Another daylight saving time (DST) has come and gone without triggering the collapse of society, but that doesn’t mean it hasn’t had an impact. Research suggests that DST can influence energy use (pdf), the prevalence of workplace accidents (pdf), and the tendency to shirk work responsibilities by looking at random stuff on the internet (a practice known as “cyberloafing.”)
One unexplored aspect of DST is how it might influence voting behavior. Before 2007, the clocks were turned back during the last weekend in October, which means that when November started on a Monday, elections took place only 2 days after the time change (rather than 9 days after.) Since 2007, clocks have been turned back during the first weekend in November, which means that elections now take place 2 days after the time change except when November starts on a Monday (in those cases the election takes place 5 days before DST.)
The question then, is whether voting behavior differs when elections are held two days after the clock change. One theory is that DST decreases voter turnout. The reasoning is that people are more likely to engage in activities when it’s light outside, and turning back the clock decreases evening sunlight. On the other hand, DST also gives people an extra hour to sleep or fulfill other responsibilities. As a result, it may create more free time or raise energy levels on the following Tuesday, and that could increase voter turnout.
So, does DST help or hurt voter turnout?
A new study by Iowa State’s Robert Urbatsch provides an answer. Urbatsch examined three different sets of data with the potential to illuminate differences between post-DST and non-post-DST elections. First, because only certain counties in Indiana observe DST, Urbatsch compared turnout among counties that did and did not have a 25-hour Sunday immediately before an election. Second, Urbatch examined a variety of states and compared turnout in years that election day occurred immediately after DST and years in which it did not. Finally, Urbatch examined voter-level data from the American National Election Study (ANES) survey to see whether individual voting behavior was different when the clocks were turned back two days prior to the election.
All three data sources suggest that turning back the clock for DST increases voter turnout. The Indiana data suggests DST results in a 2.5 percentage point increase in voter turnout, an effect that’s approximately equivalent to increasing the over-65 population by 5 percentage points. Data from individual states suggests that the increase in turnout could be as high as 4.5 percentage points, and data from the ANES suggests DST increases the odds an individual will vote by about 2 percentage points. Furthermore, according to the ANES surveys this effect is almost entirely restricted to people who are not habitual voters.
Given that more turnout generally benefits Democrats, and that non-habitual voters are particularly likely to lean left, the study is reason for Democrats to support the continuation of DST. In fact, 2010 was a rare occasion in which the election did not follow DST, and at the margin that may have contributed to the sweeping Republican victory. (And thus begins the talking point of DST being a liberal conspiracy.)
More broadly, if one extra hour two days before an election can increase turnout, it implies that proposals to extend voting hours or make election day a national holiday could have a large impact on voter turnout. If we want more people to vote, it’s probably a good idea to give them the time to do it.
————————————————————————————————————————————-
Urbatsch, R. (2014). Time Regulations as Electoral Policy American Politics Research DOI: 10.1177/1532673X14523034