Investigating human and LLM psychology by prompting LLMs to play experimental economics games: Xie, Mei, Yuan, and Jackson in PNAS

 The great science fiction writer of my youth was Isaac Asimov, who not only wrote space opera (The Foundation Trilogy), but also wrote about intelligent robots, i.e. about robots with artificial general intelligence.  So, like you and me, they had complicated psychological lives, and one of the main characters in these stories was the robopsychologist  Dr. Susan Calvin (see e.g. the short story collection I, Robot, and also several of the robot novels).

I'm reminded of this by the several papers now reporting how large language models respond when asked to play games that have been used to study human behavior.  Those papers are framed as using LLMs to learn about the human behavior on which they were trained. But they can also be read as telling us about the 'psychology' of LLMs. Here's a good one from the PNAS. 

Xie, Yutong, Qiaozhu Mei, Walter Yuan, and Matthew O. Jackson. "Using large language models to categorize strategic situations and decipher motivations behind human behaviors." Proceedings of the National Academy of Sciences 122, no. 35 (2025): e2512075122. 

Abstract: By varying prompts to a large language model, we can elicit the full range of human behaviors in a variety of different scenarios in classic economic games. By analyzing which prompts elicit which behaviors, we can categorize and compare different strategic situations, which can also help provide insight into what different economic scenarios might induce people to think about. We discuss how this provides a step toward a nonstandard method of inferring (deciphering) the motivations behind the human behaviors. We also show how this deciphering process can be used to categorize differences in the behavioral tendencies of different populations. 

 

Recent good looking market design papers I hope to read (on auctions, unraveling, and interviews)

There was a time when I could reasonably hope to have read market design papers before they appeared in print, but now there are many fine papers that I'll never have a chance to read. (I'm sure that's just because the field is growing so much...)  I haven't given up, however...   Here are three that recently caught my eye.

Hu, Edwin, and Dermot Murphy. "Vestigial tails? Floor brokers at the close in modern electronic markets." Management Science (2025).

 Abstract: The closing auction is an increasingly important trade mechanism due to the rise of passive funds that require closing price execution. We study differences in auction mechanism design on NYSE and Nasdaq that may affect closing price efficiency. Unlike Nasdaq, NYSE allows late auction orders through its floor brokers, providing traders with more flexibility to mitigate or create large last-minute auction imbalances. Price changes in the closing auction are more likely to reverse on NYSE compared with Nasdaq, suggesting greater price inefficiency in NYSE closing auctions. Larger last-minute abnormal imbalances on NYSE, particularly in stocks where auction competition may be inhibited by relatively high floor broker auction fees, explain these stronger reversals. Evidence from the NYSE floor closure during the COVID-19 pandemic supports a causal interpretation. Our results highlight an important tradeoff between auction flexibility and price efficiency.

######

Stable Matching with Interviews, by Itai Ashlagi, Jiale Chen, Mohammad Roghani, Amin Saberi

    Part of: Volume: 16th Innovations in Theoretical Computer Science Conference (ITCS 2025)
    Series: Leibniz International Proceedings in Informatics (LIPIcs)
    Conference: Innovations in Theoretical Computer Science Conference (ITCS)  

Abstract: "In several two-sided markets, including labor and dating, agents typically have limited information about their preferences prior to mutual interactions. This issue can result in matching frictions, as arising in the labor market for medical residencies, where high application rates are followed by a large number of interviews. Yet, the extensive literature on two-sided matching primarily focuses on models where agents know their preferences, leaving the interactions necessary for preference discovery largely overlooked. This paper studies this problem using an algorithmic approach, extending Gale-Shapley’s deferred acceptance to this context.
Two algorithms are proposed. The first is an adaptive algorithm that expands upon Gale-Shapley’s deferred acceptance by incorporating interviews between applicants and positions. Similar to deferred acceptance, one side sequentially proposes to the other. However, the order of proposals is carefully chosen to ensure an interim stable matching is found. Furthermore, with high probability, the number of interviews conducted by each applicant or position is limited to O(log² n).
In many seasonal markets, interactions occur more simultaneously, consisting of an initial interview phase followed by a clearing stage. We present a non-adaptive algorithm for generating a single stage set of in tiered random markets. The algorithm finds an interim stable matching in such markets while assigning no more than O(log³ n) interviews to each applicant or position."

############
Deadlines and matching, by Garth Baughman, Journal of Economic Theory, Volume 228, September 2025
https://doi.org/10.1016/j.jet.2025.106065

Abstract: Deadlines and fixed end dates are pervasive in matching markets. Deadlines drive fundamental non-stationarity and complexity in behavior, generating significant departures from the steady-state equilibria usually studied in the search and matching literature. I consider a two-sided matching market with search frictions where vertically differentiated agents attempt to form bilateral matches before a deadline. I give novel proofs of existence and uniqueness of equilibria, and show that all equilibria exhibit an “anticipation effect” where less attractive agents become increasingly choosy over time, preferring to wait for the opportunity to match with attractive agents who, in turn, become less selective as the deadline approaches. When agents are patient, a sharp characterization is available: at any point in time, the market segments into a first class of matching agents and a second class of waiting agents. This points to a different interpretation of unraveling.

The Walras-Bowley Lecture: Fragmentation of Matching Markets and How Economics Can Help Integrate Them, by Kamada, Kojima, and Matsushita

 Fuhito Kojima's 2023 Walras Bowley Lecture has just been uploaded to arxiv, and it looks to be an exciting contribution to the market design literature.  It considers the fact that administrative boundaries cause many markets to be fragmented, and hence less thick than they might otherwise be, and how some of the resulting efficiency loss can be recovered.

The Walras-Bowley Lecture: Fragmentation of Matching Markets and How Economics Can Help Integrate Them, by Yuichiro Kamada, Fuhito Kojima, and Akira Matsushita
(August 27, 2025) 

Below is the abstract and opening paragraphs. 

"Abstract
Fragmentation of matching markets is a ubiquitous problem across countries and across applications. In order to study the implications of fragmentation and possibilities for integration, we first document and discuss a variety of fragmentation cases in practice such as school choice, medical residency matching, and so forth. Using the real-life dataset of daycare matching markets in Japan, we then empirically evaluate the impact of interregional transfer of students by estimating student utility functions under a variety of specifications and then using them for counterfactual simulation. Our simulation compares a fully integrated market and a partially integrated one with a “balancedness” constraint—for each region, the inflow of students from the other regions must be equal to the outflow to the other areas. We find that partial integration achieves 39.2 to 59.6% of the increase in the child welfare that can be attained under full integration, which is equivalent to a 3.3 to 4.9% reduction of travel time. The percentage decrease in the unmatch rate is 40.0 to 52.8% under partial integration compared to the case of full integration. The results suggest that even in environments where full integration is not a realistic option, partial integration, i.e., integration that respects the balancedness constraint, has a potential to recover a nontrivial portion of the loss from fragmentation.

Introduction
Many of the most consequential markets in our societies—school admissions, medical resident matching, daycare placements, kidney exchanges—are matching markets, where centralized mechanisms are often employed to improve efficiency and fairness. Over the past several decades, the field of market design has made substantial progress in developing and implementing such mechanisms (e.g., Abdulkadiroğlu and Sönmez (2003) for school choice, Roth (1984) for medical residency matching, Kamada and Kojima (2023) for daycare placements, and Roth et al. (2004) for kidney exchanges). In doing so, it has led economists to assume a dual role as both analysts and engineers (Roth, 2002).

How are those sophisticated mechanisms implemented in practice? Typically, these mechanisms are run by individual cities, districts or institutions, and the implementation is usually confined to narrowly defined administrative or political boundaries. These boundaries can reflect long-standing institutional arrangements, localized funding responsibilities, or jurisdictional autonomy. Regardless of their origins, the consequence is that agents on different sides of a boundary are matched as if they participated in entirely separate markets—even when they live mere blocks apart.

The aim of this paper is to study the implications of fragmentation of matching markets and possibilities for integration. To do so, we begin by offering a detailed descriptive account of fragmented matching markets in practice. We observe that fragmentation is prevalent across a variety of settings globally, from public school systems and childcare allocation to medical residency assignments, foster care placements, and public housing markets, among others. We highlight how institutional boundaries and localized governance create fragmented, parallel markets. Each case underscores the potential inefficiencies due to constrained choices caused by fragmentation, motivating our inquiry into mechanisms that can integrate markets effectively.

Against this background, we then investigate public daycare assignment in Japan in detail. "

Thinking about the ethical, legal and political relationships between IVF and abortion (in JAMA)

Some people support both IVF and abortion (women's right to choose) and some people oppose both (embryos are people), but many oppose abortion but support IVF.  Here's an article that focuses on some ethical distinctions (is the intention to have a child or not), and also on some political ones (IVF patients are on average more affluent than abortion patients).

Watson K. Rethinking the Ethical and Legal Relationship Between IVF and Abortion. JAMA. Published online May 22, 2025. doi:10.1001/jama.2025.6733 

"US voters have elected a president who promised he would make the government or private insurance cover in vitro fertilization (IVF), yet takes credit for reversing Roe v Wade. These positions highlight a question that has lingered since US IVF practice began in 1981: are hospital and governmental policies that support IVF but do not support abortion ethically consistent? And if not, why is this division so common?

"Those who see IVF and abortion as ethically distinct often focus on differences in intention and outcome—having a baby vs avoiding having a baby. Others see them as comparable practices because both destroy embryos. I offer a third perspective, which is that abortion and IVF are comparable practices because both are family-building medical interventions; therefore, support for IVF access ought to lead to support for abortion access.

"Abortion was a federal constitutional right until 2022, and IVF was subject to state regulation like the rest of medicine. Yet constitutional protection did not stop many states from heavily regulating abortion, and IVF rarely faced governmental limits. A stark example of their disparate treatment occurred shortly after Roe v Wade was reversed when the traditionally antiabortion state of Indiana began its statute criminalizing abortion provision by clarifying “This article does not apply to in vitro fertilization” (IN Code §16-34-1-0.5 [2024]).

"Yet like abortion, IVF also involves embryo death. 

...

" differences in patient income, race, age, and education also suggest a stark difference in political power between the constituencies invested in IVF and abortion. Eighty-one percent of fertility patients have household incomes of more than $100 000 and 75% are White6; 72% of abortion patients have incomes less than 200% of the federal poverty line and 59% are Black or Latinx.3 Sixty-four percent of IVF patients are 35 years or older,7 while 70% of abortion patients are in their teens or 20s.3 Fourteen percent of births to women with a college or graduate degree were conceived with the use of assisted reproductive technology, but only 1.5% of births to women with some college or less were conceived with assisted reproductive technology in 2023.8 In contrast, 77% of abortion patients have some college or less.9 (There are no data documenting how many IVF patients deferred their childbearing by having an abortion when they were younger, or how many IVF patients later abort to decrease a multiple pregnancy or avoid unexpected medical problems.)

"Entities involved in providing IVF also have financial interests in preserving the legality of a practice with a median cost of $19 200 per cycle.6 Infertility care generates approximately $8 billion per year in gross revenues in the US.10 High operating margins have drawn private equity investors to many private fertility practices, and IVF is lucrative for hospitals and physicians."

Will artificial intelligence disrupt labor markets as much as electricity and computers have?

 Here's a paper that takes a long view of American occupations (and concludes that it's too early to tell about ai...)

TECHNOLOGICAL DISRUPTION IN THE LABOR MARKET by David J. Deming, Christopher Ong, and Lawrence H. Summers, NBER Working Paper 33323 , January 2025, http://www.nber.org/papers/w33323 

ABSTRACT: This paper explores past episodes of technological disruption in the US labor market, with the goal of learning lessons about the likely future impact of artificial intelligence (AI). We measure changes in the structure of the US labor market going back over a century. We find, perhaps surprisingly, that the pace of change has slowed over time. The years spanning 1990 to 2017 were less disruptive than any prior period we measure, going back to 1880. This comparative decline is not because the job market is stable today but rather because past changes were so profound. General-purpose technologies (GPTs) like steam power and electricity dramatically disrupted the twentieth-century labor market, but the changes took place over decades. We argue that AI could be a GPT on the scale of prior disruptive innovations, which means it is likely too early to assess its full impacts. Nonetheless, we present four indications that the pace of labor market change has accelerated recently, possibly due to technological change. First, the labor market is no longer polarizing-- employment in low- and middle-paid occupations has declined, while highly paid employment has  grown. Second, employment growth has stalled in low-paid service jobs. Third, the share of  employment in STEM jobs has increased by more than 50 percent since 2010, fueled by growth in software and computer-related occupations. Fourth, retail sales employment has declined by 25 percent in the last decade, likely because of technological improvements in online retail. The postpandemic labor market is changing very rapidly, and a key  question is whether this faster pace of change will persist into the future.