Diversity indices and their potential application in ethnic studies

Abstract

The diversity indices – ways of compressing information about the number and proportion of groups in a given environment into a single number – usually serve as independent variables in social science research. They are useful tools to assess the potential impacts of ethnic, linguistic, religious, etc. diversity on economic development, political radicalisation, the risk of tensions or ‘social cohesion’. However, in order to realise their potential, it is crucial to understand the theoretical background of these indicators and their benefits and shortcomings in practice.

The main purpose of this analysis is to collect and classify the main diversity indices found in the international literature, to explain the principle of their operation and clarify the terminology. (Previously, authors often used different names for the same method, leaving room for misunderstandings.) The ‘basic types’ can be grouped according to the form of diversity they measure (fragmentation or polarisation), and whether they are weighted or not. Following Schaeffer’s suggestion, this paper applies a matrix based on Welzel’s factor scores of ‘sacredsecular’ and ‘protective-emancipative’ values from the World Values Survey, including people’s attitude toward women’s equality, the importance of national or religious pride, toleration of homosexuality and so forth. The standardised Euclidian distances between the average scores of the countries serve as ‘cultural’ weights in the weighted formulas.

In addition, based on 1,000 randomly generated ethnic compositions of a hypothetical settlement, the study draws general conclusions on the relationships between the ‘basic types’ of diversity indices. Although the Pearson’s correlation coefficients have been found to be relatively high in all cases, the spread of points for medium and high values is remarkable. For example, significantly different weighted values may belong to the same unweighted values, even if the proportions of ethnic groups are the same. Recognising this phenomenon can make a meaningful difference in assessing the potential social, economic or political impact of diversity. However, the application of these indices is particularly important in dynamic analyses (i.e. value changes between two points in time), because they can provide an accurate solution for measuring the pace and directions of homogenisation or heterogenisation. The choice of the optimal method depends on the research context, but there are several aspects to consider. This paper aims to assist scholars in selecting suitable diversity indices in ethnicity research.