“Opt out” or kept out? The effect of stigma, structure, selection, and sector on the labor force participation of married women in India

• Date: 01 August 2020
• DOI: http://doi.org/10.1111/rode.12672
• Author: Suddhasil Siddhanta,Nabanita Datta Gupta,Debasish Nandy
• Published date: 01 August 2020

Rev Dev Econ. 2020;24:927–948. wileyonlinelibrary.com/journal/rode

|

927

© 2020 John Wiley & Sons Ltd

Received: 27 December 2018

|

Revised: 31 March 2020

|

Accepted: 10 April 2020

DOI: 10.1111/rode.12672

REGULAR ARTICLE

“Opt out” or kept out? The effect of stigma,

structure, selection, and sector on the labor force

participation of married women in India

NabanitaDatta Gupta1,2

|

DebasishNandy3*†

|

SuddhasilSiddhanta3

*†Deceased September 29, 2019.

1Department of Economics and Business

Economics, Aarhus University, Aarhus V,

Denmark

2IZA Institute of Labor Economics, Bonn,

Germany

3Gokhale Institute of Politics and

Economics, Pune, India

Correspondence

Nabanita Datta Gupta, Department of

Economics and Business Economics,

Aarhus University, Fuglesangs Allé 4, DK-

8210 Aarhus V, Denmark.

Email: ndg@econ.au.dk

Abstract

Using the 61st and 68th rounds of National Sample Survey

data, we investigate the role of stigma, the economy’s struc-

ture, potential selection bias, and sectoral differences in ex-

plaining the low labor force participation (LFP) of middle

and secondary educated women in India relative to low- and

high-educated women (the “U-shape”). Estimating LFP re-

gressions on a sample of prime-aged married women, we

show that, controlling for background characteristics, if a

woman’s husband works in a white-collar job in a region

with a high share of blue-collar jobs, she is less likely to par-

ticipate in the labor market (stigma). We also find a positive

effect of an increase in an index of white-collar job growth

(structure) on married women’s LFP. These effects are

present in both rural and urban sectors but are strongest in

the rural sector. However, middle and secondary educated

women are still found to have substantially lower LFP than

low- and graduate-educated women in both sectors. Indeed,

over time, the U-shape persists in the rural sector and deep-

ens in the urban sector. Because unobservables are quite

large in the urban sector, we use the method of instrumental

variables and find an increasing return at middle levels but

stagnation at higher levels.

KEYWORDS

education, female labor force participation, index of white-collar jobs,

Indian labor market, NSSO data, sector, selection, stigma effect

928

|

DATTA GUPTA eT Al.

1

|

INTRODUCTION

Across the globe, married women’s labor force participation (LFP) has increased dramatically over

the past century, in particular, in East Asian countries such as China, South Korea, and Taiwan. This

increase is generally attributed to the positive impact of the increased access to education on female

labor force participation (FLFP). According to the seminal theories of human capital and household

production of Mincer (1962) and Becker (1973), education, by increasing the opportunity costs of

time and lowering the demand for children, should increase female labor supply. However, in South

Asia, notably India, a strongly negative (Das & Desai, 2003; Kingdon & Unni, 2001) or U-shaped

relationship (Klasen & Pieters, 2015; Neff, Sen, & Kling, 2012) has been found between women’s

education and LFP. LFP is high for women with low levels of education who need to work out of

necessity, is low for women with intermediate-level education, and is high again for women with a

graduate degree. Furthermore, the relationship seems to be persisting over time.

In this paper, we use 61st and 68th rounds of National Sample Survey (NSS) data to investigate

the stagnation of FLFP in India and, in particular, the role of stigma, the economy’s structure,

potential selection bias, and sectoral differences in explaining the persistently low relative LFP of

middle and secondary educated women. We estimate LFP regressions separately by sector on a

sample of prime-aged married women focusing on the role of societal stigma and the economy’s

structure. However, even after accounting for these effects, we find that middle and secondary ed-

ucated women have substantially lower LFP compared to low- and high-educated women. We next

turn to the role of unobservables in FLFP and instrument for education where necessary, as the

level of education is to an extent self-chosen by individuals considering their expected labor market

prospects. Our paper yields many new results on the factors affecting the observed U-shape in FLFP

by education.

One explanation offered for the low LFP at intermediate levels of education is that women’s tra-

ditional role is more valuable to households than their economic role. If women in this group marry

educated, high-income men, the husbands can afford to sustain their nonparticipation and thereby

generate family status. In fact, the status of the husband’s family is even greater if a wife is withdrawn

from the labor market after marriage (Papanek, 1979; Standing, 1991). With growing numbers of

women receiving education and the expected acute shortage in the number of females in the marriage-

able age group in India, we may expect that the demand for women’s traditional role will decrease as

the price of female labor and women’ bargaining power increases. On the contrary, if rich families

continue to practice sex-specific abortions and abort female fetuses and poor families continue having

girls, then the prospects of poor women to “marry up” will increase, meaning less bargaining power

for women than before. Or, men could marry even younger women than before, meaning a decreased

opportunity for women to accumulate education and therefore lower employment (d’Albis & De La

Croix, 2012). This study, thus, points to cultural norms, marriage markets, family status production,

bargaining, and so on, for the persistence of the low LFP of low- and intermediate-educated women in

India (Desai & Jain, 1994; Kapadia, 1995; Kemp, 1986; Papanek, 1979).

A competing explanation offered for the low LFP of lower-educated women originates from the

demand side—exogenous factors such as a scarcity of white-collar jobs, that is, lack of structural

transformation in the economy (Das & Desai, 2003) constrains, in particular, the chances of inter-

mediate-educated women in finding work that they deem appropriate. This is because as education

JEL CLASSIFICATION

J16; J24; O12; O15