It's a tragic irony of SA that companies continue to shed jobs while lamenting the lack of good people to hire. Of course, it's underskilled labour that is being shown the door while the hunt for skilled workers intensifies.
There's a simple but revealing explanation for this: the structure of SA's economy is changing fast. Whereas in 1990 the primary sector, which includes mining and agriculture, accounted for 15% of gross domestic product (GDP), by 2000 the figure was 10,3%. The secondary sector, which includes manufacturing and construction, was worth 33,3%; in 2000 it accounted for 26,2%. The tertiary sector, which includes services, was worth 51,7%; in 2000 it was worth 63,5%. The action is at the top end, the much-vaunted "knowledge economy" requiring highly skilled workers.
Meanwhile, in part to redress historical imbalances, wage increases at the bottom end have been growing fast. The average minimum monthly wage rose by 10,4% in the first three quarters of 2001 (latest figures), faster than the overall average of 8,9%. This is indicative of rising wages for low-skilled workers, which diminish their attractiveness as a factor of production.
This, combined with the expected impact of HIV/Aids, is pushing firms towards capital intensification. In 1975, 11,5 workers were employed to produce R1m worth of goods and services in the private sector. By 2000, only 6,7 workers were required for the same output.
The attrition of low-skilled workers left the overall skill level of the work force higher, meaning more effective production. But the bulk of the productivity gain came from the replacement of workers with machines.
Between February and September last year, official unemployment rose from 26,4% to 29,5%, according to Statistics SA. The Mesebetsi Labour Force Survey Report, by Fafo, the Norwegian Institute for Applied Social Science, says it's higher - 32% counting just individuals who are actively looking for work and 45% if individuals who have given up the job hunt are included.
Of the 20 biggest employers in this year's ranking, eight cut back their work force last year. Of the 224 companies with comparative figures, 90 shed jobs. In all, 133 589 jobs were lost, only marginally made up by the 134 companies who hired 140 141 more workers than they fired.
According to the Reserve Bank, the annual rate of increase in the nominal compensation per worker in the formal, nonagricultural sectors of the economy picked up from 7,4% in the year to June 2001 to 8,8% in the year to September. Remuneration growth in the public sector accelerated and there was a modest slowdown in the private sector.
On average, economy-wide nominal remuneration outside the agricultural sector increased 8,9% in the first three quarters of last year compared with the same period a year before. Considering overall consumer price inflation averaged 5,7% over the period, employed workers enjoyed a generous improvement of about 3% in their real consumption wage - in practical terms meaning a real increase in spending power.
Partly as a consequence of heightened industrial action during 2001, the growth in worker productivity - measured as real output per worker in the nonagricultural sectors of the economy - fell back from a year-on-year rate of 6,5% in the second quarter of 2000 (a recent high) to 3,4% in the third quarter of 2001.
What was happening in the first three quarters of 2001 was that output growth slowed, in part a consequence of the increased incidence of work disruption, coinciding with a slowdown in the rate of labour attrition. Year-on-year growth in labour productivity in the first three quarters of 2001 was still at a relatively robust level of 4,4% but was significantly down from growth of 6,1% in 2000.
But here's the rub. Rising nominal compensation per worker and falling productivity can only spell danger for nominal unit labour costs.
Unit labour cost is derived as the ratio of nominal compensation per worker to output per worker. When productivity rises at a slower pace than nominal compensation per worker, the cost of the labour required to produce one unit of output rises. If one worker can produce 100 widgets in a day for a given wage, should his wage be increased by 10% his output - or effective productivity - ought to increase by 10% as well. If not, his employer faces a higher cost for a given level of output.
The recent slowdown in labour productivity growth is therefore the key to the pick-up in the growth of unit labour cost from a year-on-year rate of 2,9% in 2000 to 4,3% in the first three quarters of 2001. In the manufacturing sector, growth in unit labour cost rose even more steeply - from a year-on-year rate of 1,7% in the first quarter of 2001 to 7,2% in the third quarter.
It's a key indicator the Reserve Bank monitors for inflation purposes. If nominal unit labour costs are rising throughout the supply chain, the pressure grows to pass the increase down the line. Competitive pressures may force firms to squeeze margins - in effect absorbing the effects of the cost increase - but sooner or later, if unit labour costs continue to rise, firms are forced to pass the cost on in the form of higher sales prices. The ripple effect can have a severe inflationary impact.
The problem is that this fuels inflationary expectations, which in turn increase wage demands. Workers, naturally and rationally, want to protect the purchasing power of their earnings. The result is nominal remuneration per worker grows at an ever-increasing rate. If productivity gains cannot keep up, which becomes more and more difficult as wage gains accelerate, unit labour costs rise even more. And so begins an inflationary spiral.
Little wonder, then, that the Reserve Bank is so reluctant, given the broader inflationary pressures in the economy now, to allow expectations to run amok. Nominal unit labour costs, after falling throughout 1999, are again showing signs of increase, an indicator that should cause alarm to employers.
The ranking, reflecting turnover and profit in nominal terms, makes it difficult to draw any concrete conclusions. The change in turnover per employee is a measure of productivity shifts. Movements in the sales price of goods and services need to be stripped out. Rand value increases (because of currency depreciation) in turnover figures for dollar-priced goods can give productivity figures a false shine. It's impossible to say whether workers are actually producing more physical output or less.
What is interesting is to compare turnover/employee with profit/employee. Companies that are able to improve the latter at a faster rate than the former are doing well, signalling a drop in unit labour costs, assuming other input costs did not fall.