Rati Ram, Wagner's Hypothesis in Time-Series and Cross-Section Perspectives: Evidence from "Real" Data for 115 Countries, The Review of Economics and Statistics, Vol. 69, No. 2 (May, 1987), pp. 194-204

A critique of the analyses and conclusions reached by a study of the relationships between economic and demographic variables conducted by Hazledine and Moreland published in 1977 is presented. Inappropriate analyses of cross-sectional data from 82 countries resulted in parametric estimates that did not justify the conclusions reached by Hazledine and Moreland. Concerning diminished returns to scale suffered by less developed regions the author found on reexamination the scale estimates to differ only slightly across regions and also the parameter reflecting returns to scale with respect to labor and capital to produce erroneous estimates for developed countries. The reported significant positive relationship between infant mortality and birth rate is not justified for low income countries because the mortality variable had been dropped from 2 of the 3 fertility regressions for low income regions. The relationship specified therefore holds only for Asia. The large differences across regions in savings behavior is a questionable finding for several reasons including the inappropriate comparisons across regions of estimated coefficients drawn from differently specified equations. In addition various extrapolations and simulations were found to be useless because of instabilities in and poor predictive capability of the underlying estimates.

A study of the impact of government size on economic performance and growth is important. Theoretically, one point of view suggests that a larger government size is likely to be detrimental to efficiency and economic growth because, for example, (i) government operations are often conducted inefficiently, (ii) the regulatory process imposes excessive burdens and costs on the economic system, and (iii) many of government's fiscal and monetary policies tend to distort economic incentives and lower the productivity of the system. At the other extreme, one can identify some points of view that assign to the government a critical role in the process of economic development, and could argue that a larger government size is likely to be a more powerful engine of economic development. There are several arguments on which the latter point of view is based. These include, besides others, (i) role of the government in harmonizing conflicts between private and social interests, (ii) prevention of exploitation of the country by foreigners, and (iii) securing an increase in productive investment and providing a socially optimal direction for growth and development.

Comments on a study by Conger and Campbell which postulated a 6 equation dynamic model including a fertility equation, estimated by 2 stage least squares on the basis of aggregate data for the U.S. The authors present a table with 3 sets of estimates for each of the 6 structural equations specified in the earlier study; 2 stage least squares estimates obtained for 1946-70 and 1946-76 and the estimates obtained by Conger and Campbell are indicated. The authors report that they were unable to obtain the results reported by the latter; their estimates were quite different and did not seem to support the conclusions offered in the earlier study. Estimates for the longer period were substantially different from those for the shorter. The authors believe that the Conger Campbell data also do not support the Conger-Campbell conclusions. Each point is discussed, with reference to the table. The equations are used to evaluate fertility, female participation, infant mortality, income, education, and medical expenditure.

,_ The Rate of Imitation of a Capital-Embodied Process Innovation, Economica 44 (Feb. 1977), 63-69. Sahal, Devendra, Patterns of Technological Innovation (Reading, MA: Addison-Wesley, 1981). Saxonhouse, Gary R., Estimated Parameters as Dependent Variables, American Economic Review 66 (Mar. 1976), 178-183. Theil, Henri, On the Estimation of Relationships Involving Qualitative Variables, American Journal of Sociology 76 (July 1970), 103-154. , Principles of Econometrics (New York: Wiley & Sons, 1971). Walker, David, An Analysis of Financial and Structural Characteristics of Banks with Retail EFT Machines, Working Paper # 79-1, FDIC, 1979.

THE primary purpose of this work is to conduct an empirical investigation concerning the role of in consumption, and to test a few postulates of the wealth theories of consumption. The opportunity for such an investigation is provided by the recent publication by Kendrick (1976) of annual data on human and nonhuihan for the United States for the period 1929-69. Specifically, the paper (a) provides estimates of a consumption function in which a variable is included in addition to the income variable; (b) compares the responses of consumption expenditures to changes in human and nonhuman and tests empirically Friedman's hypothesis (1957, p. 17) regarding the effect on consumption of an increase in nonhuman relative to total wealth; and (c) throws some light on the stock adjustment' and habit persistence' postulates by reporting estimates for equations in which a lagged consumption term is added to the variables. The organization of the paper follows the aspects stated above. After a discussion of some methodological questions, we provide estimates of the parameters of consumption functions that include as a distinct variable in addition to income, and show that the variable has significant coefficients, and the coefficients look plausible. The income variable coefficients are substantial, and while being consistent with the life cycle' hypothesis are not necessarily inconsistent with Friedman's theory. Next, we discuss the separate effects of human and nonhuman on consumption. The evidence seems to favor Friedman's postulate that an increase in nonhuman wealth, relative to the total, increases consumption. Estimation of separate equations for (a) durable goods and (b) nondurables and services, by including a lagged consumption term in addition to the variables, suggests that the coefficient pattern can be interpreted as indicative of a stronger habit formation' effect for nondurables and services than for durables. A summarizing section concludes the paper.