The current debt crisis in Europe and North America raises the question of how to impose spending discipline on governments and politicians. A country with historically low government spending is Switzerland, which many argue is related to the high use of direct democracy. Direct democracy is also prevalent in other countries such as the United States, where more than two thirds of the population lives in a state or city with a popular initiative. Comparing data on postwar spending in states with more or less direct democracy, the empirical evidence points to a strong negative correlation between a region’s spending level and the existence of direct democracy in both the United States and Switzerland (Feld and Matsusaka 2003, Matsusaka 2004).
What is the mechanism behind this negative correlation? Do citizens, if equipped with direct democratic participation rights, cause government spending to decline? Or, are citizens in areas with strong direct democracy just fiscally more conservative than voters in other areas – and therefore have lower public spending?
Evidence from spending in the Swiss cantons, 1890–2000
To answer this question, we put together a comprehensive new data set with detailed information on direct democratic institutions in all Swiss regions over more than a century. Switzerland has a strong federalism where ‘cantons’ (comparable to the US states) play an important role. In fact, all political responsibilities remain with the canton unless they are granted to the federal government in a national referendum. In 2007, 45% of all government spending was undertaken by the cantons, 31% by the federal and 26% by local governments. Cantons have a lot of autonomy in the provision of public goods, and the authority to tax labour and capital income. As a result, there is a lot of heterogeneity in taxes, public spending, and – most importantly for this study – political institutions, across cantons. The study focuses on the direct democratic institutions most relevant for fiscal policy: the mandatory budget referendum and the voter initiative.
The mandatory budget referendum (currently in place in about 60% of the cantons) gives citizens the power to approve or reject government projects when its (one-time or recurring) expenditures exceed a certain monetary threshold (the exact threshold is set in the canton constitution). The voter initiative in turn allows citizens to propose entirely new laws – for example, limits on spending growth. While all cantons have the voter initiative, there is substantial variation in the number of signatures required to get an initiative on the ballot. Therefore, a lower signature requirement (measured in terms of eligible voters) facilitates the use of direct democracy.
The long horizon of our analysis – more than 100 years – has the advantage that almost all cantons changed their direct democratic institutions (often multiple times). Our study then estimates what happens to public spending if a canton adopts – or abolishes – the mandatory budget referendum, or facilitates voter initiatives by reducing its signature requirement. It does this by comparing cantons to other similar cantons without such reforms. We control for permanent differences between cantons that affect spending with the inclusion of canton fixed effects. We also control for other observable canton characteristics such as population composition, or the economic structure. Most importantly, we explicitly construct measures of voter preferences in a canton from several hundred voting decisions in federal referendums (see Funk and Gathmann 2011a for details).
To address the concern of reverse causality, ie the possibility that public spending results in institutional reforms rather than the reverse, we use an instrumental variable approach. Since all direct democratic participation rights are set down in the constitution, institutional reform necessarily requires a change of the canton’s constitution. Our first instrument is then the barriers to amending the canton’s constitution. As a second instrument, we use changes in direct democracy in the neighbouring cantons. Both instruments affect the provision of direct democracy in a canton, but are plausibly unrelated to a cantons’ public spending. Empirical identification of the causal effect of direct democracy on public spending is then achieved by instrumental variables combined with canton fixed effects.
We find that:
- Having a mandatory budget referendum in place reduces public spending by 12%.
- Voter initiatives that allow citizens to propose new laws also lower public spending. For every 1% reduction in the signature requirement, public spending declines by 0.6%.
- The constraints imposed by direct democracy at the state level do not result in more local spending. This result suggests that state politicians cannot avoid the disciplining effect of direct democracy by simply shifting responsibilities to lower levels of government.
In sum, this study shows that direct democracy indeed causes a decline in public spending, though the relationship is weaker than is suggested by a naive comparison of spending across states with and without direct democracy. Apparently, voters are fiscally more conservative than elected politicians, and the tools of direct democracy help them to get their preferences better represented.
Feld, LP and JG Matsusaka (2003), “Budget Referendums and Government Spending: Evidence from Swiss Cantons”, Journal of Public Economics, 87:2703–2724.
Funk, P and C Gathmann (2011a), “Preferences Matter! Voter Preferences, Direct Democracy and Government Spending”, Unpublished Manuscript.
Funk, P and C Gathmann (2011b), “Does Direct Democracy Reduce the Size of Government? New Evidence from Historical Data, 1890-2000”, Economic Journal, 121(557):1252-1280.
Matsusaka, JG (2004), For the Many or the Few: The Initiative, Public Policy, and American Democracy, Chicago: University of Chicago Press.