What Is Inflation?

Setting the frame with first principles

You are reading Economic Forces, a free weekly newsletter on economics, especially price theory, without the politics. Economic Forces arrives weekly in the inboxes of 16,000 subscribers. You can support our newsletter by sharing this free post or becoming a paid subscriber:

---

When Brian and I set out to write a newsletter on price theory, I did not expect to write so much about inflation. Nonetheless, the events of the last few years provided a number of opportunities to write about inflation. Over that time, I have grown somewhat frustrated with the discourse around inflation. This likely isn’t surprising given the number of posts we have written on inflation. However, the fundamental thing that has bothered me has been to try to pinpoint why the commentary gets things wrong. Thus, today I would like to write about why I think so much of the commentary is lacking. Not surprisingly, I am going to argue that we need price theory to get things right.

What Mistakes Do People Make?

We typically define inflation as a sustained increase in prices, on average, over time. There is not really anything wrong with that definition. In fact, there is a lot to like about a definition like that. By saying “sustained,” this implies that this isn’t a one-time increase in prices, but rather that prices continue to rise. Furthermore, by referencing that this is an overall change in prices, this distinguishes what is being observed from relative price changes. When, on average, prices are going up, this suggests that this is distinct from a relative price change.

A simple thought experiment is sufficient to capture this distinction. In the United States, everything is priced in terms of dollars. If the U.S. government came out tomorrow and announced that they were adding a “zero” to the end of every denomination of currency (e.g., a penny is now worth 10 cents, a dollar is now 10 dollars, etc.), one would expect that when this change went into effect, and holding everything else constant, dollar prices would all go up by a factor of 10.

In reality, all else isn’t equal. As a result, not all prices would increase by a factor of 10. However, on average, we would expect prices to increase by a factor of 10. It is at this point that this approach to discussing inflation seems to go awry. Since all else isn’t equal, we don’t get such clean observations.

Some of the trouble seems to be about measurement. You see this when the latest consumer price index (CPI) numbers are released. Often, the commentary surrounding an increase in the CPI involves discussion about what prices went up the most and what prices changed the least (or perhaps even declined). There is nothing that is necessarily wrong with that exercise. However, the commentary often assigns causal language to this accounting exercise. For example, one might see commentary that says, “the CPI increased by 2% last month, driven by changes in automobile prices” (Get it? Driven by. Car prices. I’ll show myself out.)

The issue here is that commentary like this is focusing on accounting and, in that sense, the price of cars went up more than the other components. Nevertheless, it is not the increase in car prices that is causing inflation. In fact, if the price of cars is increasing faster than average, that is indicative of a relative price change.

This doesn’t mean that the accounting exercise is meaningless. Suppose that there is an oil shock and everyone is aware of the oil shock. Inflation, as measured by the CPI, is higher than expected. Policymakers might want to know how much of this change is explained by the dramatic increase in the price of oil.

Unfortunately, a number of people seem to use this accounting exercise as a way of understanding inflation more generally.

A Different Way of Thinking About Things

Although I don’t have a particular problem with the definition of inflation we often use, I think that one needs to take a step back and begin with first principles.

Let’s use an alternative definition of inflation. Let’s refer to inflation as a sustained decline in the purchasing power of money over time.

At this point, you might be losing your mind. Isn’t that just a different way of saying what the previous definition already accomplished? Well, yes. All I am trying to do is re-frame the topic to provide some clarity. If we want to understand inflation and its cause, I think that this framing is better.

This is a newsletter about price theory. As such, we should start with some supply and demand. What determines the purchasing power of money? It is easy to say supply and demand, but supply and demand for what?

A common answer is for people to say “the supply and demand for money.” But that actually just leads to other questions: What counts as money? Currency? Bank account balances? This leads to a lot of other distractions like trying to define what is and isn’t money. However, the answer is quite simple — although not always simple to explain.

In my view, Scott Sumner gives the best description: the purchasing power of money is determined by the supply and demand for the medium of account.¹

I say that this is the best description, but most readers are probably thinking “what in the world is a medium of account?” This language, as far as I know is most clearly outlined by the late, great Jurg Niehans (an under-appreciated economist who used a price theoretic approach to monetary theory). Here is his description:

Money is here called a medium and not, as customary, a unit of account because, clearly, money itself is not a unit, but the good whose unit is used as the unit of account.

Even that definition might be confusing in the modern world in which the medium of account and the unit of account seem like the same thing. Thus, let’s think about this definition by way of example.

Consider a gold standard. Under a gold standard, the term “dollar” is defined to be a particular quantity of gold. For example, the dollar might be defined as one-twentieth of an ounce of gold, 9/10 fine. In this case, the dollar is the unit of account, but gold is the medium of account. As a result, the purchasing power of the dollar is determined by the supply and demand for gold.

To understand why this is the case, consider that the intersection of supply and demand curves determine the equilibrium price and quantity. But it is important to remember that the price is the relative price of the good. Since we are holding everything else constant, the price determined through supply and demand is the price in terms of other goods. This is an important distinction because the dollar price of gold is fixed by definition. Given our definition above, an ounce of gold 9/10 fine has a dollar price of $20. Fluctuations in the supply and demand for gold do not impact the dollar price, but rather the relative price of gold.

The relative price of gold is its purchasing power. It measures the number of baskets of goods (where the basket is defined by the price index) that one can purchase with an ounce of gold. The higher the price of gold relative to other goods, the greater its purchasing power. Since the dollar is defined as a particular quantity of gold, then by definition, this also determines the purchasing power of the dollar.

When the supply of gold increases, the relative price of gold declines. Since the dollar price of gold is fixed, this implies that all other prices rise to clear the gold market. When the demand for gold increases, the relative price of gold increases. Since the dollar price of gold is fixed, this means that all other prices must fall to clear the gold market.

Note here that our two definitions of inflation are identical. When the purchasing power of money is declining, prices are rising. These definitions are just two different ways of saying the same thing. Furthermore, to the extent to which some prices rise faster than others, this simply reflects a change in the price of that good relative to gold.

Nonetheless, I think that the framing is important. When we are talking about inflation, we are talking about a decline in the purchasing power of money. We measure that by constructing price indices to get a sense of the magnitude of those changes. Fundamentally, the price index is just a measurement tool.

The Medium of Account in the Current World

The gold standard no longer exists. A reader might therefore wonder why I spent so much time talking about such a world. The reason, and one that I think a lot of people miss, is that the same basic principles hold.

In the modern world, only the Federal Reserve has the ability to create dollars. Banks can only create claims to dollars. Thus, the dollar remains the unit of account, but the medium of account is the monetary base.

Given that distinction, it is a bit weird to think about the nominal price of a base money. Base money has a unit price. In other words, the nominal value of $1 is $1, by definition. The supply and demand for base money determines the relative price of base money. Since base money has a unit price, this implies that all other prices must adjust to clear the market for base money.

This helps us to understand, for example, why we didn’t experience the high rates of inflation that were predicted following the Fed’s large expansion of its balance sheet in the aftermath of the financial crisis. The Federal Reserve expanded its balance sheet while also paying interest on reserves. This payment of interest on reserves increased the demand for reserves at precisely the same time that it was creating reserves. Supply increases. Demand increases. Those two effects offset one another.

At this point, an astute reader might ask, “okay, this is all fine, but aren’t you the Divisia monetary aggregate guy? Why care about those aggregates if the monetary base is so important?” My answer is that the broader, Divisia monetary aggregates are useful because they reflect discrepancies between the supply of and demand for base money. The much-maligned money multiplier is useful here. Not the caricature of the money multiplier as some constant that when multiplied by the monetary base gives you the broader aggregate, but rather the basic principles behind that multiplier. As the demand for base money increases, this reduces the money multiplier. Thus, to the extent that the broader measure of money declines, this reflects an excess demand for base money — caused either by an increase in demand or reduction in supply. To the extent that the broader measure of money increases, this reflects either a decline in the demand for base money or increase in the supply of base money — either way, an excess supply. When we are thinking about the modern world and inflation rates, we need to think about these in terms of growth rates rather than in terms of levels, but the same principle applies. Empirically, these are useful.

Now, to return to the basic point, the purchasing power of money is determined by the supply and demand for the medium of account. The relative price of money is just another relative price. However, because everything is priced in terms of money, differences between the quantity demanded and quantity supplied of the medium of account tend to influence a large number of markets such that prices, on average, adjust to this discrepancy.

This is true regardless of how big or small the base is relative to the entire economy. Again, an analogy to the gold standard is apt. One argument that was made against the gold standard was that it was unreasonable to let the market for one commodity, which accounts a small fraction of the total economy, to have such a significant influence on prices and economic fluctuations.

Once you start thinking about this in terms of supply and demand for the medium of account, it is hard to think about these issues otherwise.

Why Does This Matter?

Re-framing inflation as a discussion about the purchasing power of money has its advantages. First, it grounds the discussion in price theory. This is important. Statistical plots, macroeconomic explanations, or blaming rising prices on rising prices simply won’t do.

Second, it helps to frame the discussion about the fundamental causes of inflation. If it is about purchasing power, certain explanations simply do not add up.

Third, this doesn’t rule out all other factors for explaining inflation. Other explanations can be correct to the extent that they affect the demand for base money (or possibly central bank decision-making).

Fourth, it forces people to think carefully about relative prices. Sure, there might be some particular commodity (e.g., oil) for which an unexpected decline in the supply can also increase a particular price index. However, it is important to ask whether this reflects a change in the purchasing power of money or if this is simply a change in relative prices.

In short, it seems important to establish some first principles. Good macroeconomic analysis requires price theoretical foundations. The recent experience of inflation has left us with muddled explanations of what happened. A common mistake people make is to engage in accounting exercises where economic analysis is necessary. In addition, there is no reason to treat each experience of inflation as though it is some unique event that requires a new theory of everything. Perhaps a re-framing is in order.

1. Scott suggested this terminology to me probably 15 years ago, although it took some time for me to fully appreciate what he was saying. The link is to one of his recent posts, which is also very good on this point.