Are voters more fiscally conservative or fiscally liberal? Do they reward politicians who engage in deficit spending or those who show more fiscal restraint? To what extent does this depend on the country-specific context? The public choice literature finds that voters in developed democracies with strong institutional checks and balances, such as the US, UK, Canada, or Sweden, reward more accountable governments which spend the public’s money responsibly (e.g. Peltzman, 1992; Lowry, Alt and Ferree, 1998; Brender and Drazen, 2008). On the other hand a number of studies done in countries such as Argentina, Brazil, Spain or Russia suggest an opposite conclusion – voters reward incumbents who spend more and more freely (e.g. Jones, Meloni and Tommasi, 2012; Sakurai and Menezes-Filho, 2008; Balaguer-Coll et al, 2015).
In my most recently published paper “Post-war voters as fiscal liberals: local elections, spending, and war trauma in contemporary Croatia”, co-authored with Professor Josip Glaurdić from the University of Luxemburg, we attempt to provide an answer to these questions in the context of a post-conflict society in which we examine how the impact of war affects citizens’ preferences towards redistribution. The title is a spin-off to Peltzman’s seminal 1992 paper on US “Voters as Fiscal Conservatives”. We argue that in Croatia, because of the recent exposure to conflict, citizens were more likely to be fiscally liberal, but socially conservative.
We use individual-level and municipality-level data from Croatia and find that both individuals and communities that were more exposed to war violence tend to demand greater economic security and fiscally expansionary policies from their local governments. Our municipality-level results show that greater exposure of a community to the war is correlated with greater support for fiscally expansionist local mayors, whereas the individual-level survey data shows that stronger feelings of war-related trauma resulted in greater demands for income redistribution and government intervention. For the aggregate-level data this relationship is tested across the past four electoral cycles (from 2005 to 2017), while individual-level data comes from an extensive cross-national survey done in 2003-2004.
The fact that war-affected areas favour more government interventionism should not be surprising. Among other things, wars cause substantial infrastructural damage to roads and buildings that only the state can repair. A politician that helps to rebuild the community quicker gets rewarded for their actions. The electoral effect can be substantial: incumbent mayors that spend more, particularly on highly visible infrastructure projects, and run budget deficits get rewarded in the polls.
We use Croatia to test our hypotheses given that it is the most recent EU member state with experiences of war on its soil; a war that lasted from 1991 to 1995. The fighting’s impact was devastating: 13,500 dead, 37,000 wounded, and approximately 800,000 people forced to migrate during the war. This makes at least one quarter of the population directly or indirectly affected by the war, which enables considerable variation in examining how the war shaped their electoral attitudes and their redistribution preferences. Furthermore, Croatia is a particularly interesting case study since it experienced war simultaneously with its economic and political transition from socialism. This could explain why, for example, voters most affected by the war tend to be more conservative and support right-wing parties, but at the same time favour more socialist economic policies. This fits well with the Tavits and Letki (2009) theory on the reverse roles of the Left and Right in Eastern Europe.
We also do comparisons of pre-war government spending and investments on a municipal level in order to control for any possible long-term factors that might have affected the internal demand for greater fiscal expansion in the absence of war. This way we can at least partially be sure that the war was indeed the source of greater demand for fiscal expansion.
One of the most interesting implications of our study is that the legacy of war delivers a long-lasting change of people’s political and economic preferences. Even twenty years after the war, the estimated effects are still strong. The deeper, more rooted impact of this long-lasting effect is that voters permanently switch their political preferences and thus tend to encourage fiscal profligacy.
Post-war societies usually suffer from weak institutions and are characterized by high levels of corruption and clientelism. For example, we found that voters in Croatia tend to favour “white elephant” projects and “bridges to nowhere.” One particular category of the budget that tends to be highly sensitive to electoral wins and losses are expenditures on non-financial assets. Non-financial assets are defined as expenditures on any construction-related work, from road repairs and maintenance jobs, to local parks and buildings, to social infrastructure spending such as schools, kindergartens or hospitals. The biggest issue with such capital expenditures is that they are vulnerable to bribes and fraudulent procurement contracts, a recurring and significant problem in Croatia. Contracts are awarded at one price, the job is done at a higher price and the difference is divided between the politicians and their clientelistic networks. However, voters do not seem to mind; whenever there is any construction work being done in the city or municipality it almost always boosts re-election chances of the incumbent.
Within such a context, having the electoral system reward incumbents for fiscal profligacy prevents such societies from fully recovering from war-time legacies. Rather than encouraging sustainable spending, it pushes them into a vicious cycle of political populism that creates public sector dependency and increases local deficits and debts. The war, therefore, did not only cause direct damage to the infrastructure and local living conditions, it also generated a long-lasting development effect on the entire society. According to our findings, developing institutional mechanisms that will prevent such outcomes is one of the most challenging aspects of post-war reconstruction for any society.