Minisymposium

A theory-informed test of reasoning in artificial intelligence (AI) combines three sequential steps to consider correct answers as the result of a reasoning process as opposed to luck of probabilistic word matching. The first step is information filtering, where an AI model that reasons must distinguish the relevant information in a prompt from trivia. In the second step, knowledge association, the AI combines implicit or explicit knowledge with the relevant prompt information. And finally in the third step of logic attribution, a reasoning AI assigns correct logic operations for deducive, inducive, and other types of logic to uncover the corret answer. In economic settings, the logic steps involve different levels of counterfactual considerations and policy-relevant thought experiments. This paper leverages insights from the large language model benchmarking literature and the social economics literature to inform the design of benchmarking tests that are challenging, robust, evolving over time and informative about any type of reasoning shortcomings. The benchmarking process can be adapted to other sciences. An accompanying training dataset is available to help AI developers improve reasoninig in their models, and interested users can submit proposals for material to create questions.

We propose new global solution algorithms for continuous time heterogeneous agent economies with aggregate shocks.We first approximate the state space so the master equation becomes a high, but finite, dimensional partial differential equation. We then approximate the value function using neural networks and solve the master equation using deep learning tools. The main advantage of this technique is that it allows us to find global solutions to high dimensional, non-linear problems. We consider two broad approaches to reducing the dimensionality of the problem: discretizing the number of agents and projecting the distribution. We demonstrate our algorithms by solving two canonical models in the macroeconomics literature: the Aiyagari (1994) model and the Krusell and Smith (1998) model.

The green energy transition is necessary within the next few decades to mitigate climate change.In the paper, I explore the effectiveness of carbon pricing and the role of technological spillovers inachieving decarbonization, with a particular focus on the challenges faced by developing countries.I develop a two-region integrated assessment model that incorporates fossil fuel and renewableenergy sources to investigate the quantitative impact of spillovers on decarbonization in developingcountries. In solving the model, I rely on the deep equilibrium nets as a global solution method. The findings indicate that technological spillovers in developing countries contribute tothe replacement of fossil fuels with renewable energy inputs. The study suggests that implementingcarbon taxation in both advanced and developing regions, along with technological spillovers, yieldsthe most favorable outcomes for the climate. However, the absence of carbon tax in developingcountries with spillovers still delivers slightly better environmental results compared to taxing bothregions without spillovers. The results emphasize the importance of considering spillovers andcarbon taxation when designing effective policies to achieve environmental goals.

This paper quantifies the general equilibrium effects of financial innovation that increases access to equity markets. I study an overlapping generations model with both idiosyncratic and aggregate risk, solved with machine learning techniques. A benchmark economy with limited stock market participation and rebalancing frictions matches the current dynamics of macro aggregates, equity and bond returns, as well as wealth and portfolio concentration. A counterfactual experiment shows how widespread adoption of target date funds would improve risk sharing, reduce inequality, and generate substantial welfare gains for households in the bottom 90% of wealth distribution. The equity premium drops from 6.3% to 2.5%, while the standard deviation of equity returns stabilizes from 24.7% to 15.2%. Full adoption of target date funds would generate around 20% average welfare gains for people in the bottom 90% at the expense of the top 10% who lose by more than 50% through the redistribution of financial wealth. Asset pricing and welfare outcomes are very close between an economy with target date funds and one without any participation costs or rebalancing frictions.