(Reprinted from HKCER Letters, Vol. 8, May 1991)
Social Insurance and Economic Growth
Editor's Note: There have been repeated calls from trade unions and other interest groups for the establishment of a central provident fund in Hong Kong. The issues have been debated extensively over the past few years. The government and legislative councillors have rejected the scheme on the very sensible grounds that it would tie up large sums of funds in a single institution, affecting the economy adversely.
More recently there is a renewed interest in a public retirement scheme. The scheme requires government, employees and employers to each contribute about 2 percent of an employee's salary to a fund. All residents over the age of 65 would receive a fixed pension from the fund. Such a scheme would not be fully funded as in the case of a central provident fund, but would instead be a pay-as-you-go scheme. Under this scheme the amount of an individual's contribution rises with his income, but the amount he gets after retirement is presumably fixed.
Very similar pay-as-you-go schemes have been adopted in many countries. The U.S. social security system is a very good example of such a scheme. The problems of such a scheme are well known. Professor Isaac Ehrlich who recently visited the Centre spoke about the long-run, damaging consequences of such a scheme for economic growth and its effects on fertility decisions and family behavior. His talk, reproduced below, deals with a very complex problem and is well worth reading.
The Explanation of Economic Growth
The subject of the talk today is family insurance, social insurance, and economic growth. The central idea is part of an emerging literature in economics titled the "New Economic Growth and Development."
In the twentieth century, we have had Robert Solow's neoclassical growth theory, which has had immense impact on academics as well as policymakers. When I say growth, I mean the rate of growth in per capita income, not the level of GNP. Solow's model was a very interesting exercise in trying to explain how levels of per capita income are reached through some sort of golden rules. The model left unexplained, however, is the rate of growth. In that model, the rate of growth was supposed to be determined by exogenous technological innovations. Hence, one important implication of that model is that different countries will tend to grow at the same rate even if their levels of income differ. This cannot explain the huge diversity in rates of growth of different nations that we observe in reality. The growth rates of Hong Kong, Taiwan, Korea, and Japan have been consistently higher than those of other countries.
That is what the new literature aims to explain: the inequality in growth rates. In order to do that we need to identify an engine of growth, some sort of factor and asset that can be accumulated over time without being subject to diminishing returns. Solow's model told us that it cannot be physical capital because if physical capital is raised relative to labor, it will be subject to diminishing returns. The new literature looks for an engine of growth that does not have to do with physical capital formation. That factor is knowledge. Knowledge can have two forms-- it can be embodied or disembodied. Disembodied knowledge refers to technical relations, while embodied knowledge is what economists call human capital. If we take technology as the engine of growth, we are again at loss in explaining the diversity in the rates of growth across countries because technology is easily transferrable. Hence, there is a stress on human capital as the engine of growth.
Human capital theory talks about decisions that individuals make privately as to how to maximize their life-cycle return by investing in education. That is a static, one-generation perspective. If each generation only undertakes to maximize their own interests, they will reach an optimal level of return on their investment. It will explain their own per capita income, but it cannot by itself explain growth.
The Family and Economic Growth
Fundamentally the problem of growth is we have to explain why the current generation cares enough about the future so that somehow forces can be unleashed in an economy that would assure that the rates of growth of per-capita income of the future generations will always be higher than that of the current generation. The main idea of the new theory is that the center of growth is the family. What links the generations within the family is a kind of a complex, very powerful, and very positive force of dependency. It goes both ways. Children are dependent on their parents for their material well-being, but more importantly, if the engine of growth is human capital, they are dependent on their parents for the growth in their knowledge.
Growth takes place essentially because parents care enough about their children. The older generation invests some of its knowledge and resources in the education of the younger generation. It may also be emotional. Parents love their children. On the other hand, there is, in principle, another side to that relation. For parents-yet-to-be, it is reasonable to expect, at least on the average, that their welfare will be dependent to some extent on contribution from their children. It may again involve material and non-material aspects as well. While there is mutual interdependency that links generations, the more powerful is this linkage, the greater is the likelihood that sufficient resources will be spent by the older generation on the education of the younger generation so that growth can be self-generating and self-sustaining.
What I have described essentially is a private insurance arrangement. I can call it a family insurance, or a private security system, where old-age dependency is basically taken care of within the family. I do want to suggest that this is not an abstract theory because we have direct evidence that such intra-family transfers do take place in reality.
Now social insurance is simply another way of taking care of this intergenerational arrangement. Incidentally, one of the reasons perhaps that the main sort of transfers from the young to their old, dependent parents are not as prevalent in affluent countries like in the United States is social security. Such transfers occur outside the family; therefore, social security does not require the family to facilitate intergenerational transfers. The question is what ramifications would a program of this sort have on the prospect of growth.
The new theory indicates that the social insurance program may affect three variables at the same time: the rate of saving; the rate of investment that parents make in their children; and therefore, the rate of growth of the economy; and finally, the fertility rate. The latter two may be affected because, in our story, the family insurance scheme hinges on the contribution that all of the children make to their dependent parents, and so both the quality and the quantity of children that parents would bear matter. To the extent a social security scheme affects the growth path or the fertility rate, we have to consider the cost of introducing such a scheme.
Theoretically, there is one very effective scheme, which I would call an actuarially fair fund, that would not interfere with growth. It may be known by the name of provident fund. Insurance funds that are actuarially fair guarantee that one will receive the expected benefits given ones investment into the fund. If children are more numerous, or if they are of higher quality so they are of higher earning power, they will pay more into the social security fund when their time comes to make these payments. If parents know that their children's contribution into the social security scheme is going to be taken into account when their private benefits are determined, they will have absolutely no incentive to lessen either the number of children that they bear or reduce the contribution that they make into their investment. Put simply, if a social security scheme is introduced that is a perfect substitute for family insurance as described above, then any amount of social security tax on the population will get absorbed immediately through the family calculus.
The Pay-as-You-Go System
In most economies where social security schemes are introduced, however, they are not introduced as a provident fund. Instead they are introduced as a pay-as-you-go system. The younger generation is taxed, and the amount of taxes collected is what is given back as benefits to those people who are fortunate enough to grow into old age and become recipients. The pay-as-you-go system is definitely not an actuarially fair system from the family members' point of view. The reason is that the law says how much one is going to receive today, a fixed amount of benefit regardless of one's actual contribution. This kind of scheme does not reward any more parents who put investment in their children than those who do not. It reduces the incentives that parents have to make the investment. I would call it a moral hazard problem, and in this case it is also known as a free rider problem. Parents may realize that if they invest less in their children or raise a smaller number of children, the fund will have fewer resources, perhaps even for themselves, by the time they get to be recipients. But they have absolutely no incentive to take that into account. The rate of return on parental investments in their children's education, or in the number of children they raise, is not going to be as high as it is under a strict family insurance scheme, and that is the problem.
The effect of a social insurance scheme may take the form of a reduction in the rate of fertility. If the reduction in the rate of fertility is sufficiently great, it is possible, theoretically at least, that the other two determinants of the growth path are not going to be affected. So in a way it becomes partly an empirical issue, to find out in countries where they did introduce a social security system which looks like a pay-as-you-go system, whether the main detrimental effect has been on the rate of fertility. To the extent that is the case, we would expect less harm to be done to the other determinants of economic growth.
The evidence shows that the fertility is affected adversely, but not dramatically. Therefore, the theory will anticipate that either the level or the rate of growth will be affected. The bottom line is that the empirical evidence indeed supports the fact that the introduction of social security systems will have a detrimental effect on the economy's rate of growth.
Another implication from this way of looking at the family and its contribution to growth is that the longevity of the population is also conducive to growth. We make here a distinction between two types of longevity, that of the young and that of the old. By that I mean the survival rates of children from childhood to adulthood, and the survival prospect of adults into old age. The theory suggests both of them can contribute to growth. When parents make an emotional and material investment in their children, they may expect to get a return. Hence, the greater the survival probability of their children, the more profitable is the investment. On the other hand, if the parents anticipate that they themselves would be living longer, they have to better take care of their old-age needs. They can do that through savings, or alternatively, they can invest in their children. They will have some incentive to make contributions to their children's education because, again, that is going to be part of the return that they will get. In comparison, the impact of aging on growth is less critical than the impact of the young people's longevity. Thus, social insurance schemes that may have the impact of raising population longevity, primarily young-age longevity rather than old-age longevity, will make a positive contribution to growth. In sum, even though social security may have some disadvantages because of its moral hazard aspect, there could also be some advantages if longevity is enhanced.
Professor Isaac Ehrlich is a Leading Professor at the State University of New York at Buffalo and the Director of the Institute for the Study of Free Enterprise Systems.