Crowding Out Brad DeLong

by Mario Rizzo  

Brad DeLong thinks that, under present circumstances, the crowding out of private expenditure by fiscal stimulus is not a live issue. The basic argument is that since neither average wages nor interest rates have risen in response to stimulus, no resources are being diverted from private to public uses. 

I am unsure what the standards of good analysis are among Keynesian macroeconomists, so I proceed with some trepidation. However, as readers of this blog will know, I am unhappy with the level of macro-aggregation usually practiced by both Keynesian and new-classical macroeconomists. So I want to disaggregate the analysis a bit.  

In the first place, expenditure is not simply expenditure in general. It is expenditure on specific things and in specific sectors of the economy. These lines of expenditure use resources appropriate to producing the output in question. They involve input combinations from many areas of the economy – not only labor, but other resource complexes including those that will have other uses. To the extent that these resources are non-specific to a specific line of production they may come from areas in which they were not formerly unemployed. 

Second, unemployment of resources, including labor, is not always pure idleness. We are living in conditions of real uncertainty. A bubble has burst, the domestic auto industry faces uncertain prospects, tax rates are on the way up – how far and in what respects in anyone’s guess – we have just faced a possible healthcare transformation with its unique costs and taxes, European debt problems are becoming manifest, and more. 

Now this is clearly not merely a problem of aggregate-demand confidence. The private spending the government hopes to stimulate must have a direction. Firms must worry about the sustainability of demands in specific areas and about costs in specific areas. Unemployment is a quite rational response to real sectoral uncertainty. To bring resources out of “unemployment” prematurely is a form of crowding out – crowding out of the economically-valuable activity of recalculation, recalibration or search.  

To anticipate a Keynesian response: Some will say that there is a difference between the rationality of the individual agent and the collective irrationality of agents, each of whom is waiting for the others to move (spend) first. This is by no means an inconceivable phenomenon. So let’s think about it.  

The phenomenon is not unlike situations that entrepreneurs face all of the time. Factors of production may be underpriced and undercapitalized relative to longer-run predicted demand. Markets may be underprovided with goods relative to longer-run demand. These provide profit opportunities to alert entrepreneurs. Those who move first can earn greater profits than others (and, of course, they open themselves up to losses).  

Third, it is difficult to believe that once recovery does occur that the new equilibrium rate of interest will not be considerably higher because of the new large amount of government debt that has been, and will be, accumulated. These higher than otherwise rates will ensure that investment projects in the future will, in effect, be crowded out by today’s stimulus.  

I have no doubt that economists can, in sufficient time, generate quantitative estimates to varying degrees of approximation on these forms of crowding out. But consider today’s political problem.  

The crowding-out consequences are indirect, complex, and hard to see without theoretical lenses. If stimulus creates jobs in construction, retains various state and local government jobs and so forth, these are more easily seen by voters. (However, there are still many theoretical and methodological issues involved in the counting. It is not a simple matter.) Therefore, as with most government programs there is a political bias toward the direct and immediate consequences, even where the indirect may be far more important. This is Frederic Bastiat’s story of the seen and the unseen. (It may well turn out that even with the political bias working in its favor, the voters still sense that the stimulus crowd has gone too far. So much the better.)  

But now dear readers: Do you think Steve Horwitz is a “clueless idiot”?

20 thoughts on “Crowding Out Brad DeLong

  1. I would say, as I want to be an ethics-free republican hack and a clueless idiot, too, that bottlenecks do not need to arise in labor markets, so the behavior of wages may be irrelevant. Considering that all raw materials have been on the rise for a year or so (not only oil), I would bet that bottlenecks are in the fixed capital supply necessary to expand production of these intermediate goods. I hope this doesn’t make me a republican, I would hate that. 🙂

  2. DeLong is the clueless one. He once dismissed the idea of regime uncertainty. I don’t recall his reason/epithet.
    Regarding the bond market and interest rates, he overlooked the fact that the Fed has done its level best to keep rates low. That inflated a bond market bubble.
    This might come as a shock to DeLong, but interest rates aren’t determined in the bond market. They are determined in the time market, and the bond market reflects that.
    Of course, the bond market also reflects other things, such as Fed policy.

    He needs to read some Austrian works and get out of the aggregation mindset that underscores Keynesian demand management (or whatever it is).

  3. Bill Stepp is on the mark about interest rates.

    In most macro model there is one interest rate, “the” interest rate. The mindless application of such models to policy is dangerous. The interest that firms and individuals are paying — not to mention the nonprice terms — are well above the zero interest are that macroeconomists are plugging into their models.

  4. Nominal and real wages for private industry have both risen throughout the recession. Some interest rates have also risen since the stimulus package passed, for example 10 year Treasury notes are over one percentage points.

    Personally, I find it remarkable that this debate would go on without at least looking at the data, even though it is highly aggregated.

    DeLong is the most guilty (there is no crowding out because wages and interest rates have not risen,) but you would think someone would at least check.

    I teach introductory Macro, and when I define “real wages” I get information from the web and show it to the class. I think it is useful for students to have some rough idea that average hourly wages have risen from 2.50 (or so) in 1984 to 18.50 (more or less) today. And that when you deflate them by some measure of the price level, they sometimes rise and fall. Of course, the concept of real wages is more important than knowing what they are or were.

    My view is that “fiscal stimulus” can increase real expenditure and increased real expenditure can increase real output. But I don’t doubt that there is some crowding out.

    Of course, Obama is talking about long run plans for green jobs and expanding exports. These will certainly crowd out alternative activities.

  5. “Nominal and real wages for private industry have both risen throughout the recession.”

    That would answer question, but raises a more difficult one. It explains the stubbornly high unemployment rate. But what is causing wages (wage rates?) to increase?

  6. Amateur asks, who’s wages, public or private employees, more in one sector than the other? Also, since it seems a large part of the stimulus, I know this for sure here in Tennessee, has gone to cover short falls in state budgets, here it was TennCare, not sure how there has been any measurable stimulating of our economy.

    The way the Feds are counting jobs saved/created is clearly iffy so exactly how much stimulation of the economy did the stimulus provide? And if it’s negligible (my own suspicion) how much crowding out is taking place?

  7. Scott Sumner has suggested that public employees with rigid contracts are responsible for a large part of the real wage increase. If their wages don’t drop (and they don’t lose their jobs, which is not likely since that sector has the lowest unemployment rate), those of the private sector must drop even more. Wage stickiness (perhaps of a lesser degree) exists there as well, so the result is unemployment.

  8. FRED says that up to end 2008 real wages were falling, then increasing, and now the CPI is on the rise fast enough to make them fall again in the near future. The average hourly earnings was in the private sector, so no public employees should have been counted.

    Considering that average hoyrly earnings in construction have been rising relative to the rest of the private sector (manufacturing earnings show no effect at all), which is utter nonsense as a free market dynamics, I think it’s the effect of credit policies aiding the construction sector and delaying the liquidation of excess capacity. No data on weekly working hours in construction is given, however, although working hours have been falling throughout the crisis and now are on the rise.

    PPI data shows the greatest variation: they have been rising 15% up to mid 2008, than falling 15% in mid 2009, and now rising again, faster than CPI. Profits margin may have been squeezing in the last few months because of this, and we are probably heading toward a stagflation. It’s possible that a financial crisis (maybe due to squeezed margins) turns it into a deflationary stagnation, and then that PPI will start rising again once the financial crisis is over because production constraints are still binding. Would be funny.

    However, fiscal policies, like monetary policies, have nothing to do with the structure of the economy (if ever, they distort it), but they can increase aggregate quantities, so maybe it’s the stimulus. I would disaggregate the effect on earnings in construction from the rest, maybe the rest is falling, but I don’t know.

  9. *Sigh*…

    You can argue that it is a *bad idea* to create jobs for people right now by boosting government spending–that if they stay unemployed until private enterprise recovers and hire them that they will get better and more productive jobs. I think that argument is largely wrong, but it is a coherent argument.

    However, that is not Horwitz’s argument. Horwitz’s argument is that boosting government spending doesn’t create any jobs now–that: “For every dollar that government spends, there is one less dollar being spent somewhere else in the economy.”

    And that is not an argument. That is simply malpractice. That is a claim that no even half-trained economist has believed for generations–not since Irving Fisher pointed out that government bond sales to finance deficits affect interest rates and that money demand is interest elastic, not since Knut Wicksell pointed out that the total flow of spending in an economy is such as to match savings to investment at the market rate of interest and that government deficits affect and detract from savings.

    *Read* Horwitz before you defend him. And if you still want to support him, go argue with Knut Wicksell and Irving Fisher and Milton Friedman, not me…

  10. Pietro,

    I’ll write a blog article on it soon.

    What are you looking at on Fred?

    I’m using the BLS series “average hourly earnings: total private industries” for nominal wages.

    And I deflate them by the monthly personal consumption expenditures price index both with and without food and energy.

    I see real wages rising the entire recession either way.

    It looks to me that the growth of nominal wages slowed a bit when things were really bad in the spring of 2009. And then the real wage movement is just a reflection of changes in the price level.

    I am sure that if you using a price level measure with enough volatility, you will see larger fluctuations in real wages.

    The way I see it, nominal wages moved to a slightly lower growth path, but inflation dropped more, so real wages rose.

    P.S. nominal wages were about 2.50 in 1964, which is the beginning of the average hourly wage series.

  11. DeLong:

    I read Horwitz. I thought his remarks about inflation were inadequate to a situation where nominal expenditure remains nearly 10 percent below trend. However, the reality of a permanent increase in the size of government is that the principle of scarcity and opportunity cost overwhelm the impact of monetary disequilibrium.

    The green jobs, the doubling of exports, and however else the administration will centrally plan the economy are going to come at the expensive of privately directed output and employment.

    The CBO estimates that nearly half of the loss of output below trend is due to less productive capacity, something that those concerned about malinvestment (like Horwitz) emphasize. It is difficult to see how government spending will solve that problem. (Except of course, trying to return to bubble-levels of housing production.)

    The remaining element of unemployment, that results from prices and nominal wages failing to drop to a low enough growth path so that the current level of nominal expenditures can purchase them, might be helped by government spending funded by money creation or, just maybe, a reduction in the demand to hold money satisfied by newly issued government bonds.

    However, that output gap could be filled by the production of private consumer and capital goods–matched by increased real expenditures either resulting from an expansion of the quantity of money by alternative means than new government spending, or through a decrease in prices and wages. Regardless of which is the better approach, those possible private goods and services (and employment producing them) remain the opportunity cost of government goods and services.

    Of course, unless there is a liquidity trap. It isn’t Wicksell or Fisher that you depend on. It is rather one possible scenario described by Keynes.

    And if that is what you want to say, “we are in a liquidity trap, so there is no way to raise private expenditures (real or nominal) and so government is the only way and there is no opportunity cost” then fine. But those of us who don’t believe this are hardly idiots.

  12. Fisher’s argument was incomplete; there are times when the demand for money is not interest elastic. Look no further than the demand for money market funds, which were invented in the 1970s, long after Fisher did his work. As for Wicksell, his argument is true in a free banking system, but doesn’t necessarily hold under central banking. In fact it usually won’t. See chap. 7 of Selgin’s book on free banking.
    Actually, even the latter point has to be modified. Why?
    Consider the aggregate (domestic) demand Keynesian formula Y = C + I + G.
    C and I are private demand that capitalists seek to satisfy by investing in capital used to produce the goods and services that compose C. They invest in anticipation of making a profit and avoiding a loss, which are the signals that their plans are correct or possibly flawed.
    What about G? Government spending is done by government bureaucrats and politicians, who don’t seek to make a profit. They are state socialists, to use the great libertarian Benjamin Tucker’s technical term. They don’t think like capitalists and entrepreneurs, nor do they have the same incentives and economic and financial constraints. They don’t have shareholders to answer to.
    They spend other people’s money, collected via taxes and money balances inflated away for the benefit of the State. (The latter is done almost always after outlawing private money, and imposing central banking.)
    That leads us to aggregate supply, AS, as a Keynesian-socialist professor of mine used to write on the blackboard, before he became an IMF honcho.
    In the Keynesian formula for AS, Y = C + I + T.
    That puts T on the same plane, morally and economically, as C + I. Hwever, it’s not on the same level either morally or economically. Forgetting the fact that T is theft and is therefore a criminal act, taxation takes resources from consumers and capitalists (and entrepreneurs) that would have been either spent on consumers’ goods or invested in capital goods done in anticipation of consumer demand, not the demand of politicians and bureaucrats.

    The bottom line is that the demand for government “services” is made by politicians, not consumers, unlike the demand for consumers’ goods, which is made by consumers. The supply of government “services” is made by taxpayers, unlike the supply of consumers’ goods, which is met by entrepreneurs and capitalists.
    Just as taxation subtracts from the private sector, so too does government spending. Government does two things: things that should be done by the private sector (supply money, schools, law, courts, roads, etc.), and things that wouldn’t be done at all in the private economy, such as war and mass murder, outlawing drugs, etc.

    So the correct Keynesian equations are
    Y = C + I – G (AD), and Y = C + I – T (AS).

    Rothbard argued, correctly in my view, that the way to view this is:
    Y = C + I – (G +T), where C + I = private production, and G + T = government depredation. Just call it state socialism.
    The Feb. 6-7 issue of the WSJ has a good article, “Opinions Split on Job Creation” (p. A4). It points out that a recent CBO report claims that the two biggest effects on employment have been by reducing payroll taxes for firms that hire new workers, and increasing aid to the unemployed. This is daft. Why not eliminate payroll taxes, or at least cut them permanently? You don’t have to accept the idea of rational expectations to know that temporary measures are just that–temporary. As for increasing aid to the unemployed, subsudizing unemployment (or anything else) doesn’t lead to less of it. (You can’t make this stuff up.)
    The article quotes the owner of Dynamic Sales Co., a St. Louis construction and supply firm, who says: “We’re just choosing to be bold and go out and sell our way out of this economic downturn. It has absolutely nothing to do with the Obama administration’s plans.”
    Tell it to the Keynesians.

  13. There are both short-run and long-run problems arising out of a bloated public sector. Propping it up with stimulus money forces more of the labor-market adjustment onto the private sector. I take that to be Scott Sumner’s argument.

    The long-run problem for state and local governments is the burden of pension costs for public-sector workers. Michael Barone addresses that in today’s Washington Examiner:

    As often happens in recessions, the short and long run are intersecting. For most of the states with severe budget shortfalls, wages, benefits and pensions for public workers are a big part of the problem.

    States that try to raise taxes lose population to lower-tax states. That is happening in a serious way in California and those exiting the state tend to be higher income residents. That loss exacerbates the volatility of revenue in the short run, and the budgetary gap in the long run.

    These microeconomic effects — unintended consequences — could easily swamp any theoretical macro effects. Or at least mitigate them.

  14. […] Brad DeLong says (in the comments below my previous post) that I got Steve Horwitz’s point about crowding out wrong and therefore my defense of Horwitz is inappropriate. Furthermore, then, I miss the importance of DeLong’s evaluation that Horwitz is incompetent. […]

  15. Pr. Woolsey:

    I was looking on Fred:

    average hourly earnings: construction
    average hourly earnings: manufacturing
    average hourly earnings: private sector
    weekly working hours: private sector
    weekly working hours: manufacturing
    consumer price index
    producer price index

    The codes of the data series are: AHECONS, AHEMAN, AHETPI, CPIAUCNS, PPIACO, AWHMAN, AWHNONAG

    CPIAUCNS (CPI) growth rate is higher than AHETPI (hourly earnings, private sector) from end 2007 to end 2008, then was lower, and has turned higher in the last months.

    I think the growth in AHE is a statistical illusion created by the differential impact of job losses at different earnings percentiles. If most of the laid off workers are low-skill, then the average hourly earning would show an increase even if all workers saw a (small) wage cut. The minimum wage has been raised, and the first decile of workers is very close to it (BLS data), besides being easier to fire than skilled and experienced workers.

    A friend of mine is now at a conferece on global equality, a very important one with the head of the World Bank and other people. She says: “news from the conference: high unemployment among low skilled workers, thus the increase in earnings we were talking about yesterday is sample selection bias; besides, there is an inverse relation between job tenure and job loss, and the impact on black men has been higher than on white men.”

    The sample selection bias can be strong in construction, because BLS says that contruction workers fell from 7.5 to 5.5 millions. The same effect for all the workers is much lower, because employment fell “only” from 113 to 105 millions.

  16. Thank you, Pietro.

    My post on nominal and real wages is here:

    Your argument about average nominal wages rising because lowest wage workers get laid off first seems plausibe enough, but there are other factors as well. For example, loss of employment by high wage autoworkers will bring down the average.

    I would think an increase in the minimum wage would be a factor raising nominal and real wages, right?

    By the way, don’t get fooled by ignoring that this net job loss during the recession is dwarfed by the new hires. It is just that there have been more quites and job losses (layoffs and firings) than new hires.

    If there were zero new hires and the only impact was layoffs of the newest first or something, then average wages would rise as you suggest. But, that isn’t the whole story by far.

    I presume that you are using that flawed series when you make claims about real wages? My guess is that our difference relates to the measure of the price level used.

    I have been using the GDP deflator, and more recently, the Consumer Expentiture Indexes (because they are monthly.)

  17. Re: “I read Horwitz. I thought his remarks about inflation were inadequate to a situation where nominal expenditure remains nearly 10 percent below trend…”

    OK. Game, Set, and Match.

    We can argue over the degree to which the current recession is casting a shadow forward, argue over the extent to which stimulus is not worth it because of the added debt burden, and argue over the extent to which the jobs created-or-saved by stimulus are on the wrong place.

    But you can’t–unless you want to be a clueless Republican hack–argue that every extra dollar spent by the government right now withdraws a dollar spent by the private sector.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s