For all of the money spent on this year’s midterm elections — $3.67 billion, according to the Center for Responsive Politics — less than half of eligible voters will actually cast ballots in the nation’s 435 House districts, if history is any guide.

House Election TurnoutPolitical scientists (and practical politicians) long have recognized that voter turnout surges in presidential election years and falls off in midterm elections. In the 2008 presidential election, for example, nearly 59% of estimated eligible voters voted in that year’s House elections. Two years later for the midterms, only about 41% of eligible voters went to the polls. (We estimated eligible voters in each district from 2006 through 2012 using data from the Census Bureau’s American Community Survey, and obtained vote totals for every House race from the House Clerk’s office. Our analysis excluded a handful of races in which unopposed candidates weren’t listed on the ballot.)

You might think there’d be some relationship between how competitive a given election is and turnout. A race where victory could go either way might spur more interest and rev up get-out-the-vote efforts from both sides; a race where one candidate is a prohibitive favorite could lead many people to conclude there’s no point in heading out to vote. But our analysis shows little, if any, correlation between a House election’s competitiveness (measured by the winner’s victory margin) and turnout.

In the Tea Party year of 2010, for example, overall turnout in House races was 40.7% of estimated eligible voters. That year, the nation’s highest turnout was in Wisconsin’s 5th District, where 62.4% of estimated eligible voters cast ballots and Rep. F. James Sensenbrenner cruised to a 42-percentage-point victory. The nation’s tightest race in 2010 was in Illinois’ 8th District, where Republican Joe Walsh edged Democrat Melissa Bean by just 290 votes; 41.7% of estimated eligible voters cast ballots in that contest. (The lowest turnout rate in 2010, 19.5%, came in Texas’ 16th District, which Democrat Silvestre Reyes won by a comfortable 21.5 percentage points.)

People who identify with the Republican Party generally are considered more likely to vote, especially in non-presidential years. In a Pew Research Center poll last month, for instance, 77% of registered voters who supported the Republican candidate in their district said they definitely would vote in the election, versus 70% of registered voters who supported the Democratic candidate. And in each of the election years from 2006 through 2012, House districts won by Republicans had slightly higher average turnout than districts won by Democrats.

Drew DeSilver  is a senior writer at Pew Research Center.