Family Planning and Fertility in South Africa under Apartheid - PowerPoint PPT Presentation

Family Planning and Fertility in South Africa under Apartheid Johannes Norling norling@umich.edu

National government expenditure on family planning (millions of 2012 Rand) 2

Total fertility rate 3

Contraception and fertility 4

Outline Politics and family planning in apartheid-era South Africa Difference-in-differences empirical strategy Results and comparison with other countries 5

Timeline National government involvement in family planning First grant to Begins to National National Population directly Family Family Development provide Planning Planning Programme services Programme Association 1963 1970 1974 1984 6

7

8

Timeline National government involvement in family planning First grant to Begins to National National Population directly Family Family Development provide Planning Planning Programme services Programme Association 1963 1970 1974 1984 1913 1948 1951 1959 late-1970s 1986 1994 First Group Areas Pass laws Bantustan Start of Pass laws Universal reserves Act & Native extended independence apartheid repealed suffrage established Laws Amd. to women (TVBC) Administrative partition into white areas and African homelands 9

Homeland departments of health “Control over health services, including family planning services, was a part of the plan for the independence for homelands. As various homelands declared themselves “territories,” they were granted self-governing rights within their borders, and responsibility for health services was duly handed over to them.” Department of Health, Annual Report , 1975 10

Per-capita expenditure on family planning: Homeland government as share of national government 11

Per-capita expenditure on family planning: Homeland government as share of national government 12

Per-capita expenditure on health: Homeland government as share of national government 13

Per-capita expenditure: Homeland government as share of national government 14

Difference-in-differences Event study First difference: Residents of white areas minus Residents of homelands Second difference: Given year minus Year before funding began Other causes of white area/homeland fertility differences LINK 15

Other causes of white area/homeland fertility differences 1960 1970 1980 1991 2.49 2.52 2.82 Per-capita income in white areas divided by homelands Ratio of men age 15 – 64 to women age 15 – 64 White areas 1.36 1.37 1.33 Homelands 0.55 0.62 0.69 Share of women age 15 – 64 who are married White areas 0.40 0.47 0.33 Homelands 0.33 0.45 0.37 Share of women age 15 – 64 who are working White areas 0.51 0.43 0.49 Homelands 0.32 0.16 0.14 Share of women age 15 – 64 employed in private residences White areas 0.27 0.19 0.25 Homelands 0.03 0.04 0.02 LINK 16

Birth history data sources 1974 Fertility Survey 1987 Demographic and Health Survey 1994 and 1995 October Household Surveys 2008 National Income Dynamics Study 17

18

19

20

Birth history data sources Birth information Sample Location 1974 Fertility Survey Children ever born Married women Residence ages 15-44 1987 Demographic and Complete birth Married women Birth and Health Survey history ages 15-49 residence 1994 and 1995 October Complete birth All women ages Birth and Household Surveys history 12-54 residence 2008 National Income Complete birth All women [Birth] and Dynamics Study history residence 21

Likelihood of giving birth 22

Difference-in-differences Event study First difference: Residents of white areas minus Residents of homelands Second difference: Given year minus Year before funding began One observation per black woman, i , per year, t • b = 1 if gave birth • L = 1 if live in a white area • 1( t = y ) = 1 if t = y for years y ≠1969 𝑐 𝑗𝑢 = 𝛽𝑀 + 𝛾 𝑧 1(𝑢 = 𝑧) + 𝜀 𝑧 𝑀 × 1(𝑢 = 𝑧) + 𝜁 𝑗𝑢 𝑧≠1969 𝑧≠1969 LINK 23

Likelihood of giving birth: difference-in-differences 24

Likelihood of giving birth: difference-in-differences 25

Likelihood of giving birth: difference-in-differences 26

Use of contraception among black women White areas Homelands 1974 32% 16% Ever 1982 used 1987-89 70% 55% 1974 24% 11% Currently using 1987-89 50% 40% 27

Use of contraception among black women White areas Homelands 1974 32% 16% Ever 1982 used 1987-89 70% 55% 1974 24% 11% Currently 1982 51% using 1987-89 50% 40% 28

Use of contraception among black women White areas Homelands 1974 32% 16% Ever 1982 used 1987-89 70% 55% 1974 24% 11% Currently 1982 51% using 1987-89 50% 40% Months used 1987-89 44 41 29

Children ever born 30

Children ever born Residents of white areas Residents of homelands Single difference: minus born in given cohort born in given cohort One observation per black woman age 40+, i , born in year group, t • c = Children ever born • L = 1 if live in a white area • 1( t = y ) = 1 if t = y for years y =1930- 34,…,1965 -69 1965−69 1965−69 𝑑 𝑗𝑢 = 𝛾 𝑧 1 𝑢 = 𝑧 + 𝜀 𝑧 𝑀 × 1 𝑢 = 𝑧 + 𝜁 𝑗𝑢 𝑧=1930−34 𝑧=1930−34 LINK 31

Children ever born: first difference 32

Timing of births: first difference 33

Outcomes in adulthood First difference: Children born in white areas minus Children born in homelands Second difference: Given year minus Year before funding began One observation per individual, i , born in five-year group, t , 1955- 59,…,1985 -89 • y = outcome in adulthood • L = 1 if live in a white area • 1( t = y ) = 1 if t = y for years y ≠1965 -69 • M = 1 if child is male 𝑐 𝑗𝑢 = 𝛽𝑀 + 𝛾 𝑧 1(𝑢 = 𝑧) + 𝜀 𝑧 𝑀 × 1(𝑢 = 𝑧) + 𝛿𝑁 + 𝜁 𝑗𝑢 𝑧≠1965−69 𝑧≠1965−69 LINK 34

Completed standards of education: diff-in-diff 35

Household income in 2008 (USD): diff-in-diff 36

Summary National government support of family planning responsible for • Up to three percentage point drop in the likelihood of giving birth • Delayed entrance into childbearing • 0.5 – 1 fewer children per woman • Higher income in adulthood for children born during the program 37

Births to black women in white areas 38

Does availability of contraception influence fertility? Absolute reduction Percent reduction Cost per in children born in children born birth averted Dates per woman per woman (2012 USD) 1970 – 1989 0.5 – 1 17 $3,761 South Africa Bangladesh 1978 – 1985 21 $384 (Simmons et al. 1991) Colombia 1964 – 1993 0.25 – 0.33 $124 – $167 5 (Miller 2009) Ethiopia 1990 – 2004 1 20 (Portner et al. 2011) Ghana 1993 – 1999 1 15 (Phillips et al. 2006) Indonesia (Gertler and Molyneaux 1994) 1982 – 1987 0.04 – 0.08 1 – 2 Iran 1967 – 2006 18 – 28 (Modrek and Ghobadi 2011) Peru 1985 – 1991 0.93 – 1.30 25 – 35 (Angeles et al. 2005) Tanzania 1970 – 1991 10.9 – 21.0 (Angeles et al. 1998) 39

Family Planning and Fertility in South Africa under Apartheid Johannes Norling norling@umich.edu

Comparison of white areas and homelands White areas Homelands Difference Number of women 33,800 21,918 28.436 26.547 1.889*** Average age (0.133) (0.081) (0.105) 0.555 0.158 0.397*** Share live in an urban area (0.005) (0.002) (0.004) 5.806 5.608 0.197*** Average standards of education (0.037) (0.022) (0.029) 0.391 0.348 0.043*** Share ever married (0.006) (0.003) (0.004) 0.546 0.499 0.047*** Share ever given birth (0.006) (0.004) (0.005) 21.117 21.157 -0.040 Age at first birth (0.081) (0.049) (0.064) 33.423 35.351 -1.928*** Age at most recent birth (0.203) (0.124) (0.161) Average number of births 2.602 2.913 -0.311*** (excluding zero births) (0.029) (0.019) (0.023) 41

Spacing between births One observation per child, i , born order j when mother was aged k • l = Months until next sibling is born • O = Order among mother’s children • K = Mother’s age at child’s birth 𝑚 𝑗𝑘𝑙 = 𝛽 𝑘 1 𝑃 = 𝑘 + 𝛾 𝑙 1 𝐵 = 𝑙 + 𝜁 𝑗𝑘𝑙 𝑘≠1 𝑙≠12 LINK 42

Months until next siblings is born 43

Months until next siblings is born 44