This is our 10th blog post for our Job Market Paper Series blog for 2025-2026.
Jijee Bhattarai is a PhD student at Binghamton University. Her interests are in the fields of labor economics, development economics, and economics of gender.
In many South Asian families that prefer sons, breastfeeding does more than nourish. It quietly times the next birth. Parents who want another try for a boy can wean earlier to resume ovulation. What happens when abortion becomes legal and families can say “not yet” in a different way? My job market paper studies Nepal’s 2002 abortion legalization and shows how access to abortion reshaped fertility decisions and closed much of the pre‑existing gender gap in breastfeeding, without large effects on vaccinations or child survival.
The question
Does legal access to abortion change how families in son‑preferring societies plan births and invest in their children? If households used breastfeeding to speed the next conception before legalization, then a law that lowers the cost of avoiding unwanted births should reduce the need to shorten breastfeeding for daughters. We should see fertility plans adjust and breastfeeding for girls rise. That is exactly what we find.
Context
Before 2002, Nepal had one of the world’s most punitive abortion regimes. Abortion was a criminal offense in all circumstances, including rape and incest, and women could be imprisoned for seeking care. In 1997, about 20 percent of women in Nepali prisons had been convicted of abortion or infanticide. The 2002 reform decriminalized abortion up to 12 weeks of gestation and later under specific conditions, with services scaling up nationwide in the years that followed. The reform arrived in a setting with well‑documented son preference and where postpartum amenorrhea and breastfeeding were key parts of fertility management.
Definitions
Firstborn-girl (FBG) families are households whose first child is a girl. Firstborn-boy (FBB) families have a boy first.
Son-biased stopping means families keep having children until they have the desired number of sons, causing more births after daughters than after sons.
Breastfeeding duration is months of any breastfeeding.
Data
We combine multiple waves of the Nepal Demographic and Health Surveys (NDHS, 1996–2022) with information on the timing of the reform. We study completed fertility, birth spacing, and breastfeeding duration, along with child health investments like vaccinations and early‑life mortality. The unit of analysis is the family, with a focus on whether the firstborn is a girl or a boy. This firstborn sex contrast provides a simple and transparent way to benchmark son‑biased fertility behavior in the population.
Research design
Families with a firstborn girl (FBG) historically had stronger incentives to keep trying for a son. Consistent with prior work and my data, first births show no evidence of sex selection; it begins at second births, once the first child’s sex is known. This supports using firstborn sex as a simple, quasi-random contrast for benchmarking son-biased behavior. Therefore, we compare children born in firstborn‑girl families to firstborn‑boy families before and after the 2002 reform. This difference‑in‑differences style comparison isolates changes that are specifically tied to the incentives created by son preference interacting with abortion access. We also examine event time around legalization and show that trends were parallel pre‑reform.
What we find
- Fertility adjusts: Legalization substantially reduced son‑biased stopping. The pre‑reform gap in completed family size between firstborn‑girl and firstborn‑boy families falls from 0.257 to 0.100 (a decline of 0.157 births per family), closing roughly three-fifths of the pre-reform disparity.
- Breastfeeding gaps close from the girls’ side: Before legalization, girls were weaned 1.7 months earlier than boys in families with a firstborn girl. After legalization, girl–boy breastfeeding gap drops, closing almost 96% of the pre-reform gap. Deconstructing this further shows that the gender gap narrows mainly because girls get more, and not because boys get less.
- Limited change in other inputs: We find small and imprecise effects on vaccinations and on child mortality. These inputs were less directly linked to the fertility‑timing channel that legalization affected.
Why this pattern makes sense
Before 2002, when abortion was not a safe or legal option, parents who wanted another birth soon had one effective lever: shorten breastfeeding to hasten the return of fertility. Once abortion became available at low legal risk and financial cost, families could avoid or postpone a pregnancy without sacrificing breastfeeding. The incentives that previously pushed parents to wean daughters earlier weakened. That channel is asymmetric, so the gap closes because girls’ breastfeeding increases.
External validity and limits
Nepal’s reform occurred in a lower‑income context with pronounced son preference. The mechanisms likely extend to similar settings where breastfeeding is a key spacing tool and where sex‑selective fertility behavior exists. At the same time, health systems, legal access, and social norms vary, so the size of effects will differ across contexts. The analysis focuses on average effects and cannot observe private costs or stigma that some households may still face when seeking abortion services.
What this means for policy in South Asia and beyond
First, abortion legalization is not only about reducing unsafe procedures. It also affects how parents plan families and allocate time‑intensive care. In places with strong son preference, improving access to safe, legal abortion can reduce the need for fertility‑driven compromises that disadvantage daughters. Second, legal abortion moves choices tightly linked to fertility timing; it does not automatically raise all child investments. Complementary policies may be needed to shift vaccination or survival where other constraints bind.
Bottom line
Legal access to abortion in Nepal reduced son‑biased fertility stopping and closed much of the gender gap in breastfeeding by raising breastfeeding for girls. These concentrated but meaningful shifts suggest that giving families a safe way to say “not yet” can bend intrahousehold choices toward greater equity without large changes in other health inputs.
