The creation and analysis of intergenerationally linked U.S. tax data has revolutionized the study of income inequality and mobility (Chetty et al. 2014b; 2014a; 2018). Building on that scientific breakthrough, this project will create, analyze, and distribute intergenerationally linked tax record-based proxy measures of wealth, rather than income, for the entire United States. Studying wealth (i.e., net worth) in addition to income is important as the two are far from perfectly correlated; wealth is much more unequally distributed, and wealth plays an independent role in the intergenerational transmission of advantage (Killewald et al. 2017; Pfeffer and Killewald 2018; Pfeffer and Schoeni 2016). However, the currently available U.S. data infrastructure does not support even basic descriptions of geographic variation in the level of wealth holdings, the degree of wealth inequality and segregation, and, even less so, the extent of wealth mobility across generations.
Based on an approved collaborative project within the Internal Revenue Service (IRS) Statistics of Income Joint Statistical Research Program, we are in a unique position to alleviate these data limitations. Through direct access to individual-level and full population IRS tax records, we will create proxy measures of the wealth holdings of all U.S. taxpayer units and link them across generations, extending pioneering work by Chetty and collaborators as well as Saez and Zucman (2016). We will analyze the level, inequality, segregation, and intergenerational mobility of wealth across all U.S. taxpayers, allowing us to answer questions such as: How does the overall distribution of wealth vary across U.S. areas and within neighborhoods? How high is the intergenerational correlation in wealth? Which U.S. areas show higher and lower intergenerational wealth mobility?
We will distribute geographic aggregates of these measures to the scientific research community and broader public. This new public data infrastructure will support a wide set of novel analyses of wealth inequality and mobility and their policy determinants, which may draw on naturally occurring experiments, policy changes, and other discontinuities at the local, regional, or state level. The public data release follows strict IRS data disclosure protocols to ensure data privacy and, in particular, to prevent the identification of individual tax records. All publicly released estimates are geographic aggregates and are subjected to a formal disclosure analysis before the IRS clears them for release. Furthermore, all project members at the University of Michigan, including those without direct access to individual-level IRS records, are required to sign the ?Policy on Safeguarding Respondent Confidentiality? that regulates data privacy at the Institute for Social Research (ISR), a research organization with nearly 75 years of experience in producing and protecting some of country?s core social scientific data.
Our project puts a particular focus on the user-friendly distribution and presentation of these estimates to make them easily accessible to a wide audience, including to local, state, and federal policymakers, community organizations, journalists, and the broader public. To accomplish this aim, we will collaborate with a data visualization vendor to build an interactive online platform that visualizes and documents our main measures, namely distributional estimates (wealth means, medians, gini coefficients, top shares), estimates of wealth segregation, and estimates of intergenerational mobility at different levels of geography for the entire United States. We will be able to distribute these estimates at the same geographic level currently included in Raj Chetty?s ?Opportunity Atlas? and potentially at an even more disaggregated level within the constraints set by the data disclosure regulations set by IRS.
Funding:
Bill and Melinda Gates Foundation
Funding Period:
07/12/2021 to 08/31/2025