Elections are critical for democracy, but electoral participation in advanced democracies has been steadily declining over the last 50 years (Figure 1), reaching record low turnout rates in important elections such as the Brexit referendum and the recent Colombian peace agreement referendum. Declining turnout rates can distort the political process if individuals who don’t turn out to vote end up being underserved by their governments, as suggested by Lipjart (1997). In fact, it has been reported that ethnic minorities, immigrants and the poor are overly represented among non-voters in Europe and the US (e.g. Timpone 1998, Gallego 2007, Linz et al. 2007). The uneven mapping between eligible voters and those who actually show up to the polls might have far-reaching distributional consequences. A nascent literature in political science and economics motivated by this issue studies how changes in turnout and the composition of the electorate affect public policy. For example, Miller (2008) finds that the enfranchisement of women in the US in the early 20th century resulted in increases in government health expenditures, a public good disproportionally preferred by women. Similarly, in a more recent study evaluating the impact of the introduction of electronic voting in Brazil, Fujiwara (2015) found that the introduction of simple and intuitive voting stations de facto enfranchised illiterate voters, which led to increases in health expenditures and a reduction in child mortality.
Figure 1 Average turnout in OECD countries, 1950-2016
Note: Author’s elaboration using data from International Idea. Average turnout (as a % of registered voters) for every decade in 1950-2016, in all elections with voluntary voting in OECD countries. We consider parliamentary elections in Austria, Canada, Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Iceland, Ireland, Israel, Italy, Japan, Netherlands, New Zealand, Norway, Poland, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey and the UK, and presidential elections in France, South Korea and the US. For the US, the only of these countries without compulsory or automatic registration, we use turnout as a % of the voting age population.
Governments have different policy tools at their disposal to increase turnout, including increasing the number of polling stations and implementing proxy or mail voting, among others. One popular way to overcome the decline in electoral participation is to make voting mandatory. Currently, 18 countries around the world have some sort of compulsory voting laws (see Figure 2), with a higher number having had CV at some point in the last fifty years (see International Idea). Even President Barack Obama, in March 2015 proposed the possibility of compulsory voting in the US, arguing: “It would be transformative if everybody voted, that would counteract money more than anything. If everyone voted, then it would completely change the political map of the country. The people who tend not to vote are young, they're lower income, they're skewed more heavily towards immigrant groups and minority groups...There's a reason why some folks try to keep them away from the polls” (CNN 2015).
A few studies in countries as diverse as Switzerland, Brazil, and Australia have shown that even with small fines for non-voting or low enforcement of penalties, electoral participation is significantly higher under compulsory voting (Funk 2007, De Leon and Rizzi 2014, Fowler 2013.) However, unlike the policies analysed in the work of Miller (2008) and Fujiwara (2015), there are no a priori reasons to believe that the preferences of voters induced to participate in elections due to compulsory voting are significantly different from those of the average voter who participates when voting is voluntary. Hence, it is unclear if increasing turnout using compulsory voting would translate into changes in public policies. This is the question we address in a new paper (Hoffman et al. 2016).
Figure 2 Countries with compulsory voting around the world
Source: International Idea.
Compulsory voting and government spending in Austria
Like many other countries, Austria exhibits socioeconomic disparities in turnout, with poor people being less likely to vote than the rich. To ensure large participation in elections, since the end of WWII, Austria’s nine states have had compulsory voting laws in place. Interestingly, these laws have changed at different points in time and for different types of elections (Figure 3). Local authorities, who are responsible for issuing fines against non-voters failing to provide a reasonable excuse for abstaining, have rarely enforced them though, and allowed for a wide range of excuses for not voting, such as illnesses, professional commitments, or “other compelling circumstances”. De facto, the enforcement of sanctions for failing to vote were very weak. The variation in the presence of CV laws at different points in time, in different states, and for different types of elections provides an excellent setting to study the impact that these laws had on turnout, electoral outcomes, and more importantly, on public policies.
Figure 3 Compulsory voting in Austria, 1949-2010
Note: Solid bars mark the period in which elections with CV were held in each state.
Using administrative data on parliamentary, presidential, and state elections, we first compare turnout rates between states with and without CV laws, as well as within states that changed compulsory voting laws during the period 1949-2010, and find that compulsory voting caused increased turnout by roughly 10%. Consistent with the hypothesis that compulsory voting might increase turnout by drawing uninterested voters to the polls, we find that compulsory voting led to a rise in the proportion of invalid ballots, but the estimates are very small. In particular, for every ten people driven to vote due to compulsory voting , only 1.5-3 of them issue an invalid ballot. These results survive a host of robustness checks. For example, one might be concerned that the direction of causality is backwards, and that states introduced compulsory voting as a response to declining turnout rates. In our paper, we show that this does not seem to be the case, i.e. compulsory voting in the past or in the future is not significantly related to current turnout. Additionally, we observe that our results hold if we only look at the effect of the federal removal of compulsory voting in 1992, in which state governments did not have any influence.
Using disaggregated data on state spending in 1980-2012, we analyse the impact of compulsory voting laws on state-level spending. Interestingly, after controlling for state specific fixed factors, country-wide year specific factors, as well as trends in state-level spending, we show that changes in compulsory voting laws did not lead to significant changes in the level or the composition of spending at the state level. In particular, our empirical results show that neither the amount of state-level spending, nor the share of this budget devoted to administrative, welfare or infrastructure spending, significantly changed with the introduction or removal of compulsory voting . We show that this zero-effect persists if we disaggregate further the spending categories, that they do not seem to be driven by trends predating changes in compulsory voting laws, and are there even if we consider the one change in compulsory voting laws that were not led by state governments (the 1992 federal removal of compulsory voting).
How could it be that compulsory voting had sizable impacts on turnout without affecting policy outcomes?
To start answering this question, we examine if compulsory voting caused changes to the results of national parliament and state elections. A potential explanation for the null effect found is that the political choices of people who turn out to vote due to the introduction of compulsory voting are, on average, similar to those who were voting when voting was voluntary, thus electoral outcomes would remain unchanged. Alternatively, if the median voter’s preferences change and impact electoral outcomes, government spending might still be unaffected due commitment or agency issues. Consistent with the former explanation, we find that for both parliamentary and state elections, compulsory voting has no impact on the vote shares of right or left-wing parties. Further, there is no response from the political supply – the number of parties running for office remains unchanged, as does the margin of victory and share of votes of the winning party.
In the final part of the paper, we shed light on the underlying mechanisms for our results. Using individual data from the 1986 and 2003 rounds of the Austrian Social Survey, we study how the composition of the electorate changed due to the introduction of compulsory voting by looking at the interaction between compulsory voting laws and voter characteristics. Exploiting the elimination of compulsory voting in parliamentary elections taking place in three states in 1992 (between the two survey rounds), we find evidence of larger impacts of compulsory voting on turnout among women and those with lower income. Impacts also seem larger among those who have low interest in politics, who have no party affiliation, and who are relatively uninformed (as proxied by newspaper reading). While suggestive, these results are consistent with a story where individuals who vote or abstain due to the introduction or repeal of compulsory voting do not have strong policy or partisan preferences (on average), thereby having little or no effect on electoral outcomes. Furthermore, if such voters are unresponsive to policy in deciding which party to support, parties may have little incentive to shape policies to suit those voters' preferences.
Our results provide evidence that even if compulsory voting increases turnout, it need not significantly affect government spending. Of course, our results are specific to Austria, though we think they would be quite relevant for other advanced democracies with high turnout, such as Germany and the Scandinavian countries. It is less obvious how they would extrapolate to other countries with lower turnout rates like the US.
CNN (2015), “Obama: Maybe it’s time for mandatory voting”, 19 March.
De Leon, F L L, and R Rizzi (2014), "A Test Tor the Rational Ignorance Hypothesis: Evidence from a Natural Experiment in Brazil", American Economic Journal: Economic Policy 6 (4), 380-398.
Fowler, A (2013), "Electoral and Policy Consequences of Voter Turnout: Evidence from Compulsory Voting in Australia", Quarterly Journal of Political Science 8 (2), 159-182.
Fujiwara, T (2015), "Voting Technology, Political Responsiveness, and Infant Health: Evidence from Brazil", Econometrica 83 (2), 423-464.
Funk, P (2007), "Is There an Expressive Function of Law? An Empirical Analysis of Voting Laws with Symbolic Fines", American Law and Economics Review 9 (1), 135-159.
Gallego, A (2007), "Unequal Political Participation in Europe", International Journal of Sociology 37 (4), 10-25.
Hodler, R, S Luechinger, and A Stutzer (2015), "The Effects of Voting Costs on the Democratic Process and Public Finances", American Economic Journal: Economic Policy 7 (1), 141-171.
Hoffman, M, G León, and M Lombardi (2016), “Compulsory voting, turnout, and government spending: Evidence from Austria”, Forthcoming, Journal of Public Economics. Barcelona GSE Working Paper 809.
Lijphart, A (1997),"Unequal Participation: Democracy's Unresolved Dilemma", American Political Science Review 91 (1), 1-14.
Linz, J, A Stepan, and Y Yadav (2007), Democracy and Diversity: India and the American Experience, 50-106.
Miller, G (2008), "Women’s Suffrage, Political Responsiveness, and Child Survival in American History", Quarterly Journal of Economics 123 (3), 1287.
Timpone, R J (1998), "Structure, Behavior, and Voter Turnout in the United States", American Political Science Review 92 (1), 145-158.