Search

This paper examines the link between a mother's autonomy—the freedom and ability to think, express, make decisions and act independently—and the nutritional status of her children. We treat 'autonomy' as a latent variable, and design a novel statistical framework to measure this. This method allows us to separate the direct associations of maternal and family characteristics in our model for nutrition, from their indirect associations that work through maternal autonomy. Using data from India, we explore the sensitivity of our estimates to endogeneity caused by sample selection in the presence of son preference. We find: (i) a one standard deviation (SD) higher autonomy score is associated with a 0.16 SD higher height‐for‐age z‐score (HAZ score); and (ii) a 10% lower prevalence of stunting (HAZ < −2 SD). The latter is equivalent to the prevention of approximately 300,000 children stunting, indicating the important role of maternal autonomy. This paper is part of the Economica 100 Series. Economica, the LSE "house journal" is now 100 years old. To commemorate this achievement, we are publishing 100 papers by former students, as well as current and former faculty. Wiji Arulampalam obtained her PhD from the LSE. [ABSTRACT FROM AUTHOR]

The relationship between health care expenditure and health outcomes has been the subject of recent academic inquiry in order to inform cost-effectiveness thresholds for health technology assessment agencies. Previous studies in public health systems have relied upon data aggregated at the national or regional level; however, there remains debate about whether the supply side effect of changes to expenditure are identifiable using data at this level of aggregation. We use detailed patient data derived from electronic neonatal records across England along with routinely available cost data to estimate the effect of changes to patient expenditure on clinical health outcomes in a well-defined patient population. A panel of 32 neonatal intensive care units for the period 2009-2013 was constructed. Accounting for the potential endogeneity of expenditure a £100 increase in the cost per intensive care cot day (sample average cost: £1,127) is estimated to reduce the risk of mortality of 0.38 percentage points (sample average mortality: 11.0%) in neonatal intensive care. This translates into a cost per life saved in neonatal intensive care of approximately £420,000. [ABSTRACT FROM AUTHOR]

The empirical literature on the determinants of charities' donation income, distinguishing the charitable cause, is small. We consider the case of development charities specifically. Using a panel covering a quarter of a century, we observe a strong fundraising effect and a unitary household income elasticity. We find evidence that the conventionally identified 'price' effect may simply be the product of omitted variable bias. Our results further suggest that public spending on development crowds in private donations for development. We find a positive spillover effect of fundraising, suggesting the efforts of one development charity may increase contributions to other development charities. [ABSTRACT FROM AUTHOR]

This paper presents a convenient shortcut method for implementing the Heckman estimator of the dynamic random effects probit model and other dynamic nonlinear panel data models using standard software. It then compares the estimators proposed by Heckman, Orme and Wooldridge, based on three alternative approximations, first in an empirical model for the probability of unemployment and then in a set of simulation experiments. The results indicate that none of the three estimators dominates the other two in all cases. In most cases, all three estimators display satisfactory performance, except when the number of time periods is very small. [ABSTRACT FROM AUTHOR]

Context In the context of changing admissions criteria and an expanding medical school intake in the UK, we analysed the determinants of the medical school dropout probability. Objectives We aimed to analyse the determinants of the probability that a student will drop out of medical school during Year 1 and to compare the results of this analysis over time. Methods We carried out logistic regression analysis for the 6 intake cohorts of 1990–92 and 1998–2000. Results Between 1990–92 and 1998–2000, there were substantial increases in both the size of the entry cohort and the proportion of students dropping out of medical school. A logit model for the 1990–92 and 1998–2000 cohorts reveals that the probability of dropping out depended on both the medical school attended and the personal characteristics of the student, including academic preparedness. Almost none of the increase in the dropout rate between the 2 cohort groups can be explained by changes in observable characteristics of the students over this period. Instead, most of the increase in the dropout rate is associated with changes at the level of the institution and in unobserved student characteristics. Conclusions University effects, rather than changes in observed student characteristics, explain most of the increased dropout rate over the time period considered. Candidate explanations behind these effects include: less effective admissions policies; changing curricula; greater costs of attending medical school, and a growing mismatch between student and school characteristics. Testing between these competing hypotheses is left for future work. [ABSTRACT FROM AUTHOR]

In the context of the 1997 Report of the Medical Workforce Standing Advisory Committee, it is important that we develop an understanding of the factors influencing medical school retention rates. To analyse the determinants of the probability that an individual medical student will drop out of medical school during their first year of study. Binomial and multinomial logistic regression analysis of individual-level administrative data on 51 810 students in 21 medical schools in the UK for the intake cohorts of 1980–92 was performed. The overall average first year dropout rate over the period 1980–92 was calculated to be 3.8%. We found that the probability that a student would drop out of medical school during their first year of study was influenced significantly by both the subjects studied at A-level and by the scores achieved. For example, achieving 1 grade higher in biology, chemistry or physics reduced the dropout probability by 0.38% points, equivalent to a fall of 10%. We also found that males were about 8% more likely to drop out than females. The medical school attended also had a significant effect on the estimated dropout probability. Indicators of both the social class and the previous school background of the student were largely insignificant. Policies aimed at increasing the size of the medical student intake in the UK and of widening access to students from non-traditional backgrounds should be informed by evidence that student dropout probabilities are sensitive to measures of A-level attainment, such as subject studied and scores achieved. If traditional entry requirements or standards are relaxed, then this is likely to have detrimental effects on medical schools' retention rates unless accompanied by appropriate measures such as focussed student support. [ABSTRACT FROM AUTHOR]

This paper estimates the impact of work-related training on wage growth over the period 1981–91, using longitudinal data from the National Child Development Study, a cohort of young men aged 23 in 1981. A hurdle Negbin model is used to control for training endogeneity. We find that training incidence has a significant positive effect on wage growth. We also find that young men with a higher level of education are not only more likely to be trained, but are also more likely to experience substantially higher wage growth as a result. [ABSTRACT FROM AUTHOR]

Focuses on a study that estimated the determinants of membership in male trade unions from 1981 to 1991 in Great Britain. Use of data from the National Child Development Study; Methodology of the study; Results and discussion; Conclusions.

Random effects probit models impose the restriction that the correlation between successive error terms for the same individual is a constant and thus known in the literature as the `equicorrelation' model. But a static model can be estimated by pooling the data and ignoring this particular correlation structure to obtain consistent parameter estimates. Not all commercially available software prove calculations of marginal effects in very routinely estimated models such as the random effects probit model. These are of practical importance to applied researchers. Comparison between different random effects probit model coefficient estimates (marginal effects), and between these and the pooled probit coefficient estimates (marginal effects) can be very misleading for the very simple reason of the normalization that is implemented in software to facilitate easy estimation. This note discusses this point and provides the necessary adjustments to enable applied researchers to make valid comparisons in terms of coefficient estimates and marginal effects across different specifications.

Explores the nexus between work-related training and labour market flexibility using the first five wave of the British Household Panel Survey (BHPS) conducted over the period 1991-95. Relationship between general education and subsequent training; Inequality in access to work-related training; Lack of involvement in any work-related training to improve skills by short-term employees.

Background: In the context of the 1997 Report of the Medical Workforce Standing Advisory Committee, it is important that we develop an understanding of the factors influencing medical school retention rates., Aims: To analyse the determinants of the probability that an individual medical student will drop out of medical school during their first year of study., Method: Binomial and multinomial logistic regression analysis of individual-level administrative data on 51 810 students in 21 medical schools in the UK for the intake cohorts of 1980-92 was performed., Results: The overall average first year dropout rate over the period 1980-92 was calculated to be 3.8%. We found that the probability that a student would drop out of medical school during their first year of study was influenced significantly by both the subjects studied at A-level and by the scores achieved. For example, achieving 1 grade higher in biology, chemistry or physics reduced the dropout probability by 0.38% points, equivalent to a fall of 10%. We also found that males were about 8% more likely to drop out than females. The medical school attended also had a significant effect on the estimated dropout probability. Indicators of both the social class and the previous school background of the student were largely insignificant., Conclusions: Policies aimed at increasing the size of the medical student intake in the UK and of widening access to students from non-traditional backgrounds should be informed by evidence that student dropout probabilities are sensitive to measures of A-level attainment, such as subject studied and scores achieved. If traditional entry requirements or standards are relaxed, then this is likely to have detrimental effects on medical schools' retention rates unless accompanied by appropriate measures such as focussed student support.