How to sell a new idea in macroeconomics
By Scott Sumner
It’s not easy to sell macroeconomists on a new idea, as I’ve discovered over the past decade. Here I’d like to discuss some mistakes that people make, and then offer a few suggestions. (Note that I have often made these sorts of mistakes, so I speak from experience.)
One mistake is to tell people that their approach is wrong, because they don’t understand certain accounting relationships. Here are a few examples of claims that I see people make:
1. There is no money multiplier, so any discussion of money multipliers is wrong.
2. The Fed doesn’t control the money supply, because the money supply is endogenous.
3. Banks don’t lend out reserves. Rather loans create new bank accounts, and the money stays within the banking system.
4. Government spending causes an increase in the money supply.
As I’ll explain in a moment, there may be a grain of truth behind some of these claims. But the dogmatic way they are presented will turn off those you are trying to convert. They should be arguing that the money multiplier is not a useful concept (I agree), or that it’s useful to think of the money supply as endogenous (I disagree), or that it’s not useful to think of banks as lending out reserves (I agree), or that it’s useful to think of government spending as boosting the monetary base (I disagree.)
1. The money multiplier is the ratio of a monetary aggregate (such as M1 or M2) and the monetary base. Since M1, M2, and the base exist, the money multiplier obviously exists. Those who deny its existence are actually denying its stability, i.e. arguing that it’s not a useful concept for policy implementation. I think that’s right. But the factors that cause the money multiplier to change over time are covered in depth in any competent textbook, such as Mishkin’s text. So if you suddenly come along and say that the profession is all wrong and there is no money multiplier, don’t expect to be taken seriously.
2. Whether something is endogenous or exogenous is not a fact of nature, like whether something is made out of copper or nickel. Rather you can treat a relationship as exogenous for some purposes, and endogenous for others. Thus if the central bank pegs interest rates over a six week period, then during that period the money supply is endogenous. Suppose that over a longer period of time they wish to increase the money supply to boost inflation and reduce unemployment, as during the 1960s. Then during that time frame you can say the money supply was increased exogenously, and that the interest rate is adjusted every 6 weeks to get the sort of money growth required to generate higher inflation and lower unemployment. I don’t think any of this is controversial. But saying someone is “wrong” to treat the money supply as an exogenous policy tool doesn’t get you anywhere. Indeed I happen to believe it is useful to view the money supply as exogenous.
3. When banks make loans, the reserves usually don’t leave the banking system. But sometimes they do, as when the loan recipient asks to receive currency from the bank that makes the loan. In my view, “loaning out money” is not a useful way to think about the issue. Rather you should think in terms of the quantity of reserves banks that wish to hold, in aggregate. If they wish to hold fewer reserves, then they make it less attractive for depositors (say by lowering the interest rate paid on bank deposits) and this makes it relatively more attractive to hold currency, causing some of the monetary base to flow out from bank reserves into cash held by the public. But even though I believe this is a more useful framework than “lending out reserves”, I try not to insult people by telling them that they are wrong when they claim that banks lend out reserves, because sometimes banks do indeed lend out reserves.
4. When the government writes a check as part of a government spending program, the money eventually gets deposited by the recipient, and adds to the monetary base. But that doesn’t mean that it’s useful to think in terms of “government spending causes an increase in the monetary base”, because usually it does not have this effect, in net terms. The actual change in the monetary base depends partly on how much the government subtracts from its account at the Fed when paying for government spending, but also how much is added via tax revenue or the funds from selling bonds. The monetary base is also impacted by open market purchases and sales by the Fed, and indeed the Fed will offset any change in the base triggered by government spending, which the Fed does not wish to occur. So as a practical matter, government spending doesn’t really affect the base in any meaningful way, despite the accounting relationship that makes it look like it does affect the base.
I have lots of heterodox ideas. When I promote them, I am most effective when I avoid telling those who disagree with me that their model is wrong, rather I argue that my approach is more useful. Thus in my Great Depression book I argued that looking at the supply and demand for gold was a more useful way of modeling the interwar price level than looking at the supply and demand for US dollars (as in Friedman and Schwartz). Interestingly, in a later book entitled “Money Mischief”, Milton Friedman tried to explain changes in the price level by looking at shocks to the supply and demand for monetary commodities such as gold and silver.
If I argue with a Keynesian, I try not to say, “IS/LM” is wrong”. Rather I might argue that monetary policy mostly works through the expectations channel, and that an easy money policy creates expectations of faster NGDP growth, which shifts the IS curve to the right, which then raises interest rates. I then argue that the market monetarist approach is a more useful way to think about monetary policy than IS/LM.
When arguing that money was actually contractionary in 2008, despite falling interest rates, I might explain to a Keynesian that the natural rate of interest was falling faster than the Fed’s target rate. Again, that’s not my preferred framework, but it’s one way to communicate.
It’s important to be able to explain what’s different about your ideas in a framework that your opponents can understand. At least if you want to win friends and influence people.