Sunday, November 29, 2009

Tantalized by Panel Data

Much conventional wisdom among econometricians and applied economists extols the virtue of "panel data": data that follows individuals or regions over time. State panel data -- data that follows each of the 50 states over time, is thought to be especially virtuous because there's never a problem with losing track of one of the states (unlike panels of individuals, from which members can nonrandomly drop). Especially lauded is the practice of including so-called "time effects" in the model, which means that changes over time are NOT examined except to the extent they occur differently among the individuals or regions.

One interesting question in industrial organization and public economics is the "incidence" of excise taxes -- that is, whether a tax on the sales of specific items will reduce what manufacturers receive for making the item, or increase what consumers pay for it. Cigarettes are an important instance of this, because the taxes are large, and policy-makers are concerned about the harmful health effects of smoking. Because the effects of taxes are determined by the interaction of supply and demand, knowledge of cigarette tax incidence would tell us about the nature of pricing in the cigarette industry, as well as effects of federal cigarette taxation on health and state revenues.

Cigarette tax incidence has usually been studied with state panel data. After all, states differ widely in terms of their use of their taxes, and the dates at which they change their rates. The state panel studies usually find that each penny of cigarette tax raises retail cigarette prices by almost exactly a penny, with little effect on the price per unit received by manufacturers.

By this logic, the $0.62 per pack federal excise tax hike this April would raise retail cigarette prices by almost exactly $0.62 per pack, reduce cigarette smoking in an amount commensurate with the $0.62, and have little effect on the amount cigarette manufacturers receive per pack shipped to U.S. retailers.

I have long been dubious of state panel data for this purpose, because the supply curve across states is very different than the national supply curve. Wholesalers in one state can pretty easily ship cigarettes to a wholesaler in a state with a different tax situation, and this possibility requires wholesale cigarette prices to be essentially the same in each state at a point in time. National pricing is very different because nothing requires the wholesale price in, say, 2006 to be the same as it is in 2009 (I'm told that cigarettes do not store well over long time periods).

We have enough data now to check whether I'm right. I found a monthly Consumer Price Index for cigarettes at The CPI is an index, telling us percentage changes in retail prices from one month to the next, but do not tell us what a pack actually costs. The aforementioned literature often gets annual (measured in November) cigarette prices from the "Tax Burden on Tobacco", so I used that to pin down the level of national average cigarette prices in Nov 2007 ($4.20 per pack; see also this article that puts average prices at $4.10 in early 2009), and then used the monthly CPI to measure retail prices for all other months Jan 2007 - Oct 2009.

The chart below displays the results. Notice that the vertical axis is scaled so that each tick is $0.62/pack -- if the state-panel studies could be used to project federal tax effects, then prices would go up by exactly one tick.

Instead, cigarette prices increased a lot more than 62 cents: more like $1. The state-panel results wildly underestimate the effect of the tax on retail prices, and thereby wildly underestimate the effect of the federal tax on smoking and excise tax revenues received by the states.

[The vast majority of states kept their excise tax rates constant during the first half of 2009. Exceptions are Arkansas, Kentucky, Mississippi, and Rhode Island.]

Saturday, November 28, 2009

Are Banks Undermining Loan Modification?

This article claims that they are:

If the terrible government loan modification program is in fact broken, let's leave it that way!

Friday, November 27, 2009

Prescott and Mulligan: Same Song, Different Verse

I just found this interesting July 2009 presentation by Professor Edward Prescott. He says that the anticipation of future taxes/bad incentives is depressing the economy. I say that the bad incentives are already here.

Now is not the time to quibble: what we both say is vastly different from the conventional wisdom, and we both recommend that government refrain from making it worse.

Wednesday, November 25, 2009

One Minimum Wage Increase With a Side of Fries, Please

Copyright, The New York Times Company

Economists have debated the employment effects of the minimum wage. A recent study of obesity now weighs in on this debate.

Many economists expect the minimum wage, if it has any effect, to (among other things) raise employer costs and therefore reduce employment, especially among people who are likely to work in minimum wage jobs like teenagers and restaurant workers.

However, inspired by a study of a 1992 minimum wage increase in New Jersey, some economists have suggested that minimum wages can increase employment, by helping to cure pre-existing problems in the labor market. In their view, a higher minimum wage could increase employment and output at employers of low-wage workers, and a lower minimum wage would reduce them.

The typical example is a fast-food restaurant.

The minimum-wage-cures-labor-markets view says that a higher wage level causes fast-food restaurants (like other employers of low-wage workers) to hire more workers, produce more fast food and sell more fast food.

More fast food sold also probably means more obesity. Thus, if the minimum-wage-cures-labor-markets view was correct, a higher minimum wage would, all things being equal, probably result in higher obesity rates.

More conservative economists would argue, though, that high minimum wages restrict employment by fast-food restaurants, which means less fast food produced, which means less obesity.

In other words, the traditional economics view implies a lower minimum wage would, all things being equal, result in more obesity.

As you may know, Americans have indeed been getting more obese over the last couple of decades, with increased consumption of fast foods contributing to that enlargement. During most of that period, the inflation-adjusted federal minimum wage had been falling.

A recent study by the researchers David Meltzer from the University of Chicago and Zhuo Chen from the Centers for Disease Control and Prevention now finds that low inflation-adjusted minimum wages are partly to blame for increased obesity.

If their study is correct, it suggests that a higher minimum wage indeed reduces employment and output at fast-food restaurants, and makes it a bit easier for Americans to adopt healthier eating habits.

As with any new study, time is needed to digest the methodology and results, and integrate them with the previous literature. But expect the fight against obesity to weigh in on the debate about low-wage labor markets.

Tuesday, November 24, 2009

What Does It Cost to Buy a Recession?

By Oct 2009, U.S. labor usage was more than 10 percent below trend. Even if it returned to trend by the end of 2010, that would put labor usage about 20 year x percentage points below trend (i.e., an average 6-7 percentage points below trend for each of three years). A year's labor income is about $10 trillion, so that's $2 trillion that labor income has been reduced over the three years.

How to Purchase a Recession
Suppose for the moment you had a lot of $ to bribe people not to work, or employers not to hire. What method of allocating the bribes would reduce employment the most? How much would it cost you to purchase a recession like this one?

If you simply paid people not to work, shrinking the labor usage by that much might cost about $3 trillion.

It can be a $3 trillion task because people who would not work anyway may take you up on your offer not to work. If you could target your bribes, you would want to target them to the weakest employment relationships -- those for which supply is closest to demand. With very well chosen targets, you could make a recession like this for a mere $100 billion.

But do not expect that you could target so well in practice, because it's difficult to know which employment relationships are the weakest, and once you started paying people for what appeared to you to be weak employment relationships, others might put on the appearance. But at least you could try to target the types of people who are generally expected to be working soon, such as persons searching for jobs (interestingly, that's what unemployment insurance does).

All together, you would be hard pressed to make a recession like this for less than $1 trillion.

UI is an Illustration, but not the Major Force
Unemployment insurance (UI) reduces the employment rate, by increasing the pay someone can earn while not being employment, and reducing the after-tax pay earned while employed. But I raised the question above to demonstrate that UI cannot be the only, or even a major, reason why employment is so low.

Recall that UI benefits are voluntary: nobody forces you to take them. Thus, even if UI had the purpose of reducing employment (which it is not), it could not be much more effective per dollar of expenditure than the hypothetical "recession purchase" discussed above.

UI will spend something like $300 billion for 2008-10, and obviously that $300 billion is not for the PURPOSE of minimizing employment. To make this recession by itself, UI would probably have had to spend more than $1 trillion. (this is the same argument I made in "Public Policies as Specification Errors" for why UI was not a major factor in the Great Depression, either).

Mortgage modification is almost a big enough operation by itself to make this kind of dent in the labor market (whether it actually does is another question). For example, if the Obama Administration achieved its goal of modifying 9,000,000 mortgages and each mortgage were written down an average of $75,000, that would be a total of $675 billion.

If you took the combination of mortgage modification, UI, big parts of the "stimulus" law, and other anti-employment policies, we probably are looking at over $1 trillion worth of spending that encourages people to have lower labor incomes.

Bottom Line
Although it's easy, and at least partly appropriate, to say that government spending of various sorts has reduced employment over the past couple of years, note that buying a recession is no cheap enterprise, and buying a recession of this size may be beyond even what governments can afford.

Investment and Housing Prices Among Various Data Released Today

A couple of housing price indices, plus national accounts revisions were released today. Much of the new data is not newsworthy, but I did notice that real nonresidential investment was even lower in Q3 than it was in Q2 -- another indicator that employment is not coming back soon.

I also noticed that the BEA price index for residential structures investment was (marginally) lower for the seventh quarter in a row. Here is a comparison of quarterly Case-Shiller, OFHEO, and BEA (all expressed relative to the PPI for housing construction).

Monday, November 23, 2009

Question About Deflation

Robert asks
"In our industry, the manufacturers claim to be holding prices, but are quietly making all kinds of deals to "help" us be more competitive.

Our competitors are taking those incentives and chasing prospects with what appears to the dealer to be lower prices across the board.

This is new behavior. For the past 8 years, no one really asked what the price was. Currently it's the prime topic.

Then I went online and ordered a pizza from pizzahut Friday. Suddenly every pizza is $10, half the price of the past few years. If that's a short term promotion were ok. If that's the new reality, are we in trouble?"

Let me rephrase Robert's post as three questions:

(1) Is deflation -- that is, a general decline of all prices (both the prices at which we buy, and those at which we sell) -- a problem? Theoretically, it is not a big problem, but just redistributes wealth from those with dollar-denominated liabilities to those with dollar denominated assets. However, this recession arguably got going because of the "underwater mortgages/foreclosure" problem -- a problem that get's better with inflation and worse with deflation. For more on this, see "Inflation, we need you!".

(2) Is there deflation right now, or will there be in the near future? I think we have a bit of inflation right now, and expect more inflation in the next couple of years (see here). It's always a bit difficult to know the inflation/deflation rate precisely, because a lot of price changes can be pretty subtle, as with the promotional discounts indicated in Robert's post. But the Bureau of Economic Analysis and the Dept of Labor have enough serious ways of measuring it that, together with the recent commodity price inflation, I am confident that we do have inflation.

(3) If not general deflation, what is Robert supposed to make of his observations? It's no surprise that various manufacturing prices have been falling a bit over 2009, after falling significantly at the end of 2008. If he's seeing more drops than a "bit" then that's some bad news for his segment of manufacturing.

(4) The pizza bargains may indicate that his region's economy is tougher than the national average, or merely that Robert hasn't purchased a pizza for a year or two (maybe the case -- that's about the time frame he wedded his lovely bride!!).

Friday, November 20, 2009

Bears Fan

Does this picture explain why I blame a large fraction of public policy mistakes (bank bailout, mortgage modification, min wage hike, etc.) on the previous administration?

Wednesday, November 18, 2009

No News Housing Construction Report

I don't see much news here:

The Minimum Wage and Teenage Jobs

Copyright, The New York Times Company

Teenage employment has fallen sharply since July. The most recent minimum wage hike may be an important factor.

Many economists expect the minimum wage, if it has any effect, to (among other things) raise employer costs and therefore reduce employment, especially among those who are likely to work in minimum-wage jobs, like teenagers and restaurant workers.

In July 2007, the federal minimum hourly wage was increased for the first time in 10 years, from $5.15 to $5.85. It was increased again a year later to $6.55, and increased yet again this July to $7.25.

Because the minimum-wage law still permits employers to pay more than the minimum, economists agree that a low minimum wage has smaller effects than a high minimum wage. The inflation-adjusted federal minimum wage had gotten to its lowest in decades by early 2007, so the July 2007 increase should have had the smallest effects of the three.

The July 2009 increase should have the largest effect, because the combination of the two previous hikes and some deflation ($6.55 bought more in June 2009 than it did the previous summer) had already gotten the inflation-adjusted minimum wage relatively high.

The chart below shows a seasonally adjusted index of the percentage of 16- to 19-year-olds with jobs. That group is especially likely to be affected by minimum-wage legislation. Of course, this is a recession period in which employment has been falling for essentially all groups, so for reference the teenage percentage has been converted to an index by dividing by the percentage for all people, with July 2009 set as the benchmark (i.e., the teenage employment rate that month has been set to 100).

The chart shows teenage employment index values greater than 100 early in the recession, which means that employment rates fell in greater percentages for teenagers even before the July 2009 increase, as it did in prior recessions (even recessions without minimum-wage increases). With the index falling somewhat less than 1 percent a month before July 2009, we would expect the index to be somewhat below 100 after July 2009 even if the minimum wage hike had no effect.

But the teenage employment after July 2009 seems sharply lower. By October 2009, the index had fallen to 92.1 — a drop of about 8 percent in just three months — whereas the prior 8 percent drop had taken more than a year. This suggests that the 2009 minimum-wage increase did significantly reduce teenage employment.

Before this recession, economists hotly debated the employment effects of the minimum wage, with special attention to a 1992 minimum-wage increase in New Jersey (this book got it started, and this book is a good source for the opposing view).

More work is needed to determine whether the 2009 experience is fundamentally different from the earlier episodes that have been studied, but next week I will describe a new study that “weighs in” on those episodes.

Tuesday, November 17, 2009

PPI For Housing Construction

The housing PPI looks pretty flat over the last six months. That is one indicator that housing prices will be flat, which means that mortgages will not be going further underwater.

The bad news is that the housing PPI in the last six months has not participated in the moderate inflation seen in the wider economy over the last six months, which suggests that housing prices have not yet participated in that inflation.

Monday, November 16, 2009

Minimum Wage Regrets?

A couple of years ago, a number of "leading economists" endorsed the federal minimum wage law that legislated the current minimum of $7.25/hr:

Henry Aaron The Brookings Institution
Kenneth Arrow Stanford University
William Baumol Princeton University and New York University
Rebecca Blank University of Michigan
Alan Blinder Princeton University
Peter Diamond Massachusetts Institute of Technology
Ronald Ehrenberg, Cornell University
Clive Granger University of California, San Diego
Lawrence Katz Harvard University (AEA Executive Committee)
Lawrence Klein University of Pennsylvania
Frank Levy Massachusetts Institute of Technology
Lawrence Mishel Economic Policy Institute
Alice Rivlin The Brookings Institution (former Vice Chair of the
Federal Reserve and Director of the Office of Management and Budget)
Robert Solow Massachusetts Institute of Technology
Joseph Stiglitz Columbia University

[read the longer list here. Interestingly, the University of Chicago does not appear on the list.]

They all said: "we believe the Fair Minimum Wage Act of 2005’s proposed phased-in increase in the federal minimum wage to $7.25 falls well within the range of options where the benefits to the labor market, workers, and the overall economy would be positive."


"the weight of the evidence suggests that modest increases in the minimum wage have had very little or no effect on employment."

An ambitious journalist might ask them whether they still believe these things, and in particular whether that they believe this July's increase had "positive" effects.

Thursday, November 12, 2009

Why is Employment Falling? How to Turn it Around?

Debated on The Kudlow Report tonight:

Welcome Kudlow Viewers!

Below are my posts on the fiscal stimulus. See also my list of reasons why government policy has been reducing employment, not raising it.

To see my posts on other economics subjects, please click on "all posts" above.

I also blog weekly at the New York Times www (one of my favorites is here).

Kudlow Show Tonight

I will be on CNBC's Kudlow Show Tonight, sometime between 7 and 730p eastern time.

What Happens Next?

I'm not sure. My version of the real business cycle model (some of my academic papers on it are here and here, and I began applying it to this recession about a year ago, such as here and here) does not give a definite answer, but it does drastically narrow the possibilities.

My model has no adverse productivity shocks, no shocks to capital markets (these variables just react to events in the labor market), no monetary policy, and no fiscal stimulus. Simply put: I view this as a one (type of) shock recession, and the labor market is ground zero for that shock.

One version of the model has a labor market distortion that gets progressively worse for two years, at which point it remains at that higher distortion forever. Specifically, the average marginal tax rate ultimately increases about 10-15 percentage points (more accurately, the after tax share is cut by 22 percent), but it takes 2 years for the full marginal tax rate hike to occur.

The model and data are shown below. Note that latest 12 months (4 quarters) of the data were not available when I first began writing about this model.

[APL is real GDP per hour worked -- labor productivity -- which in the model is in fixed proportions to the marginal product of labor]

The good news from this first scenario is that the labor market will stop getting worse in 2010 (ie, employment and hours will stop falling further below trend). The bad news is that aggregate hours will never return to that previous trend, even part way. Consumption will ultimately be further below trend than it is now.

I am still working on it, but I think there's another version of the model that would fit the same data: the labor market gets even worse for a couple of more years, but eventually will be closer to the previous trend than we are now. The bad news from this second scenario is that the labor market will not stop getting worse until beyond 2010. The good news is that consumption and aggregate hours will eventually return to their previous trends, at least most of the way.

Either way, the inference I am making from consumption behavior -- it has fallen a lot by historical standards, but far less than labor has fallen -- is that the present value of lost labor is great, but much of that loss labor has not yet occurred. Whether the remaining lost labor is spread over the infinite future (the first scenario above), or concentrated in the next couple of years (the second scenario above), I do not know.

Wednesday, November 11, 2009

A Jobless Recovery

Copyright, The New York Times Company
The Bureau of Economic Analysis recently confirmed what everyone suspected, that real spending and incomes grew again from the second to the third quarters of this year. Yet we also learned on Friday that employment still fell in October, as it had in previous months. Although employment will someday turn upward, I suspect that this divergent pattern for spending and employment will continue for a while.

One bit of conventional wisdom about this recession is that it was caused, or at least significantly worsened, by a paradox of thrift: Consumers suddenly ceased to be willing or able to spend as they once did. But I have argued against that conventional wisdom, based in part on the fact that work hours and employment have fallen much more than either consumer spending or personal incomes have.

Indeed, real personal consumption expenditure was higher in September 2009 than it was a year earlier (as was real personal disposable income), while work hours had fallen 7 percent. This recession cannot be understood merely as the consequence of low spending.

A variety of models can help explain the recession so far, and to predict where it may be going, but here I’d like to focus on the two variables emphasized in my own research: productivity, and what are known as labor market distortions.

Productivity is the amount produced per hour worked. If productivity grows, it means that output can grow faster than employment, as it did from the second to the third quarter.

During several previous recessions, productivity was falling. Yet this recession has been different, with productivity throughout 2008 and 2009 being higher than it was when the recession got started.

Productivity growth has been especially strong during the last two quarters, perhaps in part a recovery from somewhat slower (but still positive) productivity growth at the end of 2008. Productivity could surge a bit more — maybe for another quarter — without exceeding the long-term trend. If so, inflation-adjusted output would continue to grow faster than employment and hours over the next couple of quarters.

Labor market distortions are a collection of factors that hold back employment, even when employees are creating a lot of value.

These distortions include difficulties in job search, income taxes, minimum-wage laws and incentives that are eroded by means-tested government benefits (determining whether someone should receive benefits based on things like the person’s income). These factors can be difficult to quantify individually, but we know from the poor employment results that at least some of them are important.

Labor market distortions have gotten progressively worse during this recession. The federal minimum wage, for example, was increased once shortly before the recession began, a second time in the summer of 2008, and yet again this summer. The housing collapse has also had multiple harmful effects, such as impeding families who might want to move out of some of the hardest-hit regions toward areas where the economy is doing better.

These types of factors can make a bad labor market much worse.

Some of the labor market distortions will stop getting worse over the next couple of months, as housing prices stabilize and the federal minimum wage stays put, but that does not mean that labor market problems will reverse themselves.

According to my measures, labor market distortions have been getting worse at the same rate over the past couple of months as they have throughout the overall recession. Moreover, Congress appears poised to further erode incentives to earn income as an accidental byproduct of its plans reforming health care. Nor do consumers seem to be spending in anticipation of a grand employment recovery.

Thus, my humble prediction for the next several months is that real incomes and spending will continue to grow, although likely at an annual pace less than the 3.5 percent estimated a couple of weeks ago. In other words, as many have feared, this part of the recovery will be “jobless,” in the sense that employment and hours will not rise significantly, and may continue to fall.

Tuesday, November 10, 2009

Is "Fresh Water Macro" Off Track?

[Economists' Voice invited me to write this, and indicated that they liked it, but nonetheless it has languished in the editorial process, so the article premiers here. The Economists' Voice version will have the footnotes. For further, but still not exhaustive, recitation of economics errors in Professor Krugman's Sunday Magazine article, see here.]

Should macroeconomists begin again, particularly those at Chicago, Minnesota, Rochester and other freshwater schools? These days, commentators tell us that we should scrap all that we hold dear – neoclassical growth models, asset pricing models, and the efficient market hypothesis alike.

And not just run-of-the-mill journalists. No less than the Nobel Laureate Paul Krugman argued this September in the New York Sunday Magazine that we are “mistaking beauty for truth,” dismissing “the Keynesian vision of what recessions are all about,” falling “in love with the vision of perfect markets,” and blaming entire recessions on laziness.

Krugman and others are getting carried away. Allow me to defend neoclassical growth models, by providing some examples of the application of these models to the current recession, and to previous recessions. The reader can then evaluate whether Krugman’s accusations are at all accurate.

The Neoclassical Growth Model
The neoclassical growth model is an aggregate model with two basic tradeoffs: (1) current versus future and (2) market versus non-market allocations of labor. Resources are allocated over time via decisions to accumulate a homogeneous capital good, rather than consuming in the current period. People allocate their time between the market and non-market sectors via employment and hours decisions.

The model has a few equilibrium conditions. Three conditions denoted (Y), (L), and (K) relate to current consumption and work: (Y) output is produced according to capital and labor inputs, (L) the supply of labor equals its demand, and (K) the supply of capital (consumption foregone) equals its demand. The remaining two conditions are versions of (Y) and (L) for the future period.

Stated this way, the model seems to be based on the assumption that markets always clear. But twenty years of applying the model has not exactly been a love affair with perfect markets. My practice and others is to include a residual in each of the conditions: a “productivity shock” in condition (Y), a “labor market distortion” in condition (L), and an “investment” or “capital market distortion” in condition (K), which means that I expect there may be significant market imperfections or other unpredictabilities. The not-so-subtle truth is that we often suspect that markets are not functioning efficiently: one of my papers on the topic has the title “A Century of Labor-Leisure Distortions”.

Three Diagnostics
In its most basic form, the neoclassical growth model has neither money nor fiscal policy. Nevertheless, it provides some diagnostics as to how public policy variables might be affecting the private sector.

In this approach, the first step uses the macroeconomic data to suggest which of the conditions – (Y) or (L) or (K) – has the most variable residual. Much like microeconomists ask “was it supply or demand?”, as Lawrence Katz and Kevin Murphy have done with changes in relative wages, we users of the neoclassical growth model ask “Was it productivity? Labor supply? Labor demand? Capital supply? Or Capital demand?” We doubt that the complexity of the larger economy will ever be understood without some means of compartmentalizing the various behaviors, and the three “equilibrium conditions” are our means of doing so.

While a variety of tools would be appropriate for understanding the roles of monetary and fiscal policy, the neoclassical growth model’s decomposition offers some suggestions as to which approaches might help the most. For example, we might think differently about monetary policy if it depressed the labor market by inadvertently raising real wages, rather than depressing capital accumulation by adding frictions to capital markets.

Not All Recessions are the Same
Well before the current recession began, this approach led to the conclusion that recessions have various causes, and therefore that no one government policy could fix all recessions, or be blamed for all of them.

I have long been of the opinion that the labor supply residual, rather than productivity or investment shocks, was the most important of the three residuals in the Great Depression. Despite the current recession’s capital market theatrics, it again seems that much of the action is with the labor supply residual.

For both 1929-33 and 2008-9, labor supply residuals seem key because employment was low while total factor productivity and real pre-tax wages were high (or, in 1929-33, at least not commensurately low): my story, then, is not so different from the business cycle described by General-Theory-Keynes himself.

In this regard, results like mine, and those in recent papers by Lee Ohanian, Robert Shimer, and Robert Hall are quite consistent with “the Keynesian vision of what recessions are all about:” something made real wages high and employment low. But long ago we recognized that many other recessions cannot be characterized that way: real wages and employment frequently cycle together as Mark Bils has found. In these other cases, the “productivity shock” – the shock emphasized in the seminal work of Fin Kydland and Edward Prescott – seems to be pretty important. There was a good reason why old-time Keynesian models fell into disrepute soon after the 1970s stagflation.

Examination of Incentives
Given the recent time series for real wages and productivity, I doubt many of us are looking for an adverse productivity shock. But we do ask how individual incentives might be consistent with those patterns. It’s this type of reasoning that led Lee Ohanian to blame some of the Great Depression on Hoover’s industrial policy.

When it came to this recession, the neoclassical decomposition quickly led me to look further at public policies – absent from some of the other recessions – that might have caused the supply of labor to shift relative to its demand. Like others, I noticed that the federal minimum wage was hiked three consecutive times. I also turned up a major policy (the Treasury and FDIC plans for modifying mortgages) that creates marginal income tax rates in excess of 100 percent. Much research remains to be done, and undoubtedly other users of the neoclassical growth model will make convincing cases for the roles of monetary and other factors.

Paul Krugman’s scorn is all we have to suggest that marginal tax rates in excess of 100 percent are not worthy of attention, and that today’s low employment is not even partly a consequence of public policy. But, regardless of how economists ultimately interpret today’s recession, it will be notable for the basic fact that total factor productivity advanced while employment fell, and for the initial reception suffered by the basic facts in a politicized marketplace for ideas.


Barro, Robert J. and Robert G. King. “Time Separable Preferences and Intertemporal Substitution Models of Business Cycles.” Quarterly Journal of Economics. 99(4), November 1984: 817-39.

Bils, Mark. “Real Wages over the Business Cycle; Evidence from Panel Data.” Journal of Political Economy. 93(4), August 1985: 666-89.

Chari, V. V., Patrick J. Kehoe, and Ellen R. McGrattan. “Business Cycle Accounting.” Econometrica. 75(3), April 2007: 781-836.

Cole, Harold L. and Lee E. Ohanian. “The Great Depression in the United States from a Neoclassical Perspective.” Federal Reserve Bank of Minneapolis Quarterly Review. 23(1), Winter 1999: 2-24.

Cole, Harold L. and Lee E. Ohanian. “New Deal Policies and the Persistence of the Great Depression: A General Equilibrium Analysis.” Journal of Political Economy. 112(4), August 2004: 779-816.

Gali, Jordi, Mark Gertler, and J. David Lopez-Salido. “Markups, Gaps, and the Welfare Costs of Business Fluctuations.” Review of Economics and Statistics. 89, February 2007: 44-59.

Hall, Robert E. “Macroeconomic Fluctuations and the Allocation of Time.” Journal of Labor Economics. 15(1), Part 2 January 1997: S223-50.

Hall, Robert E. “Reconciling Cyclical Movements in the Marginal Value of Time and the Marginal Product of Labor.” Journal of Political Economy. 117(2), April 2009: 281-323.

Katz, Lawrence F. and Kevin M. Murphy. “Changes in Relative Wages, 1963-1987: Supply and Demand Factors.” Quarterly Journal of Economics. 107(1), February 1992: 35-78.

Kehoe, Timothy J. and Edward C. Prescott. Great Depressions of the Twentieth Century. Minneapolis, MN: Federal Reserve Bank of Minneapolis, 2007.

Keynes, John Maynard. The General Theory of Employment, Interest, and Money. London: Macmillan, 1936. (Diagnosing at p. 17 the 1929-33 period in a way similar to my own diagnosis)

Kydland, Finn and Edward C. Prescott. “Time to Build and Aggregate Fluctuations.” Econometrica. 50(6), November 1982: 1345-70.

Mulligan, Casey B. “A Century of Labor-Leisure Distortions.” NBER working paper no. 8774, February 2002.

Mulligan, Casey B. “Public Policies as Specification Errors.” Review of Economic Dynamics. 8(4), October 2005: 902-926.

Mulligan, Casey B. “A Depressing Scenario: Mortgage Debt Becomes Unemployment Insurance.” NBER working paper no. 14514, November 2008.

Mulligan, Casey B. “What Caused the Recession of 2008? Hints from Labor Productivity.” NBER working paper no. 14729, February 2009a.

Mulligan, Casey B. “Means-tested Mortgage Modification: Homes Saved or Income Destroyed.” NBER working paper no. 15821, August 2009b.
Ohanian, Lee E. “What – or Who – Start the Great Depression?” forthcoming, Journal of Economic Theory. 2009.

Parkin, Michael. “A Method for Determining Whether Parameters in Aggregative Models are Structural.” in Karl Brunner and Bennett T. McCallum, eds. Money, Cycles, and Exchange Rates: Essays in Honor of Allan H. Meltzer. Carnegie-Rochester Conference Series on Public Policy, 29, Autumn 1988: 215-52.

Prescott, Edward C. “Some Observations on the Great Depression.” Federal Reserve Bank of Minneapolis Quarterly Review. 23(1), Winter 1999: 25-31.

Shimer, Robert. Labor Markets and Business Cycles. Forthcoming, Princeton University Press, 2009.

Friday, November 6, 2009

Where's the Spending Disaster? Or the Income Disaster?

Lehman failed in September 2008, and that started the panic that got the world's attention.

So a year later, in September 2009, after living through a year of "disaster," how did real consumption expenditure (one economists' favorite measures of living standards) compare to what it was in September 2008?

What about real disposable personal incomes: the amount of income households have on hand to spend?

Both of these are HIGHER in September 2009 than they were a year earlier.

Of course, we cannot say the same thing about employment, but nobody seems to acknowledge that this recession is much more about the labor market than about drops in real incomes or spending. [prediction: leftie bloggers will ignore this sentence (and 100 other posts I've had on the labor market), and have you believe that I claim that employment also has no perceptible decline]

[The BEA may revise the September 2009 income and spending numbers, up or down, so it is possible that the revisions show real income and or spending to be slightly lower 9-09 than 9-08. Undoubtably one could find other tweaks to the series to change a slight increase into a slight decline, or vice versa. But the fact that the BEA's measurement updates and other tweaks are first order considerations when characterizing the changes is proof itself that no spending or income collapse occurred since Lehman failed. A collapse should be obvious even when viewed with blurry glasses!]

Household Survey Shows Labor Market Continuing Swiftly Down

From Aug to Oct 2009, the Household Survey shows employment falling 1.4 million (sic).

This is one of the largest two month drops during this recession. That measure of employment hasn't been this low in six years (back when the population was significantly less).

Why Employment is Continuing Down

I mentioned earlier this week that, despite the fact that GDP and consumer spending turned around months ago, employment would probably continue down. Today it was reported that October payroll employment was almost 200,000 less than it was in September.

For those who think that employment is just a reaction to consumer spending, that may come as a surprise. But, as I suggested on Wednesday, the "paradox of thrift" has been a symptom, not a cause, of this recessinon.

Your Government Is Selling Puts, Putting Your Money at Risk

The Congressional Oversight Panel has reported that the Treasury Department leveraged limited bailout money to insure assets worth many times more. These guarantees could have, and implicit guarantees may still, cost the taxpayers trillions.

Thursday, November 5, 2009

Today's Productivity Data Fed into Distortion Model

Today I was able to use the BLS productivity release to update of Figure 7 from my NBER working paper.

Of particular interest is the fact that the labor market distortion shows no signs of getting better in Q3. Until that happens, employment could continue down while real GDP rises.

Wednesday, November 4, 2009

The Recession and the ‘Paradox of Thrift’

Copyright, The New York Times Company
One bit of conventional wisdom about this recession is that it was caused, or at least significantly worsened, by a “ paradox of thrift”: Consumers suddenly ceased to be willing or able to spend like they once did. An alternative interpretation puts the labor market at ground zero, and sees the spending decline merely as a reaction to the labor market.

Was spending or the labor market the fundamental driver in this recession? Causality is always hard to disentangle, but one fact is worth noting: Consumer spending fell much less than did the labor market.

The chart below displays inflation-adjusted private consumer spending, alongside hours worked in the labor market (the product of employment and hours worked per employee), for each month of this recession.

Each series is displayed as an index, with its value at the start of the recession (December 2007) set to 100. Consumer spending normally trends up more than employment does, so I have adjusted for that by removing prior trends from consumer spending and work hours.

For example, a value of 95 for real consumer spending in September 2009 means that inflation-adjusted consumer spending in September 2009 was 5 percent below what it would have been had it continued its previous trend since December 2007.

Consumer spending fell significantly during the months of 2008, a drop that sorely hurt manufacturing and other industries. But labor fell at least as much for the first nine months of 2008, and fell a lot more since then.

Through September 2009, work hours were 11 percent below trend, while consumer spending was “only” 5 percent below trend.

The economic news first started getting widespread attention in September 2009 when Lehman failed and credit markets froze. Since that time, consumer spending only fell only two percentage points further below its trend while labor fell another eight.

While it is conceivable that a few percentage points’ decline in consumption could cause a many-fold reduction in work hours, it seems more likely that the reduced consumer spending was mainly a reaction to layoffs and hours cuts. The roots of this recession go a lot deeper than the paradox of thrift.

Tuesday, November 3, 2009

Residential and NonResidential Construction Through September

Residential construction spending was higher in September than in August, which was itself higher than in July, which was itself higher than in June.

Interestingly, I predicted last December that the housing market would turn around in the summer of 2009. By now, we have seen enough housing price and construction data to see that prediction was correct.

My prediction was based on the assumption that a credit crunch would not be an housing important factor. In other words, I expected any significant credit market restraint of the housing market to cause the market to turn around later than summer 2009. That's not to say that credit is flowing easily -- just that credit conditions are reacting to the housing market (in particular, its new and lower levels of price and construction) rather than the other way around.

But look at nonresidential construction: it's fallen significantly since spring. Why?
  • Credit crunch?
  • housing construction crowding out non-residential construction?
  • an expectation that labor will remain low for a while to come.

I am skeptical that a credit crunch is all that important -- certainly not the entire story (if it were, why did the housing market turn around as quickly as I predicted?). Housing construction right is now not significant enough to crowd out much non-residential construction. Thus, the bad news is that non-residential investment spending may be indicating more tough times for the labor market. I'm going to write more about that tomorrow on

Monday, November 2, 2009

Update on Labor Market Decomposition

Here is an update of Figure 7 from my NBER working paper.

Of particular interest is the fact that the labor market distortion shows no signs of getting better in Q3. Until that happens, employment could continue down while real GDP rises.

More than 100,000 Mortgage Modifications per Month

I found this graph on the Huffington Post.

It appears to be for Home Affordable Modification Program, and therefore omits modifications under Home Affordable Refinance Program and under the FDIC's Streamlined Modification.