Arnold Kling  

Scott Sumner Asks a Question

PRINT
Fight, Flight, Submission: War... Mea Culpa...

Scott Sumner asks,


So why does the banking crisis in 2008 cause low NGDP in 2011?

Possible answers:

1. Because Reinhart and Rogoff say that financial crises cause pain for a long time. They don't have an explanatory model, but they do have data.

2. Because the Fed made forecasting errors. Right-wingers are fond of brandishing charts showing that the unemployment rate with the stimulus is on a worse trajectory than what was forecast without the stimulus. That may or may not be evidence that the stimulus failed, but it is evidence that standard forecasts were not sufficiently pessimistic about the economy. Assuming the Fed used standard forecasts, that would explain the inadequate monetary expansion back then. It doesn't explain their reluctance to expand now, though.

3. Some Austrian economists slipped something in The Bernank's drink and hypnotized him into believing in PSST, so now he thinks that all the unemployment is structural.

4. Since the crisis began, the Fed has been focused on strengthening the balance sheets of the top banking institutions. A monetary expansion would mean more lending by banks, which would make their balance sheets more fragile. Better that they should just hold interest-bearing reserves. In that sense, the goal of strengthening the banks conflicts with the goal of restoring full employment, and the Fed has chosen one goal over the other.

My money is on (4). Although, as you know, I think that if the Fed tried really hard to restore nominal GDP, it would not do much for employment, because I believe in PSST and in Reinhart-Rogoff.


Comments and Sharing


CATEGORIES: Monetary Policy



COMMENTS (13 to date)
Ironman writes:

Here's the explanatory model for why Reinhart-Rogoff's observation seems to hold:

1. High national debt levels are often associated with an increased probability of default, especially if a significant portion of the debt is held by non-domestic interests.

2. That increased risk of default increases interest rates above the level they would otherwise be.

3. That effective increase in interest rates for the government is transmitted to the nation's businesses (assuming the government's borrowing rate is the "risk free" rate for the country), which increases the cost of doing business. As a result, less business than might otherwise occur gets done, which makes for lower GDP.

MarkS writes:

This is not a banking crisis. When will people get that? It is a household crisis. This didn't start with the banks. It started with the households. We don't need to fix the banks. We need to fix the households. That is why your monetarist gimmicks are all failing.

The bipartisan CBO already confirmed that the fiscal stimulus is helping. But now we're cutting it off and it will result in failing fiscal AND monetary policy.

This isn't rocket science.

Charles R. Williams writes:

The banking crisis and the recession are interrelated but distinct phenomenon. Yes, things would be better if we had seized the banks and recapitalized them by converting bank debt to equity. The cause of the financial crisis however was bad mortgages built into AAA-rated securities. The cause of the recession is the fall in net worth. Either of these phenomena could have happened on their own without the other.

The problem is not household debt it is household net worth. Consumption will not recover until net worth exceeds the level of 2006 plus expected growth. Households are trying desperately to save and their are few investment opportunities - hence, zero interest rates, exploding cash balances and high unemployment. Any effort to cut household debt directly will fail because John's debt is Joe's asset.

Sumner thinks that by goosing NGDP real output will increase. But why? Faster inflation will not create wealth. Destroying confidence in the dollar can cause the dollar to fall and commodity prices to rise through speculation. This has already caused the consumer to pull back. But by impoverishing people who have jobs we can put people back to work. This only works if we view labor as a commodity.

While the administration's policies have made things worse, we were fated to endure a long and painful recession because macro-economic theory offers no tools to create true wealth.

lostgen writes:

labor is a commodity, albeit one with sticky prices and those would fall if they weren't sticky. Their failure to fall in relative terms causes the unemployment. That might not sound very humanistic, but not viewing them as commodity ill just blur our understanding.

Charles R. Williams writes:

A job represents a large investment on the part of a worker and an employer. A good job is a unique combination of skills and a value-added activity created jointly by the two parties. Job creation takes time and resources. Mal-investment created by asset bubbles results in jobs that disappear when the bubble breaks. The dust must settle and people pick up the pieces and create new jobs that they hope are genuine. This process may take years and is impeded by government policies that stir up dust and cloud the landscape.

Saying that wages are sticky is saying that there are contracts, explicit or implicit, between employers and workers that have to be renegotiated or contracts that could be negotiated and that fiddling with the value of the dollar will somehow help this process. This might have been true in the 1950s. Today, inflation may be a cure for the minimum wage but that's about it.

Lewis writes:

I agree with (4).

"if the Fed tried really hard to restore nominal GDP, it would not do much for employment."

This reminds me of your writing in a previous post about manufacturing: "The extra nominal GDP would show up in part as a higher price level and in part as more jobless output..."

If PSST is true, then why judge policies mainly on the employment created? The additional "jobless output" would have to go to someone. The exchange of "jobless output" would enhance consumer and producer surplus.

Further, given the ongoing layoffs by state and local governments, it seems unlikely that higher nominal income would have little effect on unemployment. Higher transaction prices would raise sales tax collections. Higher real estate prices would raise property tax revenues. Higher commodities prices would boost royalties.

Maybe this employment is unproductive. but as a matter of science I think that higher nominal income would definitely stop some state and local layoffs.

Matthew C. writes:

Growth is dead because of the metastatic growth of the central state, crony capitalism and the TBTF nexus in Manhattan.

The current system has morphed into a Rube Goldberg contraption being constantly patched and duct taped together with money printing and based on lies. Lies like keeping social security and medicare obligations off the books (no private company could do this!). Lies like the substitution of mark to myth for mark to market -- carrying of nonperforming second mortgages at 100% when the first are underwater and in default. Lies like the periodic "tweaking" of the definition of CPI so a declining economy can be misrepresented as still growing, albeit sluggishly. Sometimes the mask even slips a bit, as when Euro group president Jean-Claude Junker admitted that the masters of the universe will lie to the public whenever they deem it to be useful: "Because the financial markets in Europe were still open and trading was still underway on Wall Street, we had to deny the existence of the meeting."

But the unfolding disaster is clear and unmistakable. The constant reduction in labor force participation and increase in food stamp utilization. The erosion of the dollar measured by objective metrics like gold. The fiscal catastrophe where 40 cents of every dollar are borrowed (and off-and-on printed up by the Fed to help support the borrowing). The ever-declining respect by Americans in our government institutions and the direction of the country. And of course this is a global phenomenon, from the colossal destruction of capital in China from building empty railroads leading to empty cities, to the fiasco unfolding in the destruction of the Eurozone. ZIRP. QE 1, 1.5, 2, 2.5, and many more to come, each designed to debase the dollar. The US Government IS virtually the entire mortgage market. Politically connected organizations like Goldman Sachs get bailed out 100% on the dollar for bad paper the market valued for pennies.

We're in an extended Wile E Coyote moment before the plunge into currency collapse and the end of an unsustainable system. Get prepared now before the global reserve currency faces its "Madoff Moment" of recognition.

Shayne Cook writes:

I'm fairly familiar with the "textbook" concepts of macroeconomics - I've been studying and teaching from them for a few years. And the proscribed methods in textbook macro for dealing with cyclical recessions - both fiscal and monetary - do (historically) seem to be effective in attenuating, if not precluding, cyclical recessions. Emphasis on cyclical.

As I noted in a previous post comment, ALL recessions have a cyclical component, against which conventional macro methods are nominally effective. I suspect that is true in this case as well. But the degree of "textbook" macro remedy - fiscal (taxes/spending) and monetary easing - that has been applied to this recession has not produced satisfactory results in terms of NGDP growth or employment.

Scott Sumner would probably call this "seat of the pants" macro on my part, but the concept of just stating that "banking system-caused recessions last longer" doesn't seem compelling in explaining the current lack of satisfactory results. Especially so given that, as Arnold noted, no one seems to be able to explain why. (Frankly, I don't buy into the notion this is/was a banking system caused recession - coincidental, not causal.) So here goes with some "seat of the pants" macro...

Again, ALL recessions have a cyclical component. Some/most are majority cyclical in nature. Historically, few have had a significant structural component. This one does have a significant structural component. The proscribed "textbook macro" treatments are effective against the cyclical components/artifacts of a recession, but it is likely those same treatments can even be counter-effective against the structural components.

Charles R. Williams above, and Arnold with his concept of Patterns of Sustainable Specialization and Trade (PSST), have identified much of the structural nature of this recession. And it is significant. Unfortunately, there are 3 subsequent problems. The first is convincing the majority economic and political communities that there even is a structural component, let alone that it is significant. The second is identifying effective remedy for the current structural artifacts (classic "textbook macro" has none). And third is evolving and incorporating structural theories/remedies into classical "textbook macro" for future economists.

J Oxman writes:

Jeffrey Hummel has good evidence to believe (4) is the case: http://www.independent.org/pdf/tir/tir_15_04_1_hummel.pdf

Keith writes:

Possible answer:

In 2008, we woke up to realize that some of our capital could no longer generate income and this led to substantial write-downs in asset values (i.e., loss of wealth).

Since 2008, we have decided to reduce private investment and increase consumption in an effort to maintain pre-2008 living standards.

The opportunity costs of allocating income toward consumption rather than investment has been to slow the accumulation of capital and our ability to grow income.

mick writes:

#2) Why not conclude a negative correlation?

Scott Sumner writes:

I don't follow point 4 at all. There is very strong data showing that the balance sheets of banks are closely correlated with NGDP growth in the crisis. Tight money in 2008-09 was a disaster for the banks. They've done better since March 2009, as NGDP growth expectations rose from negative back to positive levels.

Very few of the financial crises that R&R looked at were associated with the country's currency appreciating strongly during the worst of the crisis. Two that were (US 1931-33 and Argentina 1998-2002) were followed by rapid growth, despite really bad supply side policies.

Guy in the Veal Calf Office writes:

Why, even during their pre-2008 good times, was French, Italian, Spanish, etc unemployment twice the U.S. rate? One argument is that punishing labor regulations and negative taxes will produce consistently higher unemployment compared to the absence of such things. Is that bunk?

If not bunk, then wouldn't the increase in punishing labor regulations and negative taxes here in the US beginning in 2009 play a role in higher unemployment.

Comments for this entry have been closed
Return to top