The unintended consequences of pro-birth policies: a cautionary tale from China
As Singapore enhances incentives to tackle declining birth rates and an ageing population, a new study reveals how China’s two-child policy inadvertently caused female students to scale back their investment in higher education.
Wong Wei Chen
22 December 2025
Singapore’s birth rate dipped to a historic low of 0.97 births per woman in 2023 – substantially below the replacement rate of 2.1 – and remained stagnant at this level in 2024. Slowing reproduction, in tandem with an ageing population, was a cause for much consternation, and, in a bid to tackle this problem, the Singapore government rolled out a slew of policies, including an enhanced Baby Bonus Scheme introduced in 2023 and the doubling of government-paid paternity leave in January 2024.
While the stated objectives are to raise birthrates, rejuvenate the population and strengthen the workforce, could such interventions inadvertently trigger unintended consequences, as a new study on China’s two-child policy illustrates? The research reveals that the policy shift actually discouraged female college students from investing in their studies, thereby compromising their future competitiveness in the job market.
Background
In 2016, China implemented a countrywide two-child policy, which relaxed the previous one-child policy that was in force between 1979 and 2015. After decades of population control, the Chinese government found itself having to contend with serious demographic and economic imbalances originating from a rapidly greying population. The strict one-child limit had created a 4-2-1 family structure, where a single working-age adult was saddled with an unsustainable burden of supporting two parents and four grandparents. Concurrently, the workforce shrank as retiring labour could not be sufficiently replaced, threatening to stall the engine of the country's economic growth.
The solution seemed obvious – allow families to have more children and at the same time encourage them to do so with incentives such as extended maternity leave. Yet, a study by Choi, Gao, Lam, Li, and Qian, "Scared Away: Credit Demand Response to Expected Motherhood Penalty in the Labor Market", uncovered an externality that in fact compromised female representation in the workforce.
Analysing data from the online student loan market, Choi et al. found that post-reform loan applications from female college students decreased by a significant 15.6% compared to male students. This phenomenon was apparently fuelled by female students pre-emptively “pricing in” the motherhood penalty. Realising that employers now had to bear the costs of motherhood – including extended maternity leave – female candidates feared their employability would suffer, and were therefore more reluctant to take on debt to further their studies.
Empirical design and findings
Given that the study was done in a quasi-experimental setting, the researchers deployed a difference-in-differences (DiD) approach to isolate the causal impact of the two-child policy.
By utilising male students as a control group – whose employability was theoretically insulated from maternity-related costs – the model effectively filtered out confounding variables such as national economic fluctuations or seasonal borrowing trends, such as during school holidays.
The validity of the DiD approach critically hinged on the “parallel trends” assumption, which required the variables under study to maintain a consistent trend relative to one another prior to the exogenous shock (in context, the two-child policy).
The researchers found that prior to policy rumours which emerged between July and August 2015 – months ahead of the official implementation on 1 January 2016 – borrowing behaviours between male and female students were statistically indistinguishable. While “male” and “female” borrowings may have fluctuated randomly over time, the gap between the two groups generally remained stable, which satisfied the DiD’s criterion for a pre-shock parallel trend.
The divergence in borrowing behaviours began around July/August 2015, coinciding with a surge in keyword searches regarding the two-child policy on Baidu Search, China’s biggest and most dominant search engine. Using August as the starting point (month “0”), Choi et al. found a drop of 6.18% in female loan applications relative to male applications for the period August to December 2015. In other words, the mere rumour of the policy was sufficient to trigger a 6.18% decline, months before the policy was even rolled out.
The anticipatory dip was, however, modest compared to what followed. Following official policy implementation in January 2016, the researchers initially discovered an average drop of 12.75% among female loan applications. However, this figure underestimated the policy’s true impact. When the researchers adjusted the baseline to account for the rumour-driven decline starting in August 2015, they found that female loan applications had actually plummeted by 15.6% relative to their male university counterparts.
Additionally, given that college tuition fees range from ¥4,000 to ¥7,000, the average loan quantum of approximately ¥7,900 suggests that female students were cutting back on financing their studies, as opposed to personal consumption loans which tend to be of smaller amounts.
Choi et al. noted that the pullback was most acute among students in STEM (Science, Technology, Engineering, and Mathematics) majors, where gender disparity is already more pronounced. Hence, female talent with promising career trajectories apparently anticipated the steepest motherhood penalty from employers.
Robustness tests
To ensure that the observed decline was indeed a reaction to the policy, and not arising from an unrelated economic trend, Choi et al. conducted a battery of robustness checks to rule out alternative explanations.
Among other tests, the researchers investigated whether the decline was driven by a supply-side restriction – that is, whether lenders, rather than the students, were the ones discriminating. If lenders feared that female graduates would struggle to repay loans due to the motherhood penalty, they might have tightened credit availability for female students. A clear corollary of more stringent screening would be a lower loan delinquency rate. However, the delinquency rate remained statistically unchanged after the shock, which again supported the hypothesis that the decline in female applications originated from the demand side.
The researchers additionally deployed a staggered DiD approach to exploit the fact that different provinces implemented maternity leave extensions at different times. Since extended maternity leave directly incurs additional costs for employers, the prediction is that female students would become even more conservative in acquiring debt after their specific province enforced the new mandate.
The results corroborate this hypothesis. The study found that female college students were more reluctant to borrow immediately after the extension of maternity leave by their provincial governments. On average, the extension led to a 2.3% decrease in loan applications from female students relative to their male counterparts in the first month after the change. The effect continued to escalate up to 6.1% in the first two months following the change, and remained significant and stronger than the initial magnitude even in the third month after the reform.
Implications for Singapore – the case for shared parenting
The findings from China are sobering and offer a cautionary tale for Singapore. They demonstrate that well-intentioned policies to support mothers can inadvertently backfire if they increase the perceived cost of hiring women. Rational actors, such as female college students covered in the study, will therefore adjust their behaviours accordingly.
However, by granting government-paid paternity leave and doubling it in 2024, the Singapore government is pivoting towards a more gender-neutral policy stance, where both mother and father have child-raising responsibilities and hence also enjoy the benefits. Perhaps the most salient point from Choi et al. is this – to boost birthrates without compromising female representation in the workforce, policy must move beyond “helping mothers” to “normalising parenthood” for both sexes.
Choi, Darwin is an associate professor in the Department of Finance, HKUST Business School, The Hong Kong University of Science and Technology.
Gao, Zhenyu is an associate professor of finance in CUHK Business School, The Chinese University of Hong Kong.
Lam, Sing-Sen is an assistant professor of finance at Lingnan College, Sun Yat-Sen University.
Li, Tianyi is an assistant professor of finance at the Department of Decisions, Operations and Technology at The Chinese University of Hong Kong.
Qian, Wenlan is the Ng Teng Fong Chair Professor in Real Estate and Professor of Finance and Real Estate at the NUS Business School, National University of Singapore.