Lies, damned lies and flawed statistics

Dependable statistics are critical to modern government. Without them delusion, wishful thinking and policy chaos reign. As is happening in the land affairs ministry.

Recent data reveal a serious flaw in the way the government is tackling land ownership redress. It shows that the restitution process distorts critically the measure that the African National Congress has chosen to assess progress on this critical and emotionally charged issue.

At present, successful claimants are given an option between land – for practical reasons, not always the land they were deprived of – or equivalent monetary compensation. This ”money or the box” choice has significant disadvantages.

The first is that it makes virtually impossible to achieve the goal of 30% of ”white” land – asserted to be 87% of South Africa, but a contested figure – to be transferred into black ownership.

So far 2.8-million hectares have been redistributed at a cost of R13-billion or R4 693 per hectare. This represents a mere 8% of the total agricultural land and amounts to a dismal 26% of the government’s target.

Failure to reach that self-imposed target is the reason why the ANC is contemplating compulsory expropriation. But it emerges from an Institute of Race Relations survey that on top of the R13-billion, another was R6-billion was paid in financial compensation to black claimants.

If this had not been paid out, but instead had been used to buy land, an additional 1.3m hectares could have been transferred, taking the government’s progress towards its goal to over the 40% mark. That might not be enough to satisfy the ANC’s land-hungry support base, but it’s not too shabby, given that the education department pegs 40% as a reasonable pass for matric students.

It is, of course, the democratic right of successful claimants to take the money instead of the land, but their decision needs to be factored into land-ownership percentages. Not doing so skews the picture.

A second disadvantage of the current restitution process is that paying out these claimants slakes land hunger temporarily, but it does not resolve it. Cash that is not invested in productive assets tends to evaporate on consumer products and in a few years the money is likely spent. Also, since most land was originally shared communally, when a pay-out is made it is to the minority who organised and sussed enough to lodge a claim; many who would have a historical right to re-settle restitution land miss out completely from the financial compensation.

Whiteys, however, are still very visibly sitting on the land, which makes double-dipping politically irresistible. That is one of the reasons why there is pressure on the ANC to re-open the land-claim process, which had as its cut-off date dispossessions going back to 1913, the year the Land Act was passed.

The complexion of who owns the land and how to keep feeding a burgeoning population requires that the ANC balance emotion with pragmatism. Statistical distortions – either deliberate, by politicians of every hue, or inadvertent, stemming from out-of-date assumptions and flawed extrapolations – should not be allowed to bedevil matters further.

For example, if the ANC used a range of performance criteria, instead of reduced land reform to the single, simplistic measure of 30% ”white” land reclaimed, it might pat itself on the back. By March 2011, government had settled 76 229 claims involving more than 1.6-million beneficiaries, representing an impressive 96% of total valid claims.

On the other side of the fence, farmer organisations argue, somewhat disingenuously, that if all state land – an estimated 25% of SA – was previously ”white”, it now is ”black”, in that government could redistribute it at the stroke of a pen. Add into the equation the already restored land and redistributed land – separate processes with different budgets — as well as the unknown but apparently significant amount of land transferred privately, and black land ownership conceivably is in the 40%-50% range.

But we actually just don’t know.

Whatever procedural changes there might need to be – like controversially abandoning the ”willing buyer, willing seller” policy – Land Affairs needs to work with credible statistics. A full land audit would be a good start.