|Turbulence or Orderly Change? Teacher Supply and Demand in South Africa - Current Status, Future Needs and the Impact of HIV/AIDS (CIE, 2000, 36 p.)|
South Africa is lucky, relative to other middle-income countries, in having a reasonably large, well-conducted, and recurrent set of household surveys that measure the socio-economic condition of the country. It may be fashionable at times, and in certain circles, to complain about StatsSA. But the reality is that the October Household Surveys, amongst other data series, are a useful source of information, sufficiently large to allow quite a lot of statistical power, and reasonably well carried out. We put together the 1995, 1997, and 1999 surveys and tried to glean what comparative data we could. There are limitations and dangers in doing this but we believe the risks are worthwhile.
An initial foray into the data produces the following results. Annex Table A1 summarises some of the results of statistical analysis. The conclusions below are stated with at least a 95% degree of confidence (though mostly with 99% confidence).
· The teaching force is 20 to 25 percentage points more feminine than the rest of the working labour force and is becoming more so. The entire working labour force became more feminized over the period 1995 to 1999. However, the teaching force did not become feminized at a faster or slower rate than the rest of the working labour force. There was no statistically significant difference between the rate of feminization of the teaching force and that of the whole working labour force. There is a generally declining preference for being a teacher; if anything, womens preference for being a teacher declined faster than mens, even though the teaching force became more feminine. In other words, women were increasingly taking up opportunities in non-teaching roles faster than in teaching, at the margin.
· Teachers say they work fewer hours per week than the rest of the working labour force. Self-reported data indicate teachers work some 17% fewer hours per week than those in the general labour force. Moreover, this does not take into account periods during the year when schools are closed - the data only refers to hours per week in an actual workweek. Furthermore, while the rest of the working labour force increased its rate of work over the period 1995 to 1997, the teaching labour force eased up its pace of work, in number of hours per week.3
3 Note that all this is by own declaration of the respondent.
· Teachers report earning much higher income than other employed persons - some 64% more. We should note that teachers are much more highly educated than workers as a whole and that this largely explains their higher income.
· Since the OHS measures income in nominal (inflation-uncorrected) terms, it is not possible to comment on whether salaries were changing in real terms without careful analysis of price inflation relative to wage inflation. However, it is valid to say that average nominal salaries for the teaching force increased faster than for others in the workforce.4 (If real salaries in fact declined after adjustment for inflation, they declined less for teachers than for others in the labour force).
4 For the other population segments the sample size was too small to enable very firm conclusions. Also note that it is not of interest to judge whether average salaries were increasing in real terms for teachers, at least in the context of this analysis. It is possible to judge whether the rate of change was higher for teachers than for non-teachers, but not whether it was higher or lower (for either teachers or non-teachers) than some index of price inflation.
· Teachers are far more educated than other employed workers-they have some 56% more years of education on average. We cannot tell from OHS data whether the average years of education possessed by teachers increased over the period, either in absolute terms or in comparison with other workers.
· Given their lower rate of monthly work hours, work-hour-adjusted salaries for teachers are even higher than nominal salaries, relative to other workers. Work-hour-adjusted salaries also grew faster for teachers than for the rest of the working labour force.
· Teachers were more unionised than the rest of the working labour force at any given point, and while the working labour force as a whole became more unionised over the period, teachers became unionised at a faster rate than the rest of the working labour force.
· The average age in the working labour force declined, but average teachers age increased.
· The most surprising factors are those relating to ethnicity or population segments. It appears that white participation in the teaching labour force increased relative to participation in the working labour force as a whole, whilst that of Africans either decreased, or stayed constant in the teaching force but increased in the rest of the working labour force. It is important to be clear what this data implies, since this is a surprising statement. The conditional probability of being a teacher is much higher if one is African than if one is white. But this conditional probability is decreasing for Africans and increasing for whites. A plausible hypothesis is that this is because the rest of the formal economy is opening up at a faster rate for Africans than are opportunities in teaching for Africans. It may be that opportunities for whites (relative to others) have waned in the rest of the formal economy faster than in teaching, either in reality or in terms of preferences by whites. Similarly, the conditional probability of being white, if one is a teacher, declined, whereas that of being African increased. But the conditional probability of being white, if one is in the rest of the labour force, decreased significantly faster than did the probability of being white if one is in the teaching labour force, whereas the probability of being African, if one is a teacher, increased more slowly than did the probability of being African if one is employed in the rest of the labour force. The trends observed may be due in part to the influence of teaching posts created by School Governing Bodies (SGBs) in public schools, or independent schools. The OHS is not restricted to teachers employed only by the public sector. Unfortunately, it does not contain information to allow comparisons between publicly and privately employed teachers.5
5 This possibility was suggested by Kuben Naidoo.