Periodic Reporting for period 4 - BAYNET (Bayesian Networks and Non-Rational Expectations)
Okres sprawozdawczy: 2021-01-01 do 2022-06-30
The main objective of this project is to make causal reasoning an integral part of economic theory. The basic idea is that when people make decisions, they establish a causal mapping from actions to consequences, which is based on an intuitive subjective causal model. For example, when a parent decides how much to invest in his child's education, he may have in mind an intuitive causal model that can be represented by the following graphical diagram: investment in child's education --> child's performance at school --> child's subsequent performance in the labor market. The decision maker then uses this subjective causal model to make sense of statistical regularities in his environment and form beliefs that guide his actions. In the above example, this means that the parent will measure the observed correlation between investment and school performance as well as the observed correlation between school performance and subsequent labor-market performance, and he will combine these correlations in accordance with his causal model. If the decision maker's causal model is incorrect - e.g. when he reverses the true direction of causality between two variables, or when he omits certain confounding variables from his model - then his beliefs may exhibit systematic biases. This in turn can lead to interesting behavioral effects. In the above example, the parent's subjective causal model neglects the possibility that both school and labor-market performance may be affected by unobserved exogenous characteristics of the child. This error may lead to over-investment in the child's education.
In its exploration of behavioral implications of erroneous causal reasoning, this project is related to the behavioral economics and bounded rationality literatures. In particular, it adds a novel dimension to the literature on "non-rational expectations". While behavioral economics has studied a variety of biases in belief formation, it has not yet dealt with the problem of erroneous causal reasoning. However, causal reasoning is fundamental to decision making, especially in modern societies. We are constantly bombarded with observational data and correlations between variables, and whether we are able to draw correct causal conclusions from this type of data is an important question. The motto "correlation does not imply causation" is so well-known precisely because people do tend to draw wrong causal inferences from mere correlations. These errors could make them vulnerable to exploitation by third parties.
From a technical point of view, the project draws on the rich literature in statistics and artificial intelligence (AI) on so-called "Bayesian networks". This literature offers tools that enable us to formalize types of errors of causal reasoning and analyze their behavioral consequences. Thus, if successful, the project may also contribute to the influence of graphical probabilistic models in economics.
In its exploration of behavioral implications of erroneous causal reasoning, this project is related to the behavioral economics and bounded rationality literatures. In particular, it adds a novel dimension to the literature on "non-rational expectations". While behavioral economics has studied a variety of biases in belief formation, it has not yet dealt with the problem of erroneous causal reasoning. However, causal reasoning is fundamental to decision making, especially in modern societies. We are constantly bombarded with observational data and correlations between variables, and whether we are able to draw correct causal conclusions from this type of data is an important question. The motto "correlation does not imply causation" is so well-known precisely because people do tend to draw wrong causal inferences from mere correlations. These errors could make them vulnerable to exploitation by third parties.
From a technical point of view, the project draws on the rich literature in statistics and artificial intelligence (AI) on so-called "Bayesian networks". This literature offers tools that enable us to formalize types of errors of causal reasoning and analyze their behavioral consequences. Thus, if successful, the project may also contribute to the influence of graphical probabilistic models in economics.
The paper that formed the basis for the original grant proposal (eventually published as Spiegler (QJE 2016)) presented a model of decision makers having a misspecified subjective causal model. The model has several non-trivial properties (mainly the possibility that individual decision making exhibits "equilibrium" effects that we normally associate with interactive models), which the formalism of Bayesian networks helps illuminating.
Spiegler (JEEA 2020) and Eliaz et al. (AERI 2021) used Bayesian-network tools to study qualitative and quantitative features of the belief distortions that arise from misspecified causal models: (1) which wrong causal models lead to systematically biased predictions of individual variables? (2) how large is the distortion of estimated pairwise correlations that a Gaussian causal model can generate? The papers explored the economic significance of these results. For example, Spiegler (JEEA 2020) included a "monetary theory" application, in which a central bank uses monetary policy to get long-run effects on the real economy, by inducing systematic underestimation of inflation by the private sector. Spiegler (EER 2022) further developed this theme.
Spiegler (Restud 2017) proposed an alternative interpretation of the beliefs that result from the above behavioral model. According to this alternative model, the decision maker does not have a prior causal model. Instead, he has access to partial data regarding the probability distribution that characterizes his environment. He uses a behaviorally motivated procedure to extrapolate a probabilistic belief from the partial data. Under certain conditions on the type of data the decision maker receives, the outcome of this belief-extrapolation procedure looks as if the decision maker fits a subjective causal model to the true distribution. Spiegler (JET 2021) used this alternative formulation to develop a novel framework to model strategic interactions in which players have limited "data access". Thysen et al. (JET 2021) apply it to a model of persuasion, in which strategic use of data for interpreting messages is part of the speaker's strategy.
Eliaz and Spiegler (AER 2020) employed the Bayesian-networks formalism to initiate a new research agenda: modeling the role of narratives in economic and political environments. The paper developed a model of competing political narratives, based on the idea that political positions become successful if they are backed by a "hopeful narrative".
Schumacher and Thysen (TE 2022) applied the Bayesian-network formalism to contract theory. A principal offers a contract to an agent who has an incorrect model of the causal effect of his effort on outcomes. The question is whether the principal can exploit this causal misperception and whether this leads to novel features of the optimal contract.
The Annual Review of Economics solicited a review article from the PI (eventually published in 2020) about this modeling approach. Other instances of successful dissemination include an inter-disciplinary workshop on "the psychology and economics of causal reasoning" was that took place at UCL in 2019, and multiple invitations to present project outputs in keynote lectures and winter/summer schools.
Spiegler (JEEA 2020) and Eliaz et al. (AERI 2021) used Bayesian-network tools to study qualitative and quantitative features of the belief distortions that arise from misspecified causal models: (1) which wrong causal models lead to systematically biased predictions of individual variables? (2) how large is the distortion of estimated pairwise correlations that a Gaussian causal model can generate? The papers explored the economic significance of these results. For example, Spiegler (JEEA 2020) included a "monetary theory" application, in which a central bank uses monetary policy to get long-run effects on the real economy, by inducing systematic underestimation of inflation by the private sector. Spiegler (EER 2022) further developed this theme.
Spiegler (Restud 2017) proposed an alternative interpretation of the beliefs that result from the above behavioral model. According to this alternative model, the decision maker does not have a prior causal model. Instead, he has access to partial data regarding the probability distribution that characterizes his environment. He uses a behaviorally motivated procedure to extrapolate a probabilistic belief from the partial data. Under certain conditions on the type of data the decision maker receives, the outcome of this belief-extrapolation procedure looks as if the decision maker fits a subjective causal model to the true distribution. Spiegler (JET 2021) used this alternative formulation to develop a novel framework to model strategic interactions in which players have limited "data access". Thysen et al. (JET 2021) apply it to a model of persuasion, in which strategic use of data for interpreting messages is part of the speaker's strategy.
Eliaz and Spiegler (AER 2020) employed the Bayesian-networks formalism to initiate a new research agenda: modeling the role of narratives in economic and political environments. The paper developed a model of competing political narratives, based on the idea that political positions become successful if they are backed by a "hopeful narrative".
Schumacher and Thysen (TE 2022) applied the Bayesian-network formalism to contract theory. A principal offers a contract to an agent who has an incorrect model of the causal effect of his effort on outcomes. The question is whether the principal can exploit this causal misperception and whether this leads to novel features of the optimal contract.
The Annual Review of Economics solicited a review article from the PI (eventually published in 2020) about this modeling approach. Other instances of successful dissemination include an inter-disciplinary workshop on "the psychology and economics of causal reasoning" was that took place at UCL in 2019, and multiple invitations to present project outputs in keynote lectures and winter/summer schools.
The project's academic contributions can be classified as follows:
1. Creating modeling frameworks that provide a language and some basic results that will guide theorists who want to analyze the implications of causal misperceptions in various economic settings
2. Providing analytical results about the situations in which wrong causal models lead to wrong beliefs, and some results about the magnitude of such belief errors
3. Applications to various fields of economics (monetary theory, contract theory, strategic communication)
4. Initiating a new long-term research agenda concerning the role of causal narratives in economics and politics
All these are novel developments that went significantly beyond the state of the art in equilibrium modeling with non-rational expectations. They started an entirely new area of study, namely the role of imperfect causal reasoning in decision making. They introduced the language of DAGs into economic theory, and demonstrated the use of some basic tools from the Bayesian-network literature.
1. Creating modeling frameworks that provide a language and some basic results that will guide theorists who want to analyze the implications of causal misperceptions in various economic settings
2. Providing analytical results about the situations in which wrong causal models lead to wrong beliefs, and some results about the magnitude of such belief errors
3. Applications to various fields of economics (monetary theory, contract theory, strategic communication)
4. Initiating a new long-term research agenda concerning the role of causal narratives in economics and politics
All these are novel developments that went significantly beyond the state of the art in equilibrium modeling with non-rational expectations. They started an entirely new area of study, namely the role of imperfect causal reasoning in decision making. They introduced the language of DAGs into economic theory, and demonstrated the use of some basic tools from the Bayesian-network literature.