計量経済学的モデルによる地域的労働力人口変動の解析 : わが国およびアメリカ合衆国の地域的労働力人口の解析

概要

論文の詳細を見る
This paper is written as a further discussion of the previous papers in which the similar problems are treated [Keisuke Suzuki: "An Econometric Analysis of the Variation of Regional Population." The Journal of the University of Transportation Economics (Ryutsu Keizai Ronsyu), Vol.4, No.2, 1969, pp.69-80 (in Japanese). and Keisuke Suzuki: "The Variation of Regional Population in Japan." Journal of Regional Science, Vol.10, No.3, 1970, pp.335-51 (in English)]. As already mentioned in the previous papers, Klaassen has constructed a model for explanation of the variation of regional labor force and regional level of wage. The present writer has tried to construct some models of regional population based on the Klaassens model in the previous papers. In this paper, several models of regional labor force are constructed based on the Klaassen's model and are applied to the data of Japan and United States of America. The structural and reduced forms of the Klaassen's model are shown by equations (2.1)-(2.7). The structural forms of 5 models constructed in this study (Derivative Model I-V) are shown by equations (3.1)-(3.35), where, L: labor force of a region at time 0. L_<-1>: labor force of a region at time-1. L_s: labor force supplied of a region at time 0. L_D: labor force demanded of a region at time 0. L_<DN>: labor force demanded in non-basic activities in a region at time 0. L_<DB>: labor force demanded in basic activities in a region at time 0. Δ^nL: natural increase in labor force of a region during the period from-1 to 0. M: net migration during the period from -1 to 0 (inflow of labor force minus outflow of labor force). w: level of wage of a region at time 0 or -1. w^^-: level of wage of all region at time 0 or -1. A: index of the level of production in all activities(basic and non-basic activities) of a region at time 0 or -1. A^^-: index of the level of production in all activities of all regions at time 0 or -1. a: increase rate of the index of the level of production in all activities of a region during the period from -1 to 0. a^^-: increase rate of the index of the level of production in all activities of all regions uring the period from -1 to 0. α, β, γ, η, λ, and π: parameters The reduced forms of these structural forms of the models are shown by equations (3.46)-(3.55), where C_<10>, C_<11>, C_<12>, C_<13>, C_<20>, C_<21>, C_<22>, and C_<23> are parameters which are constructed by parameters contained in the structural forms. Before the models are applied to the actual data, the correlations between the number of persons employed in all sectors of industries L and the numbers of persons of some sectors of industries were observed to discuss the existence of the basic and nonbasic activities. The correlation coefficient r between the L and the total number of wholesalers and retailers L_<WR> was very high (r=0.972) as shown in Figure 1, which suggested that commercial activities may be regarded as typical non-basic activities. On the other hand, the correlation coefficient between the L and the numbers of persons employed in mining industry was very low as shown in Figure 2, which suggested that mining industry may be egarded as a typical basic activity. In order to apply the models to the actual data, the variables in the models are slightly modified. Consequently, the forms of the reduced form are also slightly changed. Firstly, L is replaced by l which is the increase rate of labor force of a region during the period from time -1 to 0. Secondly, (L_<-1>+Δ-nL) is replaced by pn which is the increase rate of population of a region during the period from -1 to 0. These replacements are easily done by supposing that every region has a same quantity of labor force at time -1 (for example, every region has 1,000 persons for labor force at time -1, as Klaassen did), and the ratio of labor force to the population in a region is stable during a short period of time. After these replacement, the modified reduced forms of the derivative models which are shown by equations (5.21)-(5.30) are ob- tained. These modified reduced forms are applied to the actual data. In these modified reduced forms, the value of the parameters with one asterisk (^*) are negative, and those with two asterisks (^<**>) can be negative or positive. All of the modified reduced forms are applied to the data of Japan by prefecture. When the reduced forms are applied, the increase rate of the number of persons employed in all sectors of industries during the period from 1960 to 1965 l is used as l in the modified reduced forms of the derivative models, the nominal wage per person at 1965 w_N, real wage per person at 1965 W_R, the index number of wage W_N, W_N (the wage w_N of all Japan is 100), and that of W_R, r_R (the wage w_R of all Japan is 100) are used as w, the natural increase rate of population during the period from 1960 to 1965 p is used as p^n, the amount of product per person at 1960, Y_<35>, that at 1965, Y_<40>, the index number of Y_<35>, y_<35> (the amount of product per person of all Japan at 1960 is 100), and that of Y_<40>, y_<40> (the amount of the product per person of all Japan at 1965 is 100) are used as A, and the increase rate of the amount of product per person during the period from -1 to 0, a is used as a. One of the derivative models, Derivative Model III, is applied to the data of U.S.A. by region. The increase rate of nonagricultural employment during the period from 1960 to 1965 is used as l, the average hourly earnings of manufacturing industries at 1965 as w, the natural increase rate of population during the period from 1960 to 1965 as p^n, and the increase rate of personal income during the same period as a. In the results of the application of the models (derivative models) to the data of Japan, we can find the fact that all the equations for l and w (equations (6.1)-(6.29)) are fitted to the data. High multiple correlation coefficients are obtained. Except a few exceptions, the parameters K_<21>'s which mustbe negative are not significantly different from 0. If the parameter γ is near 1, this fact may be obtained. And in the results of the application of the Derivative Model III to the data of U.S.A. (equations (7.1)-(7.2)), the equation for l is fitted to the data. owever, the equation for w is not fitted. The reason why the equation for w is not fitted may be that γ is near 1 and 0 is approximatly equal to (1-γ)π. According to these results, we may draw the following conclusion. First of all, the derivative models are approximately fitted to the data, especially the equations for l are completely fitted. Secondly, the increase rate of labor force l is influenced by the natural increase rate of population p^n (or that of labor force l^n), the level of activity of production A or the increase rate of the level of activity of production a. Lastly, the fact that l is influenced by A and a supports both the wage theory in migration of labor force proposed by J.R. Hicks and job opportunity theory proposed by T.W. Schultz and J. Robinson.

著者

鈴木啓祐
流通経済大学

計量経済学的モデルによる地域的労働力人口変動の解析 : わが国およびアメリカ合衆国の地域的労働力人口の解析

スポンサーリンク

概要

著者

関連論文

スポンサーリンク