Search

Home ownership is the largest component of wealth for most households and its intergenerational transmission underpins the production and reproduction of economic inequalities across generations. Yet, little is currently known about ethnic differences in the intergenerational transmission of housing tenure. In this paper we use linked Census data covering 1971-2011 to document rates of intergenerational housing tenure mobility across ethnic groups in England and Wales. We find that while home ownership declined across all ethnic groups during this period, there were substantial differences between them. Black, Pakistani and Bangladeshi households experienced the strongest intergenerational link between parent and child housing tenure, and Black individuals had the highest rates of downward housing mobility. In contrast, those of Indian origin had homeownership rates similar to White British families, and a weaker link between parent and child housing tenure. These patterns are likely to exacerbate existing gradients in other dimensions of ethnicity-based inequality, now and in the future. Juli 2023

Using a large and novel administrative dataset, this paper investigates variation in returnsto different higher education 'degrees' (subject-institution combinations) in the United King-dom. Conditioning on a rich set background characteristics, it finds substantial variation inreturns, even within subject, across universities with very similar selectivity levels, suggest-ing degree choices matter a lot for later-life earnings. Selectivity is weakly related to returnsthrough most of the distribution but strongly positively correlated at the top end. Other thanselectivity, returns are poorly correlated with observable degree characteristics, which has im-plications for student choices and the incentives of universities.

In this paper we use linked census data to assess whether an academically selective schooling system promotes social mobility, using England as a case study. Over a period of two decades, the share of pupils in academically selective schools in England declined sharply and differentially by area. Using a sample of census records matched to administrative data on selective system schooling within local areas, we exploit temporal and geographic variation to estimate the effects of the selective schooling system on absolute and relative social class mobility. Our results provide no support for the contention that the selective schooling system increased social mobility in England, whether considered in absolute or relative terms. The findings are precisely estimated and robust to a comprehensive battery of robustness checks.

Social mobility—the extent to which social and economic position in adulthood is facilitated or constrained by family origins—has taken an increasingly prominent role in public and policy discourse. Recent studies have documented that not only who your parents are, but also where you grow up, influences subsequent life chances. We bring these two concepts together to study trends in social mobility in England and Wales, in three post-war generations, using linked Decennial Census data. We estimate rates of occupational social class mobility by sex and region of origin. Our findings show considerable spatial variation in rates of absolute and relative mobility as well as how these have changed over time. While rates of upward mobility increased in every region between the mid-1950s and the early 1980s, this upward shift varied across different parts of the country, and tailed off for more recent cohorts. We also explore the role of domestic migration in understanding these temporal and spatial patterns, finding that those who stayed in their region of origin had lower rates of upward mobility compared to those who moved out, although this difference also narrowed over time. While policy discussion has focused almost entirely on national-level trends in social mobility, our results emphasise the need to also consider persistent spatial inequalities.

Using German register data for the period between 1970 and 2010, this paper studies the labor market consequences of being born during the German food crisis after World War II. The paper finds that those born during the first half of 1946---about nine months after the period of most severe malnutrition---have significantly lower earnings than those born shortly before and after. The effect persists throughout their entire labor market careers, so that their total earnings losses accumulate to about one and a half year of average annual earnings. In addition, those affected by the malnutrition early in life have lower employment prospects and retire earlier. The paper provides novel insights to better understand how nutritional conditions early in life influence economic outcomes over the entire life course.

This paper re-examines the wage returns to the 1972 Raising of the School Leaving Age (RoSLA) in England and Wales using a high-quality administrative panel dataset covering the relevant cohorts for almost 40 years of their labour market careers. With best practice regression discontinuity methods we find at best a zero return to the additional education for men. However, we contend that regression discontinuity methods in this context will give unreliable estimates of the return. Using the panel data to correct for this we find a local average treatment effect of 7% over the lifetime for this additional year of education.

This paper re-examines the wage returns to the 1972 Raising of the School Leaving Age (RoSLA) in England and Wales using a high-quality administrative panel dataset covering the relevant cohorts for almost 40 years of their labour market careers. With best practice regression discontinuity methods we find at best a zero return to the additional education for men. However, we contend that regression discontinuity methods in this context will give unreliable estimates of the return. Using the panel data to correct for this we find a local average treatment effect of 7% over the lifetime for this additional year of education.

This paper examines the consequences of introducing a normally distributed effect into a system where the dependent variable is ordered and the explanatory variable is ordered and endogenous. Using simulation techniques we show that a naive bivariate ordered probit estimator which fails to take a mixed effect into account will result in inconsistent estimates even when identification conditions are optimal. Our results suggest this finding only applies to non-linear endogenous systems.

Social, or 'generalized', trust refers to beliefs that people hold about how other people in society will in general act towards them. Can people in general be trusted? Or must one be careful in dealing with people? Research on the antecedents of social trust has typically relied on cross-sectional regression estimators to evaluate putative causes. Our contention is that much of this research over-estimates the importance of many of these causes because of the failure to account for unmeasured confounding influences. In this paper we use longitudinal data assess the causal status of a particularly prominent mooted cause of trust: the degree to which individuals are socially integrated via formal membership of civic organisations and through friendship networks. We fit a range of regression estimators to repeated measures data from the UK for the period 1998 to 2008. Our results show little support for the widely held view that social trust results from integration within social networks, of either a formal or an informal nature.

This paper investigates the returns to lifelong learning, which is interpreted as the attainment of qualifications following entry into the labour market. For a number of reasons our analysis of the British Household Panel Survey (BHPS) represents an important addition to the existing evidence base. We allow for financial and non-financial returns to lifelong learning by using as dependent variables both (i) hourly earnings and (ii) CAMSIS score. A fixed effects specification counters the potential biases that arise from unobserved individual heterogeneity and the inclusion of lags allows estimation of how the returns to lifelong learning evolve over a ten year period after the qualification is obtained. We find evidence of earnings and occupation status returns using a broad categorisation of lifelong learning for both men and women, but more variability in returns when disaggregated NVQ-equivalent categories of qualification are considered. Our findings are broadly in line with existing evidence within the UK, which is mostly based on the analysis of cohort studies. 0f particular interest is the finding that returns to women materialise much sooner after the attainment of a qualification, than is the case for their male counterparts.

This paper investigates the relationship between educational attainment and truancy. Using data from the Youth Cohort Study of England and Wales, we estimate the causal impact that truancy has on educational attainment at age 16. Problematic is that both truancy and attainment are measured as ordered responses requiring a bivariate ordered probit model to account for the potential endogeneity of truancy. Furthermore, we extent the naive bivariate ordered probit estimator to include mixed effects which allows us to estimate the distribution of the truancy effect on educational attainment. This estimator offers a more flexible parametric setting to recover the causal effect of truancy on education and results suggest that the impact of truancy on education is indeed more complex than implied by the naive estimator.

In this paper, we discuss the derivation and application of a bivariate ordered probit model with mixed effects. Our approach allows one to estimate the distribution of the effect (gamma) of an endogenous ordered variable on an ordered explanatory variable. By allowing gamma to vary over the population, our estimator offers a more flexible parametric setting to recover the causal effect of an endogenous variable in an ordered choice setting. We use Monte Carlo simulations to examine the performance of the maximum likelihood estimator of our system and apply this to a relevant example from the UK education literature.

Using American panel data from the National Educational Longitudinal Study of 1988 (NELS:88) this paper investigates the effect of working during grade 12 on attainment. We exploit the longitudinal nature of the NELS by employing, for the first time in the related literature, a semiparametric propensity score matching approach combined with difference-in- differences. This identification strategy allows us to address in a flexible way selection on both observables and unobservables associated with part-time work decisions. Once such factors are controlled for, insignificant effects on reading and math scores are found. We show that these results are robust to a matching approach combined with difference-in-difference-in-differences which allows differential time trends in attainment according to the working status in grade 12.