Germane to some of the ongoing debates over fiscal policy effectiveness [1] [2]:
Fiscal policy multipliers are central to Keynesian macroeconomics. In this paper I explore a
possible microeconomic foundation for one fundamental theory of income determination, the
‘Keynesian cross’. My model deviates from a Walrasian equilibrium model only by the assumption
of imperfect competition in the goods market. I show that textbook fiscal policy multipliers arise as a
limiting case.
Under imperfect competition, firms are always eager to sell an additional unit of output, since
price exceeds marginal cost. This profit margin creates the potential for the multiplier. An expansionary
change in fiscal policy increases aggregate expenditure, which increases profits, which in turn
increases expenditure, and so on.
The theme that imperfect competition may be crucial to macroeconomic issues is increasingly
prevalent. See, for example, the work of Weitzman (1982) Hart (1982) Solow (1984), Blanchard and
Kiyotaki (1985), and Startz (1986). The purpose of the model presented here is partly pedagogical. I
therefore do not hesitate making strong (yet not implausible) assumptions about the economic
structure: Cobb-Douglas utility, constant marginal cost, and constant mark-up pricing. There is no
reason to suppose, however, that the sorts of effects highlighted here are specific to these assumptions.
While the model is in some ways surprisingly similar to the standard Keynesian model, in other
ways it differs greatly. In particular, it incorporates both an equilibrium labor market and a static
environment. These features are chosen for simplicity rather than realism. The goal is not to provide
a complete reformulation of Keynesian economics, but only to illustrate what sort of Keynesian
results one can obtain with a small movement away from Walrasian equilibrium in the direction. of
imperfect competition.…
Both increases in government purchases and reductions in taxes increase welfare. In standard
analysis, increases in G or reductions in T are financed by future generations. Here, these changes
are financed by reductions in government workers W. In neither case is it surprising that the welfare
of current individuals increases.
4. Conclusion
The model examined here is surprisingly similar to both Walrasian models of general equilibrium
and Keynesian models of income determination. It deviates from a standard general equilibrium
model only by the assumption of imperfect competition in the goods market. As competition in the
goods market becomes less perfect, the fiscal policy multipliers approach the values implied by the
Keynesian cross.
The model could be usefully extended in several directions. First, the labor market might be made
less classical. One could posit imperfect competition among workers, for example. Some of the rents
generated by expansionary fiscal policy would therefore accrue as labor income. The multiplier
would thus work through both labor income and firm profits.
Second, the model could be made intertemporal. The impact of debt-financed fiscal policy
obviously cannot be studied in a static model. That saving and investment play an important role in
standard Keynesian analysis also suggests extending this model to a dynamic setting.
My description of this paper as one my favorites is not false flattery; I used to teach this paper regularly to my PhD students (syllabus here). A working paper version is available here. The paper was published in Economics Letters (1988), reprinted in New Keynesian Economics, and was written by Greg Mankiw.
There’s no investment and all the “pain” of deficit spending comes at the expense of government workers who don’t get included in the welfare analysis.
So, yes, if you sweep all the costs of deficit spending under the normative rug then its a no brainer, spend away.
I liked that paper too, and taught it. But (IIRC) it is very definitely NOT Keynesian.
Keynes’ GT lists three “Classical postulates”:
1. real wage = marginal product of labour
2. real wage = marginal disutility of labour
3. Say’s Law
Keynes accepted 1, and denied 2 and 3.
Greg Mankiw’s model denied 1 (because of imperfect competition in output markets), but accepted 2 and 3.
That’s exactly the opposite of Keynes’ GT.
He got the multiplier from differentiating 2 (the labour supply curve). Increased Government spending reduced workers’ wealth, and caused them to supply more labour, so output increased. Decidedly un-Keynesian.
What is interesting about this paper? The model was out of date when the paper was written. Even ignoring the oversimplifications already noted above by pushmedia1, there is no intertemporal dimension. Specifically, there is no life cycle behavior, which drastically affects the magnitude of the multiplier. In a credible model, households must be treated as doing some kind of intertemporal maximization. As a result, it is important to distinguish between permanent and temporary changes in government spending (see Barro, JPE 1980 or 1981, on the output effects of government purchases). Also, there are no lump sum taxes in reality, and essentially all tax changes must have distortion effects, totally absent from this model. The Keynesian multiplier emerges in the “limiting” case of mu=1, where the profit margin is 1, is of no practical significance whatsoever. Most economists regard the economy as either monopolistically competitive with small profit margins because of entry (e.g., Peretto-style second generation models of endogenous growth) or sufficiently close to perfect competition that the imperfections can be ignored (most Chicago and Minnesota treatments). We can imagine what life would be like if the gravitational constant went to infinity or the coefficient of friction went to zero, but that’s not too useful for predicting behavior in the world we live in. The same is true of profit margins of 1, especially when embedded in a grotesquely oversimplified model.
John: (IIRC) it got a multiplier without assuming nominal rigidities. That was interesting.
Actually – Mankiw was not denying the Keynesian results if we assume unemployed resources. What Mankiw was saying is that one could get the Fama results if one assumed we were at full employment. Alas – as DeLong noted – Fama’s oped explicitly assumed unemployed resources.
Good post and thanks for those links, Menzie.
Off topic, DFA’s Small Cap Value fund performed well in 2008 and over the past 10 years, it is ranked in the top 12% of funds in its category by Morningstar. David Booth, who runs the firm, just gave $300 million to the UofC. Not too shaby for a believer in efficient markets.
It’s interesting how early in Mankiw’s career he seemed to be predominantly motivated by academic success and respect, so he would be willing to say intelligent things that went strongly against Republican ideology, like in one of his early textbooks where he referred to supply-siders as charletons and cranks, and like in this article where he wrote, “Both increases in government purchases and reductions in taxes increase welfare.”
It’s striking how different the Mankiw of today is, the professionally embarrassing things he says to please the Republicans, the constant stream of intentional disinformation. Of course, the reference to charletons and cranks was removed from later editions of his text.
So as a non-economist, what am I supposed to think when on one side Paul Krugman is saying that the bigger the fiscal stimulus the better and on the other Eugene Fama says that any fiscal stimulus will have zero impact with the absolute certainty of the simple equation PI = PS + CS + GS.
They couldn’t be any farther apart if they were two astronomers arguing whether the earth circles the sun or the sun circles the earth. Yet this apparently is the state of the field of economics today.
If this were the field of astronomy and some astronomer claimed that the sun circled the earth, there would be a hue and cry from all astronomers and the offender would be ridiculed and ostracized.
Even in a field as contentious as global climate change, the layman has been able to detect a general consensus of thousands of climate scientists and the marginalization of the fringe naysayers.
But in economics that never seems to happen. Two completely opposing views seem to persist for decades or even centuries with no perceivable resolution. And even if the implications of the disagreement have profound implications for the general public, the great majority of economists seem to maintain polite silence on the matter, allowing a few headline economists to fight it out in the arena.
It makes it hard for a non-economist to sort out. On the one hand you have Krugman who has a Nobel Prize under his belt and specializes in macroeconomics. Perhaps he should be given greater weight in this discussion than Fama who is more a microeconomics and finance guy. But it’s harder to dismiss Edmund Prescott who also has his Nobel.
What a curious field of study economics is. I can’t think of any other science in which so fundamental differences can persist for so long with such apparent equanimity.
It’s hard to figure out the effects of monetary and fiscal policy because you almost never (thank God) get the random fluctuations in monetary and fiscal policy you would need to figure it out clearly.
The side which is better at ostracism and ridicule isn’t always the side that’s right.
Fama is wrong, because an increase in government demand for loans will create an excess demand for loans, which will put upward pressure on interest rates, which will cause an increase in the supply and/or a decrease in the demand for money, which will cause an excess supply of money, and that excess supply of money will be loaned out, meeting the government demand for loans without taking loans away from private borrowers. The excess supply of money creates an excess demand for goods, so the government’s borrowing to spend can increase aggregate demand.
What is embarrassing to me as a macroeconomist is that even good macroeconomists, who know Fama is wrong, cannot properly explain *why* he is wrong. Many of them just went off talking about inventories, which is a red herring. (It would make no difference if all output were services, which cannot be stored in inventories, for example).
Richard, you are misquoting Mankiw. He maintains his integrity in my book. This NY Times op ed was balanced:
http://www.nytimes.com/2008/11/30/business/economy/30view.html http://www.nytimes.com/2008/11/30/business/economy/30view.html
I have perhaps an even more basic question than the value of the multiplier. I would like to understand why it matters whether a fiscal stimulus is spent or saved. Surely, in order to save it is necessary to find someone to lend to (even just holding a banknote is effectively an interest-free loan to the government). And they are not going to borrow unless they have a use for the money, so any money that is saved must be spent anyway. Thanks for any explanations.
As a natural science student myself before I studied economics, I have the same impression of economics as Joseph, except that I think that science would have an even better response than he suggests. If an astronomer claimed that the sun orbited the earth, the other astronomers would point out how this theory was inconsistent with observations. Scientists seem to narrow discussion down to points of disagreement, resolve the disagreement and move on.
Rebel:
I think the simplest explanation of why savings provides no stimulus, is that the “safe” place to invest savings is always government bonds. Thus, suppose you have a $100 billion stimulus. Suppose this is all put into government bonds. All this does then, is finance the stimulus. The government borrows the money to finance the “stimulus”, sends out the checks, and then gets the money right back. Where has any stimulus occured?
The problem is that increased savings does not really automatically translate to increased investment. To get a genuine increase in capital investment, that is investment in future production of goods and services, there needs to be some anticipation of demand for those goods and services, for there to be some expectation of profitable investment.
When you get too great a disparity in wealth and income, what happens is that too little is in the hands of those who create future consumer demand, and too much is in the hands of those with a desire to save and invest. Instead of going into capital investment, the excess savings starts to seek for other places to store wealth; you get bubbles in real estate, in commodities, or speculative financial creations which never lead to real capital investment.
This is where we are today. There is little willingness to fund capital investment by business engaged in production of future goods and services; some of these businesses are having difficulty getting loans. Yet interest rates for government borrowing are near zero. If we extend this to a global analysis, you also have inefficient capital markets in places like China, leading the Chinese also to invest their money in safe US government bonds, rather than in capital investment in their own country.
The solution thus needs to be direct capital investment by the government, policies that encourage actual capital investment (such as capital investment tax credits), as well as policies which increase incomes for working people, stimulating demand. The more redistribution you get, ultimately, the less you will need government spending to maintain full employment.
And, we should consider making these ambitious capital investment plans global in scope. Our large trade deficits with China would also be easily sustainable if we could use most of the excess to make capital investments there which create jobs and further stimulate demand.
acerimusdux,
Thanks for trying, but I do not think your answer solves the problem. The government has to sell bonds (unless it raises taxes, which I assume is not what is proposed) to raise the extra money it is spending, so someone has to “put money into government bonds” anyway.
Yes, that was my point.
You can have:
a> money going into government bonds doing nothing (say temporary tax cuts).
or
b> money going into government bonds being used to produce a stimulus.
Only putting it towards capital investment, creating jobs, or boosting consumption will produce the stimulus.
You can, of course, have increased demand for government bonds without the government borrowing; what you have then is just even further downward pressure on interest rates, which are already too low.
The government isn’t having any difficulty at all right now selling new bonds. They are able to borrow at very low rates. Private capital markets collectively have apparently decided that the government can invest their money more wisely than they can themselves.
acerimusdux,
Are you saying that a temporary tax cut is not a stimulus? Many (including the British government) would argue that a temporary tax cut is the most stimulating form of fiscal policy. As for me, I do not see much difference between (a) and (b). But perhaps someone else can offer their view.
Why do all of these multiplier arguments ignore the fact that the future demand is pulled forward, leaving a gaping hole where that future demand used to be? We would laugh an economist off of the stage if he proposed that you could infinitely borrow from the future via credit cards in order to fund today’s spending. Credit cards and Keynesian deficit spending are not monetary perpetual motion machines, and there is no reason to believe that future demand will magically recover when you spend it today.
Well, no I’m not saying a temporary tax cut is never an effective stimulus. I’m saying that in current current conditions it isn’t. If you go back to the “stimulus checks” that were sent out in the spring, less than half of them were spent, and they provided a very weak stimulus. But, even that was before we had clearly entered into liquidity trap conditions and a deflationary spiral.
At present time, with interest rates and inflation expectations near zero, and prices across the board continuing to plunge, while more and more layoffs are announced, even less of those checks would be spent. There’s just too much economic uncertainty, and those dollars seem to get more valuable as long as you hold them.
Now, I suppose as long as there is some positive stimulus, you could argue that you just need to make the cuts large enough to have the same effect. But if that’s the case, I think we would need to be talking about a stimulus on the order of more than $4 trillion over the next 2 years. Problem is, total government revenues come to only about $2.6 trillion a year, and nearly a trillion of that is dedicated to social insurance programs. Even eliminating all other taxes would only get you just over $3 trillion total over the next 2 years.
Meanwhile, on the spending side, it isn’t so hard to find the places that $1 trillion in spending is needed. A bipartisan group of centrists, including Schwarzenegger, Rendell, and Bloomberg, since even before the crisis hit, has been promoting the fact that we are badly in need of well over $1 trillion in new infrastructure spending. These are things we need anyway, repaired roads, bridges, levees, a modern electrical grid, more modern communication and transportation systems. Why not finance them now, when government is able to borrow at record low rates? Why not hire the needed labor now, when you are getting them off the employment line, rather than hiring them away from productive jobs in private industry? Why not build these things now, when demand for the resources and materials needed is lax and prices are low?
Yes, we are shifting that demand from the future; we are shifting from a future strong economy when it would likely be harmful, inflationary, and crowd out private industry, to a present when it is needed, and it will employ unused resources, and stimulate growth.
I understand the desire for smaller government. And if this were 1979, with over investment, high inflation, high taxes, over stimulated demand, high interest rates, restricted trade, and excessive government regulation, then yes I would likely support tax cuts (and even some deregulation). In such an environment they would still have some stimulative effect.
But if Republicans really wanted a sensible smaller government they should have recognized
*Some constant moderate level of redistribution is needed to sustain demand with lower levels of government spending. Falling real wages and the growing gap between rich and poor since 1979 are contributing to the shortfall in demand.
*One of the most effective ways to help achieve this without the heavy hand of government, would be a moderate inflation rate. Inflation is effectively a tax on wealth. At stable, moderate levels, it can even stimulate capital investment (as capital investments pay off in future goods rather than future dollars). A sensible long term inflation target might be around 3.5%, rather than 1.5%.
*Declining fixed investment, both private and governmental, has been another problem. Some kinds of government spending are needed and contribute to economic productivity. If you want to maintain a healthy modern economy, you need to maintain the infrastructure.
*Some kinds of reasonable regulation are also needed. There’s a difference between eliminating red tape, and allowing the financial shenanigans that have led to scandal after scandal over the last 30 years.
*We need to maintain a sustainable trade deficit. Trade can be beneficial to a degree, but not at the cost of massive trade deficits, and a loss of US manufacturing, to the point that we are consuming far more than we produce and have to go into debt to sustain it.
So if you really want a smaller government, in the long run, you need to fix some of the above. Don’t neglect infrastructure, fix the trade deficit (with moderate trade restrictions), and support more equity in compensation and taxes. But, in the short run, we are still going to need something of a jump start, with where we are now.