3rd May 2011
The demand for men regulates the production of men
Recently, in the O'Rourke's book discussed yesterday, I discovered this quotation from Adam Smith which effectively pre-empts Gary Becker's work by two centuries.
"The demand for men, like that for any other commodity, necessarily regulates the production of men." [Book 1, Wealth of Nations]
This also reminded me of my doctoral dissertation of 12 years ago, that tests in detail this demand for men (demand for children). The dissertation is available here should you wish to download and have a detailed look. The concluding chapter of the dissertation is posted below. I do wish I had spent more time in editing the dissertation, but the key message is there.
My dissertation helped me see the human economy as an interactive strategic network of anticipations and reactions. That is pretty much what Hayek or Adam Smith would say, and in that sense Smith preceded the rational expectations school by two centuries as well. My model of human fertility can be extrapolated into a consideration of the social contract (which is what I do, briefly, in DOF).
Concluding chapter of my dissertation
6.1 Review of the study
There has been considerable criticism about the standard models of fertility ignoring the strong effects of uncertainty and institutional factors such as information asymmetries and the moral hazard arising from this asymmetry. Many attempts have been made to overcome this noticeable weakness in theory and some progress has been independently by many authors by the introduction of ‘real-life’ complexities like social norms, strategic bequests, access to markets, etc. In this study we made an attempt to integrate, in Chapter 1 and 2, the existing theory of the life-cycle with some of these more difficult and intangible aspects of the social reality. The mathematical model of Chapter 3 was indicative of the range of possibilities that arise in the simple determination of fertility, and though some ground was covered, much remains to be done in order to incorporate the institutional issues mentioned therein.
The study then moved on to empirical tests of the role of institutional variables, and it has was interesting to find that many of the effects hypothesized in the literature do have some plausibility. This is consistent with life-cycle optimization with bequests by primarily ‘selfish’ individuals, as the simple model presented in Chapter 3 illustrates, and does not require altruism to explain most of the observed relationships.
Does this mean that we are now in a better position to predict the levels and direction of future fertility changes? The answer would be a qualified yes and no. We now can say with more certainty that macro-economic and institutional variables have strong effects on fertility. If the developing world sees strong economic growth including the provision of alternative markets for savings, and reduces infant mortality, rapid declines in fertility would be possible in most societies. If it is a goal of the policy maker to reduce fertility in a society, the bulk of the focus should be on the demand side
of the fertility equation. Once the demand for reduced the number of children builds up, it does not appear that the supply of contraception would form a major hindrance or barrier to the fulfillment of this reduced demand, since there were and continue to be many non-modern methods of contraception in most societies. Indeed, to be more precise, without the build-up of an appropriate demand, the mere act of supplying modern contraceptive devices would prove entirely meaningless
(as indeed has been found in the case of Pakistan, for example, where family planning workers have been known to dump condoms supplied from the ‘headquarters’ into the rivers
At the same time, we have much more to learn. We are only barely beginning to understand the magnitude and direction of the impact of these institutions. Therefore, the goal of achieving accurate prediction of the magnitude of changes in fertility in the world, while definitely nearer to us in terms of being fulfilled than it was a few decades ago, remains distant even now. Changes in national fertility still continue to take most astute observers and demographers by surprise despite the many millions of dollars of research that has gone into this area since the past four or five decades. This calls for greater multi-disciplinary cooperation in the study of this area through the involvement of economists, demographers, sociologists, anthropologists and even psychologists.
6.2 Limitations and recommendations
It is clear by now that not only are there many theoretical issues that need further study, but there are many limitations to the empirical part of this study because of the nature of the data set and the difficulty in creating some needed variables, apart from some definitional and measurement issues. If we are to learn more about the ‘mystery’ of the formation of fertility desires, as a first step, a panel data set needs to be created, spread over at least 15 years with measurements taken every 5 years, commencing from the time of marriage of the couples. It is true that we had observations on the young and old populations, but since the crux of the study of formation of fertility desires is the study of the formation of expectations, these have to be traced out carefully, over time, to ensure that we are capturing exactly how these desires are formed, changed and impacted by institutional and other factors. In an essentially cross-sectional study of this nature, the exact range and impact of security concerns does not come out in an ideal fashion. Given that there will be attrition in the panel, and in order to minimize possible loss of statistical significance, the number of observations must be increased substantially from that of the current study.
The measurement of desired fertility remains very problematic and the questionnaire must be redesigned to eliminate as many potential problems as possible. One of the most important findings of this study was that husbands need to be studied with equal care as the wives, since they play such an important role in the determination of completed fertility. Hence all husbands should be surveyed in complete detail, and if husbands are missing, a clear effort should be made to determine why they are missing and how the environment of the woman changes in that process. It is important also to trace the actual marriage patterns of the household, to isolate the decisions taken in the first, second, and additional marriages, if any, of either member of the couple. It would be important also to elucidate information on the level of communication and solidarity within a household, and its relative power structure.
Expectational variables are very difficult to rely upon as being formed exogeneously. A careful redesign of the survey would help minimize possible measurement errors. Another difficult set of variable which needs more care in its measurement is household income and the income of the various members of the household. So also the issue of labor force participation needs very careful analysis, since so many of the components of this seem to vary at very short duration.
At the community level, the effects of alternative assets markets, including insurance markets and the nature of the insurance contracts available, need more careful measurement. A separate measurement and analysis of social norms would be very useful, in order to arrive at an understanding of the actual components of the norms set. For example, it would be important to know the quantitative nature of the ‘norm’ of exchange of bequests and parental services in the society.
Despite these and many possible limitations this study is perhaps one of the very few that has examined the formation of demand for children at the micro-level and has also thrown light on the possible relationships of fertility with labor force participation, savings, marital solidarity, and bequests.
6.3 Policy implications
If the results of the study are anything to go by, there are some significant implications that invite the attention of policy makers, particularly those who are searching for ways to reduce the growth of population in their countries. While it is always difficult to generalize across countries from studies in a single country, some regularities seem to emerge which might possibly hold across many nations.
It appears that the key ‘side’ of the demand-supply equation for children is the demand side.
To the extent possible, the policy maker interested in reducing fertility has thus to try to decrease the demand for children first.This demand is crucially dependent on three things: the growth of the economy and the consequent increase in the likelihood of a higher level of future economic returns and insurance from children, the likelihood of survival of children into the future, and the reliability of social norms which underly the upholding of the implicit intergenerational contract.
The first two are relatively easily impacted by the policy maker, while the third one is more problematic.
It would obviously be a good idea to focus on policies that increase the likelihood of a child’s survival and future income. At the same time, it might be a good idea to slow down the onset of governmental social security programs in rural areas which might give a false sense of security about the possibility of government stepping in to support the elderly in their old age
Such programs can only speed up the breakdown of the extant, efficient, social norms for care of the elderly, and it is very difficult to find any example of economically efficient social security programs across the world. Most programs are ill-managed and are a drain on their economy.
As social norms break down, we have observed that there is likely to be an inverted U-shaped effect on fertility. At the very beginning, when norms are strong, favorable conditions will induce people to have fewer children and to focus on quality. Sometime later, as the norms start breaking down, people will want more children hoping that they can use strategic methods to enforce return. But as the breakdown of norms declines even more, and the reliability of children becomes lower than that of alternative insurers, one would find that OAS motives lead to a notable decline in the demand for children since they are no longer good serve any major economic purpose. Thailand seems to be distant from the second stage of increasing fertility, but if social norms were to completely break down, then fertility might show a blip in the positive direction before declining again.
Once a declining demand for children has set into place, it must be supported by a provision of access to family planning services. For societies that have very poor family planning services, it might be a good idea to give the supply of contraception a boost in order to fulfil the felt need of contraception. Thailand has had a very successful family planning program (see, for example, World Bank, 1993:18). It is a good idea to keep it going.
One of the findings of this study is that it is perhaps ill-advised to focus resources on adult education unless one is looking for possible short-term effects on economic growth. Parent’s education does not appear to have much bearing on the demand for children. People are generally good at rational decision making irrespective of the level of formal education they have had. This vindicates the statements cited from Knodel (1984) in the Box at the beginning of this paper.
A key policy implication of this study is that it is very important to increase access of rural people to savings institutions. To the extent that this is possible, in a socially efficient manner, it is very likely that fertility will decline. On labor force participation also, it was seen that more women will begin to work in non-farm activities as their education and opportunity increases. That would have further positive effects on economic development in general, as well as cause a ‘virtuous’ spiral upwards into better living conditions and reduced fertility.
Sadik (1991: 110) pointed out that “As yet … there is little understanding of how social, economic, political and cultural structures and institutions interact to produce the motivation for reducing fertility.” The current study has shed considerable light on some of these effects, but many questions remain unanswered, partly because some were not raised earlier. The old age security theory (OAST) is much better supported by data than alternative and often simplistic, world views. These findings are seen to be generally robust to alternative measures and specifications. We can therefore say with some confidence that there is indeed substitution between children, capital assets and insurance under the incomplete markets conditions of rural Thailand.
It is clear from the tremendous significance of the OAS effects that their neglect by theorists has cost economic demography much in terms of the richness of analysis and predictive power. But before the OAS theory of Leibenstein and others takes its rightful place as the major theory of fertility, these findings need to be replicated in other societies, using even more sophisticated and comprehensive surveys. One possible fruitful direction of further research would be to create a panel of a similar nature to this data set over time. The time has come to take an institutional and comprehensive approach to fertility. Issues such as bounded rationality, the role of information, the role of expectations and strategies, and the role of power distribution in decision-making are intangible issues not readily susceptible to formal modeling. These need to be understood in a better way in order to increase the probability of predicting the path of fertility in a given society. Demographers (particularly economic demographers) will now have to look at macroeconomic policies and social institutions in a much more holistic way.
In the meanwhile, the key message of this study appears to be that governments in traditional societies should focus on economic growth and education of children (apart from reducing infant mortality and increasing access to family planning). There is a strong synergy in the pursuit of wealth and education on the one hand and the successful ‘management’ of population growth on the other.
 This has been documented by Afaque Ahmed, who worked in the Pakistan government, and who was a Fellow of the Population Reference Bureau, 1997-98.  The Planned Parenthood Association of Zambia reported that “Ninety percent of Zambians have acquired family planning knowledge, while only 26 percent are practicing family planning methods” (as reported in Popline, vol. 20, Nov-Dec 1998). This further confirms that unless the demand factors for reduced fertility come into play, it is virtually impossible to achieve significant declines in fertility merely by expanding the supply side.  The five recommendations for this purpose in World Bank (1994:67), with which one would generally agree, are (1) avoid biases against traditional agriculture, (2) improve communication between rural and urban areas, (3) consider special programs for widows, (4) try to have formal systems complement informal ones, and (5) go slow on formal systems.  The fact that public pensions could “crowd-out” private care by children has some empirical support from Cox and Jiminez (1992). They find in Peru “that private transfers from the young to old would have been nearly 20 percent higher without social security benefits.”  Changes in expectations from children are definitely taking place in developing countries. For example, Hermalin (1995:13) states, “In Taiwan fertility surveys regularly include a question on whether respondents expected to be supported by their sons in old age. The sharp declines in positive responses to this question that occurred between 1973 and 1986 (from 51 to 19 percent) underscore the ongoing transformation in expectations” (data from Chang and Ofstedal, 1991). At the same time, there is possibly a substitution at work in terms of greater financial support of parents who do not co-reside with children. The Program on Population (1998) states, “Data from Taiwan suggest that increasing financial support may be partially substituting for declining co-residence. The proportion of married women age 20-39 reporting that they regularly give money to their husband’s parents went up from 32 percent in 1973 to 42 percent in 1986.” Similar effects might be taking place in Thailand.