Theory and Empiricism in the Study of Sociological Development: The Philippines

Ares, Julian Arcenyas, Jr.

     This study is a participation in the furor, loudest perhaps in the mid-1960s, over what 'social indicators' should mean and how they ought to be applied in national development studies. It attempts to criticize the various theories and empirical methods involved thus far in the furor. It appears that the dialectical opposites in the theoretical controversy may be identified now as the structural- functionalists at one extreme and the system theorists at the other. There are shades in between. The former argues for a treatment of societal development as resembling that of an organismic unfolding, an entelechical relationship between structure and function preordained. The latter, on the other hand, shuns the organismic line and posits a view of development as product of systemic interaction . . . among all parts of a system. Whether or not society may be likened to a biologic organism, seems to be the crux of the matter. At issue in the empirical field are the limits of aggregation, or disaggregation, of the measures -- or indicators -- of development and the choice of, and statistical methodology to use in evaluating, these measures. The aim is that of arriving at meaningful cross- spatial as well as cross-temporal comparisons. This study suggests a theoretical approach, akin to system theory, which may be called 'total systems-humanist' approach; and a statistical methodology, a variant of the UN- developed correspondence analysis, which may be termed 'relative positions' analysis. The total systems-humanist approach centers on the 'total man,' in terms of Sismondi's ideas, as focus of hypothesizing, and emphasizes man -- as opposed to state -- as the ultimate beneficiary of development. Correspondence analysis (as well as relative positions analysis), a multivariate method, operationalizes the system concept of society. Mathematically, it can admit any number of socioeconomic variables, from which it can identify the salient ones -- the core indicators -- or those that best explain the phenomenon of development.
     In this study a set of 231 variables was tried. The result of this reduction process is similar to that of factor analysis. Relative positions analysis, however, while using an almost identical reduction procedure, uses the median as datum for measuring cross-spatial variation -- while correspondence analysis uses the mean. The use of the median as datum magnifies inter-territorial differences and minimizes the effect of the outliers or extreme scores. These characteristics lend themselves to recognition of small inequalities and to more accurate estimation of missing scores -- as demonstrated in this study, compared to other methods. The above-mentioned theory and empirics were applied to the Philippine socioeconomic situations of 1960 and 1970 with very satisfactory results which support the following hypotheses: (1) that development is temporally unique for each territory as well as spatially unique vis-a-vis all others; (2) that the structure of the agriculture industry is largely responsible for the vast inequities of the quality of life among the different classes of Philippine society; (3) that population density relates positively to development in rural areas, but negatively to the same in urbanized ones; (4) that prime industry production is directly related to urbanization; (5) that urbanization enhances sectoral diversification, given the same population density; and (6) that spurts of economic growth exacerbate inter-class inequalities.