WHEN President Jacob Zuma contracted a pool company to build a fire-pool at his Nkandla mansion, they arrived, pottered around for days, got paid and left. He never followed-up by checking if the pool was in the right place, if it leaked or if the filter worked. It got worse. He did not notice that water dissolved the paint or that the walls collapsed and left a ghastly crater in his garden.
Something similar occurred when Bill Gates started up that celebrated company, Microsoft. He neither bothered to see if MS_DOS worked, nor did he debug errors.
That was reminiscent of the Second World War. There was no "military precision" about the D-Day invasion. It was not preceded by consideration of alternatives or research into why it might succeed. There was no estimate of anticipated troop losses or the benefits of landing. Military commanders got no feedback on what happened after the troops embarked, let alone whether they reached France.
None of the above is true, of course. Yet it is true of the surreal world of government. Governments routinely squander resources and erode liberty by passing laws without prior evidence that they are likely to be effective. They neither want nor get feedback on real world consequences. The machinery of government is designed to make laws regardless of need or impact. Legislatures legislate, executives execute and judiciaries adjudicate. There is no institutionalised error-correction. None of the three branches of government has an organ of state designed to monitor the extent to which intentions materialise. None of the institutions of government is designed to identify and repeal or reform failed measures, which partly explains the conspicuous incompetence of governments.
It also explains why regulatory impact assessments were invented. Socio-Economic Impact Assessments (SEIAs) became mandatory in 2016. It could be one of the most important developments in our history, yet scarcely anyone knows it happened. Hitherto, laws were passed at the behest of bureaucratic or political whim without objective evidence of probable or actual consequences.
We now have a SEIA Unit in the planning department to ensure that SEIAs precede laws and policies. About 120 have been done. Few have been released and all were done by the worst people to do them, officials with selfish bureaucratic interests.
We are in SEIA infancy. Things will improve when politicians realise they will get the information they need for informed decisions only if SEIAs are produced with public participation by or to the satisfaction of a properly empowered autonomous agency. Since government departments cannot interfere with each other, the SEIA Unit should be elevated to the status of an auditor-general or the public protector.
One of the first published SEIAs is perfect ... perfectly bad. Every proposed law used to be accompanied by an "explanatory memorandum" which superficially summarised political rhetoric. SEIAs are meant to replace rhetoric with substance. Instead of the "Impact Study of the Twin Peaks Reforms" on the Financial Sector Regulation Bill being a proper SEIA, it is glorified rhetoric. It motivates measures not mentioned in the bill and ignores what is in the bill. All that politicians who bother to read it get, is discombobulated twaddle.
The National Health Insurance (NHI) SEIA, on the other hand, follows the official guidelines. Its form is excellent. Its substance lacks plausible cost-benefit predictions, but it could easily be tweaked to do so.
Since their infancy, impact assessments have progressed to the point where they enable politicians to make objectively justified decisions and explain why they do not always proceed with populist options. If we learn from our early mistakes and the world’s experience, we could progress towards having the world’s best policies. That will happen when all policies are preceded by independently and expertly produced SEIAs and followed by monitoring and error-correction.
• Louw is executive director of the Free Market Foundation
This article was first published in Business Day on 7 September 2016