Thursday, November 8, 2012

... and Statistics

There is an old saying that there are “lies, damn lies, and statistics”.  There is also the chestnut about a politician using statistics “the way a drunk uses a street light – more for support than illumination”.  Of course the problem is not with the numbers (assuming they are accurate) but with their interpretation.  

One of my favorite examples is the story of the Fever trees.  This tree is actually a species of acacia with yellow pom-pom flowers and attractive greenish-yellow colored bark.  It got its name because in the days of the exploration of the hinterland of South Africa, it was noticed that people who camped under the trees often came down with a bad fever.  What’s more, sometimes their skin took on a yellowish tinge (caused by liver complications) which reminded people of the tree bark.  Once that connection was noticed, people started avoiding the trees and – although the fever still occurred – it did seem somewhat less prevalent.  Needless to say, the trees did not cause of the fever.  What was happening was that cause was being confused with effect.  Indeed there was a correlation: the trees love to grow along water courses.  The mosquitoes that carry the yellow fever virus also like to be near the water where they breed.  So the water was the connection.  People who camped near water – a sensible thing to do to have a ready supply – were as much at risk as those who camped under the shade of the trees.

This is a long introduction to what I really want to discuss.  The South African census of 2011 was released by Statistics South Africa last month.  The context is that we need to be careful about how we interpret the data.  I promise I won’t make this long!

Here are two deductions from the census data:  
1     1      If you are white, you are more than four times as likely to have a higher education qualification (i.e. post-secondary school) than if you are black, 
2     2  Nearly two-thirds of the people with higher education qualifications are black.

Here’s another pair (money translated to US dollars).
3   Over the ten years since the last census, white households increased their average income by $20,000 per annum, while blacks increased theirs by only $4,500,
4.       Over the ten years since the last census, black households have increased their income by 50% on average, while white income on average has stayed the same.

It seems to me that statements 1 and 3 might convince a casual reader that the government has done little to improve the lot of black people in South Africa, and has merely allowed itself to be co-opted by the wealthy white group.  On the other hand, statements 2 and 4 might convince the same casual reader that the government has done well and is strongly supporting black empowerment.
Yet these four statements are all supported by the census data. Here is the data:
  • ·         36.5% of whites have a higher education qualification
  • ·         8.3% of blacks have a higher education qualification
  • ·         There are some 52 million people in South Africa and 80% of them are black. So using the bullet points above, 2 million blacks have higher education qualifications.  The total number of people with higher education is 3.6 million.

·         In the income figures, it depends on whether you take the real value of money into account.  Over the ten years, incomes would have had to grow at around 80% in order to keep up with inflation.  White incomes grew by about that amount so stayed flat in real terms.  Black incomes increased by 50% (in real terms) BUT they started from a very low base.  In real terms they grew from about $5,000 to $7,500.

There’s comfort in the census that a lot has been achieved, as the positive take on the above figures suggests.  Also 91% of households now have water on tap (80% in 1996) and 85% have access to electricity (58% in 1996).  The percentage of people in formal housing has also increased from 65% to 78%.  This last fact is much more dramatic when expressed in terms of the actual numbers of people: in 1996 26,4 million people had formal housing.  Now the number is 40 million. (And the number of people per household has decreased by 1 on average.)

Much has been done well.  So why the dreadful events at Marikana?  Here is another statistic. Something like 60% of working age people are employed in the formal or informal economies. About 50% of 20 to 30 year olds are employed. That’s a lot of unemployed people.  The government mooted a plan to subsidize the employment of new school leavers.  But the trade union movement blocked it.  The same movement that was shunned by the miners at Marikana.

Thanks for bearing with me.

Michael - Thursday

1 comment:

  1. I feel overwhelmed by statistics this week. Between incessant analyses on the US election results and unemployment figures in Greece, I'll settle for a night under the fever trees...the hallucinations will be a relief.