Arnold Kling  

Macroeconomic Disconnects

PRINT
Steinbrenner U.... Paul Collier on Democracy...

Scott Sumner brings up some difficulties for anyone who would argue that monetary policy is not a sufficient tool for fighting the recession. He is not arguing from his yet-to-be-articulated new paradigm. He is stating traditional macro.

Traditional macro says that an increase in the money supply causes higher nominal GDP. It says that in a recession, higher nominal GDP leads to greater real GDP. Higher real GDP leads to greater employment.

The case for fiscal stimulus rests in part on the view that there is a disconnect between the money supply and nominal GDP. Keynes had two stories for this. One is the liquidity trap, in which at low interest rates the economy is willing to absorb gobs of money as a store of value. The other is interest inelasticity of investment demand, in which risk-takers are so moved by their animal spirits (or lack thereof) that they ignore changes in interest rates in making investment decisions.

I am willing to argue for a liquidity-trappish point of view, which is that money and government debt are pretty close substitutes. Therefore, when the Fed swaps the two, not very much happens in the broader financial markets. This is my interpretation of Fischer Black's perspective. One big problem with that perspective, at least as far as I can tell, is that it is not able to give as straightforward an explanation for the inflation of the 70's and the disinflation of the 80's as is the monetarist view.

Sumner's main point is that low interest rates do not necessarily produce a liquidity trap. Instead, the Fed can generate expectations of inflation. (On this point, he may need help from his new paradigm.)

As Sumner points out, I have been arguing most for a different form of disconnect. That is, I see a disconnect between nominal GDP and real GDP. We can raise nominal GDP without raising real GDP, because in a Recalculation the economy acts as if it had suffered a supply shock. Under this view, either monetary or fiscal stimulus will produce more inflation and less employment than one might hope.

David Leonhardt is puzzled by the fact that real wages are still edging up, even though we have double-digit unemployment. One story is that real wages are up because nominal wages are sticky and we have had an unusually large drop in prices. That would be consistent with Scott's point of view. Another story is that real wages are up because unemployed workers from the housing and mortgage sectors are not exerting much downward pressure on wages in health care and education. That would be consistent with my point of view.

In recent months, perhaps the most interesting disconnect has been between real GDP and employment. With measured real GDP increasing and employment falling, measured productivity is soaring. If these data are reliable, it suggests that profits are recovering. That in turn should provide the basis for a rebound in employment.

Keep in mind the Garrett Jones view that workers are hired not to produce widgets but instead to build organizational capital. In recent months, firms have sharply cut back on their investments in organizational capital, in order to remain profitable and survive. With profits up, the cutbacks should be fewer, and we should start to see some expansion. The question becomes, again, whether the unemployed workers from declining sectors are capable of supplying organizational capital to the expanding sectors.

If conventional macro is correct, then as long as nominal GDP expands, real GDP will expand. The disconnect between real GDP and employment will be erased by rapid increases in employment.

If the Recalculation story is correct, then rapid increases in employment and real GDP are not possible. If policy makers succeed in expanding nominal GDP, much of that increase will go to inflation, and relatively little will go toward real GDP.


Comments and Sharing


CATEGORIES: Macroeconomics



COMMENTS (12 to date)
David Pearson writes:

Scott Sumner is uncharacteristically vague when it comes to explaining the link between nominal and real response to monetary stimulus. Best I can tell he makes two arguments for real growth following nominal:

1) In the event of a deflationary collapse, real growth would suffer. Therefore, preventing deflation also raises real growth.

2) Inflation lowers real wages, and lower real wages are a spur to growth.

The first item is fairly unarguable, but it unfortunately only applies in the narrow circumstance of a financial panic, now averted.

The second item ignores the state of consumer balance sheets. Yes, falling real wages are a stimulus to production. But in a service/consumer dominated economy, it is not clear that rising corporate profits will generate growth outside of exports (and even then, the export stimulus will be small if real wages are suffering globally). This recession is different in the sense that 50% of job losses have occurred in services, compared to 15% or so in the previous two recessions. Do lower real wages in services have the same impact on "production" than they do in manufacturing? If I pay my waiters 10% less in real terms, does my higher profit cause me to open another restaurant? Not if consumers are refusing to spend on discretionary items as a result of their excessive debt and low liquid savings. This is a narrow example, but instructive.

Daniel Kuehn writes:

Re: "One big problem with that perspective, at least as far as I can tell, is that it is not able to give as straightforward an explanation for the inflation of the 70's and the disinflation of the 80's as is the monetarist view."

But isn't that precisely because there is no binding liquidity trap in the 70s and 80s? It seems to me that was the whole point of monetarism - to fill in those gaps left by Keynes. Keynes himself remarked to Hayek over dinner once that his theory only really applied to deflationary periods.

Re: "Sumner's main point is that low interest rates do not necessarily produce a liquidity trap. Instead, the Fed can generate expectations of inflation."

Which, despite the fight between the two recently, is something that Krugman has always agreed on.


Couldn't both stories be correct, leaving this to be an essentially empirical question? Generally I agree with Sumner, and I buy the liquidity trap disconnect. But I also buy your recalculation disconnect between NGDP and GDP. But it seems to me you don't have to throw the baby out with the bath water. The NGDP-GDP relationship is weakened to the extent that the recalculation story holds true - but it doesn't have to be entirely eliminated. And the extent to which the recalculation story holds true over time or the NGDP-GDP connection story holds true over time is certainly likely to vary.

Lee Kelly writes:

If a 3% growth rate for nominal expenditure were targeted, then the price level would have a higher growth rate when real productivity losses are experienced. On the other hand, exceptionally high real productivity gains might produce a stable (or even falling) price level. The economic troubles usually associated with a stable or falling price level would not occur, because nominal expenditure would remain on its growth path, i.e. there would be no shortage of money depressing prices, but merely greater productivity.

In other words, I think Arnold is right. In the aftermath of a economy wide misallocation, in the process of recalculation, attempting to meet a nominal expenditure growth rate would produce a greater rise in the price level. Much of the new money would just push prices up rather than redeploy idle resources, because those resources cannot be so easily put to new uses. In consequence there have been real economic losses. However, this is not a bug but a feature! Consider a world in which nominal expenditure was completely stable: price level changes would primarily be the result of productivity gains and losses. What is different in a world with a 3% growth rate in nominal expenditure, is that changes in the growth rate of the price level would primarily be the result of productivity gains and losses. Thus the rate of change in the price level actually contains information about true gains and losses in productivity.

The 3% growth rate for nominal expenditure (which is actually what Bill Woolsey proposes) would also create a better economic environment for the Great Recalculation to occur. The idea is to allow relative prices adaptions to do their thing while preventing monetary disequilibrium.

Steve Roth writes:

"the Garrett Jones view that workers are hired not to produce widgets but instead to build organizational capital"

This little blog-o-meme seems to be based solely on a one-sentence tweet.

Have we come to that?

Scott Sumner writes:

Arnold, My magnum opus (which I just finished and will post soon) has nothing on how to get out of a liquidity trap. Looks like I better add something.

David, I don't say much about that issue because mainstream economics has a reasonable answer. When in a deep recession a policy to boost NGDP will result in higher RGDP, with perhaps some inflation. When at full employment a policy of boosting NGDP will just create inflation. There are all sorts of sticky wage, sticky price, misperceptions explanations. I like sticky wages best, but think others also play a role.

Daniel and Lee, The way to test that question is have the Fed boost NGDP ten percent and see what happens. If its 7% inflation and 3% real growth, then Arnold's right and I will quit my blog. If its 7% real and 3% inflation, then I'm right.

Steve, I wondered about that too.

Doc Merlin writes:

@Steve.

No, it has to do with several well placed and respected members of the econ-blogosphere taking it up as it resonated with them.

Lee Kelly writes:
Daniel and Lee, The way to test that question is have the Fed boost NGDP ten percent and see what happens. If its 7% inflation and 3% real growth, then Arnold's right and I will quit my blog. If its 7% real and 3% inflation, then I'm right.
I don't understand why you'd quit your blog.

There has been a misallocation of resources, because there was a bubble, and that's what bubbles do -- misallocate resources. Therefore, a recalculation is necessary. In consequence, in a nominal expenditure targeting regime, we should expect to see more growth in the price level than usual. Do you deny there was a bubble? The question is surely more empirical: how much will the recalculation drag down potential productivity? But regardless of the answer to that question, it seems to me that targeting a nominal expenditure growth path would get closer to that potential.

In a nominal expenditure targeting regime, the rate and direction of price level changes would tend to reflect productivity gains and losses. In other words, a central bank can't target both nominal expenditure and the price level at the same time.

Larry writes:

If I remember some of Scott's earlier posts, he concurs that there was a bubble, and that the bubble popped, and that neither had to do with mismanaged money. I.e., there was a real shock, and real shocks have consequences.

His departure is to argue that poor policy amplified that shock, and turned what would have been an ordinary recession into the end of the world as we know it.

Part of the demonstration is that after other real shocks, economies quickly returned to above-trend growth (7% RGDP, etc.) for a period. That does not appear to be in prospect now, in part because credit remains crunched...in part because monetary policy is extremely tight, despite the misleading nominal interest rate indicator.

Lee Kelly writes:

Larry,

I agree that poor monetary policy turned an ordinary recession into the Great Recession. (But I would add that poor monetary policy probably inflated the bubble in the first place).

In any case, resources have been squandered. Surely this reduces the potential for economic growth, right? The opportunity to use those resources to more productive ends is gone; however much ground is made up afterward, potential growth must be lower than before.

In a nominal expenditure targeting regime, squandering resources will induce a faster rise in the price level. I see no way to avoid that, even if growth exceeds the trend afterward.

How much a 10% increase in nominal expenditure would be seen in real productivity versus the price level seems to depend on just how much potential growth has been reduced, i.e. on just how costly the recalculation really is.

Am I missing something?

Larry writes:

We may just have to disagree about monetary policy's contribution to crash-phase-1 (I think that micro issues were the problem, and that if BB had attempted to employ monetary policy that he would no longer hold his august position).

Yes, capital was misallocated. But that's happened many times before without consequences remotely like what we're now experiencing.

I agree that squandererd resources are often gone. I agree that some recalcs are more costly than others, but I don't see this one as particularly bad.

Again, the problems we see have both causes (real and monetary.) But we're screaming because of the monetary errors. We'd merely be groaning otherwise.

Scott Sumner writes:

Lee Kelly, I agree with Larry. I think the data shows that relatively little of the unemployment is directly due to resources being misallocated into housing. The vast majority is due to the recession. If AD rises quickly those resources should go back to work (although as Arnold pointed out in another post they are less likely to go back to their previous job than decades ago when millions of unemployment steel and autoworkers would wait around to get their old job back.)

VangelV writes:

It seems to me that many economists are over-thinking the problem and avoiding the simple and obvious explanation. When central the planners at Treasury and the Fed distort market forces by manipulating the supply of money and credit they create incentives for individuals to take advantage of the distortions. By doing so, the planners and the individuals that react to their policies ensure that capital is misallocated in ways that cannot happen in the absence of such distortions.

The bottom line is that individuals react to incentives and when governments and central banks pursue certain policies the incentives may lead to activities that work against the goals of the central planners. When free credit was being handed out those that had access to it had every incentive to pursue their own short term goals regardless of what that meant for the aggregate long term effect on the economy. Cheap money and credit gave many money managers the incentive to use massive leverage to provide great returns by doing pursuing a simple strategy like playing spread differentials. The compensation packages and bonuses were so large that it made sense to play the game for a few years even if one expected a total collapse over the long term. At the same time the cheap money allowed others to make huge bets that things would go terribly wrong as they expected constant frequent losses to be replaced by one very large win that would make large amounts of money for themselves and clients willing to take such small losses in order to hedge their reckless strategies on the long side of the markets.

While the loose money policies that allowed the market manipulation to be successful worked for quite some time, when it stopped working the real economy needed to adjust to the actual reality. Sadly, the central banks and governments were unwilling to permit such an adjustment for political reasons. As such, they tried to 'fix' the problem by doing what created it in the first place. By doing so, they prevented individuals from taking prudent step to recognize reality and rewarded the degenerate gamblers who created the mess in the first place. That ensures that the destruction of the Federal Reserve Note will come faster and that the eventual readjustment will have to be even more painful. While we should see a correction in the dollar index due to technical reasons, the fundamentals ensure that the FRN and the other fiat currencies will lose massive amounts of purchasing power over the next decade or so.

Comments for this entry have been closed
Return to top