Who Do You Trust? Ethnicity and Trust in Bosnia and Herzegovina

Abstract

This article examines the question of trust in Bosnia and Herzegovina with a special focus on the role of ethnicity. We find generalised trust to be low and declining in Bosnia and Herzegovina. Moreover, generalised trust is negatively affected by the degree of ethnic heterogeneity in the region. However, a further examination of trust reveals a more complex relationship between ethnicity and trust: people tend to show low levels of trust in all other people irrespective of their ethnic belongings. We argue that ethnic distribution might capture some other regional specific characteristics that also affect the level of trust.

Notes

¹See Kolankiewicz (1996) for a different opinion.

²Other obstacles for trust in Bosnia include a lack of transparency in government procedures (UNDP 2003a) and a weak civil society (UNDP 2003b).

³World Value Survey, 1998, available at: www.worldvaluesurvey.org, accessed 7 September 2005.

⁴More informal types of social interactions have also been emphasised in later work.

⁵For more information on the survey, see: www.prismresearch.ba/eng/sind_research/sind_research_f2f_omnibus.htm, accessed 25 May 2007. See also P. Håkansson & S. Hargreaves, Trust in Transition: Generalised Trust in Bosnia and Herzegovina (Sarajevo, Balkan Analysis Group, 2004), Appendix 4, available at: http://www.balkansanalysis.org/reports/trust-in-transition.pdf, accessed 25 May 2007.

⁶It should be noted that Montenegro and Serbia were one country at the time of the study.

⁷The Central Bosnia and Neretva Cantons are divided in two regions each according to the ethnic majority.

⁸Our result of sub-national differences in trust corresponds to a similar finding in Romania where social capital is significantly higher in Transylvania compared to the rest of the country (Badescu & Sum 2005).

⁹The standard error measures the distance from the mean. If one group dominates, for example its proportion of the population is 0.9, and two other groups account for 0.05 of the population each, the standard error will turn out greater than if each of the three groups account for 0.33 of the population.

¹⁰Other studies have calculated ethnic homogeneity in a different way. For example, Sullivan (1991), and Zak and Knack (2000), used the proportion of the largest ethnic group. Such a measure might be less suitable with several ethnic groups. However, we recalculated our ETHNIC variable according to this method, with little effect on the results (not shown).

¹¹Today Gorazde belongs to the Croat and Bosniak dominated Federation (FBiH). Gorazde was of specific interest during the peace negotiation in Dayton (see for example Holbrook 1999).

¹²We follow the terminology used in Bosnia where Serbs, Bosniaks, and Croats are commonly, and slightly misleadingly, referred to as different nationalities.

¹³People with other nationality, All + Most: 13.6%. People with a different way of life, All + Most: 14.6%. The levels are not significantly different from the 14.7% of generalised ‘trusters’.

¹⁴See Gagnon Jr. (2004) for a similar view in the context of ethnicity and conflicts in the Balkans.