Comparing apples and oranges – the case for regulatory impact assessment

5th November 2013

Policymaking is difficult. It involves opportunity costs, trade-offs, and unintended consequences. Ideally the process should be open and transparent, and should include a vigorous debate on all aspects of the policy being developed: its purpose, cost, mechanics, alternative designs, the likelihood of success, and the like. This is the only way to ensure that fair and balanced policies are developed that take into consideration the welfare of all South Africans and not only that of narrow interest groups.

In South Africa the policymaking process often does not live up to these ideals. While extensive stakeholder consultation is often involved, the parties consulted are not always privy to all the information, arguments and assumptions feeding into the policy design. This is not ideal as ‘stakeholders’ is not a homogenous grouping and policies typically benefit some stakeholders whilst negatively impacting on others. Without seeing the actual facts and figures used, it is not possible to know whether stakeholders’ inputs have been interpreted correctly, or what weights have been applied to different arguments.

At present there are a number of policies under consideration that illustrate the intricate nature of policymaking. Government has just announced an increase in the tariff on imported chicken. This is good news for chicken producers (since imported chicken, which is a substitute to locally produced chicken, will now be more expensive), but not so great for chicken importers and consumers who will now pay more for both imported and (probably) locally produced chicken. In addition, the Minister of Rural Development and Land Reform is contemplating ways of addressing the impact of sugar imports on local production. Again, if import tariffs end up being the preferred instrument to address this issue, it is probably quite appealing to upstream producers of sugar (and their suppliers and employees).  This will however lead to an increase in the cost of sugar to downstream users (like the food and beverages industry) and final consumers.

Similarly, the Minister of Economic Development recently issued a policy directive that prevents ferrous and non-ferrous scrap metal from being exported unless it has first been offered to local scrap metal beneficiators, at a discount to international export prices. Local users of scrap metal, and downstream consumers of products manufactured using scrap (like iron and steel), are justifiably upbeat. Scrap metal collectors and merchants, however, have reason to be concerned since they will now receive lower prices for the scrap metal they sell. Both sides have raised a number of reasons for supporting or objecting to the regulations. The two relevant studies that are in the public domain, one by Econex and one by Conningarth Economists, indicate that the export restrictions may do more harm than good. [1] Presumably the Economic Development Department has conducted its own analysis indicating that the policy will be beneficial to the economy. [2] Crucially, however, how the arguments of the different sides, and the studies supporting these, were weighed up has not been disclosed. What is clear is that this would have been no easy task. Different assumptions and interpretations of how the world works, even when based on the same data, can lead to very different conclusions.

Fortunately there is a tool that is readily available to decision makers to assist with these complex policy choices (providing a coherent, transparent and impartial way of weighing the expected costs and benefits of new policies): the regulatory impact assessment (RIA). It is an evidence-based process widely used internationally to equip policymakers to assess the real consequences - both positive and negative - of government action. It ensures that laws and regulatory interventions actually achieve what is promised, while reducing harmful hidden costs and trade-offs. RIA is also a means of managing conflict between stakeholders by aligning policy interventions with existing policies and objectives.

The RIA process is designed to allow a critical analysis of all available and relevant data, whether quantitative or qualitative, to ensure that policy interventions have as good a chance as possible of meeting their objectives with minimal costs. RIAs thus encourage policymakers to consider the wider net costs and benefits of their proposals, rather than focusing simply on the narrow interest of certain stakeholders.

If integrated into the design of policy (as good practice suggests it should be), RIAs can increase the effectiveness of policymaking. RIAs place an emphasis on understanding the context in which a policy will be implemented, and the mechanisms through which it is expected to be successful. This removes ‘mental leaps’ in policy design. It is not sufficient to simply say what a policy will do (as is the case in the Government Gazette notice outlining the rationale for the scrap metal export restrictions and regulated prices), the way in which it will achieve these ends also needs to be spelled out. Once the mechanisms have been spelled out, they can be interrogated – which in turn leads to more informed commentary and more robust policies.

In theory the use of RIAs should be a common occurrence in South Africa. Cabinet approved the use of RIAs as far back as February 2007, and the Presidency issued guidelines for the implementation of RIAs in February 2012. [3] In the preface to the guidelines, the rationale for RIAs is clearly described (Presidency, 2012: 2):

RIAs are an extension of a broader commitment to the quality of government through evidence based policy making. The advantage of instituting RIA is that RIA adds structure, predictability, and methodological clarity to assessment while also ensuring that the right information is available for decision making.

The document calls for high-level RIAs to be undertaken for new primary legislation (unless prohibited by national security concerns), subordinate legislation, significant regulations and policy proposals, and when the intention to revise any existing legislation or regulations and policies arise. If significant costs of policies are found, this is to be follow-up with more in-depth RIAs. The Presidency also provides a number of resources to assist with the implementation of RIAs, from a RIA template to training materials (link here).

In practice, however, RIAs are relatively rare in South Africa, and are only implemented in an ad hoc fashion. Also, unlike the Presidency’s guidelines suggest, they are often implemented too late in the policy development process (when it is so far advanced that significant changes or delays cannot happen without a policy’s backers losing face) or are not made public (which means the assumptions underlying the assessment cannot be effectively interrogated).

Institutionalising RIAs as a normal part of the policymaking process will make policymaking a less adversarial process, and will help government to weigh the arguments of proponents and opponents of a policy in a common language. Hopefully this will replace the current model, whereby public relations battles are fought in the press, with a more informed and structured debate – and better policy outcomes.

Written by Brent Cloete, DNA Economics


[1] Summaries of the studies can be found online here and here.

[2] The General Notice (33 of 2013) in the Government Gazette of 25 January 2013 outlining the rationale the policy states a number of expected benefits from the policy, but does not refer to any research supporting the assertions in the Notice. Tellingly, it also does not mention any of the likely negative impacts of the policy.

[3] The Presidency: Republic of South Africa (2012) Guidelines for the Implementation of the Regulatory Impact Analysis/Assessment (RIA) Process in South Africa. Available online here.
Acknowledgement: The discussion of RIAs draws on the work and insights of my colleague, Sarah Truen.