The difference between Keynes and Hayek

Gene Callahan and Jonathan Catalan have weighed in. Gene agrees with me that the real difference is in the way they thought about interest rates, and he suggests that which was right is an empirical question and may vary from time to time. Jonathan thinks that a potentially more important difference is between the Ricardo effect and the multiplier (I'm not sure why these are being contrasted with one another, personally).

But most of Jonathan's post is about interest rates again. If I'm reading him right, I think he's suggesting that the interest rate theories are not a "crucial difference" because they ultimately provide you the same conclusion. He writes:

"If the classical loanable funds theory is correct, the aggregate demand for factors of production equals exactly the amount of money saved. If Keynes’ liquidity preference theory is correct, there runs the chance that the rate of interest won’t accurately reflect society’s time preference. In both cases, the prices of the factors of production should reflect the nature of the rate of interest.
In other words, even if the rate of interest is higher than what it should be given society’s time preference, the prices of the factors of production will reflect total aggregate demand for them. That is, the prices of the means of production would be lower than they would be if the rate of interest were lower. We can roughly conceive of this as an equilibrium ratio between the prices of producers’ goods, the rate of interest, and expectations of future income."

I think this is right. You could talk about both theories with an AD-AS graph (just ask Roger Garrison!). This is a good point to make because it speaks to my initial reaction to Gene's post. Gene wrote:

"And they were both partially correct: the interest rate is influenced by both of these factors. So here is where we must get empirical: To the extent that the interest rate is determined by liquidity preference, to that extent Keynes was correct. To the extent the interest rate is determined by the supply and demand for loanable funds, Hayek was correct."*

But what is the empirical test??? We don't have data on liquitiy preference schedules (although presumably we could come up with something). For me, the real testable difference comes back to Wicksell. Both Hayek and Keynes are heavily influenced by Knut Wicksell, specifically the idea that there is a "cumulative process" (and Keynes and Hayek both have thoughts on what that cumulative process is... and perhaps this is what Jonathan was thinking about when discussing the multiplier and the Ricardo effect together) that are set off when the interest rate departs from some "natural rate".

But this works differently for Keynes and Hayek. Keynes suggests that the departure from the natural rate comes during the crash, either because of changes in the interest rate or MEC or both. Before the crash (aside from perhaps some asset bubbles that might induce the panic), the economy is operating normally. For Hayek the interest rate goes below the natural rate before the crash. That sets of an unsustainable cumulative process that ultimately has to end because of the irreversibility of capital investments, the distortions set off by the Ricardo and Cantillon effects, and revisions in entrepreneurs understanding of what's going on.

So here's the heart of all that: what is the empirical test we could do that Gene suggests we do?

If we had a good estimate of the natural rate, we could compare it to the interest rate over the business cycle and assess the strong and weak points of each description of reality.

So why do I think interest rate theory is so important? Because Keynes's liquidity preference theory and the whole constellation of observations about financial markets in the General Theory provides us with a reason to think that the interest rate is higher than the natural rate during the downturn. It motivates the whole description of what we observe.

I am thinking about all these things right now, and Gene and Jonathan should both be prepared to be pestered for thoughts on a draft in a couple months.

* I actually don't hold a pure liquidity preference theory myself (and I wonder if Keynes really would have if he thought about it). So in that sense I'm in between Gene's two poles to begin with. When a person saves rather than consumes, that is obviously informed by their time preference. So even if the marginal decision between keeping money liquid and putting it at interest is primarily a liquidity preference decision, the size of the pool of liquid assets influences the ultimate decision and the size of that pool is clearly determined by time preference! In addition, financial institutions that pay interest are going to be informed by the demand for loanable funds, which is going to be determined by firms' internal rate of return. So yes - interest is definitely the price for parting with liquidity. But the amount of liquid funds available and the demand for the liquid funds put at interest are both determined by time preference. A pure liquidity theory seems far less important to me than the simple fact that liquidity preference is in there at all, introducing a "loose joint".