The Brookings Institution Puts the Brookings Papers on Economic Activity Online for Free
Daniel Davies: the "-ist" in "Economist" Is Like the "-ist" in "Trotskyist," Not Like the "-ist" in "Scientist"

Lecture: Seven Sects of Macroeconomic Error, Part II

Lecture: Seven Sects of Macroeconomic Error, Part II

J. Bradford DeLong
U.C. Berkeley
March 2011

20110315 ias 107 ppt

Picking up from last time. Let me run through what these schools are, why people believe in them, and why you should not:

& & &

Zero Marginal Product Workers

20110315 ias 107 ppt 1

The first wrong model is the idea that we have high unemployment now because our educational system has failed. We in America have produced 12 million workers who effectively have no useful and productive skills that make it worth anybody's while to pay them to do anything. These 12 million people have a zero marginal product--or at least a marginal product that is less than the minimum wage. Representative thinkers who advocate or have advocated a version of this model include Niall Ferguson and Tyler Cowen.

Tyler Cowen of George Mason, especially, likes to talk about "unemployment" among horses in the early twentieth century. Back in the late 19th Century America had roughly one horse per person. Horses were extremely useful: you could ride them, they could pull things, you could put them on a treadmill and make them power things--with steam engines and windmills as the only other non-human-muscle power sources, we had an awful lot of horses.

20110315 ias 107 ppt 2

Then we developed alternative technologies: diesel engines, gasoline engines, electrical engines linked to power plants via high-voltage transmission lines--and all of a sudden the marginal economic product of most of the horses who had existed in 1900 was zero. Horses became unemployed. Horse-breeding operations shut down.

Why can’t we conclude that the same isn’t true of the 12 million excess American workers who are now unemployed? The coming of computers and the decline in manufacturing today have brought the same reduction in the marginal economic product of low-skilled human workers that the coming of electrical, gasoline, and diesel engines brought to the marginal economic product of horses a century ago.

The first rebuttal is: Although these extra 12 million surplus workers are unemployed now, they were employed back in 2007. Are we supposed to believe that changes in technology have proceeded so rapidly that the marginal economic product of all these 12 million workers crashed in two short years?

The counter is that their marginal economic products had been zero for a long time--but that we had not realized it. Because of irrational exuberance, we thought that they could be productively employed in housing construction because we all overestimated the long-run value of the houses that they were building. Now, however, we have discovered that houses are not terribly valuable things--especially not houses in the swamps of Florida and in the desert between Los Angeles and Albuquerque. We used to think they had a high marginal product working in construction. Now we recognize that they do not, and that they never had. And given that the only industry in which they might have been productive enough to earn their keep was construction, now we recognize that the American economy has a 12 million oversupply of excess zero marginal product workers . The big empirical problem with this theory is that employment on construction payrolls in the United States has only fallen by 2 million since 2006. And of this 2 million decline in payrolls perhaps 500,000 are people who have gone back to Mexico. Of the 12 million excess unemployed currently here in the United States, only something like 1.5 million can be attributed to the decline in construction employment.

Where did the other 10 1/2 million come from?

Even if we do have 1 1/2 million permanently-unemployed zero-marginal-product workers as a result of the decline in the construction sector, why have they carried an extra 10 1/2 million other workers into unemployment as well when the other 10 1/2 million seemed to have perfectly good non-construction skills doing what they were doing in 2007? Why has each construction job that has vanished relative to trend carried another seven jobs along with it? The zero-marginal-product workers story has absolutely no answers to these questions.

This is the decisive argument against the idea that we now have high unemployment because we have 12 million workers who do not have the skills and the mental attitude and the willingness to show up that make it worth anyone’s while to pay them to do anything.

& & &

The Structural-Unemployment Labor-Shortage Economy

20110315 ias 107 ppt 3

A second wrong model is that of Minneapolis Federal Reserve Bank President Narayana Kocherlakota, who argues not that we have 12 million permanently unemployable excess workers but rather that we have 12 million excess unemployed whom it will take a long time to retrain and find other jobs. The idea is that because of structural shifts in the economy we have workers who had the skills for declining industries but do not have the skills for the currently-expanding industries, and that it will take a long time for them to acquire the skills that they need to fit the labor requirements of the new economy.

The decisive rebuttal to this is that if the problem is indeed on the supply side rather than the demand side--if the problem is a lack of workers with the right skills for expanding industries rather than a lack of demand and so a lack of expanding industries in the first place--we should see signs of expanding industries and labor shortages in the data. We should see a lot of vacancies in the economy, as the expanding industries that want to hire workers but can’t find any who are qualified search for workers to fill the holes in their staffing patterns. We should see substantial wage increases--even if not in the economy as a whole, we should see them in the expanding industries as firms in industries that are short of labor try to take qualified workers away from one another by promising higher wages. We should see these things, but where are they?

Job openings are a good 600,000 above their absolute minimum mid 2009 trough. But job openings are still far below their normal non-recession levels. If the structural unemployment theory is correct, they should be above normal non-recession levels. And you simply cannot say that there is upward pressure on the peoples’ wages: four years ago your average wage increased by 4% in dollar terms in a year; now nominal wages are rising at a rate of 2% per year in dollar terms and economy-wide wage increases are kissing zero in real terms.

This is not a labor shortage economy.

& & &

Overaccumulation of Capital

The third wrong model is a perennial: that our problem is the overaccumulation of capital.

This is the oldest alternative theory of all. This was a favorite of Chicago economist Frederick Hayek. It was the theory on which Andrew Mellon and Herbert Hoover ran the U.S. economy during Hoover's presidential term at the start of the Great Depression. This theory dates all the way back to Karl Marx--who talked about how economists like James John Stuart Mill and the politicians they advised like Robert Peel thought that you could deal with a financial crisis and a depression by simple strategic interventions in financial markets. This was, Marx said, an attempt to play a game of economic three-card-monte with the economy--to make the economy live beyond its means on its wits. And, Marx said--and Mellon, Hoover, Hayek, and their present-day epigones fellow them--this cannot work. Increasing the quantity of liquid or safe or savings-vehicle assets in the economy cannot permanently repair aggregate demand because, Marx and company said, the real problem is not a shortage of aggregate demand but rather a surplus of aggregate supply: the economy has made too much capital for the market system to absorb, and needs a "prolonged liquidation" of high unemployment, bankruptcies, and scrapped factories before the process of capitalist economic growth can resume.

I confess it is not clear to me why the Marx-Hayek-Mellon-Hoover axis is so certain that there is no financial sector cure to the problem of high unemployment. But let me try to outline Marx's version of the argument, because it is the one that seems to me to make the most sense--or, at least, the least nonsense.

Capitalists, Marx said, own the capital and hire the workers to produce output which they then sell. Output comes in two forms--consumption goods which they sell to workers, and capital goods which they sell to capitalists. Capitalist economies expand over time: each year the previous year's production of capital goods is added to the economy's capital stock and production becomes more capital-intensive and, hence, larger.

Demand for consumption goods, Marx said, grows only slowly: workers are poorly paid, as the economy's capital stock grows capital is substituted for labor and that puts downward pressure on wages, and so forth. Workers cannot increase the amount of money they spend on consumption goods because they don't have the money and are not earning much more money over time. So as production grows a greater proportion of production must be sold to the capitalists--who buy it because they expect to be able to make bigger profits next year by producing on an even larger scale than they produced this year.

Say that we are in Britain in 1860, with nominal net domestic product at £600 million a year and a capital stock of £1.5 billlion, with £500 million of consumption goods produced and £100 billion of net capital goods produced to add to the capital stock. And let's say that the production function and technological progress are such that production is always 40% of the capital stock, Then next year, in 1851, you will have a capital stock of £1.6 billion, production of £640 million, and of that £500 million will be consumption goods and £140 million will be capital goods--which capitalists must buy because they expect to be able to use it all to make further profits in subsequent years.

Extend this economy's trajectory out a decade, to 1860...

Microsoft Excel 75

You see that by 1855 the economy's productive capacity will have grown significantly--instead of £600 million of net domestic product we will have £1.04 billion. But with consumption still at £500 million, net purchases of capital goods by capitalist must be not the £100 million they were in 1850 but rather £540 million: capitalists must be willing to purchase not £100 million in the expectation that if they add this amount to their capital stock they will be able to sell all the extra output they can produce but they must be willing to purchase £540 million and plan to sell all the production. By 1860 the level of output will be not £500 million but £3.39 billion--and capitalists must be so exuberant that they expect to sell not £100 million to their fellows but rather £2.89 billion.

Now this, Marx says, is unsustainable. Capitalists may well be willing to add 7% to their capital stocks each year to service expanding markets. But will they be willing to spend their money expanding their capital stocks by 34% in a year in the expectation that there will be demand for the full output of their factories next year because next year capital stocks will grow by 36%? No. Sooner or later capitalists will recognize that full demand for factories next year is a chancy thing. So they will cut back on their plans for expansion this year--and when they do so everybody will find that it is this year and not next year that demand falls short of supply.

The workers are unable to buy the "surplus" output because they have no extra money. The capitalists are unable to buy the "surplus" output because there is no point in expanding their capital stocks when they cannot even find markets for all they can produce this year. The result is crisis: overproduction, mass unemployment, and bankruptcy. The weakest firms fail. Their capital is scrapped. Other factories stand idle. Their capital rusts away. The depression continues until the waste and scrapping of capital has proceeded so far that capitalists once again start to believe that it is time to invest in building up capacity once again--and when they do so decide, they find that there is not a surplus but a shortage of capacity and so the cycle of expansion, exuberance, investment, irrational exuberance, overproduction, crisis, and bankruptcy and unemployment starts all over again.

The root problem, Marx thought, is that a market economy can only run at full employment if the surplus value received by employers is then ploughed back into investing to increase their capital stock. But if the surplus value is invested in increasing the capital stock--well, then, who are you going to sell the next round of increased production to because there will be even more surplus value in the future?

Lowering interest rates via financial manipulation, Marx thought, would not help because artificially propping up capitalist demand for investment goods for a year or two simply magnifies the overproduction of capital and magnifies the eventual crash.

Public works financed by taxes on the rich, Marx thought, would not help because the public works would have to be paid for and taxes raised from the capitalists would diminish their confidence and their willingness to invest in further expansion.

Public works financed by taxes on the poor, Marx thought, would not help because the public works would have to be paid for and you can't get blood from a stone.

The only way out is through universal bankruptcy: "prolonged liquidation." That is the only way out, Marx thought--unless you resort to socialism: the abolition of the capitalist class, the overthrow of the market economy and the wages system, public ownership of the means of production, and the creation of a free and democratic society of associated producers.

Mellon, Hoover, Hayek and their epigones have not been so keen on the "socialism", the "abolition of the capitalist class", and the "free society of associated producers" parts of Marx's argument, but otherwise they buy it: irrational exuberance leads to overaccumulation which has to be cured by a "prolonged liquidation" requiring persistent high unemployment. Mellon, Hayek, and Hoover tended to say that this is an unfortunate drawback of what is otherwise a pretty good economic system.

There always seemed to me to be two big questions here.

The first is Paul Krugman’s question. Capitalists don't have to spend all their money buying capital goods. They are, historical experience teaches us, perfectly happy buying consumption goods as well. The only underlying point that is even potentially valid point is not that full employment requires an unsustainable explosion of capital good production but rather that sometimes the economy gets itself wedged into a configuration in which capital goods production is too high for the level of total demand. The result would be unemployment among workers in capital goods industries--and when they get unemployed they lose their jobs and when they lose their jobs their incomes fall and you get a recession.

Krugman's question is this: Sometimes the economy does indeed get too much capital for the sustainable level of demand. But equally there are times when the economy has too much consumption goods for the sustainable level of demand as well--some times there is not an overaccumulation of capital but instead an overproduction of consumption goods. When that happens the people who make consumption goods do lose their jobs and have to go find jobs someplace else. But that process does not produce a recession: it produces a boom. No period of high unemployment is generated by the redeployment of workers from consumption goods to capital goods industries. High unemployment arises only when workers are switching out of capital goods and into consumption goods industries.

This is a decisive argument against the "overaccumulation of capital" theory in its Hoover-Hayek-Marx form. The logic of the argument requires that overproduction of capital and overproduction of consumption goods have similar effects on unemployment if the argument is true, and they do not.

20110315 ias 107 ppt 4

Second, and more important, where is the overaccumulation of capital right now? Look at private construction spending in the United States. It grows smoothly throughout the 1980s and the 1990s, We get this housing bubble in the 2000s. And since 2007 housing and other construction has been depressed below its trend level.

No matter how you draw the housing-bubble triangle and the housing-depression trapezoid, the trapezoid is larger than the triangle. The housing market has spent more years depressed by more than it’s been elevated back during the boom. We don’t have a surplus of houses relative to trend right now. What we have is a shortage.

Admittedly, we do not have houses we would really want. What we would really want to have would be more two-bedroom condos in Venice Beach rather than five bedroom houses with swimming pools beyond San Bernardino. But the fact that a bunch of our houses are in the wrong place means that we should be more eager to build houses right now, not less willing.

The idea that we have an overaccumulation of capital right now is simply wrong.

What we do have is we have a high housing vacancy rate because a lot of people have lost their jobs and so are doubling up in their apartments with their relatives.

So at the empirical level the "overaccumulation of capital" theory simply does not fly. You might--I doubt it but you might--have been able to use it to justify a recession in 2008 and 2009, but it will not explain high unemployment today and even less high unemployment next year.

& & &


20110315 ias 107 ppt 5

The fourth wrong model--I don't hear this from economists as much as from Republican politicians, and maybe from a very few economists who want jobs in the next Republican administration too much, and from Alan Greenspan--is that there been an enormous increase in uncertainties since the election of Obama. This enormous increase in uncertainty is due to government deficits and overregulation. It has led businesses cut back on their spending. We cannot get spending up again until we do something to make businesses less uncertain about the future.

Here I would say that there are five considerations:

The first is that this is a version of the correct, aggregate demand argument--only instead of placing the deficient demand for goods and services and the excess demand for financial assets on the shock caused by the collapse of the subprime bubble it places it on the shock caused by Obama's election. Thus the cure is the same as in my standard story, for the cure does not depend on and is the same no matter what the cause of the flight to quality: rebalance financial markets to leave investors satisfied with their holdings of safe and liquid assets and they will invest in businesses that want to expand. Restoring business confidence is one way to rebalance financial markets by diminishing the demand for safe and liquid assets, but it is not the only way: rebalancing financial markets by expanding the supply works as well.

Second, this theory should never even get out of the gate. The timing is simply wrong. It was not Obama's election or expectations of Obama's election that caused the flight to quality. The flight to quality was caused by the collapse of the subprime bubble.

Third, if businesses are uncertain and are worried about overregulation and overtaxation, shouldn’t they be saying that that is what worries them? Ask businesses what they are worried about right now, and they will tell you that they are unusually worried about poor sales and not about Obama-generated uncertainty. This theory says that it is worries about future government policies that have hit the minds of business entrepreneurs that are the cause of the recession. If it is true, that is what business entrepreneurs should say that they are worried about when you ask them. And they are not.

Fourth, if businesses are genuinely worried about inflation due to government deficits, shouldn’t we see this in financial markets?

Shouldn’t the price in financial markets for insuring yourself against inflation over the next 10 years be high right now? Right now it is a break-even trade at an inflation rate of less than 3% per year. There are no signs that businesses are worried about inflation and are taking steps to insure themselves against it. And as to high interest rates in general because of a fear that government deficits will starve private businesses of capital--Treasury interest rates continue at their extraordinarily low levels. This, too, simply fails to meet the plausibility test: if people were uncertain, and if that was causing the recession, we wouldsee signs of uncertainty in financial markets and surveys. We do not.

It is entirely possible that uncertainty about future policy can generate a recession.

But uncertainty is not generating this one.

& & &

Inflation Around the Corner

20110315 ias 107 ppt 6

Fifth, closely linked to the argument that "uncertainty" is the cause of the recession is the argument that we definitely do not need stimulative policies right now--an argument I hear most these days from John Cochrane of the University of Chicago. Further stimulative policies will set off a burst of inflation that will harm the economy in the long run and it’s causing that further stimulative policies will raise interest rates and crowd out private investment and slow long term government growth. On the internet this morning, once again, who was it, somebody had a rant about how worries that Japan, Japanese construction after the earthquake, about how that will push up interest rates and so diminish private investment and harm the global economy as the Japanese government borrows and spends a whole bunch of money to deal with the crisis caused by the earthquake and the tsunami. A rant about how that just doesn’t apply because what we saw in the aftermath of the tsunami and the earthquake was not an increase but a decrease in the interest rates on Japanese and American government bonds, that people are more willing to buy them, not less. That we’re not crowding out government investment that we’re crowding it in.

If there were a genuine need to control inflation well you’ll expect to see wages increase or you’d expect to see the inflation rate implicit in government bond prices, this red line rising rapidly and no, we’d expect to see this blue line which is the real interest rate rising rapidly and it’s not. That we see a situation in which an economy is approaching an inflation danger looks like and once again it doesn’t look like this.

& & &

Banking and Fiscal Policy Unnecessary

20110315 ias 107 ppt 7

The sixth wrong model is advocated by those who believe that we do not need to even think about fiscal and banking policy because Milton Friedman was right all along: simply stimulative monetary policy through normal open market operations can do the job. Here we have Nobel Prize winner Robert Lucas, who simply claims to not see the point of either banking policy to lower risk premia when the short-term safe nominal interest rate is already at zero and cannot go any lower or the point of expansionary fiscal policy. Lucas annoys me because he has spent some time trashing Berkeley’s own Christina Romer in an unfair, ignorant, and incoherent way. So I want to stress that when you actually listen to what Lucas is saying--as, for example, at the March 2009 conference of the Council on Foreign Relations--the immediate first impression is that he simply has not thought any of the issues through. Take Robert Lucas back to 1829, put him in a room with John Stuart Mill and Jean-Baptiste Say, have Lucas give his argument for why it was that increased government purchases or that bank rescues would not help an economy in financial crisis--if we were to do that I think that both Mill and Say would classify him as an unarmed man in a battle of wits--they would ask: "this is supposed to be a macroeconomist?"

Back in 1829 economists had got it straight what the relationship was between the soundness of the banking sector and the level of overall employment. How come Bob Lucas in 2009 has not?

& & &

Invariance "Results"

20110315 ias 107 ppt 8

The seventh wrong model is a line of reasoning that you see from finance economists--a set of claims that a whole bunch of policies have no effect on anything.

For example, consider Eugene Fama of Chicago, and his claim that fiscal policy has to be ineffective because whenever government spends it has to borrow, and whenever it borrows it has to take money away from a single private individual who is also spending. Thus whatever the government does to plan to spend money has by definition to be offset by an equal and opposite decrease on what the private sector plans to spend on goods and services.

That is simply wrong. When planned government spending goes up, planned private sector spending does not have to go down. When private sector income and wealth goes down, it does not have to plan to cut its spending if it is willing to hold fewer financial assets. And one way that increased government spending boosts the economy is that government debt issue boosts the supply of safe and liquid financial assets--and so makes people more comfortable holding the financial assets that they have and less anxious to cut back on their current spending on goods and services.

Consider Myron Scholes of Stanford, who criticizes Federal Reserve Chair Ben Bernanke for believing that the Federal Reserve can buy long-term bonds for cash and so affect long-term interest rates. Scholes's argument appears to be that when the Federal Reserve buys risky bonds there is a chance that the Federal Reserve will go bankrupt and then that taxpayers will be taxed in order to pay for the losses on those risky bonds, and so you actually have no removed any risk from investors via the Federal Reserve's purchases. That is simply wrong. Taxpayers are a very different group of people from investors. Investors are rich and old. Taxpayers are middle class and much younger.

When we shift risk off of investors who fear that they are already bearing too much risk and on to taxpayers is a very plausible way to change the economy-wide price of risk and thus to affect interest rates.

& & &

The Seven Sects of Macroeconomic Error

All seven of these lines of argument I have outlined seem to me to be completely, obviously, and trivially fallacious. There are simply not enough unemployed construction workers to account for the recession. There are no signs of the high vacancies you would have in a high-structural-unemployment economy. The overaccumulation of capital simply is not there because the overbuilding triangle is smaller than the post-crisis under-building trapezoid. If the recession were due to uncertainty over overregulation and high future taxes then people would be worried about overregulation and high future taxes rather than being worried about poor sales and slack demand. The need to control inflation in order to avoid a double dip as people get scared of inflation requires that people be actually scared of inflation--which means that asset market signals should show that inflation is expected to rise. Conventional monetary policy is not the only appropriate policy tool because--as the classical economists knew back in 1829--banks are special, and liquidity, confidence, and the solvency of the banking sector are closely linked. If you don't understand things about the centrality of the financial sector that other economists have known for 180 years, why are you on the stage talking rather than in the audience listening? If you have not figured out that planned economy-wide spending goes up when people plan to dump and down when people plan to accumulate financial assets and thus that plans for the government to raise do not immediately entail plans for the private sector to equally cut spending, why are you in the room at all? And if you have not noticed that investors are a different group than tax payers and that shifting the burden of risk off of overburdened investors on the taxpayers may well affect the risk premium--well, you simply have not thought about the issues at all.

And that’s where I want to conclude our depression economics section.

I want you to hold tight to the idea that we could push our current unemployment rate down relatively rapidly if only we had the political will to undertake the stimulative fiscal and monetary policies to do so.

Of course, that might generate inflation. And next lecture we will start on inflation economics: where inflation comes from, how in proceeds, and how to control it.