In this final reply, I will correct an error, account for a new finding and try to give a principled answer to the replies Brian Kantor and Loane Sharp, whilst explaining some of the basics that underpin my argument that in South Africa, wages are too low, profits too high and labour legislation must be defended (see original debate here).

First of all I find myself with the unpleasant but necessary task of pointing out that in my Business Report article (13/12), I made a calculation error that led to an impossibly high figure for the total increase in SA labour productivity for the period 1990-2011, ten times the true figure. According to the data used by the Reserve Bank, the total increase in labour productivity from 1990-2011 in SA is about 65 percent. I sincerely apologise for this, also to my critics who did not mention the error in their responding articles (14/12, 22/12).

Labour productivity - measured by the Reserve Bank as total value added, or as gross domestic product (GDP), per formal employee - fell during 1990-1993, but has increased from the start of democracy at a little more than 3% on the average, year on year. From 1998 and until today, on average, annual real wage increases have been one percentage point lower than the increase in labour productivity.

A 3% annual increase of GDP per employee - labour productivity - is above increases in labour productivity in many industrialised high income countries. "Compared to other middle-income countries, South Africa has relatively strong average labour productivity", writes Organisation for Economic Co-operation and Development (OECD) in a July 2008 Policy Brief.

Indeed, a diagram that compares labour productivity development in 18 countries in the 2011 Budget Review from the Treasury (page 45) shows that for a developing country, a 65% labour productivity increase over two decades would lie between normal and very good.

Strangely however, that Treasury diagram (see below) reports that SA productivity growth 1990-2008 was only about 1% per year, or about 20% for that whole period! A worried Treasury comments: "Over the past two decades, real wage growth has outpaced growth in labour productivity, which has been relatively slow." This completely contradicts both the Reserve Bank and the 2008 OECD Policy Brief. How can this be?

Treasury diagram using incorrect data for SA:

As I explained in my first article (13/12) the original labour statistics from Stats SA (called QES) have been revised by the Reserve Bank. The Bank instead uses the revised data for its discussions on productivity in the Quarterly Bulletins. Stats SA has also made an internal revision of the original data series.

In total, about one million persons suddenly appear in the old data during the 2000s in three upward jumps - so called "structural breaks". This has no resemblance to actual reality. When the unrevised data is used, it creates an absurd drop in labour productivity of minus 23% from 2000 to 2007. This defunct statistic still continues to be published giving rise to a great deal of confusion.

The Treasury cites the International Labour Organisation (ILO) as its source for its 2011 Budget Review diagram, not the Reserve Bank. The responsible officer at Stats SA says: "Both the ILO and OECD use the ‘official data'. The series with the breaks", because this is the data the Stats SA supplies them with.

The OECD officer responsible for the 2008 report, quoted from earlier, says however: Of course, the OECD staff did not use the data series with the "structural breaks" when they wrote that labour productivity development in SA is relatively strong (and that employment is the main problem for SA, not productivity).

The only explanation for the Treasury's mistaken assessment of trends in SA labour productivity is that its source was the ILO which, in turn, relies on the faulty data series from Stats SA. Through this circuitous route this flawed data has found its way into one of the most important policy documents of the government. In other words, in the 2011 Budget Review Treasury's argument for a quasi wage freeze in SA is underpinned by a crazy data series.

It is in this context that AIDC in September was forced to issue a press statement responding to Adcorp and the Centre for Development and Enterprise (CDE). They too have been using the wrong and unrevised data in their debate about productivity. This they have yet to dispute in this debate. Labour broker economist Loane Sharp has remained silent on this crucial issue in his two contributions.

Having dispensed with the issue of the use of the defunct statistics, we can now turn to the more substantial issues underpinning the debate around labour productivity.

I wrote that Sharp's methods are completely idiosyncratic (13/12) and that his references to OECD or the US Bureau of Labour Statistics were irrelevant or wrong (19/12). None of this he is able to refute. In his reply, titled "Excessive wage hikes a recipe for disaster" (22/12), he introduces new issues and writes that "the intention behind productivity enhancement is not higher wages for workers, but the increased profitability of the firm".

No economist, regardless of the economic school they belong to or their political sympathies, can agree with such a narrow framing concerning the question of productivity.  

The benevolent purpose of supporting productivity increases is to increase the total income of society; more precisely, the growth of gross domestic product (GDP) per capita. This means: the amount of goods and services that on the average would be available for every inhabitant, should the national income be divided equally. The starting point for discussions regarding distribution of this national income (GDP) is: "How much do we have per person in this country every year, no matter if we are children, in working age or pensioners?"

For example in The Wealth of Nations (1776) Adam Smith wrote: "Consumption is the sole end and purpose of all production; and the interest of the producer ought to be attended to, only so far as it may be necessary for promoting that of the consumer".

The book is regarded by many as the beginning of Economics. Smith's concern is not focused primarily on the naked interest of the capitalist firm (unlike Mr Sharp), but society in general, i.e. all consumers.

GDP is not the same as company profits. It is a much broader measure. In the national accounts of a country, GDP is more or less equal to profits plus wages. The wages and profits of a firm together comprise the value added by that firm to the economy, i.e. to GDP.

Stripped from juridical and political restraints, this basic wages-profit division of the value a firm contributes to GDP - and the same division of the whole national income - is often seen as a choice between consumption and investment.

Profits, or the surplus after paying wages, are the part designed to be reinvested. Ironically, Mr Sharp mentions the cooperative. But in organisations with general ownership, as opposed to privileged minority ownership, the reality of that basic choice is self evident. Society has a choice to make with new value created. Shall the new income be used for consumption or for investment? Who shall control how much of it?

In the case of a large private company, the debate at end of the day must become quite similar. The less the profits under the exclusive control of private owners, are reinvested, the less they can be taxed and finance public services - or, the more ruthlessly owners try to increase them, the more indefensibly profits are maximized without concern for environment or employees, the more profits are expatriated legally and illegally, hoarded on bank accounts or supporting an extreme upper class lifestyle - the less weighs the argument that profits are too small and the heavier weighs the arguments for increasing wages and supporting other goals. As Adam Smith observed: "What improves the circumstances of the greater part can never be regarded as an inconveniency to the whole. No society can surely be flourishing and happy, of which the far greater part of the members are poor and miserable".

In his reply "Job protection leads to shrinking workforce" (22/12), Mr Kantor believes that "the Left" thinks it is an arithmetic impossibility for average wages to rise when the wage share of the gross domestic product (GDP) is falling. He has misunderstood the issue. Nobody argues this.

Mr Sharp's example of the average real wage increase 1995-2011 in fact illustrates what this is about. If we take his figures for granted, then there has been an average and after-inflation (i.e. real) wage increase from "R9 378 a month to R12 564" over the 16 year period 1995-2011. This is a total increase of 34%. The average real wage increase per year then becomes about 1.8% (the nominal increase is irrelevant here).

The first point to note is that the 1.8% per year increase in real wages, as stated by Sharpe, is more than 1% below the average yearly growth in labour productivity during the period. That is, real wage growth has increased less than labour productivity. By the law of arithmetic, this is why the wage share of the national income is decreasing in SA.

Mr Sharp's explanation that this is wrong and the wage share instead falls because "employment has been falling relative to wages" is only an unclear acceptance of the same mechanism. Unclear, because we must keep track of what is produced per employee before we look at the wages. I shall explain this step-by-step.

If employment falls a lot and production falls a little, or remains the same or even increases, then those who remain employed produce more per person. This means that labour productivity has increased, according to any statistical bureau, methods manual or economic research institution in the world.

If the remaining workers in the example above are paid lower wages, the same wages, or even get real wage increases, but which are lower than the increase in labour productivity, then the wage share of the national income must fall. It's a simple and incontestable fact. Full stop.

Secondly, the 1.8% annual increase derived from Mr Sharp's example is an average for the whole labour market, including top executives and board members of companies. No doubt, tens of thousands of ordinary workers get real wage cuts today in a labour market that different sources claim is 30 percent controlled by labour brokers.

Mr Sharp from Adcorp informs us in the CDE booklet Jobs for Young People that "over the past eight years an entry-level employee in our organisation has earned the same nominal wage. So an 18-year-old employed by us today earns the same as an 18-year-old with equivalent qualifications did eight years ago". This is a decrease with 35% in the real buying power of that entry wage, due to 8 years of inflation.  

Thirdly, half of all employees in South Africa earned less than R2,500 per month and one third earned less than R1,000 per month in 2008, according to the National Planning Commission. If real wages really started to increase by 1.8% per year for millions of ordinary South Africans, it would take 40 years for their buying power to double from today's R1000 to R2000, or from R2500 to R5000.

Even R5000 per month is arguably not a living wage. The absence of measures to recover from the history of cheap wages (something that apartheid was designed to ensure) and to put our hopes in an extremely outdrawn process is not politically sustainable, or compatible with democracy.

Surely, faster change is needed. As for the unions, they should consider demanding wage increases in rand instead of percent. This would lead to rapid growth of the lowest wages and also address inequality. The broader solution is already known and coincides with the fastest and most sustained growth of the global economy.

Before World War 2, the wage share of GDP in countries like the USA and Britain was like it is in South Africa today. But after the war, real wages increased greater than labour productivity. The wage share of the national income increased drastically. The very foundations of inequality and imbalance, was addressed and the rapid growth of the world economy roughly between 1948 and 1975 is history, so to speak.

That example answers in passing Brian Kantor's objection (22/12), which is common, that it is the increase in capital intensity in SA that explains 15 years of falling wage share. If democracy develops, the ownership argument is defeated as time passes. A general sense of ubuntu, in an ever growing economy, must win.

The obvious fact that technical progress of any kind, all machines, tools, are products of labour and human ingenuity; invented, manufactured and repaired by employees everywhere, inevitably wins over narrowly pointing to who it was that had money to buy all the gear that everybody is increasingly using. We have had 250 years of technological development in the industrialised countries.

If Mr Kantor is right, should not this by now have put the wage share of the national income in countries like Germany, France or UK close to zero? The absolute opposite is a fact: The wage share of the national income in the old, capital intensive and highly industrialised countries with much more equipment per employee is 15-20 percent higher than it is in South Africa. This is what democratic development does in a country and what it also should mean here.

The attacks on labour laws and regulation finally, are part of an international political movement driven by the employers and the ruling elites. Labour laws have become too worker friendly in India and in England, everywhere... in a race to the bottom provoked by competition: "We must do this, because others are doing this".

In China, the first companies are now starting to move to Cambodia and Vietnam. They have come under pressure from the emerging labour movement. Will this first sign of capital flight deter Chinese wage earners from demanding decent wages and human rights? Should it? To pose such questions is to answer them. The labour laws now under fire are a product of the fight for democracy, in South Africa and in the whole world. The solution to unemployment cannot be to erode them, abolish them and roll back history.

Page S-150 of the National Accounts, describes in numbers what I and others are pointing to. In 1994, the wage share to GDP was 55.9%. In 2010 it had fallen to 50.6%, or with 5 percentage points. This year, goods and services will be produced in South Africa to the value of about 3 000 billion rand. The drop in the wage share of the growing national income between 1994 and 2012 corresponds to R150bn not paid out this year to ordinary employees, and which therefore cannot boost demand, to the benefit of small businesses and industry.

When conservative economists argue for higher profits as a cure for unemployment they disregard that this has been tested in practice for well over a decade with disastrous results for the real economy. They argue for more of the same - as if there is no history.

To combat growing inequality at its source, and to curb unemployment, real wages must instead start to increase substantially. There can be no diversified local industry if not even the overwhelming majority of people with jobs have disposable incomes that pay for essentials, as well as provide basic improvement in the quality of life.

The extreme and growing income inequalities as well as the increasing profit share of the national income, together deprive South Africa of local markets and jobs. In conclusion, and in the political language of this remarkable country, a crucial ingredient in the fight against mass unemployment in South Africa is Broad Based Working Class Consumption, BBWCC. Maxhalanga, ndityeni!

Dick Forslund

Alternative Information and Development Centre.

Click here to sign up to receive our free daily headline email newsletter