Fractionalization and inter-group differences

Publisert i

Kyklos 60/2007, pp. 123-139

Sammendrag

Fractionalization is usually defined as the probability that two randomly chosen persons belong to different groups, be it ethnic, religious, linguistic, or other groups. High fractionalization leads to more corruption (Mauro 1995), low growth rates and bad policies in general (Easterly and Levine 1997), low provision of public goods (Alesina et al. 1999), less redistribution (Alesina et al. 2001,Lind 2006), less socialmixing and activity (Alesina and LaFerrara 2000), lower voluntary contributions to schools (Miguel 2004, Miguel and Gugerty 2005), and higher prevalence of civil war (Elbadawi and Sambanis 2002, Montalvo and Reynal-Querol 2005, Reynal-Querol 2002)1.

There are twomajor problems with themeasure of fractionalization used in virtually all studies of the effects of fractionalization. First, the choice of which groups to consider is often done in a more pragmatic than rigorous way. Second, the measure implies that two persons are either identical (belong to the same group) or totally different (belong to different groups). This misses the important point that some groupsmay be closer to each other than others. The objective of this paper is to develop a method to construct correct measures of
fractionalization that takes these two objections into account.

Fulltekst

By Jo Thori Lind
Published Mar. 10, 2008 9:52 AM - Last modified Nov. 20, 2017 2:35 PM