There is a trivial yet telling authorial slip in Claire Adida's book. It comes on page 8, where she talks about the violence that broke out against foreigners in South Africa in ‘the spring of 2008’. It may well have been spring in North America, but in South Africa, as indeed in all of the southern hemisphere, May is, of course, an autumn month. I say that this is telling because a similar US-centrism marks the entire book. Processes of immigration and immigrant integration are depicted as ‘normal’ or ‘intuitive’ when they conform to processes of immigration or attitudes towards immigrants in the United States; counter-intuitive when they do not. The book's entire methodology is also very much reflective of contemporary social science in the United States, which has seen a return to quantitative methodology and statistics in ways reminiscent of the 1970s. Although welcome as a counter-balance to qualitative case study research, the adoption of statistical methods can lead to an unduly reductionist approach that seeks answers to questions, or rather proofs of hypotheses, framed in terms of its own inherent logic of indicators, variables and regression analyses.

In her analysis of immigration in West Africa, the independent variable is ethnicity, and in particular the cultural overlap between immigrants and their host societies, and the dependent variable the degree of immigrant exclusion. Through a study of two Nigerian ethnic groups, the Hausa and Yoruba, in the three West African cities of Accra in Ghana, Cotonou in Benin and Niamey in Niger, Adida presents findings from her own surveys and interviews to show that immigrants with higher levels of cultural overlap with their host communities are more, rather than less excluded by the host society, and also demonstrate higher levels of attachment to their own immigrant community. The two-group, three-country research design allows multiple cross-comparisons to be drawn, adding robustness to the empirical findings. The research strategy is further strengthened by separate surveys of immigrant and host society members, complemented by interviews with immigrant community leaders.

Immigrant in-group attachment was measured through an index incorporating answers to survey questions about voting practices, media use, passport held, travel and remittance behaviour, country of children's education, whether their current job was obtained through co-ethnic networks, and whether they identified as a Nigerian (i.e. foreigner) or host-country national. The high-overlap group (Yoruba in Accra and Cotonou, Hausa in Niamey) consistently showed higher levels of immigrant in-group attachment. To demonstrate immigrant exclusion by members of the host society, Adida presents a two-city comparison between Accra and Niamey, based on asking host society members what distinguishes them from Yoruba or Hausa and if they would vote for a Yoruba or Hausa presidential candidate. Respondents were consistently less likely to vote for members of higher-overlap than lower-overlap groups. Compared with the immigrant attachment index, this seems a rather crude indicator to adopt as a measure of exclusion, making the claims of this chapter rather less convincing. Adida explains high-overlap immigrants' exclusion, including self-exclusion, in terms of local-level bargaining, giving a central role to immigrant leaders who ‘resist immigrant assimilation because they benefit both socially and materially from their leadership position over a distinct immigrant community’ (p. 146).

Chapter 5 presents a separate, macro-scale analysis based on compilation of a data set of mass immigrant expulsions from African countries, subjecting these to a series of tests to identify any patterns. Few would argue with Adida's conclusion that expulsions are the product of a combination of economic, nation-building and socio-demographic factors, but this chapter is where the analysis becomes simply too reductionist, employing dubious proxy indicators, problematic data and questionable categorisations. Most curiously, she adopts oil exports and precipitation as proxy indicators for immigration flows, assuming that they acted as pulls to immigrant labour and thus contributed to socio-demographic pressure. Adida concludes that the ‘typical expelling country is a country whose ethnic minority leader, following the pull of an agricultural boom year, faces difficult economic conditions’ (p. 138).

Adida's analysis is indisputably thorough and systematic within its own quantitative framework, but in my view she strays too far in that direction. The bulk of the book is on the details of methodology and presentation of statistical analysis, leaving insufficient room for deeper interpretation and broader contextualisation, historical or political. Colonial legacies are dealt with in sweeping and over-generalised fashion. She provides scant reference to the literature on contemporary immigration in other African contexts. If she had read more widely, and less American-ly, she might have been less inclined to represent her findings as counter-intuitive or even especially novel. Anyone familiar with the literature on African immigration and ethnic relations, or who was paying attention to the targets of xenophobic violence in South Africa in 2008, would be entirely unsurprised to see further evidence of what David Matsinhe (Reference Matsinhe2011), drawing on Freud, describes as the narcissism of minor differences. Also open to debate, or at least moderation, is her repeated assertion that immigration is an economic positive, and that ‘African economies have perhaps the most to gain from immigrant economic activity’ (p. 145). This may well be true in certain places and circumstances, but is neither universal nor inevitable.

The book's title infers that the findings might be taken to apply more broadly in Africa than just the three case study cities. This is at best accidentally misleading and at worst an intentional marketing ploy. It also masks the fact that this is essentially a PhD dissertation. As such, the book perhaps has a place on the shelves of those few well-endowed university libraries that can afford to buy a copy of almost everything published. Individual scholars would be better advised to read Adida's already-published journal articles instead.