Labor supply consequences of family taxation: Evidence from the Czech Republic☆
Introduction
The choice of an appropriate family tax treatment is an important part of an optimal tax design. While individual taxation systems tax each individual's income separately, systems of joint taxation either tax the sum of the family income as a whole or tax each spouse individually based on half of the total income (Stephens and Ward-Batts, 2004). Joint taxation meets the requirement for equal treatment of households with the same total income—the tax liability of a married couple is the same regardless of how income is divided between spouses (Cigno et al., 2011). However, joint taxation equalizes the marginal tax rates of the spouses and thereby decreases the marginal tax rates of primary earners (usually men) and increases the marginal tax rates of secondary earners (usually women). The effect of joint taxation on the labor supply of married men is ambiguous, because the substitutions and income effects work in opposite directions, but the theoretical effect on the labor supply of married women is unambiguously negative. This study is concerned with the empirical investigation of these labor supply aspects of family tax treatment.
Countries are not unified in their choice of tax unit. Even though individual taxation is in force in the majority of EU countries, a tax law often contains features that provide incentives similar to those of a joint taxation system, and tax systems based on joint taxation are not exceptional either.1 Fig. 1 shows that, indeed, countries with systems of (truly) individual taxation (Sweden, Denmark, Finland, the United Kingdom, etc.) tend to have higher female employment rates (for a given level of male employment) than countries with joint taxation systems or systems with ‘joint’ features.
Although economic theory predicts a negative female labor supply effect of joint taxation, there is little empirical evidence as a result of the lack of recent policy changes with respect to family taxation. Two studies have estimated the impact of joint taxation on the labor supply of married women using family taxation reforms: LaLumia (2008) and Selin (2009). Although both studies provide a comprehensive analysis of the changes in the tax treatment of families in the U.S. and Sweden respectively, their results are based on tax reforms that are more than 40 years old. Among others, Blau and Kahn (2007) show that the female labor supply elasticities and behavioral responses to tax reforms have changed significantly since the 1980s, pointing to the need for more up-to-date evidence.
This paper exploits the most recent family taxation reform, the introduction of joint taxation in the Czech Republic in 2005, to estimate the labor supply effect of joint taxation.2 From January 1, 2005, married couples raising at least one child could have taken the opportunity for joint taxation in the Czech Republic. Since the actual usage of joint taxation among eligible couples is unknown, what I estimate here is the intention-to-treat effect of this reform.3
I apply a difference-in-differences approach with several alternative treatment and control groups to evaluate the effect of joint taxation on the married women's and married men's labor supply. The whole analysis is conducted separately by gender. First, I compare married individuals with children (all eligible) with unmarried individuals and married individuals without children (all ineligible). Next, I use the discontinuity in the eligibility rule—children are defined by a strict age threshold in the Czech tax code, which is 18 years, or 26 years in the case of full-time students. Therefore, I focus on a more homogeneous subset of the sample and compare married individuals with children aged 10–17/25 and married individuals with children aged 18/26–30. Furthermore, I apply a local difference-in-differences estimation around the two age thresholds—comparing married individuals with children aged 16–17 vs. those aged 18–19 (not in education), and married individuals with children aged 24–25 (in education) vs. those aged 26–27. Finally, I provide several robustness checks including the triple differences estimation (with an additional control group of Slovak married individuals with children)4 and two placebo tests to check the validity of the estimation approach.
This project sheds new light on the effect of the family tax treatment on the labor supply of married men and women with children. The estimates show that joint taxation decreases the labor supply of married women with children—it is associated with a decline of 2.9 percentage points in their employment rate. Moreover, I show that those women who experienced the highest decline in work incentives did indeed respond with the largest decrease in employment probability (by 5.5 percentage points).
The estimated effect for married men is largely insignificant at the extensive margin, supporting the findings of LaLumia (2008), who also did not find any effect of joint taxation on the labor supply of married men. The effect of joint taxation on hours worked by married men is negative and significant in most specifications, but of a very small magnitude, which is consistent with the income effect slightly outweighing the substitution effect.
The remainder of the paper is organized as follows. The next section reviews the relevant literature, then the institutional background of the Czech reform analyzed in this paper is introduced, with an ensuing discussion of the methodology and identifying assumptions of the chosen approach. Finally, the paper presents the results and concludes.
Section snippets
Literature review
Recently, there has been an expansion in the literature that simulates the effect of a switch from joint to individual taxation on female labor supply (among others, see Steiner and Wrohlich, 2004; or Haan, 2010). However, these microsimulation studies face common problems connected to the estimation of labor supply effects. Blundell et al. (1998) argue that the “[l]abor supply effects have been notoriously difficult to estimate in a robust and generally accepted way” (p. 827). The main reason
Institutional background
The policy change of interest in this study is the introduction of joint taxation in the Czech Republic in 2005. Before 2005, there was a progressive individual tax system with four tax brackets.6 Married couples were taxed individually based on each individual's income. Joint taxation
Simple model of family labor supply
My empirical strategy is based on a simple model of family labor supply, which is often referred to as a unitary model (Samuelson, 1956). This model treats the household as a single decision-making-unit assuming that spouses pool their resources and maximize joint utility.15
Difference-in-differences approach
In this section, I present the main estimation results of the effect of joint taxation on the extensive and intensive margins of the labor supply of married men and women with children. Table 4 reports the difference-in-differences coefficients (the interaction of the treatment group dummy with joint taxation years) for women, where Columns 1 to 4 correspond to the four treatment and control groups. The effect on employment decisions of married women with children is negative and significant
Robustness analysis
In this section, I provide additional evidence supporting the validity of the difference-in-differences approach and the conclusions drawn from its results. In particular, I provide results of the triple differences approach with Slovak married individuals serving as additional control group and of two difference-in-differences placebo tests.
Conclusion
The theoretical impact of joint taxation on the labor supply of secondary earners (usually women) is negative, while the impact on primary earners (usually men) is ambiguous. Despite the fact that married couples file jointly in many European countries, the magnitude of the joint taxation effect on labor supply remains unclear as the empirical literature is limited by a lack of recent policy changes with respect to family taxation.
This paper utilizes the most recent family taxation reform, the
References (23)
- et al.
Taxes and the labor market participation of married couples: the earned income tax credit
J. Public Econ.
(2004) A multi-state model of state dependence in labor supply: intertemporal labor supply effects of a shift from joint to individual taxation
Labour Econ.
(2010)The effects of joint taxation of married couples on labor supply and non-wage income
J. Public Econ.
(2008)- et al.
Identification of causal effects using instrumental variables
J. Am. Stat. Assoc.
(1996) - et al.
Individual versus joint taxation in models with household production
J. Polit. Econ.
(1999) - et al.
Unnatural experiments? Estimating the incidence of endogenous policies
Econ. J.
(2000) Gender unemployment gaps: evidence from the new EU member states
- et al.
Labor supply after transition: evidence from the Czech Republic
Czech J. Econ. Financ.
(2011) - et al.
Changes in the labor supply behavior of married women: 1980–2000
J. Labor Econ.
(2007) - et al.
Estimating labor supply responses using tax reforms
Econometrica
(1998)
Introduction to the symposium on taxation and the family
CESifo Econ. Stud.
Cited by (9)
The Taxation of Couples
2023, SSRNTill taxes keep us apart? The impact of the marriage tax on the marriage rate
2023, International Tax and Public FinanceLabor supply responses to income tax free and bracket expansions
2022, Journal of Economic StudiesRedistributive effects of the taxation of couples and families: A microsimulation study of income tax
2021, Economie et Statistique
- ☆
This research was supported by the Grant Agency of the Charles University, grant no. 595812, and by the Grant Agency of the Czech Republic, grant no. 14-36154G. This paper has greatly benefited from comments by Richard Blundell, Alena Bicakova, Stepan Jurajda, Patrick Gaule, and two anonymous referees. I also thank the participants of the 2013 EALE and ESPE Conferences for the helpful suggestions. This paper is based on data from Eurostat, EU Labour Force Survey 2002–2007. The responsibility for all remaining errors and conclusions drawn from the data lies entirely with the author.