1. Gender and the Demographic Dividend Karen Oppenheim Mason East-West Center

2. N ational T ransfer A ccounts 2 Outline 1. What is the demographic dividend? 2. What role might gender play in the demographic dividend? 3. What is the evidence about gender and the demographic dividend? 4. Report on ICRW Study

3. 1. What is the demographic dividend?

4. N ational T ransfer A ccounts 4 The demographic dividend ► Refers to the economic growth that results when fertility declines ► Fertility decline temporarily reduces the dependency ratio ► If the “extra” income is invested wisely, per capita income will grow

5. N ational T ransfer A ccounts 5 FemaleMale India 2005 Source: United Nations Population Division 2006.

6. N ational T ransfer A ccounts 6 S. Korea 2005 FemaleMale Source: United Nations Population Division 2006.

7. N ational T ransfer A ccounts 7

8. 8 What are wise investments? ► Physical capital: savings accounts, equity markets, etc. ► “Human capital:” education, health care, nutrition ► Both government & household invest-ments will work ► But whether investment occurs depends on policies

9. N ational T ransfer A ccounts 9 Policies that encourage investment ► Economic and political stability ► Reliable banking and investment systems ► Openness to trade ► Incentives to invest in human capital (health, education) ► Type of pension system (asset-based versus pay-as-you-go)

10. 2. What role might gender play in the demographic dividend?

11. N ational T ransfer A ccounts 11 Gender systems ► Refers to the socially prescribed roles & rights of women and men ► Culturally agreed on and enforced, so it’s a feature of groups ► Important for the DD are restrictions on women’s freedom of movement, ability to operate in public venues (e.g., schools, factories), and ability to make decisions within the household & family

12. N ational T ransfer A ccounts 12 Five possible effects of non- restrictive gender systems 1. Increased investment in female education 2. Rising female labor force participation rates 3. Increased investments in children’s human capital 4. Increased household savings or investment 5. Declining familial support of the aged (an incentive to save for old age)

13. N ational T ransfer A ccounts 13 1. Investment in female education ► Fertility decline reduces the value of girls’ domestic labor (fewer siblings to care for) ► Rising income & smaller family size also reduce parents’ need to ration their investments in children (a well-established finding) ► So parents are more willing to invest in girls’ schooling (unless gender norms seclude girls) ► Rising educational levels of both sexes encourage later ages at marriage ► Later age at marriage allows parents to reap the monetary rewards of investing in girls’ schooling ► So where girls’ freedom of movement is not severely restricted, “extra” income may be used to invest in female education

14. N ational T ransfer A ccounts 14 2. Rising female labor force participation rates ► Low fertility & increased life expectancy reduce the length of the child-rearing years  so the number of years women are available for labor force participation increases ► Better educated workers can command higher wages, so rise in female schooling makes their employment more attractive ► Once low fertility cohorts reach working age, the growth of the labor force slows, wages rise, and more women are attracted into the work force ► All these effects, however, depend on a gender system that does not severely restrict women’s freedom of movement

15. N ational T ransfer A ccounts 15 3. Investment in children’s human capital ► Studies from a variety of settings show that women are more likely than men to invest in children’s health, nutrition and schooling  Studies include Bangladesh, Brazil, Canada, Cote d’Ivoire, Ethiopia, Indonesia, South Africa, Taiwan, and the U.S. ► So rise in women’s employment may encourage greater investments in children’s human capital (daughters’ and sons) ► But this effect depends on whether women are able to determine how money is spent, i.e., on a gender system that gives wives some degree of economic autonomy or power

16. N ational T ransfer A ccounts 16 4. Increased household savings? ► Women face greater economic insecurity in old age than men do ► Gender bias in the ownership of land and other real estate undermines women’s ability to fund old age support from accumulated wealth ► So with rise in women’s employment, women may save for their old age more than men do ► Households may also see women’s earnings as “extra” income and may be more inclined to save or invest such income than an equal amount earned by the husband

17. N ational T ransfer A ccounts 17 5. Declining familial support of the aged ► Upward intergenerational transfers from family members or the state are thought to reduce the propensity to save for old age ► Studies find less intergenerational co-residence with rising female education and employment ► Daughters-in-law are less available to care for elderly parents when they are working ► So rising female employment may undermine upward transfers from family members (although increased income may help to fund such transfers)

18. 3. What is the evidence about gender and the demographic dividend?

19. N ational T ransfer A ccounts 19 General picture ► Much evidence that gender equality contributes to economic growth (see World Bank, Engendering Development) ► But little evidence on the extent to which gender is important in creating a demographic dividend ► Will review some of the general evidence

20. N ational T ransfer A ccounts 20 Closing the gender gap in schooling promotes economic growth 0 1 2 3 4 Sub-Saharan Africa South AsiaMiddle East/ North Africa Average annual growth in per capita GNP, 1960-1992 (percent) Actual Predicted

21. N ational T ransfer A ccounts 21 East Asia vs. the Middle East ► East Asia has enjoyed a very large demographic dividend ► The Middle East/North Africa region has not thus far, despite declining fertility ► Why? ► One reason may be the restrictive gender systems found in MENA vs. the less restrictive ones found in East Asia

22. N ational T ransfer A ccounts 22

23. N ational T ransfer A ccounts 23

24. N ational T ransfer A ccounts 24

25. N ational T ransfer A ccounts 25

26. N ational T ransfer A ccounts 26 “It is difficult to imagine East Asia’s labor- intensive, export-led industrialization occurring without the efforts of female workers, whose labor fueled the growth in manufacturing and helped to moderate wage growth.” “It is difficult to imagine East Asia’s labor- intensive, export-led industrialization occurring without the efforts of female workers, whose labor fueled the growth in manufacturing and helped to moderate wage growth.” John Bauer 2001, pp. 366-7

27. N ational T ransfer A ccounts 27 Needed research ► How do changes in age structure affect female education and employment? ► How does female employment affect household saving rates? ► How do changes in female education and employment affect familial support for the elderly?

28. 4. Report on an ICRW Study

29. N ational T ransfer A ccounts 29 Reference Based on a paper by Jeffrey Edmeades, Janna McDougall, Anju Malhotra & Margaret Greene, “Gender Equality and the Demographic Dividend,” presented at the PAA Meeting, New Orleans, April 17-19, 2008

30. N ational T ransfer A ccounts 30 Research Design ► Country-level growth regressions that build on the Bloom-Canning-Malaney (2000) model ► Hypothesis is that gender inequality in education (a) slows economic growth, and (b) reduces the positive effects of a favorable age structure on growth ► Reasoning is that inequality represents lost human capital and productivity

31. N ational T ransfer A ccounts 31 Data ► Data for 82 countries covering six five-year periods,1965-1999 (N = 470) – Ghana & S. Africa removed as “outliers” ► All data time-varying with independent variables measured in the base year for each period ► Dependent variable is the logged annual average percentage growth of real GDP per capita measured in PPPs

32. N ational T ransfer A ccounts 32 Data, continued ► Gender inequality is measured as the ratio of female to male years of completed secondary schooling in the population 15+ years of age (model also includes “total schooling” as stock measure) ► Inequality is treated as a dummy variable classification: 1. Female/male > 1.0 2..75 = or < Female/male = or <1.0 3. Female/male <.75 ► Reason for this choice is not discussed in the paper

33. N ational T ransfer A ccounts 33 Models ► Basic model predicts growth from age structure (log [15-64/N]), base-year GDP, tropics dummy, landlocked dummy, quality of institutions, openness of economy, total schooling, growth in total population, % of GDP from agriculture, and period dummies ► Second model adds the educational inequality dummies ► Third model adds the interaction between base-year GDP and age structure (I’ll ignore this model) ► Final model adds interaction terms for (a) gender inequality & age structure, (b) gender inequality & base-year GDP, and (c) gender inequality & pop growth ► They also run separate models within each of the three gender inequality groups

34. N ational T ransfer A ccounts 34 Selected Results 1

35. N ational T ransfer A ccounts 35 Selected Results 2

36. N ational T ransfer A ccounts 36 Questions ► What would results look like if there were fewer interactions in the models? ► How did the “outliers” affect the results – and why did removing only 12 observations out of 482 change the results substantially? ► Is gender inequality in education what is driving these results or is it some other correlated variable?

37. N ational T ransfer A ccounts 37 Acknowledgement Support for this project has been provided by the following institutions: ► the John D. and Catherine T. MacArthur Foundation; ► the National Institute on Aging: NIA, R37-AG025488 and NIA, R01-AG025247; ► the International Development Research Centre (IDRC); ► the United Nations Population Fund (UNFPA); ► the Academic Frontier Project for Private Universities: matching fund subsidy from MEXT (Ministry of Education, Culture, Sports, Science and Technology), 2006-10, granted to the Nihon University Population Research Institute.

38. The End