The great monetary economist Milton Friedman declared “Inflation is always and everywhere a monetary phenomenon.” He wrote this in 1963, in the wake of two waves of European hyperinflations that followed each of the World Wars. Central Bankers relearned this lesson during the “Great Inflation” of the 1970s and 80s. Since then, this principle has been the underlying assumption of the inflation targeting policy frameworks of most Central Banks, the SARB included.
Unfortunately, many of us have no idea what “monetary” means in relation to “inflation”, other than it must mean that things cost more money, right?
Even more unfortunately, most of us (economists included) are ignorant of the fact that Friedman later clarified his position, to say that only persistent inflation is always and everywhere a monetary phenomenon. In the short term, supply shocks can also change the price level.
It is such an important clarification, but it’s not a snappy soundbite. It’s also strangely technical, as though this is a point of theory rather than a point of real world practicality.
Not all inflation is monetary
In fact, in practice, we all understand that inflation isn’t monetary most of the time. Prices tend to go up when something is hard to find, and come down when something is plentiful. For example, when apples are in season, they’re on special. When they’re out of season, they are not on special. That adjustment between discounted price and full price, which reflects the abundance and scarcity caused by the movement of the earth on its axis as it revolves around the sun, falls into the definition of inflation because it is an increase in price.
But today, simply because Friedman once said that inflation is a monetary phenomenon everywhere and always, you will find folks on Twitter declaring that the higher price of out-of-season apples is a function of rampant money-printing and consequent consumer exuberance. And worse, that the SARB is obligated constitutionally to raise interest rates to control all that monetary chaos.
Even thought it’s just that apples are out of season and hard to find. Monetary policy has nothing to do with it.
The shifting supply of a particular good (in our case, apples) is not the only example of “normal” inflation. Microeconomic theory is almost entirely a study of how demand and supply curves are generated and how they affect the price of a good. And even macroeconomic theory courses devote a chunk of time to analysing how aggregate demand and aggregate supply interact to affect the overall price level.
And while I definitely believe that those theories are out of date and need a hard refresh – there is still real truth and insight within them.
The inflation problem
Let’s go back to that “always and everywhere” monetary inflation. Friedman’s key insight is that one particularly pernicious type of inflation is a function of the money supply (which is why we call it a “monetary” phenomenon), and that this variety is especially persistent. Why? Because when inflation is a function of the money supply being manipulated, the dynamics are abnormal and absurd, and they overtake all the usual rules of demand and supply.
We want to avoid monetary inflation, for certain. That is well within a Central Bank’s control, as they ultimately control the levers of money supply.
But should a Central Bank be trying to control inflation that comes from elsewhere? If there is a shortage of wheat, and the price of bread goes up, is that inflation really caused by excessive debt creation in the banking sector? If there is a new craze for an energy drink like Prime, and the producer puts their prices up to take advantage of the demand, should the SARB intervene to dampen enthusiasm for energy drinks?
Of course not. A shortage of wheat is a legitimate supply shock. An intervention on energy drinks is a political decision (if the drink is considered socially unacceptable but still legal, it gets disincentivized fiscally with a sin tax; anything beyond that gets legislative bans or prosecutions for anti-competitive behaviour).
We do not need a Central Bank to be worrying about those kinds of interventions.
So how do we create the dividing line between inflation that needs to be targeted, and inflation that ought to be left alone?
We need to start by talking about “price”
Before we can get to a change in price, we need to talk about how a price is set. Economic theory spends a fair bit of time here speculating about utils and marginal utility and the derivation of a demand curve. But we can probably get away with saying:
- A producer of a particular good looks at their costs to produce that good, decides on a reasonable margin to charge for making it, and then they go out into the marketplace and try to sell their product to customers at their target price.
- If people don’t buy their product, it may be that their price is too high, and so they’ll reduce their margin on it a bit. If that doesn’t work, they’ll stop making it because it would be clear at that point that there is no market for their product.
- But if they do find enough buyers at or near their target price, they have a market. The producers will play with their pricing for a while (with increases and promotions) and see how it impacts sales. At some point, the price will settle at a point where the producer is happy to produce and the market is happy to buy.
- That is demand and supply in practice.
How then do prices change?
Two obvious ways:
- The producer’s costs go up (say because an unreliable power supply from the public utility means they have to work with generators and spend a fortune on diesel to run them). They increase prices because they have to. Economic theory calls this “cost-push inflation” when it isn’t calling all inflation “monetary”.
- Alternatively, the product becomes very popular for some obscure reason, and people are crowding to get it. The producer sees their product racing off the shelves, and lifts pricing to slow down the pull of the demand, and to make extra profit. They increase prices because they can. Economic theory calls this “demand-pull inflation” when (again) it isn’t calling all inflation “monetary”.
And then you get true “monetary” inflation.
Now this part may require some explaining. And I’m going to come back to a point that I have been making for years.
Money is not a thing – it is a measurement tool
When our producer was setting his price, he didn’t say: “I need 12 kg of this material; 14 yards of this other material; about three hours of time from my designer; 20 kilowatt hours of electricity; 3 litres of water and about a half gallon of fuel – and if if I put that all together, my target price is three grams of silver and a block of cheese.”
No. The producer took all those inputs and ascribed each of them a value that was denominated in monetary terms.
This is money’s primary purpose: to be able to price things. That is the extraordinary innovation of having a common “medium of exchange”. It takes all that guesswork out of relative value.
But like with most measurement tools, we get into quirky territory when we start to think about them too deeply. But we need to, so that I can give you an idea of what monetary inflation really is (people talk about it as money “losing value” – but I think that is unhelpful, and I will get to why).
So let’s talk about other measurement bases:
- A second of time is our starting point for all measurements of duration. It governs almost everything that we do every day. How do we define how long a second is in the first place? As I write this, it is the fixed numerical value of the caesium frequency, ΔνCs, the unperturbed ground-state hyperfine transition frequency of the caesium 133 atom, to be 9192631770 when expressed in the unit Hz, which is equal to s−1. None of us know what that really means. But it sets the time for every moment of our lives.
- A metre is our starting point for all measurements of length. How do we define how long a metre is in the first place? For a while, it was an actual metal bar kept somewhere in France. We’ve reset that a few times since. Today, we define a metre as the length of the path travelled by light in a vacuum during a time interval of 1/299,792,458 of a second (which, as you now know, we define as the fixed numerical value of the caesium frequency, ΔνCs, the unperturbed ground-state hyperfine transition frequency of the caesium 133 atom, to be 9192631770 when expressed in the unit Hz, which is equal to s−1).
Now, why have I given you that seemingly irrelevant information? So that I can illustrate with distance what monetary inflation is like.
The Alice in Wonderland world of measurement
Imagine you are going to run a 10km race. We know that the race is 10,000 metres long. You start running, and as you start, the definition of a metre starts changing. It’s changing because I am gradually increasing the definition of the interval of a second. You check your watch, and it feels to you like time is moving more slowly, and the race is taking forever. Your body is telling you that you should be at the halfway mark, but your Apple Watch tells you that you’ve only been running for 5 minutes and that you’ve barely run your first km. You keep running because you’ve run so many 10km races in your life. Time ticks more slowly, and the race becomes endless. You die of exhaustion. According to your watch, you were only 20 minutes in.
A variation of this would be where the race ends at a finish line that is 10km away. As you’re running, I keep shifting out the finish line to reflect the change in the measurement base of a metre.
This should give you a sense of the loopy world of persistent monetary inflation. The normal rules of demand and supply set the start line and the finish line of the race. But somewhere in the middle, the money measurement base gets eroded, and everyone seems to lose. The consumer gets less than they should have and the producer earns less than they should have. Who secures the “true” value that was lost? It is the government that got a discount on the real services received through spending the money that it printed for free. It is the borrower that took the money to buy an asset, and the measurement base of their debt makes it start to disappear in real terms.
And we do start talking in “real terms”. Because when 10km no longer means 10km, we may start to talk about 6.21 Roman miles. Or the base distance between other more fixed points. That is: we would find a more stable base of measurement.
But we would never say, in my race scenario above, that the metre is “losing distance” when I start to debase it. That would be an oxymoron. The underlying base reality of the physical space between objects would remain the same. The metre itself would simply be losing its value as a measurement of distance.
And that is the same with money. We must be careful when we say that money “loses value”, as though money is both a measurement tool for value and that it holds value in itself. This is confusing, and conflates “value in use” with “value in exchange”. Rather, we should say that money is de-based. It therefore loses its usefulness as a measurement of relative value.
I also want to contrast this idea against folks saying that money has “lost its buying power” over time.
Let’s avoid saying “buying power” too much
Changes in “buying power” are also treated as “inflation”, and from my perspective, it’s something completely different.
If you go and look at the exchange rate of the South African Rand against the United States Dollar going back to the 1970s, you will see a strong and steady depreciation of the ZAR over time. It’s a graph that the cryptocurrency community love to hold up as evidence of the failure of our current banking system.
But this is just a misunderstanding of how economies grow over time, and what that means for a measurement base that is fixed in nominal terms.
I’m going to go back to using distance, because that’s got a lot less baggage attached to it. In ancient times, a “cubit” was the common measure for length, and it was calculated as the distance from the elbow to the tip of the middle finger. The first question you might ask is: “Okay, but whose arm?”
In some cases, this was set as a “royal” cubit, referencing the arm length of the ruler. So, we might have started with a cubit that was equal in length to elbow-arm length of Julius Caesar, and then it would be reset by each new Emperor that followed.
In general, if every new ruler was an adult man at full maturity, you’d be within a general range. And if you had nearby kingdoms following the same principle, it would all sort of roughly work out.
But imagine if hereditary rules meant that a new ruler took power as an infant. The cubit would be a baby’s armlength. And the measure of distance for his reign would be a bizarre anomaly, and completely out of touch with the kingdom’s neighbours.
This is an absurd example, but it gives you an idea of what happens with national currencies. At some point in time, when a currency is first introduced, its first exchange rate is set. The value of that money is fixed in pure numerical (or nominal) terms. Only, you get some countries that are already matured, and their economies grow very slowly. And you get some countries that are still quite young and developing, and their economies grow quite fast.
As those economies progress from that arbitrary point in time, their differences can compound into long-term deviations. Developed economies that grow at similar rates to each other tend to have stable pricing internally and consistent exchange rates to each other. Developing economies have “inflation” internally and depreciating exchange rates relative to stable countries. But this is not a bad thing – it is a consequence of healthy growth.
You might mistake that long-term adjustment for inflation.
It’s just that there are more things to measure
We forget that “money supply” is just a short-hand way of saying all the things that money is needed to measure. And developed economies have simply already done the work of “unlocking” the value of their resources.
Here’s an example to help:
- In an underdeveloped economy, I am a subsistent farmer, and I work on a piece of land. I have $10 in my pocket to pay for seed. $10 is my contribution to the money supply.
- As the economy develops, I go to a bank, and ask them to value my piece of land, so that I can finance my farming production against it, and thereby be less subsistent and more entrepreneurial. They value my land at $10,000 and give me a $5,000 loan against it in my bank account. I still have $10 in my pocket. My contribution to the total money supply is now $5,010.
- The land is still the same land. But its value has been unlocked for the purposes of growth and development, which has a direct impact on the money supply.
Over time, that process of growth drives the relative pricing levels to re-adjust upwards. Both salaries and prices go up. This is “inflation” in the sense that pricing has adjusted relative to an earlier starting point. But if you can buy the same things with your income, then you are no worse off.
All the domestic inflations
At this point, we’re left with inflation being all the following:
- A consequence of a producer facing higher costs to produce
- A consequence of consumers being exuberant
- A consequence of underhanded debasing of the money supply
- A consequence of the process of unlocking value that comes with economic growth
Already, it should be so clear that we desperately need more words than “inflation” to describe these economic phenomena that are all indicated by an increase in nominal prices. They are very different.
And I haven’t yet accounted for the fact that we are only talking here about domestic matters. Because we now live in a world of international trade and relatively free flows of capital.
Imported inflation
You know the line “America sneezes, and the rest of the world catches a cold?”
When the Developed World initiated quantitative easing and near-zero interest rates in the wake of the 2008 subprime mortgage crisis, that extra liquidity spread out as international capital into emerging markets, strengthening currencies across Latin America, Africa and Asia.
When QE ended, the flows reversed. Almost all emerging market currencies crashed, sparking waves of price adjustments and (if we’re honest) general political turmoil.
We are now in the second interation of that policy reversal, with the US, the UK and the Eurozone (and other developed countries) raising their interest rates to dampen their own domestic inflation. As that happens, international capital is flowing back there to secure the higher interest yields, creating general currency weakness in emerging markets. Central Bankers in the developing world are forced to raise interest rates in lockstep to stave off currency depreciation and prevent import-cost driven inflation.
So even if South Africa’s present inflation is ultimately a monetary phenomenon driven by excess liquidity – who created that excess liquidity, exactly? Can a South African Central Bank realistically tackle the fallout of the Federal Reserve’s efforts to control an American monetary phenomenon?
Developing countries like South Africa face an appalling choice:
- Raise interest rates to control imported inflation, and risk political instability caused by cost-of-living crises due to excessive debt burdens
- Accept imported inflation, and risk political instability caused by cost-of-living crises due to high costs of key imports (like fuel)
- Directly manage the exchange rate, and get cut off from international capital investment, thereby abandoning the economic growth compact and accepting a decline into nepotistic authoritarianism
The five phenomena that produce inflation
Now we’re up to five different types of inflation:
- Cost-push inflation where a producer facing higher costs to produce
- Demand-pull inflation where consumers are being exuberant
- Monetary inflation caused by underhanded debasing of the domestic money supply
- Economic growth inflation caused by the process of unlocking value that comes with economic development and financialization
- Exchange rate inflation caused by global capital flows, often driven by excess liquidity in dominant reserve currencies.
This is where our language lets us down. Inflation ought to be a neutral phenomenon. We should rebrand it as “price adjustment”, which is a much less weighted term.
But whether we come up with new words or terms, surely it is clear that any “inflation-targeting” framework really needs to be careful to discuss which inflation it is targeting?
If I can give one more analogy here, we could use “belly distention” as the key indicator for obesity, and then adopt a calory-controlled diet as a policy response. Only, belly distention in children is also a symptom of kwashiorkor, a disease caused by severe malnutrition. A calory-controlled diet is exactly the opposite of what is required in that case.
Appropriate Central Bank policy responses for each type of inflation
- Cost-push inflation is (usually) a non-monetary phenomenon. When it happens at scale (like changes in the oil price, which affects all producers at once), it is still a non-monetary price adjustment. And any demand-pull inflation that is seasonal is also non-monetary. Any intervention on these is really a fiscal policy question. The Central Bank has no business implementing fiscal policies and should stay out of it.
- Economic growth inflation is fundamentally positive and long-term. Any intervention by the Central Bank would be depressive, and it’s usually avoided (generally by the establishment of a target band of low and stable inflation).
- Exchange rate inflation interventions are fundamentally exchange-rate targeting regimes, and we should call that type of policy what it is. It’s also dangerous and you usually need to break your economy to sustain it. We should debate that on its merits.
- Only monetary inflation is under the Central Bank’s direct influence, and should be their predominant concern.
And here is my main gripe.
The most recent string of interest rate hikes has been overtly described by the SARB as being focused on shoring up the exchange rate to prevent further inflation.
This is not inflation-targeting. That is exchange rate management.
And we should be clear about that.
Because managed exchange rates are a whole other kind of policy intervention. Especially as history keeps demonstrating that managed exchange rates don’t work, and that they’re extremely damaging when they inevitably fail.