Saving, dependency and development

Allen C. Kelley, Robert M. Schmidt

This study examines the impact of dependency on savings between 65 less developed countries (LDCs) and 23 developed countries over time and cross-sectionally since 1960. The study tests a modified Leff model and the Mason life-cycle framework. Empirical estimates address potential simultaneity between savings and output growth. The price indices of Summers and Heston are used because each country's national accounts are converted from nominal into purchasing-power variables. This eliminates the problems with using exchange rates which vary systematically by level of development with a "true" index of purchasing power. Savings (S/Y) is the percentage share of gross national saving in gross domestic product. Ygr is the growth of per capita income. Y/N gr is the growth in the per capita gross domestic product. Analysis is based on ordinary least squares (OLS) and two-stage least squares techniques, treatment for heteroscedascity, aggregation periods, several definitions of savings, different country samples, and aged dependency and youth dependency. Findings support the Mason variable-growth life-cycle framework that shows that changes in demographic factors accounted for a large part of savings. The relationships in the modified Leff-type model were weak, with the exception of the mildly negative youth and elderly dependency impact in the 1980s. The rate of growth of youth dependency was negative and significant in all cross-sections for the full sample, all panel estimates for both LDCs and the full sample, and in the 1980s for LDCs. In the OLS model, life-cycle effects were weaker, but direct dependency effects were stronger. S/Y over time became slightly more sensitive to changes in life cycle impacts but less sensitive to youth dependency. Demography's impact on savings over time is attributed to the increase in the pace of youth dependency decline and secondarily to its increasing sensitivity to life-cycle effects.