<P> Much discussion among researchers centers around defining and measuring two key but related variables: ethnicity and diversity . It is debated whether ethnicity should be defined by culture, language, or religion . While conflicts in Rwanda were largely along tribal lines, Nigeria's string of conflicts is thought to be--at least to some degree--religiously based . Some have proposed that, as the saliency of these different ethnic variables tends to vary over time and across geography, research methodologies should vary according to the context . Somalia provides an interesting example . Due to the fact that about 85% of its population defined themselves as Somali, Somalia was considered to be a rather ethnically - homogeneous nation . However, civil war caused ethnicity (or ethnic affiliation) to be redefined according to clan groups . </P> <P> There is also much discussion in academia concerning the creation of an index for "ethnic heterogeneity". Several indices have been proposed in order to model ethnic diversity (with regards to conflict). Easterly and Levine have proposed an ethno - linguistic fractionalization index defined as FRAC or ELF defined by: </P> <Dl> <Dd> 1 − ∑ i = 1 N s i 2, (\ displaystyle 1 - \ sum _ (i = 1) ^ (N) s_ (i) ^ (2),) </Dd> </Dl> <Dd> 1 − ∑ i = 1 N s i 2, (\ displaystyle 1 - \ sum _ (i = 1) ^ (N) s_ (i) ^ (2),) </Dd>

Three strands in the development of the definition of economics