Black Diamonds and the trouble with statistics

Black Diamonds: not all they seem

In 2001 South Africa was suffering from weak economic growth.  High interest rates (broaching 25%) were causing business and home-owners sleepless nights.

It didn't help that Stats SA at the time had miscalculated some important numbers.  They got inflation wrong.  They got the rate of job creation wrong.  They got production numbers wrong.  And it wasn't until others pointed out inconsistencies that the methodology was reviewed and corrected.

Turned out the economy was doing much better than they claimed.  The high interest rates we experienced probably lasted a year longer than they needed to because of these mistakes.

Talking up the numbers is as bad as talking down the numbers.  Either can lead to unfortunate investment decisions.

Which is why it is so important that when any authority makes claims about the economy those claims be subjected to peer review.  This crucial process ensures that other experts get to look at the source information as well as the methods used to calculate those claims.

The UCT Unilever Institute of Marketing has come up with a whimsical and compelling brand: the Black Diamonds.  These, they say, are the middle-class black South Africans powering the economy.

There are 2.6 million Black Diamonds, we are told, and their numbers are growing at 30% per year.  We are also told that 12 000 households are leaving the townships for the suburbs every month.  These are big numbers indeed and are important for developers and retailers wanting to supply goods and services to a growing market.

Yet there is a lingering concern about these figures.  They don't relate to other economic data.  Their definition of what constitutes "middle-class" used in the report is left vague.  Most economists regard middle-class as being any household that earns above R 7 000 per month.  The Black Diamond study starts off at around R 1 000 per month.  The definition, on its own, may be contributing to the massive growth they report.

Most of the research is astonishingly mundane.  Who knew that teenaged girls like shoes?  Or that wealthy middle-aged men fancy buying a big-screen TV?  But, if you look beyond the good cheer, the research is flimsy.

Each of four groups is probed:  young adults loathing their studies and partying much as young adults do (average income R 1 241); recent graduates and self-starters planning their assaults on the corridors of power much as any of their peers do (average income R 6 292 per month); established middle class people complaining they don't know their neighbours and getting back to their roots, much as the middle class do all over the world (average income R 7 712 per month).

And one last group.  Young parents.  Most of whom are single mothers.  This group is exploding at 60% growth (from 440 000 to 710 000 people) and earning R 5 751 per month.  There's something wrong with this category.  Research all over the world shows that, as women become middle-class, their careers become more important and reproductive rates fall.  Any study that shows an explosion of single-motherhood in the middle-class has some explaining to do.

When I asked Professor John Simpson, who presented the findings, "How do you explain this data?  The explosion of single mothers seems unusual?"  he glared at me, turned to someone else and said to me, "I'm speaking to this person now."

The information I was supplied with lacked source numbers, and academic economists I spoke to said their requests for data had fallen on similarly deaf ears.

I have no doubts that the middle class is growing.  I do have my doubts as to whether the UCT Unilever Black Diamond research can help anyone sell better products to them.