Counterfactuals and causal inference: Methods and principles for social research

"Most quantitative empirical analyses are motivated by the desire to estimate the causal effect of an independent variable on a dependent variable. Although the randomized experiment is the most powerful design for this task, in most social science research done outside of psychology, experimental designs are infeasible. (Winship & Morgan, 1999, p. 659)." This quote from earlier work by Winship and Morgan, which was instrumental in setting the groundwork for their book, captures the essence of our review of Morgan and Winship's book: It is about causality in nonexperimental settings.

Lewis’ Theory of Counterfactuals and Essentialism

La logique contemporaine a connu de nombreux développements au cours de la seconde moitié du siècle dernier. Le plus sensationnel est celui de la logique modale et de sa sémantique des mondes possibles (SMP) dû à Saul Kripke dans les années soixante. Ces dans ce cadre que David Lewis exposera sa sémantique des contrefactuels (SCF). Celle-ci constitue une véritable excroissance de l’architecture kripkéenne. Mais sur quoi finalement repose l’architecture kripkéenne elle-même ? Il semble bien que la réponse soit celle d’une ontologie raffinée ultimement basée sur la notion de mondes possible. Ce mémoire comporte quatre objectifs. Dans un premier temps, nous allons étudier ce qui distingue les contrefactuels des autres conditionnels et faire un survol historique de la littérature concernant les contrefactuels et leur application dans différent champs du savoir comme la philosophie des sciences et l’informatique. Dans un deuxième temps, nous ferons un exposé systématique de la théorie de Lewis telle qu’elle est exposée dans son ouvrage Counterfactuals. Finalement, nous allons explorer la fondation métaphysique des mondes possible de David Lewis dans son conception de Réalisme Modal.

Examining moderators of the hindsight bias in the context of civil legal decision -making: Counterfactuals, causal proximity, and self -referencing

The current research sought to clarify the diverging relationships between counterfactual thinking and hindsight bias observed in the literature thus far. In a non-legal context, Roese and Olson (1996) found a positive relationship between counterfactuals and hindsight bias, such that counterfactual mutations that undid the outcome also increased participants’ ratings of the outcome’s a priori likelihood. Further, they determined that this relationship is mediated by causal attributions about the counterfactually mutated antecedent event. Conversely, in the context of a civil lawsuit, Robbennolt and Sobus (1997) found that the relationship between counterfactual thinking and hindsight bias is negative. The current research sought to resolve the conflicting findings in the literature within a legal context. ^ In Experiment One, the manipulation of the normality of the defendant’s target behavior, designed to manipulate participants’ counterfactual thoughts about said behavior, did moderate the hindsight effect of outcome knowledge on mock jurors’ judgments of the foreseeability of that outcome as well as their negligence verdicts. Although I predicted that counterfactual thinking would increase, or exacerbate, the hindsight bias, as found by Roese and Olson (1996), my results provided some support for Robbenolt and Sobus’s (1997) finding that counterfactual thinking decreases the hindsight bias. Behavior normality did not moderate the hindsight effect of outcome knowledge in Experiment Two, nor did causal proximity in Experiment Three. ^ Additionally, my hypothesis that self-referencing may be an effective hindsight debiasing technique received little support across the three experiments. Although both the self-referencing instructions and self-report measure consistently decreased mock jurors’ likelihood of finding the defendant negligent, and self-referencing instructions decreased their foreseeability ratings in studies two and three, the self-referencing manipulation did not interact with outcome knowledge to moderate a hindsight bias effect on either foreseeability or negligence judgments. The consistent pattern of results across the three experiments, however, suggests that self-referencing may be an effective technique in reducing the likelihood of negligence verdicts.^

Examining Moderators of the Hindsight Bias in the Context of Civil Legal Decision-Making: Counterfactuals, Causal Proximity, and Self-Referencing

The current research sought to clarify the diverging relationships between counterfactual thinking and hindsight bias observed in the literature thus far. In a non-legal context, Roese and Olson (1996) found a positive relationship between counterfactuals and hindsight bias, such that counterfactual mutations that undid the outcome also increased participants’ ratings of the outcome’s a priori likelihood. Further, they determined that this relationship is mediated by causal attributions about the counterfactually mutated antecedent event. Conversely, in the context of a civil lawsuit, Robbennolt and Sobus (1997) found that the relationship between counterfactual thinking and hindsight bias is negative. The current research sought to resolve the conflicting findings in the literature within a legal context. In Experiment One, the manipulation of the normality of the defendant’s target behavior, designed to manipulate participants’ counterfactual thoughts about said behavior, did moderate the hindsight effect of outcome knowledge on mock jurors’ judgments of the foreseeability of that outcome as well as their negligence verdicts. Although I predicted that counterfactual thinking would increase, or exacerbate, the hindsight bias, as found by Roese and Olson (1996), my results provided some support for Robbenolt and Sobus’s (1997) finding that counterfactual thinking decreases the hindsight bias. Behavior normality did not moderate the hindsight effect of outcome knowledge in Experiment Two, nor did causal proximity in Experiment Three. Additionally, my hypothesis that self-referencing may be an effective hindsight debiasing technique received little support across the three experiments. Although both the self-referencing instructions and self-report measure consistently decreased mock jurors’ likelihood of finding the defendant negligent, and self-referencing instructions decreased their foreseeability ratings in studies two and three, the self-referencing manipulation did not interact with outcome knowledge to moderate a hindsight bias effect on either foreseeability or negligence judgments. The consistent pattern of results across the three experiments, however, suggests that self-referencing may be an effective technique in reducing the likelihood of negligence verdicts.

Antipassives in Yukulta

Yukulta’s antipassive construction is obligatorily used to code transitive propositions that involve counterfactuals or marked A-O relationships. Its use and the use of active transitive constructions are strictly regulated by these grammatical features, to the extent that the two constructions have a complementary distribution. The particular functions of Yukulta’s antipassive and its highly conditioned usage are atypical for antipassives cross-linguistically.

Machine ethics via logic programming

Machine ethics is an interdisciplinary field of inquiry that emerges from the need of imbuing autonomous agents with the capacity of moral decision-making. While some approaches provide implementations in Logic Programming (LP) systems, they have not exploited LP-based reasoning features that appear essential for moral reasoning. This PhD thesis aims at investigating further the appropriateness of LP, notably a combination of LP-based reasoning features, including techniques available in LP systems, to machine ethics. Moral facets, as studied in moral philosophy and psychology, that are amenable to computational modeling are identified, and mapped to appropriate LP concepts for representing and reasoning about them. The main contributions of the thesis are twofold. First, novel approaches are proposed for employing tabling in contextual abduction and updating – individually and combined – plus a LP approach of counterfactual reasoning; the latter being implemented on top of the aforementioned combined abduction and updating technique with tabling. They are all important to model various issues of the aforementioned moral facets. Second, a variety of LP-based reasoning features are applied to model the identified moral facets, through moral examples taken off-the-shelf from the morality literature. These applications include: (1) Modeling moral permissibility according to the Doctrines of Double Effect (DDE) and Triple Effect (DTE), demonstrating deontological and utilitarian judgments via integrity constraints (in abduction) and preferences over abductive scenarios; (2) Modeling moral reasoning under uncertainty of actions, via abduction and probabilistic LP; (3) Modeling moral updating (that allows other – possibly overriding – moral rules to be adopted by an agent, on top of those it currently follows) via the integration of tabling in contextual abduction and updating; and (4) Modeling moral permissibility and its justification via counterfactuals, where counterfactuals are used for formulating DDE.

US Monetary and Fiscal Policies - Conflict or Cooperation?

Most of the literature estimating DSGE models for monetary policy analysis ignores fiscal policy and assumes that monetary policy follows a simple rule. In this paper we allow both fiscal and monetary policy to be described by rules and/or optimal policy which are subject to switches over time. We find that US monetary and fiscal policy have often been in conflict, and that it is relatively rare that we observe the benign policy combination of an conservative monetary policy paired with a debt stabilizing fiscal policy. In a series of counterfactuals, a conservative central bank following a time-consistent fiscal policy leader would come close to mimicking the cooperative Ramsey policy. However, if policy makers cannot credibly commit to such a regime, monetary accommodation of the prevailing fiscal regime may actually be welfare improving.

Organizational Choice and Environmental Change

We explain the choice between franchising and vertical integration by estimating a model of relative performance in a sample of 250 Spanish car distributors, controlling for self-selection and including environmental factors. The method allows us to estimate performance counterfactuals. Organizational choice seemingly aims to contain moral hazard for both distributors and manufacturers but it is subject to start-up constraints and switching costs. While the market for franchises remained underdeveloped, information asymmetries led to the opening of integrated outlets. Their subsequent conversion into franchised outlets probably involved prohibitive transaction costs. Consequently, they performed worse than would have been expected had they been independent, as confirmed by the systematic improvement observed when they were in fact converted. The timing of such conversions suggests that switching costs were prohibitive until firms developed a substantial cushion of temporary contracts, previously forbidden by regulation.

Democracy over governance

Varieties of institutional economics are available to evaluate varieties of capitalism. These methodologies dig behind preferences and technology to arrive at the ground on which agents make choices. The individual is at the foundation of these edifices, neoclassical and otherwise. Consequently, the denouement of all these models is that the market knows best in the absence of effective counterfactuals. A natural corollary is that the task of the government is to set effective mechanisms in place in order to approach the best outcomes. In contrast, we propose a framework which contends with the modern economy as an aggregate that evolves in historical time. Problems like effective demand failures are endemic to capitalist economies. Therefore, systematic State intervention is essential to their functioning. In particular, political economy teaches us that intervention must be in the interest of wage earners. In contrast to the earlier model, the fabric of norms and conventions that facilitate the growth and development of economies must emerge from the consciousness and practices of the working class.

Imagination et perception morale

Dans cette thèse en psychologie morale, je m’intéresse au rôle de l’imagination dans la perception morale. Je soutiens que l’imagination y a une fonction épistémique dans la mesure où – en s’accompagnant ou non d’émotions – elle nous révèle des normes, des valeurs ou des vertus morales qui seraient autrement passées inaperçues. En simulant des croyances et des perceptions, l’imagination nous permet d’accéder à ces caractéristiques d’une situation moralement pertinentes, mais perceptuellement non saillantes. J’identifie trois modes de « perception morale imaginative » : 1) la prise de perspective qui consiste à endosser le point de vue d’autrui, 2) le cadrage imaginatif qui désigne le fait de voir un élément d’une situation comme autre qu’il n’est et, 3) la comparaison imaginative qui, grâce à la pensée contrefactuelle, éclaire le monde actuel à partir d’un monde possible imaginé. Chacun de ces modes contribue à enrichir notre connaissance morale, et partant, à améliorer notre délibération morale. J’appuie ma démonstration sur des travaux récents en philosophie de la psychologie, en psychologie cognitive et sociale, en neuropsychologie et, bien évidemment, en psychologie morale.

The National Audit Office uses OR to assess the value for money of public services

Supreme audit institutions (SAIs) have an important role in assessing value for money in the delivery of public services. Assessing value for money necessarily involves assessing counterfactuals: good value for money has been achieved if a policy could not reasonably have been delivered more efficiently, effectively, or economically. Operations research modelling has the potential to help in the assessment of these counterfactuals. However, is such modelling too arcane, complex, and technically burdensome for organisations that, like SAIs, operate in a time- and resource-constrained and politically charged environment? We report on three applications of modelling at the UK's SAI, the National Audit Office, in the context of studies on demand management in tax collection, end-of-life care, and health-care associated infections. In all cases, the models have featured in the audit reports and helped study teams come to a value-for-money judgment. We conclude that OR modelling is indeed a valuable addition to the value-for-money auditor's methodological tool box.

Partially identified treatment effects under imperfect compliance: the case of domestic violence

The Minneapolis Domestic Violence Experiment (MDVE) is a randomized social experiment with imperfect compliance which has been extremely influential in how police officers respond to misdemeanor domestic violence. This paper re-examines data from the MDVE, using recent literature on partial identification to find recidivism associated with a policy that arrests misdemeanor domestic violence suspects rather than not arresting them. Using partially identified bounds on the average treatment effect I find that arresting rather than not arresting suspects can potentially reduce recidivism by more than two-and-a-half times the corresponding intent-to-treat estimate and more than two times the corresponding local average treatment effect, even when making minimal assumptions on counterfactuals.

Geographical externalities and the Pacifying Police Units Program in Rio de Janeiro

The paper quantifies the effects on violence and police activity of the Pacifying Police Unit program (UPP) in Rio de Janeiro and the possible geographical spillovers caused by this policy. This program consists of taking selected shantytowns controlled by criminals organizations back to the State. The strategy of the policy is to dislodge the criminals and then settle a permanent community-oriented police station in the slum. The installation of police units in these slums can generate geographical spillover effects to other regions of the State of Rio de Janeiro. We use the interrupted time series approach proposed by Gonzalez-Navarro (2013) to address effects of a police when there is contagion of the control group and we find that criminal outcomes decrease in areas of UPP and in areas near treated regions. Furthermore, we build a model which allows to perform counterfactuals of this policy and to estimate causal effects in other areas of the State of Rio de Janeiro outside the city.

Theoretical and empirical essays on microeconometrics

This Master Thesis consists of one theoretical article and one empirical article on the field of Microeconometrics. The first chapter\footnote{We also thank useful suggestions by Marinho Bertanha, Gabriel Cepaluni, Brigham Frandsen, Dalia Ghanem, Ricardo Masini, Marcela Mello, Áureo de Paula, Cristine Pinto, Edson Severnini and seminar participants at São Paulo School of Economics, the California Econometrics Conference 2015 and the 37\textsuperscript{th} Brazilian Meeting of Econometrics.}, called \emph{Synthetic Control Estimator: A Generalized Inference Procedure and Confidence Sets}, contributes to the literature about inference techniques of the Synthetic Control Method. This methodology was proposed to answer questions involving counterfactuals when only one treated unit and a few control units are observed. Although this method was applied in many empirical works, the formal theory behind its inference procedure is still an open question. In order to fulfill this lacuna, we make clear the sufficient hypotheses that guarantee the adequacy of Fisher's Exact Hypothesis Testing Procedure for panel data, allowing us to test any \emph{sharp null hypothesis} and, consequently, to propose a new way to estimate Confidence Sets for the Synthetic Control Estimator by inverting a test statistic, the first confidence set when we have access only to finite sample, aggregate level data whose cross-sectional dimension may be larger than its time dimension. Moreover, we analyze the size and the power of the proposed test with a Monte Carlo experiment and find that test statistics that use the synthetic control method outperforms test statistics commonly used in the evaluation literature. We also extend our framework for the cases when we observe more than one outcome of interest (simultaneous hypothesis testing) or more than one treated unit (pooled intervention effect) and when heteroskedasticity is present. The second chapter, called \emph{Free Economic Area of Manaus: An Impact Evaluation using the Synthetic Control Method}, is an empirical article. We apply the synthetic control method for Brazilian city-level data during the 20\textsuperscript{th} Century in order to evaluate the economic impact of the Free Economic Area of Manaus (FEAM). We find that this enterprise zone had positive significant effects on Real GDP per capita and Services Total Production per capita, but it also had negative significant effects on Agriculture Total Production per capita. Our results suggest that this subsidy policy achieve its goal of promoting regional economic growth, even though it may have provoked mis-allocation of resources among economic sectors.