丁香园AV

Dongwoo Kim

Semi-nonparametric modelling of multidimensional matching

This research project aims at improving our understanding of how people find employment and life partners. The project focuses on developing new mathematical models and statistical techniques to analyze how workers and firms, as well as potential spouses, match with each other. This knowledge is crucial for developing effective labor market policies and understanding income inequality.

Economists have been dealing with this problem using many simplifying assumptions. For example, Gary Becker (1992 Nobel Prize winner) of University of Chicago proposed a mathematical model of spousal matching, assuming matching between males and females hinges on their 鈥渁bility scores鈥� which aggregate numerous characteristics like age, income, education level, religion, race, and appearance to a single number.  However, this assumption is unrealistic, as it would deem two individuals with vastly different attributes (like appearance and income) equally desirable if they have the same ability score. This project aims to develop a theoretical model that does not rely on such unrealistic assumptions, while keeping it easy to apply to real-world data.

Additionally, the project will investigate patterns in labor and marriage markets, including the impact of technological progress and changing preferences on income inequality. Our preliminary findings from the U.S. labour market data suggest that wage polarization (phenomenon featuring stronger wage growth in the bottom and upper tails of the wage distribution relative to the median) has been mainly driven by technological progress favouring cognitive abilities over manual skills. As we navigate the era of artificial intelligence, will this trend accelerate? We seeks to answer this question using our model.

The findings of this project will help policymakers create more effective employment programs and inform strategies to address income inequality. Moreover, it will provide insights into the dynamics of labor and marriage markets, shedding light on how people form partnerships and find employment. Overall, this research has the potential to make a significant impact on our understanding of these important issues.

Alexander Karaivanov

Digital Finance Platforms and Mechanism Design - Theory, Data and Applications

Advances in computing and data processing technologies can have a transformative effect on economic activity, financial markets, government policies and safety nets. Digital platforms, including blockchain-based, can store and process vast amounts of real-time and historical granular information about individuals or firms, with cryptographic protection for sensitive data. Smart contracts and dApps can automatically execute code mapping individual or aggregate input data (information, reports, messages) into outputs (payments, transfers, loans), quickly perform complex computations, and embed multidimensional contingencies. Digital property rights, including escrow and collateral, can be secured and enforced without relying on trusted third parties or imperfect institutions.

My research proposal addresses several objectives. First, I provide formal analysis of the algorithmic tools offered by digital technologies focusing on their strengths and limitations vis-a-vis the building blocks of contract design: property rights, (asymmetric) information and commitment/trust. Second, I propose and describe digital platform applications for improved risk-sharing. Examples under development include Thai rural households, credit constrained firms, and cross-country sharing of unemployment or macroeconomic risks. Third, implementing and scaling up digital platforms requires making them robust to unobserved heterogeneity, private information and default. This ties back to mechanism-design theory, as these challenges must be addressed respecting incentive compatibility and voluntary participation.

Krishna Pendakur 

Linear Models for Scale Economies and Resource Shares

My proposed research project aims to create a toolkit to identify and estimate scale economies and resource shares. Scale economies describe the relative needs of people living in different types of households and are used to make redistribution (such as welfare benefits) horizontally equitable across household types. Resource shares, defined as the fractions of consumption accruing to each member of a household, are also essential to the measurement of inequality and poverty.

The objective is to provide models that can be estimated by transparent econometric methods with off-the-shelf household expenditure data of the type already collected by central banks and statistical agencies in more than 100 countries around the world, including Canada.

Pendakur鈥檚 models will imply linear reduced forms that are easy to estimate, in contrast to the standard nonlinear structural models that are traditionally used. His models will also allow for a better picture of within-household inequality.  

This research project will:

1. provide a class of structural models that imply linear reduced form models whose parameters are identified and easily estimated;
2. provide alternative identifying assumptions that neither require that we observe the demands of all people as singles, nor require that all people have demands that are the same regardless of their household type.
3. provide models that are straightforward to estimate and permit averaging of unobserved heterogeneity into reduced form parameters;
4. and, take into account the restrictions implied by marriage markets.

Anke Kessler

Understanding Female Earnings During and After Marriage - Is There a Divorce Premium?鈥� 

Women earn less than men on average, and it well-documented that the gap in pay increases further when women get married (鈥渕arriage penalty鈥�) or have children (鈥渕otherhood penalty鈥�). We know less about the impact on women鈥檚 earning of other significant life events such as divorce, however. This limitation is important given the possibility that divorce can further exacerbate gender differences, especially among vulnerable groups of women.

Our research will explore how marital dissolution affects the income of women by developing a novel theoretical framework that captures essential elements of household separation, and by examining new evidence from Canada and Italy. One possible effect we want to study in particular is that some women will work and earn more following a divorce, implying a  "divorce premium鈥� that may be partially reversing the existing marriage penalty for women with high-earnings potential.

The research will contribute to our understanding of the determinant of divorce as well as the trajectory of (female) earnings, labour supply, and related outcomes, following household dissolution. It will also advance what we know about what factors shape women's labour supply over the life-cycle, gender-identity and relative earnings, and the division of labour of households pre- and post- separation. Increasing our knowledge base in these areas is not only of interest for economists, but also relevant for a wide audience in social sciences, policy makers, and the general public.