|
Article Excerpt
Abstract--We examine the tax-deduction incentive for parents to have babies in December rather than January, focusing on whether or not parents get "shotgun married." We choose this focus because non-shotgun-married parents are more likely than shotgun-married parents to time conception and, consequently, the birth of their children. Results show that in the 1976 forward sample, the tax deduction is positively correlated with December births among non-shotgun-married parents, who are likely to time conception for tax purposes, but not among shotgun-married parents. In other samples, the differential effect between shotgun- and non-shotgun-married parents, is consistent as predicted.
INTRODUCTION
Taxes are levied to encourage or discourage individuals from engaging in various activities; needless to say, these activities are not limited to those that are traditionally considered to be related to economics. A few studies have attempted to determine whether and to what extent taxes provide incentives to change human behavior, such as, the timing of human births, which is the subject of our paper. Several demographic studies show that some environmental factors, such as nutrition, temperature, and photoperiods, are related to the monthly distribution of births (see Lam and Miron (1996), and Bronson (1995, 2004)). Even if this is true, however, it does not imply that economic factors are unrelated to the pattern of human births. In fact, Dickert-Conlin and Chandra's (1999) seminal study on taxes and birth timing uses data from the National Longitudinal Survey of Youth (NLSY) and discovers that tax-deduction benefits result in a surge in childbirth from the first week of January to the last week of December; this is due to the tax-deduction system in the U.S., according to which a child born on December 31 receives all the tax benefits in that year, whereas a child born one day later, in the next calendar and tax year, does not receive any tax benefits for the previous year. In addition, Gans and Leigh (2006) analyze the effect of a $3,000 "Baby Bonus" announced by the Australian government on May 11, 2004 for children born on or after July 1, 2004, to postpone the timing of births in Australia. (1)
In this paper, we examine whether or not a tax deduction for dependents provides an incentive for the timing of births, using Japanese micro data from the 1998 and 2003 administrations of the "Nationwide Survey on Families" (in Japanese, Kazoku ni tsuiteno Zenkoku Chousa), conducted in January 1999 and January 2004.
One feature of our paper is that we focus on whether or not the parents get "shotgun married." A shotgun marriage or a shotgun wedding is commonly understood as a rushed marriage or wedding because the bride is pregnant. The reason we focus on such parents is that parents who get shotgun married are presumably less likely to time conception and, consequently, the birth of their child; therefore, their pregnancies are less likely to be planned. On the other hand, parents who do not get shotgun married are more likely to time conception; thus, their pregnancies are more likely to be planned. We incorporate the interaction term of tax-deduction benefits and shotgun status into our estimations, which enables us to examine our main hypothesis: parents who do not get shotgun married and stand to gain more in terms of higher tax-deduction benefits for dependents are more likely to have a baby in December than in January.
By breaking down the monthly distribution of births in our data based on whether or not parents are shotgun married, we can observe an important feature with regard to December births. Empirically, we consider marriages in which babies were born within six months or less after marriage as shotgun marriages and those in which babies were born twelve months or more after marriage as non-shotgun marriages (see the first subsection of the fourth section for details on our shotgun/non-shotgun classification). In Figure 1, the bottom line--depicting the difference between the share of babies born to non-shotgun-married parents and the share born to shotgun-married parents in each month--shows a large spike in December. In December, the share of babies born to non-shotgun-married parents is 3.4 percentage points higher than the share born to shotgun-married parents. (2) This presents circumstantial evidence that a tax deduction for dependents or other socioeconomic factors provide non-shotgun-married parents with incentives to have babies in December rather than January. More precisely, it suggests that environmental factors, such as nutrition, temperature, photoperiods, and the number of holidays decreased the number of December births of children of both shotgun- and non-shotgun-married parents, which may cause a large dip in the number of babies born to shotgun-married parents in December (the dotted line of the top two lines). On the other hand, socioeconomic factors, such as a tax deduction for dependents, increased December births only of children of non-shotgun-married parents (the solid line), thereby countervailing the dip pertaining to shotgun-married parents.
Our paper contributes to the literature on the effect of tax on fertility in the following three ways: first, focusing on whether or not parents are shotgun married provides a different and interesting source of variation since we take account of the difference of the effect of tax deductions on the timing of births between shotgun-married parents and non-shotgun-married parents. In other words, the parameter of our interest is the average effect of tax policy on the timing of birth for parents who are likely to time conception, rather than the average effects of the tax policy on the timing of birth for all parents, regardless of type (shotgun-married or not). The latter is the parameter adopted by Dickert-Conlin and Chandra (1999). The difference between our paper and theirs is the fact that they do not attempt to identify which parents were more likely to time birth, while we do. The reason for this is that they focus on the timing of delivery by Caesarean section (C-section) and labor induction rather than the planning at the time of conception, nine to ten months prior to delivery. All parents have the option to time birth through a Caesarean section or labor induction around the time of delivery. On the other hand, we focus on the timing at the point of conception, which is an option available only to non-shotgun-married parents. Second, we can think of shotgun-married parents as being a control group. That is, since only non-shotgun-married parents have the tax incentive to time conception, they are treated as having a tax incentive for conceiving in time to get the tax benefit. However, both shotgun-married and non-shotgun-married parents are subject to noninstitutional and external factors (such as temperature, photoperiods, and the number of holidays). Finally, focusing on whether or not parents are shotgun married clarifies how much power they have over the timing of birth. The timing of birth may be affected not only by parents, but also by doctors and/or hospitals. The preferences of doctors and/or hospitals affect the decision to manipulate the timing of delivery by Caesarean section and/or labor induction. However, we are interested in the effect of the above factors on the parents' decision-making process rather than that of doctors and/or hospitals. Since both types of parents can manipulate the timing of delivery through medical intervention, we have to note that if the use of a medical intervention makes the timing of delivery differ between shotgun-married and non-shotgun-married parents by unobservable characteristics, then the estimates will be biased. However, as we will see in the next section, in Japan, these medical interventions do not have much impact on the timing of birth.
[FIGURE 1 OMITTED]
Our results show that in the case of the 1976 forward sample, a tax deduction is positively correlated with December births among non-shotgun-married parents, who are likely to time conception for tax purposes, but not among shotgun-married parents, who are unlikely to do so. That is, non-shotgun-married parents who stand to gain ten thousand yen in terms of higher tax-deduction benefits are 0.9 percentage points more likely to have a baby in December than in January. In the case of the full sample and the 1991 forward sample, the differential effect between shotgun-married and non-shotgun-married parents is consistent with non-shotgun-married parents reacting to tax incentives as predicted. That is, non-shotgun-married parents who stand to gain ten thousand yen in terms of higher tax-deduction benefits are only 0.6 and 1.5 percentage points less likely to have a baby in December than in January, while shotgun-married parents are 4.4 and 20.4 percentage points less likely to do so.
This paper is organized as follows. In the next section, we explain the controllability of the timing of birth and the personal income tax system in Japan. The third section describes the data source.
The fourth section presents variable definitions and the calculation method. The fifth section describes the sample selection. The sixth section presents the estimation model and the estimation method. The seventh section presents the descriptive statistics. The eighth section presents our estimation results, and the final section concludes the paper.
THE TIMING OF BIRTH AND THE TAX DEDUCTION FOR DEPENDENTS
Controllability of the Timing of Birth by Shotgun- and Non-shotgun-Married Parents
The main assumption in this paper is that non-shotgun-married parents plan the timing of childbirth for tax purposes, while shotgun-married parents do not. In order to examine the validity of this assumption, we examine whether babies born to non-shotgun-married parents are more likely to be planned than others, based on the practice of obstetrics and gynecology and on cultural views in Japan.
First, several pieces of evidence point out the high costs and stigma associated with out-of-wedlock childbearing in Japan. According to the "International Attitude Survey on the Society...
|
