Eduardo is a statistician/econometrician who uses secondary data (as well as experiments) to understand the development human traits (particularly health and cognition), from childhood to the old age. His econometric and statistical skills are wide ranging, but the focus of his recent publications has been on causal inference and nonparametric methods (with Regression Discontinuity and Partial Identification occupying a significant amount of his recent work).  Much of his work can be framed within a school of thought emphasising that statistical and econometric modelling ought to rely on weak, credible and intelligible assumptions that allow data to speak for themselves. He has contributed, nonetheless, to the parametric literature. With Richard Hofler, he has proposed the Poisson Log-Half Normal distribution (a useful starting point for modelling under- or over-reported count data), and earlier on (2007) he developed the first Stochastic Frontiers models for discrete outcomes.

Eduardo’s applied work has covered health services in rural China, domestic abuse, the development of strategic sophistication in children, and the effect of early life experiences on adult outcomes. However, a significant amount of his work has been devoted to studying the effects of retirement on health, cognition and other outcomes. 

Eduardo has published his research in high impact international journals (such as the Journal of Political Economy, Journal of the Royal Statistical Society, Statistical Papers, Health Affairs, Health Economics or Journal of Productivity Analysis). You can see Eduardo's publications and some of his working also in his Google Scholar profile. 

Before joining Manchester, Eduardo was a Senior Researcher at the University of Oxford (Blavatnik School of Government and Health Economics Research Centre) and a Lecturer in Economics at the University of Strathclyde. Eduardo completed a Ph.D. in Econometrics at the University of Manchester thanks to the support of Fundación Ramón Areces.

Current projects.

Inference in medical cost effectiveness analysis with heavy tailed data

The bootstrap appears to be well suited for performing accurate inference in medical cost-effectiveness analysis, where data are generally drawn from a randomized experiment. However, the literature to date has given priority to asymptotic approximations. Cost data, though, often exhibit heavy tails and this presents challenges to asymptotic methods. In this paper, we study the performance of these methods for cost effectiveness analysis in medical research and compare them with inference based on the standard bootstrap. We find both methods fail to deliver reliable inference in the presence of heavy tails. We therefore consider two alternative resampling schemes: the wild bootstrap and randomization inference. Both methods provide accurate inference, providing the underlying statistic has finite expected value. In general, these methods provide inference with little size distortion and power above that achievable with asymptotic methods or the standard bootstrap. 

Retirement and Mental Health: Partial Identification with Panel Data

We study how retirement affects mental health   using a partial identification framework for panel data. We consider indentifying assumptions that restrict the process of selection, bind the amount of variation in counterfactual moments or limit the amount of interference within units along the time dimension.  The identification regions that we estimate are then transformed into quantitative policy recommendations by adopting a minimax-regret  decision rule.  We find that retirement can have a moderate, positive effect on younger individuals whose pre-retirement mental health is below average of the corresponding age cohort. Our minimax-regret analysis further suggests that while retirement can be delayed without affecting the mental health of the majority of the population, it is advisable to facilitate the retirement of people with poor mental health.

Exposure and contemporaneousness: What can we learn about the effect of drought on children's cognitive development?

With Y-Ling Chi (Imperial College)

In this paper, we combine a comprehensive Indonesian household survey with detailed meteorological data to explore what can be learned about the effect of drought on the long term cognitive development of Indonesian children living in rural areas. We face a common problem of latent exposure with observable contemporaneousness. The problem is compounded by plausible endogeneity and likely confounding. To estimate the effects of drought on Indonesian children's scores in a fluid intelligence test, we consider a battery of different identification assumptions which vary in credibility and power. Our most powerful assumptions point identify the effect of contemporaneousness, however they have debatable credibility. Our most credible assumptions, on the other hand, convey little information about the effect of contemporaneousness. In between these two extreme, we consider a range of   middle-of-the-way assumptions  which partially identify the effect of contemporaneousness. Specifically, we characterise some of the assumptions required to establish the sign of the effect of contemporaneousness. Our results reveal differential effects of drought across sexes, however we find at least two competing explanations which would explain these difference: natural selection, on the one hand, and family dis-investing on girls in the fact of hardship, on the other.

What can we learn about the Average Treatment Effect of retirement on consumption?

We present a nonparametric bounds analysis of the Average Treatment Effect (ATE) of retirement on domestic expenditure. We consider identification under a wide catalogue of assumptions which restrict either the potential outcomes or the process of selection into treatment. We put forward new assumptions that exploit the longitudinal information in panels to refine existing results regarding the partial identification of ATEs. The tightest identification region suggest that retirement could lead to a drop of up to 7% in expenditure. However, the sign of the actual effect is not identified unless one restricts the variation in potential outcomes across households or one assumes that, on average, retirement is detrimental for consumption. The latter assumption might be suitable only in environments like the United States or the United Kingdom, where the pension replacement rate are low. We  find that savings and education have mitigating effects on the magnitude of the maximum potential drop in expenditure.

I will be happy to supervise Ph.D. students in any of the following areas:

• Econometrics and Statistics (Theory or applied: Causal inference; Partial identification; Nonpametric methods)
• Experimental Economics and Experimental Pyschology (strategic thinking; social norms)
• Ageing and retirement.
• Childhood development in developed and developing countries.