Modelling social indicators: some problems, prospects, and illustrations

Although the compilation of systematic socioeconomic indicators can be dated in centuries rather than in years, a veritable explosion of social monitoring activities has occurred the post-world war ii era. these data, which pertain to matters as diverse as presidential popularity and chronic physical ailments, provide the basis for developing a dynamic model of society. although several series, especially economic and demographic ones, can be studied over a substantial period of time, many of the key variables which would obviously enter into a model of the society such as educational attainment, church attendance, the occupational division of labor, and geographic mobility have been recorded on an annual basis in our society only for the past two or three decades. for the past three years we have been systematically exploring the expanding pool of annual social indicators in an effort to assess the feasibility of constructing a dynamic model of american society which complements existing economic models by virtue of its inclusion of social, demographic, and political indicators. in the course of our efforts, we have encountered numerous technical and methodological obstacles. our purpose in this essay isto discuss some of these difficulties and to provide substantive illustrations of them. to the best of our knowledge, all of the problems discussed herein are well known to econometricians; the main purpose of this essay is simply to pose and illustrate these difficulties in the context of sociopolitical, rather than purely economic indicators. in preparing this survey of some methodological issues encountered in the analysis of time series and the construction of structural equation models based upon them, we have attempted to reduce technical details to a bare minimum in order to make this essay accessible to the widest possible audience. more formal treatments are already available in most standard econometric texts; little would be served by duplicating those formalities herein. although this essay is more broadly conceived, we should also note that many of the technical problems broached herein have also been raised in the sociopolitical literature by russett (1971), in an essay which leans solely upon the analysis of defense expenditures for illustrative material. his essay can, in our judgement, be profitably read in conjunction with the present review.;