BOULU RESHEF Beatrice

The energy transition cannot be achieved without a better understanding of the energy consumption behaviors of individuals. The first chapter describes the characteristics, the stakes and the effectiveness of energy demand control policies. The second chapter focuses on financial incentives to reduce energy consumption. Based on laboratory experiments, our results show that financial incentives (presented as a bonus in this thesis) serve as benchmarks for subjects in their energy consumption. They encourage the least sober individuals to strive for energy sobriety but, on the contrary, partially crowd out the intrinsic motivations of the most sober.The last chapter studies the impact of the form of communication. The last chapter examines the impact of the form of communication, which can either emphasize negative externalities or stress the positive externalities created by consumption and encourage prevention. We study the impact of this framing, with a non-repeated linear public good game experiment with simultaneous decisions, also taking into account the effect of risk on decisions. Our results show that participants cooperated more in a risky universe than in a certain one, and even more with the framing that emphasizes losses. Uncertainty favorably modifies extreme behaviors, especially for women.

This thesis explores the impact of information on job search behavior and demand for vocational training.The first chapter evaluates the impact of the Bob Emploi website, which aims to deliver information to job seekers about the labor market. The results indicate that there is no impact on job seekers' search effort and geographic and sectoral scope of the search. However, job seekers who use Bob Emploi mobilize their personal network and public employment services more. Finally, there is no effect on well-being and return to employment.The second chapter examines the role of information on entering job training. Results indicate that receiving an email with a message emphasizing post-training return-to-work opportunities more than doubles the likelihood that job seekers will call the training center again. However, callback rates are low in absolute terms (less than 1%) and there is no impact on training enrollment. Our results suggest that the impact detected on calls is due more to the increase in importance placed on training information than to the updating of job seekers' beliefs.Finally, the third chapter also studies the demand for vocational training, but takes into account behavioral constraints. Distinguishing between "external" (world) and "internal" (self) beliefs, the results show that jobseekers are financially constrained from joining a training program, and that they underestimate the proportion of available training that is funded. Internal barriers related to self-efficacy, inter-temporal preferences, self-esteem, and organizational skills are mentioned equally by jobseekers indicating internal barriers to enrolling in training. Based on this diagnosis, the last part is devoted to the design of a randomized controlled trial, with interventions based on the transmission of information via online courses, and interactive sessions by groups of job seekers. These courses aim to target external beliefs, internal beliefs, or both simultaneously.

Standard economic models assume that individuals collect and process information in a way that gives them a relatively accurate perception of reality. However, this assumption is often challenged. Evidence shows that individuals often form positive biases about themselves, which can have adverse economic consequences. This thesis aims to explain the persistence of overconfidence in social interactions by showing the existence of strategic benefits of overconfidence that outweigh its social cost.Using a series of laboratory experiments, this thesis shows that (i) overconfidence occurs primarily when it provides an advantage in social interactions (Chapter 2) and (ii) identifies situations in which overconfidence is likely to harm society (Chapters 3 and 4). This thesis contributes to the literature by improving our understanding of the situational determinants of overconfidence in social interactions and lays the groundwork for improving policies to prevent or limit negative effects.

Voluntary contributions are often solicited in sequential and public settings where information on the quality of the fundraising project unfolds with the sequence of decisions. This paper examines how the different sources of information available to potential donors in such settings influence their decision-making. Contrary to most of the leadership literature, neither leaders nor followers in these settings have certainty about the quality of the fundraising project. We explore whether leaders remain influential, the extent to which they use their influence strategically, and the consequences on followers when leaders are misinformed. We combine an information cascade method with a modified public goods game to create a “Voluntary Contributions in Cascades” paradigm. Participants sequentially receive private signals about the state of the world, which determines the potential returns from the public good, and take two public actions: an incentivized prediction about the state of the world and a contribution to the public good. We find that participants' predictions mostly align with Bayesian predictions, and find no evidence for strategic or misleading predictions. Leaders' contributions are positively correlated with followers', suggesting they remain influential despite their limited informational advantage. This influence takes a tragic turn when leaders happen to be misinformed, as most misinformed leaders end up unintentionally misleading followers. We find that having a misleading leader is associated with a reduction in gains from contributions roughly twice as large as the reduction that stems from dividing the marginal-per-capita-return by two. Our results stress the significance of having well-informed leaders.

No summary available.

The term "money illusion" was first coined by the American economist Irving Fisher in 1928 to describe the phenomenon whereby individuals do not perceive the increase or decrease in value of monetary units. Over the following decades, economists, psychologists and behavioural scientists first debated the existence of the money illusion, before refining and expanding its definition. This thesis builds on previous research on money illusion, with the goal of identifying, using experimental methods, the characteristics that explain the occurrence of money illusion at the individual level, as well as the factors that may help reduce its occurrence. This thesis first summarizes the evolution of the concept of money illusion from Fisher's discovery to the latest advances in behavioral science on this subject. It then presents the results of three separate experiments conducted to identify several factors that we have hypothesized contribute to the money illusion or to its reduction. The first focuses on the relationship between money illusion, financial education, and computation. The second focuses specifically on computation and numerical skills in general as a means of overcoming money illusion. and the third on whether temporal perspectives can explain and/or mitigate money illusion.

Parochial altruism-individual sacrifice to benefit the in-group and harm an out-group-undermines inter-group cooperation and is implicated in a plethora of politically-significant behaviors. We report new experimental findings about the impact of variation in social distance within the in-group together with variation in social distance between the in-and out-groups on parochial altruism. Building from a minimal group paradigm setup , we find that differential social distance has a systematic effect on individual choice in a setting of potential inter-group conflict. In particular, parochial altruism is stimulated when individuals' distance to both their in-and out-group is high. A long-standing finding about behavior in contexts of inter-group conflict is that low social distance facilitates collective action. Our results suggest that the effects of high social distance may create a potential additional pathway to group-based individual action. Research on inter-group conflict and collective action can be advanced by investigating such effects.

This article proposes an experimental approach that makes it possible to identify the individuals most likely to form a cartel: the "hotheads." The experiments test the deterrent effectiveness of different anti-cartel schemes by comparing the propensities of individuals to collude under different types of sanctions-fine, leniency, compliance, and debarment-and detection-detection with exogenous probability or whistleblowing in the case of leniency. Compliance and debarment are particularly deterrent; leniency is not. The deterrent effect of high fines is limited for "hotheads," who are more influenced by the magnitude of the risk of detection. Differences in gender and risk aversion affect behavior among other participants but not among individuals described as "hotheads.

In this thesis, we use the Nudge approach to solve behavioral problems that companies may face. In the first chapter, we first outline some of the problems that classical economics struggles to solve, then we introduce the Nudge approach and explain how it can be an effective alternative. In the second chapter, we use Nudge to change the wording of invitations to participate in internet surveys in order to increase the participation rate. We find that most Nudges succeed in convincing individuals to leave their email address, but that only those that acknowledge their efforts improve the participation rate. In the third chapter, we use Nudge to develop very short satisfaction surveys, administered monthly, and compare the results to those of a longer survey at the end of the course. We find that satisfaction in the first month is strongly correlated with final satisfaction, which makes it possible to detect possible problems very early. In the final chapter, we use Nudges to make a simple task more fun, an approach called "gamification". Nudges increase productivity as much as monetary incentives, without costing as much as monetary incentives. Furthermore, Nudges improve intrinsic motivation when they are not implemented at the same time as incentives. We conclude with some practical advice for decision makers who want to try Nudge.

Parochial altruism-individual sacrifice to benefit the in-group and harm an out-group-undermines inter-group cooperation and is implicated in a plethora of politically-significant behaviors. We report new experimental findings about the impact of variation in social distance within the in-group together with variation in social distance between the in-and out-groups on parochial altruism. Building from a minimal group paradigm setup , we find that differential social distance has a systematic effect on individual choice in a setting of potential inter-group conflict. In particular, parochial altruism is stimulated when individuals' distance to both their in-and out-group is high. A long-standing finding about behavior in contexts of inter-group conflict is that low social distance facilitates collective action. Our results suggest that the effects of high social distance may create a potential additional pathway to group-based individual action. Research on inter-group conflict and collective action can be advanced by investigating such effects.

This article proposes an experimental approach that makes it possible to identify the individuals most likely to form a cartel: the "hotheads." The experiments test the deterrent effectiveness of different anti-cartel schemes by comparing the propensities of individuals to collude under different types of sanctions-fine, leniency, compliance, and debarment-and detection-detection with exogenous probability or whistleblowing in the case of leniency. Compliance and debarment are particularly deterrent; leniency is not. The deterrent effect of high fines is limited for "hotheads," who are more influenced by the magnitude of the risk of detection. Differences in gender and risk aversion affect behavior among other participants but not among those labeled "hotheads.

This thesis is organized around two themes: social norms and production networks. The first chapter focuses on a case study where social norms are used in the fight against online hate speech. Using machine learning methods, I show that denouncing hateful opinions is a way to deter further hateful speech. This effect is explained by the fact that this form of contradiction serves to communicate the presence of a social norm or accentuates its importance. Beyond anti-discrimination, this chapter can shed light on how norms influence political behavior more generally. The second chapter focuses on the role that social image tastes play in explaining the effect of social norms on behavior. Many studies show that these tastes affect people's behavior on average, but we do not yet know which individuals are most likely to adjust their behavior. I present a novel experiment designed to fill this gap. It computes an individual measure of image concern, shows that there is substantial heterogeneity, and analyzes its correlation with other social preferences. The last chapter studies production networks. Vertical integration of firms can lead to anticompetitive behavior. I discuss one such behavior, called foreclosure, in which vertically integrated firms cut off the supply of essential inputs to their competitors. I use new data on production networks to identify mergers and acquisitions between vertically related firms. I show these mergers affect the supply chains of their competitors and interpret this as evidence of foreclosure.

Several recent studies have shown that investors in asset markets form their expectations by extrapolating the most recent evolution of an asset's price. The aim of the experiment is to confront individuals assumed to be "rational" with "extrapolators" in order to characterize forecasting behavior and to see if the presence of "extrapolators" affects the behavior of "rational" individuals. To do this, we simulated the price of an asset over fifteen periods that fluctuates around its fundamental value. The experiment consists of five treatments. In order to differentiate between the two types of participants, we vary the information available to them to make their price prediction: the so-called "extrapolators" only have access to a window of five previous periods, while the so-called "rational" participants have access to the entire history up to the forecast period. Therefore, we posit two modes of remuneration: one based on the accuracy of the forecasts and the other based on the deviation from the average forecast. Thus, one group of each type of participant will have a different remuneration mode . corresponding to four treatments. Finally, the last group will be composed, in equal proportions, of "extrapolators" and "rationals" and will have an objective of deviation from the average forecast. This will allow us to see if the "rationals" correct their forecast when they are in the presence of "extrapolators". [source: abstract].

Several recent studies have shown that investors in asset markets form their expectations by extrapolating the most recent evolution of an asset's price. The aim of the experiment is to confront individuals assumed to be "rational" with "extrapolators" in order to characterize forecasting behavior and to see if the presence of "extrapolators" affects the behavior of "rational" individuals. To do this, we simulated the price of an asset over fifteen periods that fluctuates around its fundamental value. The experiment consists of five treatments. In order to differentiate between the two types of participants, we vary the information available to them to make their price prediction: the so-called "extrapolators" only have access to a window of five previous periods, while the so-called "rational" participants have access to the entire history up to the forecast period. Therefore, we posit two modes of remuneration: one based on the accuracy of the forecasts and the other based on the deviation from the average forecast. Thus, one group of each type of participant will have a different remuneration mode . corresponding to four treatments. Finally, the last group will be composed, in equal proportions, of "extrapolators" and "rationals" and will have an objective of deviation from the average forecast. This will allow us to see if the "rationals" correct their forecast when they are in the presence of "extrapolators". [source: abstract].

This paper proposes an experimental approach to the effectiveness of anti-cartel schemes, which result from individual decisions to engage in these anticompetitive cartel practices. Our experiments compare individual propensities to form cartels under monetary sanctions, leniency, compliance and exclusionary schemes. Our study thus assesses the impact of sanctioning schemes, their levels and different probabilities of detection, and identifies the influence of certain individual characteristics, such as gender and risk aversion, on cartel formation. Our results show that the debarment scheme, and to a lesser extent the compliance scheme, are the most effective in deterring cartel formation. The level and probability of punishment have the expected effects, but in non-linear ways. Leniency, on the other hand, increases the effectiveness of fines. Finally, gender and risk aversion influence the propensity to choose to engage in a cartel, but not their rate of formation. The implications of these results for regulators and firms are substantial: they provide insight into how to better deter these illegal practices.

Abstract When taking into account the chances of success, strategic mating motivations do imply a bias not toward the most attractive individuals, but toward average or mildly attractive individuals, undermining the explanation of Maestripieri et al. at a fundamental level. This leaves open the possibility of alternative explanations and calls for a full-fledged explicit model of courtship behavior.

We study a finitely repeated public goods game (based on the voluntary contribution mechanism) played under complete uncertainty about the marginal benefit of the public good relative to the private consumption (commonly known as the marginal per capita return): neither one's marginal per capita return nor other players' marginal per capita returns are known at the time of decision-making. We show that contributions are equivalent when the rate of return is predetermined and when it is uncertain, and display a similar decay over time. Combined with the previous experimental findings, our results suggest that the cooperation in public goods games is sensitive to the source of uncertainty about marginal per capita return.

We study a finitely repeated public goods game (based on the voluntary contribution mechanism) played under complete uncertainty about the marginal benefit of the public good relative to the private consumption (commonly known as the marginal per capita return): neither one's marginal per capita return nor other players' marginal per capita returns are known at the time of decisionmaking. We show that contributions are equivalent when the rate of return is predetermined and when it is uncertain, and display a similar decay over time. Combined with the previous experimental findings, our results suggest that the cooperation in public goods games is sensitive to the source of uncertainty about marginal per capita return.

No summary available.

My main research areas are experimental economics and managerial economics. My work is at the interface of economics and management because the issues I study have both an economic and a managerial dimension. My research activities focus on individual behaviors in organizational contexts. In all my work, published and in progress, I use laboratory experiments and/or framed-field experiments for the empirical part. The experimental procedures I develop aim at describing precisely the information, incentives and interaction sequences as they are in the contexts I study. My general research theme is the question of resource allocation behavior in organizational contexts. More specifically, my research is organized around the following main themes: First, I study the effect of identity, affiliation, social distance, and leadership on resource allocation behaviors in organizational contexts; second, the effect of uncertainty and risk aversion on resource allocation behaviors in organizational contexts; and third, the effect of information utilization methodologies on resource allocation behaviors in organizational contexts, including limited-capacity firms and entrepreneurial firms.

No summary available.

To make better decisions today, companies and other economic agents are interested in getting accurate predictions of future events. Prediction markets can, at least potentially, give those accurate forecasts for the probability of the event by aggregating information from traders. However, formal studies highlight that the risk attitudes of market participants may bias the market equilibrium prices, and consequently make the prediction unreliable. This research examines the effect of participants' risk attitudes on prediction market prices, through a framed field experiment on the two semifinals at the 2015 NCAA Men's Division Basketball Tournament. The results of the experiment show a significant price difference between the risk-averse group and the less risk-averse group. The large price discrepancy between markets with participants with varying risk aversion suggests that risk aversion deserves a critical consideration in future prediction-market research and implementation.

Prediction- and control-based strategies are the two main hypotheses of how entrepreneurs deal with uncertainty in theories of entrepreneurship. Prediction-based strategies focus on estimating unknowns via sampling methods, whereas control-based strategies focus on shaping unknowns via proactive behavior. These strategies may lead to different propensities to undertake uncertain prospects, as they differ in terms of cognition and involvement. In an experimental test, we study the conditions under which prediction- and control-based strategies lead subjects to accept bets in ambiguous environments. Individuals who use control methods to mitigate uncertainty are more likely to accept the bet after a favorable outcome compared to those who use predictive methods. These results revert in the presence of unfavorable outcomes. We discuss the implications for entrepreneurship theory and practice.

Organizations often match leaders with followers in order to foster cooperation, mitigate free riding, and thereby accomplish tasks effectively. This paper studies the effect of organization styles on leaders' choices of leadership strategy, and the effect of this choice on free-riding behavior. In controlled experiments, leaders are asked to choose messages from a message set that induces a leadership style and to send it to followers in a repeated and finite horizon public goods game. When provided with collegial leadership style messages, leaders perform better than their top down counterparts, but only when targeted private communication is not allowed. When it is allowed, leaders who are instructed to be top down perform better by focusing on individuals, not the group, and by leading while accounting for contributor types. The paper uncovers core and understudied mechanisms of top down leadership, and challenges the consensus that collegiality is best for leading groups.

This paper studies the effect of firm boundaries on inter-individual cooperation in corporate settings. I conduct a framed-field linear public good game experiment in the natural environment of the workplace. I use information about the actual boundaries of two firms, which are a parent firm and its subsidiary. I document differences in the overall level of cooperation across the parent firm, its subsidiary, and an intra-corporate group that comprises players from both the parent firm and the subsidiary. In stark contrast to previous results, I find that cooperation is increasing over time within the parent firm, indicating that firms can foster stronger cooperation within their boundaries. In other treatments the standard declining cooperation is not rejected. Overall, I find higher levels of contribution to the public good relative to conventional laboratory experiments. I estimate the importance of individuals' identity and find that higher weight placed on identity is associated with higher contributions, but only within the parent firm. Higher levels of self-awareness can help explain patterns of cooperation within the boundaries of the firm.

Despite the boom in research on identity in economics, the question of how to depict personal identity in a fashion that combines potentially conflicting social identities has so far been largely ignored. This paper introduces a framework that relies on the structure of the social realm that is exogenous to the individual in order to depict investment decisions as per social identities. The structure of the environment is identified using organizational boundaries and identity production functions are used to combine social identities. The inclusion of the various social spaces that are relevant to personal identity strategies enables one to simultaneously study identification strategies and individuation strategies. It also allows the depiction of social identities that may be conflicting, due to different social commitments and/or norm valuations. This externalist conception of identity helps discuss the limitations of internalist conceptions of identity and accounts for the heterogeneity of identity strategies across individuals.

No summary available.

This thesis develops identity theory in economics by using the boundaries of the firm to study the effect of firm identification on intrafirm cooperation. We build on John B. Davis' definition of personal identity and propose a theoretical framework that conceptualizes the relationship between identity and cooperation. In this framework, personal identity is the product of the management of social identities that an individual develops and maintains through efforts to invest in these identities. We use the boundaries of the firm to distinguish between social identities that are linked to the firm and those that are not. We test the theoretical predictions with field experiments in real firms and with their employees, and control for the effect of different degrees of firm affiliation, including studying cooperation between individuals in two firms belonging to the same group. Consistent with theory, we find that firm affiliation has a positive and progressive impact on cooperation in a voluntary contribution game. Moreover, greater social distance between individuals implies less cooperation. Our theoretical and experimental strategy overcomes the many criticisms of social identity approaches. It accounts for the heterogeneity of individual strategies in managing social identities, the impact of institutions on individual behavior, and intra-firm cooperation. Personal identity adds to the traditional study of identification with the firm the issues of individuation and the effect of multiple social identities. We identify the structure of this new approach to the link between cognition and motivation in the firm. We explain how the issues of this extended theory of identity are related to the central issues of the nature of the firm, internal organization, and the boundaries of the firm. We analyze the implications of including the analysis of individuation and multiple social identities on the analysis of cooperation. We also study the effects of learning and corporate culture on identity strategies in the firm.