Lies, damned lies and statistics

by May 29, 2012All Articles

Lies. damned lies and statistics. It’s a cliché that has been used several times in this column over the years. I have used it because statistical data, especially about employment and the cost of living, can influence attitudes, opinions and policy decisions. As such, it is readymade for use as propaganda in ideological battles and power struggles.
Because such data can be misused and abused, all statistics should be treated with caution and can and should be open to constant critical analysis. However, for the most part, statistics, both official and from academic sources, are produced in a transparent manner and without any overt ideological bias. But seldom do such statistics present radical or startling information: official
inflation may rise a little more or less than expected by commentators and jobs and jobless figures may be quibbled about.
However, the unions complain that the media usually accepts figures produced by Stats SA or one of the university economics departments without critically analysing them, even when additional data are available. So something like an increase in the per capita income of South Africa to R80 000 a year can be hailed without noting that the income gap has increased.
That gap means that the wealthiest 10 per cent of the population enjoys nearly 45 per cent of the wealth generated annually, while the bottom 10 percent share just 1.3 percent. In round figures, the wealthiest 5 million people in the country share roughly R248.5 billion while the poorest 5 million get by on R7.2 billion, an anything but glowing picture.
But over the past year a new element has emerged on the statistical front that, several unions maintain, is “pure propaganda”. This is the “employment index” produced by Adcorp, the country’s largest and most diverse labour brokerage.
The unions have come late to this battle, but this week there were allegations across the labour movement that an ideologically motivated attack is underway against them. The weapon, Cosatu says, is “bogus statistics” that are being published by a complicit or, at the very least, malleable media. At the same time, the same statistics continue being attacked by academic economists, discounted by the government’s statistical agency and have caused some caustic comments from sources such as the Commission for Conciliation Mediation and Arbitration (CCMA). The evidence reveals that the media is certainly at fault although under resourced newsrooms and inexperienced journalists are probably where the blame lies, rather than with any conspiracy.
Perhaps because the “Adcorp employment index” is distributed free of charge its results are usually published and broadcast widely. But also uncritically — and often accompanied by commentary from a company executive, Loane Sharpe, who is frequently described only as “a labour economist”. “But he is in no sense neutral. He is a role player with an agenda,” complains Cosatu spoksperson Patrick Craven. “To back up its political views, Adcorp is playing political and self-serving games around the employment crisis by publishing figures that cannot be objectively verified.”
Cosatu’s critique came in the wake of the Adcorp March employment index that maintained that unions are losing members and that a decline in income through subscriptions was a cause of labour unrest. Craven points out that this does not make sense: “Unions are organisations of workers, they are not separate from their members and are driven by workers.” Various recent comments by Sharpe also prompted Nerine Kahn, the director of the CCMA to note diplomatically this week that the Adcorp executive seems to “lack an understanding of the labour market”. Yet such an understanding, she says, should be a minimum requirement for a labour economist.
She gives as an example of this apparent lack of understanding the insistence by Sharpe that the CCMA is also responsible for many strikes. Yet the CCMA, which provides a free mediation and arbitration service, is widely regarded as playing a major role in diffusing industrial unrest. Particularly galling to Kahn and CCMA commissioners is a statistic frequently quoted by Adcorp regarding labour brokerage referrals to the CCMA. The source for this is a “very rough, back-of-the-envelope” estimate made two years ago in response to a request from a business representative.
This CCMA “guesstimate” was all that could be provided because it was — and remains — unclear in most cases whether workers, and especially contract labourers, in dispute with an employer are connected to a brokerage. Complaints to Adcorp and efforts to clear up this matter have so far come to nothing. However, Sharpe’s comments and the statements about the CCMA contained in the monthly index, have largely gone unchallenged in the mass media. “Not once did anyone in the media ask the CCMA for comment or contact us to verify the facts,” says Kahn.
It is not as if there has been no criticism of this index and the comments of its main promoter. Academic journals have raised serious questions about the claimed statistics and leading academics such as Professor Martin Wittenberg of the University of Cape Town, have published highly critical articles. However, such criticism concerns mainly the methodology employed in compiling the index and tends to appear in academic journals. Arguments at this level can be complex but, as a number of critics point out, it is unnecessary to go into such detail: it is enough to know that neither Sharpe nor Adcorp have revealed how they come to the conclusions they publicise.
This was made clear at a recent debate at the Mapungubwe Institute for Strategic Reflection. Several economists present told Sharpe they were at a loss to work out how the figures produced in the index were arrived at. They wanted answers. They failed to get them.
But perhaps facts are not the issue. Nor politics, although the index certainly plays to anti-union prejudices among elements in the busines; marketing sector. The more prosaic view may be correct: Adcorp deserves an award for shrewd marketing, that can be summed up as, never mind the detail — give us the promotion.
Share this article:


Latest issue

Amandla Issue #92