Friedman’s Thermostat is a control mechanism that attempts to keep an output constant but is mistakenly believed to be causing the external variation that it is controlling for.

A slightly more formal description might help.

Given a system with a feedback controller that adjusts the degree of an intervention C to maintain a target output O despite external variation V, we observe:

• C and V are strongly negatively correlated.

• C and O are uncorrelated.

• V and O are uncorrelated.

An uninformed observer, ignoring the feedback loop, might falsely infer that C has no effect on O, and that C instead “causes” changes in O.

Origin[edit]

The origin appears to be a blog post by Nick Rowe on the Worthwhile Canadian Initiative economics blog[1][2] where he names it that while referencing an Op-Ed post in the WSJ by Milton Friedman where he argues about the mechanism that the Fed has available to be a good thermostat[3]. Friedman in his article is not arguing for the point that will come to be named after him: he is merely likening the Fed’s good performance on inflation-control post the ’80s to being akin to that of a good thermostat controlling the temperature. Introducing the mistaken cause-effect relationship as the interesting part is an invention of Nick Rowe.

If a house has a good thermostat, we should observe a strong negative correlation between the amount of oil burned in the furnace (M), and the outside temperature (V). But we should observe no correlation between the amount of oil burned in the furnace (M) and the inside temperature (P). And we should observe no correlation between the outside temperature (V) and the inside temperature (P).

An econometrician, observing the data, concludes that the amount of oil burned had no effect on the inside temperature. Neither did the outside temperature. The only effect of burning oil seemed to be that it reduced the outside temperature. An increase in M will cause a decline in V, and have no effect on P.

A second econometrician, observing the same data, concludes that causality runs in the opposite direction. The only effect of an increase in outside temperature is to reduce the amount of oil burned. An increase in V will cause a decline in M, and have no effect on P.

But both agree that M and V are irrelevant for P. They switch off the furnace, and stop wasting their money on oil.

— Nick Rowe, Worthwhile Canadian Initiative[1]

Rowe is aware that this concept was not invented by Friedman.

Yeah. The really neat thing about this actually, is that it’s not Milton Friedman’s at all. It’s an old idea. Actually I found it, an old Keynesian Maurice Peston used very much the same argument. He was talking about fiscal policy, rather than monetary policy and real GDP not the price level. But the idea’s pretty simple. What’s the easiest way to explain it? If you’ve got a house with a good thermostat, in a climate like Canada and you’ve got the thermostat set on a steady 20 degrees Celsius and the house stays at 20 degrees Celsius, what do you see? You see the outside temperature going up and down all over the place, you see it burned in a furnace going up and down all over the place. But you don’t see any effect at all on the inside temperature of the house. It just stays the same.

— Nick Rowe on Monetary Basics, Milton Friedman’s Thermostat, and More[4]

He then goes on to explain the idea, while relating it to Friedman’s metaphor:

That was basically Milton Friedman’s metaphor. If you’ve got a good Central Bank, you’ll see the money supply varying, that’s the amount of oil being burned in the furnace, you’ll see all sorts of other shocks in the economy, that’s the outside temperature, but you don’t see any effect on inflation, that’s the inside temperature, or on the price level. So if you want to look and try and find where, is there an effect of money on inflation? You’re not going to see it if you’ve got a house with a good thermostat, or a country with a good Central Bank. It’s trying to make sure that you can’t see any such relation. That’s the idea behind it. And it’s a really, really simple point. It seems to get missed a bit.

— Nick Rowe on Monetary Basics, Milton Friedman’s Thermostat, and More[4]

This seems to exaggerate the degree to which Milton Friedman was exercising the metaphor, particularly because his article seems primarily to focus on the fact that:

The MV=Py key to a good thermostat was there all along.

— Milton Friedman, The Fed’s Thermostat[3]

I can find no earlier reference to it, so we must conclude that Friedman’s Thermostat as terminology and succinct expression of this error is due to Nick Rowe.

Subsequent Uses[edit]

The article and the term have since become fairly popular among econ bloggers, and new references to it pop up every so many years[5][6][7][8][9]. In a case of Stigler’s law of eponymy, this mistaken characterization of what is cause and what is effect is now named after Friedman, though his contribution to the entire thing is solely to compare the influence of the Fed to have gotten more effective over time to using a better thermostat over a worse one, using the analogy of inflation rate to indoor temperature.

Examples[edit]

Typical examples of this come from economics and the Fed’s influence on inflation and prices. This makes sense due to the origin of the term. But the concept itself is not so limited. While the original speaks about economists misunderstanding inflation, this misapprehension of cause and effect is common in other spheres among laymen as well.

“The math isn’t mathing, right?” Armstrong said. “If incarceration was the thing that made people safe, we would be the safest state in the country.”

— Andrea Armstrong, Loyola University[10]

If incarceration worked to secure safety, we would be the safest nation in all of human history.

— Danielle Sered, Until We Reckon[11]

As we have observed in San Francisco, as crime rate increases, politicians are placed under pressure to increase enforcement, increase prosecution, and increase judgment. Consequently, enforcement C aims to preserve some rate of crime O, while encountering variations in crime due to external causes like sudden unemployment, masking, etc. V.

One I could imagine being similar could be a belief that places with heavier lockdowns during the first few years of the COVID-19 had the same rate of deaths as those with softer lockdowns. Presumably, health authorities could be reacting to high death-rate with heavier lockdowns because of some other underlying variation (one place having more elderly or disease-vulnerable or so on). However, I didn’t search too hard for whether this belief was widespread or whether it was unjustified (both necessary for it to be a Friedman Thermostat belief) and so we must leave it unquoted.

Notes[edit]

1. ↑ 1.0 1.1 Rowe, Nick (2010-12-22). “Milton Friedman’s Thermostat”. Worthwhile Canadian Initiative (in English). Typepad. Archived from the original on 2014-05-28. Retrieved 2025-11-02.
2. ↑ Rowe, Nick (2012-07-27). “Why are (almost all) economists unaware of Milton Friedman’s thermostat?”. Worthwhile Canadian Initiative. Typepad. Archived from the original on 2014-03-10. Retrieved 2025-11-02. Posted by Nick Rowe on July 27, 2012 in Econometrics, Fiscal policy, Macro, Monetary policy.
3. ↑ 3.0 3.1 Friedman, Milton (2003-08-19). “The Fed’s Thermostat”. The Wall Street Journal (in English). Dow Jones & Company. Retrieved 2025-11-02.
4. ↑ 4.0 4.1 Beckworth, David; Rowe, Nick (2016-08-15). “Nick Rowe on Monetary Basics, Milton Friedman’s Thermostat, and More”. Mercatus Center. Macro Musings (in English). George Mason University. Retrieved 2025-11-02.
5. ↑ Romanchuk, Brian (2024-01-22). “Why The “Friedman Thermostat” Analogy Should Be Uncomfortable For The Mainstream“. Bond Economics (in English). Bond Economics. Retrieved 2025-11-02.
6. ↑ Christensen, Lars (2011-11-05). “Friedman’s thermostat and why he obviously would support a NGDP target”. The Market Monetarist (in English). The Market Monetarist. Retrieved 2025-11-02.
7. ↑ Beckworth, David (2010-12-22). “Salvaging the Equation of Exchange”. Macro Musings Blog (in English). Blogger. Retrieved 2025-11-02.
8. ↑ Shalizi, Cosma Rohilla (2021-03-26). “Regression, Thermostats, Causal Inference: Some Finger Exercises”. Three-Toed Sloth (in English). Cosma Rohilla Shalizi. Retrieved 2025-11-02.
9. ↑ Kimball, Miles Spencer (2012-07-28). “Milton Friedman’s Thermostat: An Econometric Cautionary Tale”. Confessions of a Supply-Side Liberal (in English). Miles Spencer Kimball. Retrieved 2025-11-02.
10. ↑ Hylton, Antonia; Berk, Emily (2024-08-28). “In one of America’s most troubled jails, a mental health unit has managed to thrive”. NBC News (in English). NBCUniversal News Group. Retrieved 2025-11-02.
11. ↑ Goodman, Amy; Sheikh, Nermeen; Sered, Danielle (2019-03-14). “Until We Reckon: Mass Incarceration, Violence & the Radical Possibilities of Restorative Justice”. Democracy Now! (in English). Democracy Now! Productions. Retrieved 2025-11-02.