The real value of forecasting is not measured by decimal points

The focus is on understanding how economies change in response to a kaleidoscope of forces

Danish physicist and Nobel laureate Niels Bohr is credited with having said prediction is difficult, especially if it is about the future. This statement could not be more applicable than it is today given the significant changes to the functioning of the global economy.

Those like me who forecast matters economic are often criticised — unfairly, I maintain — when our predictions turn out wrong, but the real value of forecasting is not in their precision. This article is my attempt to explain what economic forecasting is and what it is not, and why recent forecasts relating to the SA economy have tended to miss the mark, at times by wide margins.

Let me begin with the global factors that complicate the job of forecasting. The global economy is a complex system, comprising heterogeneous economic role players that cannot simply be lumped into one. Countries’ economic structures differ in terms of natural resource endowments, human capital development, technological advancement and geographic location, among many other factors. All of these affect the performance of an economy.  

Major supply-side factors can also disrupt the functioning of an economy. The Covid-19 lockdowns and Ukraine war are major supply shocks that have possibly changed the structure and functioning of the global economy forever. Many of the assumptions we took as given before the Covid-19 pandemic and Russia’s invasion of Ukraine are no longer valid. All need reassessment. 

For example, poorer countries could not get Covid-19 vaccines as quickly as they would have liked, even though some rich countries stockpiled more vaccine doses than they needed. The war in Ukraine was also a wake-up call to many countries in Europe and elsewhere. Much of Europe had come to depend on oil and gas supplies from Russia, but when Russia was slapped with sanctions in response to its aggression in Ukraine it retaliated and reduced the supply of gas to Europe.

Friend-shoring

The breakdown of long-held understandings and social contracts affecting how countries relate to each other changes how countries approach investment in critical sectors of the economy. There is now a realisation that in certain key areas, such as health, there is a need to build domestic capacity even if that might not be efficient financially.

Another way to achieve this goal is what is now referred to as friend-shoring, which entails bringing the manufacturing of key products or manufacturing inputs closer to home, to countries that are more reliable friends. This is less efficient than when globalisation was in full force and many goods consumed in America could be produced in, say, China or Vietnam, where the cost of labour was low. Friend-shoring is a compromise solution in a world where such relationships no longer guarantee security of supply.  

Policymakers have distorted how the economy should work. Phenomena such as negative interest rates and quantitative easing were anathema before the global financial crisis. Their introduction has changed how the global economy works, along with the basic assumption that inflation was never going to rise aggressively again and interest rates would remain low forever.

As we now know, this was not to be the case in a world after US tariffs imposed on China and after Covid — and now in the midst of a European war. Global value chains are changing as are inflation and growth drivers. Monetary and fiscal policy will have to continue to change going forward.  

These are some of the global dynamics that make it difficult to forecast how things will unfold for the domestic economy. There was no way to anticipate many of these events, and once they were understood they necessitated a new approach to the way of doing things.  

With this complexity globally the domestic economy also became more complex to forecast, simply because SA is a small, open economy that is greatly influenced by global developments. Being small does not mean less complex. It simply means the domestic economy is far more exposed to global shocks and responds more dramatically than a larger economy would. This unpredictability and complexity makes it even more difficult to make accurate economic forecasts.  

Then we come to domestic factors. Take February and March’s inflation outcomes, which caused many economists to question their assumptions. For both months inflation surprised on the upside, the unexpected part mainly concerning food prices. While the historical experience has been that domestic food prices tend to follow global trends, that has not been the case in recent months.

Unusual factors

Global agricultural input and commodity prices have moderated and food price inflation has generally followed suit. Yet in SA food prices have continued to rise. Part of the story is the cost of energy in SA; agribusinesses in SA’s food value chain need to recoup the cost of diesel and generators by passing them through to consumer prices. There is no exact way to estimate the impact of load-shedding and incorporate this into economic forecasts.  

When economic forecasts diverge significantly it is usually because of an unusual factor that does not have a standard way of being incorporated into economic models. For the first time in a long time the National Treasury and SA Reserve Bank forecasts for economic growth in 2023 look quite different. In January the Bank forecast economic growth of 0.3%, while in February the Treasury predicted growth of 0.9%. The difference is assumptions around load-shedding and its impact on the economy.  

The structure of the post-Covid SA economy has also changed. Seasonal factors in the data have been affected by a change in consumer behaviour, which is not yet captured in the economic models that are used for economic forecasting. Consequently, for some time to come many predictions are likely to miss the mark.

In the meantime, more focus will have to be placed on understanding how the economy has changed and how different sectors relate to each other, rather than fixating on the decimal point accuracy of forecasts. Economic growth forecasts of 0.1% for SA (IMF) or 0.2% (Reserve Bank) both mean the economy is stagnant and the risk of recession is high. Inflation forecasts of 6% or 6.1% both mean inflation is too high, well above the midpoint of the Bank’s target range.  

It is always good to get a forecast right, but that is not where the real value of the exercise lies. The true benefit of economic forecasting is in explaining the dynamic interaction between different economic role players and sectors of the economy. It is more useful to get the direction right than to be accurate to within a decimal point.