Abuses of socioeconomic and demographic data

In a recently published article, Ekanem (1972) presents the data, used for a crosssection study of the effects of economic development on fertility. The basic data are for crude biirth rate, illiteracy, labor force in agriculture, urbanization, infant mortality, and per capita gross national product in 32 less developed countries for two points in time: the 1950 decade and the 1960 decade. In a subsequent note, Janowitz (1973) uses Ekanem's data to study the relation between changes in the indices of socioeconom.ic development and changes in fertility between the two points in time and questions the validity of the results of crosssection studies for predicting trends in birth rates. However, upon a, casual examination of Ekanem's data, one observes several implausible figures. For example, there is a change from, seven percent to 84 percent for the labor force in agriculture between the 1950 and 1960 decades in Madagascar. The Moroccan infant mortality rate is chara.cterized as low as 54.2 in 1935 and 46.7 in 1963, and there are illiteracy rates of 16.9 and 11.8 percent for the two points in time reported for Honduras. Reported data divergent from expected values led us to check Ekanem's data with the data in the sources cited in Ekanem (1972). Surprisingly enough, we discovered a host of serious mistakes in the reported data. Conclusions based on such defective data are suspect. Admittedly, the available data on indices of socioeconomic developmnent and fertility in the less developed countries are fraught with problems of unreliability, inaccuracy, and incomparability. Ekanem quite adequately makes note of all these problems and warns his reader to keep in mind the limitations of the data when interpreting the results of his study. However, we want to point out in this note the failure on the part of the users to exercise the necessary ca.utions in the selection, collection, and evaluation of the available data, not the inherent defects of such data. Since the mistakes we have discovered in the da.ta are too numerous to be described individually, we will simply group them into three broad categories of (a) inconsistency in the selection of the dates and the bases of estimation of the data (e.g., whether crude birth rate is based on registered birth data. or is estimated from analyses of population censuses and surveys); (b) incomparability in age groups and populations, and (c) clerical mistakes of transcription. These categories will be referred to as we describe below some of the most important mistakes for each of the variables of socioeconomic development and fertility that we were able to cross-check. The following procedure was used for checking the data reported by Ekanem (1972). In all cases the sources of data we used were those cited by Ekanem. Using the standa.rd cited sources we attempted to collect data for the specific years indicated by Ekanem. Where there was a discrepancy between Ekanem's data. and the correct data, an effort was made to ascertain the inature

Geography · Population · Socioeconomic status · Socioeconomics · Sociology · Demography · Gender, Labor, and Family Dynamics · Income, Poverty, and Inequality · Poverty, Education, and Child Welfare

• Consanguinity in Spain, 1911–1943

  J Pinto-Cisternas, J Pinto‐Cisternas et al.•Social Biology•1979

  Incluir item

• A further note on the relation between economic development and fertility

  Open Access•Ita I Ekanem•Demography•1972

  Incluir item

• Cross-Section Studies As Predictors of Trends in Birth Rates

  Open Access•Barbara S Janowitz, B Janowitz•Demography•1973

  Incluir item

• The Study of Population. An Inventory and Appraisal

  Open Access•Philip M Hauser, Otis Dudley Duncan•Population•1959

  Incluir item