How can tariffs be inflationary?

An econ-nerd response to a friend who questions whether inflation, "a monetary phenomenon", can be inflationary without more money in the system.

My good friend who, to protect his identity, I’ll call J, offers these thoughts to which I respond below:

I would like Chairman Powell to answer two questions. Tariffs are taxes on consumers. Since when is a tax inflationary? What Keynesian fantasy are your economists conjuring? Or is this just more of their MMT hubris?

Second, inflation is dollar devaluation caused by excess money supply. When are you going to really start addressing inflation? The FOMC’s insistence on using rates to control inflation is only hurting the private (real) economy, not stopping inflation.

If the Fed is serious about reducing inflation / dollar devaluation, then drain the excess supply of money the Fed created through QE.

I’ll respond to the Fed/FOMC stuff first and quickly, then get to tariffs and inflation:

I agree that the Fed is behind the curve, as they always are. I agree that money supply is too high (and what they did during COVID was an economic sin) but I don’t think the Fed will speed up the process of reducing it.

Part of the reason they won’t is because of their ridiculous “dual mandate” which includes not only price stability but also maximum employment. Especially if you get Keynesian FOMC members they’ll be reluctant to do things that they think might slow the economy and reduce employment because, note, the mandate is not “optimal employment” but “maximum employment”. After all, sometimes the policy required to move inflation in the right direction is the opposite of the policy required to move employment in the desired direction which, apparently is always up. (At least to the extent that you assume that Fed policy is even effective, but that’s a topic for another day.)

Now to tariffs and what I think is an error/oversight in J’s thinking:

Tariffs can be inflationary in a couple of different ways. I think the second one is more important than the first but will mention both:

First, depending on just how inflation is measured, CPI could miss price declines that one might theorize would offset price increases elsewhere.

Tariffs not only raise the price of imported goods but also tend to raise the prices of 1) domestic goods made with imported parts and 2) domestic goods whose prices have been constrained by foreign competition.

Of course, without more money in the system – and I think this is J’s argument and is correct to a point – extra money being spent on one thing would cause less money to be spent on other things, thus offsetting inflation increases.

It’s surely possible that those other things decline in price due to lower demand.

Let’s say I must buy steel and may buy chocolate — and those are the only items in the CPI — and I’m not getting more money to spend. I buy more expensive steel so I have less to spend on chocolate.

In that case, if the price of chocolate dropped by just the right amount as chocolate sellers cut prices to maintain sales volumes, CPI could theoretically be unchanged even with the tariffs. But what if the CPI is constructed in a way that gives more weight to things that are impacted by tariffs and less to things that aren’t? What if the CPI “basket” overweights steel and underweights chocolate? In that case, CPI could overstate inflation.

I’m not saying this is certainly the case; it’s a theoretical argument which might or might not be right. And of course, it’s also possible that CPI understates inflation. I’ve seen reasonable arguments on both sides, though some of them seem to have political motivations, by people who want to increase or decrease cost-of-living increases in entitlement programs. (Overstating inflation is insanely expensive for the American taxpayer as it compounds cost-of-living increases for the Ponzi schemes we call “entitlements” and further burdens our children under a disgusting, dangerous, and immoral level of debt.)

One argument, though I’m a bit skeptical of the idea that asset price “inflation” is an important driver of actual inflation; I think that’s more of a result than a cause: How CPI understates real inflation

The other side (arguing that CPI overstates inflation): Are There Good Alternatives to the CPI?

In a less theoretical vein:

What happens, though, if those other things, i.e. my chocolate, don’t go down (or not down much) in price because producers/sellers cannot make sufficient profits at lower prices to stay in business? So instead of cutting prices enough for CPI to be unchanged, companies with lower sales due to the reduced quantity of money available to be spent on their products/services after people spend more on things impacted by tariffs (steel), either only cut a little bit or just shut down? I mean, what if for CPI to be unchanged, chocolate prices need to drop by 10% but at least some chocolate makers can’t stay in business at those prices? Some may cut 3% or 7% to try to keep customers. Some may just close down. You don’t want to be “the car dealer who loses money on every sale but makes it up in volume.”

Then there’s little or no offset to price increases. And inflation (at least as measured by CPI, and probably in how it really feels to consumers) rises. And yet there was no increase in the money supply. This is not a Keynesian argument.

So, this is the error that I think J and plenty of others make when referring to the “inflation is a monetary phenomenon” line. The other common way of phrasing it is “too much money chasing too few goods.” J’s oversight is only focusing on the “too much money” part rather than on the “too few goods” part.

Thus, the second way tariffs can cause inflation: if the quantity of goods (or services) available to purchase declines then you can have the same money chasing fewer goods, and that’s inflation.

The real risk of tariffs is stagflation. Higher prices with a lower total quantity of stuff being bought. The higher prices may or may not increase employment in those industries but the lower quantity of stuff being bought will very likely cost jobs, sometimes even in “protected” industries with higher prices (depending on the elasticity of demand) and certainly in the other stuff, like my chocolate.

If Trump really stuck with his insane “reciprocal tariffs”, we’d have stagflation within a few months. If he stays at the (still harmful but not immediately massively destructive) 10% level, it’s true that the foreign exporters and domestic importers and sellers will cut their margins to some degree to maintain market share, and that would significantly reduce the inflationary/stagflationary effect. (It would reduce profits within the system and potentially returns/dividends to American business owners and shareholders but it’s not necessarily the government’s job to protect those.)

But just because the effect is reduced doesn’t mean it’s zero and I think it’s absolutely certain that tariffs will have at least a small inflationary effect. Will it be big enough to notice? If only 10%, maybe not very much because some is absorbed in lower margins and there are so many other things that impact the economy such as tax policy and, as noted above, interest rates. So we could see other things lower inflation, masking tariff-caused inflation and allowing the supporters of tariffs to say that they’re not inflationary. But that won’t be true and doesn’t mean it isn’t bad policy.

I also note that at 10% tariffs Trump certainly can’t achieve his stated goal of massively increasing American manufacturing. He can only take billions of dollars out of the private sector and give it to government, like any other tax.

Anyway, J, I think one need not be a Keynesian to make a very strong argument for how tariffs can be inflationary without a change in the money supply as long as you stop assuming that the supply of goods/services cannot decline.