Husband’s Unemployment and Wife’s Labor Supply: The Added Worker Effect across Europe

Abstract

This article investigates the responsiveness of women’s labor supply to their husband’s job loss—the so-called added worker effect. The authors contribute to the literature by taking an explicit internationally comparative perspective in analyzing the variation of the added worker effect across welfare regimes. Using longitudinal data from the European Union Statistics on Income and Living Conditions (EU-SILC) survey covering 28 European countries from 2004 to 2013, they find evidence of an added worker effect, which, however, varies over both the business cycle and the different welfare regimes in Europe. The latter result might be explained, in part, by differences in the design of the unemployment benefit system across countries, which create different incentives for the labor supply of wives of unemployed men.

Keywords

added worker effect labor supply unemployment cross-country analysis

Theoretical models of family labor supply predict that the unemployment of one spouse should increase the labor supply of the other spouse. To offset the expected income loss associated with the partner’s job loss, inactive spouses newly enter the labor market and become so-called added workers and already-participating spouses increase the amount of hours they work. We investigate these theoretical predictions by focusing on the responsiveness of women’s labor supply to their husband’s job loss. Previous empirical literature on this topic mainly concentrates on a single country and provides mixed results. Cross-country evidence on the added worker effect is scarce.

Despite this lack of cross-country evidence, it seems safe to assume that wives’ response to their husband’s job loss varies across welfare regimes. Even within the European framework, countries differ largely with respect to their institutional settings, their social policies, and the structure of their labor markets; therefore, they offer different incentives for women to adjust their labor supply. As Bentolila and Ichino (2008) argued, the role of family support and thus wives’ reactions to their husband’s job loss should be stronger whenever the welfare state fails to mitigate the consequences of unemployment. In this regard, Reher (1998) showed a dividing line between southern European societies, with their history of depending on strong family networks, and northern European societies, with their weaker family systems and greater reliance on extended welfare states. Following this argument, we expect the behavioral response of wives to their husband’s unemployment to be stronger the lower the generosity of the welfare system.

To test this hypothesis, we take an explicit internationally comparative perspective and analyze whether the added worker effect varies across the welfare regimes in Europe. In doing so, we use longitudinal data from the European Union Statistics on Income and Living Conditions (EU-SILC) covering 28 European countries over the period 2004 to 2013. Observing households over the time of the Great Recession, which forced many families to devise strategies to cope with negative income shocks due to job loss, provides a fresh opportunity to investigate couples’ labor supply. Whereas previous evidence on the added worker effect during recessions is based on before–after comparisons for single countries, we are able to investigate the role of the added worker effect in Europe’s economic crisis by explicitly analyzing its variation in relation to the countries’ economic conditions.

In addition, we contribute to the literature by considering a variety of behavioral responses of wives to their husband’s job loss, covering reactions at both the extensive and the intensive margins of women’s labor supply. Previous literature has mainly concentrated on analyzing the labor market entry of nonparticipating wives and has mostly ignored the labor supply adjustments of already-participating wives. Given that female labor force participation rates have increased remarkably over the past decades and that the countries within Europe vary largely with respect to the structure of their labor markets, addressing this issue in an internationally comparative perspective is of particular importance.

Theoretical Framework and Literature

The theory underlying the notion of spousal labor supply as insurance against unemployment has been developed in, among others, Ashenfelter (1980), Heckman and MaCurdy (1980), and Lundberg (1985). According to the life-cycle model of family labor supply (e.g., Stephens 2002), the household jointly maximizes utility, which depends on the leisure of the husband and the wife and total household consumption over its lifetime. In this framework, the husband’s unemployment will affect the wife’s labor supply, through both the cross-wage and the income effect. Whereas the impact of the cross-wage effect is ambiguous, depending on whether the couple’s leisure times are substitutes, complements, or independent of each other, the income effect will clearly increase the wife’s labor supply. Note that an increase in the wife’s labor supply is not only induced by the immediate reduction in household income caused by the husband’s unemployment but also is brought about by a reduction in expected lifetime income due to the decline in the husband’s future wage offers.

As outlined by Stephens (2002), the wife’s labor supply response will depend on whether the husband’s job loss is anticipated by the household and on the magnitude of the resulting income loss. If the unemployment of the husband is unanticipated by the household, the permanent earnings loss associated with the husband’s unemployment will permanently increase the wife’s desired labor supply. If, however, the husband’s unemployment is anticipated by the household, for example, because certain characteristics of the husband make him very likely to become unemployed, it will not induce an added worker effect because the expected income loss will already have translated into respective adjustments in household consumption and/or labor supply. In addition, the wife’s response will depend on the magnitude of the expected loss in lifetime income, being strongest in families that experience the largest permanent income reductions.

Despite the theoretical well-known effects, the existing empirical literature analyzing the added worker effect misses a clear consensus on its magnitude or even its existence. Most of the early empirical literature focused on the labor supply of nonparticipating women in the United States. In that scenario, the added worker effect was usually found to be small or nonexistent (Lundberg 1985; Maloney 1987, 1991; Spletzer 1997). Those studies that did uncover an added worker effect usually concluded that the small responses were optimal because the husband’s unemployment only led to a transitory reduction in earnings, which were considered to be small in a life-cycle framework (Heckman and MaCurdy 1980).

Furthermore, Spletzer (1997) argued that the added worker effect is expected to be less present during times of economic prosperity. In these economically prosperous phases, the absence of liquidity constraints may enforce other opportunities of smoothing family income, for example, couples are more able to rely on credit or savings to maintain their consumption (Sullivan 2008). Moreover, when employment rates are high, job losses as well as the associated expected income losses are likely to be transitory. It is therefore not surprising that previous literature concluded that the added worker effect tends to be more prevalent in periods of economic downturns (Parker and Skoufias 2004; Mattingly and Smith 2010; Bryan and Longhi 2013).

Another factor lowering the magnitude of the added worker effect is the unemployment benefit system. For the United States, Cullen and Gruber (2000) found that the added worker effect was partly crowded out by unemployment benefits and that the labor supply response of women whose husbands became unemployed was 30% larger in the absence of these benefits. Ortigueira and Siassi (2013) reached a similar conclusion and further showed that the crowding-out effect of unemployment insurance was stronger among liquidity-constrained households.

Some more-considerable effects were found by studies that focused on the intensive margin of wife’s labor supply. For the United States, Stephens (2002) found that women whose husbands had been displaced significantly increased their paid working time. Similar effects were found by Kohara (2010) for Japan and by Gong (2011) for Australia.

Cross-country evidence on the existence of the added worker effect, however, is still scarce. Exceptions are McGinnity (2002), which compared Britain and West Germany, and Prieto-Rodriguez and Rodriguez-Gutierrez (2003), which analyzed the added worker effect for 11 European countries; both studies focused on the extensive margin of women’s labor supply responses. McGinnity (2002) found evidence of an added worker effect in West Germany, but identified no effect for Britain. An explanation for the non-presence of an added worker effect in Britain is given by the country’s unemployment benefit system, which is largely based on means-tested benefits and therefore sets disincentives for women to enter the labor market after their husbands become unemployed. Prieto-Rodriguez and Rodriguez-Gutierrez (2003) revealed that the added worker effect was present in only a few countries in the European Union, including Italy and, to a lesser extent, Germany, the Netherlands, Portugal, and Spain.¹

While the existing cross-country studies limit their analysis to wives’ entries into the labor market because of their husband’s unemployment, we assume that the wives’ behavioral response varies across countries. Whereas the female labor force participation rate is relatively low in most Mediterranean countries, it is higher in most Western societies. In 2014, the average female labor force participation rate for the EU-28 was 66.5%. It was lowest in Malta (52.1%), Italy (54.4%), and Greece (59.0%) and highest in Sweden (83.4%), Norway (75.9%), and Denmark (75.0%) (Eurostat 2015). Therefore, it is not surprising that most of the empirical literature that identifies an added worker effect deals with countries in which the labor force attachment of women is comparatively low (Prieto-Rodriguez and Rodriguez-Gutierrez 2000; Başlevent and Onaran 2003; Bentolila and Ichino 2008; Ayhan 2017). In most Western societies, the ability of married women to newly enter the labor market and become additional workers is limited, because most women already participate in the labor market. In these countries, wives’ reactions to their husband’s job loss are more likely to be observed in terms of an increase in their hours of work.

Although differences in econometric frameworks across studies are not the focus of this article, we share further methodological aspects by which previous studies differ. Whereas early studies of the added worker effect simply analyzed the relationship between husbands’ labor force status and their wife’s labor supply (e.g., Lundberg 1985; Maloney 1987), more recent literature specifically investigates the labor market dynamics around husbands’ unemployment. In addition, studies differ with respect to the type of unemployment considered. Most of the literature does not distinguish between different types of job losses, yet some papers look exclusively at the labor supply responses to unexpected shocks to the household, such as unemployment induced by a husband’s displacement (Stephens 2002) or by an involuntary job loss (Kohara 2010). Last, studies differ regarding the timing of the measurement of wives’ behavioral response to their husband’s job loss. Whereas most studies investigate women’s employment response directly following husbands’ job loss, usually using quarterly or yearly data, Stephens (2002) considered wives’ responses both before and several years after husbands’ job loss. He found that wives responded to their husband’s displacements with small increases in their labor supply prior to displacement and much larger increases once the displacement occurred.

The main conclusion we draw from previous literature is that the existence and the magnitude of the added worker effect are highly dependent on the considered circumstances. Although every prior study provides a valuable hint on which circumstances matter, the literature lacks a comprehensive empirical investigation of the responsiveness of wives’ labor supply to their husband’s unemployment. Our aim, therefore, is to unify the previous literature and reconcile the varying results by providing a large-scale investigation of the added worker effect. Analyzing its variation across welfare regimes and its fluctuation over the business cycle while also considering a variety of behavioral responses of the wife at both the extensive and the intensive margin of labor supply should provide a better understanding of the circumstances that facilitate or hamper spousal labor supply as an insurance device against unemployment shocks.

Empirical Strategy and Data

Econometric Model

To test the added worker hypothesis for the European case, we estimate different probit models of the form

Δ Y_{i t}^{m} = Φ (X_{i t}^{'} β^{m} + γ^{m} Δ E_{i t} + Σ ϕ_{j}^{m} C_{j} + Σ θ_{t}^{m} T_{t} + M_{j t}^{'} α^{m} + ε_{i t}^{m})

which describe women’s behavioral response in household i in year t in country j. The above models mainly differ with respect to their dependent variable as denoted by the superscript m, with m = (1,…,5). First, for m = 1, $Δ Y_{i t}$ indicates a binary variable that equals unity if the wife was out of the labor force (IA) in t–1 and is in the labor force (A) in t, that is, $Δ Y_{i t} = (I A_{t - 1} \to A_{t} | I A_{t - 1})$ . In a second step, we distinguish between two types of labor market activity. For m = 2, the dependent variable equals unity if the wife is unemployed (UE) in t and for m = 3, it equals unity if the wife is employed (E) in t, given that she was out of the labor force in t–1.² In a third step, we acknowledge that the individual’s self-defined economic status captures only the person’s own perception of their main activity and may thus differ from the strict criteria of the International Labor Organization (ILO) concept. Thus, for m = 4, the dependent variable equals unity if the wife was not searching for a job in t–1 and is searching for a job in t ( $Δ J S$ ). Finally, for m = 5, $Δ Y_{i t}$ is set to unity if the wife was employed part-time (PT) in t–1 and is employed full-time (FT) in t.

The vector $X_{i t}$ includes a set of individual and household characteristics as described in more detail below. The vector $C_{j}$ contains a full set of country dummies and the vector $T_{t}$ contains a full set of year dummies.³ $M_{j t}$ is a vector of macroeconomic conditions of the country, which vary over time.

The variable $Δ E_{i t}$ is the variable of main interest and in the following is referred to as the “added worker dummy.” This variable is a binary indicator that equals unity if the wife’s spouse became unemployed from t–1 to t, and 0 if he stayed employed. Its coefficient is expected to be positive and significant in each specification if an added worker effect is present in the particular sample. The magnitude of its marginal effect can be interpreted as the increase in wife’s probability of adjusting her labor supply as a response to her husband’s unemployment.⁴

In identifying a causal added worker effect, it is crucial to discriminate between permanent and transitory factors leading to the husband’s unemployment (Maloney 1991). On the one hand, the unemployment of the husband might proxy for predominantly transitory factors that are unrelated to the personal characteristics of the household, such as the closure of a plant that directly results in the layoff of the husband. On the other hand, the unemployment of the husband might proxy for predominantly permanent characteristics of the household. The husband’s unemployment propensity might be correlated with unobserved characteristics of the household, such as the sorting mechanism that initially formed the household, which matches spouses with similar levels of human capital or similar preferences for leisure. In the latter case, we are likely to underestimate the true added worker effect, since wives of frequently unemployed husbands are likely to face low market wage rates themselves and thus to show similarly low labor supply patterns as their husbands. To identify a causal effect of husband’s unemployment on wife’s labor supply, it is therefore important to disentangle permanent and transitory factors leading to the husband’s unemployment. Although we aim to accomplish this goal by controlling for a variety of individual and household characteristics correlated with husbands’ unemployment probability, and we conduct a series of sensitivity analyses to verify the robustness of our results, we cannot entirely rule out that unobserved heterogeneity still biases our estimation results.⁵ We keep that in mind when interpreting our estimation results.

In addition to our baseline regressions, we further explore the heterogeneity of the added worker effect. First, we seek to identify whether the magnitude of the added worker effect varies with the macroeconomic conditions of a country. To do so, we add an interaction of the added worker dummy and the macroeconomic indicators ( $Δ E_{it} \times M_{jt}$ ) to the model. Second, we separately estimate Equation (1) for several subsamples of countries to test whether the added worker effect differs across the welfare regimes in Europe. For this, we group countries according to a modified Esping-Andersen welfare regime typology (Esping-Andersen 1990).

To ensure representativeness, we use the longitudinal weights delivered with the EU-SILC data in all regressions. These weights not only correct for different selection probabilities of individuals within each country as well as panel attrition but also ensure that each country is represented in proportion to its actual population size.

Data

The data used in this study are taken from the EU-SILC covering the period 2004 to 2013. The EU-SILC data include all European Union member states as well as Norway and Iceland. Because of insufficient data quality, however, Iceland and Malta are excluded from our analysis, which provides a sample of 28 countries. Since we are interested in wives’ labor supply adjustments as a reaction to their husband’s unemployment, we use the longitudinal version of the EU-SILC data. The longitudinal version is a four-year rotating panel, which allows us to follow households and individuals for a maximum of four years.

The data were collected by Eurostat for the first time in 2004. In the first wave, 13 countries were surveyed and most of the other countries—except for Bulgaria (2006), Romania (2007), and Croatia (2010)—followed in 2005. Whereas the majority of countries were surveyed through 2013, some countries either left the survey (Germany in 2006) or had not yet provided any data for 2013 (Croatia, Greece, Romania, and Sweden).⁶

In our analysis, we restrict the sample to married or cohabiting couples in which both individuals are between 16 and 65 years of age and neither is retired or unable to work.⁷ For the analysis of wives’ labor supply responses at the extensive margin, we further restrict our sample to “traditional couples,” that is, we condition on the husband being employed and the wife being out of the labor force in t–1. In analyzing the labor supply adjustments of wives already participating in the labor market, the sample is restricted to couples in which the wife is working part-time and the husband is employed in t–1.

We obtain details on a husband’s and wife’s labor market status from a variable that contains information on the self-defined current economic status of an individual, distinguishing between full-time and part-time employment, unemployment, and different types of inactivity (e.g., schooling, retirement, fulfilling domestic tasks). This variable is used to define labor market transitions of the wife. First, we ignore the type of labor market activity and define a variable that equals 1 if the wife enters the labor market (i.e., if she either becomes employed or unemployed), and 0 otherwise. In a second step, we explicitly distinguish between the two types of labor market activity to discriminate between mechanisms occurring on the supply and the demand side of the labor market. We create two variables that take value 1 if the wife enters into employment and unemployment, respectively, and 0 otherwise. In a third step, we acknowledge that the individual’s self-defined economic status captures only the person’s own perception of their main activity at present. It therefore differs from the strict criteria of the ILO concept, which defines the unemployed as those who are without work, currently available for work, and seeking work. For instance, some people who consider themselves unemployed may not take active steps to find work and to be immediately available. By contrast, some people may actively search for a job, but consider themselves not in the labor force, as, for example, students who have not yet entered the labor market. In our sample, 38% of the women who consider themselves unemployed state that they are not actively searching for a job, whereas the proportion of those who consider themselves inactive but do search for a job is much smaller (5%). Therefore, we further use information on the individual’s job-search behavior with a question that asks respondents whether they have been actively looking for a job within the last four weeks. The respective variable takes value 1 if the wife has not been searching for a job in t–1 but is doing so in t, and value 0 if she has not been searching for a job in either period. Last, we use information on the individual’s self-defined current economic status to define a variable equal to unity if the wife has been working part-time in the period t–1 and is working full-time in the period t. This variable is equal to 0 if the wife continuously remains in part-time employment.

Instead of using information on the current employment status, husband’s labor market transitions are identified by using retrospective information on the husband’s employment history in the past 12 months. In doing so, a husband is considered to be unemployed if he had at least one unemployment spell within the past 12 months. This method means that a husband might be considered as being unemployed even if he is currently employed. The reasoning behind using this criterion to define husband’s unemployment is that we assume that even small or transitory reductions in household income might change the optimal behavior of the household and thus result in individual labor supply responses.⁸

In our regressions, we control for a variety of individual and household characteristics.⁹ At the household level, we control for whether the couple is married, the number of children, and whether the youngest child is aged 0 to 3 years and 4 to 6 years, respectively. To capture the couple’s financial background, we include the logarithm of the household’s equivalized disposable income as a regressor.¹⁰ Moreover, we include a binary variable indicating whether the household currently has to repay some non-housing-related debts, and we control for the dwelling type the couple inhabits, distinguishing between couples living in a detached house, a semi-detached house, and an apartment or a flat.

At the individual level, we include both spouses’ age and its square and control for their highest level of education (grouped per the United Nations’ International Standard Classification of Education [ISCED]), distinguishing between low-skilled (ISCED 0–2), medium-skilled (ISCED 3–4), and high-skilled (ISCED 5) individuals. Furthermore, we control for the husband’s occupational status in t–1 in all models and for the wife’s occupational status in t–1 when considering wives who actively participate in the labor market, that is, when analyzing women’s transitions from part-time to full-time employment. In doing so, we differentiate (using the International Standard Classification of Occupations [ISCO]) between white-collar high-skilled (ISCO 1–3), white-collar low-skilled (ISCO 4–5), blue-collar high-skilled (ISCO 6–7), and blue-collar low-skilled (ISCO 8–9) individuals.¹¹

As outlined earlier in the Econometric Model subsection, it is important to discriminate between permanent and transitory factors leading to the husband’s unemployment. A standard method is to control for the husband’s (and the wife’s) labor market experience. Although the EU-SILC data contain information on the individual’s years in employment, in some countries this information is not surveyed for all household members and is answered by only one person, the “selected respondent.” This approach is followed in all Nordic countries, as well as Ireland, the Netherlands, and Slovenia. As a result, the EU-SILC data do not allow us to control for both partners’ labor market history, and even if only the husband’s years of employment are included, the number of observations for the above-named countries is significantly reduced. We therefore exclude this variable from our basic regression but conduct a sensitivity analysis in which the husband’s labor market experience, as measured by his share of years in employment in all years since entering the labor market, is additionally controlled for. In these regressions, we further control for the husband’s previous job type, that is, whether the job was permanent or temporary. Such information is available only for selected respondents. The weights have been adjusted to account for this new data structure, so that the remaining observations are still representative for the population of each country.

In addition to analyzing the existence and the magnitude of the added worker effect in general, we investigate its variation with the countries’ economic conditions. In contrast to previous literature, we not only compare its magnitude in times of economic up- and downswings but also apply a more flexible approach in interacting the added worker dummy with time-variant macroeconomic indicators, namely the country’s GDP growth rate, its unemployment rate, and its female labor force participation rate. GDP growth and unemployment rates both capture the country’s state of the economy at present and are as such strongly correlated. Nevertheless, it is plausible to consider both factors in a single regression. Whereas the GDP growth rate proxies the country’s economic situation in general, the unemployment rate explicitly captures the current situation of the labor market. As the Great Recession demonstrated, not every downturn of the economy (directly) translates to increasing unemployment rates. If the economy struggles, firms may have other ways to control costs, such as cutting back on investments or resorting to short-time work.¹² It is therefore important to distinguish between the current situation of the economy in general and the conditions of the labor market in particular, and to separately analyze their impact on the existence and the magnitude of the added worker effect.

As outlined above, we additionally estimate our model separately for specific subsamples of countries to test whether the added worker effect differs across the welfare regimes in Europe. The subsamples are chosen according to a modified Esping-Andersen welfare regime typology (Esping-Andersen 1990) suggested by Bonoli (1997). Bonoli’s typology is based on a two-dimensional approach that classifies countries according to the quantity and the quality of welfare provision. Quantity and quality of welfare provisions are measured by social expenditure as a proportion of gross domestic product (GDP) and by contribution-financing as a proportion of social expenditure, respectively.

Following Bonoli’s classification system, we distinguish between four types of welfare states: 1) high quantity/high quality countries: Denmark, Finland, Norway, and Sweden (referred to as Nordic countries); 2) high quantity/low quality countries: Austria, Belgium, Germany, France, Luxembourg, and the Netherlands (referred to as Continental countries); 3) low quantity/high quality countries: Ireland and the United Kingdom (referred to as Anglo-Saxon countries); and 4) low quantity/low quality countries: Greece, Italy, Portugal, and Spain (referred to as Mediterranean countries). Since the countries of Central and Eastern Europe are not covered by Bonoli’s typology, we add a fifth category that includes Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia. Austria, too, was not covered by Bonoli’s original classification. We decided to categorize this country to the Continental countries, according to both its values on the above-named indicators and its geographical position.

Results

Basic Results for the Pooled Sample

The results of the estimation of our basic model (Equation (1)) are shown in Table 1. With respect to our control variables, the results are overall as expected from economic theory. Married women and women with a higher number of children are less likely to increase their labor supply, irrespective of which labor market transition is considered. Women whose youngest child is younger than three years are less likely to enter the labor market, to start searching for a job, and/or to increase their working hours. Women whose youngest child is of preschool age (four to six years) are more likely to enter the labor market, though the presence of preschool children does not affect women’s job search and part-time to full-time transitions.

Table 1.

Pooled Regression of Women’s Labor Market Transitions

Explanatory variables	IA_t-1 → A_t	IA_t-1 → UE_t	IA_t-1 → E_t	Δ JS	PT_t-1 → FT_t
Household characteristics
Married	−0.031***	−0.012***	−0.021***	−0.020***	−0.028***
	(0.007)	(0.004)	(0.007)	(0.006)	(0.007)
Number of children	−0.024***	−0.007***	−0.016***	−0.009***	−0.019***
	(0.003)	(0.001)	(0.002)	(0.002)	(0.003)
Child age 0 to 3	−0.014**	−0.005	−0.009	−0.024***	−0.017**
	(0.007)	(0.003)	(0.006)	(0.004)	(0.008)
Child age 4 to 6	0.037***	0.010***	0.027***	0.004	−0.010
	(0.007)	(0.004)	(0.006)	(0.004)	(0.007)
Log. equiv. disposable income (in thsd.)	0.011***	−0.005***	0.017***	−0.007***	0.001
Log. equiv. disposable income (in thsd.)	(0.004)	(0.002)	(0.004)	(0.002)	(0.005)
Repayment of debts	0.031***	0.009***	0.021***	0.020***	0.007
	(0.005)	(0.002)	(0.004)	(0.003)	(0.005)
Dwelling type (ref.: detached house)
Semi-detached house	−0.010*	0.002	−0.010*	0.005	0.004
	(0.006)	(0.003)	(0.005)	(0.004)	(0.006)
Apartment or flat	−0.010*	0.006***	−0.016***	0.013***	0.014**
	(0.005)	(0.002)	(0.004)	(0.003)	(0.006)
Wife’s characteristics
Age	−0.004***	−0.001***	−0.003***	−0.003***	0.000
	(0.001)	(0.000)	(0.000)	(0.000)	(0.001)
Education (ref.: medium skilled)
Low skilled	−0.036***	−0.005**	−0.031***	−0.009***	−0.001
	(0.005)	(0.003)	(0.004)	(0.003)	(0.006)
High skilled	0.079***	0.003	0.073***	0.026***	0.026***
	(0.007)	(0.003)	(0.007)	(0.006)	(0.006)
Occupation (ref.: white-collar high)
White-collar low	—	—	—	—	−0.041***
					(0.006)
Blue-collar high	—	—	—	—	0.054***
					(0.011)
Blue-collar low	—	—	—	—	−0.036***
					(0.008)
Husband’s characteristics
Age	−0.003***	−0.001*	−0.002***	0.000	−0.005***
	(0.001)	(0.000)	(0.000)	(0.000)	(0.001)
Education (ref.: medium skilled)
Low skilled	−0.016***	−0.004	−0.014***	−0.004	−0.008
	(0.005)	(0.003)	(0.005)	(0.004)	(0.006)
High skilled	−0.014**	0.001	−0.014***	−0.006	0.000
	(0.006)	(0.003)	(0.006)	(0.004)	(0.006)
Occupation (ref.: white-collar high)
White-collar low	0.001	0.004	−0.001	0.008	0.027***
	(0.007)	(0.004)	(0.006)	(0.005)	(0.007)
Blue-collar high	0.010*	0.005*	0.006	0.009**	−0.020***
	(0.006)	(0.003)	(0.005)	(0.004)	(0.006)
Blue-collar low	0.000	0.005	−0.003	0.012***	0.012*
	(0.006)	(0.003)	(0.006)	(0.004)	(0.007)
Country characteristics
GDP growth rate	0.003***	0.000	0.003***	0.000	0.005***
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
Unemployment rate	0.004***	0.000	−0.001	0.002**	0.000
	(0.001)	(0.001)	(0.001)	(0.001)	(0.002)
Female LFP rate	−0.010***	−0.007***	−0.005**	−0.006***	−0.004
	(0.003)	(0.001)	(0.002)	(0.002)	(0.003)
Added worker dummy	0.036***	0.030***	−0.001	0.048***	0.060***
	(0.009)	(0.005)	(0.008)	(0.007)	(0.014)
Δ% due to AWE	19.69%	72.50%	—	70.46%	35.54%
Pseudo-R²	0.101	0.098	0.112	0.089	0.095
Observations	87,416	87,416	87,416	76,133	73,891

Source: EU-SILC; authors’ calculations.

Notes: Results represent average marginal effects, calculated as average effects over all individuals in the respective sample, and robust standard errors (clustered at the household level) obtained from probit estimations of women’s labor market transitions on a set of individual, household, and country characteristics. IA_t-1 → A_t refers to women’s transitions from inactivity to activity; IA_t-1 → UE_t refers to women’s transitions from inactivity to unemployment; IA_t-1 → E_t refers to women’s transitions from inactivity to employment; ΔJS refers to women’s job-search transitions; PT_t-1 → FT_t refers to women’s transitions from part-time to full-time employment. The added worker dummy takes value 1 if the husband becomes unemployed from t–1 to t and 0 if he stays employed. Δ% refers to the percentage change in women’s probability of adjusting their labor supply due to their husband’s unemployment. Percentage changes are shown for significant added worker effects (5% level) only. Both country and year fixed effects are additionally included in the regressions. LFP, labor force participation.

Asterisks denote statistical significance * at the .10 level; ** at the .05 level; *** at the .01 level.

The household’s disposable income has a diverse effect on women’s labor supply: Whereas household income is positively correlated with women’s employment transitions, it is negatively correlated with their unemployment and job-search transitions. This result is likely driven by unobserved heterogeneity, in a sense that unobserved characteristics exist, such as the couples’ preferences for leisure or their productivity in the labor market and in household production, that are correlated with both household income and wife’s attachment to the labor market. A similar diverse effect is found for the dwelling type the couple inhabits. In couples who live in an apartment or flat, wives are less likely to enter employment and more likely to become unemployed and to start looking for a job than are wives in couples who live in a detached house. In households that have to repay non-housing-related debts, women are significantly more likely to enter the labor market or to start searching for a job, though the repayment of debts is uncorrelated with wives’ changes from part-time to full-time work. This result is in line with the theoretical argument that labor supply adjustments are more common among households that are financially constrained.

Women’s probability of entering the labor market decreases with their age and increases with their level of education. Moreover, women working in low-skilled blue-collar or white-collar jobs are less likely, and women working in high-skilled blue-collar positions are more likely, to change from part-time to full-time employment than are high-skilled white-collar workers. This result might be explained by the fact that as compared to high-skilled jobs, low-skilled jobs offer less flexibility in terms of enabling women to increase their working hours in the short term. The age and the education of the husbands are correlated only with women’s transitions into employment and are uncorrelated with their unemployment or job-search transitions. Overall, women are less likely to make labor market transitions the higher their husband’s occupational status, suggesting that women are more likely to stay out of the labor market the higher their husband’s earning potential.

The country’s GDP growth rate has a diverse effect on women’s labor supply transitions. As the economy grows, women are more likely to become employed or to change from part-time to full-time employment, but their unemployment and job-search transitions are uncorrelated with the GDP growth rate. This finding suggests that the country’s economic condition, as measured by its GDP growth rate, does not affect the individual decision to participate in the labor market itself, but rather it contributes to the success in finding a job and entering into employment given that the labor supply decision has already been made. As the unemployment rate increases, however, women are more likely to enter the labor market and to start searching for a job. This result contradicts the hypothesis of the “discouraged worker effect,” which states that individuals who would otherwise have been looking for work tend to remain out of the labor market as the unemployment rate increases and their chances of getting a job fall. The country’s female labor force participation rate is negatively correlated with all transition probabilities considered except for the part-time to full-time transitions; that is, the more women already participate in the labor market, the less women are entering into it.

Our result of main interest is the estimated effect of the added worker dummy, which indicates whether the husband became unemployed between t–1 and t. To compare the magnitude of women’s behavioral response across our different outcomes, not only do we present the estimated marginal effect of the added worker dummy, we also calculate the percentage change in women’s probability of adjusting their labor supply in response to their husband’s unemployment.¹³

The results suggest that women whose husbands lost their job at any time during the last 12 months have a 3.6 percentage point (20%) higher probability of entering the labor market than those with a continuously employed husband. This effect, however, is driven only by wives’ changes into unemployment. Women with an unemployed husband are 3 percentage points more likely to enter into unemployment and 4.8 percentage points more likely to start searching for a job, which corresponds to relative changes in transition probabilities of 73% and 71%. Women’s probability of becoming employed, however, is not significantly affected by the husband’s employment status. This result is consistent with the findings of Lundberg (1985), who showed that married women in the United States were more likely to enter the labor market when their husbands were unemployed, but were even less likely to become employed. This finding suggests that husband’s unemployment indeed affects the wife’s willingness to work in the labor market. As stressed by Maloney (1991), however, some wives may have the will to enter the labor market but may not be able to find a job in the short term to offset the associated loss in household income.

We further find a strong behavioral response at the intensive margin of women’s labor supply. Women whose husbands became unemployed have a 6 percentage points (36%) higher probability of changing from part-time to full-time employment than do women with a continuously employed husband. That we find no evidence for an added worker effect in terms of women’s employment transitions, but a strong effect in terms of their part-time to full-time transitions, may be explained by the fact that part-time work provides greater scope for labor supply adjustments, as it is harder for women to increase their labor market activities by entering the labor market than it is by increasing working hours when they are already working. This result is consistent with the finding of Gong (2011), who found evidence for the existence of an added worker effect for married women in Australia, but also showed that this effect was mainly driven by part-time to full-time transitions of already-participating wives.

Overall, the results for our pooled sample covering all European countries reveal the existence of an added worker effect at both the extensive and the intensive margin of wives’ labor supply. This finding suggests that in Europe, marriage still functions as an intra-household risk-sharing mechanism to smooth inter-temporal income shocks (Attanasio, Low, and Sánchez-Marcos 2005; Ortigueira and Siassi 2013).¹⁴

A remaining concern in our analysis, however, is the problem of unobserved heterogeneity. In particular, it is not clear whether the husband’s job loss is exogenous to the wife’s labor supply. First, the husband’s unemployment might be anticipated by the household, either because it is voluntarily chosen or because it is of a more frequent nature and thus does not represent an unexpected shock to the household. Such anticipated unemployment would not induce an added worker effect, because it would already have translated into household adjustments in either consumption or labor supply. Furthermore, as pointed out by Maloney (1991), the frequent nature of the husband’s unemployment might be correlated with the characteristics of the wife. Given assortative matching in the marriage market, wives of frequently unemployed husbands are likely to face low market wage rates themselves and thus to show similarly low labor supply patterns as their husbands.

One way to overcome these problems is to search for exogenous variation in husband’s unemployment, for example, by focusing on unemployment caused by plant closures, which are assumed to be exogenous to the characteristics of the husband and the household, respectively. Although this approach is not possible in our study, we control for individual characteristics that are highly correlated with the husband’s risk of becoming unemployed in order to proxy for the unobservable factors being correlated with the recurring nature of the husband’s unemployment. Specifically, we add controls for the husband’s labor market experience, as measured by his share of years in employment and his previous job type, that is, whether the job was permanent or temporary. The respective estimation results are shown in online Table B.6. The results reveal that the more stable the husband’s employment, as measured by his share of years in employment, the less likely his wife enters the labor market or changes from part-time to full-time employment. Moreover, wives of husbands who had a temporary job in the previous year are significantly more likely to enter employment or to increase their working hours than are those whose husbands had a permanent position. The estimated marginal effects of the added worker dummy, however, remain significant and decrease only slightly in magnitude, suggesting that unobserved factors that are correlated with the husband’s unemployment probability do not impose a major problem for our analysis.¹⁵

Second, the husband’s unemployment might not be involuntary, but instead voluntarily chosen by the husband. In his decision to quit his job, the husband might therefore have already taken his wife’s labor supply decision into account. In this case, we would overestimate the true added worker effect due to reverse causality and joint decision-making within the household, respectively. Conversely, one might argue that voluntary job losses are long known by the household, such that the observed added worker effect is an underestimate of the true effect, since some women might already have adjusted their labor supply to the husband’s expected job loss. Although the data do not allow us to distinguish between voluntary and involuntary job losses, we try to test whether reverse causality caused by voluntary job losses imposes a problem for our analysis. We do so by assuming that voluntary job losses should, on average, last shorter than involuntary job losses, because those quitting their jobs have more time to search for a job or might already have found a new job before giving notice. If we therefore condition on at least three months instead of one month of husband’s unemployment in defining our added worker variable, thereby reducing the share of voluntary job losses, we would expect the estimated added worker effect to decrease if reverse causality is indeed a problem in our analysis. In fact, we find the opposite: When defining husband’s unemployment as the husband having had at least three months of unemployment within the last year, the estimated added worker effect increases for all outcomes (see online Table B.5), suggesting that reverse causality is of minor relevance in our analysis. Note, though, that the fact that the added worker effect increases when only longer periods of husbands’ unemployment are considered could also indicate that some women are induced to adjust their labor supply only when their husband’s unemployment lasts longer and the associated income loss is expected to be higher.

Last, note that although we cannot rule out that unobserved heterogeneity might lead us to under- or overestimate the true added worker effect in general, there is no reason to believe that this sort of unobserved heterogeneity varies over the business cycle or differently affects the estimation results in the respective country groups. Hence, our analysis of the variation of the added worker effect—both over time and across countries—which is the main focus of this article, should not be affected by unobserved heterogeneity.

Variation over the Business Cycle

Many explanations have been given for why the added worker effect may depend on the economic context. Previous literature has concentrated on comparing the added worker effect in times of economic up- and downturns, arguing that wives’ responsiveness to their husband’s job loss should be higher during recessions because of both the reduced ability to borrow against income losses and the more permanent nature of unemployment shocks during recessions. It is also possible, however, that the added worker effect decreases during times of economic downturn. Whenever unemployment rates are high, the chance of getting a job, and thus the expected wage of those without jobs, both fall. People who would otherwise have been looking for work might therefore become discouraged in a recession and tend to remain out of the labor market.¹⁶ Accordingly, we would expect the labor supply response of wives to their husband’s job loss to be smaller if unemployment is high. Moreover, more than the country’s general economic situation might affect the presence of the added worker effect. If the share of women already participating in the labor market is high, the potential of inactive wives to newly enter the labor market is low, suggesting that the size of the added worker effect in its traditional sense should be small whenever female labor force participation rates are high.

To determine whether the magnitude of the added worker effect varies with the countries’ macroeconomic conditions, interactions of the added worker dummy and 1) the GDP growth rate, 2) the unemployment rate, and 3) the female labor force participation rate are further added to the model.¹⁷ The average marginal effects of the added worker dummy at each point of the countries’ GDP growth rate are shown in Figure 1. Overall, we find no clear relationship between the added worker effect and the country’s GDP growth rate. If anything, women’s probability of starting to search for a job decreases slightly and their probability of changing from part-time to full-time employment increases slightly as the economy grows. Although the latter result contradicts the previous literature, which finds the added worker effect to be more present in times of economic downturns, our finding is quite intuitive. As the economy shrinks, firms might first cut down the working hours of those already employed, before having to rely on personnel layoffs to reduce their overall costs. As the economy recovers and GDP grows, women might therefore find it easier to increase their working hours and in this way expand their labor supply.

Figure 1.

Marginal Added Worker Effects over the GDP Growth Rate

For the interactions of the added worker dummy with the country’s unemployment rate (Figure 2), however, a different pattern emerges: As the unemployment rate rises, women are more likely to increase their labor supply as a reaction to their husband’s unemployment. Except for women’s part-time to full-time transitions, which hardly vary over the unemployment rate, this finding holds for all labor supply responses considered but is most pronounced for women’s unemployment and job-search transitions. Whereas these results contradict the discouraged worker hypothesis, they are consistent with the findings of Parker and Skoufias (2004), Mattingly and Smith (2010), and Bryan and Longhi (2013), who all found that the added worker effect is more present in periods of economic downturns. Bryan and Longhi (2013), in particular, showed that women in the United Kingdom substantially increased their job-search activity following a partner’s job loss during the 2008 to 2011 recession, though the increase in search during the boom was smaller and did not appear to translate into more success in finding work. These findings support the hypothesis that in times of high unemployment, husband’s job losses are less likely to be transitory and therefore more likely to result in a behavioral response of the wife.

Figure 2.

Marginal Added Worker Effects over the Unemployment Rate

The respective interaction effects for the country’s female labor force participation rate are shown in Figure 3. Overall, the added worker effect appears to decrease with the country’s female labor force participation rate. That is, the greater the participation rate of women in the labor market, the less likely it is that a wife enters the labor market in response to her husband’s unemployment. This relationship is particularly pronounced for women’s employment transitions and their job-search transitions, whereas women’s part-time to full-time transitions hardly vary over the distribution of the female labor force participation rate. That women’s labor supply adjustments at the extensive margin are more strongly related to the country’s female labor force participation rate than are their adjustments at the intensive margin is quite intuitive. We find that the ability of women to newly enter the labor market is lower when the share of women already participating in the labor market is higher, and that women’s ability to increase their working hours should hardly be affected by the female labor force participation rate.

Figure 3.

Marginal Added Worker Effects over the Female Labor Force Participation Rate

Variation across Country Groups

In the final part of our analysis, we separately estimate our basic regression for specific subsamples of countries to test whether the added worker effect differs across the welfare regimes in Europe. As outlined in the Data section, we distinguish between five welfare regimes, namely 1) the Nordic countries, 2) Continental Europe, 3) the Anglo-Saxon countries, 4) the Mediterranean countries, and 5) Central and Eastern Europe. The estimated marginal effects of the added worker dummy obtained from these subsample regressions are shown in Table 2 (and full estimation results are provided in Tables B.9 through B.13 in the online Appendix). The results reveal large differences in both the existence and the magnitude of the added worker effect across Europe.

Table 2.

Added Worker Effect by Country Group

	IA_t-1 → A_t	IA_t-1 → UE_t	IA_t-1 → E_t	Δ JS	PT_t-1 → FT_t
	Nordic countries
ME	0.089**	0.038*	0.041	−0.005	0.020
StdE	(0.042)	(0.020)	(0.042)	(0.027)	(0.038)
Δ%	25.76%	—	—	—	—
Observations	7,339	7,339	7,339	4,641	11,228
	Continental Europe
ME	−0.015	0.006	−0.024	0.015	0.058***
StdE	(0.020)	(0.007)	(0.019)	(0.013)	(0.020)
Δ%	—	—	—	—	63.05%
Observations	21,099	21,099	21,099	17,355	36,482
	Anglo-Saxon countries
ME	−0.064	0.020	−0.089**	0.024	0.069
StdE	(0.043)	(0.018)	(0.039)	(0.029)	(0.044)
Δ%	—	—	−39.07%	—	—
Observations	4,005	4,005	4,005	3,180	5,867
	Mediterranean countries
ME	0.066***	0.044***	0.018*	0.062***	0.095***
StdE	(0.012)	(0.009)	(0.009)	(0.011)	(0.027)
Δ%	47.10%	76.49%	—	82.73%	37.23%
Observations	29,232	29,232	29,232	27,028	11,920
	Central and Eastern Europe
ME	0.038***	0.029***	0.004	0.051***	−0.028
StdE	(0.015)	(0.009)	(0.013)	(0.012)	(0.035)
Δ%	20.86%	62.45%	—	82.68%	—
Observations	25,828	25,828	25,828	23,929	8,394

Source: EU-SILC; authors’ calculations.

Notes: Results represent average marginal effects, calculated as average effects over all individuals in the respective sample, and robust standard errors (clustered at the household level) obtained from probit estimations of women’s labor market transitions on a set of individual, household, and country characteristics. IA_t-1 → A_t refers to women’s transitions from inactivity to activity; IA_t-1 → UE_t refers to women’s transitions from inactivity to unemployment; IA_t-1 → E_t refers to women’s transitions from inactivity to employment; ΔJS refers to women’s job-search transitions; PT_t-1 → FT_t refers to women’s transitions from part-time to full-time employment. The added worker dummy takes value 1 if the husband becomes unemployed from t–1 to t and 0 if he stays employed. Δ% refers to the percentage change in women’s probability of adjusting their labor supply due to their husband’s unemployment. Percentage changes are shown for significant added worker effects (5% level) only. Full estimation results are shown in online Tables B.9 to B.13.

Asterisks denote statistical significance * at the .10 level; ** at the .05 level; *** at the .01 level.

In the Nordic and Continental European countries, we find only weak evidence for the existence of an added worker effect. In the Nordic countries, nonparticipating women are more likely to enter the labor market when their husband becomes unemployed, an effect that is mainly driven by wives’ changes from inactivity to unemployment (significant at a 10% level only). For women’s employment transitions, job-search transitions, and part-time to full-time transitions, there is no significant effect of their partners’ job loss. This result is consistent with the findings of Hardoy and Schøne (2014), who investigated wives’ behavioral responses to their husband’s job displacement in Norway. The authors found hardly any added worker effect at the intensive margin, but showed that three years after the husband’s displacement, previously nonworking wives of displaced husbands had labor market earnings that exceeded those of wives of nondisplaced husbands by approximately 5%.

In Continental Europe, by contrast, women paired with newly unemployed men are more likely to change from part-time to full-time employment, yet we do not find any behavioral response at the extensive margin of women’s labor supply in these countries. The difference in the type of behavioral response between the two country groups might be explained by differences in the structure of the workforce. Whereas the Nordic and the Continental European countries are both characterized by comparatively high female labor force participation rates, the share of part-time employment in all employment is particularly high in the Continental European countries and as such, part-time work may provide a greater scope for labor supply adjustments in these countries. The share of part-time employed women is particularly high in the Netherlands (76.7%), followed by Germany (46.3%), Austria (46.3%), and Belgium (41.2%) (2014 values; Eurostat 2015). As shown by Blau and Kahn (2013), such cross-country differences in female labor force participation rates and part-time employment rates are explained, in part, by differences in family-friendly policies across countries. Parental leave and benefit policies, the access to publicly provided childcare services, and regulations with respect to the right for part-time work differ largely across the European countries, which may explain the high variation in female labor force participation and part-time employment rates and thus the differential behavioral response of women to their husband’s unemployment across the different welfare regimes.¹⁸

Moreover, many of the Continental European countries (i.e., Germany, France, Luxembourg, and, partially, Belgium) are characterized by tax systems of income splitting, which might create disincentives for women to enter the labor market.¹⁹ If the difference between the husband’s and the wife’s (potential) earnings is high and the unemployment duration of the husband is expected to be low, then the couple might not be better off if the wife enters the labor market, as her gained earnings might be completely offset by the reduced amount of tax savings.

In general, the limited responsiveness of wives to their husband’s job loss in the Nordic and the Continental European countries might be explained by the high level of social protection guaranteed in these countries. Among the European countries, the Nordic and the Continental countries rank highest with respect to both the length and the amount of unemployment benefits, and it might be the generosity of the welfare state that partly crowds out the family as an insurance device.²⁰ The hypothesis that the state plays an important role in smoothing out income fluctuations caused by external shocks is also supported by Hardoy and Schøne (2014), who showed that the initial negative wage effect of husband’s displacement is reduced by approximately 65% after adjusting for welfare benefits and lower tax payments. This finding suggests that in a generous welfare state, households are well insured against negative shocks in the labor market.

In the Anglo-Saxon countries, we also find very little evidence for the existence of an added worker effect. Indeed, we even find a negative added worker effect. Women in these countries are significantly less likely to become employed when their husband becomes unemployed. Though this result might be driven by unobserved heterogeneity, in a sense that spouses with low labor market prospects or similar preferences for leisure select each other, it might also reflect the disincentive effect of the unemployment benefit system on wives’ labor force participation. A disincentive effect of a husband’s unemployment-related benefits on his wife’s labor supply arises if the husband’s benefits are linked to the wife’s earnings, either because benefits are means-tested on the basis of family income or because any part of the benefit is withdrawn when the wife is working or earning. The size of the adverse effect on wife’s labor force participation thereby depends on both the rate of withdrawal and its amount (Dex, Gustafsson, Smith, and Callan 1995). As such, the effect is arguably strongest in the Anglo-Saxon countries—the only countries in which the benefits received through both unemployment insurance and unemployment assistance contain elements that involve some kind of means-testing and the rate of withdrawal of benefit is particularly high. Usually, unemployment insurance represents an individual non-means-tested benefit, the size of which (given eligibility) depends solely on previous contributions. In Ireland, however, a substantial part of the benefit, the so-called dependent’s allowance, is paid only to individuals with a non-employed partner, and it is withdrawn if the partner earns more than a certain permitted amount. In addition, in both the United Kingdom and Ireland, unemployment allowance, which is paid to individuals who are not (or who are no longer) eligible for unemployment insurance, is withdrawn at a rate of almost 100% if the partner receives some sort of income, thus creating a strong disincentive to the labor force participation of wives of unemployed men.²¹

Such a disincentive effect is consistent with the findings of Kell and Wright (1990), who found large negative effects of means-testing on the labor force participation of wives married to unemployed husbands in the United Kingdom. Dex et al. (1995) came to a similar conclusion in their cross-country comparison of the labor force participation of married women in the United Kingdom, Ireland, the United States, Sweden, and Denmark. They found that in unemployment benefit regimes that take a wife’s earnings into account in allocating benefit, there was a significant negative effect on those wives’ labor force participation.

In the Mediterranean countries, by contrast, we find a strong and significant added worker effect. In these countries, wives whose husbands became unemployed within the past 12 months are significantly more likely to become employed (10% level), to enter unemployment, to start searching for a job, and to change from part-time to full-time employment than are wives with a continuously employed husband. In fact, the Mediterranean countries are the only countries in which an added worker effect at both the extensive and the intensive margin of women’s labor supply is observed. Moreover, the effects are quite large, ranging from a 37% increase in women’s probability of entering full-time employment to an 83% increase in their job-search probability.

The finding of a strong relationship between husband’s and wife’s labor supply in the Southern European countries supports previous literature on this topic (Prieto-Rodriguez and Rodriguez-Gutierrez 2000). The Mediterranean welfare states are characterized by offering a low level of social protection and by a strong reliance on the family. In his analysis of family ties across societies, Reher (1998) showed a dividing line between southern European societies, with their history of depending on strong and extended families to care for the elderly and the poor, and northern European and North American societies, with their weaker family systems and greater reliance on public and private organizations to provide social assistance. The low level of social protection in the Mediterranean countries is not only reflected in the short duration and low level of unemployment benefit provision, which creates a strong incentive for wives of newly unemployed husbands to increase their labor supply, but is also evident from the limited entitlement to paid parental leave after childbirth. In the Mediterranean countries, parental leave is usually granted for a short duration and is either unpaid (as in Greece and Spain) or characterized by low benefits (as in Italy and Portugal) (Gauthier 2011), which strongly incentivizes inactive mothers of young children to enter the labor market in case of their husband’s job loss. The strong added worker effect in the Mediterranean countries might therefore be explained by low social protection and a strong reliance on the family in these countries.

In the countries belonging to Central and Eastern Europe, we also find some evidence for the existence of an added worker effect. In contrast to the Nordic and the Continental European countries, however, wives’ responsiveness to their husband’s job loss is reflected only in their increased likelihood of entering unemployment and starting to search for a job. Women’s probabilities of entering employment or changing from part-time to full-time employment, however, are not affected by husband’s unemployment. This finding suggests that women in Central and Eastern Europe are willing to increase their labor supply in response to their husband’s job loss, but may be limited from the demand side of the labor market in that they are not able to find a job or to increase their working hours in the short term in order to offset the associated loss in household income.

Though we cannot claim that the estimated added worker effects as shown in Table 2 represent causal effects, we argue that the difference in the size and the direction of the added worker effect between the country groups is attributable solely to differences in the countries’ institutional and macroeconomic conditions and can thus be interpreted accordingly. Although assortative mating or reverse causality might lead us to over- or underestimate the added worker effect in general, there is no reason to believe that this sort of unobserved heterogeneity affects the country groups differently and is thus able to explain the difference in the added worker effect across welfare regimes.

A last concern, though, is that our results are driven by transitory shocks to the household, which affect the employment probability of husbands and wives alike. As outlined by Maloney (1991), a closure of a local plant, for instance, might directly result in the layoff of the husband, and indirectly lower the market wage rate or employment opportunities of the wife, thereby masking the existence of her behavioral response. In contrast to unobserved heterogeneity in general, such local transitory shocks are likely to vary over the country groups and are thus able to explain the observed difference in the added worker effect across welfare regimes.

To rule out that local (unemployment) shocks are the main driver of our results, we conduct two robustness checks. First, we re-estimate our country–group regressions by adding country–year fixed effects instead of single country and year fixed effects to our model to allow for country-specific shocks to the labor market. The respective estimation results are shown in online Table B.14. Overall, the estimated added worker effects remain stable in both significance and magnitude, suggesting that the difference in wives’ behavioral responses across countries is not simply a result of country-specific unemployment shocks.

Although country–year fixed effects account for all of the variance in women’s labor market transitions caused by time-variant differences between the countries, they might not fully capture shocks to the households’ local labor market. In a second step, we therefore add region–year instead of country–year fixed effects to our regressions. Information on the households’ place of residence within their country is available on the NUTS-2 or NUTS-1 level, which subdivide countries into smaller administrative units in the size of regions or provinces.²² As can be seen from online Table B.15, adding region–year fixed effects does not alter our results substantially—the estimated added worker effects remain stable in both significance and magnitude.²³ These results make us confident that the cross-country variation in the added worker effect documented in this article is not only an artifact of region-specific transitory shocks to the households but indeed reflects differences in the macroeconomic and institutional conditions between the welfare regimes.

Conclusion

In this article, we analyze the responsiveness of women’s labor supply to their husband’s loss of employment—the so-called added worker effect. Whereas previous empirical literature on this topic mainly concentrates on a single country, we take an explicit internationally comparative perspective and analyze whether the added worker effect varies across the countries in Europe. We follow the argumentation of Bentolila and Ichino (2008), who pointed out that the role of family support should be stronger whenever the welfare state fails to mitigate the consequences of unemployment.

In our analysis, we use longitudinal data from the European Union Statistics on Income and Living Conditions (EU-SILC) covering the period 2004 to 2013. As we observe households over the time of the Great Recession, we are further able to investigate the role of the added worker effect in Europe’s economic crisis by analyzing its variation with the countries’ economic conditions. We contribute to the literature by considering a variety of behavioral responses of wives to their husband’s unemployment, covering reactions at both the extensive and the intensive margins of labor supply, which is of particular interest and importance in any international comparative framework.

For our pooled sample of 28 European countries, we find evidence for the existence of an added worker effect. Women whose husbands become unemployed show a significantly higher probability of entering the labor market than do women whose husbands remain employed. This effect is driven primarily by wives’ changes from inactivity to unemployment and increases in their job-search efforts, whereas wives’ probability of becoming employed appears to be independent of the husband’s job loss. We find, however, that wives are more likely to increase their working hours in reaction to their husband’s unemployment. These results suggest that in Europe, marriage (or cohabitation) still functions as an intra-household risk-sharing mechanism to smooth inter-temporal income shocks (Attanasio et al. 2005; Ortigueira and Siassi 2013).

Our results further reveal that the magnitude of the added worker effect varies with the countries’ economic conditions. Whereas wives’ likelihood of increasing their labor supply as a response to their husband’s job loss increases with the country’s unemployment rate, no clear relationship exists between wives’ responsiveness and the country’s GDP growth rate. This finding suggests that it is the current conditions of the labor market rather than the country’s economic situation in general that affects couples’ labor supply behavior. In addition, we are able to show that wives’ probability of entering the labor market in response to their husband’s unemployment decreases with the country’s female labor force participation rate. As female labor force participation rates have increased remarkably over the past decades in most developed countries, this result might provide one explanation why more recent studies find hardly any evidence for the existence of an added worker effect in its traditional sense (Prieto-Rodriguez and Rodriguez-Gutierrez 2003; Gong 2011).

Furthermore, we show that the existence and the magnitude of the added worker effect largely varies over the different welfare regimes within Europe. Overall, the added worker effect is strongest among couples living in the Mediterranean countries and is less present in the Continental European and the Nordic countries. Although we are the first to provide comprehensive evidence on the added worker effect across Europe, our results are in accordance with previous literature, which tends to find no or small added worker effects in high-welfare countries, such as Norway (Hardoy and Schøne 2014), but stronger effects for low-welfare countries, such as Italy (Prieto-Rodriguez and Rodriguez-Gutierrez 2003), Spain (Prieto-Rodriguez and Rodriguez-Gutierrez 2000), and the United States (Stephens 2002). Hence, our results support the view that the role of the family as an insurance device against unemployment might be crowded out by the generosity of the welfare state. In addition, our finding of a negative added worker effect in the Anglo-Saxon countries, which are characterized by unemployment benefit systems that create a large disincentive for the labor force participation of wives of unemployed men, highlights the important role of the unemployment insurance system in compensating for income losses caused by involuntary job losses while maintaining incentives for intra-household labor supply adjustments.

Finally, we find large differences in the type of behavioral response to husbands’ job loss across countries. Whereas women in the Nordic countries are more likely to increase their labor supply at the extensive margin, women in Continental Europe are more likely to do so at the intensive margin. Furthermore, we find that women in the Central and Eastern European countries are highly limited from the demand side of the labor market, in that they respond to their husband’s unemployment in terms of increased job-search activity, but that these attempts do not translate into more success in finding work. These results stress the importance of considering different behavioral responses of wives to their husband’s job loss, including measures of both the extensive and the intensive margin of labor supply, in providing a meaningful comparison of the added worker effect across countries.

Footnotes

We are grateful to Ronald Bachmann, Barbara Petrongolo, Sandra Schaffner, and participants at the 3rd European User Conference for EU-LFS and EU-SILC, the 8th Workshop on Labour Markets and Demographic Change, the 7th RGS Doctoral Conference in Economics, the 2nd NEUJOBS workshop, the 28th ESPE conference, the 26th EALE conference, and the 20th Spring Meeting of Young Economists for helpful comments and suggestions.

The online Appendices discussed in this article may be found at .

For information regarding the data and/or computer programs utilized for this study, please address correspondence to the lead author at julia.bredtmann@rwi-essen.de.

Notes

References

Ashenfelter

Orley

. 1980. Unemployment as disequilibrium in a model of aggregate labor supply. Econometrica 48(3): 547–64.

Attanasio

Orazio

Low

Hamish

Sánchez-Marcos

Virginia

. 2005. Female labor supply as insurance against idiosyncratic risk. Journal of the European Economic Association 3(2-3): 755–64.

Ayhan

Sinem H.

2017. Married women’s added worker effect during the 2008 economic crisis—The case of Turkey. Review of Economics of the Household. doi: 10.1007/s11150-016-9358-5.

Başlevent

Cem

Onaran

Özlem

. 2003. Are married women in Turkey more likely to become added or discouraged workers? LABOUR 17(3): 439–58.

Bentolila

Samuel

Ichino

Andrea

. 2008. Unemployment and consumption near and far away from the Mediterranean. Journal of Population Economics 21(2): 255–80.

Blau

Francine D.

Kahn

Lawrence M.

2013. Female labor supply: Why is the United States falling behind? American Economic Review: Papers & Proceedings 103(3): 251–56.

Bonoli

Guliano

. 1997. Classifying welfare states: A two-dimension approach. Journal of Social Policy 26(3): 351–72.

Bryan

Mark

Longhi

Simonetta

. 2013. Couples’ labour supply responses to job loss: Boom and recession compared. IZA Discussion Paper No. 7775. Bonn: Institute of Labor Economics.

Burda

Michael C.

Hunt

Jennifer

. 2011. What explains the German labor market miracle in the Great Recession? Brookings Papers on Economic Activity 42(1): 273–335.

10.

Cullen

Julie B.

Gruber

Jonathan

. 2000. Does unemployment insurance crowd out spousal labor supply? Journal of Labor Economics 18(3): 546–72.

11.

Dex

Shirley

Gustafsson

Siv

Smith

Nina

Callan

Tim

. 1995. Cross-national comparisons of the labour force participation of women married to unemployed men. Oxford Economic Papers 47(4): 611–35.

12.

Esping-Andersen

Gosta

. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press.

13.

Eurostat. 2015. Statistics Database. Accessed at http://epp.eurostat.ec.europa.eu/ (July 15, 2015).

14.

Gauthier

Anna H.

2011. Comparative family policy database, Version 3 [computer file]. The Hague: The Netherlands Interdisciplinary Demographic Institute and Max Planck Institute for Demographic Research (distributors). Accessed at http://www.demogr.mpg.de (November 18, 2016).

15.

Gong

Xiaodong

. 2011. The added worker effect for married women in Australia. Economic Record 87(278): 414–26.

16.

Hardoy

Inés

Schøne

Pål

. 2014. Displacement and household adaptation: Insured by the spouse or the state? Journal of Population Economics 27(3): 683–703.

17.

Heckman

James J.

MaCurdy

Thomas E.

1980. A life cycle model of female labour supply. Review of Economic Studies 47(1): 47–74.

18.

Kell

Michael

Wright

Jane

. 1990. Benefits and the labour supply of women married to unemployed men. Economic Journal 100(400): 119–26.

19.

Kohara

Miki

. 2010. The response of Japanese wives’ labor supply to husbands’ job loss. Journal of Population Economics 23(4): 1133–49.

20.

Lundberg

Shelly

. 1985. The added worker effect. Journal of Labor Economics 3(1): 11–37.

21.

Maloney

Tim

. 1987. Employment constraints and the labor supply of married women: A reexamination of the added worker effect. Journal of Human Resources 22(1): 51–61.

22.

Maloney

Tim

. 1991. Unobserved variables and the elusive added worker effect. Economica 58(230): 173–87.

23.

Mattingly

Marybeth J.

Smith

Kristin E.

2010. Changes in wives’ employment when husbands stop working: A recession-prosperity comparison. Family Relations 59(4): 343–57.

24.

McGinnity

Frances

. 2002. The labour-force participation of the wives of unemployed men: Comparing Britain and West Germany using longitudinal data. European Sociological Review 18(4): 473–88.

25.

Nordblom

Katarina

. 2004. Cohabitation and marriage in a risky world. Review of Economics of the Household 2(3): 325–40.

26.

[OECD] Organisation for Economic Co-operation and Development. 2010. OECD Employment Outlook 2010. Paris: OECD Publishing.

27.

[OECD] Organisation for Economic Co-operation and Development. 2016a. Benefits and wages: Policies. Accessed at http://www.oecd.org/els/benefits-and-wages-policies.htm (February 12, 2016).

28.

[OECD] Organisation for Economic Co-operation and Development. 2016b. Benefits and wages: Statistics. Accessed at http://www.oecd.org/els/benefits-and-wages-statistics.htm (February 12, 2016).

29.

Ortigueira

Salvador

Siassi

Nawid

. 2013. How important is intra-household risk sharing for savings and labor supply? Journal of Monetary Economics 60(6): 650–66.

30.

Parker

Susan W.

Skoufias

Emmanuel

. 2004. The added worker effect over the business cycle: Evidence from urban Mexico. Applied Economics Letters 11(10): 625–30.

31.

Prieto-Rodriguez

Juan

Rodriguez-Gutierrez

Cesar

. 2000. The added worker effect in the Spanish case. Applied Economics 32(15): 1917–25.

32.

Prieto-Rodriguez

Juan

Rodriguez-Gutierrez

Cesar

. 2003. Participation of married women in the European labor markets and the “added worker effect.” Journal of Socio-Economics 32(4): 429–46.

33.

Reher

David S.

1998. Family ties in western Europe: Persistent contrasts. Population and Development Review 24(2): 203–34.

34.

Spletzer

James R.

1997. Reexamining the added worker effect. Economic Inquiry 35(2): 417–27.

35.

Stephens

Melvin

. 2002. Worker displacement and the added worker effect. Journal of Labor Economics 20(3): 504–37.

36.

Sullivan

James X.

2008. Borrowing during unemployment: Unsecured debt as a safety net. Journal of Human Resources 43(2): 383–412.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.36 MB