Does the Welfare State Destroy the Family? Evidence from OECD Member Countries

• Date: 01 April 2016
• Author: Mario Lackner,Martin Halla,Johann Scharler
• Published date: 01 April 2016
• DOI: http://doi.org/10.1111/sjoe.12144

Scand. J. of Economics 118(2), 292–323, 2016

DOI: 10.1111/sjoe.12144

Does the Welfare State Destroy the Family?

Evidence from OECD Member Countries∗

Martin Halla

University of Innsbruck, A-6020 Innsbruck, Austria

martin.halla@uibk.ac.at

Mario Lackner

University of Linz, A-4040 Linz, Austria

mario.lackner@jku.at

Johann Scharler

University of Innsbruck, A-6020 Innsbruck, Austria

johann.scharler@uibk.ac.at

Abstract

We study the effect of the size of the welfare state on demographic trends in OECD member

countries. Exploiting exogenous variation in public social spending, due to varying degrees

of political fractionalization (i.e., the number of relevant parties involved in the legislative

process), we show that an expansion in the welfare state increases fertility, marriage, and

divorce rates with a quantitatively stronger effect on the marriage rate. We conclude that the

welfare state supports family formation in the aggregate. Further, we find that the welfare

state decouples marriage and fertility, and therefore alters the organization of the average

family.

Keywords: Divorce; fertility; marriage; risk sharing; welfare state

JEL classification:D1; D62; H31; H53; J12; J13; J18

I. Introduction

Family and kinship traditionally provided services such as care for the

young and the elderly, and provided insurance against unforeseen events

such as illness and unemployment. Nowadays, governments in industrial-

ized countries provide, or at least subsidize, these services, for instance,

through public health and unemployment insurance. Thus, the role of the

family might have become less important in this respect (Anderberg, 2007),

and the incentives to form a family might have decreased along with the

∗For helpful comments, we would like to thank (among others) two anonymous referees.

The usual disclaimer applies. This research was funded by the Austrian Science Fund (FWF)

National Research Network S103, the Austrian Center for Labor Economics, and the Analysis

of the Welfare State.

CThe editors of The Scandinavian Journal of Economics 2015.

M. Halla, M. Lackner, and J. Scharler 293

introduction of comprehensive welfare state arrangements. Nevertheless, the

implementation of welfare state arrangements can incorporate (implicit or

explicit) subsidies of certain family arrangements. If, for instance, welfare

state regulations promote marriage and fertility, then the welfare state can

also exert a crowding-in effect on families. Consequently, the overall effect

is less clear.

Does the welfare state destroy or support the family? We study this

issue by examining OECD member countries in the period from 1980 until

2007. Thus, in contrast to the existing body of literature, which examines

specific welfare arrangements and reforms, we study the impact of welfare

state arrangements on an aggregated level. To measure the extent of the

welfare state, we mainly use public social spending as a percentage of the

gross domestic product (GDP), but we also use more specific measurements

such as public social spending on the family. Thus, we evaluate the effect

of the average implementation of the welfare state in the sample of OECD

member countries in the given time period. Demographic trends of family

outcomes are captured by marriage, divorce, and fertility rates.

To obtain exogenous variation in the size of the welfare state, we turn to

the literature on the political economy of public spending. This body of lit-

erature stresses the importance of political fractionalization for the level of

public spending.1We exploit varying degrees of political fractionalization

– in particular, the number of relevant parties involved in the legislative

process – as an instrumental variable. Our first-stage estimations show a

highly significant effect of within-country variation in political fraction-

alization on public (social) spending. The identifying assumption of our

instrumental variable strategy is that political fractionalization in the par-

liament affects family behavior only through the channel of public social

spending. While this assumption is not testable, there are few determinants

of family behavior that are reasonably correlated with political fractional-

ization. We discuss a number of potential mediating links (such as govern-

ment’s ideology or polarization, immigration, and taxation of families) and

show that our estimated effects are very robust to the inclusion of these

potential confounders. We hope that this discussion convinces the criti-

cal reader that using political fractionalization as an instrument for public

social spending is a useful identification strategy in our context.

1As we discuss in detail, the link between fractionalization and public spending can be

derived from different political economy theories of public finance. While the so-called

common pool theory holds that highly fractionalized systems are generally prone to increase

spending, and thus lead to an expansion of the welfare state, Primo and Snyder (2008)

argue that fractionalization can lead to lower spending. Intuitively, the benefits of smaller

and cheaper projects can be easier to internalize by specific groups, leading to lower overall

spending. For empirical evidence on the role of fractionalization, see, for example, Volkerink

and de Haan (2001).

CThe editors of The Scandinavian Journal of Economics 2015.

294 Does the welfare state destroy the family?

Our results indicate that a larger welfare state increases the turnover in

the marriage market by increasing both marriage and divorce rates. Because

the effect on marriage is stronger than that on divorce, an increase in the

size of the welfare state increases the stock of married individuals. Further,

we observe an increase in the fertility rate, which is particularly pronounced

for non-marital fertility (as compared to marital fertility). Hence, while

the welfare state supports the formation of families, it crowds out the

traditional organization of the family by increasing the divorce rate and

the number of children born out of wedlock. All the estimated effects are

highly statistically significant and their quantitative importance increases

when we use a narrower measurement of the welfare state.

Three guideposts can be used to put this analysis in the context of the

existing body of literature. First, we add to the literature that studies the

consequences of the welfare state in a general sense (Castles et al., 2010).

Second, we add to the literature on demographic trends (Stevenson and

Wolfers, 2007b).2With the exception of the out-of-wedlock ratio, compa-

rably little attention has been paid to the influence of the welfare state

on demographic outcomes.3In the public debate, the dominant view is

that the welfare state has to adjust to changing demographic patterns (as,

for instance, in the case of an aging society with extensive pensions sys-

tems). Our finding points to the reversed link, where the welfare state has

the capacity to shape demographic outcomes. Finally, we provide empiri-

cal support for Gary S. Becker’s claim that the organization of the family

changes as the state supplements or replaces traditional family functions.

In this regard, our empirical evidence complements the large body of liter-

ature examining the effects of specific US welfare arrangements on family

outcomes at less aggregated levels.4

Typically, scholars use variations in welfare benefit levels across time

and states to identify the effects on family outcomes. By and large, micro

analyses confirm theoretical expectations. Moffitt (1997) concludes that

many existing studies find a negative effect of public transfers on marriage,

2Economic scholars have studied the role of the economic independence of women (Isen

and Stevenson, 2011), access to abortion and oral contraception (Akerlof et al., 1996; Goldin

and Katz, 2000, 2002), changes in divorce law (Rasul, 2006; Wolfers, 2006; Matouschek

and Rasul, 2008), and reforms of custody law (Halla, 2013).

3A notable exception is Bj¨

orklund (2006) who identifies a positive effect of the Swedish

family policy on the overall fertility level.

4Earlier papers evaluate the effect of Aid to Families with Dependent Children (AFDC) and

more recent papers look at Temporary Assistance for Needy Families (TANF). In the case

of AFDC, where the disposable income of recipients increases with children and decreases

with marriage, one expects a decrease in marriage, an increase in divorce, and an increase

in fertility, especially out of wedlock (Becker, 1993). TANF replaced AFDC (in 1996) and

reversed some of the incentives of welfare arrangements. In fact, the major stated goals of

this legislation included reducing out-of-wedlock births and increasing marriage.

CThe editors of The Scandinavian Journal of Economics 2015.