The Effects of Differential Income Replacement and Mortality on U.S. Social Security Redistribution

Abstract

We study redistribution via the U.S. Social Security retirement system for cohorts of men born during the second half of the 20th century. Our focus is on redistribution across race and education groups. The cohorts we study are younger than cohorts studied in previous, similar research and thus more exposed to recent increases in earnings inequality. All else equal, this should increase the degree of progressivity of Social Security redistribution due to the structure of the benefit formula. However, we find that redistribution is only modestly progressive for individuals born as late as 1980. Differential mortality rates across race and education groups are the primary explanation. While Black–White mortality gaps have narrowed some in recent years, they remain large and dull progressivity. Mortality gaps by education level are also large and unlike the gaps by race, they are widening, which puts additional regressive pressure on Social Security redistribution.

Transparency, Reproducibility, and the Credibility of Economics Research

“…BITSS has established an active training program for the next generation of economists and other social scientists, as well as an award to recognize emerging leaders in this area, the Leamer-Rosenthal Prize for Open Social Science. 46 Other specialized organizations have also emerged in economics: the Replication Network aims promote the publication of replication studies, Project Teaching Integrity in Empirical Research (TIER) has developed a curriculum to teach computational reproducibility to economics undergraduates, and the Meta-Analysis of Economics Research Network has developed guidelines for meta-analysis (Stanley et al 2013). Similar organizations play analogous roles in other disciplines, including the COS, which is most active within psychology (although it spans other fields), and the EGAP group.…”

“…Bartling et al, AER 2012 (35) Charness and Dufwenberg, AER 2011 (36) Chen and Chen, AER 2011 (37) de Clippel et al, AER 2014 (38) Duffy and Puzzello, AER 2014 (39) Dulleck et al, AER 2011 (40) Ericson and Fuster, QJE 2011 (41) Fehr et al, AER 2013 (42) Friedman and Oprea, AER 2012 (43) Fudenberg et al, AER 2012 44Huck et al, AER 2011 (45) lfcher and Zarghamee, AER 2011 (46) Kessler and Roth, AER 2012 47Kirchler et al, AER 2012 48Kogan et al, AER 2011 (49) Kuziemko et al, QJE 2014 50British Medical Journal 340: C950. Longhi, Simonetta, Peter Nijkamp, and Jacques Poot.…”

Imputing Top‐Coded Income Data in Longitudinal Surveys*

“…The approach builds on the non‐parametric empirical Bayes framework used in Gu and Koenker (2017), and is effectively a non‐parametric random effect method in my application. This Gu and Koenker (2017) framework has been shown to be quite effective in various applications to project income over time when unobserved income heterogeneity in levels and volatility is important (Gu and Koenker, 2016, 2017, ; Tan and Koedel, 2019).…”

Essays on Pension Policy

“…Thus, they do not consider any forecasting exercises or conduct accounting exercises with actual data on contributions and benefits. Tan and Koedel (2019) use data from the Survey of Income and Program Participation (SIPP) to predict Social Security contributions and benefits for cohorts of men born in the second half of the 20th century, and calculate IRRs for race-education groups. They find that OASI is only modestly progressive and that differential mortality is the main driver of missing redistribution.…”