This paper asks the question: "Given the objective of maximizing per capita income, are resources invested in population control liable to yield a rate of return which justifies their being used for this purpose rather than for conventional investment?" It is, of course, recognized that a population policy must have many other objectives besides increasing per capita income of a nation or region. For the sake of simplicity of analysis, however, this paper confines its discussion to an examination of population policy as a means of increasing per capita income in comparison with conventional investment. It should be pointed out at the onset that this paper does not attempt to answer the question raised above. Instead, it presents several approaches to answering this question and evaluates the relative merit of each approach. The simplest approach is the numerical one. According to this approach, one may start by looking at the per capita income as a ratio, that is, national income divided by the number of the nation's population. One may seek to raise this ratio by increasing its numerator - investing in physical and human capital - in order to increase the annual output. Or one may seek to increase the ratio by decreasing the denominator. Which method is better can, then, be decided by the relative economic effectiveness of the two methods. The economic effectiveness depends on which method raises the ratio more, given the same amount of resources spent. Professor Stepher Enke often relied on this method to state his position. In estimating the effect of investment on output, the above ratio method overlooks the so-called multiplier effect completely. It also mistakingly equates the number of adults to whom a contraceptive device is made available for a year with the annual number of births prevented. These are, however, minor flaws of the method. The basic flaw lies in the fact that calculation of this kind is over-sim plistic because it ignores so many of the economic and demographic interactions. For example, it excludes the effect of conventional investment of birth rate through the induced change in output. This kind of number's game. appears to have been played in vacuum. The second and more commonly used approach is the cost-benefit analysis of birth prevention. Its basic assumption is that a new born child in a typical less developed country consumes output more than he or she produces in his life time, that is, his birth decreases the nation's savings and, thereby, not only per capital income but also gross output. The basic flaw in this kind of cost-benefit analysis lies in the assumptions behind it. For example, the output foregone is usually assumed to be that of a marginal worker. Since the marginal product is always less than the average product except in case of an economic where an increasing return to scale exists, the benefit always exceeds the cost. If the impact of the birth prevention programme falls mainly on middle-class tamnies, however, the basic assumption of the calculation does not hold. If the middle-class children over their lifetime produce in total more than they consume, and if they provide the major part of the savings, the effect of the birth prevention programme would be to decrease savings per head and potentially decrease the rate of growth of the economy. Furthermore, if the assumption that the impact of the birth prevention programme falls on the marginal workers only does not hold, it may reduce the so-called "residual" contribution to growth. The estimates of the residual contribution to growth range from 50% of the growth of national income to 90%. In the literature of economic growth and development, the explanation for the residual factor, named for the difference between the total measured growth of inputs and outputs, has increasingly been sought in the improvement in human capital or the quality of labor rather than in technical changes. Accordingly, it is likely that most of residual contribution is provided by non-marginal workers. If the birth prevention programme affects mostly "non-marginal" births, it is quite possible that the programme may actually worsen the condition of the economy. In a more philosophical vein, one may ask whether avoiding consumption is really to be construed as a benefit to an economy. After all, consumption is usually viewed as the main purpose of economic behavior. Furthermore, many families may consider children as consumption goods, if such a way of looking at the problem is permissible. If a household prefers an additional child to the additional income per household member, there is a gain for the household in question. This brings us to the whole question of private us. Social benefit and cost. The above discussion indicates that the two approaches have serious limitations, if they are to be used to analyze whether an investment in population control yields a rate of return which justifies their being used for this purpose rather than for conventional investment. A sound approach must take account of economic and demographic interactions which have bearings on the relationship between population growth and economic growth. In other words, the vacuum mentioned above must be filled. The third approach attempts to do this by basing their analysis on the results of statistical studies of the key parameters which specify the relationship between population growth and economic growth. One of these studies indicates that an investment necessary to reduce population growth by X percentage point has an effect on the economic growth of equivalent magnitude as an increase in the investment of Y percentage points as a fraction of GNP. Another study shows how much increase in labor productivity is required to offset the unfavorable effects of population growth on per capita income. Studies of this kind provide policy makers with a basis for evaluating an investment in population control as a means of increasing per capita income in comparison with that in physical and human capital. An example of the above studies is that conducted by p. Sommers and D. suits. In their study, a three-equation model of economic growth was built and the model was tested with the data covering a cross section of 100 countries compiled and published by the United Nations. This three-equation model has the advantage of forming a system whose solution is given simultaneously. Thus, the chain of causes and effects is determined simultaneously. For example, investment is expressed as a function of per capita GNP, and investment in turn is expressed as the factor which, together with population growth, determines the rate of growth of per capita income. The rate of population growth is expressed as a function of per capita GNP and the rate of population growth is in tum expressed as the factor which, together with investment, determines the growth rate of per capita GNP. According to this study in order to offset the effects of a rise of 1 percentage point in the rate of population growth, an increase of about 1.7 percentage points in the fraction of GNP invested is needed. Sommers and Suits' three equation model is simplistic. For example, expressing the rate of capita formation and also that of population growth as a function of per capita income alsone is oversimplification. Simplistic as it is, however, the model focusses on the key variables relating population growth to economic growth. Therefore, it provides a basis for a more refined evaluation of the relative effective-ness of population control as a means of increasing per capita income. Another simple model was used to conduct a statistical study of the same kind by Enterline and Stewart. The model used was John H. Power's full employment growth model. The results of the statistical study was used to examine what is required to offset unfavorable effect on per capita income of population growth. Instead of expressing the offsetting requirements in terms of investment out-lays, this study expresses it in terms of the required increase in labor productivity which is, of course, determined by investment in physical and human capital. For example, an economy with a capital-output ratio of 3 to 1, the study estimates that an increase of 4.6 percent in output per worker is required to maintain per capita income, if life expectancy is to increase from 50 to 60 years with relatively constant fertility rate. Population growth is expressed here in terms of increasing life expectancy with constant fertility. It can easily be expressed in terms of an increase in population which results from a change in life expectancy and or fertility rate. The magnitude of the offsetting increase in labor productivity required depends on the capital-labor ratio of the economy in question. The fourth approach is to take a comprehensive look at the complex economic and demographic interaction. One way to do this is to build a simulation model which includes all recognized economic and demographic interactions involving population growth. Another way is to build a simple format which theorize the results of various changes in the key economic and demographic variables. One may base his evaluation of the relative effectiveness of a population policy as a means of increasing per capita income on such a format by inserting numerical values into key variables in the format. An example is Kuznets' format which explicity includes. age composition, dependency problem and capital-output ratio as well as investment components in the parameters through which population growth affects per capitc income. The Kuznets, formet suggests that raising the rate of population growth from 1.3 to 3.0 percent per year can presumably be accommodated by a reduction of about a seventh in consumption per unit; and that a few more percentage points taken from ultimate con-sumption would permit a much higher rate of growth of per capita product. Note that analysis following Kuznets' calculation follows the traditional lines of the economic discipline - even if in a crude form (using capital output rations and simple assumptions concerning labor inputs, rather than linear production functions, whose use, however, would not change the results significantly). The results could easily be modified within a limited range either by raising or lowering the fractions or consumption that would have to be foregone as a result of accelerated population growth. As Kuznets states, the major weakness of using his format lies in the assumption that physical capital is the sole agent of increase in per capita product, and that the input of labor is proportional to numbers in the labor force. As we pointed out earlier. human capital is an important element in increasing output per man-hour and, therefore, per capita income. The pressure of rapid population growth is expected to fall mainly on educational and health resources, and the improvement of human capital depends mainly on these resources. In conclusion, this paper has examined various approaches to analyzing the question of whether resources invested in population control are liable to yield a rate of return to increasing per capita income which justifies their being used for this purpose rather than for conventional investment. It has been shown that there are many pitfalls in each approach and some erroneous assumptions have been made in some approaches. Therefore, it is advised that extreme caution should be employed in using any of the several approaches evaluated above. In evaluating the relative merit of each approach, one thing merits attention. It is that any method of analyzing the relative effectiveness of population control to raise per capita income should include economic and demographic interactions and avoid relying on numerical exercises in vacuum. Discussant Lee, Hyong Joon, Professor of Economics, Korea University The paper makes contribution in pointing out the relative strengths and weaknesses of the commonly used methods of analyzing the effectiveness of population policy. It is regrettable, however, that the paper does not go beyond evaluating the various methods. The paper could have applied some of these methods of analyze the current population policy of Korea and thus, have provided valuable information for policy makers. It should also be pointed out that a population policy designed solely for increasing per capita income lacks relevance for Korea. In the Fifties, Korea has had a low rate of economic growth and using population control as means of increasing per capita income was a viable option for policy makers. Recently, however, Korea has had an average growth rate of about nine percent per annum for some time. In a such rapidly growing economy, it is unlikely that the investment in population control could yield a rate or return comprable to that achieved by the inve