Power-sharing and the quality of democracy

There is evidence that power-sharing supports democratic transitions. Yet, how do the resulting regimes fare in terms of democratic quality? In our new article in European Political Science Review, we systematically investigate this question for 70 democracies since 1990.

From Bosnia and Herzegovina, over North Macedonia and Kosovo, to South Africa, power-sharing has been a widely-propagated measure to introduce and safeguard democracy. Moreover, several long-established democracies, such as Switzerland and Belgium, rely on power-sharing to reconcile the majoritarian tendencies of democracy with minority accommodation. Evidence is mounting that power-sharing can indeed help introduce and uphold democratic regimes. However, we know little about the resulting quality of democracy. Given an increasingly vocal critique that power-sharing may come with important illiberal short-comings that limit democracy, this is a critical oversight. Addressing this gap, we provide a comprehensive discussion of how power-sharing affects the quality of democracy, going beyond specific individual aspects of democracy. Moreover, we offer a first systematic, large-N analysis of this relationship.

Our findings offer a more optimistic picture than previous case-based research. We find that power-sharing institutions are associated with clear advances for several aspects of democratic quality, including the representation of diverse social groups, liberal rights, and political checks and balances. In contrast, in our sample, the shortcomings that country experts have identified in power-sharing democracies are not systematically different from those in majoritarian democracies. Partial exceptions are a somewhat more passive public and limited external instances of control.

Our argument distinguishes between four aspects of democratic quality. We expect power-sharing to affect these aspects in different ways, entailing critical trade-offs in the process:

1. Representation: We argue that power-sharing provides for a more inclusive representation of the main ethnic segments. Conversely, we expect it to deemphasize the representation of non-dominant groups that do not identify with the main ethno-national divide, such as women, micro-minorities, and lower economic classes. A famous example for these tendencies is the three-headed presidency of Bosnia and Herzegovina. It consists of a representative for each of the three constitutionally recognized ‘constituent peoples,’ the ethnic Bosniaks, Serbs, and Croats, thereby limiting other citizens’ representation. This arrangement also deemphasizes inequalities that cross-cut ethno-national divisions, such as class or gender.
2. Liberal rights: Similarly, we expect power-sharing to boost the political, cultural, and economic rights of the ‘main’ ethnic segments, especially those that make up the power-sharing system. Conversely, we expect it to partly infringe on individual rights, especially those that transcend group boundaries. One example for this is South Tyrol’s regional power-sharing system, which provides comprehensive rights to ethnic Germans, Italians, and Ladins. However, it requires individuals to identify with one of these segments to access political functions and public services, thereby limiting individually-based liberal rights.
3. Vertical control: Power-sharing relies on a model of group representation: its institutions encourage the representation of each segment by its respective elites in a cartel-like government. To govern effectively, these elites require a large degree of autonomy. By concentrating power in this way, we expect power-sharing to have substantial drawbacks for vertical control: most notably, it should reduce democratic competition, public participation, and the transparency of the political process. One example is Macedonia, where electoral participation is high, but characterized by a low confidence in the political system and by an ethnically-segregated civil society and newspapers. Another example is Belgium, where power-sharing is implemented through linguistic cartel parties with comparably demobilized followerships. Belgium’s ‘partitocracy’ has further also shown a rather high reliance on patronage and clientelism. For instance, compared to other Western European countries, it disproportionately deemphasizes meritocratic criteria in favor of political ones to fill administrative and judicial positions.
4. Horizontal checks and state capability: Power-sharing almost by definition enshrines strong horizontal checks and balances that are under the control of the governing political parties. Conversely, we expect it to deemphasize external instances of control that remain outside their influence. Although partisan veto points are conducive to compromises in decision-making, they risk producing reform blockages and thereby reducing state capability. One example is Switzerland, where extensive checks and balances contrast with the limited prerogatives of the federal court. In 1848, these were circumscribed with the express purpose of upholding segmental autonomy. In addition, Switzerland’s judiciary is comparably dependent on executive or legislative power in its appointment procedures, putting it under partial control of the political elites

In on our quantitative analysis, we find a robust association of power-sharing with several aspects of the quality of democracy. We find that it provides for the better representation of diverse social groups compared to majoritarian democracies. Notably, this association goes beyond ethnic segments, which are particularly privileged by power-sharing. We find no evidence that power-sharing restricts the participation and representation of other groups. To the contrary, women – one of the ‘other’ groups not considered by power-sharing – tend to be better descriptively represented. We also find a strong association of power-sharing with formal liberal rights. In addition, we find that power-sharing comes with additional political checks and balances that enable segmental elites to control one another.

In our sample, we attain only two drawbacks: First, the public tends to be more passive in power-sharing democracies, at least as regards informal participation. However, this is not mirrored by lower public engagement in the civil society, thus contradicting arguments that powersharing would demobilize the public across the board. Second, we find that power-sharing is associated with weaker external instances of control that remain outside the influence of segmental elites.

All our findings are based on a quantitative assessment, using the new Constitutional Power-Sharing Dataset, which I collected. Our dependent variables are taken from the Democracy Barometer. This enables us to disentangle the contradictory consequences of power-sharing for various aspects of democratic quality.

In sum, our empirical assessment highlights the democratic qualities of power-sharing, rather than its illiberal shortcomings. However, these qualities appear mostly limited to the core domains where power-sharing has a direct effect, and do not go much beyond.

For detailed findings and methodology, please check out the article here.

Power-sharing and the quality of democracy
Bochsler, Daniel & Juon, Andreas (2021). European Political Science Review 13 (4): 411-430.
Description | PDF | Supplement | DOI