Productivity is rising even as wage share dips, data show

by May 23, 2012All Articles

Last week, the Reserve Bank (SARB) issued its latest Quarterly Bulletin. It covers economic development in South Africa up to June 2011 and economic prospects for the coming period, as the SARB views it. As usual, the bank also discussed developments in labour productivity. Data underpinning that discussion are published in a supplement called Key Indicators, together with data about many other things. The document reveals that average labour productivity was up 1.1% in June, compared to June last year.
During the past year, both labour productivity data and the very concept labour productivity have become increasingly surrounded by confusion and spin. Economic commentators and think-tanks have intervened in the debate about wages and protective labour laws with claims that the productivity of labour is going down, is “the lowest in 40 years”, et cetera.
Given the attention that is currently being paid to the so called “New Growth Path Social Dialogue”, which is concerned with the establishment of a productivity accord, it is essential that this propaganda and these easily refuted statements are challenged. Indeed, as the latest Reserve Bank figures again make evident, it is a National Employment Accord that is really needed and not a productivity accord.
Labour productivity is defined by the OECD as “output per unit of labour input”. For output, the SARB looks at the recorded value of all goods and services produced in South Africa during a 12 month period. For labour input, the SARB uses the number of employees producing that output.
Calculated in that way, there has been a total 660 per cent increase in output per employee in South Africa since 1990 [This shall be 65 per cent. Note Dick Forslund’s correction below]. This means that the average South African formal employee today produces 6.6 times more per year than he or she did only 20 years ago [See correction below]. Labour productivity stands at a level never before experienced in history. This is so in South Africa and in the whole world.
Of course, one might add that a steady increase in labour productivity over time is normal. In recent decades, the only time that an increase in labour productivity was not achieved in South Africa was during the volatile and difficult years immediately before 1994.
So how is it that different “experts” arrived at contrary conclusions this year? Firstly, debaters, like the Centre for Development and Enterprise, have been using two statistical tables in which historical under reporting from employers completely distorts the numbers. Corrections made by Statistics SA in 2002, 2004 and 2006 added over a million formal employees to these tables in three batches, creating illusory drops in labour productivity during the 2000’s. Many more workers suddenly seem to be producing almost the same amount of goods and services. The “new” employees had been there all the time, but were not reported. Stats SA and the SARB are therefore using revised data for the period 1990-2006; the SARB to get a truer account of labour productivity before 2006.
In addition to making this grave error, so called “labour unique productivity” has been marketed in media during 2011. There is no such thing. All economic schools recognise that the productivity of our labour over the past is crucially dependent on our tools and machines (i.e. capital). It cannot be “stripped out”, without arriving at the most peculiar conclusions. In one bizarre press statement from the labour broker Adcorp, the public is informed that real GDP increased by 6.6% in 2011, employment by 2.6%, but that labour productivity is going down. When production increases faster than employment, the difference incontrovertibly signifies an increase in labour productivity. The most zealous search will not turn up a serious economist who thinks otherwise. In this case, labour productivity has increased by 4 %, because 6.6 – 2.6 = 4.
What should the trade unions and the general public make of this? Firstly, the official numbers are the natural starting point for a critical discussion. It is the table “Labour in the non-agricultural sectors” that is used by the SARB in its productivity reporting. According to the SARB, labour productivity has increased at an average rate of about 3%, quarter by quarter in 12 month periods, during the last ten years. Real wages has increased at an average yearly rate of only a little more than 2% during the same period or at a 0.9 percentage points lower rate (Diagram).
Because real wages increase at a lower rate than labour productivity, the wage share of the national income has been falling since 1998. The profit share of the national income rises correspondingly. The latter fact SARB reports in “National Accounts: Ratios of selected data”.
Diagram: Average labour productivity increases 2001-2011, compared to changes in real wage, quarter by quarter, year on year, in the non-agricultural sector. Source: SARB’s Key Information table ”Labour in the non-agricultural sectors”. The diagram is made by the author.
Labour productivity is indeed growing every year in South Africa. The task of a progressive government must be to use the steady productivity gains made in many sectors, by transferring surplus resources created there to useful public jobs, like housing, or to sectors like health care – i.e. sectors which by their nature is very labour intensive and where productivity development therefore always is much lower. At Cop 17 in Durban, one million public jobs to protect the climate were suggested.
In general we must recognise that the economic elites in SA inherited extreme profit shares of the national income from apartheid and colonialism. Those profit shares are growing, instead of decreasing. In August, the investment company Stanlib reported that big business in South Africa currently sits on a R479 billion mountain of cash, equalling 18% of the GDP, or close to half of the entire government budget for this year. A wait-and-see practice, of hoarding excess profits is bank accounts, has currently reached the same levels that prevailed in 1995.
This is an indication that the trade unions are gaining much too little in their fight for decent work conditions and higher wages. Every year, wage increases should on average correspond to the increase in labour productivity plus the expected rate of inflation. This is a minimum. The wage share of the national income must go up. South Africa’s local markets for food, everyday items and semi-durable goods will otherwise continue to be strangled by extreme inequality and deplorable incomes for the vast majority.
There can be no diversified local industry without local mass markets. Broad Based Working Class Consumption is needed to break away from mass unemployment in South Africa.
Dick Forslund is an economist and researcher at Alternative Information and Development Centre in Cape Town.
Foot note added by Dick Forslund (January 16, 2012): Due to a calculation error, the “660%” or “6.6 times” numbers (the latter changed to “7.6” by the Business Report editors), which illustrated the ever growing labour productivity in SA and elsewhere, became impossibly high. According to the data series used by the Reserve Bank, the total increase in labour productivity 1990-2011 is about 65%. This is a normal increase. I discuss this further in my final reply. The error was NOT mentioned by the other debaters, Mr Sharp and Mr Kantor.
Reply by Loane Sharp to Dick Forslund, originally published in Business Report, December 14, 2012:
Productivity myths and labour illusions
Dick Forslund argues that labour productivity in South Africa has risen sharply in recent years (“Productivity is rising even as wage share dips, data show”, December 13 2011). He concludes, among other things, that “trade unions are gaining much too little in their fight for decent work conditions and higher wages”.
Forslund’s argument hinges on the output-per-unit-labour measure of productivity. If a firm produces 100 units using 100 workers, average output per worker is 1 unit (i.e. 100/100). If, by installing a new machine or adopting a new technology, the firm produces 120 units with 80 workers, retrenching the other 20, average output per worker rises to 1.5 units (i.e. 120/80), an increase of 50%. In this simple example, it is clear that the additional production was achieved, not by workers, but by the introduction of a new machine or technology. This is the labour productivity illusion: attributing to workers what is, in fact, attributable to capital or technology.
In contrast with Forslund’s claims, the OECD defines productivity as “the ratio of a volume measure of output to a volume measure of input”. As such, labour productivity varies as a function of both other input factors (management, arable land, natural resources, physical capital, information and other technologies, etc.) and the efficiency with which the input factors are used. The US Bureau of Labor Statistics, which calculates productivity in the United States, prefers the “marginal productivity” measure: the change in output that results from changing one of the inputs by one unit, all other factors remaining constant. An exact statistical procedure for doing so is described on Adcorp’s web site (contact for details). Using the OECD’s and BLS’s preferred methods, South Africa’s labour productivity recently fell to the lowest level in 40 years.
Attributing all production to workers, as Forslund does, is closely connected to the Marxist conception of value. This conception, the labour theory of value, was demolished long before Marx’s writings by one of its early proponents, the 18th-century economist David Ricardo, who, writing in his later years, showed that a bottle of wine lying in a cellar increases in value with no additional labour input. Later, John Maynard Keynes dismissed Marx’s Capital, in which these ideas were most elaborately expounded, as “an obsolete economic textbook which I know to be not only scientifically erroneous but without interest or application to the modern world”. Yet Marx’s ideas are enjoying a surge in popularity: his German publishers have reported 10-fold growth in orders since the global financial crisis. Forslund, whose institute’s slogans “People before profits” and “Resisting the WTO” hanker after those errors, has an ideological axe to grind.
Productivity, in the sense of making better use of available resources, is a paramount economic goal. It is the only known way of producing improved living standards, and the various input factors, including labour, should be applied to this end. But a raft of laws and regulations undermines labour productivity in South Africa. In the 15 years since the Labour Relations Act was introduced, real (after-inflation) wages increased by 28.8%, which is nearly treble the increase in the preceding 25 years. Per unit of productivity, real wages have increased at an astonishing average annual rate of 7.6% – or 200% in total – since the LRA was introduced. The economy’s capital intensity is rising sharply for the simple reason that the productivity of capital exceeds the productivity of labour, relative to their respective costs. This (not wavering on trade unions’ part, as Forslund claims) is why labour’s share of national income has fallen to the lowest level since reliable records began 50 years ago.
Loane Sharp is labour economist at Adcorp
In September 2011 AIDC released statements questioning the accuracy of productivity data used by Adcorp and the Centre for Development and Enterprise (see here).
In November 2011 Adcorp released a report stating that “this year, labour productivity (i.e. labour’s unique contribution, separating out the productivity that is rightly attributable to capital equipment and machinery) fell to the lowest level in 40 years” (see here).
A debate then started in Business Report and on the PoliticsWeb about wages and labour productivity in SA. Below follows the first article by Dick Forslund  from AIDC.
Originally published in Business Report, December 13, 2012:
Share this article:


Latest issue

Amandla 92