Human history has been characterised by both shocking levels of political violence as well as by the quest to curb it. The 20th century has been a particularly salient example of this. On the one side, it has been marked by two World Wars, several dozen episodes of mass killings of civilians, devastating purges carried out by a series of totalitarian regimes, as well as many recurrent ethnic civil wars. On the other hand, there has been an important surge in democratisation and increased efforts in peacekeeping initiatives.
One promising idea, that reaches far back and has been hitting the news today, has been the idea of sharing power to prevent violence. Sharing power between distinct religious or linguistic groups has been in place for many years in countries such as Switzerland, Belgium, and Lebanon, and likewise in Northern Ireland since the 1998 Good Friday Agreement. It is under consideration for various other ethnically, linguistically, or religiously diverse countries, such as Iraq. The public debate is often heated with proponents of power sharing stressing its potentially pacifying virtues while opponents pointing out the risk of a lack of accountability, and of political competition in grand coalition governments.
Much anecdotal evidence and journalistic accounts suggest a potentially important role for power-sharing to curb conflict, and there is a clear tendency for some ethnically or religiously divided countries to adopt some form of power sharing. As shown in the qualitative work of Lijphart (1999), many successful and peaceful ethnically and religiously divided countries chose the so-called consensus model of democracy, characterised by power sharing and the decentralisation of power on all levels.
Still, while historical examples tell us that several ethnically and politically divided countries adopted power sharing, and that this correlates with peace and prosperity, this is a long way from showing systematic statistical evidence that the adoption of power sharing results in a reduction of the risk of conflict. In fact, there is surprisingly little hard statistical evidence linking power sharing to peace. Exceptions include Cederman et al. (2013), for example, who find that groups included in government show less propensity to engage in insurgency. This dearth of evidence is in spite of a recent literature which points towards a strong role of democratic institutions in reducing the incentives to engage in violence (for a theoretical argument, see Besley and Persson 2011; for evidence on rent extraction along ethnic lines, see Burgess et al. 2015). Recent years have been marked by a renewed interest in understanding power-sharing better (for a recent review, see Francois et al. 2015), but there are still only very few quantitative micro studies of the effect of power sharing on violence.
The case of Northern Ireland
To address this gap, we study the impact of local-level power-sharing on the risk of conflict in Northern Ireland (Mueller and Rohner 2017). We link local power-sharing measures from administrative records to detailed data on violence which was gathered in the context of the so-called Troubles. This allows us to use quasi-random variation in election outcomes in a subset of politically balanced districts to study the effect of changes in political power on violence.
Our empirical analysis draws on panel data for Northern Ireland’s 26 local district councils over the 1973-2001 period, and we make use of two approaches. The first corresponds to simple panel regressions, where district fixed effects filter out time-invariant district characteristics and time effects to account for global shocks. The identification strategy relies on the time variation in power-sharing agreements at the district level to see whether violence decreases when power sharing is put in place, and whether it increases when it stops.
The second approach carries out a further refinement step in the identification strategy by instrumenting the power sharing in a given district and year with the absence of a majority for sectarian parties – the logic being that in a hung parliament, where no side in the conflict can govern alone, power sharing is more likely to emerge. To avoid comparing apples with pears, we restrict the sample to districts with voting results close to the 50% threshold, where it is arguably close to random if a political block reaches a majority or not.
In both approaches, we find a statistically significant and quantitatively substantial impact of power sharing to curb the scope for killings. In particular, comparing a situation of no power sharing with power sharing, fatalities per capita would be increased in the absence of power sharing by roughly three times the baseline risk. Our estimates show that several hundreds of lives could have been saved if power sharing had been put in place in all districts and years. This result is robust to the use of alternative data on population, alternative sample compositions, alternative ways of measuring power sharing, alternative construction of the dependent variable, as well as of the inclusion of various demographic or political control variables.
In terms of the underlying mechanisms in place, while we do not find robust heterogeneous effects of power sharing on peace with respect to the religious composition of the ward, we show that our findings both apply to killings carried out by Republican (Catholic) as well as Loyalist (Protestant) paramilitaries. The fact that the impact of power sharing is quite symmetrical, applying to various sub-groups of society could imply that the fact of having to talk with each other in local government could create a common ground to sort out issues in a non-violent way. This is also in line with the recent findings of Saia (2017), which show that random seat allocation in parliament can bring seat neighbours closer together in terms of political views.
A noticeable finding is that the positive virtues of power sharing only persist as long as it is in place. Once a group is disenfranchised from power, the old patterns of violent contestation break through anew, and the level of violence rises to the pre-power-sharing levels. This means that power sharing does not establish local trust or change social norms once and for all, but that political cooperation and inclusion need to be maintained in the long run if the benefits of lower violence are to continue.
An important caveat to keep in mind when interpreting our results is that Northern Ireland is a relatively rich country without many natural resources, and that one should be cautious when extrapolating the current findings to developing countries with other risk factors for conflict, such as poverty and natural resource wealth. Future empirical research on the impact of power sharing on conflict and other outcomes is strongly encouraged. The road to this next layer of evidence will go through the collection of further fine-grained data on power sharing for developing countries.
Besley, T and T Persson (2011), "The logic of political violence", Quarterly Journal of Economics 126: 1411–1445.
Burgess, R, R Jedwab, E Miguel, A Morjaria and G Padró i Miquel (2015), "The value of democracy: Evidence from road building in Kenya", American Economic Review 105: 1817–1851.
Cederman, L-E, K Skrede Gleditsch and H Buhaug (2013), Inequality, grievances and civil war, Cambridge: Cambridge University Press.
Francois, P, I Rainer and F Trebbi (2015), "How is power shared in Africa?", Econometrica 83: 465–503.
Lijphart, A (1999), Patterns of democracy, New Haven: Yale University.
Mueller, H and D Rohner (2017), “Can power-sharing foster peace? Evidence from Northern Ireland”, Paper presented at the 66th Panel Meeting of Economic Policy, October.
Saia, A (2017), “Random interactions in the chamber: Legislators’ behavior and political distance”, working paper, University of Lausanne.