The unintended consequences of education policies on South African participation and unemployment

Literature › Works

Rulof Burger; Servaas van der Berg; Dieter von Fintel (2012). The unintended consequences of education policies on South African participation and unemployment. Working paper 289.

Type	working paper
Scope	Post-1994 South African economy
Method	econometric
Period studied	1995–2007
Regions	South Africa, Eastern Cape, Limpopo, Free State, Mpumalanga, Western Cape
Themes	labour markets wages, inequality poverty unemployment, education human capital, transition post apartheid, methods data sources
Citation key	burger2012

Argument

The paper argues that late-1990s Department of Education policies pushed hundreds of thousands of over-age learners out of school and into the labour market. These were the exclusion of learners more than two years above grade-age and a limit of one repeat per schooling phase. This helps explain the sudden 1996-2002 jump in labour force participation and the rise in broad unemployment, which was a supply-side shock rather than a collapse in job creation. It uses administrative enrolment data and household surveys, simulates policy scenarios with a dynamic schooling progression model, and runs an age-period-cohort decomposition of participation. The authors conclude that the unemployment increase reveals disguised unemployment that already existed in the mid-1990s, not a deterioration in labour market conditions.

Evidence and method

Department of Education administrative enrolment data (via Kraak 2008); Stats SA October Household Surveys (1995-1999) and Labour Force Surveys (2000-2007); matric candidate numbers; provincial and national policy documents (1995 White Paper, 1998 admissions notice, Western Cape circular, 2000 EFA assessment, OECD 2009 review); and the secondary literature on participation and cohorts.

Findings

• The policies coincided with a large fall in enrolment, estimated at 400,000 to 900,000. (p. 2)

  these policies coincided with a decrease in school enrolment of at least 400,000 and possibly as many as 900,000 learners.

• Labour force growth outpaced job creation, so unemployment rose even though employment grew. (p. 4)

  job creation could not keep up with the 4.6% per year growth in the labour force.

• Unemployment within a given birth cohort was stable, so rising unemployment is tied to new entrants. (p. 6)

  the unemployment rate for any particular birth cohort remained remarkably stable over this period.

• Provincial directives barred admission of learners more than two years over grade-age. (p. 9)

  schools in the Western Cape were to deny admission to learners who were more than two years older than the appropriate grade-age

• Matric candidate numbers fell sharply, consistent with the exit of over-age learners. (p. 9)

  the number of matric candidates decreased by fully 20% between 1998 and 2003

• Departmental data show a fall in total school enrolment between 1998 and 2003. (p. 10)

  the number of pupils in school (Grades 1 to 12) decreased by almost 425,000 between 1998 and 2003

• Household surveys show a much larger fall in the number of over-age learners in school. (p. 10)

  the number of over-aged learners decreased by more than 865,000 between 1999 and 2003.

• The simulation of observed outcomes shows enrolment falling by roughly 830,000 students. (p. 13)

  total school enrolment decreased from 38% of 15-30 year-olds to less than 32% in 2001, which amounts to approximately 830,000 fewer students.

• Age restrictions, not faster promotion, drove most of the enrolment drop. (p. 14)

  demonstrates that age restrictions played a much larger role in explaining the sudden drop in learners.

• The shift of over-age learners out of studying accounts for about half of the rise in participation. (p. 15)

  This constitutes approximately one-half of the total increase in the participation rate over this period

• Over-age students who left school were pushed into unemployment rather than other forms of inactivity. (p. 17)

  It is therefore likely that instead of making the transition out of education into other forms of inactivity

• The policies exposed hidden unemployment previously absorbed by schools. (p. 17)

  a large group of the unemployed were no longer allowed to hide in the school system.