Are South Africa’s teachers among the best paid in the world? Using household assets as a proxy for monetary pay

Teachers tend to be the largest group of workers in any country with wages driven by government policy. Teacher unions play an important role in determining these wages, but so does research into what levels and systems of teacher pay are best for educational outcomes in a specific political and economic context. One element of this research is international comparisons, and one purpose of these comparisons is to gauge whether teachers in general, or specific categories of teachers, are paid too little or too much in one country. These comparisons are often unreliable due to inconsistent use of definitions, unreliable data on what workers in countries actually earn, and complexities in determining purchasing power parity (PPP) exchange rates between currencies. Some widely publicised comparisons are misleading, yet why this is the case can be unclear because of insufficient transparency around methodologies. It seems PPP complexities often make monetary comparisons difficult. A key contribution of the paper is to demonstrate the use of household assets as an alternative to monetary wages, in this case through use of the IPUMS dataset of the University of Minnesota. Such an approach seems to produce intuitively correct comparisons. This is especially so for South Africa, the country the paper pays special attention to. The household assets approach is shown to be useful both for absolute comparisons of teacher purchasing power across countries, and for the calculation of the within-country advantage of teachers, which can then be compared internationally. Comparing teachers to other professionals in the same country is the basis for the UN’s preferred teacher pay indicator. Though this indicator uses monetary wages, household assets are shown to be useful for this indicator too. Household assets are also used to estimate conditional purchasing power premiums for teachers in multiple regressions. The paper concludes that existing findings that South African teachers enjoy the purchasing power of teachers in, say, Denmark are incorrect, and that their purchasing power is not that different to teachers in other middle income countries.

C13, I28, J31

South Africa, teacher pay, IPUMS, household assets