Governor Lesetja Kganyago: Public lecture at Wits School of Governance

ublic lecture by Lesetja Kganyago, Governor of the South African Reserve Bank, at the Wits School of Governance Johannesburg
Keeping it simple: monetary policy, growth and jobs in South Africa
Good morning, ladies and gentlemen.
Central banks are important, long-standing expressions of a universal need for stability in social and economic affairs. Their goals centre on achieving some definition of price stability, and in more recent decades, their methods have fixed primarily on inflation targeting.
Where they directly target inflation, central banks’ primary tool is the policy rate, normally defined as a very short-term or overnight borrowing rate. At this rate, banks can borrow from the central bank to address overnight needs for liquidity and this marginal borrowing rate sets the basis for all other lending rates in the economy. A secondary policy tool is a blend of communications about current economic conditions and the policy rate level. A third encompasses the requirements and flexibility of the policy framework – the target itself and how it is measured.
In more recent years, and under the impact of deflationary conditions, a few advanced economy central banks have adapted their approaches with secondary targets, such as inflation averages over time and unemployment rates.
Today, I want to spend some time unpacking why it is useful for the South African Reserve Bank (SARB), and most emerging economies generally, to target inflation, and why achieving that goal ensures that monetary policy includes economic growth and job creation.
I will look at recent changes in policy targets for the moment, before returning to current conditions and the relevance of jobs and growth targets for monetary policy.
Financial distress, deflation and policy exceptionalism
It is useful to start with the global economic shocks of the past 15 years and their monetary policy implications.
While the global financial crisis (GFC) and the COVID-19 pandemic had different origins, the economic policy responses to them were similar. Advanced economies reacted to both by dropping interest rates as far as possible and using quantitative easing (QE). During the GFC, QE programmes kept asset prices of various kinds higher than they would otherwise have been, preventing asset price deflation from causing even worse economic outcomes. During the pandemic, QE additionally supported spending, as mobility and many face-to-face economic transactions were curtailed, while also protecting financial institutions.
Job protection was also a major focus for major central banks and job creation became a measure of their policy effectiveness. More liquidity would help keep interest rates low, enabling firms to keep paying wages and to restart the economy.
What impact have these efforts made on policy frameworks? Certainly, where countries faced collapsing growth and weaker inflation together, they could all move policy in the same direction. However, when the GFC ended and the pandemic ended, not all faced the same policy trade-offs. The GFC ushered in an extended period of ongoing QE and low rates, but this was not true for most emerging market economies. The end of the pandemic has been different. It has thrown us all back into a pre-Great Moderation world in which inflation is super-sensitive to supply and demand shocks.
In particular, and with the benefit of hindsight, the current global inflation finds much of its origin in a too aggressive use of QE and in negative real rates as the pandemic started to wind down and economic activity rose. As in emerging economies post- GFC, many countries now, post-pandemic, find their output gaps to be badly measured and giving off incorrect signals about their policy stance.
As, and if, the current surge in global inflation wanes, some advanced economies may very well return to lower inflation for structural and demographic reasons. These central banks may retain other metrics in their policy frameworks – such as economic growth rates or the prices of specific assets (e.g. house prices in New Zealand), or in the case of the United States (US), employment levels.
But it is highly unlikely that emerging market economies will decide to change their policy frameworks. Many had reverted to much higher inflation even before the pandemic eased, while others, like South Africa, are now caught up in broader inflation. In that more normal context, the standard monetary policy approaches make as much of a positive impact on economic growth and jobs as they can. However, the real solutions to faster growth and job creation lie in other policy domains.
Why the difference between advanced and emerging economies? Why might the rules governing objectives differ somewhat? The answer has to do with pure economic theory, actual experience, and with the different economic conditions − cyclical and structural − faced by different central banks.
To begin, it should be clear that for quite some time, essentially since the early 2000s, central banks of advanced economies have faced stubbornly low inflation, despite low interest rates, for a range of structural reasons related primarily to demographics and high savings. So, when they took an accommodative stance to raise growth and employment levels during the Great Moderation of the 2000s and the period into the pandemic, inflation still remained modest.
As the pandemic hit, because inflation remained low, there was as yet no contradiction between the inflation targets and boosting growth. Policy expansion could help get people who had dropped out of the labour market back in. Job creation was efficient. Expansionary stances worked well with strategies that lowered the cost to firms of retaining people in jobs.
In South Africa’s case, the pandemic also facilitated expansionary policy, precisely because inflation had trended lower from 2017. This allowed us to lower policy rates sharply in 2020 to confront the shorter-term damage done to the economy. The repurchase (repo) rate averaged 6.7% between 2017 and early 2020, before dropping to 3.5% in March/April of that year.
A deeper dive into jobs and growth targets
As in advanced economies, our expansion did little to immediately bring back jobs. Many were lost as lockdowns were extended, while some new ones were created in various services linked to the shift in consumption patterns. Expansion did support the recovery of both pre-pandemic spending patterns and many of the jobs associated with them. Of course, there is also much further to go, as some sectors remain constrained.
The open question is whether sustained expansion in an environment of high debt levels and rising inflation could live up to the hopes of those that argue for a ‘jobs’ mandate for monetary policy. The short answer is ‘no’, but let me explain why.
Our basic problem is that while growth creates job, inflation does not. Not only does this fatally challenge the belief many hold in the existence of usable Phillips curves, but it also limits what macroeconomic policy can achieve in terms of job creation. As we have noted many times in the past, the solutions to our unemployment problem lie well outside the realm of monetary policy, and in fact the failure to employ those solutions directly limits the positive contribution monetary policy can play.
Let’s look at what the historical data tell us about Phillips curves. We see two correlations. One is that as inflation rises, unemployment rates rise. This characterises the late 1990s and into the latter half of 2003. In 2002/03, for example, inflation reached double digits even while employment was falling – when unemployment breached 30%.
From about 2003 to 2007, however, we see another correlation, where inflation falls and employment rises. After the GFC, from about 2011 to 2019, we see something different. Inflation first came off the highs of the GFC and then accelerated back up to around 6%. There was some initial recovery in jobs, but as time went on, the acceleration was increasingly located in the public sector rather than the private sector. Economic growth weakened quite sharply from 2013 to 2015 and then more gradually slowed through to the pandemic. From 2017 onwards, inflation decelerated and so did job creation.
The GFC and the pandemic were relatively clear instances when policy could respond in textbook ways to support the economy and see-through inflation shocks. Before and after these crises, we see more transparently the longer-term relationships between inflation and job creation. First, inflation does not create jobs. And second, on balance, expansionary policy prompts more inflation than growth or job creation.
This tells us that South Africa’s Phillips curve is near-vertical at a low rate of positive economic growth.
This is strong evidence that the basic job creation mechanism is being impeded by things other than aggregate demand.
This adverse relationship between policy expansion and inflation kicks in when employment levels rise above what is called the non-accelerating inflation rate of unemployment, or NAIRU. Evidence for South Africa shows that this happens when our unemployment rate is still very high. This is exacerbated by level changes in the NAIRU caused by structural impediments to job creation. For instance, load-shedding or higher transport costs indirectly increase the cost of creating jobs and push up the NAIRU.
A range of long-standing factors contribute to our very high NAIRU. Perhaps the most important of these is where people live and the cost of supplying their labour, and the skills they have compared to their cost.8 Regulatory costs of our economy also feature in various ways, including housing, where businesses can ply their trade, and compliance − which is a much larger burden for small and medium-sized businesses compared to that of larger firms.
Part of the difficulty is that we have done little to make it easier for less-skilled workers to find jobs, or to make it less costly to employ them.
Another part of the failing job creation machine is simply weaker real economic growth compared to the 2000s, when job creation was quite strong.
This should not, however, lead us down the path of grasping at another seemingly ‘easy’ answer. Higher inflation will not give us higher sustainable growth.
Instead, higher inflation undermines short-run growth by increasing interest rates on borrowing, affecting consumers’ buying on credit and business owners who want to use credit to invest. Higher interest costs reduce short-run cash flow, reducing all future consumption spending.
While surprise inflation reduces the real value of the existing stock of debt owed, the trade-off is lower economic growth in subsequent years. The short-run benefit of a surprise lower interest rate is transformed by higher inflation into a long-run cost to growth.
When inflation is higher than that of our trading partners, we suffer a continuous loss of competitiveness. High inflation overall also generally means higher inflation for poorer and less-skilled South Africans than for the wealthy. This increases inequality further and worsens the already low standard of living for those households, making them costlier to employ.
For a few years before the GFC in 2008, the relationship between growth and employment was better – if economic activity grew by 1%, employment grew by around 0.62%. But since then, up to 2018, each percentage point improvement in growth only gives us 0.37% more jobs. Our low employment problem overlaps entirely with our low growth problem.
In this context, the best a central bank can do is stabilise unemployment at the rate consistent with price stability. If a central bank attempts to get unemployment below the NAIRU, the result will be larger quantities of inflation but only small and temporary quantities of jobs as the supply curve becomes more vertical. The same is true for economic growth.
Assuming that most of our unemployment problem is structural, are we at least sure that the residual cyclical unemployment is being reduced by monetary policy? Is a dual mandate better at addressing these cyclical drivers?
At the SARB, we use an alternative measure of economic activity to the NAIRU for understanding where the economy is relative to its tipping point into more inflation or into deflation. This is the output gap, or the distance between the possible, or potential, growth rate of the economy and the actual or realised rate of growth.
Similar to the US Federal Reserve (Fed), we use a Taylor rule to help tell us what the interest rate level should be, given the output gap and the distance between the inflation rate we want (the target) and where it currently is. 13 The Taylor rule tells us that we should set an interest rate that includes how quickly or slowly the economy is growing, compared to the speed at which it could be growing.
If the output gap is a good representative measure of the cyclical unemployment rate, then we are fully covered – the monetary policy equation we use to get to the interest rate includes, as best we can, the relationship between those rates and employment. If we look at how the Fed operates, we can take further comfort in our own framework. The Fed has an inflation target of 2%, averaged over time. This averaging introduces some flexibility to not respond to inflation shocks that are temporary, just as in our flexible inflation-targeting framework. But the Fed framework does not say that when unemployment rises it will aim for a higher inflation rate. In other words, the dual mandate recognises that in the long run, the highest cyclical employment rate is consistent with low inflation. In the policy debate at present in the US, there may be a rise in cyclical unemployment to get inflation back down and to achieve a lower sustainable unemployment rate.
We need also to be cautious in our current circumstances. With surging inflation in advanced economies, it is likely that the adjusted policy frameworks of recent times will shift back to simpler and clearer formulations.
Three takeaways from this discussion stand out. One is that inflation-targeting central banks also consider real variables such as growth and unemployment when they make monetary policy decisions. Second, that central banks don’t drop targeting inflation, even if they have an employment mandate. It simply means they respect the NAIRU and discuss more directly its level and what can happen to inflation when the speed limit implied by it is exceeded. And third, that despite some policy innovation in recent times in some advanced economies, these may very well be scaled back to simpler frameworks.
Finally, we need to recognise, for the sake of solving our low employment problem and to keep monetary policy focused on the right things, that many issues sharply limit the transmission from policy to job creation outcomes. One is the job intensity of growth which we discussed. Another is the supply of skills that are coming into the market, and their cost.
Some might argue that in that case, it means that monetary policy should be pushed much further and harder to get the expected growth or jobs. Unfortunately, when inefficiencies and constraints exist, pushing harder on monetary policy is like pushing the accelerator to the floor on a curvy, icy road over a mountain pass. At present, many central banks are skidding on that ice, with global inflation sharply higher and persistent. Even our own relatively benign recent inflation experience has become much more challenging very quickly.
Keeping it simple
We have seen now that having two targets certainly does not mean double the benefit. Instead, it means that there are times and certain conditions when one policy tool helps to achieve both. It also means that under other conditions, one tool cannot achieve both, and much of the world has now entered this territory.
When policy becomes overloaded with too many and contradictory objectives, then negative outcomes are more likely.
As inflation rises and growth slows, a central bank that fails to respond to rising prices will face the prospect of compounding inflationary shocks. Currencies depreciate and investment falls. 14
So where, in all of this, does South Africa find itself? We have not reached the ‘end’ of our policy rate space. We do not have stubbornly low or close to zero inflation. We are experiencing rising inflation rates. Inflation expectations, for the most part, are proving to be more responsive to current inflation outcomes than we would like, and less anchored around the midpoint of our target.
If the expectations that firms and households hold for future inflation stray from the inflation target, then higher nominal wages and consumer prices are likely to emerge. In other words, once inflation expectations rise, they become a self-fulfilling prophecy, and the central bank has a more difficult and longer-term problem on its hands. A more expansionary policy stance runs that risk by enabling first-round inflation to embed into second-round inflation.
This implies that we need to continue the normalisation of interest rates, moving them closer to the level that is consistent with more stable inflation rates and sustainable economic growth. At present, our repo rate is at 6.25%, still below long-term levels, but rising to a more sustainable long-term level that is consistent with inflation stabilising at 4.5%.
As we have discussed today, neither growth nor high inflation lead directly to job creation in our economy. Much of our employment challenge lies in encouraging the return of economic activity in sectors that have been hardest hit by the pandemic. And we will also need to recognise that some jobs may be permanently lost as firms do things differently and as consumption patterns shift. Other jobs will be gained, however, and permanently impacting on employment levels requires approaches that have nothing to do with monetary policy. Just adding jobs targets will not get us there, and indeed part of our inflation problem stems from efforts to achieve such targets at a global level.
We should be careful not to add to our policy objectives in a way that surely would push us into sharply higher inflation.
If these are the primary constraints to job creation defined by the literature, and monetary policy has no impact on them, then the claim that more expansionary policies will solve the unemployment problem is simply an empty promise, backed up by little more than ideology and wishful thinking.
We have an unemployment problem that needs more credible solutions.
Employment and growth are both limited by factors that are beyond the reach of the central bank’s toolset.
The best chance we have with monetary policy to get faster, more job-rich growth is to maintain our focus on price stability with flexible inflation targeting − a proven framework. This enables the SARB to help maintain a stable environment that is conducive to economic growth, and because credibility is high, for it to create the necessary flexibility to ignore short-term inflation shocks.
A credible monetary policy will also keep borrowing costs lower than they would otherwise be. This is a central benefit to long-term economic growth and job creation. When inflation rises and stays high, investment decisions are distorted towards short- term investments that carry with them short-term jobs. For this reason, low inflation is a sound developmental policy. It encourages firms across the private and public sector to make long-term investment decisions that imply productivity growth over time. This is critical, indeed a prerequisite, for sustainable jobs and income growth.
Consumption based on rising debt levels cannot be a sustainable growth and development strategy.
Having faced the unique threat of the COVID-19 crisis, we confronted that challenge with relatively low stable inflation and policy rate space. We were able to soften the damage of the crisis with the policy rate, while still protecting the value of the rand, and in so doing, were able to play our part in maintaining macroeconomic stability.
Now that the global economy is recovering and inflation in many countries, including our own, is rising, we have learned from experience that we must not be tempted to loosen our grip on inflation, or to fall behind our peers as rates are normalised – the consequences would be too costly.

Source: Government of South Africa