As China relaxes its draconian fertility laws, a key question for policymakers is how the resulting increase in fertility will affect economic performance – in particular, whether it will lower household savings. This concern is motivated by the observation that China’s rapid rise in household savings rates coincided with a drastic reduction in fertility that began in the mid 1970s as a result of what came to be known as the one-child policy. During the 1990s, household savings rates reached highs of 34%.
Comparing savings rates between households with different numbers of children in China, recent research has found evidence to support the conventional wisdom that a key cause of the rise in savings rates is the decline in fertility (Banerjee et al. 2011, Choukmane et al. 2014, Ge et al. 2012, Jin 2013). The pattern over time can be seen in Figure 1, which is taken from Modigliani and Cao (2004: Fig 3). It shows that the savings ratio and the ratio of the employed population to minors (E/M) both begin to rise in the mid-1970s.
The most commonly cited logic for why increased fertility will lower savings is based on two mechanisms.
- The first is often called the transfer channel. Parents in developing and transitioning economies see children as an important source of old age support. This could be for psychological or financial reasons. Note that in rural China there are no state-provided pensions. In urban China, state-provided pensions were relatively generous by the standards of the pre-reform era. In the post-reform era, pensions drastically declined in net present value because they have not been indexed to the high growth rates experienced by the Chinese economy (and the accompanying rise in prices). In the absence of adequate pensions – and unable to save via children – Chinese increased their private saving rate.
- The second is the expenditure channel. When parents have more children, they must spend more to raise their children on items such as school fees, food, and clothes. Since savings is the difference between income and expenditures, having more children will therefore reduce savings.
While logical, the transfer and expenditure mechanisms cannot be applied to policy concerns regarding aggregate fertility change without taking other factors into account. In particular, the channel whereby increased aggregate fertility reduces savings – because parents expect future transfers from adult children – can be smaller when accounting for the fact that an increase in aggregate fertility will trigger other changes in the economy that affect savings. In Banerjee et al. (2014), we illustrate this point using data from urban Chinese households to show that for the reasons discussed in existing studies, an increase in the fertility of a household may reduce that household’s savings. But an increase in aggregate fertility – i.e. the fertility of all households in the economy – may have little effect or even increase household savings. This is because aggregate fertility growth can, for example, reduce future capital-to-labour ratios, which in turn will increase interest rates and reduce wage growth (Barro and Becker 1989, Galor and Weil 1996). Since higher interest rates reduce the value of future transfers from children, parents will save more today. Similarly, reduced future waged growth reduces the future earnings of children, and therefore the support they can provide elderly parents. This will also induce parents to increase savings today.
We argue that not accounting for other changes due to the change in fertility will cause researchers and policymakers to significantly overstate the negative effect of an increase in aggregate fertility on savings. The exact magnitude depends on factors such as how the percentage of income transferred to elderly parents changes when an adult child has more siblings. In our illustrative example, we use household-level data from 19 cities to show that failing to account for the effect of increased aggregate fertility on interest rates will overstate the reduction in savings caused by the change in fertility by two-thirds.
For policymakers, this means that abandoning the one-child policy will not necessarily lead to large reductions of household savings.
Banerjee, A, X Meng, and N Qian (2011), “The life cycle model and household savings: Microevidence from urban China”, Yale University Working Paper.
Banerjee, A, X Meng, N Qian and T Porzio (2014), “Aggregate Fertility and Household Savings: A general equilbrium analysis using household data” NBER Working Paper 20050.
Barro, R J, and G S Becker (1989), “Fertility choice in a model of economic growth”, Econometrica pp.481–501.
Choukhmane, T, N Coeurdacier, and K Jin (2013), “The One-Child Policy and Household Savings”, LSE working papers, London School of Economics.
Curtis, C C, S Lugauer, and N C Mark (2011), “Demographic Patterns and House-hold Saving in China”, NBER Working Papers 16828, National Bureau of Economic Research, Inc.
Galor, O, and D N Weil (1996), “The Gender Gap, Fertility, and Growth”, The American Economic Review, pp.374–387.
Ge, S, D T Yang, and J Zhang (2012), “Population Policies, Demographic Structural Changes, and the Chinese Household Saving Puzzle”, IZA Discussion Papers 7026, Institute for the Study of Labor (IZA).
Modigliani, F, and S L Cao (2004), “The Chinese Saving Puzzle and the Life-Cycle Hypothesis”, Journal of Economic Literature, 42(1), 145–170.