Skip to content

Labour Market Policy and Poverty: Exploring the Macro-Micro Linkages of Minimum Wages and Wage Subsidies

Literature › Works

Karl Pauw (2009). Labour Market Policy and Poverty: Exploring the Macro-Micro Linkages of Minimum Wages and Wage Subsidies. PhD thesis, School of Economics, University of Cape Town. University of Cape Town.
Type thesis
Scope Post-1994 South African economy
Method mixed
Period studied 1995–2009
Regions South Africa
Themes labour markets wages, inequality poverty unemployment, methods data sources, macro fiscal finance debt, transition post apartheid, comparative theory
Citation key pauw2009

Argument

The thesis asks whether minimum wages and wage subsidies can reduce poverty in post-apartheid South Africa. It builds a sequential macro-micro framework: a CGE model (STAGE, 2000 SAM) linked to micro-incidence and micro-simulation models. Both policies reduce poverty only modestly. Minimum wages work better when the wage elasticity of labour demand is low and wage subsidies work better when it is high. Results are also sensitive to price effects, financing, productivity (efficiency-wage) assumptions and how job gains and losses are allocated among individuals.

Evidence and method

IES/LFS 2000 merged household and labour force survey, OHS/LFS 1995-2005, the PROVIDE 2000 SAM (22 sectors, 88 labour groups, 162 household groups), published wage-elasticity estimates (Fallon and Lucas, Fields et al.), minimum-wage sectoral determinations, and National Treasury wage-subsidy costings. Multinomial logit and earnings regressions are estimated on the IES/LFS data.

Findings

  • Poor households rely heavily on non-wage income, so wage-based policies must raise wages substantially to affect poverty. (p. 30)

    since the average poor household only derives about half its income from wages, wages would have to rise significantly to have an impact on poverty.

  • Unemployment rates are far higher in the bottom quintiles, and poverty and joblessness are strongly linked. (p. 28)

    The broad unemployment rate among poor people was 58.7 per cent in 2000 (the weighted average of 62.8 and 47 per cent), compared to only 19.8 per cent among non-poor people.

  • Rising unemployment mainly reflects labour-force growth outpacing job creation rather than simply jobless growth. (p. 35)

    The Labour Force Surveys suggest that the broadly defined labour force grew by 46 per cent between 1995 and 2005 (from 13.8 million to 20.1 million)

  • The 2 million new jobs claim is partly a data artefact, since much of the growth came from informal and subsistence work and definitional change. (p. 41)

    over 1 million new jobs were created in informal self-employment (760 000) and subsistence agriculture (300 000) alone

  • Minimum wages redistribute income from the top third to the bottom third of households, but aggregate disposable income falls. (p. 164)

    At the benchmark simulation (short run), households in the bottom group experience a 3.1 per cent increase in disposable income.

  • Partial equilibrium models understate minimum-wage job losses at low elasticities because they ignore price effects. (p. 168)

    Partial equilibrium model job losses 163,757 272,928 382,100 545,856

  • Assumptions about who loses jobs strongly affect poverty results: the micro-simulation concentrates losses on poorer households far more than the micro-incidence model. (p. 173)

    if we assume the bottom four deciles to represent the poor, the micro-incidence model shows that 40 per cent of job losses will impact on this group, compared to 66 per cent in the micro-simulation model.

  • A general wage subsidy costing about R30bn would create roughly one million low-skilled jobs at the benchmark elasticity, with cost per job falling sharply as elasticity rises. (p. 192)

    On average, about 500 000 low-skilled jobs are created in the low elasticity case (η = 0.3) compared to one million when η = 0.7 and 1.5 million when η = 1.0.

  • Income-tax financing is preferable to a capital tax for the wage subsidy: capital taxes divert capital from subsidised sectors in the long run. (p. 197)

    a capital tax imposed on capital used in wage subsidy sectors will in the long run tend to divert capital stock

  • Youth-targeted micro-simulation gives much larger income gains for the poor than the micro-incidence approach. (p. 210)

    the average increase in disposable income in the bottom four deciles