«When a Son is Born: The Impact of Fertility Patterns on Family Finance in Rural China∗ Weili Ding Yuan Zhang Queen’s University Fudan University ...»
When a Son is Born: The Impact of Fertility Patterns on
Family Finance in Rural China∗
Weili Ding Yuan Zhang
Queen’s University Fudan University
This paper examines the impact of an observable shock to households in rural China, the
oﬀspring gender structure, on household ﬁnancial activities. We develop theoretical channels
that endogeneously generate heterogeneity in the levels of ﬁnancial activities on the basis of a child’s gender, even if the parents do not possess discriminatory tastes. Using nationally representative household data collected in 300 rural Chinese villages and econometric models that account for endogenous fertility and sex selection, we present strong evidence that having a son signiﬁcantly increases both the amounts that a family will loan or give to relatives as well as increase the amounts of gifts they receive from others. Having a son increases the amount of gifts received from others by over 50% and is also found to increase household investments in both agricultural activities and family businesses. Finally, we present evidence that these family structure variables should not be treated as exogenous and demonstrate the robustness of our results to a number of criteria used for sample construction, speciﬁcation and to account for alternative selection biases. Taken together these results suggest that social norms or convention play important roles in household ﬁnancial decisions that extend beyond the traditional role of budget constraints and consumption shocks. This has clear implications for policies that aim to address rising sex imbalance amid economic growth and discriminating investment to female children in developing countries.
PRELIMINARYPlease Do Not Quote * We are grateful to Steven Lehrer for helpful discussions would also like to thank Li Rui for generously providing information on the data used in the study. Ding wishes to thank SSHRC for research support.
Please direct correspondance to Weili Ding at firstname.lastname@example.org.
1 1 Introduction Understanding the role of rural ﬁnance has been an important research agenda in development economics. The prevailing consensus is that informal network-based loans and transfers provide insurance against negative shocks in consumption, production and health (Fafchamps 1992; Rosen- zweig 1988 1993; Udry 1994; Townsend 1994). Most empirical testing of this theoretical consensus has shown that such insurance is only partially achieved (Morduch 1991; Grimard 1997; Fafchamps and Lund 2003; De Weerdt and Dercon 2006). A detailed examination of some informal ﬁnancial networks reveals that they were largely kinship based and geographically constrained, thus limiting the organizations’ ability of risk hedging (Fafchamps and Gubert 2007). The mechanisms and costs required to enforce the informal contracts might be responsible for the limited scales observed in order to sustain the networks (Murgai et al. 2002). Thus this line of research has completed a satisfying sequence of explanations from the demand side on the role of informal network-based ﬁnance.
What is missing from this literature is household heterogeneity, that households of diﬀerent types might have diﬀering demand for loans and transfers when faced with the same shock and same budget constraints. Symmetrically, these households are likely to provide diﬀerent supply of loans and transfers when faced with the same environment. The addition of household heterogeneity could potentially augment the picture on rural ﬁnance in a signiﬁcant way. For example, exogeneous negative or diﬀering income, debt or asset levels are no longer required to generate network-based lending and borrowings.
Although rare in rural ﬁnance,1 the aforementioned heterogeneity is often studied in many streams of development economic literature. Researchers have reported signiﬁcant gender diﬀerences in consumption and human capital investment patterns of families with boys versus girls as 1 The exception includes Rahman(1999) that documents females, while less likely to obtain loans, were more likely to return loans. La Ferrara (2003) ﬁnds that borrowers who have children in kinship band networks in Ghana are less likely to default, which means that the family structure has a signiﬁcant eﬀect on borrower’s repayment decision.
2 well as within families (Jacoby 1994; Doelalikar and Rose 1998; Behrman, 1988 1992 1998; Rose, 2000; Chowdury and Bairagi 1990; Das Gupta 1987; Kishor 1993; Muhuri and Preston 1991). If families with girls consume and purchase education diﬀerently from families with boys, one might reasonably expect families of diﬀerent oﬀspring structures also vary in their ﬁnancial activities in formal and informal networks. The focus of our paper is to examine a major household heterogeneity, the oﬀspring gender structure, on household ﬁnancial activities. We employ a nationally representative data of households in rural China as the cultural and institutional features there generate distinctively diﬀerent expectations and incentives for parents from the birth of a boy over a girl.
Children in China are viewed by many parents as the most important contributor of their old-age care. This is particularly the case in rural China where social security or community-based old-age care system has been scarce. However, like in many developing and developed countries, not all children are equal when it comes to old-age care (Astone et al. 1999). Cultural norms in rural China have it that male adult children are primarily responsible for the care of their elderly parents, while female adult children are mainly responsible for the care of their elderly in-laws. Together with the common belief that female laborers are less productive than their male counterparts in agricultural production, these conventions provide powerful economic incentives for parents to favor sons over daughters. Rural Chinese parents have stronger incentives to invest into the physical capital, human capital and social capital of their sons over their daughters due to the expected higher returns sons would bring over parents’ lifetime. This paper bases its primary analysis on the increased incentives to make intergenerational investments when the family starts to have son.
The son preference, however, is not monotonically increasing with the number of sons. Conditional on having a son, the arrival of a daughter brings at least two beneﬁts to a rural Chinese family: the bride price that the family receives when they marry oﬀ their daughter, which usually helps ﬁnancing the bride price that family has to pay towards the marriage(s) of their son(s) and 3 some of the wage income from the daughter before she is married oﬀ that helps the family to feed, cloth and educate her younger siblings, especially son(s) (Greenhalgh 1994; Lin 1993; Parish and Willis, 1993; Tatyana and Vaithianathan, 2008). Thus, although the preference of the ﬁrst son over any daughter is strong, conditional on having a son, a daughter, especially an elder daughter who would start working earlier and marry oﬀ earlier, could be more welcome than son(s) in some situations.
The strong son preference in rural China not only manifests in discriminating investments but more prominently results in sex selection in fertility.2 There is a rich literature in demography documenting sex selection in developing countries, especially in East Asia (Chu 2001; Hull, 1990;
Banister 2004; Kim 2005; Murphy 2003; Yi et al. 1993; Johnson 1996). The main method of sex selection in China has changed to sex selective abortions from abandoning female infants and female infanticide with the widespread adoption of ultrasound machines in China from late 80’s to early 90’s (Chu 2001; Murphy, 2003; Yi, et al., 1993; Johnson, 1996; Ebenstein 2008). Thus the main empirical challenge of this paper is to properly account for the possibility of sex selection for each rural family when investigating the impact of a son, especially the ﬁrst son, on household ﬁnancial activities.
We propose a speciﬁc mechanism in the style of statistical discrimination on how the arrival of a son makes a family invest diﬀerently. There is strong social convention in rural China that sons should provide ﬁlial support for their elders.3 Parents who value this support are incentivized to invest more into their son and thus deeper engagement in the rural ﬁnancial market despite the lack of an explicit “taste” against daughters. Parents may also be incentivized by cultural and social 2 Sen (1990) is a well cited reference that documents the high ratios of males to females in China and concerns it generates.
3 Cameron and Cobb-Clark 2001) and Das Gupta et al. (2003) both note that the persistence of son preference is driven by greater anticipated old age support from sons relative to daughters and the absence of formal ﬁnancial mechanisms for families to save for retirement.
4 reasons to favour sons. Non-pecuniary incentives may make parents invest more into a son even when he is not expected to deliver more monetary beneﬁt to parents compared to a daughter. We prefer the economic explanation not only because we believe it is the more important mechanism when it comes to household heterogeneity in rural ﬁnance in China but also is due to its capability of producing empirically testable predictions.
This paper is organized as follows. Section 2 provides an overview of the prominent culture, institution and fertility history of rural China and a review of the literature on rural ﬁnance and sex selective fertility. The data is described in Section 3. In Section 4 we state the theoretical mechanism and empirical hypotheses before presenting and discussing the empirical results. We present strong evidence that having a boy increases both the amounts that a family will transfer outside and to relatives as well as increase the amounts they receive from these sources. Not only are they more active in transfers but they also invest more in agricultural activities and family businesses. We ﬁnd that having an additional child increases the amount of funds received but if that child is a boy one receive a 50% premium. We also present evidence that these family structure variables should not be treated as exogenous and that the results are incredibly robust to a number of criteria used for sample construction, accounting for alternative selection biases and speciﬁcation.
Section 5 is the concluding section.
1.1 The Literature on Rural Finance and Sex Selection
This paper relates to branches of the development economics literature that examine formal and informal mechanisms in rural ﬁnance, fertility and evidence for sex-selection. Within rural ﬁnance, it is well established that when confronted by a negative income or consumption shock, rural household have limited access to formal mechanisms. As such, informal channels are used to generate funding from others and social networks have played a large role in these activities. Not surprisingly, these activities have led to both theoretical and empirical investigation within economics. The 5 main empirical challenge in this area is trying to identify diﬀerent kinds and timing of shocks that household face.
In general, researchers present evidence that members within a social network are more likely to obtain loans and insurance which is then used to smooth consumption.4 There is also mounting evidence that the credit, gifts, and other economic transactions provide insurance for social network members.5 Lastly, evidence indicates that these ties between households in social network provide similar returns to that which would have been achieved by purchasing insurance contracts that protect against the consequences of adverse events such as earnings losses and illness (Caldwell et al., 1986; Rosenzweig, 1988; Rosenzweig and Stark 1989). Thus, these kinship and marital ties that exist within informal mechanisms in rural ﬁnance have been termed “‘insurance’ capital” by Rosenzweig (1993). Our study contributes to this research program by investigating the impact of an accurately observed and important shock to rural families - the arrival of a son.
The arrival of a son has been argued to be preferred in rural regions where households’ livelihood depends mainly upon agricultural production. Further, parents typically depend upon their sons 4 For example, using longitudinal household data from rural India, Rosenzweig (1988, 1993) ﬁnds that interhousehold ﬁnancial transfers play a small but signiﬁcant role in contributing to consumption-smoothing. Using data from northern Nigeria, Udry (1994) reports that within informal credit institutions there is a great deal of activity by individuals on both sides of the credit market. Speciﬁcally, within a single year he ﬁnds that approximately 75% of households made loans, 65% of households borrowed (50% participated as both lenders and borrowers), and 97% of the loans (weighted by value) were between neighbors or between relatives. Lastly, the role of credit as a smoothing device has long been recognized in the sovereign debt literature (e.g., Eaton and Gersovitz, 1981; Kletzer, 1984;
Grossman and Van Huyck, 1988).
5 For example, Fafchamps (1992) presents evidence that solidarity systems are usually organized around delayed reciprocity contingent upon need and aﬀordability. In other words, solidarity is a form of mutual insurance and can provide protection against many sources of risk. Fafchamps and Lund (2003) using detailed data on gifts, loans, and asset sales in the rural Philippines,) ﬁnds that income and expenditure shocks have a strong eﬀect on gifts and informal loans, but little eﬀect on sales of livestock and grain. Mutual insurance does not appear to take place at the village level; rather, households receive help primarily through networks of friends and relatives.