So what does central bank independence mean? What does South Africa’s inflation target framework actually mean for the poor? Why is the content of the public protector’s recommendation (appropriate or otherwise) worth debating?
The importance of monetary policy
It is hard to exaggerate the importance of monetary policy (the setting of interest rates at which you and I can borrow money) for the direction of the economy. It determines the availability and price of credit in the economy. That has a decisive effect on household investment, i.e., private sector housebuilding, which drives demand for everything from construction to concrete; for consumer credit, and hence cars and furniture and appliances; for farm credit, and hence farm investment; and on companies’ decisions to invest, since any investment must make more than the cost of capital to be viable, so the costlier capital becomes, the less investment there will be.
Raising or lowering the interest rate by a percentage point ripples across the economy with enormous effect. When borrowing is difficult and rates are high, policy is said to be “tight”, or “hard”; when easy, policy is “loose”. Determining whether the long-term bias of rates will be towards tighter monetary conditions (higher rates) or looser conditions (lower) exercises an effect on the economy several orders of magnitude greater than almost any other discrete policy decision.
This outsize importance is reflected in history. The original Progressive movement was founded on opposition to tight monetary policy. Neoliberal economics began in monetary policy, and always considered it to be the most powerful of all levers in an economy. The most enduring and powerful intellectual currents of the left and right not only began in monetary policy, but held it as their most important battleground. Corruption must be stopped, but monetary policy has an order of magnitude of greater importance for our economy and for our lives.
Central Bank independence and its mandate
The idea of “central bank independence” gained popularity from the 1970s onwards and became dominant in the 1990s. It refers to the idea that a central bank should have a goal (its “mandate”) set for it, such as keeping inflation or unemployment down, and then be left alone to try to achieve that goal, free of interference, but with a requirement to explain its decisions to the public. The crucial point is that independence is within the mandate.
This was clearly stated by Mark Carney, the Governor of the Bank of England, formerly head of the Canadian central bank, counted by many as among the best central bankers in the world. On the stage of the London School of Economics in January this year, he stated that the Bank of England follows an inflation target, but it does so because that has been set in legislation, as the preferences of the people. If the preferences of the people change, and that is changed in law, the Bank’s objective will change. There is little in the ideas of the world’s leading central bankers, or in the theory of independent central banks, to support the idea that advocating a change in the Bank’s mandate is interfering with its independence.
Of course, those who benefit from the existing mandate will invoke the spectre of “interference” when that mandate is questioned. But this is an attempt to ward off a legitimate debate, and hence must beg the question of quite why someone would be so very eager to avoid that debate.
Inflation targeting and the poor
The South African Reserve Bank (SARB) has a mandate to “protect the value of the currency”. It interprets that mandate to mean it must keep consumer price inflation within a range of 3-6%. In the South African economy at the moment that means, and has meant for the last 20 years, a strong bias towards tight monetary policy – a preference for keeping rates high, and raising them at the slightest risk of inflation, while waiting very long to reduce them.
With inflation targeting, the devil is in the detail – specifically, how inflation is calculated. SARB has chosen “headline consumer price inflation”, or CPI, as its target. That is calculated by averaging over a long list of goods and services, with weightings for each determining how much they “count” for overall inflation. The weightings are available online and easily accessible to anyone commenting on inflation. The present methodology is skewed towards the consumption of the middle class, not the poor.
So, South Africa’s headline CPI gives bread a weighting of 3 points out of 100. Car purchases have a weighting of 6 out of 100. Running expenses for private cars (not taxi fares, which are counted separately) have another 5 points’ weight. Private medical aid premiums have a weighting of 7 out of 100. The cost of domestic workers has a weighting of 2.45.
Within a rounding error, therefore, the inflation targeting framework cares as much about the “price” of domestic workers as it does about bread. While it is true that the “food” category overall counts for 15 points, much of that is spread across food categories that barely overlap with the consumption of the poor. On the other hand, middle-class goods – private medical aid, a car, a domestic worker, DStv, and other such goods and services – count for at least 25-35 out of 100, or almost 10 times as much as bread. Another 13 points is derived from rents, not of shacks in backyards, but of “flats, townhouses and houses” obtained “through private letting agents”.
It may be argued that the inflation framework does not directly target bread, but that generalised inflation will affect bread prices and therefore the mandate still affects bread prices. The US Federal Reserve, however, disagrees. It deliberately strips food out of the inflation index it tracks (“core CPI”). It does so because food prices are volatile for reasons having little to do with monetary policy. First, the demand for basic food is rarely affected by interest rates. One columnist suggested the absurdity that people would spend less on bread if they spent more on cars, though this may have been a joke. The second is that the supply of food is heavily dependent on weather conditions and hence on the size of harvests, at home and abroad. The monetary policy committee does not affect the frequency of rain.
There is one other argument for a relationship between inflation targeting and the price of bread. It is made in technical arguments, though rarely in the media. It is known as “expectation anchoring”. It relies on the idea that people’s beliefs about what inflation will be are self-fulfilling – they raise prices if they think prices will rise – but can be “anchored” by strict inflation targeting. Recent research has cast a great deal of doubt on the theory even as it applies to companies in small, rich economies; how it would hold in a large, diverse, middle-income country, among the makers and buyers of bread, is a matter more of faith than reason.
Inflation targeting and basket cases
None of the above holds if the economy falls into hyperinflation or economic collapse. Some have suggested this is the necessary result of any deviation from strict inflation targeting. There is, however, no evidence for such a claim.
The United States has never had a strict inflation target. Instead, the Federal Reserve (the “Fed”) has a dual mandate: both “full employment” and “price stability”. When these mandates conflict, the Fed’s Board must decide on the trade-off, and justify its decision to Congress. In 2009-12, the dual mandate enabled the Fed, led by Ben Bernanke, to implement monetary stimulus. In the same period, in 2011, the European Central Bank (ECB) – which does have a strict inflation target – tightened rates because of what turned out to be a temporary increase in inflation. The former decision is now widely praised, the latter considered disastrous. The divergence contributed significantly to Europe’s recovery being so much more delayed than the US’ after the global financial crisis.
The scale of the ECB’s mistakes before and after the crisis has been a principal contributor to the crumbling of the intellectual orthodoxy behind strict inflation targeting. That is not confined to developed countries.
In 2010, the Governor of the Bank of Malaysia, speaking at the Bank of International Settlements (central banks’ international association, or version of Fifa), said: “Prior to the recent financial crisis, the predominant view was that … the best role for the Central Bank … was to ensure that inflation remains low. This recent crisis has changed this perception. Maintaining price stability and financial stability is a means to an end. The ultimate goal is surely to achieve sustainable economic growth and prosperity.” Malaysia does not have a strict inflation target at all. Malaysia’s inflation rate has hovered between 2-4% for the last 10 years, while its GDP has grown by 5% per year.
A few decades ago there was a strong argument that a single mandate focused on inflation was beneficial to long-term growth. That argument has crumbled. But there was never any argument, or indeed any evidence, that a dual or multiple mandate necessarily leads to hyperinflation or even sub-optimal growth.
Inflation targeting and growth
There is evidence that an overly strict interpretation of the mandate can damage growth. The ECB example described above is one instance. There is also the case closer to home.
In 2006-8, the group of Harvard and Princeton economists who advised Thabo Mbeki’s AsgiSA task force strongly advocated relaxing the interpretation of the mandate. They pointed out that our Reserve Bank’s adherence to strict inflation targeting resulted, for complex reasons, in an extremely volatile currency that was often too strong for viable and durable export growth. Their recommendation was flatly rejected by the forces of orthodoxy in Treasury and SARB, in part based on a strict interpretation of the current mandate.
It is extremely difficult to see a path towards rapid growth that is consistent with the current interpretation of the mandate. If we are to reach 6-8% GDP growth, either domestic demand and investment must boom, or exports must. The Harvard team concluded that rapid export growth was almost impossible under current monetary policy. On domestic demand, an investment boom could in the medium term create enough supply in the economy to allow rising demand without rising prices. But in the transition stage a rapid rate of growth would encounter an economy with decades of deficient investment behind it. Under current orthodoxy, SARB would step in to raise rates sharply, damping demand before it gained steam.
As a third element, one of the key constraints on our growth is a banking oligopoly that earns excess profits and restricts entry. This is, again, not a radical view. It was expressed by the deputy head of the IMF, David Lipton, in a speech in 2016. High on his list of needed reforms he stated: “In the finance sector, there are only a few retail banks in operation, and their fees are high. Small enterprises have trouble accessing banking services, though there has been some improvement of late. Barriers to entry into the industry favour existing institutions.”
In its earlier responses to the draft public protector report, and in public pronouncements over the years, SARB has explicitly invoked its stability mandate in defence of its de facto capture by the large, existing banks. That this capture is a fact is, again, hardly radical, but a truism among technology startups. It is the reason why most financial technology incubators sit inside the existing banks – attempting to take them on from outside almost guarantees eventual regulatory shut-down by SARB. This capture does not necessarily flow from the mandate – many central banks with inflation targeting mandates are not nearly as captured – although it should be noted that higher interest rates mechanically raise banks’ profits. SARB itself has chosen to mount its defence on the mandate, and so has itself brought the mandate into debate.
The Mandate and the Constitution
To sum up: changing the inflation mandate is not interfering with the independence of the central bank; doing so would not make us a basket case, and is probably a necessity for long-term growth; and the current inflation target has little to do with the price of bread, or the lives of the poor.
It is true that the motivation for the public protector’s recommendation is opaque. Some of the protector’s specific findings, with references to state banks, are as incoherent and ill-informed as anything written by columnists in the last few days, albeit no more so.
But the fundamental issue is what the mandate is doing in the Constitution in the first place. We are the only country, to my knowledge, that has the mandate in the Constitution. In the 1990s it was understandable why reasonable and well-intentioned negotiators and officials would have accepted the argument for a strict mandate. But that could have been accepted, as in all other countries, in legislation. Instead, it was inserted in the Constitution, where it would be almost impossible to change. It was inserted through closed-door negotiations, at the back of the document. It has embedded, deep in the heart of our economic governance, a bias towards the status quo, as it existed in 1994.
Everywhere else, as Governor Carney mentioned, the mandate is set by legislation. Alone among our peers, in contrast to every developed country, the goal of monetary policy in South Africa is set not “by the preference and instruction of the people”, but by the cunning and fearmongering of entrenched interests 20 years ago. That cunning and fearmongering has been much in evidence again these last two days. It is hoped that a few facts may help us move past it. DM