The Nakahara prize was established in 1995 and is funded by a donation from Mr. Nobuyuki Nakahara. The purpose of the prize is to honor and encourage young researchers under the age of 45 who have published internationally recognized research.

Professor Kitagawa is a leading figure in theoretical econometrics. He has made several creative and fundamental contributions to various fields of econometrics. He has published seminal papers in the relevant literature, and they influence subsequent research including those by the top econometricians. His research on treatment allocation problems is a momentous contribution.

Kitagawa and Tetenov (2018) develop and theoretically justify the Empirical Welfare Maximization (EWM) method. In the last decades, we have seen a significant development of causal inference methods in econometrics and their applications in various areas of economics. Arguably, we have now achieved a good understand of how to estimate treatment effects. A natural question, that comes next and would be policy relevant, is how one can decide whether the treatment should be implemented or to whom it should be assigned. EWM allocates the treatment by maximizing an estimated welfare. In traditional plug-in approaches, the treatment effect is first estimated and then the treatment is assigned to maximize the welfare. Instead, EWM directly targets the welfare. This approach has several advantages in terms of both theory and applications. Theoretically, Kitagawa and Tetenov (2018) make connections to rate optimality discussed in the statistical and machine learning literature. In terms of application, it can handle policy constraints in simple manners. Such applicability is appealing compared to another theoretically sophisticated approach based on minimax regret. EWM has caught attention from many researchers, as evidenced by the fact that Econometrica has already published two papers (Athey & Wager, 2021; Mbakop & Tabord-Meehan, 2021) that extend EWM. Prof. Kitagawa himself extends EWM to other contexts in, for example, Kitagawa and Tetenov (2021) and Kitagawa and Wang (2023).

Giacomini and Kitagawa (2021) propose a new Bayesian method in settings with partial identification. Partial identification means that multiple parameter values are consistent with the distribution of observables. It has been one of the most popular topics in econometrics in the last decades. In such settings, the frequentist approach and the Bayesian approach do not coincide even asymptotically, although they do in the standard case of point identification. The proposal by Giacomini and Kitagawa (2021) solves this discord. Their approach is to use multiple (possibly infinitely many) priors for unidentified parameters. This paper contributes to the foundations of statistics and provides useful statistical tools particularly for macroeconomic applications.

Professor Kitagawa also makes other contributions. Kitagawa (2015) develops a test for instrument validity in the causal inference framework. Andrews et al. (2021) establish valid inference methods after structural break estimation using the tool in the selective inference literature.

Selection Committee:

Koichiro Ito, University of Chicago.

Hideshi Itoh, Waseda University.

Takashi Kamihigashi, Kobe University.

Yuichi Kitamura, Yale University.

Scott Kominers, Harvard Business School.

Ryo Okui, University of Tokyo (Chair).

Takashi Ui, Hitotsubashi University.