of opinion on the Court's decisions, especially in cases that attract little attention. The Court also needs public acceptance, because it depends on other branches to enforce its rulings.

Policy mood and the thermostat

Individual opinions are noisy, but aggregate opinion moves in sensible ways. Stimson's measure of policy mood, built from hundreds of survey questions, shows the public drifting more liberal or more conservative over periods of years. Christopher Wlezien's thermostatic model, published in 1995, adds that the public reacts to what government does: when spending on a programme rises, demand for more spending falls, and when spending falls, demand rises. Like a thermostat, the public signals that policy is too hot or too cold. This helps explain why parties that win power often see opinion turn against their direction.

Common mistakes to avoid

First, assuming officials simply obey polls; most act as politicos, and presidents often try to change opinion rather than follow it. Second, confusing the bandwagon effect, a drift towards apparent winners, with the underdog effect. Third, claiming exit polls decided an election; evidence of turnout effects is weak, though the worry changed network practice. Fourth, thinking the Supreme Court ignores opinion entirely; studies find it rarely strays far for long. Fifth, reading the thermostatic model as fickleness; the public is reacting to changes in policy.

Key takeaways

Delegates follow constituents, trustees follow their own judgement, and most officials act as politicos in a permanent campaign. Polls shape campaigns through coverage, donations, debate access and bandwagon or underdog effects; horse-race coverage magnifies them. Early network calls in 1980 and 2000 showed the risks of projections. Presidents use approval and going public to lead opinion as much as to follow it. Dynamic representation links shifts in public mood to policy, through members adjusting their votes and elections replacing members. Courts are insulated but rarely far from sustained opinion, and the public reacts to policy like a thermostat.

Flashcards

Flashcard

Delegate

A representative who votes as constituents wish

Polls tell delegates what that is.

Test your understanding

Question 1 of 5

A representative votes against a bill most of her constituents support, saying her own judgement of the public good must decide. Which model is she following?

Question 2 of 5

Which is the best example of a bandwagon effect?

Question 3 of 5

According to the thermostatic model, what happens to public demand for spending on a programme after the government sharply increases that spending?

Question 4 of 5

According to research on dynamic representation, by which two routes does policy in Congress follow shifts in the public's mood?

Question 5 of 5

What changed after networks projected the 1980 election before West Coast polls closed?

← Previous topicNext topic →