Effects of future pension benefits on pre‐retirement labor supply: Evidence from Chile

• Published date: 01 February 2023
• Author: Oscar Becerra
• Date: 01 February 2023
• DOI: http://doi.org/10.1111/rode.12950

REGULAR ARTICLE

Effects of future pension benefits on

pre-retirement labor supply: Evidence

from Chile

Oscar Becerra

Department of Economics, Universidad

de los Andes, Bogot

a, Colombia

Correspondence

Oscar Becerra, Department of Economics,

Universidad de los Andes, Calle 19A No

1-37 Este, W-824, Bogot

a, Colombia.

Email: obecerra@uniandes.edu.co

Abstract

In this paper, I estimate the effect of future pension

benefits on pre-retirement labor supply for a represen-

tative sample of Chilean workers. Using nonlinear

patterns in pension benefit formulas and a reform that

permanently changed non-contributory pensions, I

estimate the effect of pension accrual and expected

pension wealth on labor force and contributory-sector

participation, labor earnings, and hours worked. I find

that the effect is concentrated on the impact of pension

accrual on the probability to contribute to the pension

system. The effect is heterogeneous and is concentrated

among middle-aged workers, low-skilled workers, and

workers with higher financial literacy.

KEYWORDS

financial literacy, labor supply, pension benefits

JEL CLASSIFICATION

H21, J26, J32

1|INTRODUCTION

Do people take into account their future pension benefits when making their labor supply

decisions? The structure of the pension system has effects on the composition of the labor force. On

one hand, pension contributions are a special kind of mandated benefit, in which workers must con-

tribute in the present to be entitled to a benefit in the future. If workers do not perceive the linkage

of current contributions to future pension benefits, pension contributions are perceived as a dis-

tortionary tax on labor (Auerbach & Kotlikoff, 1987). On the other hand, although non-contributory

Received: 7 July 2021 Revised: 21 July 2022 Accepted: 11 September 2022

DOI: 10.1111/rode.12950

198 © 2022 John Wiley & Sons Ltd. Rev Dev Econ. 2023;27:198–219.wileyonlinelibrary.com/journal/rode

pensions have a positive impact on poverty reduction among the most vulnerable, they may also

reduce incentives to work and to contribute (Piggot et al., 2009). Workers’responses along those

margins are especially relevant for developing economies, where the large size of non-contributory

(unregulated) labor markets may amplify those effects (Levy, 2008).

The empirical literature on labor supply responses to future pension benefits is extensive

and focuses on the financial incentives to retire for workers near retirement age (Krueger &

Meyer, 2002; Krueger & Pischke, 1992). The literature studies the effects of financial incentives

affecting the decision to contribute at the margin and incentives affecting a person's lifetime

income on the probability of retirement. The marginal incentives to contribute are typically

measured by the pension accrual (i.e., the expected marginal gain from contributing to the

pension system),

1

while the incentives on the lifetime income are measured by the expected

pension wealth (i.e., the present value of pension entitlements minus contributions). Most of

these studies related to pre-retirement labor supply focus on workers who are near retirement

age, typically over 50 years, since they face less uncertainty and have a better understanding of

the pension system rules (Liebman et al., 2009; Liebman & Luttmer, 2012).

In this paper, I estimate the effects of future pension benefits on individuals' labor market

choices before retirement. I use exogenous variation from nonlinear patterns in the Chilean pension

system and an unanticipated pension reform undertaken in 2008 to estimate the effects of future

pension benefits along four margins of labor supply: labor force participation, contributory-sector

participation, labor earnings, and hours worked per week. I estimate these effects using the Longitu-

dinal Social Protection Survey (LSPS), a representative sample of Chilean workers between 2002 and

2015 (Ministerio de Trabajo y Protecci

on Social, 2015). Since pension accrual and expected

pension wealth are not observable, I use age patterns of contributions and earnings to predict

pension incentives for each individual in the sample. Then, I run regressions of labor supply

outcomes on the estimated accrual and expected pension wealth, where I control for observable

determinants of labor supply and unobservable heterogeneity using the person's fixed effects.

I find a significant relationship between future pension benefits and pre-retirement labor

supply. The effect is concentrated on the probability that a worker contributes to the pension

system. Consistent with forward-looking behavior, my results show that pension accrual has a

positive and significant effect on contributory-sector participation. On average, a one-standard

deviation increase in pension accrual increases the probability of contributing to the pension

system by 0.025 on a basis of 0.71. Similarly, the effect of expected pension wealth on

contributory-sector participation is positive and significant, although relatively small. On aver-

age, a one-standard deviation increase in the expected pension wealth increases the probability

of contributing to the pension system by 0.01 on a basis of 0.71. Although the estimated effect

of expected pension wealth is consistent with forward-looking behavior, its sign is the opposite

to what is expected under the standard life-cycle model setting. The aggregated results do not

show responses of labor supply to pension system variables along the labor force participation,

labor earnings, and hours worked margins. The estimated effect is heterogeneous across

workers. The heterogeneity analysis suggests that the effects of pension accrual and expected

pension wealth on contributory-sector labor supply are concentrated on men, workers between

40 and 59 years, and less-educated workers.

This paper contributes to the literature by studying the effects of the financial incentives of

the pension system on labor supply. By studying an economy with a sizable non-contributory

sector (about 30% of the labor force), the paper provides empirical evidence of the financial

incentives to avoid paying pension taxes by switching from contributory to non-contributory

jobs, a relevant dimension for developing economies. Previous literature has found evidence

BECERRA 199