this paper we consider two fundamental issues concerning Robins' theory. Firstly, do his assumed relations (between observed and unobserved---factual and counterfactual---random variables) place restrictions on the distribution of the observed variables. If the answer is yes, adopting his approach means making restrictive implicit assumptions---not very desirable. If however the answer is no, his approach is neutral. One can freely use it in modelling and estimation, exploring the consequences (for the unobserved variables) of the model. This follows the highly succesful tradition in all sciences of making thought experiments. In what philosophical sense counterfactuals actually exist seems to us less relevant. But it is important to know if a certain thought experiment is a priori ruled out by existing data

We describe a computer program to assist a clinician with assessing the efficacy of treatments in experimental studies for which treatment assignment is random but subject compliance is imperfect. The major difficulty in such studies is that treatment efficacy is not "identifiable", that is, it cannot be estimated from the data, even when the number of subjects is infinite, unless additional knowledge is provided. Our system combines Bayesian learning with Gibbs sampling using two inputs: (1) the investigator's prior probabilities of the relative sizes of subpopulations and (2) the observed data from the experiment. The system outputs a histogram depicting the posterior distribution of the average treatment effect, that is, the probability that the average outcome (e.g., survival) would attain a given level, had the treatment been taken uniformly by the entire population. This paper describes the theoretical basis for the proposed approach and presents experimental results on ...

Abstract not found

Four alternative but related approaches to empirical evaluation of policy interventions are studied: social experiments, natural experiments, matching methods, and instrumental variables. In each case the necessary assumptions and the data requirements are considered for estimation of a number of key parameters of interest. These key parameters include the average treatment effect, the treatment of the treated and the local average treatment effect. Some issues of implementation and interpretation are discussed drawing on the labour market programme evaluation literature.

How large are the benefits of transportation infrastructure projects, and what explains these benefits? To shed new light on these questions, I collect archival data from colonial India and use it to estimate the impact of India’s vast railroad network. Guided by six predictions from a general equilibrium trade model, I find that railroads: (1) decreased trade costs and interregional price gaps; (2) increased interregional and international trade; (3) eliminated the responsiveness of local prices to local productivity shocks (but increased the transmission of these shocks between regions); (4) increased the level of real income (but harmed neighboring regions without railroad access); (5) decreased the volatility of real income; and that (6), a sufficient statistic for the effect of railroads on welfare in the model accounts for virtually all of the observed reduced-form impact of railroads on real income. I find similar results from an instrumental variable specification, no spurious effects from over 40,000 km of lines that were approved but never built, and tight bounds on the estimated impact of railroads. These results suggest that transportation infrastructure projects can improve welfare significantly,

An important objective of empirical research on treatment response is to provide decision makers with information useful in choosing treatments. This paper studies minimax-regret treatment choice using the sample data generated by a classical randomized experiment. Consider a utilitarian social planner who must choose among the feasible statistical treatment rules, these being functions that map the sample data and observed covariates of population members into a treatment allocation. If the planner knew the population distribution of treatment response, the optimal treatment rule would maximize mean welfare conditional on all observed covariates. The appropriate use of covariate information is a more subtle matter when only sample data on treatment response are available. I consider the class of conditional empirical success rules; that is, rules assigning persons to treatments that yield the best experimental outcomes conditional on alternative subsets of the observed covariates. I derive a closed-form bound on the maximum regret of any such rule. Comparison of the bounds for rules that conditional on smaller and larger subsets of the covariates yields sufficient sample sizes for productive use of covariate information. When the available sample size exceeds the sufficiency boundary, a planner can be certain that conditioning treatment choice on more covariates is preferable (in terms of minimax regret) to conditioning on fewer covariates.

this paper I (1) examine three levels of inferential strength supported by typical social science data-gathering methods, and call for a greater degree of explicitness, when HMs and other models are applied, in identifying which level is appropriate; (2) reconsider the use of HMs in school effectiveness studies and meta-analysis from the perspective of causal inference; and (3) recommend the increased use of Gibbs sampling and other Markov-chain Monte Carlo (MCMC) methods in the application of HMs in the social sciences, so that comparisons between MCMC and better-established fitting methods---including full or restricted maximum likelihood estimation based on the EM algorithm, Fisher scoring or iterative generalized least squares---may be more fully informed by empirical practice.

Causal effect modeling with naturalistic rather than experimental data is challenging. In observational studies participants in different treatment conditions may also differ on pretreatment characteristics that influence outcomes. Propensity score methods can theoretically eliminate these confounds for all observed covariates, but accurate estimation of propensity scores is impeded by large numbers of covariates, uncertain functional forms for their associations with treatment selection, and other problems. This paper demonstrates that boosting, a modern statistical technique, can overcome many of these obstacles. We illustrate this approach with a study of adolescent probationers in substance abuse treatment programs. Propensity score weights estimated using boosting eliminate most pretreatment group differences, and substantially alter the apparent relative effects of adolescent substance abuse treatment. Experimental studies offer the most rigorous evidence with which to establish treatment efficacy, but they are not always practical or feasible. Experimental treatment evaluations can be expensive to field and may be too slow to produce answers to pressing questions. In some cases

Causality is a very intuitive notion that is difficult to make precise without lapsing into tautology. Two ingredients are central to any definition: (1) a set of possible outcomes (counterfactuals) generated by a function of a set of ‘‘factors’ ’ or ‘‘determinants’ ’ and (2) a manipulation where one (or more) of the ‘‘factors’ ’ or ‘‘determinants’’ is changed. An effect is realized as a change in the argument of a stable function that produces the same change in the outcome for a class of interventions that change the ‘‘factors’ ’ by the same amount. The outcomes are compared at different levels of the factors or generating variables. Holding all factors save one at a constant level, the change in the outcome associated with manipulation of the varied factor is called a causal effect of the manipulated factor. This definition, or some version of it, goes back to Mill (1848) and Marshall (1890). Haavelmo’s (1943) made it more precise within the context of linear equations models. The phrase ‘ceteris paribus’ (everything else held constant) is a mainstay of economic analysis

In many cases, up to a third of male youth (including children) are drawn into armed groups, making soldiering one of the world’s most common occupations for the young. Little is known, however, about the impacts of military service on human capital and labor market outcomes due to an absence of data as well as sample selection: recruits are usually self-selected and screened, and may also selectively survive. We assess the impacts of participation in civil war using an original survey from Uganda, where a rebel group’s recruitment method provides arguably exogenous variation in conscription. Contrary to the prevailing view that participation in war leads to broad-based ‘traumatization’, we find that military service primarily hinders long-term economic performance because it is a poor substitute for civilian education and work experience. The most significant impact is upon a recruit’s skills and productivity: schooling falls by nearly a year, skilled employment halves, and earnings drop by a third. These impacts are highly robust to relaxation of the assumption of exogenous conscription. Effects are greatest for child soldiers, who lose the most education. There is no observed impact on social capital, and adverse impacts