Estimating the Role of Social Reproduction in Economic Growth

Abstract

Do investments in social reproduction, or the time and commodities that it takes to produce and maintain the labor force, actually matter for the rate of economic growth? Using a Kaleckian macroeconomic model that incorporates gender and care provisioning, this article seeks to empirically evaluate this question. With panel data for a set of 121 countries between 1991 and 2015, the article uses principal component analysis to generate estimates of social reproduction regime by country, and then applies these estimates in growth regression analysis. Results indicate that the pressure on women’s care time that comes with their increasing labor-force participation—absent strong social and more gender-egalitarian supports for care provisioning—compromises investment and growth. In economies where those supports for social reproduction exist, the increasingly outward-oriented and market-driven macro structures and policies that prevail across a variety of countries, including those associated with financialization, are shown to constrain investment in human capacities and long-run productivity growth. In mutual social reproduction regimes, greater gender equality in the labor market and in the distribution of responsibilities for care also stimulates economic growth, while regimes built on the exploitation of women’s labor in these domains generate lower growth.

Keywords:

• Feminist economics
• structuralist macroeconomics
• social reproduction
• gender and growth
• gender and macroeconomics

JEL CLASSIFICATIONS:

• B5
• O4
• J1

Acknowledgments

We thank the organizers, supporters, and participants of the Care Work and the Economy Project (CWE), an initiative of the Program on Gender Analysis in Economics (PGAE) at American University. Generous financial support for this work was provided through PGAE by The William and Flora Hewlett Foundation. We benefited from important feedback at the CWE Berlin workshop, the 2019 Allied Social Science Association meetings, the 2019 Eastern Economic Association meetings, and at seminars at American University and the University of Massachusetts Boston, as well as the journal’s anonymous reviewers. All mistakes are ours.

Notes

1 For references to this work, see the following literature reviews: Kabeer and Natali (2013) and Seguino (2020).

2 Michał Kalecki (1971), a contemporary of Keynes, developed a model that investigates how the income distribution affects output, employment, and growth in demand-constrained economies. The academic literature using this theoretical approach has expanded over the last two decades into a rich literature that addresses dynamics in open as compared to closed economies, and has taken into consideration the role of economic structure.

3 For recent work on gender segregation internationally, see Borrowman and Klasen (2020). For a review of the efficiency argument for gender equality, including the literature on women’s spending on household needs, see Braunstein (2011).

4 For a more extensive discussion of these regimes in terms that are embedded in particular country cases, see Braunstein (2015).

5 The HDI’s education index is based on mean years of schooling for adults older than 25 years and expected years of schooling for children entering school: health by life expectancy at birth (UNDP 2013).

6 With observations from more than 80 countries from the United Nations Statistical Division, the correlation coefficient between the two is −0.52.

7 For more detail on this approach, including methods to deal with missing values, see Braunstein, Bouhia, and Seguino (2020).

8 This mean does not exactly equal zero because not all observations used in the score standardization were used in the regressions, and Table 6 includes only those observations included in the regressions.

9 Some additional econometric details: Country fixed effects are typically the default in panel analyses, but we use pooled OLS both because the long time period specification cannot accommodate country fixed effects, and because the variation we are interested in is differences across countries, not within countries over time. This is in line with the fact that there is little variation in the key variables of interest within countries over time, which includes just three time periods. On the question of multicollinearity, independent variables are correlated, but this is not a problem for the coefficient estimates as evidenced by examining variance inflation factors (VIFs), all of which average less than 5.

10 Human capital stock is highly correlated with social reproduction, but the presence or absence of secondary schooling in the regressions does not affect other coefficient estimates.

11 In the parlance of the Marglin and Bhaduri model (1990) that gives the structural inspiration for this analysis, this is the contrast between a stagnationist (care-led) regime and an exhilarationist (inequality-led) regime.

12 Running the regression with one weighted regime at a time gives the same results only with generally larger magnitudes and greater statistical significance on the regime coefficient estimates. We present the simultaneous estimates approach because it seems the most analytically straightforward.

13 Some research suggests that while wage squeeze can stimulate growth in the short run, longer run growth is hampered due to the negative effects on human capacities and productivity growth, although our data do not permit a long enough time series to explore that effect.

14 There are additional variables one could consider to measure financialization—for instance, the utilization of credit by the household sector to measure financialization. However, data availability limited our choices.

Additional information

Notes on contributors

Elissa Braunstein

Elissa Braunstein is a professor and chair of economics at Colorado State University and Editor of the journal Feminist Economics. She also worked as a senior economist and head of the Unit on Economic Cooperation and Integration Among Developing Countries at UNCTAD. Her research focuses on the international and macroeconomic aspects of development, with particular emphasis on growth, macro policies, social reproduction, and gender. She publishes widely in both academic and policy venues, and has done consulting work for a number of international development institutions, including the ILO, the World Bank, UNDP, and UN Women.

Stephanie Seguino

Stephanie Seguino is a professor of economics at the University of Vermont, and a research associate at the Political Economy Research Institute. Her research explores the relationship between intergroup inequality by class, race, and gender on the one hand and economic growth and development on the other. Seguino also studies the impact of globalization on income distribution and well-being, with a particular emphasis on Asian and Caribbean economies. She has been an advisor or consultant to numerous international organizations, including the World Bank, UNDP, ADB, and USAID, and publishes regularly in refereed journals, including World Development, Journal of Development Studies, and Feminist Economics.

Levi Altringer

Levi Altringer is a PhD candidate in economics at Colorado State University. His research interests include stratification economics, the political economy of race and gender, social reproduction, applied microeconomics, and public economics. His dissertation work seeks to better understand the role of care in economic outcomes.