WOMEN PREFER LARGER GOVERNMENTS: GROWTH, STRUCTURAL TRANSFORMATION, AND GOVERNMENT SIZE

Abstract

I. INTRODUCTION In developing and developed economies alike, growth of income per capita is generally accompanied by a rise in the share of government expenditure in gross domestic product (GDP), as well as an increase in the participation of women in the labor force. We argue that these phenomena are causally related, propose a model that motivates that relation, and provide evidence in its support. The share of total government spending in GDP has increased by 10% in Organization for Economic Cooperation and Development (OECD) countries, in the last two decades alone (see World Bank 2001), and this rise comes in the wake of significant secular increases in government spending in the 20th century. (1) Apace with this rise in the government's share of GDP, female labor force participation has also risen from 28% to 41% in the same countries and the same period. (2) As for the long-run, Goldin (1995) reports an increase in female labor force participation in the United States between 1900 and 1980, from 3.1% to 50% of the labor force, (3) while according to Tanzi and Schuknecht (2000) the share of government spending rose in the same period from 8% to 30%. Figure 1 plots the change in government size as a share of GDP versus the change in female labor force participation from 1972 to 1999 for a sample of selected countries. The twin increases in female labor force participation and government size are not a coincidence; they are causally related. The mechanism at work is twofold; as wages rise with development and women choose to work more hours in the market, they increase their demand for services that tend to be provided by the government, including education, healthcare, care for the elderly and the unemployed. The additional provision of government services further encourages female labor market participation as it decreases the time necessary to perform those home and family activities whose burden falls disproportionately on women. (4) This circle of causation leads to higher female participation and larger governments. Our growth model extends Galor and Weil (1996). In addition to female participation in market activities, fertility, and output, government size is also endogenously determined. The rise in hourly wages that accompanies development and capital accumulation leads to an increase in female labor force participation as the opportunity cost of caring for the children increases. In our model, government spending decreases the time cost of performing household chores. One obvious connection is the provision of extra school hours and child care, which decrease the time cost of child rearing. As couples--women in particular--participate more in market activities, they choose higher government spending to compensate for time spent in the market. Under certain conditions, which we make explicit, they also choose higher tax rates, with the result that female labor force participation and government size increase apace. (5) As female labor force participation leads to a larger government, the total cost of raising children is reduced for two reasons: the total time at home decreases, and the time cost per child also falls as public spending partly substitutes for home production of child care, education, and other activities. (6) The decision to work outside the home is likely to lead women to demand wider and better provision of the public services that alleviate their unequal burden as caregiver at home. In addition, the role of women as insurer of family risks may lead to a demand for more government intervention as women become less present at home. There is an additional mechanism through the composition of the public labor force. Government jobs in teaching and health--which are a large portion of total public jobs--tend to be filled mostly by women. (7) Our model draws on the solid literature on growth, fertility, and female labor force participation, which is supported by microevidence. …