Local autonomy and national–local turnout gap: Higher stakes, higher turnout?

ABSTRACT

This paper offers a cross-national comparison of the aggregate turnout rates in local elections across 21 European countries between 1990 and 2014. The study confirms that the aggregate turnout in local elections is almost always lower than the turnout in the preceding parliamentary election; yet, the national–local turnout gap varies considerably across European countries. The level of local autonomy, i.e. the amount of power and discretion available at the local level turned out to be only a weak explanation of this participation gap. The Local Autonomy Index (LAI) had a positive effect on the local voter turnout, which is in line with the ‘less at stake’ argument of the second-order elections theory, yet the magnitude of this effect is rather small. The synchronization of electoral cycles plays a much more significant role in explaining the aggregate local voter turnout.

KEYWORDS:

• Local elections
• voter turnout
• electoral participation
• local autonomy
• second-order elections

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 The number of eligible voters in local and national (parliamentary) elections may slightly differ due to migration, natural population changes, as well as eligibility rules. Historically, in many European countries universal democratic rights have been introduced from above, i.e. first in national elections (to deal with local affairs has been the privilege of the local elites). Usually local elections are more inclusive (e.g. the citizens of EU countries are entitled to vote in local elections in any of the member states; similar regulations apply among Nordic countries). Yet, the electorate in parliamentary elections often includes citizens living abroad who are not allowed to vote in local elections. Unfortunately, the sources of historical data on electoral participation do not usually report the size of the electorate precisely, but only turnout rates.

2 Data were collected initially for 23 countries, yet the models exclude two countries. Belgium was excluded from further analyses due to the use of mandatory voting both in national and local elections. Sweden was excluded because of the simultaneity of local and national elections. Not surprisingly, both countries represent high values of turnout and very small national–local turnout gaps.

3 I based the calculations on the LAI dataset containing the raw expert scores in the 1990–2014 time-series (disaggregated at the regional level for the federal and asymmetrically decentralized states). The synthetic index of local autonomy takes 7 dimensions of local autonomy into account, ascribing weights from 1 to 3 to each of them (see previous section).

4 Including mixed proportional (MMP) and STV.

5 The baseline specification includes SE clustering for each of the German Lands. However, models brought similar substantial results if all observations from Germany were treated as one cluster. The main result did not change substantially even when the German data were excluded from the dataset.

6 The closer inspection of data suggests that simultaneous presidential elections have the most visible impact on local voter turnout; yet, this result should be interpreted with caution as it is based on two observations only.

Additional information

Funding

This work was supported by Ministry of Science and Higher Education [Mobility Plus grant number 1665/MOB/V/2017].