Monday, 2 December 2013

Preliminary evidence on the impact of voluntary voting in Chile: Not good news

By Carlos Fortin

In a blog in early 2012, I reported on the debate in Chile about replacing the existing system of compulsory vote with voluntary vote in national elections. I cited the case of the Netherlands, where a similar change in 1970 led to negative impacts in the shape of a fall in overall participation and a proportionally higher fall in the participation of women, the young, the less educated and the poorer. Not, I concluded, exactly what Chilean democracy needs today.

The change was however duly approved and we have just had the first presidential and parliamentary elections under the new system. While the evidence is not yet in to pass a definitive judgment on all the aspects mentioned, some data on at least two of them seem to lend credence to the fears I was reporting on.

Figure 1 shows the level of abstention in the presidential elections in Chile since the return to democracy. Abstention in the first democratic election in 1989 was at a historical low of 13.2% and afterwards rose steadily but fairly gradually (except for a more pronounced increase in 1999) to reach 41.2% in 2009, the last election with compulsory vote. In the 2013 election it jumped nearly 10 percentage points to 50.7%. While there are no doubt many other factors at work, it is hard to escape the conclusion that the shift to voluntary voting has led to a decrease in electoral participation.

Figure 1. Chile: Abstention in presidential elections since return to democracy
Non-voters as percentage of eligible voters
Source: Gonzalo Contreras and Patricio Navia, Participación Electoral en Chile, 1988-2010, Buenos Aires, Instituto de Investigaciones Gino Germani, 2011

On the differential impact of lower turnout, there is some evidence concerning the impact by socio-economic status; it is less conclusive, but the direction it points to is also clear. Figure 2 plots all the municipalities of the Santiago Metropolitan Region in terms of the incidence of poverty in their populations in 2012 and the voter turnout. The negative correlation is apparent: the more the poor in a municipality, the lower the electoral participation (the correlation coefficient is -.496, significant at the 0.01 level).

Figure 2. Chile: Participation and poverty in the 2013 presidential election
Municipalities in Santiago Metropolitan Region
Source: Gonzalo Contreras and Mauricio Morales, Precisiones sobre el sesgo de clase con voto voluntario, 2013

This second finding is not conclusive firstly because it covers only the country’s capital; impressionistic evidence for the rest of the country is mixed, with at least three other regions showing similar tendencies to Santiago but others apparently showing no differential impact.

More seriously, the finding could be criticised for falling prey to the ecological fallacy: it does not tell us whether it was the poor within each of the municipalities that abstained to a larger extent.

More research is therefore needed to arrive at firm conclusions. In the meantime, though, the evidence presented is a serious warning signal; in fact, there are already voices in the Chilean Parliament calling for a reversal of the change, and they include some who originally supported voluntary vote in the hope that it would lead to higher, and more committed electoral participation.

Carlos Fortin is an IDS Research Associate currently working on the relationship between the emerging international trade regime and human rights.