Voters Respond to Economic Woes

September 10, 2008

After the Gulf War, President George H. W. Bush was riding high. In March, 1991, some 89 percent of Americans approved of his job performance. By August of the following year, his approval rating had toppled to 29 percent.

What happened? Recession.

Political strategist James Carville capitalized on the downturn, and he kept Bill Clinton's election workers on track by hanging a sign in campaign headquarters with three key messages on it. It read:

  1. Change vs. more of the same
  2. The economy, Stupid
  3. Don't forget health care

Although originally intended for campaign insiders, "It's the economy, Stupid" became an important slogan during Clinton's first presidential run. That's not surprising. Several studies have shown that, indeed, voters are sensitive to prosperity and, especially, to economic setbacks when they're picking a president.

But, are some voters more sensitive than others? Does the local economy matter more to voters than the national one? How far back do voters' memories go when looking at economic data? One year? Two? All the way back to the beginning of an incumbent party's administration?

These are some of the questions behind research conducted by Jonathan Ketcham, an assistant professor and economist at the W. P. Carey School of Business, who teamed up with Daniel Eisenberg of the University of Michigan to conduct his investigation. Although Ketcham notes that "many factors go into determining people's votes," he also says his findings confirm conventional political wisdom: The economy does matter.

"Part of what triggered this analysis is the fact that there were county-level voting data available for the first time," Ketcham explains. Up until that point, he says, data sets only presented information on a state or national level. County-level data offered opportunities to link voter behavior back to local economic conditions and demographics.

Study results suggest that what matters most to voters is the national economy, not state or local ones. The researchers also found voters hold incumbent parties responsible for what happened in the year directly preceding a presidential election.

In local news …

To perform their investigation, Ketcham and Eisenberg examined nationwide, county-by-county data on voting between 1972 and 2000 and national data from 1932 to 2000. The two also tapped data from the U.S. Bureau of Labor Statistics, the Dow Jones Industrial Average, the Consumer Price Index, and the Bureau of Economic Analysis to see if and how the economy filters into the voting booth.

That local snapshot provided Ketcham with a few surprising pictures. One was that counties with larger black populations appeared more sensitive to economic conditions. "The conventional wisdom is that black voters are strongly affiliated with the Democratic party," Ketcham says.

Had that been true, those voters would have voted along party lines, not in response to economic troubles or triumph, "But, that isn't what we found at all," Ketcham notes. Other research has shown that economic downturns hit people with low socioeconomic status harder, and these results suggest that it makes a difference in how black voters cast their ballots.

Elderly voters were less sensitive to unemployment statistics than others, the researchers found. "That makes sense," Ketcham says. "They're retired and not in the workforce."

Different economic factors have differing impact, too. For example, if national income grows by 1 percent during the two years prior to an election, county vote share for the incumbent party grows 1.6 percent. If that growth is on a state level, it only results in a 1.1 percent bump in vote share for the incumbent party. At a county level, the impact is even more minimal: A .08-percent increase in vote share rewards that 1 percent of income growth.

Unemployment figures have a more dramatic effect, but only on a national level. With each 1 percent year-over-year increase in unemployment, voter support for the incumbent party declines by 9.2 percent.

The national scene

According to the research results, county-level economic conditions have little impact on voter behavior, and state-level ones have only a slightly stronger punch. In a paper covering their research, Ketcham and Eisenberg state: "Voters believe the president has little effect on their local economy." Ketcham finds this a reasonable stance on the part of voters. It makes some sense, he says, "because there are other layers of government in place that could share credit or blame" for how the local economy stands at election time.

Apparently Truman was right when he said, "The buck stops here." The researchers found that economic voting was not more pronounced for incumbent candidates than for incumbent parties running a new candidate. And, economic voting wasn't affected by whether the incumbent president's party also controlled Congress.

Timing of economic news is important, too. "We found that the most recent economic factors are the most important in terms of influencing people's voting patterns."

In fact, the research suggests that voters discount the first three years of an administration, and vote in accordance with economic circumstances in the last year before an election. "It's quite possible that voters ignore the first three years at their own peril," Ketcham says. "It could create bad incentives for politicians." For instance, it could prompt incumbents to resort to short-term problem solving, such as beefing up law enforcement in an attempt to lower crime rates or providing a one-time tax rebate to boost sickly consumer spending.

On the other hand, Ketcham does see some logic in ignoring early years in an administration. "If you think that the economy doesn't turn sharp corners but rather goes around a gradual curve, this may be a rational way to look at what presidents are responsible for," he says.

Then and now

Given Ketcham's findings, how might voters respond to this year's circumstances? As of August 31, 46 percent of survey respondents fingered the economy as the most important issue in this year's presidential election, according to a Diageo/Hotline Poll conducted by Financial Dynamics. Next in importance: America's energy policies. Some 19 percent of voters found energy of concern.

A week earlier, 43 percent of those who participated in a USA Today/Gallup poll called the economy issue number one. Another 15 percent called the situation in Iraq the most important issue voters face.

Such figures suggest that the economy may be even more important this year than it has been in the past, Ketcham says.

As this story was being written, U.S. stocks were tumbling downward in response to a spike in the national unemployment rate, which reached a five-year high of 6.1 percent in August.

How might these factors play out in battleground states? On August 25, The Wall Street Journal ran an article headlined, "Obama Targets Swing States like Colorado to Tilt Election." In the story, Jefferson County, an area south and west of Denver, was offered as one of the key counties candidates will woo.

There, the county's 350,000 or so voters are almost evenly split among Republican, Democratic and Unaffiliated designations, although an extra percentage point or two has, for years, gone to the group belonging to the GOP. For decades, Republican candidates had an advantage.

But, as The Wall Street Journal noted, that's changing.

Dick Barkey, head of the Jefferson County Democratic Party, recent crunched the numbers on new-voter registrations in his area. "Jeffco" has seen a net gain of 8,563 voters since March, he says. Of those, 78.5 percent registered as Democrats, while 19.2 percent registered as "unaffiliated" to a party. The Republican Party gained one lone Jeffco voter in that six-month time period, Barkey reports. "If I were a Republican, I'd be feeling a little sick to my stomach right about now," he says.

Paired with these kinds of numbers, Ketcham's findings don't bode well for the incumbent party. "The economic performance announced over the next eight weeks could very well make the difference," he says. Yet, he still sees his investigations as delivering bottom-line good news.

"People respond to the economy and hold the president responsible for it," he says. "This means the president's interests are aligned with the peoples'. If the economy tanked and it made no difference in voter behavior, elected officials would have no incentive to represent voter interests."

Bottom Line:
  • The economy matters in presidential elections.
  • Voters hold incumbents and their parties responsible for economic conditions in the year prior to an election.
  • National economic conditions are more likely to sway voters than state or local ones.
  • Unemployment figures have more impact than those concerning income growth.