Compulsory voting and voter turnout
It is the right of American citizens to vote in presidential elections, with exceptions based on age, incarceration, and perhaps a few other conditions that don’t occur to me right now. Millions of Americans regularly exercise that right, and other millions regularly don’t.
What if voting were compulsory in the U.S., as it is in many other countries? That is, what if it were treated as a legal obligation rather than as a right? How would that alter citizens’ “cost-benefit calculus” as they think about whether to vote or not”? Would turnout rise because many people want to be law-abiding and/or don’t want to get caught not being law-abiding?
In a recent article titled “The Calculus of Voting in Compulsory Voting Systems” (Political Behavior 30 2008, 455-467), here, gated Costas Panagopoulos notes that countries with compulsory voting laws vary widely in the penalties they impose for non-compliance and the stringency with which they enforce these penalties.
In some nations, including Greece and Mexico, voting is constitutionally required as a civic duty, but this requirement is largely symbolic and there are no formal sanctions imposed on defectors. Non-voters face a fine sanction in other societies. Fines vary across countries and can range, for example, from 200 pounds in Cyprus to 20 dollars in Australia to 10-20 pesos in Argentina. Laws in some compulsory voting states threaten non-voters with possible imprisonment. In others, defectors may face civil rights infringements or disenfranchisement. In Belgium, for example, citizens lose their right to vote after they abstain in four elections within a 15-year period. In Peru, proof of electoral participation is required to obtain certain public services; and in Argentina, offenders cannot work in public service for 3 years following an abstention.
Countries also vary in the degree to which they enforce these penalties. In some countries, non-compliance routinely goes unpunished, but in other countries voting requirements are strictly enforced.
So what are the consequences? To what extent do compulsory voting laws boost turnout?
In the countries in Panagopoulos’s analysis, voter turnout averaged 42.9% during the 1990s. How does that compare to countries with compulsory voting laws? The answer: It depends. Here’s how it all falls out, in terms of the predicted turnout levels in the nine types of nation.
In words: In countries where there is no or low enforcement of the compulsory voting law, turnout is virtually the same as in countries with no such law; in those countries, it doesn’t matter how big the penalty is because it’s so unlikely to be enforced anyway, and people presumably know that.
In countries where enforcement is weak but existent, turnout varies as a function of the severity of the penalty – all the way from 40.3% for countries in the “no/low penalty” category to 55.7% for those in the “high penalty” category.
And in countries where enforcement is strict, turnout varies even more dramatically as a function of the severity of the penalty – ranging from 37.4% in the “no/low” penalty category to 67.8% in the “high” category.
In sum, compulsory voting laws don’t necessarily increase voter turnout, but compulsory voting laws with teeth do.
Having said that, I’ll also register my surprise that even in countries where the penalty for voting is high and so is the strictness of enforcement, a third of the eligibles, on average, don’t vote. Apparently, in order to get more people to the polls, governments would have to take even more drastic steps to “force them to be free.”
Comments
Not voting is both a form of political expression and political freedom. It provides incentives for politicians to make me care. Therefore non-voting should not be illegal.
Posted by: jdk2714 | February 8, 2009 11:44 PM
Those numbers don’t match what I’ve read in other places. They even look low compared to countries with no compulsory voting, like the 54% in the US. (Haven’t read the paper though, so maybe there’s an easy answer there).
Posted by: itsadok | February 9, 2009 01:21 AM
This graph doesn’t even have any points, just trend lines! Seems pretty suspicious to me.
Posted by: Aaron Swartz![[TypeKey Profile Page]](http://www.themonkeycage.org/nav-commenters.gif)
|
February 9, 2009 08:16 AM
And are there really only 3 values on the x-dimension? And does the graph really mean that “strict enforcement” of “low penalty” means turnout is lower than where the law provides for no penalty at all?
OK, I shall desist now in further criticism of a paper I have not read.
Posted by: MSS | February 10, 2009 04:00 PM
Aaron: The graph has points, obviously; you can’t draw lines without ‘em.
MSS: Yes, the researcher categorized nations on the two dimensions (penalties and enforcement), using three categories apiece; so there are nine types altogether. On your question about the difference between two of the points, don’t make anything at all out of that difference, which is way too small to matter or to be taken at face value. If it will make you happier to be told that they’re essentially the same, I’ll say exactly that: they’re essentially the same.
itsadok: These numbers do look low, but I was hoping that someone who knows more about comparative electoral systems than I would weigh in to explain that.
Posted by: Lee Sigelman | February 10, 2009 04:23 PM