Welcome to another in my occasional series on useful, interesting or controversial findings from academic studies of our elections.
Today it’s turnout and the question, “Is voting habit forming?” In other words, if you persuade someone to go out and vote in one election, do you get a bonus benefit in that they are also then more likely to vote in future elections?
That’s the question David Cutts, Edward Fieldhouse and Peter John studied for their article in Journal of Elections, Public Opinion and Parties, Volume 19 Number 3. The study was a follow up to a practical experiment carried out in the Wythenshawe & Sale East constituency in 2005.
During the 2005 general election, some voters were targeted with non-partisan phone calls urging them to vote, some received personal calls on the doorstep and a control group received neither. Those who received phone calls or doorstep visits had a turnout rate seven percentage points higher than the control group. That is a reassuring finding for those of us who have spent time on polling day reminding people to vote. The sample sizes were not large enough to say with confidence whether personal visits or phone calls were more effective than the other.
This study looked at the knock-on impact in the 2006 local elections. No further get out the vote (GOTV) activity was tried, so it was a test of whether the effect of the activities in 2005 still lasted through to 2006.
The conclusion of the detailed statistical analysis is clear:
If you persuade 100 people out to vote today, you may be gaining an additional 50 votes in next year’s election with no additional effort.
The one slight flaw in the work is that it sidesteps the issue of there being two different possible effects. Voting in 2005 may have caused a habit which made people more likely to vote in 2006. Or, the memory of the messages received in 2005 may have still lingered in people’s minds in 2006.
This distinction matters because there are other circumstances which raise turnout (e.g. a one-off particularly controversial election) which would result in raised turnout in the future if the former is the case, but not if the latter is. However, the paper makes the assumption for its statistical analysis that the latter effect does not exist.
Overall though the findings are (another) good reason why elections plans at different levels should fit together into one complete plan. Decide to give the elections for one body a pass from a proper polling day operation because you’re not so interested in them? You then miss out on the opportunity to help get your supporters in the habit of voting in the election where you do care more about their votes. Voting, and voting Lib Dem, is something that is best built up repeatedly over different elections.