Swapping saving schemes: are pensions the new children?

While new parents are mastering the art of nappy-changing, or bouncing their bundle of joy on their knee, they are unlikely to think of themselves as handling an 'investment good'. But before the inception of taxpayer-funded pensions and private savings products, having children was just about the only way to ensure a comfortable old age.

Fertility has declined markedly in developed countries over the past century. In the US for example, each woman had an average of 3.2 children in 1920 and that had slipped to 2.1 by 2000. There have been falls of a similar magnitude in France and Germany.

One explanation commonly cited by economists for this dramatic social change is that many people can now rely on state-funded assistance in their retirement - so they do not need to have children to look after them. It used to be that children would repay their parents' kindness by supporting them in old age. In effect, this was a way of 'saving'. However, once other saving mechanisms exist, this is no longer necessary. In effect, pensions have replaced children as a means of preparing for a comfortable old age. In a new CEPR paper, CEPR Research Associates, Vincenzo Galasso and Roberta Gatti, and their co-author, Paola Profeta, test whether this argument is true.

In order to analyse the relationship between saving, the availability of pensions, and the decision to have children, the authors develop an OLG - overlapping generation - model. This uses equations to formalise the decisions parents and children have to make: for parents, choosing to have children and bring them up is costly; but they receive a return in their old age when their offspring support them financially. The children pay a certain proportion of their income in social security contributions, and later, receive a pension in return.

Crucially, Galasso et al. include in their model a measure of how developed the local financial markets are. Instead of relying on their children to provide them with an income in old age, the parents could choose to invest their money themselves and rely on the return they will receive. But inefficient, or under-developed markets are likely to be more difficult and costly to access, in effect reducing the return on a given investment. If parents really are thinking of their children partly as an investment for the future then the availability of this market-based alternative should also influence their decisions, as well as the possibility of a state-funded retirement scheme.

When the authors solve their economic model to reveal the optimal choices for the decision-makers represented in it, they find that the more inefficient are the local financial markets - and the worse the return parents would get on their investments - the more worthwhile it is for them to have children. In other words, where financial markets are not very developed, the appearance of a state-funded pension system offering security in old age, should exert a larger downward effect on fertility decisions.

With this key prediction in hand, Galasso et al. then use data on fertility rates and pensions saving from more than 80 countries to test whether it is reflected in reality. They carry out a regression analysis to probe the connection between the availability of state-funded pensions and fertility, and control for a number of other factors that could skew the results, including the size of the elderly population in each country and GDP per capita. As a measure of the availability of social security they use World Bank data on the share of the population covered by state pensions, and also test alternative measures from the IMF and ILO to ensure that one particular source of data is not biasing their findings.

In order to examine whether the efficiency of local markets is important, as they suspect, the authors also include a measure of the level of private credit in the economy as a proportion of GDP.

Across the eighty plus countries they examine, in general, the authors find that the availability of a social security system does indeed tend to reduce average fertility. On its own, their separate measure of private market efficiency is not significant in determining how many children a mother chooses to have. Crucially, however, the empirical evidence supports their prediction that a state-funded pension is all the more important in depressing fertility where markets are weak.

This link with the option of other pensions saving outside the state system is important corroborating evidence for the idea that, when fertility declines after the institution of social security it is not simply coincidence, but a result of the fact that taxpayer-funded pensions make it easier to plan financially for a lifetime, without relying so much on the generosity of family members. One reason mothers in developed countries with sophisticated social safety nets have fewer children appears to be that - strange as it may seem - they no longer need them.

The authors suggest that the links they have uncovered help in understanding not just the history of wealthy nations with established social security systems, but the implications of future policy decisions in a much wider range of economies. In China, for example, where pension coverage is relatively low at around 10%, expanding the availability of taxpayer-funded retirement scheme could be a powerful lever for keeping fertility low, without the need for the controversial 'one child policy'.

DP 6825 Investing for Old Age: Pensions, Children and Savings

Vincenzo Galasso, Roberta Gatti and Paola Profeta