Bryan Caplan  

Macro Miscommunication: How Keynesians and the Right Misunderstand Each Other

Donations vs. Taxes... Raico on the Lop-Sided Treatme...
I know lots of Keynesian economists and lots of right-wing non-economists.  I've spent decades in both worlds.  In the process, I've noticed some recurring macroeconomic miscommunications between the two camps.  Each takes something so much for granted that it can't imagine that the other disputes it.  As a result, they talk past each other, and fail to address the other side's main concerns.

Most right-wingers take it for granted that printing money increases demand.  As a result, they fail to realize that Keynesians often dispute this very point.  When Keynesians claim that "quantitative easing won't work," they're usually denying that printing money will actually lead to additional spending. 

Most Keynesians, on the other hand, take it for granted that, when there's significant unemployment, increasing demand will increase employment and output.  As a result, they fail to realize that right-wingers often dispute this point.  When right-wingers claim that "quantitative easing won't work," they're usually denying that additional spending will lead to additional employment and output.

My own view is that both groups are largely correct about what they take for granted.  Right-wingers are correct to take the money-->demand connection for granted.  Keynesians are correct to take the demand-->employment/output connection for granted.  In each case, there's a simple, compelling mechanism at work.  The money-->demand connection ultimately holds because when people have more purchasing power than they want, they spend it.  The demand-->employment/output connection ultimately holds because nominal wage cuts hurt morale, and thereby productivity.

In any case, though, the path to better debate begins with understanding what the other side takes for granted and what it disputes.  Hope this helps.

Comments and Sharing

COMMENTS (20 to date)
Jacob Oost writes:

Actually, this just confuses me even more. Are you saying that printing money leads to greater demand, or just greater nominal spending? I don't often hear right-wing non-economist commentators say this, I normally hear them complain that printing money inexorably leads to inflation and risks hyperinflation.

And are you saying that when Keynesians talk about "increasing demand" they are really just talking about "increased spending"? I think *you* are talking past *me*!

I always thought Keynesians had too shallow a view of demand, and they seem to think that government spending simulates (yes, I mean "simulate" not "stimulate" here) *genuine* demand enough to stimulate the economy (and there I meant "stimulate"), when as we all know, genuine demand leads to greater spending by the demandERs which gets the cycle of economic growth going. I think that, at best, government spending by itself can just rearrange (or distort) the nature of employment and output (output as measured by its utility), not make it actually go up.

andy writes:

Not sure where should I place myself, but I never quite understood what - actually - is 'aggregate demand'. The same with the 'circular flow' model which seems to me a little pointless. I always found using the 'P' (price level) at the model level very unsystematic (P is by definition a number generated by the 'Ministry of Statistics'; it's as if you put 'index of economic liberty' into mathematic models; I have never seen a consistent definition of P that wouldn't rely on 'Ministry of Statistics).

And, finally, I am not sure what 'would work' should mean in this case. Maybe that's because nobody said what is actually 'broken'.

If I heard in private sector 'let's spend a billion of dollars on something and the problems will go away', I would bet, that the company will not solve any problem this way. Yet, that's more or less what Keynessians say.

As for the wage rigidity: I did search for the studies and I am confused:
- most of the studies study behaviour of firms in normal economic conditions. In a crisis many firms behave differently and you did actually see many firms cutting significantly;
- the studies often conclude that the wage rigidity applies to different groups differently; you have a lot of self-employed where it obviously does not apply; it mostly doesn't apply to recently-hired people and it applies more to senior positions;

Given these findings, I would say that there is a huge group of people to whom the wage rigidity does not apply; people can move between the groups. Why should such a rigidity cause large-scale unemployment? It shouldn't be that hard to arbitrage it away..

jsalvati writes:

@Jacob Oost
(Aggregate) Demand is a nominal concept, and basically equals nominal spending. There are lots of someone informed people who think otherwise, but as far as I know (and I know a fair amount about this topic) Aggregate Demand only makes any coherent sense as a nominal concept.

Joe Cushing writes:

I don't think you need crazy monetary policy or crazy government spending to pull out of a down turn. What you need is to clear the market as quickly as possible. When a firm goes bankrupt, no wealth is lost--it merely changes hands. When a house is lost to foreclosure the same situation occurs. No wealth is lost in the foreclosure; the house just changes hands. Where wealth is lost, is in lost productivity and accelerated depreciation during the time it takes to change hands. Productive assets derive their value from production. If they are not producing, their value is lost for the time they are shut down. The accelerated depreciation comes from the fact that assets in limbo are not guarded and maintained. Think of the house with fish in the swimming pool and squatters inside. There is nobody to fix the leaky roof which destroys the house. The key to getting out of an economic fix has nothing to do with Keynes or monetary policy. What you need to get the markets cleared as fast as possible. It's like pulling off a bandage. The faster you do it, the less painful it will be. What governments do during downturns is to try and prevent this painful process from occurring at all. All they accomplish is dragging out the pain and destroying wealth.

Nick Rowe writes:

There's an aggregate demand curve and an aggregate supply curve in {P,Y} space. P on the vertical axis is the price level, Y on the horizontal axis is real GDP. And P.Y is nominal GDP.

1. Does monetary policy (or fiscal policy) shift the AD curve right?

2. Is the AS curve vertical?

Most of the confusion comes from shifts in curves vs movements along curves. Add in the real vs nominal confusion, and you can have great confusion, long before you get to any substantive issues.

Two Things writes:

Decades in the biz or not, I think you're wrong about what "right-wing non-economists" think. We actually think printing money increases prices (inflation), not demand. Moreover, most of us think printing money is a substitute for increasing taxes.

Since printing money (inflation) transfers resources from private actors to the government and its cronies (e.g., big banker execs) it decreases demand overall because of deadweight loss and transaction costs. (Of course we understand that government workers and government cronies may demand more luxuries, but their gains come at the expense of normal people who are left with fewer resources so they demand less).

You are correct, however, that we think QE won't work because increased spending of inflated fiat money will not lead to increased employment or output. We think QE will merely push up prices and transfer productive resources to government workers and cronies, who will blow said resources on luxuries-- or at the least awful, buy assets to be mismanaged by themselves instead of being put to their highest and best uses by private actors. Since inflation, like taxation, comes with deadweight loss etc., we think QE will result in less employment and output.

Joe Cushing writes:

Oh, another point. The inflation that comes from Monetary policy clears markets by transferring wealth from lenders to borrowers.

Bill Woolsey writes:


Firms fail all the time without there being a recession.

You are making an error in assuming that a recession is the same thing as a single firm failing. As soon as the firm is reoganized then it can start again. Or something like that.

Well, really, what happens when particular firms fail, usually, is that they permanently shrink in size--use fewer resources and produce less product. Resources, including labor, are shifted to expanding areas of the economy. This is an economy that is prospering. This is _normal_.

A situation when may sectors of the economy are shrinking, freeing up resources, but few if any sectors are growing--needing the additional resources--is a recession.

The problem is that spending in total is too low. However, the only way that can exist is if the demand to hold money is greater than the quantity of money.

One solution is for the price level (including wages) to fall enough so that the real quanity of money equals the amount people want to hold. At the same time, real expenditures on goods and services will rise to match the productive capacity of the economy. That _means_ that the demands for some goods and services will rise faster than the demands for other goods and services fall. That is, that the resources freed up in shinking industries will be needed to produce goods in the growing industries.

The other solution is for the nominal quantity of money to rise enough so that it equals the demand to hold money, and so money expenditures grow. The demand for some goods can still fall, some firms can go bankrupt, some workers laid off, but the whole point is to free up the resources to produce other goods and services, those for which spending rises, those which people want to buy the most.

Or, of course, there is a third possibility. People could just choose to hold less money. And that, I think, is what you are implicitly assuming. People are holding more money for now, and then, once we work out these problems that are like bankruptcy by a single firm, then people will choose to hold less money, and so spending will recover. We just need to wait for this adjustment.

Your error is to fail to see that the adjustments that are needed is a shift in resources, including labor, between industries. And that there is no need to wait for a time before starting the adjustment. Let's just pause, and have many poeple do nothing, and then, after a while, we can start expanding something else.

The market process at a given quantity of money is the lower price level. That works slowly, and the alternative is what you call "crazy" monetary policies.

Nick Rowe writes:

Two things: "We actually think printing money increases prices (inflation), not demand."

Voila! A perfect illustration of what I was talking about. A failure to distinguish a rightward shift in the demand curve from an increase in quantity demanded.

Two things implicitly agrees that printing money shifts the AD curve to the right (increases demand), but the AS curve is vertical, so the only effect is on prices, with no effect on quantity demanded.

Two Things writes:

In {P,Y} space printing money is a bit like relabeling the divisions along the P axis. The AD curve doesn't go anywhere, really.

Of course that's too simple, because we don't just give everyone the same proportion of new money like with a stock split. We give some people more and others less. Aggregate demand stays about the same (maybe shifting a little because of deadweight loss, etc.) but if we could disaggregate it we would see many shifts in demand by various actors.

(Inflation tends to transfer real goods from private actors to government/crony actors, from employees to employers, and from creditors/savers to debtors; e.g., from depositors to banks. Tax "bracket creep" exacerbates the first effect.)

Although money acts, in the short run, as "the most liquid commodity" (medium of exchange, etc.), in the long run, fiat money is just a rubber ruler for measuring the relative values of real goods and services, its length constantly distorted by central-bank machinations.

So many people so desperately wish money to be a "real" commodity so they can save "money" rather than stockpile goods or make investments,* that many of them simply delude themselves that what they wish is true. Then they assume changes in money supply right out of their analyses of changes in price levels. In fact, only relative prices (that is, among various goods and services) are very meaningful in a fiat-money economy, though we mustn't neglect people's money holdings at any given time.

The worst are the pretenders-to-wisdom like Paul Krugman, who really seems to think that newly printed money has some intrinsic value which can be traded for new real goods and services. Of course he is delusional; printing money can change the distribution of goods and services but not the quantity, except to the modest extent that some people (foolishly, wrongly) feel richer when they have more quatloos of less value so they inadvertently (spend their savings/eat their seed corn) in an Austrian fashion producing a temporary boom.

*Of course bank-account/money-market balances are "investments" but many people don't understand that very clearly and thanks to the government such accounts pay virtually no interest/dividends at the moment. For good reasons, such investments wouldn't pay much even without government manipulation.

Bob Murphy writes:

I like Bryan's attempt to mediate the confusion; I agree with him that a lot of times Keynesians and non-Keynesians talk past each other.

But is this a fair summary of Bryan's point?

"I agree with both camps in their premises for why QE can't work, but they are wrong in their conclusions: QE *does* work, because printing money increases demand, and increased demand leads to higher employment."

Mike Sproul writes:

The right-wingers are wrong to think that printing money increases demand. The Fed issues $100, and gets a $100 bond from the public. Nobody's net worth is changed, so there's no change in demand for goods.

The Keynesians are wrong because, well, where to begin? For starters, they think the government can make us richer by burying bottles of cash in coal mines.

Lee Kelly writes:

Mike Sproul,

The purpose of open market operations is not to increase the asset seller's net worth. The purpose is to increase their money balances above that which they desire to hold. Unless the bond sold and the cash received are considered perfect substitutes, some fraction of money received will be spent. All else being equal, this will increase total spending, and -- supposing there are idle resources to be put to work -- tend to increase employment and output. If the economy is already at its natural level of unemployment, then such monetary policy will merely create inflation.

The typical "right-wing" view of the recession is that the high level of unemployment is entirely natural. That is, resources are already fully employed in their most valued uses. The problem, in this view, is entirely the preceding boom, because it rendered a large part of the capital structure unemployable for any productive end. The slow process of re-transforming such resources for sustainable production cannot be hurried along by monetary policy.

For what it is worth, both are right, depending on the cause of unemployment.

Mike Sproul writes:

Lee Kelly:

You have to ask why that money was issued in the first place and why people chose to spend it, rather than, say, returning it to the Fed or letting it sit in a drawer.

If a certain community used nothing but 1 oz. silver coins, and if, for some reason, someone minted 100 new coins, then somewhere in the economy, 100 coins would be melted back to bullion or stamped into spoons, forks, candlesticks, etc. Nothing would happen to the number of coins spent at the grocery store, and nothing would happen to the price of groceries. That's the Law of Reflux in action. The reflux works even more easily for paper money, since paper can easily reflux to the issuer without the expense of melting silver.

Lee Kelly writes:

Mike Sproul,

You assume monetary equilibrium. Suppose the demand for money is increasing, i.e. people wish to hold larger cash balances in aggregate by 100 coins. All else being equal, silver coins would then rise in value against all other goods and services.

Prices fall, right? But instead suppose that someone melts spoons, forks, candlesticks, etc. and issues 100 new coins. The increase in the supply of silver coins exactly offsets the increase in demand. No coins are melted anywhere else in the economy and prices do not fall.

Think about it. If the demand for silver coins rises relative to other goods that use silver, then silver will be drawn out of other uses and toward coinage.

Lee Kelly writes:

Mike Sproul,

M*V = P*Y = AD

All else being equal, AD rises when the Fed increases M. To hold AD constant, you implicitly assume that V falls in proportion, but that implies the bonds sold were perfect substitutes for cash. This is false. Even short-term notes with near 0 interest rates are not perfect substitutes, but just closer substitutes. Long-term bonds with higher interest rates are certainly not perfect substitutes. While sellers may hold some of the money received from the Fed purchase, it is highly unlikely that all will go unspent.

Since the dollar is a fiat paper currency, it cannot be melted for a precious metal, and it cannot be redeemed at the Fed for any valuable asset. There is no reflux of notes to the Fed that forces it to contract its issuance. The Fed can issue as many, or as few, of its notes as it likes, and aside fluctuations in money demand, has almost complete authority over the level of AD.

Lee Kelly writes:

The typical story of the recession espoused by libertarians and conservatives has an interesting implication: unemployment has risen because that is the most efficient use of labour. In the short run, such labour can only be reemployed at a loss, which is why neither fiscal or monetary policy can help. The transformation of labour for profitable reemployment is slow, and the process cannot be hurried along by government. In a sense, he current high level of unemployment is as it should be.

While I believe their is some truth to this story, it cannot account for the large decline in NGDP in 2008. Unemployment is far too high to be explained by the Austrian malinvestment story alone.

Mike Sproul writes:

Lee Kelly:

You give a fine description of the operation of the Law of Reflux for silver coins. If there is increased need for silver coins in trade, then bullion will be stamped into coins, and there will be a small rise in the price of silver. A fall in the need for silver coins will have the reverse effect. But of course this means that if someone just minted 100 coins on a whim, then that many coins would reflux to bullion, and the price of groceries would not change.

But let's focus more on paper money. Suppose a free-banking world where people trade with dollar bills, each redeemable by its issuer for either 1 oz. of silver or for other goods (land) with a market value of 1 oz. of silver. When business picks up and people need more dollars, they sign over title to 1 oz. worth of land to the issuer of dollars. If business slows, and people need fewer dollars, they bring their dollars back to the issuer and get back their land titles. Would you agree that in this case, the issuance and reflux of dollars would happen with only negligible effect on the value of land, silver, and dollars?

(I also have plenty to say about MV=PY, and about fiat money, but I'll keep it short for now.)

Jacob AG writes:

"In any case, though, the path to better debate begins with understanding what the other side takes for granted and what it disputes. Hope this helps."

This is something that Keynes did a lot of. If you read the General Theory, you will find him constantly describing the state of macroeconomics in his day, asking about what assumptions are needed for that paradigm to hold true, and attacking/adjusting those assumptions left, right and center.

Sonic Charmer writes:

Lee Kelly,

The typical story of the recession espoused by libertarians and conservatives has an interesting implication: unemployment has risen because that is the most efficient use of labour.

Indeed, that's my view. With one minor adjustment: it's the most efficient use of labor given the web of minimum wages, taxes, regulations, etc. etc. that make labor far more expensive to 'purchase' than might otherwise be the case.

For many if not most unemployed people, you can fairly ask: what, exactly, would it be worth paying them minimum wage + benefits + paperwork + etc. to do? Send emails to each other? No. The answer in many cases is: 'not very much'. (If you had a better answer, you'd implement it and make money. But you don't so you don't; no one does.)

As for whether unemployment is 'too high' to be well explained by this factor, that's not obvious. Says who?

Comments for this entry have been closed
Return to top