Angry Men and Civic Women? Gendered Effects of Conflict on Political Participation in Kosovo

Abstract

This article studies the effect of the 1998–99 Kosovo war on current political participation, disaggregating the analysis by the type of conflict experience – namely death or injury to self or a family member or displacement – and by gender. The results show that experience of conflict is associated with more political participation but with important distinctions between genders by the form of participation and the type of conflict experience. Displacement is associated with more voting among women, but not among men, and with more demonstrating by men but weaker or no effects for women; death and injury are associated with higher political party membership for men but not women. While experiences of conflict increase levels of political participation, the form that this takes varies by gender, with effects on private, civic, action among women, and effects on direct, public, and more emotionally heightened engagement among men.

HIGHLIGHTS

• The view that conflict victims are more politically active than non-victims needs nuancing.

• In Kosovo, women’s war displacement is only associated with an increase in voting.

• But men will join a political party (if injury or death in the family) or demonstrate (if displaced).

• This implies that victimization does not contribute to challenging gendered social norms.

• The accepted “post-traumatic growth” hypothesis is insufficient to explain these findings.

KEYWORDS:

• Conflict
• political participation
• voting
• gender
• Kosovo

JEL:

• Codes: D74
• O17

ACKNOWLEDGMENTS

The authors are grateful for feedback from Skerdilajda Zanaj, Elena Nikolova, and Eren Arbatli. We also benefited from participants’ comments in a research seminar in the Economics department in March 2021 at the University of Swansea, UK, the 10th ICSID conference on “Political Economy in a Changing World” (HSE, Moscow; June 2021), the 23rd Annual Conference of the Association for Heterodox Economics (online July 2021), the AFK Workshop on “Empirical Research on peace and Conflict” organized by the University of Potsdam and the University of Frankfurt (online September 2021), and the 17th Annual Workshop of the Households in Conflict Network (online, October 2021), during an internal seminar presentation at UCL SSEES, London, UK (January 2023), at the First Welfare and Policy Conference in Bordeaux, France (May 2023), and at the 7th European Workshop on Political Macroeconomics at the Bank of Finland, Helsinki (June 2023). We are also grateful for feedback from Elena Nikolova and to our discussants Eren Arbatli (ICSID) and Roos Haer (AFK). Finally, we want to sincerely thank our reviewers and the editor of the journal for their helpful and constructive feedback: our article has been much improved following their advice. Any remaining errors are entirely ours.

SUPPLEMENTAL DATA

Supplemental data for this article can be accessed http://dx.doi.org/10.1080/13545701.2024.2323657.

Notes

1 See Bateson (2012) for analysis based on an exhaustive set of data sources and focusing on crime victimization.

2 See Bauer et al. (2016) for a recent review of the field focusing specifically on violent conflict victimization.

3 War has been credited for building strong states in modern Europe (Tilly and Ardant 1975; Tilly 1985).

4 We also indicate whether victimization was self-reported or measured from an external source and note that victimization seems more often associated with positive change in political participation when it is measured as individual-level, self-reported victimization, credibly implying that it is personal experience that matters rather than exposure to contextual conflict.

5 There are few other examples of experimental approaches that explore the impact of conflict on gender. Cassar, Grosjean, and Whitt (2013) finds no evidence of gender differences in the case of Tajikistan; Gilligan, Pasquale, and Samii (2014) in their study of Nepal do not control for gender.

6 See for example the paper by Schlozman, Burns, and Verba (1994) highlighting the role of income in explaining part of the gender differences in engagement in the US.

7 This is particularly the case now that the European Union has officially recognized Kosovo as a potential candidate country for accession.

8 According to the “Electoral Democracy Index” of V-Dem, immediately prior to (and during) the conflict Serbia (and then Kosovo) ranked lower than neighboring countries of Albania and North Macedonia, which hosted large proportions of those displaced from Kosovo (V-Dem 2022).

9 We note that we have also run fully interacted models as robustness checks. The results obtained were consistent with those presented here. Split sample regressions were chosen over interacted models for ease of presentation.

10 Unfortunately, the survey data does not record where the respondent was displaced to or for how long, which means we cannot explore the extent to which any effects of displacement on political participation might be due to exposure to stronger democratic regimes and more progressive gendered norms.

11 The survey includes separate questions relating to having a household member injured versus killed during the conflict; this would, thus, allow us theoretically to investigate these two forms of victimizations separately. Twenty-one percent of the respondents report having experienced injuries, and 11 percent have experienced the killing of a family member, but 78 percent of those reporting a killing have also experienced injuries. We, thus, investigate these two forms of victimization jointly. In regressions conducted with separate indicators for killed and injured, our findings appeared to be carried by the experience of injuries. At the suggestion of a reviewer, we also tested a model with an interaction term between injured/killed and displacement to explore if there was any attenuating or strengthening effect of having experienced both types of victimization. The coefficient of the interaction was not statistically significant in any specification.

12 These are the odd numbered regressions in our tables.

13 Estimations with region-level fixed effects are available upon request.

14 Douarin, Litchfield, and Sabates-Wheeler (2012) use this data to build an index capturing the degree of damage at the municipality level to relate conflict intensity to livelihood choices after the war.

15 We note that this is true over all specifications presented Tables A3 and A4 in the Online Appendix.

16 Regarding strikes, demonstration, and signing a petition, we reproduced the analysis but with an indicator equal to 1 if the respondents had participated or would consider participating in these actions and 0 if they had never done so (see discussion in the data section). Results are available upon request.

Additional information

Notes on contributors

Julie Litchfield

Julie Litchfield is Senior Lecturer in the Department of Economics at the University of Sussex, UK. She is Development Economist with a particular focus on the microeconomic impacts of conflict and of migration in low and middle-income countries.

Elodie Douarin

Elodie Douarin is Associate Professor in Economics at the School of Slavonic and East European Studies, University College London, UK. Her research focuses on institutional and cultural economics. She is particularly interested in better understanding informal institutions, as culture and values, and the effects of shocks on values and behaviors. She has recently co-edited the Palgrave Handbook of Comparative Economics with Oleh Havrylyshyn.

Fatlinda Gashi

Fatlinda Gashi is a former graduate student in the Department of Economics, University of Sussex and is currently Senior Economic Policy Analyst in the Ministry of Finance, Labor and Transfers of the Republic of Kosovo.