nep-exp New Economics Papers
on Experimental Economics
Issue of 2017‒11‒12
25 papers chosen by

  1. Nudging Generosity: Choice Architecture and Cognitive Factors in Charitable Giving By Schulz, Jonathan F.; Thiemann, Petra; Thöni, Christian
  2. I (Don't) Like You! But Who Cares? Gender Differences in Same Sex and Mixed Sex Teams By Leonie Gerhards; Michael Kosfeld
  3. Laws and Norms: Experimental Evidence with Liability Rules By Bruno Deffains; Claude Fluet; Romain Espinosa
  4. An Experimental Test of Gender Differences in Charitable Giving: Empathy Is at the Heart of the Matter By van Rijn, Jordan; Quinones, Esteban J.; Barham, Bradford L.
  5. Field experiments on the development of time preferences By James Andreoni; Michael Kuhn; John List; Anya Samek; Charles Sprenger
  6. Testing consumer theory: Evidence from a natural field experiment By Adena, Maja; Huck, Steffen; Rasul, Imran
  7. Salience of Law Enforcement: A Field Experiment By Robert Dur; Ben Vollaard
  8. Creating an efficient culture of cooperation By Ernst Fehr; Tony Williams
  9. Does Exposure to Unawareness Affect Risk Preferences? A Preliminary Result By Wenjun Ma; Burkhard Schipper
  10. Measuring and Bounding Experimenter Demand By Jonathan de Quidt; Johannes Haushofer; Christopher Roth
  11. Efficiency Consequences of Affirmative Action in Politics: Evidence from India By Das, Sabyasachi; Mukhopadhyay, Abhiroop; Saroy, Rajas
  12. Social norms and energy conservation beyond the US By Andor, Mark Andreas; Gerster, Andreas; Peters, Jörg; Schmidt, Christoph M.
  13. Status Inequality, Moral Disengagement and Violence By Armin Falk
  14. An experimental test of reporting systems for deception By Sascha Behnk; Iván Barreda-Tarrazona; Aurora García-Gallego
  15. Do Professional Norms in the Banking Industry Favor Risk-taking? By Alain Cohn; Ernst Fehr; Michel André Maréchal
  16. Animal Welfare and Human Ethics: A Personality Study By Konstanze Albrecht; Florentin Krämer; Nora Szech
  17. How is the Trade-off between Adverse Selection and Discrimination Risk Affected by Genetic Testing? Theory and Experiment By David Bardey; Philippe De Donder; Cesar Mantilla
  18. How Do Employers Use Compensation History?: Evidence from a Field Experiment By Moshe A. Barach; John Horton
  19. Crime and Violence: Desensitization in Victims to Watching Criminal Events By Rafael Di Tella; Lucía Freira; Ramiro H. Gálvez; Ernesto Schargrodsky; Diego Shalom; Mariano Sigman
  20. Measuring success in education: the role of effort on the test itself By Uri Gneezy; John List; Jeffrey Livingston; Xiangdong Qin; Sally Sadoff; Yang Xu
  21. Generalization in the Tropics: Development policy, randomized controlled trials, and external validity By Peters, Jörg; Langbein, Jörg; Roberts, Gareth
  22. Efficient Institutions and Effective Deterrence: On Timing and Uncertainty of Punishment By Johannes Buckenmaier; Eugen Dimant; Ann-Christin Posten; Ulrich Schmidt
  23. Institutional Choice and Cooperation in Representative Democracies: An Experimental Approach By Schories, Fanny E.
  24. Disguising Lies - Image Concerns and Partial Lying in Cheating Games By Kiryl Khalmetski; Dirk Sliwka
  25. Estimation and Inference of Heterogeneous Treatment Effects Using Random Forests By Wager, Stefan; Athey, Susan

  1. By: Schulz, Jonathan F. (Harvard University); Thiemann, Petra (Lund University); Thöni, Christian (University of Lausanne)
    Abstract: In an experimental setup we investigate the effect of two different choice architectures on donation decisions. In the treatment group, subjects can either specify a charity of their choice, or select one from a list of five well-known charities; in the control group we do not provide a list. In a sample of 869 subjects we find a large treatment effect: Offering a list of default charities doubles the fraction of donors, as well as the revenue for charities. We find that the treatment intervention particularly affects subjects who tend to make intuitive choices.
    Keywords: charitable giving, donation, choice architecture, defaults, affective reactions
    JEL: C93 D64 H41 L3
    Date: 2017–10
  2. By: Leonie Gerhards; Michael Kosfeld
    Abstract: We study the effect of likability on female and male team behavior in a lab experiment. Extending a two-player public goods game and a minimum effort game by an additional pre-play stage that informs team members about their mutual likability we find that female teams lower their contribution to the public good in case of low likability, while male teams achieve high levels of cooperation irrespective of the level of mutual likability. In mixed sex teams, both females’ and males’ contributions depend on mutual likability. Similar results are found in the minimum effort game. Our results offer a new perspective on gender differences in labor market outcomes: mutual dislikability impedes team behavior, except in all-male teams.
    Keywords: gender differences, likability, experiment, team behavior
    JEL: C90 J16
    Date: 2017
  3. By: Bruno Deffains; Claude Fluet; Romain Espinosa
    Abstract: We conduct an experiment where participants choose between actions that provide private benefits but may also impose losses on strangers. Three legal environments are compared: no law, strict liability for the harm caused to others and an efficiently designed negligence rule where damages are paid only when the harmful action causes a net social loss. Legal obligations are either perfectly enforced (Severe Law) or only weakly so (Mild Law), i.e.,material incentives are then nondeterrent. We investigate how legal obligations and social norms interact. Our results show that liability rules strengthen pro-social behavior and suggest that strict liability has a greater effect than the negligence rule.
    JEL: C91 K13 D03
    Date: 2017–10–30
  4. By: van Rijn, Jordan (University of WI and Credit Union National Association); Quinones, Esteban J. (University of WI and Center for Demography and Ecology); Barham, Bradford L. (University of WI and Center for Demography and Ecology)
    Abstract: This study uses a dictator game with a charitable organization as the donation recipient to test whether inequality aversion, empathic concern and feelings of manipulation explain gender differences in giving found in the literature. We first explore whether we can evoke these feelings in the lab by exogenously varying the content of a charitable appeal video. Then we examine whether the evoked feelings help explain heterogeneity in giving between males and females. We find that females donate significantly more than males in the treatments that include personal stories from children, with females donating 63 percent more than males in these treatments. Using instrumental variable (IV) methods, we also show that empathic concern that results from the videos with the children's personal stories increases average donations among females but not males. Although we evoke feelings of empathic concern and inequality aversion among males, this does not translate into increases in donations; on the other hand, empathic concern among females that is evoked via treatments with children's personal stories does lead to increases in average female donations. Our study is novel in demonstrating that females not only have larger stocks of empathic concern than do males, but also donate more in response to empathic concern that results from an emotional charitable appeal featuring children's stories. This highlights the importance of empathic concern in explaining gender differences in giving found in the literature.
    Date: 2017–09
  5. By: James Andreoni; Michael Kuhn; John List; Anya Samek; Charles Sprenger
    Abstract: Time preferences have been correlated with a range of life outcomes, yet little is known about their early development. We conduct a field experiment to elicit time preferences of nearly 1,000 children ages 3-12, who make several inter temporal decisions. To shed light on how such primitives form, we explore various channels that might affect time preferences, from background characteristics to the causal impact of an early schooling program that we developed and operated. Our results suggest that time preferences evolve substantially during this period with younger children displaying more impatience than older children. We also find a strong association with race: black children, relative to white or Hispanic children, are more impatient. Interestingly, parents of black children are also much more impatient than parents of white and Hispanic children. Finally, assignment to different schooling opportunities is not significantly associated with child time preferences.
    Date: 2017
  6. By: Adena, Maja; Huck, Steffen; Rasul, Imran
    Abstract: We present evidence from a natural field experiment designed to shed light on whether individual behavior is consistent with a neoclassical model of utility maximization subject to budget constraints. We do this through the lens of a field experiment on charitable giving. We find that the behavior of at least 80% of individuals, on both the extensive and intensive margins, can be rationalized within a standard neoclassical choice model in which individuals have preferences, defined over own consumption and their contribution towards the charitable good, satisfying the axioms of revealed preference.
    Keywords: natural field experiment,revealed preference
    JEL: C93 D01 D12 D64
    Date: 2017
  7. By: Robert Dur; Ben Vollaard
    Abstract: We conduct a field experiment to examine whether the deterrent effect of law enforcement can be strengthened by making law enforcement activities more salient. Our focus is on illegal disposal of household garbage in residential areas. At a random subset of 56 locations in a city in the Netherlands, law enforcement officers supplemented their regular enforcement activities by the practice of putting bright warning labels on illegally disposed garbage bags saying that the item was “Found by law enforcement; fine minimally 90 euros†. We find evidence for a substantial reduction in illegal disposal of garbage as a result of the treatment at locations with garbage disposal containers, but not at locations with glass/paper disposal containers. Overall, the estimated treatment effect is negative, but imprecisely estimated.
    Keywords: law enforcement, deterrence, perception, salience, disorder
    JEL: C93 K42
    Date: 2017
  8. By: Ernst Fehr; Tony Williams
    Abstract: Throughout human history, informal sanctions by peers were ubiquitous and played a key role in the enforcement of social norms and the provision of public goods. However, a considerable body of evidence suggests that informal peer sanctions cause large collateral damage and efficiency costs. This raises the question whether peer sanctioning systems exist that avoid these costs and whether other, more centralized, punishment systems are superior and will be preferred by the people. Here, we show that efficient peer sanctioning without much need for costly punishment emerges quickly if we introduce two relevant features of social life into the experiment: (i) subjects can migrate across groups with different sanctioning institutions and (ii) they have the chance to achieve consensus about normatively appropriate behavior. We also show that subjects universally reject peer sanctioning without a norm consensus opportunity –an institution that has hitherto dominated research in this field – in favor of our efficient peer sanctioning institution or an equally efficient institution where they delegate the power to sanction to an elected judge. Migration opportunities and normative consensus building are key to the quick emergence of an efficient culture of universal cooperation because the more prosocial subjects populate the two efficient institutions first, elect prosocial judges (if institutionally possible), and immediately establish a social norm of high cooperation. This norm appears to guide subjects’ cooperation and punishment choices, including the virtually complete removal of antisocial punishment when judges make the sanctioning decision.
    Keywords: Cooperation, punishment, endogenous institutions, public goods
    JEL: D02 D03 D72 H41
    Date: 2017–10
  9. By: Wenjun Ma; Burkhard Schipper (Department of Economics, University of California Davis)
    Abstract: One fundamental assumption often made in the literature on unawareness is that risk preferences are invariant to changes of awareness. We study how exposure to unawareness affects choices under risk. Participants in our experiment choose repeatedly between varying sure outcomes and a lottery in 3 phases. All treatments are exactly identical in phase 1 and phase 3, but differ in phase 2. There are five different treatments pertaining to the lottery faced in phase 2: The control treatment (i.e., a standard lottery), the treatment with awareness of unawareness of lottery outcomes but known number of outcomes, the treatment with awareness of unawareness of outcomes but with unknown number of outcomes, the treatment with unawareness of unawareness of some outcomes, and the treatment with an ambiguous lottery. We study both whether behavior differs in phase 3 across treatments (between subjects effect) and whether differences of subjects' behavior between phases 1 and phase 3 differs across treatments (within subject effects). We observe no significant treatment effects.
    Keywords: Unawareness, Awareness of unawareness, Risk aversion, Experiments
    JEL: C91 C92 D81 D87
    Date: 2017–05–02
  10. By: Jonathan de Quidt; Johannes Haushofer; Christopher Roth
    Abstract: We propose a technique for assessing robustness of behavioral measures and treatment effects to experimenter demand effects. The premise is that by deliberately inducing demand in a structured way we can measure its influence and construct plausible bounds on demand-free behavior. We provide formal restrictions on choice that validate our method, and a Bayesian model that microfounds them. Seven pre-registered experiments with eleven canonical laboratory games and around 19,000 participants demonstrate the technique. We also illustrate how demand sensitivity varies by task, participant pool, gender, real versus hypothetical incentives, and participant attentiveness, and provide both reduced-form and structural analyses of demand effects.
    Keywords: experimenter demand, beliefs, bounding
    JEL: B41 C91 C92
    Date: 2017
  11. By: Das, Sabyasachi (Ashoka University); Mukhopadhyay, Abhiroop (Indian Statistical Institute); Saroy, Rajas (Indian Statistical Institute)
    Abstract: We examine how overall delivery of public goods (i.e., efficiency) is affected by affirmative action in elections, i.e., restricting candidate entry in elections to one population group. We argue that when group identities are salient, such restrictions on candidate entry need not necessarily reduce electoral competition. In fact, when group sizes are asymmetric, affirmative action may increase electoral competition and consequently, improve provision of public goods. This happens because in an open election, the (best) candidate from the large group facing a minority candidate suffers from a moral hazard problem. Affirmative action eliminates this problem and increases within-group competition. We study a randomized caste based quota policy in village elections in a large state in India to test these claims. Consistently, we find that electoral quotas for a caste group (OBCs) increased provision of public goods in villages with high OBC population shares. We show that this did not happen due to changes in politicians' preferences or quality, and the increased provision of public goods did not disproportionately benefit the OBCs. Further, using election data, we show evidence in favor of our mechanism: win margins are narrower in quota elections relative to open elections in villages where OBC group is large. Our results highlight that efficiency concerns regarding affirmative action in politics may need reevaluation.
    Keywords: electoral competition, reservation, public goods, Gram Panchayat
    JEL: D72 D78 H41 O12
    Date: 2017–10
  12. By: Andor, Mark Andreas; Gerster, Andreas; Peters, Jörg; Schmidt, Christoph M.
    Abstract: The seminal studies by Allcott and Mullainathan (2010), Allcott (2011), and Allcott and Rogers (2014) suggest that social comparison-based home energy reports (HER) are a cost-effective non-price intervention to stimulate energy conservation. The present paper demonstrates the context-dependency of this result. We show that, outside the US, electricity consumption levels and carbon intensities are typically much lower and, hence, HER interventions can only become cost-effective when treatment effect sizes are substantially higher. Yet, our evidence from a large-scale randomized controlled trial in Germany suggests that effect sizes are actually much lower than in the US.
    Keywords: social norms,energy demand,external validity,randomized field experiments,non-price interventions
    JEL: D12 D83 L94 Q41
    Date: 2017
  13. By: Armin Falk
    Abstract: This paper studies the causal effect of status differences on moral disengagement and violence. To measure violent behavior, in the experiment, a subject can inflict a painful electric shock on another subject in return for money. We exogenously vary relative status in the realm of sexual attractiveness. In three between-subject conditions, the assigned other subject is either of higher, lower or equal status. The incidence of electric shocks is substantially higher among subjects matched with higher- and lower-status others, relative to subjects matched with equal-status others. This causal evidence on the role of status inequality on violence suggests an important societal cost of economic and social inequalities.
    Keywords: morality, violence, status, inequality, laboratory experiments
    JEL: A13 C91 D03 Z13
    Date: 2017
  14. By: Sascha Behnk (Department of Banking and Finance, University of Zurich, Switzerland); Iván Barreda-Tarrazona (LEE and Economics Department, Universitat Jaume I, Castellón, Spain); Aurora García-Gallego (LEE and Economics Department, Universitat Jaume I, Castellón, Spain)
    Abstract: We use a repeated sender-receiver game in which sender behavior is revealed to future counterparts by (i) standardized computer reports or by (ii) individual reports composed by the receivers, representing a common form of consumer feedback. Compared to our baseline without reporting, computer reports reduce deception in all payoff scenarios while the effect of individually written reports is lower and in some scenarios only marginal. This comparably weaker impact can be explained by the senders’ anticipation of a high number of missing or deficient receiver reports that we find. We conclude that the precision of a reporting system has a higher importance for reducing deception than its personal character via individual feedback. Surprisingly, the reliability of computer reports is not correctly anticipated by receivers, who trust individually written reports more in the beginning and hence seem to back the wrong horse initially.
    Keywords: deception, trust, reporting systems, reputation, experiment
    JEL: D03 D63 K42
    Date: 2017
  15. By: Alain Cohn; Ernst Fehr; Michel André Maréchal
    Abstract: In recent years, the banking industry has witnessed several cases of excessive risk-taking that frequently have been attributed to problematic professional norms. We conduct experiments with employees from several banks in which we manipulate the saliency of their professional identity and subsequently measure their risk aversion in a real stakes investment task. If bank employees are exposed to professional norms that favor risk-taking, they should become more willing to take risks when their professional identity is salient. We find, however, that subjects take significantly less risk, challenging the view that the professional norms generally increase bank employees’ willingness to take risks.
    Keywords: risk culture, banking industry, experiment
    JEL: G02 M14 C93
    Date: 2017
  16. By: Konstanze Albrecht; Florentin Krämer; Nora Szech
    Abstract: We revisit the long-standing question whether there is a relation between animal welfare and human ethics. Therefore, we elicit concern for animal welfare in an incentivized, direct, and real setup: Subjects choose between intensive farming versus organic living conditions for a hen. Guaranteeing organic living conditions is costly, but implies organic feed, access to daylight, and more space. We compare the interest in animal welfare with morally relevant dispositions in subjects, relying on well-established measures such as Machiavellianism scores and the Big 5 personality test. The data confirm a strong, positive relation between caring for animal welfare and moral dispositions.
    Keywords: animal welfare, human ethics, experiment, sustainability
    JEL: D01 D62 D69
    Date: 2017
  17. By: David Bardey; Philippe De Donder; Cesar Mantilla
    Abstract: We compare two genetic testing regulations, Disclosure Duty (DD) and Consent Law (CL), in an environment where individuals choose to take a genetic test or not. DD forces agents to reveal the test results to their insurers, resulting in a discrimination risk. CL allows agents to withhold that information, generating adverse selection. We complement our model with an experiment. We obtain that a larger fraction of agents test under CL than under DD, and that the proportion of individuals preferring CL to DD is non-monotone in the test cost when adverse selection is set endogenously at its steady state level.
    Keywords: consent law, disclosure duty, personalized medicine, test take up rate, pooling health insurance contracts
    JEL: C91 D82 I18
    Date: 2017
  18. By: Moshe A. Barach; John Horton
    Abstract: We report the results of a field experiment in which treated employers could not observe the compensation history of their job applicants. Treated employers responded by evaluating more applicants, and evaluating those applicants more intensively. They also responded by changing what kind of workers they evaluated: treated employers evaluated workers with 7% lower past average wages and hired workers with 16% lower past average wages. Conditional upon bargaining, workers hired by treated employers struck better wage bargains for themselves. Using a structural model of bidding and hiring, we find that the selection effects we observe would also occur in equilibrium.
    Keywords: field experiments, compensation, search and screening
    JEL: J01 J30 M50 M51
    Date: 2017
  19. By: Rafael Di Tella (Harvard Business School, Business, Government and the International Economy Unit); Lucía Freira (Universidad Torcuato Di Tella); Ramiro H. Gálvez (Universidad Torcuato Di Tella); Ernesto Schargrodsky (Universidad Torcuato Di Tella); Diego Shalom (Universidad Torcuato Di Tella); Mariano Sigman (Universidad Torcuato Di Tella)
    Abstract: We study desensitization to crime in a lab experiment by showing footage of criminal acts to a group of subjects, some of whom have been previously victimized. We measure biological markers of stress and behavioral indices of cognitive control before and after treated participants watch a series of real, crime-related videos (while the control group watches non-crime-related videos). Not previously victimized participants exposed to the treatment video show significant changes in cortisol level, heart rate, and measures of cognitive control. Instead, previously victimized individuals who are exposed to the treatment video show biological markers and cognitive performance comparable to those measured in individuals exposed to the control video. These results suggest a phenomenon of desensitization or habituation of victims to crime exposure.
    Keywords: crime, biological markers, experiment, victimization, desensitization
    JEL: K42
    Date: 2017–07
  20. By: Uri Gneezy; John List; Jeffrey Livingston; Xiangdong Qin; Sally Sadoff; Yang Xu
    Abstract: Tests measuring and comparing educational achievement are an important policy tool. We experimentally show that offering students extrinsic incentives to put forth effort on such achievement tests has differential effects across cultures. Offering incentives to U.S. students, who generally perform poorly on assessments, improved performance substantially. In contrast, Shanghai students, who are top performers on assessments, were not affected by incentives. Our findings suggest that in the absence of extrinsic incentives, ranking countries based on low-stakes assessments is problematic because test scores reflect differences in intrinsic motivation to perform well on the test itself, and not just differences in ability.
    Date: 2017
  21. By: Peters, Jörg; Langbein, Jörg; Roberts, Gareth
    Abstract: When properly implemented, Randomized Controlled Trials (RCT) achieve a high degree of internal validity. Yet, if an RCT is to inform policy, it is critical to establish external validity. This paper systematically reviews all RCTs conducted in developing countries and published in leading economic journals between 2009 and 2014 with respect to how they deal with external validity. Following Duflo, Glennerster, and Kremer (2008), the hazards to external validity we scrutinize are Hawthorne effects, general equilibrium effects, specific sample problems, and special care in treatment provision. Based on a set of objective indicators, we find that the majority of published RCTs does not discuss these hazards and many do not provide the necessary information to assess potential problems. The paper calls for including external validity dimensions in a more systematic reporting on the results of RCTs. This may create incentives to avoid overgeneralizing findings and help policy makers to interpret results appropriately.
    Keywords: policy evaluation,systematic review,internal validity,external validity,randomized controlled trials
    JEL: C83 C93
    Date: 2017
  22. By: Johannes Buckenmaier; Eugen Dimant (Philosophy, Politics and Economics, University of Pennsylvania); Ann-Christin Posten; Ulrich Schmidt
    Abstract: Reducing criminal acts in society is a crucial duty of governments. Establishing punishment structures to attain this goal involves high costs. Typically, both theorists and practitioners resort to the adjustment of severity and/or certainty of punishment as effective deterrents of criminal behavior. One more cost effective, but scientifically understudied mechanism for effective deterrence is the swiftness of punishment. We carry out the first controlled economic experiment to study the effectiveness of swiftness of punishment along the following two dimensions: the timing of punishment and the timing of the resolution of uncertainty regarding punishment. Our results indicate an inverted u-shaped relation between the delay of punishment, the delay of uncertainty resolution regarding the detection of deviant behavior, and any resulting deterrence. In fact, institutions that either reveal detection and impose punishment immediately or maintain uncertainty about the state of detection and impose punishment sufficiently late deter individuals at equal rates. We conclude that the same institutional settings that are capable of reducing recidivism are also the ones deterring deviant behavior in the first place. Our results yield strong policy implications for designing effective institutions in mitigating misconduct and reducing recidivism.
    Keywords: Experiment, Illicit Behavior, Institutions, Punishment, Uncertainty
    JEL: C91 D02 D81 K42
    Date: 2017–10
  23. By: Schories, Fanny E.
    Abstract: This paper examines whether an institution has a differing impact on cooperation if it is introduced by a representative of the affected parties rather than exogenously imposed. The experimental design is able to control for selection effects arising from the democratic policy choice. I find evidence of a large democracy premium in the sense that endogenously implemented institutions lead to more cooperation than iden- tical exogenous institutions. Especially the subjects who initially did not prefer the policy comply if it was brought about by an elected representative. The results have implications for the analysis of decision-making processes and policy recommendations in general.
    Keywords: Laboratory Experiment,Representative Democracy,Collective Decision-Making,Social Dilemma,Legitimacy
    JEL: C9 D02 D72
    Date: 2017
  24. By: Kiryl Khalmetski; Dirk Sliwka
    Abstract: We study equilibrium reporting behavior in Fischbacher and Föllmi-Heusi (2013) type cheating games when agents have a fixed cost of lying and image concerns not to be perceived as a liar. We show that equilibria naturally arise in which agents with low costs of lying randomize among a set of the highest potential reports. Such equilibria induce a distribution of reports in line with observed experimental patterns. We also find that higher image concerns lead to an increase in the range of reported lies while the effect of the fixed cost of lying is the opposite.
    Keywords: cost of lying, image concerns, cheating game, truth-telling, deception
    JEL: D03 D82 D83 C72
    Date: 2017
  25. By: Wager, Stefan (Stanford University); Athey, Susan (Stanford University)
    Abstract: Many scientific and engineering challenges--ranging from personalized medicine to customized marketing recommendations--require an understanding of treatment effect heterogeneity. In this paper, we develop a non-parametric causal forest for estimating heterogeneous treatment effects that extends Breiman's widely used random forest algorithm. In the potential outcomes framework with unconfoundedness, we show that causal forests are pointwise consistent for the true treatment effect, and have an asymptotically Gaussian and centered sampling distribution. We also discuss a practical method for constructing asymptotic confidence intervals for the true treatment effect that are centered at the causal forest estimates. Our theoretical results rely on a generic Gaussian theory for a large family of random forest algorithms. To our knowledge, this is the first set of results that allows any type of random forest, including classification and regression forests, to be used for provably valid statistical inference. In experiments, we find causal forests to be substantially more powerful than classical methods based on nearest-neighbor matching, especially in the presence of irrelevant covariates.
    Date: 2017–07

General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.