Larry Summers Interview: A Cynic's Interpretation

Larry Summers seems to be saying:

• Obama is a central planner: his teams is trying to guide the U.S. economy to grow the "right" industries, and shrink the "wrong" ones.
• The economy is going to get worse. How does he know? Maybe he knows of plans to unveil yet more policies that further destroy work incentives, in addition to those unveiled already: like means-tested mortgage modification, mean-tested student loan modification, unemployment insurance extensions, state income tax hikes, and marginal federal tax rate hikes on the "rich"!

There has been so much talk of "fiscal stimulus" and ways the government can help, but I think the ultimately the debate about public policy and the 2008-9 economy will be whether the government made things worse, or was just impotent.

There are few economists now saying that government caused the housing bubble (quite an exaggeration, I think), but I seem to be the only one saying that, even with a housing crash, the depth of this recession would have been significantly less if the government had not done its damage too.

An Ill-Timed and Ill-Targeted Stimulus

Copyright, The New York Times Company

Last week’s disappointing employment report has raised the question of whether the stimulus package is working. Construction data over the last couple of years supports the view that federal spending would never provide much “stimulus,” in part because it was doomed to be ill-timed and ill-targeted.

The chart below graphs monthly construction spending, from 2003-2009, for three sectors: private residential, private nonresidential and public (of which residential is negligible). In order to adjust for the inflation and economic trends of the period, each spending series is expressed as a percentage of personal income less transfers.



Hindsight shows that housing construction sharply reversed trend in early 2006. Hindsight also shows that the housing boom used up significant numbers of workers and equipment, and the housing crash freed up those resources. That created opportunities for non-residential construction projects in the private and public sectors.

Although private construction did not significantly respond to these opportunities during 2006, private non-residential construction ramped up throughout 2007. That private non-residential activity remained high, and perhaps increased further, throughout 2008.

Meanwhile, the public sector still kept its construction activity at the low level that prevailed throughout the housing boom. Unlike the private sector, it did essentially nothing to put unemployed workers and equipment to work.

In January 2009, Congress passed a stimulus bill that was supposed to put some of the unemployed back to work. But January 2009 was almost three years after the spending opportunity presented itself, and two years after many projects were under way by those in the private sector who saw these opportunities.

Moreover, typing and signing a bill is not the same as doing construction work. The public construction series shown in the chart goes through May 2009 — four months after the bill became law — and shows only a tiny increase by that time. We are told that we will actually notice stimulus bill spending sometime in 2010.

One might argue that the housing crash presented the public sector with little opportunity because residential construction workers and equipment are not useful for the kinds of transportation construction that dominate public construction spending. That is at most partly true, because some housing workers have responded to the housing crash by re-qualifying for highway work.

More importantly, that argument (if correct) only provides another reason public spending cannot both put unemployed housing resources back to work and provide value to the public: The housing resources are ill-suited for public sector work.

The public sector — especially at the federal level — is not known for its agile and effective responses to market conditions. In authoring the stimulus bill, Congress and the president relied too much on federal spending and too little on the private sector.

C-S/OFHEO comparison from Professor Nunes

Professor Nunes sent me this chart comparing the Case-Shiller national housing price index with the OFHEO index for the Pacific region only. The close correlation between the two is consistent with the hypothesis of Rebecca Wilder that the "national" Case-Shiller index does not represent well the nation as a whole.

A New Comparative Analysis of Housing Price Indices

I applaud Rebecca Wilder for doing some hard work. I have not yet studied the details, but she seems to be saying that some of the Western cities are getting too much weight in the Case-Shiller index, and as a result the OFHEO index is giving a more accurate picture.

That agrees with my rough impression. If we are right, housing prices suggest that the housing market has bottomed, although construction spending is not yet supporting that conclusion.

Can Unemployment be Blamed for the Foreclosure Crisis?

Unemployment and foreclosure rates have skyrocketed over the last 1-2 years. The regions experiencing the most unemployment seem also to be the ones that have the highest foreclosure rates.

Both economic theory and data suggest that unemployment by itself cannot create much foreclosure. Negative home equity -- having the mortgage "under water" as they say -- causes foreclosures.

Theory: a homeowner always has the option to stop paying his mortgage. Although state laws are somewhat different, to a good approximation the worst case scenario for a homeowner who stops paying is that he can no longer own or occupy the house, and may suffer a reduction in his credit rating that might raise his costs of future borrowing. But if the combined present value of these costs were less than the present value of his promised mortgage payments, he can do better than paying in full. That’s probably an important reason why, as of early 2009, more than five million homes were already either in foreclosure or their owners were delinquent on their mortgage payments.

“Inability to pay” is probably not enough by itself to create a foreclosure, because a homeowner unable to pay but with positive home equity may want to sell his home to pay the loan (and thereby retain his home equity) rather than invite foreclosure. Conversely, a person who is quite able to pay might rationally invite foreclosure, for the reason cited in the text above.

Liebowitz (2009) finds that negative equity was a much more important factor than unemployment in causing the foreclosures that occurred in the second half of 2008.

With that said, unemployment probably magnifies the effect of negative equity. If everyone had remained employed, and falling housing prices were all that had happened, then lenders would probably work out a deal with their borrowers so that their negative equity was brought back closer to zero, which would prevent the borrowers from seeing foreclosure as their best option. But when some people are unemployed and others are not, lenders will make more (lose less) by discriminating among borrowers: likely foreclosing on those in the most financial trouble, working out a deal with those with mediocre incomes, and demanding full payment from those most able to pay.

As I have pointed out, this rational discrimination among borrowers also raises the unemployment rate, so foreclosures may have a bigger an effect on unemployment than unemployment has on foreclosures.

Big Surprise: The Stimulus Isn't Working

Right now the unemployment rate is 9.5 percent. The administration had said that, thanks to their fiscal stimulus, it would be less than 8 percent.

I don't blame them for missing a forecast. Forecasting is difficult. And this labor market has been surprisingly bad.

I do blame them for misleading the public by saying that the stimulus package would improve the unemployment rate (relative to whatever admittedly unpredictable outcome it would have been without stimulus). They claimed that, by June, the unemployment rate would be improved by almost 0.5 percentage points. That clearly has not happened.

Although we did not know the future of the economy, we knew the stimulus package wouldn't help.

Women's Employment Share Reaches 49.8%

in May 2009. Pretty much the pattern has been: job losses are primarily male, so male employment is falling towards female employment.

Labor Market Continues Down

Employment is still falling pretty sharply.


While GDP has hardly surprised me, I did not anticipate last fall that employment would get so low. Even with the benefit of hindsight, I am not sure I see a complete explanation.

Part of the story is the plethora of means-tested public policies: they discourage work and raise employer costs:

• Mortgage modification
• Student loan forgiveness
• IRS enforcement of prior tax debts
• extended unemployment benefits

But likely there is more to the story: too bad I don't know more than that.

Construction Spending Released Today

Today the Census Bureau revised its estimates of construction spending January 2007-April 2009. The revision says that more of the INCREASE in non-residential construction spending occurred during 2007 than occurred in 2008 (to see the time series prior to the revision, see here). Still, non-residential construction is higher now than it was when the recession began.



Today the Census Bureau also released an estimate of May 2009 construction spending. Residential spending continued to fall. Private non-residential spending increased, but little.

I have noted that construction could bottom out and recover without a rise in private construction, because the private activity would be crowded out by public spending. However, the public spending fell too.

Housing prices would likely stop falling, and begin to rise, before construction spending recovered because prices can adjust more quickly. Nevertheless, a housing recovery cannot be declared until residential construction spending stops falling.

Inflation and Investor Sentiment

Copyright, The New York Times Company

The easy monetary policy at the end of 2008 has set up our economy for inflation, but the timing depends in part on how investors behave.

Last week I showed how the Federal Reserve dramatically expanded the monetary base (that is, the value of currency, coin and Federal Reserve deposits) at the end of 2008, and how nothing like this occurred during the onset of the Great Depression of the 1930s.

Still, even though monetary policy is so different in this recession as compared with the policies of the 1930s, inflation has not yet been very different.

During most of our lifetimes, there has been inflation: The prices of things we buy have generally increased over time. Only on rare occasions have consumer price trends suddenly changed directions.

One of those occasions was 1929.

Consumer prices were pretty constant in the 1920s. The chart below picks up the story in January 1929 with the red line. That line measures the (seasonally unadjusted) consumer price index in each month through July 1930, normalized so that October 1929 is 100 (for example, the value of 97.9 in April 1929 means that prices then were 2.1 percent lower than they would be in October).

In the fall of 1929, the inflation stopped (incidentally, the stock market crashed in late October of that year) and prices headed down, falling almost every month for almost four years.

For the first 15 months or so of this recession, consumer prices have followed a similar pattern. The blue series in the chart shows the consumer price index for 2008 and 2009. Like the 1929 series, the 2008 series is normalized so that October is 100.

The chart shows how consumer prices also rose in the spring and early summer of 2008. Inflation had stopped by the fall (there was a stock market crash in October 2008, too), and consumer prices headed down. In fact, the deflation at the end of 2008 brought prices down more than 4 percent in a couple of months, as compared with a 1 percent drop at the end of 1929.

The actions of the Federal Reserve and its chairman Ben S. Bernanke guarantee that we will not experience a four-year deflation like that of the Great Depression. But investor sentiment is an important reason why the short-run inflation patterns have been similar in 2008-’09 to what they were in 1929-’30.

During both episodes, investors had a sudden reduction in their willingness to hold private sector debt and equity and to purchase goods, and a sudden increase in their desire to hold “quality” assets like Treasury bills.

An increase in Treasury bill prices is one way markets adjust to this change in demand — and we saw this in September through December of last year — but another market adjustment is for the prices of goods, equities, and private sector debt to fall (or rise less than they would have) as investors pull their money out of these categories. Deflation is, by definition, a drop in goods prices.

Part of the next inflation may be the reverse of this process: Investors suddenly shift their demands from “quality” assets back to equities, private sector debt and goods. As some of the commenters explained last week, a sudden investor shift like this will be associated with a sharp reduction in the value of the dollar.

If I could predict exactly when investor demands will shift away from “quality” assets, both I and the readers of this blog might get as rich as the billionaire financier Warren E. Buffett. But recognizing the role of investor sentiment at least helps us appreciate why the timing of the next inflation is so uncertain.