The Independent Research Fund Denmark (DFF) International Postdoc at University of Toronto

Matching and Welfare Analysis of the Danish Mortgage Market, with Ronald Wolthoff and John Kennes

Press Coverage:

Buying a home is the most important financial decision that most people will make in their life. The primary sources of finance for households when buying a home are banks and mortgage credit institutions (MCIs). The types of mortgage products offered by these institutions deeply affect individuals’ lifetime well-being. This project investigates how the structure of, and changes in, the mortgage market affect both individual and overall social welfare.

While it is a matter of course that households ultimately choose their institutions and mortgage products, it is also important to recognize that the institutions themselves have incentives to choose customers with specific characteristics. These incentives are sometimes influenced by regulatory concerns. For example, the Danish FSA introduced in 2016 restrictions on the number of loans a bank can serve to customers for which the Loan to Income (LTI) ratio exceeds a factor of 4 in the Copenhagen metropolitan area and Aarhus. This is an example of a macroprudential policy which has become a focus point for policy makers since the financial crisis in 2007. To assess the welfare impact of new and potential developments in the mortgage market, it is important to consider both sides of the market and how they interact. In this project, we

1) determine what drives the terms of trade, i.e. pricing and other details of a mortgage contract, and the assignments between any particular individual and institution, and

2) determine how individuals and institutions will behave in response to structural and regulatory changes in the mortgage market.

We do this by extending state-of-the-art theoretical and empirical methodologies and applying it to unique data covering the population of Danish individuals and the universe of credit institutions and mortgage transactions.

This project will help to improve the quality of the Danish mortgage market by identifying direct and indirect trade-offs associated with policy proposals aimed at managing systemic risks. For example, policy makers would be interested in a counterfactual policy experiment that removes a particular institution (e.g. Bank A goes bankrupt) or a type of mortgage product (e.g. interest-only loans), to determine how the portfolio of loans for the remaining institutions will be impacted, and which customer types will experience the greatest losses/gains in the mortgage market.

Theoretical foundation: We will analyse this pattern of banking relationships and the allocation of gains within these relationships using a two-sided revealed preference framework. Our framework assumes individuals and institutions can compensate each other using transfer payments, which the literature refers to as transferable utility (TU). Moreover, participants on both sides of the market operate competitively and can freely choose trading partners subject to finding agreeable terms of trade. Our theoretical foundation is the “two-side stable matching model with TU” framework which is well known in the market design literature, pioneered by the Nobel Prize winners Alvin Roth and Lloyd Shapley, also known as the Becker model referring to Becker (1973). To estimate this Becker model, we will extend recent developments in econometric methods presented in Fox 2007, 2010, 2018, Akkus et al., 2016, and Fox, Yang & Hsu, 2018.

Teacher and Principal Labor Markets and School Performance, with Arnaud Dupuy, John Kennes and Peter Rohde Skov

Measuring and improving the quality of school leadership and teaching is of fundamental importance in education. This project uses a two-sided multidimensional matching model to provide a full cost-benefit analysis of public policies that impact the assignments of principals/teachers to schools/classrooms. In particular, we will (1) estimate the potential amenities of teaching/managing "easy-to-serve classrooms/schools" for each type of teacher/principal in terms of their wages; (2) estimate how much inequality exists across schools with regard to the quality of teachers and principals; (3) evaluate the impact of counterfactual policy proposals, such as a change in the wage cap of teachers for one group of schools, on the assignments of more effective teachers to disadvantaged schools; (4) obtain new estimates on value-added measures for teachers and principals that take into account the selection of teachers and principals into schools, and (5) compare estimates on the selection and performance of teachers within schools with estimates from other studies. We extend state-of-the-art theoretical and empirical methodologies and apply it to the new and unique matched teacher-to-classroom and principal-to-school data from Denmark. A country that has decentralized school systems and school performance can be measured through national test scores. Our project will develop operational tools for the policymakers to improve the quality and governance of public schools. We aim to provide proof of practice for the general application of our methods to other education systems. Our project is multidisciplinary, involving a combination of education, pedagogy, and developing of quantitative methods. It includes the transfer of knowledge to both academic and public sector, and the training of researcher in new advanced econometric methods. This study contributes to the EU strategy to develop school education systems and strengthen European identity through high-quality education.