Journal Article A View of the Aggregation Problem Get access M. H. Peston M. H. Peston London Search for other works by this author on: Oxford Academic Google Scholar The Review of Economic Studies, Volume 27, Issue 1, October 1959, Pages 58–64, https://doi.org/10.2307/2296051 Published: 01 October 1959

Early history, 374. — World War II and after, 377. — The Korean period, 382. — The present structure of the industry, 384. — Postwar price behavior, 388. — A supplement to antitrust policy? 390.

IN the current debate on the financing of education, estimates of the money benefits which school attendance confers on ex-students are occasionally invoked. In an effort to shed some further light on this subject, this note presents some crude and limited calculations on the relation between education and income. The approach to the estimation of life-time income chosen here is the so-called crosssectional one, which involves the analysis of incomes received by people of different ages and educational histories during a single year. Relevant data are available from the I950 Census of Population,' which tabulates total money incomes received in I949 by a 3 /3 per cent sample of the population aged I4 and over. Only males, irrespective of color, are considered in this paper.la The principal difficulty raised by this source of information 2 is that mean incomes are not stated; the table only gives the frequency distribution and the median for each educationage group. The median is clearly not the appropriate type of average for the present purpose; since the distributions are all positively skew, it is uniformly less than the (estimated) mean. Moreover the ratio of mean to median is larger at higher levels of education, indicating that the distribution is more unequal there than at lower levels. Hence a calculation based on medians would not even give the right proportions between life-time incomes for varying school attendance. It was consequently necessary to estimate the mean incomes. For every income group (irrespective of age and education) a representative income was selected by inspection of the income distribution.3 The figures used were as follows: