What is ethnicity? There’s lots of debate about the impacts it has on a country, but it strikes me that a fundamental question is what it is and how it might be measured or discussed. And that the answer might be less objective than we might hope, and depend on who’s asking/ answering the question.
Ethnicity is a complicated concept, and it is related to a number of different issues including nationality and religion. In part because it’s something people self-identify with, in addition to being an academic and political concept – it can be very personal. Our discussion in class about Equality Monitoring Forms reminded me of the connections between the labels presented to people, and the ones they might give themselves – or the ones they might reject. For example, British forms would not allow someone to identify as Kashmiri but would give the option of being of Indian origin.
Alesina et al discuss the difficulty in defining ethnicity, distinguishing between ethnic and linguistic variables, and how ethnic composition can change over time. They also find that while ethnic and linguistic fractionalisation are associated with negative outcomes (like Easterly and Levine), religious fractionalisation is not. I find this really interesting, because religion is often associated with conflict. They suggest that this is because religion is endogenous and can be banned or easier to hide in order to avoid repression. Therefore, higher religious fractionalisation can be a sign of a more tolerant/ democratic government. I don’t know, though, if this relationship is linear or if the data needs some disaggregation.
All the ways of trying to measure ethnic fractionalisation are interesting because they’re trying to make scientific sense of something that’s so political and personal to people. Posner discusses the problems with the commonly used ELF measure of ethnic diversity and proposes his own measures – the PREG. He is concerned that ELF data is 40 years out-of-date and some of the groupings are inaccurate. I find it interesting that this question of time comes up again (Scarritt also points to the ways ethnic identities are constructed and reconstructed over time), as potentially researchers are making big presumptions about the static composition of countries. This is probably particularly a problem in big cross-country studies. But we might think about how ethnicity is discussed in some countries and the problems of getting new data. We might think about Rwanda and how difficult it would be to examine the current ethnic breakdown of the country. But there are problems in the North too – in Britain ethnicity questions were dropped from the 1981 Census because of very high non-response in the 1979 Test Census resulting from fear of deportation.
Ethnicity is usually going to feature as part of politics and nation building – whether in an attempt to overcome it or as a way of dividing people. The trouble in investigating it is that there’s not much you can change as a result. It doesn’t tell actors in an ethnically diverse society much more than they might have a harder time than other countries. While it can explain, there probably needs to be more emphasis on what to do about it. And how some countries are able to sustain ethnic diversity and achieve their development goals.
Nicola's Development Politics Blog 