_ Agnes Fauske, Norwegian Institute of Public Health, Department of Health and Inequality; Rannveig K. Hart, Norwegian Institute of Public Health, Department of Health and Inequality, Centre for Fertility and Health and Centre for Evaluation of Public Health Measures. Oslo, 24 February 2020.
This article escribes the results of a systematic review of the literature of policy effects on fertility after 1970 in Europe, USA, Canada and Australia. Empirical studies were selected through extensive systematic searches, with subsequent literature list screening. Inclusion was conditional on implementing an experimental or quasi-experimental design. 57 published papers and recent working papers were included, covering the topics of parental leave, childcare, health services, universal child transfers and welfare reforms. Childcare and universal transfers seem to have the most positive effects on fertility. Few effects were found for parental leave, but this could be linked to these reforms not lending themselves to efficient (quasi)experimental evaluation. Withdrawing cash transfers to families through welfare reforms has limited fertility effects. Subsidizing assisted reproductive technologies show some promise in increasing birth rates of women above age 35.
We have summarized studies of the effect of policies on fertility, based on an extensive and systematic search of both published articles and working papers (> 17 000 screened). We have found five groups of policies that are evaluated with respect to whether they influence fertility: Parental leave, childcare, universal transfers, health services and welfare reforms. Of these, especially childcare, universal transfers and some types of health services tend to have positive effects on fertility.
Concerns about falling fertility are mostly linked to concerns about decline in future labour supply, and countries who aim to increase fertility often tend to simultaneously want to preserve or even increase maternal labour supply (Thévenon 2011). Which of the evaluated policies unite these goals? The most obvious policy is accessible and reasonably priced childcare. In contexts where childcare coverage is high, one can speculate that improvements on accessibility, such as opening hours compatible with non-standard work hours, or quality could have further positive effects. We do not have empirical studies that assess these dimensions, however.
Universal transfers also seem to increase fertility, at least when they are substantial. A challenge with large universal transfers is that they may act as a disincentive to paid work, particularly among women with many children and lower levels of education or work experience. However, when transfers are given as tax breaks, they will to a lesser degree work as a disincentive to paid work.
Finally, a targeted intervention that may have a small but significant effect is offering subsidized assisted reproductive treatment at all ages where the success rate is of meaningful size. That such services increase birth rates if available at a low cost to couples likely to benefit from them is founded in data.
Second, what does not work? The welfare cutbacks seen in the last decades in the USA and UK seem to have very limited impact on fertility. Whether such policy packages in “reversed form” have the potential to increase fertility is dubious, and a reform package that increases fertility while reducing labour supply at the lower end of the income distribution also does not seem politically feasible. To the extent that family income is important for child health and wellbeing, such an attempt would also raise some ethical concerns.
Our largest knowledge gap seems to be on the effect of parental leave. This is not due to a lack of studies, but rather because the nature of parental leave reforms makes them difficult to evaluate. Given that long parental leaves are costly, evaluating them in an experimental design akin to the US tradition would provide important insights. It would, however, only be politically feasible to randomly allocate additional parental leave benefits or rights, meaning that effects would be evaluated at yet another margin.
Finally, we note that several studies point towards announcement effects of policies. Policies perceived as supportive to families that signal childbearing as valuable to society tend to have more positive fertility effects. In contrast, if increased (female) labour supply is the main policy aim, fertility effects (if any) will be heterogenous across groups, and sometimes negative in sum.