Measuring Regional Ethnolinguistic Diversity in Sub-Saharan Africa: Surveys vs. GIS

Abstract

This paper compares two approaches to measuring subnational ethnolinguistic diversity in Sub-Saharan Africa, one based on censuses and large-scale population surveys and the other relying on the use of geographic information systems (GIS). The two approaches yield sets of regional fractionalization indices that are moderately positively correlated, with a stronger association across rural areas. These differences matter for empirical analysis: in a common sample of regions, survey-based indices of deep-rooted diversity are more strongly negatively associated with a range of development indicators relative to their highest-quality GIS-based counterparts. AbstractThis paper compares two approaches to measuring subnational ethnolinguistic diversity in Sub-Saharan Africa, one based on censuses and large-scale population surveys and the other relying on the use of geographic information systems (GIS). The two approaches yield sets of regional fractionalization indices that are moderately positively correlated, with a stronger association across rural areas. These differences matter for empirical analysis: in a common sample of regions, survey-based indices of deep-rooted diversity are more strongly negatively associated with a range of development indicators relative to their highest-quality GIS-based counterparts.

Approaches to the Study of Human Capital within a Set of Socio-Economic and Socio-Cultural Factors on the Example of Yenisei Siberia

“…Moreover, the factors' integration within the human capital assessment as well as relative importance of these factors may also depend on the initial level of data aggregation. In this context, following (Gershman, Rivera, 2018, 2018a and (Desmet, Ortuno-Ortin, Wacziarg, 2012, 2017, we believe that there is some optimal level of aggregation that allows providing sufficient statistical validity of findings and recommendations without excess costs on collecting and processing primary data.…”

Subnational diversity in Sub-Saharan Africa: Insights from a new dataset

“…Since IWI supplements the data on the ownership of private goods with some elements reflective of local public goods provision (access to electricity and water), the negative association between the latter and deep-rooted diversity is presumably driving this result. 47 A more comprehensive analysis of this important societal cleavage, including its interaction with ethnic divisions, is beyond the scope of this paper, but is undertaken in Gershman and Rivera (2017).…”

“…Although fractionalization and polarization indices have dominated the empirical literature on ethnolinguistic and religious diversity, numerous other measures have been offered.30 To see whether the inclusion of the "none" group makes a big difference, we recalculated our indices by dropping those respondents. The correlation coefficients between pairs of RF and RP indices based on alternative classifications are equal to 0.89 and 0.85, respectively.InGershman and Rivera (2017), we construct subnational indices capturing the overlap between ethnic and religious cleavages and examine their relationship to conflict. In the same vein,Desmet et al (2017) construct country-level measures of the overlap between ethnicity and culture.…”

“…In addition, our methodology is in many ways preferable to the approach based on combining digital maps of ethnolinguistic groups with disaggregated population data, which is prone to measurement error due to inaccurate "homeland" boundaries, ad hoc aggregation of groups, noisy imputed regional population shares, and inability to capture high diversity in urban areas(Gershman and Rivera, 2018).3Miguel and Gugerty (2005),Alesina and La Ferrara (2005),Habyarimana et al (2007),Esteban and Ray (2011),Ashraf and Galor (2013), among many others, discuss various mechanisms plausibly linking diversity to social and economic outcomes.…”