Chapter 5 rigorously tests the observable implications of the argument with longitudinal analysis. I conduct two panel analyses that isolate the effects of educational expansion from the “contaminating” effects of affirmative action policies. First, drawing on annual household surveys conducted by the census bureau, I construct a synthetic panel of birth cohorts to test the hypothesis that better-educated Brazilians situated in the lower classes are mostly likely to self-darken over time. The analysis supports this hypothesis and finds that this relationship holds across diverse cultural regions of Brazil. Next, I introduce an original panel dataset of Brazilian municipalities in 2000 and 2010 to explore whether spatial variation in educational expansion causes higher rates of reclassification within Brazil. Fixed-effects analysis again supports the hypothesis, showing that greater rates of high school and university attendance correlate with greater black identification. Additional analysis indicates that the hypothesized patterns are clearest in urban centers, and are not conditional on the presence of state-level affirmative action policies.

The familys economic and socio-cultural capital and how it is shared among its members influence a person’s capabilities and choices. This chapter posits that the family acts as a collective conversion factor, and presents a case study in Mali. A typology of household configurations that best expresses the diversity of family forms is built, then used to see the relation to the overall quality of life of household members, measured by goods (household assets) and opportunities (child education and women’s autonomy). The association between these configurations and children’s access to schooling, controlling for the household standard of living, is then considered, regression analysis results showing that access to education is correlated with the household standard of living, but there is also a household configuration net effect. Extended households seem better off and better suited to develop solidarity strategies that facilitate access to schooling. But the priority given to education also appears to play a role in differences between households, shown by a higher education rate of children in rural households headed by an educated man and in urban ones that are female-headed.

Chapter 3 reports on the evidence-gathering undertaken in Wales. By interviewing selected politicians and a range of senior civil servants charged with the formulation and implementation of language policy, an evaluation is made of current thinking on the relevance of the new speaker phenomenon within official language policy. These enquiries are supplemented by interviews with civil society policy formulators, decision-makers from national organisations and academic specialists who have studied the phenomenon. Several innovative interventions are identified and evaluated at both national and local level. But the neat division between governmental and civil society policy formulators is misleading. Neither category is impervious to the ideas of the other for in so many minority language communities there is a close relationship between them. In the most promising contexts this interaction is mutually stimulating, if not necessarily co-dependent, as together they fight for the survival of their threatened languages.

In this chapter I examine the statistical effect of industrialization on ethnic change. I first take Soviet-era cross-national data measuring ethnic diversity by country in 1961 and 1985 and regress the change in ethnic diversity across these twenty-four years on change in carbon emissions per capita over the same time period. The results demonstrate a strong relationship between decreasing ethnic diversity and increasing levels of industrialization, a result which is robust to the inclusion of several control variables and the use of various sub-samples, as well as alternative measures of industrialization such as cement production and urbanization. I also show that carbon emissions are robustly correlated with change in the percentage identifying with the largest ethnic group per state. I then use an alternative original dataset consisting of individual country censuses between 1960 and 2019, and show that the same effect holds, both as regards the effect of carbon emissions on ethnic fractionalization as well as robustness checks and multiple alternative measures such as electricity consumption and the share of labour in both agriculture and industry.

Here I examine in more detail my first case study, namely mid-twentieth-century Turkey. The presence of high-quality census data from Turkey between 1935 and 1965 allows me to replicate my statistical analysis from the previous chapter at the provincial level to show that urbanization is correlated with increasing levels of people identifying as Turkish; moreover, this relationship is robust to the inclusion of a number of different controls and the use of sub-samples. Qualitative examination of evidence from Turkey suggests that this process was a consequence of economic structural change and incentives provided by the state, such that incentivized assimilation took place in urban Turkey but not in the countryside. Indeed, I show that a lack of industrialization in Kurdistan was responsible for the lack of Turkish identification in the region, despite both violent and non-violent attempts by the Turkish state at assimilating its Kurdish population.

More than seventy years of England and Wales census data is available to search electronically. This chapter uses the digitized census data on London’s penal, semi-penal and voluntary institutions on census night 1881 to explore the social composition of incarcerated women. The census data shows that the prison population only counts a very specific category of female ‘deviants’, as they were predominantly young, unmarried and had low-status, unskilled and insecure occupation. Women in their mid-thirties and older, the married and widowed on the other hand only constituted a small minority of the prison population. This chapter argues that these women can be found in much greater numbers among other major state institutions like the workhouse and the public asylums. While men may have faced the brunt of penal discipline, deviant women were more often taken care of by semi-penal institutions, before but also sometimes after their conviction.

This chapter examines the evidence of language shifts to English in post-apartheid South Africa. We first present a brief overview of the language situation in South Africa and offer a review of the literature on language shift. A number of scholars have expressed concern that the maintenance of language diversity in South Africa is threatened by the dominance of English in schooling and the economy. However, other studies find that despite an increased use of English, mother-tongue languages have retained their cultural value as languages of identification. Moreover, recent quantitative work shows that among Africans the increased use of English is associated primarily with an increase in bilingualism rather than with the displacement of Bantu languages by English. We add to this literature with an analysis of Population Census data from 1996 to 2011 to investigate possible language shift from Afrikaans to English, particularly among Coloureds. We show that, despite a small decline in the use of Afrikaans as L1 and a corresponding rise in the use of English, the overall use of Afrikaans increased, because of the considerable growth in English–Afrikaans bilingualism.