Shifting costs, and blame, is no way to fix infrastructure
SOUTH Africans are getting used to headlines about large cost overruns on public infrastructure projects. Where, initially, the main problem of the infrastructure roll-out programme was the inability of public entities to spend their capital budgets, many state-owned enterprises (SOEs) now appear to suffer from a chronic inability to deliver projects on time and within budget. Cost overruns of more than 50% and delays of multiple years are not uncommon. Examples include Eskom’s three largest power station construction projects (Medupi, Kusile and Ingagula), Transnet’s multiproduct pipeline and Airports Company South Africa’s King Shaka International Airport.
The government’s plans for infrastructure roll-out are ambitious. For the next three years, R827bn has been budgeted for public sector projects. Beyond that, public sector-led projects worth R4-trillion are up for consideration. The government has leaned heavily on the retirement industry to step up the investment of pension funds into infrastructure projects. Given the poor performance of many SOE projects, it is no wonder that more pension fund investment is not forthcoming — Eskom’s most recent credit downgrade being a case in point. South Africa cannot afford to repeat past mistakes on this much larger scale. The imperative to improve on project outcomes is overwhelming.
Infrastructure projects are often large and typically highly capital intensive. The economic viability of these projects is determined at the outset: by the way the project is defined, specified and executed. Errors at this stage can usually not be remedied by smart operating or commercial practices later and consumers and taxpayers will pick up the bill.
Several causes of these problems have been identified. First, politicians expect state agencies to achieve a multitude of, often conflicting, political objectives that are ancillary to their main purpose and increase the complexity and cost of infrastructure projects. Second, large public-sector executed projects provide many opportunities for rent-seeking by public officials and private-sector participants, or worse.
Researchers have further identified a third, more fundamental and perhaps less recognised cause of the problem. Infrastructure projects are often large, very complex and will be exposed to uncertainty about unknown long-term operating circumstances and competing future technologies. Devising appropriate infrastructure strategies under these circumstances is exceedingly difficult.
Experience has shown that public sector decision makers are poorly placed to grapple with this problem. They tend to underestimate project complexity and uncertainty about conditions for future project performance and therefore typically have a poor grasp of project risks. A small group of decision makers, who control the project information, are insulated from the consequences of potential negative outcomes and thus have weak incentives to undertake adequate due diligence investigations or share information with stakeholders. This results in projects that are poorly designed for the uncertainties they will encounter during their execution and commercial operating phases.
This problem bedevils investments in complex infrastructure projects and is acute in the public sector, where projects are not adequately scrutinised by an army of financial analysts and rating agencies — as would be the case if private investors were taking sufficient project risk. Often idealised public sector projects are given the go-ahead because, upfront, they appear cheaper than their private sector alternatives, which have to reflect a more realistic view of the cost of internalised risks.
The literature has identified a number of investment strategies that can be adopted to reduce the effect of risk and uncertainty on project viability. Incrementalism involves choosing strategies that involve numerous smaller steps over time, rather than large upfront commitments. This allows for plans to be adjusted midway when the anticipated circumstances do not materialise, rather than being stuck with uneconomic sunk investments. Flexibility involves choosing strategies that allow greater midstream adjustment if required. Diversity increases the resilience of a system by spreading investment over a range of options and locations and is suited to accommodating a plurality of stakeholder interests. By emphasising human limitations, this approach highlights the desirability of strategies for coping with uncertainty and poses a challenge to the culture of mastery-via-understanding and grand central planning schemes.
How is this best achieved? The government should specify overall programme objectives, maximise transparency and information availability and devise ways to increase risk shifting onto competing project developers, who should have the freedom to specify, design, finance, build and operate projects to minimise uncertainty and achieve concomitant investment returns. In cases where natural monopoly infrastructure is better provided by the state, turnkey developers should bid to undertake projects at their risk. When investors bear more of these risks, they tend to go to much greater lengths to properly identify and understand uncertainties upfront and to devise appropriate strategies in response.
To date, the government’s response to the crises in infrastructure provision has, with one or two promising exceptions, not adequately engaged with these fundamental problems. The Eskom, Transnet and Department of Public Enterprises investigations tend to focus on the details of the project execution and do not address the fundamental information asymmetry and the incentive problems that bedevil state project delivery.
The National Energy Regulator of South Africa Energy’s (Nersa’s) response is to do after-the-fact cost studies with the threat of disallowing past asset expenditure that is deemed "not prudent" when it sets tariffs. This will do nothing to reduce the problem of cost overruns and merely shifts costs from consumers to taxpayers. Further, the legal basis on which Nersa might disallow these costs tariffs is doubtful. Overall, this response increases regulatory risk, will contribute to further SOE credit-rating downgrades and reduces the financing capacity of the state sector. More fundamental changes will be required.
Without great fanfare, the Treasury and the Department of Energy have implemented an important alternative with their Independent Power Producers (IPPs) procurement programme. Competing project developers provide price bids for providing renewable energy. While achieving similar levels of price competition for conventional power will pose greater challenges, they are also sensibly forging ahead with preparing a conventional power IPP procurement programme. This initiative represents the most fundamental break from South Africa’s power generation policies that have been in place since Eskom was established in 1922.
Private sector provision of infrastructure is not a panacea that relieves the government of its responsibilities. Critically, the state has to provide the coherent policies, market design, procurement programmes and regulation to ensure private provision of public infrastructure that meets society’s needs. The time for toying with lame-duck "developmental state" rhetoric has run out. South Africans need the government and stakeholders to act urgently to establish an inclusive infrastructure roll-out programme that leverages all the resources available to us.
• Steyn is an infrastructure economist with Meridian Economics.