The Intergenerational Effects of Recessions

• Published date: 01 December 2023
• Author: Diana Alessandrini,Bharat Diwakar
• Date: 01 December 2023
• DOI: http://doi.org/10.1111/roiw.12619

bs_bs_banner

Review of Income and Wealth

Series 69, Number 4, December 2023

DOI: 10.1111/roiw.12619

THE INTERGENERATIONAL EFFECTS OF RECESSIONS

BY DIANA ALESSANDRINI∗

Department of Economics, St. FrancisXavier University

AND

BHARAT DIWAKAR

Department of Humanities and Social Sciences, Indian Institute of Technology

Several studies analyzethe contemporaneous effects of recessions on individuals and their children. This

paper shows that recessions also affect futureoffspring, not born yet. By linking fathers and offspring

from the Panel Study of Income Dynamics, we nd that a percentage-point increase in the national

unemployment rate experienced bymen between ages 16 and 20 reduces their future offspring’s annual

wages by 4 percent, occupationalprestige by 3 percent, and education by 0.2 years. Wethen investigate

the mechanisms explaining our ndings.

JEL Codes: 62, E32

Keywords:business cycles, education, intergenerationaleffects, occupational prestige,permanent income

1. INTRODUCTION

A growing body of literature documents the negative and long-lived impact

of recessions on labor market outcomes. Entering the labor market during a reces-

sion may reduce employment and wages (e.g Gangl 2006;Kahn2010; Oreopou-

los and Salvanes 2011; Schwandt and Von Wachter 2019), and may increase the

chances of accepting a job for which one is overeducated (e.g. Summereld and

Theodossiou 2017). Downturns can also accelerate job polarization(Jaimovich and

Siu 2012), which in turn could negatively affectlabor market opportunities for

workers in routine jobs.

An important question that is still unanswered in the literature is the extent to

which these effects impact future generations. This paper shows that the effects of

recessions experienced by individuals extend to their future offspring not born yet,

suggesting that the impacts of downturns can last longer than expected. We exploit

the intergenerational design of the American Panel Study of Income Dynamics

(PSID, 1968– 2019) to recover information on parents and offspringregarding edu-

cation, occupationalprestige, wages, and health. To proxyfor business cycles, we use

Note: We thank the referees and seminar participants at Acadia University for their insightful

comments and suggestions.

*Correspondence to: Diana Alessandrini, Department of Economics,St. Francis Xavier University,

Antigonish, Nova Scotia B2G 2W5, Canada (dalessan@stfx.ca).

© 2022 International Association forResearch in Income and Wealth.

1060

Review of Income and Wealth, Series 69, Number 4, December 2023

three macroeconomic indicators thathave been consistently estimated over time for

two generations: the national unemployment rate, de-trended national GDP, and

de-trended state personal income.

We nd that a percentage-point increase in the national unemployment rate

experienced by men between ages 16 and 20 reduces their future offspring’s annual

wages by 4 percent and occupational prestigeby 3 percent. Offspring are also more

likely to fall belowthe poverty line in adulthood. Moreover,we nd a modest reduc-

tion in offspring’s educational attainment and health. These effects are in addition

to any impacts caused by recessions experienced bymothers or directly experienced

by the offspring after they are born. Using de-trended national GDP or de-trended

state personal income leads to similar conclusions. We also nd that labor mar-

ket conditions between ages 21 and 25 do not generatesignicant intergenerational

effects.

Several mechanisms could explain our results. First, recessions can have

long-lasting effects on workers’ occupation, employment status, and permanent

income. In turn, these effects could inuence outcomes of future offspring by

affecting family resources, the family’s social network, and children’speers. Second,

the effects may depend on howrecessions alter family dynamics (e.g. timing of fam-

ily formation and female labor force participation), with consequences for future

children. Third, recessions may limit individuals’ ability to move to geographic

areas with better economic opportunities, which could affect offspring’s outcomes

later on. Given the richness of PSID data, we are able to directly investigate these

mechanisms. We nd that the rst two hypotheses can explain most of the effects

found in this paper.

As previously recognized in the literature, endogenous migration could bias

estimates of recession impacts.As individuals can choose their residence, local eco-

nomic conditions may be endogenous.For this reason, when using state-level busi-

ness cycles, we adopt an instrumental variable (IV) approach used in the literature

on the contemporaneous effects of recessions.1We instrument business conditions

in the state where the fatherlived at ages 16– 25 with business conditions in the state

where the father spent most of his childhood and early teenage years (ages 1– 14).

Wend that the estimated coefcients from IV are very similar to the ordinary least

squares (OLS) coefcients.

We also address the possibility that intergenerational estimates may be

affected by changes in fertility decisions caused by adverse economic conditions.

If high-ability individuals choose to have fewer children during downturns, then

recessions would lead to an increase in low-ability children in our sample. This in

turn could explain the negative effects that we nd on education and labor market

outcomes. We provide evidence showing that our results are not driven by changes

in fertility decisions during recessions.

We contribute to the literature in a number of ways. We study and quantify

to what extent recessions have repercussions for future generations. Studying this

topic is important because downturns may generate large effects on future gen-

erations. Recessions happening before conception can signicantly alter parental

1See e.g. Kawaguchi and Kondo (2020), Summereld and Theodossiou (2017), Maclean

et al. (2016), Kondo (2015), and Kahn (2010), among others.

© 2022 International Association forResearch in Income and Wealth.

1061