Skip to content

Intergenerational Networks, Unemployment, and Persistent Inequality in South Africa

Literature › Works

Jeremy R. Magruder (2009). Intergenerational Networks, Unemployment, and Persistent Inequality in South Africa. University of California, Berkeley (Dept. of Agricultural and Resource Economics).
Type working paper
Scope Post-1994 South African economy
Method econometric
Period studied 2002–2005
Regions Cape Town, Western Cape
Themes labour markets wages, inequality poverty unemployment, transition post apartheid, methods data sources, demography urbanisation
Citation key magruder2009

Argument

Using the Cape Area Panel Study, Magruder tests whether parents act as job-search network connections for young adults in post-apartheid South Africa. He compares changes in children's employment with employment trends in the parent's baseline industry, using individual fixed effects. Daughters and children of absent fathers serve as control groups for wealth effects, giving a triple-difference design. Fathers who live in the province help sons, but not daughters, find jobs, and mothers do not appear useful. The result is robust to attrition, industry-specific human capital and correlated networks. The paper concludes that network-based job allocation reinforces unemployment and poverty traps across generations.

Evidence and method

Cape Area Panel Study (CAPS) waves 2002-2005, a panel of young adults aged 14-22 in Cape Town (analysis sample: black and coloured respondents, about 4,100 individuals). Provincial two-digit industry employment series built from Statistics South Africa September Labour Force Surveys 2002-2004. Descriptive figures from the 1993 October Household Survey and a 1996 manufacturers' survey (Standing, Sender and Weeks).

Findings

  • Only sons with present fathers are more likely to work when the father's industry grows. The effect is large. (p. 18)

    if employment increases by 10% in a father’s industry, then the son is 3% more likely to work if his father is nearby.

  • Sons with absent fathers and all daughters respond negatively, consistent with wealth effects on labour supply. This supports the control-group design. (p. 18)

    Sons whose fathers live away and all daughters are negatively e¤ected by increases in fathers’employment trends, consistent with the hypothesis of a negative wealth e¤ect on labor supply.

  • Post-apartheid mobility out of unemployment has deteriorated: long-term and never-employed shares among the unemployed rose between 1993 and 2003. (p. 10)

    Among the prime-aged broadly de…ned unemployed in 2003, 77% have not worked in at least 3 years, including 58% who have never held a job.

  • Aggregate unemployment context: very high broad unemployment in the early 2000s. (p. 2)

    42% of prime-aged adults were unemployed over the period 2002-2004

  • Young men with present fathers are far more likely to work in their fathers' industry. (p. 19)

    11.5% of young men with present fathers who ever work do so in their fathers’two-digit industry at least sometimes, which contrasts with 4.7% of young women with present fathers who ever work

  • The network effect is not explained by industry-specific human capital: employment trends in the father's historical industry do not matter. (p. 24)

    conditioning on log employment in the fathers’historical industry does not impact the parameters of interest, nor is it itself signi…cant.

  • Mothers do not serve as effective network connections for daughters, and the weak son effect vanishes once agricultural mothers are excluded. (p. 21)

    There is no discernible e¤ect on daughters of changes in mothers industries.

  • Back-of-envelope calculation: father-son networks explain a large share of the employment gap between sons of present, working fathers and other young adults. (p. 26)

    In the estimation sample, 20.7% of person-years are spent working; if networks increase working by .06 per year for those who can use them this would suggest a gain of 1/3.

  • Most young men lack the advantage of a present, working father. (p. 26)

    Since 55% of black and coloured young men in this region have absent, unemployed, or deceased fathers, this suggests that the majority of these young adults are at a large disadvantage.