Economic Influence on Public Sentiment and Voter Behavior
Introduction
This chapter explores the subject of economic influence on citizen opinions and voter behavior. We consider the presidential vote, Congressional House and Senate elections, the voter participation rate, macropartisanship, consumer sentiment, and the social happiness index. The chapter explains three theories of economic influence on voter behavior. They are the responsibility hypothesis, the issue hypothesis, and the salient goal hypothesis. The chapter also discusses electoral efficiency in relation to economic influence on public attitudes. In addition, we consider the political capital effect. This is the impact of the economy on presidential approval and the corresponding spillover effect on public support for the president’s political and legislative agenda. Finally, we examine some important noneconomic effects on voter sentiment.
Macroeconomic Accountability and Electoral Efficiency
Voters hold the president and the in-party to the White House accountable for the health of the economy. Macroeconomic events, however, are not solely attributable to the actions of the president and the in-party. The president does not fully determine macroeconomic policies or the economic outcomes that arise from those policies.
The political interaction among the president and the Congress in association with the right and left political parties determines fiscal policy, while the central bank directly determines monetary policy. Several other factors also affect the economy, such as energy costs, commercial technology, macroeconomic policy lag and uncertainty, business cycle momentum, exogenous shocks, special interest influence on policymakers, and the frequent occurrence of partisan gridlock among the fiscal policymakers. Additionally, war, the weather, and natural disasters impact the economy.
Regardless of the various influences on the economy, voters generally hold the president and the in-party to the White House accountable for what happens. The president receives strong public approval if high economic growth occurs. This high approval rating tends to happen regardless of whether the incumbent’s policies create the favorable macroeconomic outcome. The president receives low approval from the electorate if a weak economy takes place. This tends to occur regardless of whether the slow economy is attributable to the president’s policy actions.
Democratic or electoral efficiency occurs if citizens elect the political candidate to the presidency whose policy agenda yields the greatest long-term economic benefit for society. Actual voter behavior, however, inevitably exhibits some amount of inefficiency. This occurs because of imperfect information and imperfect rationality among voters. Electoral inefficiency takes place if citizens elect a politician whose policies do not yield the greatest long-run economic benefit for society. This happens if voters support a president’s policy that ends up making the economy worse. Citizens, for example, may be fooled into supporting a policy of macroeconomic overstimulation that creates only a temporary improvement in unemployment that comes at the cost of greater inflation in the long run.
Democratic inefficiency also arises if voters approve of the president because of a strong economy that is due to factors other than the president’s policies. Suppose high economic growth takes place because of commercial technological advancements or cheap energy costs rather than the macroeconomic program of the president. Presidential approval, nevertheless, is likely to be high, as voters give credit to the president for the economic outcome. Electoral inefficiency similarly develops if the public disapproves of a president’s policy that improves the economy in the long run, but with short-term adverse effects. Presidential approval, for example, is likely to decline if contractionary macroeconomic policy causes a temporary rise in unemployment, but ultimately makes the economy better off because of a permanent decline in inflation.
Electoral inefficiency also happens if presidential approval decreases because of a slow economy that is caused by factors other than the administration’s policies. Suppose a recession transpires because of an energy shock or a financial crisis rather than poor macroeconomic management by the president and the in-party. Presidential approval, nevertheless, falls, despite the weak economy being unrelated to the administration’s actions.
Electoral inefficiency, in other words, occurs if the public praises or condemns the president for macroeconomic events that the administration’s policies do not create. Voters tend to judge an administration for economic conditions that are unrelated to the president’s actions. Some ambiguity, however, exists concerning the part of economic performance that is attributable to the president versus the portion of economic performance that is from other causes. Macroeconomic influence on voter opinions is not a perfectly efficient process. Inefficiency occurs to the degree that voters cast votes based on economic events that transpire, without recognizing if the in-party’s policies caused the outcomes or not.
A vigilant media and farsightedness among politicians and other opinion leaders can help to promote a rationally informed economic society and greater electoral efficiency. Some degree of democratic inefficiency, however, seems inevitable. Some citizens cast votes based on misinformation, habit, emotion, bias, or other irrationalities. This might be reduced but probably cannot be completely eliminated because of human nature. Nevertheless, to the extent that voter behavior becomes more rationally informed on the economy and other issues is the degree to which electoral efficiency increases.
Economic and Noneconomic Effects on Presidential Elections and Presidential Approval
Presidential job approval is an estimate of the percentage of the public that approves of the incumbent’s handling of his or her job. Several organizations, such as the Gallup Poll, conduct regular surveys of presidential approval and various other measures of public opinion on the incumbent. Besides presidential approval, the most substantive public judgment of the incumbent is presidential elections. The in-party presidential vote share is the percentage of the two-party vote in favor of the in-party candidate in a presidential election. Although not identical measurements, both presidential approval and the in-party presidential vote have parallel characteristics. Both indicators reflect public perceptions on the president’s effectiveness. Accordingly, some of the explanatory variables are similar for both presidential approval and the in-party presidential vote share, while other determinants are different between the two indicators.
Economic Influence on Presidential Elections and Incumbent Popularity
The economy affects presidential job approval and presidential election outcomes in a similar manner. Three main theories have been proposed to explain economic influence on government popularity, presidential elections, and presidential job approval (Carlsen 2000). The three models are the responsibility hypothesis, the issue hypothesis, and the salient goal hypothesis.
Responsibility Hypothesis of Economic Influence on Voter Behavior
The conventional view of economic influence on voter behavior has been called the responsibility hypothesis (Carlsen 2000) or the score model (Swank 1990). This theory asserts that a strong economy causes high presidential approval and a high in-party vote share in a presidential election. A strong economy is characterized by low unemployment, high real economic growth, and low inflation. A slow economy causes low incumbent approval and a high vote share for the out-party candidate in a presidential election. The responsibility theory of voter behavior underlies the median voter model and the electoral PBC effect that are discussed in previous chapters.
The partisan PBC effect, as presented in Chapter 7, is also theoretically in line with the responsibility hypothesis of economic influence on voter actions. An implication occurs regarding the partisan PBC effect and the responsibility hypothesis. Assume the median voter model holds. The median voter theory asserts that reelection votes are highest if the in-party adopts policy that matches with the median voter’s preference. Divergence of economic policy from the median voter’s most preferred outcome causes presidential reelection votes for the in-party to be lower.
Suppose the president adopts a partisan economic policy instead of appealing to the median voter. The median voter’s preference lies between the two partisan preferences of the left and right political parties. The conservative preference occurs to the political right of the median preference. The liberal policy preference occurs to the political left of the median voter. The in-party’s probability of reelection to the White House declines if the president adopts policy based on partisan goals instead of the median voter’s opinion.
Liberal presidential administrations are relatively unemployment-averse according to the partisan influence model. Expansionary policies that seek to reduce unemployment, however, can cause higher inflation if macroeconomic over-stimulus occurs. Liberal presidencies consequently risk losing the White House if high inflation arises prior to a presidential election. Citizens might choose to vote for a conservative inflation-averse president instead.
Alternatively, suppose the conservative party is in the White House. Conservative stabilization policy is more inflation-averse than the median voter’s preference according to the partisan influence model. In pursuit of low inflation, unemployment may worsen because of the short-run inflation–unemployment trade-off. If this takes place prior to a presidential election, then the conservative party might lose reelection because the contractionary policy takes the economy too far to the political right of the median voter’s preference. A conservative presidency is at risk to lose reelection because of the short-run effect of high unemployment from disinflationary policies. A conservative administration that focuses too much on inflation might induce voters to prefer a more liberal unemployment-averse president to be elected to the White House.
Another type of economic influence on the president may be referred to as the political capital effect. This is the impact of the economy on presidential approval and the presidency’s ability to achieve its political and legislative goals. A strong economy usually causes presidential popularity to be high. This creates political capital or high public support for the administration to fulfill its political goals on issues like immigration, tax reform, health care, international trade, national defense, and the environment.
If presidential approval is high because of a strong economy, then the Congress is inclined to support the president’s legislative and political program. Congressional opposition to the proposals of a popular president adversely affects the reelection prospects of legislators. Citizens might vote out of office those legislators who disagree with the plans of a popular president. A slow economy, on the other hand, causes low presidential approval. Consequently, the political capital or public support for the administration to accomplish its political and legislative agenda is likely to be weak. Congress may be inclined to challenge or oppose the proposals of an unpopular president.
Evidence on Time Consistency: The Presidential Vote versus Presidential Approval
The concept of dynamic macroeconomic consistency is compatible with the responsibility hypothesis of economic influence on voter behavior. Macroeconomic consistency occurs if the median voter’s unemployment target equals the natural unemployment rate. As discussed in Chapter 2, the efficient level of unemployment is the natural unemployment rate.
If the unemployment target equals the natural rate, the median voter is rationally informed and farsighted. The median voter’s preferred macroeconomic outcome is dynamically consistent because the preference is in line with what the economy can sustain. If the median voter is misinformed or shortsighted, then the median unemployment target is less than the natural rate. This preference is dynamically inconsistent because the preferred outcome is not in line with what the economy can maintain.
The research on presidential elections and presidential approval yields somewhat mixed results on whether the median voter’s unemployment target is dynamically consistent or dynamically inconsistent. Some empirical findings on presidential elections imply that the median voter’s macroeconomic preference is shortsighted and dynamically inconsistent (e.g., Fox 2013). This suggest citizens are willing to embrace a short-term improvement in the economy that comes at the cost of greater long-term inflation. The possibility of macroeconomic inconsistency by the median voter enables the electoral cycle effect to occur, as discussed in Chapter 6.
Some of the research in the presidential approval literature seems to imply the reverse result. These empirical findings suggest the median macroeconomic preference may be farsighted and dynamically consistent (Fox 2003, 2009; Smyth and Dua 1989). If voters are farsighted, they oppose opportunistic policies by the president to temporarily boost economic growth which comes at the cost of higher long-term inflation. If the median voter’s unemployment target is dynamically consistent, then opportunistic policies cause the presidential reelection vote share for the in-party to decrease rather than increase. Voters are not fooled into supporting opportunistic policies. Instead, voters penalize presidents who attempt such strategies with lower presidential approval and lower reelection votes. Overall, the empirical results are mixed on whether the median voter’s preference is dynamically consistent or inconsistent. However, if voters lack understanding on how the macroeconomy works, then the risk of dynamic inconsistency is at least present.
Clientele and Salient Goal Hypotheses of Economic Influence on Voters
The second theory of economic influence on presidential elections and presidential approval has been called the clientele hypothesis (Carlsen 2000) or the issue model or the partisan vote model (Swank 1990). This theory asserts that voter behavior emphasizes the different macroeconomic priorities of the left party and the right party.
The issue model maintains that citizens cast votes based on which of the two main parties is best suited to resolve the more important economic problem at a particular time. This theory of voting behavior is compatible with the partisan influence model that liberal administrations are relatively unemployment-averse and conservative presidencies are relatively inflation-averse. According to the issue model, citizens vote for the left party if unemployment is high compared to inflation. Voters prefer liberal presidencies if unemployment is high because of the liberal party’s reputation of unemployment aversion. Citizens vote for the conservative party in presidential elections if inflation is high compared to unemployment. This occurs because of the conservative party’s reputation of emphasizing low inflation in macroeconomic policy.
The issue model of voting behavior is supported by the empirical findings of Swank (1995). His analysis found that high unemployment causes public approval to increase for Democratic presidencies, while high inflation causes popularity to increase for Republican administrations.
The third model of economic influence on presidential approval and presidential elections has been called the salient goal hypothesis (Carlsen 2000). This theory of voter behavior is also compatible with the partisan PBC effect that conservative presidencies focus on reducing inflation, while liberal presidencies focus on reducing unemployment. The salient goal model asserts that voters judge presidencies by how well they attain their partisan goals. Voters approve of presidencies that succeed in their partisan economic objectives. Citizens disapprove of incumbencies that fail in their partisan priorities.
Voters weigh high unemployment more heavily against Democratic presidencies. Citizens disapprove of Democratic presidencies that fail in their liberal objective of unemployment aversion. Citizens weigh high inflation more heavily against Republican administrations. Citizens disapprove of Republican presidencies that fail in their partisan macroeconomic goal of inflation aversion.
Noneconomic Effects on Presidential Elections and Incumbent Popularity
Several noneconomic considerations weigh on voter opinions of the president. One important determinant is war. Two types of war effects impact voter attitudes. They are the soldier casualty effect and the war rally effect. The casualty effect is the adverse influence of soldier deaths on citizen sentiment of the president. The larger the number of soldier casualties, the lower the incumbent approval rating and the lower the vote share in favor of the in-party in a presidential election.
The soldier casualty effect is more adverse against presidents who are war initiators compared to presidents who are war inheritors (Fox 2013). A war initiator is a president who starts a major military conflict that ends up being long and costly in terms of military fatalities over time. A war inheritor is the subsequent president of the opposing political party who inherits a long military conflict from a war-initiator administration as the result of a presidential election. Voters penalize war-initiator presidencies with a greater casualty effect on presidential approval and reelection votes than war-inheritor presidents.
An example of the war initiator effect was G.W. Bush and the Iraq War. As soldier casualties mounted in the Iraq War, presidential approval for G.W. Bush gradually declined, especially during his second term in office. This casualty effect became a major issue that contributed to the Republican loss of the White House in the 2008 election. An example of the war inheritor effect was Barrack Obama and the Iraq War that he inherited from G.W. Bush. Iraq war casualty deaths had a much smaller negative impact on presidential approval for Obama than G.W. Bush. The war-casualty effect also had little or no adverse impact on the 2012 presidential reelection victory for Obama, since he was a war inheritor.
A further example of the war-initiator and war-inheritor effects was the impact of soldier casualties on the approval ratings of President Johnson and then President Nixon. Johnson was the war-initiator, while Nixon was the war-inheritor. Vietnam War causalities severely hurt Johnson’s job approval to the degree that he chose not to run for reelection in 1968. Subsequently, Nixon from the opposing political party won the White House. Since Nixon was a war-inheritor, Vietnam war casualties had a much smaller negative impact on his popularity and he easily won his reelection in 1972, despite the Vietnam War not being over.
The war rally effect is the second type of war-related influence on voter sentiment of the incumbent. War rallies are major war-related events that create a transitory boost in presidential approval. War rallies create temporary spikes in presidential popularity connected with major military victories and other major war-related events. One type of war rally is the nationalistic support for a president at the start of a war. In the G.W. Bush presidency, three war rally effects caused transitory boosts in presidential popularity. The effects were the 9/11 terrorist attack, the start of the Iraq War, and the capture of Saddam Hussein (Fox 2009). War rallies cause presidential approval to temporarily spike upward but then gradually dissipate. In contrast, the accumulation of war casualties causes presidential approval to incrementally worsen over time.
Another political influence on the in-party presidential vote share is political party duration. This refers to the length of time (the number of consecutive 4-year presidential terms) that the in-party occupies the White House. After a political party controls the White House for two or more consecutive terms, voters increasingly prefer a change of the political party in the presidency (Fair 2009).
The party-duration effect on the in-party presidential vote share does not generally occur after just one term in office. Instead, a positive incumbency effect takes place in favor of the incumbent in presidential elections after just one term. Voters exhibit a small to moderate bias in favor of the incumbent in presidential elections after one term in office, perhaps because of familiarity with the candidate. After two or more consecutive presidential terms for the in-party, the favorable incumbent effect fades, and the negative party duration effect increasingly dominates public attitudes. Voters increasingly prefer to elect the presidential candidate from the out-party (Fair 2009).
In some instances, the adverse party duration effect can cancel out a positive economic effect on a presidential election outcome. This may have been an issue in Al Gore’s presidential election defeat in 2000. Although the economy was relatively strong at the time, many voters sought a change of the political party in control of the White House. In this instance, the Democrats had occupied the Oval Office for two consecutive terms under Bill Clinton. Despite the strong economy, G.W. Bush from the opposing Republican Party won the presidency in 2000, partly because of the party-duration effect.
The party-duration effect was also probably a factor in Donald Trump’s presidential victory in 2016. In that election, Obama and the Democrats had occupied the White House for two consecutive terms. Despite the recovering economy in aftermath of the Great Recession, voters chose to elect the opposing Republican Party to the Oval Office partly because of the party-duration effect.
Three additional noneconomic determinants affect presidential approval but have little or no impact on the presidential vote. They are the honeymoon effect, the scandal effect, and the opinion inertia effect. The presidential honeymoon occurs during the president’s first year in office. According to the honeymoon effect, presidential popularity is relatively high immediately after an incumbent takes office after an election victory. Over a period of about one year, the initial high approval rating gradually dissipates as the election victory euphoria fades (Smyth and Dua 1989).
Presidential scandals are another noneconomic influence on presidential popularity. Some examples are the Watergate scandal on Nixon approval, the Iran–Contra scandal on Reagan approval, and the Lewinski scandal on Clinton’s job approval. The Watergate and Iran–Contra scandals negatively impacted presidential approval for Nixon and Reagan. The Lewinski scandal, however, had a positive impact on Clinton’s popularity. In this case, many voters viewed the impeachment against Clinton as excessive, and subsequently the Senate acquitted Clinton of the House impeachment. After the acquittal, Clinton’s popularity spiked upward.
The opinion inertia effect is a third noneconomic influence on incumbent popularity. Because of opinion persistence or inertia among voters, presidential approval tends to change gradually in response to changes in the economy. A sustained improvement in the economy has a small initial impact on presidential approval. This small effect increases in magnitude over the subsequent months if the strong economy persists.
Economic Influence on Congressional Elections
The economy also influences Congressional House election results. The Congressional House vote share for the in-party is the percentage of the two-party vote in favor of in-party candidates in House of Representatives elections. Two types of Congressional House elections occur, which are on-term elections and midterm elections. The on-term House vote refers to Congressional elections that take place in the same years as the presidential vote. The 2016 Congressional election was an on-term vote year because a presidential election took place. The midterm vote refers to Congressional House elections that occur between presidential elections. The 2018 Congressional vote was a midterm election because a presidential election did not take place that year.
The economy influences Congressional House election outcomes for both on-term and midterm elections. The impact of the economy, however, is different for on-term elections versus midterm elections. Economic influence on on-term House elections occurs indirectly through the presidential coattail effect. Economic influence on midterm House elections takes place indirectly through the presidential approval effect.
The on-term House vote share for in-party political candidates is directly correlated with the in-party presidential vote share. This is the presidential coattail effect. If the in-party wins a presidential election because of a strong economy or other factors, then the in-party Congressional House vote share likewise tends to increase. The more the votes for the presidential candidate of the incumbent party, the greater the in-party House vote share. If a low vote share occurs for the presidential candidate of the in-party because of a weak economy or other factors, then the in-party Congressional House vote share similarly tends to decline. The House vote share, in other words, increases for the political party that wins the presidential election. If a Democrat wins the White House, then Democrats are likely to gain votes and seats in the House of Representatives. If a Republican wins the presidency, then Republicans likely gain votes and seats in the House.
In midterm elections, the economy indirectly impacts the in-party House vote share through the presidential approval effect. The stronger is the economy, the greater is presidential approval, and the larger the vote share for in-party candidates in midterm House elections. The weaker is the economy, the lower is presidential approval, and the lower is the in-party midterm House vote share. If presidential approval for a Republican president is high because of a strong economy or other factors, then Republicans may gain seats in the House in a midterm election. If presidential approval for a Republican president is low because of a weak economy or other factors, then Democrats may gain seats in the House in a midterm election.
A noneconomic determinant on midterm House elections is the balancing effect. The midterm balancing effect partially offsets or counteracts the on-term presidential coattail. Through the coattail effect, the House vote share in on-term elections increases for the political party that wins the White House. For midterm elections, the in-party tends to lose House votes and seats as citizens reduce their support for the in-party. Citizens tend to vote for the out-party in midterm elections. Perhaps this occurs because some in-party campaign promises from the on-term election tend to go unfulfilled by the time of the subsequent midterm election. The public tends to vote for the out-party in midterm House elections through the balancing effect, whereas in on-term elections citizens tend to vote for House candidates who belong to the political party that wins the Oval Office through the presidential coattail effect.
The economy affects Senate election outcomes similar to the way the economy influences House elections. Campbell and Sumners (1990) found that the economy indirectly affects the in-party Senate vote share in on-term elections through the presidential coattail. If the in-party wins the White House because of a strong economy or other factors, then the in-party Senate vote share likewise increases. For midterm elections, Abramowitz and Segal (1992) found that the economy indirectly affects Senate election outcomes through the presidential popularity effect. If the economy causes presidential approval to rise prior to the midterm vote, then the in-party’s share of Senate seats may go up in midterm elections.
Their analysis also found a balancing effect in midterm Senate elections analogous to the effect on House elections. Other things held equal, the in-party often suffers seat losses in both the Senate and the House during midterm elections. Voters tend to increase support for the out-party in midterm elections. This balancing effect in favor of out-party Congressional candidates in midterm elections partially or wholly negates the coattail effect in favor of in-party Congressional candidates in on-term elections.
Other Measures of Public Sentiment Relating to the Economy
Besides economic influence on election outcomes, macroeconomic events affect other indicators of citizen attitudes. Consumer sentiment, for example, is a measure of public perceptions and expectations on the health of the economy. The Survey of Consumer Sentiment by the University of Michigan is the most well-known measurement. The index of consumer sentiment is based on survey responses from households. Consumer expectations about the economy are optimistic if consumer sentiment is high. Consumer expectations about the economy are pessimistic if the consumer sentiment index is low. The state of the economy affects consumer sentiment, especially unemployment and inflation. Consumer confidence tends to be high or optimistic if unemployment and inflation are low. Consumer confidence among households becomes pessimistic if inflation and unemployment are high.
Some other measures of public sentiment that are affected by the economy are macropartisanship, voter turnout or the voter participation rate, and societal happiness. Macropartisanship is a political indicator of the distribution of aggregate voter partisanship across the population. This partisanship indicator estimates the percentage of citizens who identify with each of the two major political parties. The percentage of voters who identify with the in-party tends to rise if the economy is strong. The percentage of the public who identify with the in-party tends to decline if economic performance is low. If, for example, the president is a Democrat and economic growth is high, then the percentage of voters who identify with the Democratic Party will probably rise (MacKuen, Erikson, and Stimson 1989).
Voter turnout or the voter participation rate is the percentage of the adult populace who decide to vote in presidential (and other) elections. An economic determinant that has been found to influence voter turnout in presidential elections is unemployment. The level of voter turnout in presidential elections tends to be low if unemployment is low. Low unemployment may cause voters to feel satisfied with their economic situation. Consequently, they may be less inclined to vote because of economic contentment. The voter participation rate tends to be high if unemployment is high. A high unemployment rate causes political and economic dissatisfaction among voters. Consequently, the public has a greater urgency to vote to motivate politicians to remedy the economic distress.
The social happiness index is an additional measure of public opinion that is affected by the economy. The social happiness indicator, based on survey responses, provides an estimate of the level of public well-being. The higher the happiness index, the greater the attitude of well-being among the population. A strong economy—such as low unemployment, low inflation, and high income growth—tends to cause the happiness index to rise. A weak economy causes the social happiness index to decrease (Frey and Stutzer 2002).