By Jens Olav Dahlgaard, Assistant Professor, DBP
There is a strong concordance in political behavior between parents and their children. This empirical relationship is well established through numerous studies of both political attitudes, such as vote choice, and political participation, such as voter participation.
Previous research from Denmark has looked into how turnout is much higher for young adults Danes whose parents vote than for young adults whose parents do not vote. For the children living at home youth turnout is around five times higher in the household where both parents vote than in household where no parents vote. For the children no longer living at home turnout is “only” about twice as high in the group whose parents vote.
We can observe similar strong correlations in other household relations such as concordance between partners, and in other social relations in general. What I have spent a good deal of my research figuring out is what causes these correlations.
Why you are so similar to your friends and family
Why do people behave more in concordance with their close social relations than we would expect by chance alone? Broadly, we can point to three reasons:
1) People in close social relations are exposed to, if not the same then very similar, environments that could have a joint influence on their behavior.
2) People do not randomly choose their social relations. Instead they tend to choose to be around like-minded people and such people happen to tend to have similar attitudes and behaviors.
3) People act in concordance due to genuine social influence where one changes behavior as a response to some social relation
Young voters affect their parents’ participation
In a paper just published in the American Political Science Review, I focus on the third factor. Specifically, I explore the correlation between parents’ and their children’s behavior. I look at it from a perspective opposite to what most people usually think of when they consider this correlation.
Instead of trying to identify how much children follow their parents’ cues, I ask how much parents follow the cues of their children. I show that children affect whether parents vote or not. The idea of so-called trickle-up political socialization is not new (see here, here, and here).
The innovation in my paper is that I am able to causally identify an effect, which runs from children to their parents. Establishing a cause and effect relation in political socialization studies is really challenging for the reasons listed above. How do we know if any correlation is due to common exposure to surroundings, sorting into relations, or genuine social influence?
What I do is to focus on a shock in the parent-child link that opens a pathway for social influence in some parent-child relations, but not in other parent-child relations that are in all other respects comparable.
When children in Denmark turn 18 they can vote in elections (there are no registration requirements). When they are 17 they cannot. Since elections happen on one specific day, there are young people who turn 18 right around the election who are in any respect comparable, except for whether or not they can vote (ask yourself: does someone who is 18 years and seven days old differ that much from someone who is seven days shy of turning 18?)
As an implication of the discontinuous jump that children take from ineligibility into the electorate their parents experience a discontinuous jump, too. On Election Day some will be parents of a young voter, while others will be parents of young people not old enough to vote. The parents’ right to vote is unaffected by the change, the only change is if their child can vote, and perhaps encourage them to do the same.
The figure shows parental turnout aggregated by the day and week their child turns 18 in the area around Election Day averaged over four elections in Denmark (the 2009 and 2013 municipality elections, the 2014 European Parliament election, and the 2015 national election).
In every election there is a jump right in the cutoff, which means that parents with children old enough to vote turn out at higher rates. Based on the figure, the effect may not seem that impressive, but the analyses in the paper show that on average the effect is 2.8 percentage points, which is quite substantial considering that around 75 % of the parents vote. Another way of thinking about the effect size is that out of nine parents who would not have voted, one changed his or her mind because of his or her child turning 18.
The result almost begs the question of why. Why do parents become more likely to vote? I try to test two plausible mechanisms: Parents could vote because they want to instill a norm of voting in their children. Or parents could vote because their children are motivated to participate and effectively mobilize their parents. There is support for both mechanisms, perhaps slightly stronger for the second mechanism of young voters mobilizing their children.
More parents would vote if the voting age was lowered (probably)
What are the implications of the results? First of all it speaks to an academic debate about both political socialization and social influence. Secondly, there is an additional result, which speaks to an ongoing debate about the appropriate voting age.
The voting age in Denmark, as in most democracies, is 18. Countries around the world have experimented with a voting age of 16 or 17 in mostly local elections, though Scotland allowed 16-year-olds to vote in the independence referendum of 2014 and Brazil has allowed 16-year-olds to vote in presidential elections since 1988.
In my paper, I show that it is only parents who still live with their children who are mobilized to vote. Since children start leaving home around the age of 18, a larger proportion of parents would live with their voting age child in at least one election if the voting age was lowered to 16. In addition to the short-term benefit, there could also be a positive long-term effect on parents’ political participation because voting is habitual.
Obviously, there are many other arguments both for and against lowering the voting age. My paper just finds an auxiliary benefit of lowering it. In a recent blog post on the political science blog The Monkey Cage, published on the Washington Post, I elaborate on the voting age discussion and how my research relates to it.