Scott Sumner Goes Too Far

As I have said many times, Scott Sumner is the world’s greatest economics blogger. What makes him such a great blogger is not just that he is smart and witty, a terrific writer and a superb economist, but he is totally passionate about economics and is driven to explain to anyone who will listen why our economy unnecessarily fell into the deepest downturn since 1937 and has been needlessly stuck in the weakest recovery from any downturn on record. Scott loves economics so much, you might even think that he studied economics at UCLA. So the reason Scott is the greatest economics blogger in the world is that no one puts more thought, more effort, more of everything that he’s got into his blog than Scott does. So, Scott, for your sake, I hope that you get a life; for our sake, I hope that you don’t.

The only downside from our point of view about Scott’s obsession with blogging is that sometimes his enthusiasm gets the better of him. One of the more recent ideas that he has been obsessing about is the insight that fiscal policy is useless, because the Fed is committed to keeping inflation under 2%, which means that any fiscal stimulus would be offset by a monetary tightening if the stimulus raised the rate of inflation above 2%, as it would certainly do if it were effective. This insight about the interaction between fiscal and monetary policy allows Scott to conclude that the fiscal multiplier is zero, thereby allowing him to tweak Keynesians of all stripes, and especially his nemesis and role model, Paul Krugman, by demonstrating that fiscal policy is useless even at the dreaded zero lower bound. Scott’s insight is both clever and profound, and if Kydland and Prescott could win a Nobel Prize for writing a paper on time inconsistency, it’s not that big of a stretch to imagine that a few years down the road Scott could be in the running for the Nobel.

Okay, so having said all these nice things about Scott, why am I about to criticize him? Just this: it’s fine to say that the Fed has adopted a policy which renders the fiscal multiplier zero; it’s also correct to make a further point, which is that any estimate of the fiscal multiplier must be conditional on an (explicit or implicit) assumption about the stance of monetary policy or about the monetary authority’s reaction function to changes in fiscal policy. However, Scott in a post today has gone further, accusing Keynesians of confusion about how fiscal policy works unless they accept that all fiscal policy is monetary policy. Not only that, but Scott does this in a post in which he defends (in a manner of speaking) Bob Lucas and John Cochrane against a charge of economic illiteracy for believing that fiscal stimulus is never effective notwithstanding the results of the simple Keynesian model. Scott correctly says that it is possible to make a coherent argument that the fiscal multiplier implied by the Keynesian model will turn out to be zero in practice. But Scott then goes on to say that the textbook understanding of the Keynesian model is incoherent, and the only way to derive a positive fiscal multiplier is to assume that monetary policy is operating to make it so. Sorry, Scott, but that’s going too far.

For those of you who haven’t been paying attention, this whole dust-up started when Paul Krugman approvingly quoted Simon Wren-Lewis’s attempt to refute Bob Lucas and John Cochrane for denying that fiscal stimulus would be effective. Scott provides a reasonable defense of Lucas and Cochrane against the charge that they are economically illiterate, a defense I have no problem with. Here’s where Scott gets into trouble:

Wren-Lewis seems to be . . . making a simple logical error (which is common among Keynesians.)  He equates “spending” with “consumption.”  But the part of income not “spent” is saved, which means it’s spent on investment projects.  Remember that S=I, indeed saving is defined as the resources put into investment projects.  So the tax on consumers will reduce their ability to save and invest.

Scott, where is savings “defined as the resources put into investment projects?”  Savings is not identically equal to investment, the equality of savings and investment is an equilibrium condition. Savings is defined as that portion of income not consumed. Investment is that portion of expenditure not consumed. Income and expenditure are not identically equal to each other; they are equal in equilibrium. One way to see this is to recognize that there is a lag between income and expenditure.  A tax on consumers causes their saving to fall, because they finance their tax payments by reducing consumption and their savings. Investments are undertaken by businesses and are not immediately affected by the tax payments imposed on consumers. Scott continues:

So now let’s consider two possibilities.  In the first, the fiscal stimulus fails, and the increase in G is offset by a fall of $100 in after-tax income and private spending.  In that case, consumption might fall by $10 and saving would have to fall by about $90.  That’s just accounting.  But since S=I, the fall in saving will reduce investment by $100 $90.  So the Wren-Lewis’s example would be wrong, the $100 in taxes would reduce private spending by exactly $100.

Consider what Scott is saying here: assume that Wren-Lewis is wrong about the fiscal stimulus, so that the fiscal stimulus fails. Given that assumption, Scott is able to prove the very surprising result that “Wren-Lewis’s example would be wrong.” Amazing! If we assume that Wren-Lewis is wrong, then he is wrong. Now back to Scott:

I’m pretty sure my Keynesian readers won’t like the previous example.

What’s not to like?

So let’s assume the bridge building is a success, and national income rises by $100.  In that case private after-tax income will be unchanged.  But in that case with [we?] have a “free lunch” where the private sector would not reduce consumption at all.

I don’t know what this means. Does calling the increase in national income a free lunch qualify as a refutation?

Either way Wren-Lewis’s example is wrong.

There is no “either way.” If you assume that the example is wrong, there is no way for it not to be wrong.

If viewed as accounting it’s wrong because he ignores saving and investment.  If viewed as a behavioral explanation it’s wrong because he assumes consumption will fall, but that’s only true if the fiscal stimulus failed.

Viewing anything as accounting doesn’t allow you to prove anything. Accounting is just a system of definitions with no explanatory power, regardless of whether saving and investment are ignored or taken into account. As for the behavioral explanation, the assumption that consumption falls is made with respect to the pre-stimulus income. When the stimulus raises income enough to make post-stimulus disposable income equal to pre-stimulus disposable income, post-stimulus consumption is equal to pre-stimulus consumption.

Scott continues with only a trace of condescension:

Now that doesn’t mean the balanced budget multiplier is necessarily zero.  Here’s the criticism that Wren-Lewis should have made:

Cochrane ignores the fact that tax-financed bridge building will reduce private saving and hence boost interest rates.  This will increase the velocity of circulation, which will boost AD.

Scott may be right about this assertion, but he is not talking about the standard Keynesian model. Scott doesn’t like it when Keynesians insist that non-Keynesians accept their reasoning or be dismissed as ignoramuses, why does Scott insist that Keynesians accept his view of the world or be dismissed as not “even know[ing] how to defend their own model?”

It does no good to “refute” Cochrane with an example that implicitly accepts the crude Keynesian assumption that savings simply disappear down a rat-hole, and cause the economy to shrink.

The Keynesian assumption is that there is absolute liquidity preference, so the savings going down the rate hole is pure hoarding. As I pointed out in my post criticizing Robert Barro for his over the top dichotomy in a Wall Street Journal op-ed between Keynesian economics and regular economics, Keynesian fiscal stimulus works by transferring idle money balances in exchange for bonds at liquidity trap interest rate and using the proceeds to finance expenditure that goes into the pockets of people with finite (rather than infinite) money demand.  In that sense, Scott is right that there is a deep connection between the monetary side and the fiscal side in the Keynesian model, but it’s different from the one he stipulates.

The point of all this is not to be critical of Scott. Why would I want to be critical of one of my heroes and a potential Nobel laureate? The point is just that sometimes it pays to take a deep breath before flying off the handle, even if the target is Paul Krugman.


49 Responses to “Scott Sumner Goes Too Far”

  1. 1 Kevin Donoghue January 12, 2012 at 1:35 am

    As I remarked on his blog, if Scott can define saving as the resources put into investment projects then I suppose he can define an identity as an equation which holds for just one value of the variables. But if he is totally passionate about economics shouldn’t he get to grips with the fact that other people have been using quite different definitions for a very long time?


  2. 2 Nick Rowe January 12, 2012 at 5:58 am

    David: “Income and expenditure are not identically equal to each other; they are equal in equilibrium.”

    No. Actual income and actual expenditure are identically equal to each other.

    “One way to see this is to recognize that there is a lag between income and expenditure.”

    If you mean that an individual’s desired expenditure depends on his lagged income, that is a behavioural assumption that may or may not be true. I would say it depends both on his lagged and on his expected future income.


  3. 3 Andy Harless January 12, 2012 at 8:26 am

    Nick is right: income and expenditure ARE identically equal, even out of equilibrium. There is no time lag, because one person’s expenditure is someone else’s income.

    However, Scott is wrong about the definition of savings. S=I is an identity that can be proven once you have defined savings, but it is not the definition.


  4. 4 Andy Harless January 12, 2012 at 8:27 am

    My original comment had a link to a blog post I did on the subject, but I guess it got caught in the spam filter.


  5. 5 David Glasner January 12, 2012 at 8:50 am

    Kevin, Scott is entitled to use any definition he wants to, but he’s not entitled to impose his definition on everyone else. Obviously I agree with you that he ought to accept that other people have the right to define terms differently from the way he wants to define them, but I don’t see why that would depend on how passionate he is about economics.

    Nick, Sorry, but we disagree on this. Income measures payments made to factors of production. Expenditure measures payments for final goods and services. Those magnitudes may well be equal, but why are they necessarily equal except because you choose to define them so that they must be equal? It’s not the same as purchases and sales, where we are talking about two sides of the same transaction. In the basic Keynesian cross diagram we measure expenditure vertically and income horizontally. If you are defining income and expenditure to be identically equal, you are saying that every point is on the 45 degree line. You have an infinite number of possible points. Every state of the world is equally possible, because in every state of the world income and expenditure are identically equal. How do you get a theory that determines what the equilibrium income and expenditure are unless income and expenditure are defined in a way that allows them not to be equal. That’s how the model works by defining an aggregate expenditure function as a function of income and selecting the unique point of equilibrium at which income and expenditure are equal.

    E = E(Y) is the expenditure function

    E = Y is the equilibrium condition

    This makes no sense if E = E(E).

    All this and more is wonderfully explained in R. G. Lipsey’s paper, “Foundations of the Theory of National Income,” in Essays in Honour of Lord Robbins, and reprinted in Macroeconomic Theory: Selected Essays of Richard G. Lipsey the best paper I have ever read about the meaning and proper interpretation of the Keynesian model.


  6. 6 Kevin Donoghue January 12, 2012 at 9:11 am

    David, my feeling is that if a man is passionate about economics he ought to use existing terminology where possible, for the benefit of readers who may not know that he is inventing his own language. I fully agree with your response to Nick. Students would have a lot less trouble if economists were a bit clearer about whether their equations are to be read as accounting identities or equilibrium conditions.


  7. 7 Benjamin Cole January 12, 2012 at 9:37 am

    Hmm. I will comment later.


  8. 9 Nick Rowe January 12, 2012 at 9:54 am

    David: “Nick, Sorry, but we disagree on this. Income measures payments made to factors of production. Expenditure measures payments for final goods and services. Those magnitudes may well be equal, but why are they necessarily equal except because you choose to define them so that they must be equal?”

    Unless you wish to subtract depreciation, retained earnings, payments to foreigners, and indirect taxes, actual income and expenditure are just two ways of looking at the same thing.

    Ignoring those little caveats above, income means “revenue from the sale of newly-produced goods and services”

    The standard Keynesian Cross diagram, like Dick Lipsey’s, has actual income on the horizontal axis, and desired/planned expenditure on the vertical.

    For a micro analogy, it’s like a diagram with actual quantity of apples bought-and-sold on the horizontal axis, and quantity of apples demanded on the vertical. It’s an Engel Curve, for people who get all their income from selling apples, and demand other people’s apples.

    That is the big insight of Keynes (via Clower). Quantity demanded depends on quantity actually bought-and-sold.


  9. 10 Nick Rowe January 12, 2012 at 9:56 am

    Let me just add one word to this paragraph I just wrote to make it clearer:

    Unless you wish to subtract depreciation, retained earnings, payments to foreigners, and indirect taxes, actual income and actual expenditure are just two ways of looking at the same thing.


  10. 11 Andy Harless January 12, 2012 at 10:52 am


    Can you give a conceivable example where “payments for final goods and services” do not equal “payments made to factors of production” (in the sense in which you are using those terms)? I submit that any such example you give will require you to redefine either “income” or “expenditure” in order to remain consistent with the way the terms are normally used.

    The only way what you’re saying here makes any sense to me is if you are excluding inventory investment from expenditure. In that case, if I’m a widget producer, and I hire a worker to make a widget for my inventory, there is income for the worker but no expenditure on my part. But that concept of expenditure is a very odd one. Surely anyone, either a lay person or an economist (other than, perhaps, you), would say that I have “spent” money on producing that widget. And if we regard inventory production as something other than expenditure, we have a very strange world in which certain kinds of investment count as expenditure and others don’t.


  11. 12 bill woolsey January 12, 2012 at 11:54 am

    I am always a bit troubled by the profit firms earn from the goods that they purchased themselves as “unplanned inventory investment.”

    The way I have always looked at it, unplanned inventory investment is the residual between output and planned expenditure (which are what we normally think of as purchases.) And profits are a residual between factor incomes paid and output. When the residuals are added back, you get equality.

    But I have recently been motivated to think about economies without storable goods, which makes the somewhat weasily “unplanned inventory investment” unnecessary.

    Realized income equals realized expenditure, so realized investment equals realized saving. Right Nick?


  12. 13 Nick Rowe January 12, 2012 at 2:15 pm

    Bill: That’s how i think about it too. And our national income accounting ought to be able to handle a pure service economy, or one where goods are only produced to order. There are no inventories of haircuts.


  13. 14 rob January 12, 2012 at 3:50 pm

    Quick question:

    What about hoarding (defined as the portion of an individuals income that is not spent on either consumer or capital goods). This reduction in expenditure leads to a corresponding fall in (someones else’s) income, but is this money taken out of the system in this way to be counted as savings or not ? Does this correspond to Bill’s unplanned inventory investment ?


  14. 15 Benjamin Cole January 12, 2012 at 5:08 pm

    I think on the big picture, Sumner is right. The Fed can flummox fiscal policy, and monetary policy alone is better anyway (as fiscal policy is just monetary policy in drag, if accommodated).

    I suppose there are situations in which all businesses and consumers are just determined to sit on their hands, and then fiscal policy stimulus makes sense. These are rare episodes.


  15. 16 David Glasner January 12, 2012 at 6:18 pm

    Kevin, I think we are mostly in agreement on the merits and we needn’t belabor Scott’s passion.

    Benjamin, I’m all ears.

    Gene, I saw your comment which I mostly agree with, and see below for more about the identify of income and expenditure.

    Nick, Are you saying that because we can choose to ignore the items (depreciation, retained earnings, payments to foreigners, and indirect taxes) that may cause actual income and expenditure to differ, it is ok to say call them identical?

    “The standard Keynesian Cross diagram, like Dick Lipsey’s, has actual income on the horizontal axis, and desired/planned expenditure on the vertical.”

    I disagree and I think Lipsey does as well. The expenditure function tells us the amount of expenditure actually undertaken at every level of income. There is no difference between planned and actual expenditure because all expenditure plans are fulfilled.

    Here’s Lipsey (p. 12):

    Let us first consider planned and unplanned behavior in the schedule sense. The plans of investors to spend I* on fixed investment are fulfilled each period; the plans of households, (i) to spend a Yt-1, and to consumer goods to this value during period t and (ii) to consume in period t an amount equal to (1-a) Yt-1 always fulfilled, and the plans of producers to keep inventories unchanged are also fulfilled, and the plans of producers to keep inventories unchanged are also fulfilled. In each period no one’s plans are unfulfilled in the sense that no one is off his schedule relating his income to his desired expenditures on consumption goods, investment goods, and purchases of bonds.

    And here is Lipsey (p. 19) on the analogy between the income-expenditure model and the supply-demand model.

    We may now consider the erroneous but nearly ubiquitous analogy between savings-and-investment macro analysis, and demand-and-supply competitive price theory.

    In the simplest theory of price we can write
    D = D (p) (1)
    S = S (p) (2)
    D = S (3)

    Where D is desired purchases and S is desired sales. We now say that actual purchases are the same thing as actual sales:

    purchases ≡ sales

    and it is apparent that (1) and (2) describe two sets of plans about the same magnitude. Thus we can say that the plans of purchasers and the plans of sellers cannot be simultaneously fulfilled unless we are in equilibrium (i.e. unless (3) is satisfied). The difference between this model and the national income one is that, in the present price-theory case, both sets of real plans (buyers and sellers) are laid about the same magnitude (the amount of goods to be transferred in ownership), while in the national income case the two sets of real plans (savers’ and investors’) are laid about two different magnitudes. Thus the analogy often drawn between the two theories in respect of plans and realized quantities is an incorrect one.

    About Keynes and Clower, I think that they were actually anticipated by Say. Clower, at least, had the decency to call it Say’s Principle.

    Andy, Nick mentioned four. Isn’t that enough? Everyone these days seems to be very perturbed about all the cash that corporations are accumulating while disposable income is falling. If people are financing consumption out of accumulated cash or selling securities does expenditure equal income? No, I am not excluding inventory investment from expenditure.

    Bill, Unplanned inventory investment is a myth. It has no necessary place in the Keynesian model and the attempt to identify points away from the intersection of the expenditure function and the 45-degree line with unplanned inventory investment is just a mistake. The simple Keynesian cross is a timeless model, and points other than the intersection have no economic interpretation. To analyze the disequilibrium behavior of the system you have to introduce a lag structure and do a period analysis leading to equilibrium. That period analysis need not have any unplanned inventory investment. On all this, see Lipsey.

    Rob, In the Keynesian model, hoarding is treated as saving. And it has nothing to do with unplanned inventory investment, which does not exist in the Keynesian model, though it may play a role in some attempts to tell a story about what happens when aggregate demand changes. But the story is not grounded anywhere in the model.

    Benjamin, On the big picture, Scott is always right. I just like to nitpick on the details.


  16. 17 Nick Rowe January 13, 2012 at 5:28 am

    David: Interesting.

    Does Dick Lipsey have Y(t-1) on the horizontal axis, and E(t) on the vertical??

    If so, what he says is logically correct, but I strongly disagree with his way of doing it. He has snuck in an empirical assumption about a lag, and that’s why it looks different from supply and demand. It’s sort of like a cobweb model of S and D.


  17. 18 Nick Rowe January 13, 2012 at 5:35 am

    Here’s the analogy.

    Think of a Counot duopoly model, with two reaction functions.

    One way to think of the equilibrium is to assume that the firms take turns to choose Q, and each chooses its Q based on the other’s Q in the last period. So we cobweb slowly towards the equilibrium. That’s what Dick Lipsey is doing.

    The second way, and the way game theorists do it nowadays, is a one period simultaneous game where both jump instantly to the Nash equilibrium. That’s the way i think of the Keynesian Cross.

    Both ways are problematic, under uncertainty. If we make the plausible assumption that demand depends on expected current income, we have to say something about how those expectations are formed. Dick Lipsey is essentially assuming adaptive expectations.


  18. 19 David Glasner January 13, 2012 at 5:38 am

    Nick, No the standard Keynesian Cross is timeless and lagless and E is a simple function of Y. His point is that you can’t tease out of a timeless model any inference about the behavior of the system out of equilibrium so that there is no basis in the model for the unplanned investment. In the model you can’t have a solution unless E = Y, but that is because the model has no solution other than equilibrium. To get any idea of what the model does out of equilibrium, you have to introduce at least one lag, but there are a number of possible lags to choose from and one needn’t assume that unplanned inventory investment occurs to trace out an adjustment path from one equilibrium to another. As I said, it’s just a wonderful article.


  19. 20 Kevin Donoghue January 13, 2012 at 5:42 am

    Does Dick Lipsey have Y(t-1) on the horizontal axis, and E(t) on the vertical?

    I’ll be surprised if it turns out that he does. Supposing he doesn’t, is he infringing any rule?


  20. 21 Kevin Donoghue January 13, 2012 at 5:45 am

    Got David’s answer while I was posting my question. Quick service!


  21. 22 Robb Lutton January 13, 2012 at 5:58 am

    Forget about theoretical hairsplitting over the labels on mythic curves. What about the actual real economy as Scott sees it.

    Isn’t it true that if income equals expenditures then there is no problem with AD? Isn’t the lack of AD precisely that expenditures lag income?


  22. 23 Barry January 13, 2012 at 7:05 am

    David: “As I have said many times, Scott Sumner is the world’s greatest economics blogger. ”

    I’m totally willing to put him up against Paul Krugman – who’s been right, and whose been wrong; who’s been more honest, and who’s been less.

    David, you own article seems to refute your thesis, unless the bar for “world’s greatest economics blogger ” is pretty low. As for enthusiasm, go to some Austrian and/or goldbug site if you want to see enthusiasm. Enthusiasm doesn’t mean much by itself.


  23. 24 David Glasner January 13, 2012 at 7:20 am

    Nick, The way to read Lipsey (or better the way I read Lipsey, you will read him for yourself and draw your own conclusions) is to provide an interpretation of the simple Keynesian model in a way that does not commit gross logical or methodological erros and does not confuse identities with equilibrium conditions and accounting identities with causal assertions. He is not himself making theoretical assumptions about what is going on he is just showing us how to work with the model and derive whatever implications can be derived from the limited theoretical content embedded in the model. You are right that when he provides an example of how to introduce a lag structure he is assuming adaptive expectations but the exercise is pedagogical not substantive. That’s why the essay is entitled “The Foundations of the Theory of National Income.”


  24. 25 David Glasner January 13, 2012 at 7:20 am

    Kevin, Only to a select few.


  25. 26 David Glasner January 13, 2012 at 7:26 am

    Robb, No it is absolutely not true that if income equals expenditure, there is no problem with AD. Within the four corners of the Keynesian model, all that is implied by income equalling expenditure is that there is an equilibrium in the sense that there is no tendency for income and expenditure to change. That level of income and expenditure is not necessarily sufficient to generate full employment. Deficient AD is not the result of a lag, but because (again within the Keynesian model) because households are trying to save (whatever that means) or businesses are not willling to invest.


  26. 27 David Glasner January 13, 2012 at 7:40 am

    Barry, Scott is a great blogger because he happens to have a lot of very creative ideas about economics, keeps coming up with new ones, and is willing to keep writing (apparently non-stop) interesting stuff about them, and because he is constantly responding to his readers (way more promptly than I do) and engaging them on every point that they raise and by the force of his ideas, persistence, humor, and, yes, enthusiasm has almost single-handedly forced the rest of the economics establishment to sit up and take notice. You compare him to Austrians and goldbugs. I don’t spend much time on those sites, but I would be surprised if you could name a single one that comes within miles of Scott’s in the quality or quantity of output over the past three years since he started blogging. Just because I found one post of his posts (there have been a few others over the past three years) a little bit deficient doesn’t mean that he’s not a great blogger. Even Babe Ruth didn’t hit a home run with every swing.


  27. 28 Peter K. January 13, 2012 at 9:07 am

    As a non-economist long-time reader of Krugman and proponent of NGDP level targeting, I find this discussion fascinating.

    Benjamin Cole commented “The Fed can flummox fiscal policy, and monetary policy alone is better anyway (as fiscal policy is just monetary policy in drag, if accommodated).”

    Why is one “better”? Stimulus/demand is added via banks instead of via government? We have the concrete case of Bernanke repeatedly asking for help from other parts of the governments, meaning fiscal policy in my mind. The problem was that Republicans were blocking fiscal policy.

    Lately the Fed has been weighing in on government housing policy and asking for help in that area. The most successful aspect of Obama’s stimulus was aid to state and local governments, which helped short-circuit the downdraft in demand. I believe Bernanke has said the stimulus helped and one way fiscal policy helps is to take some of the political heat off of central bank activism.

    During the financial crisis every major government engaged in fiscal stimulus, including China which had a stimulus proportionally larger than Obama’s. Also isn’t it widely accepted that World War II helped pull the global economy out of the Great Depression?


  28. 29 Andy Harless January 13, 2012 at 2:19 pm

    “If people are financing consumption out of accumulated cash or selling securities does expenditure equal income?”

    Yes, because the people who sell to the consumers are receiving income (or else they’re disinvesting, making a negative expenditure).


  29. 30 Benjamin Cole January 13, 2012 at 5:30 pm

    Peter K-

    You raise an interesting question, and I am not a PHD economist. I will give you my answer.

    My view is that fiscal policy should be what we need the federal government to do. Provide for national defense, for example—the level of defense spending should be determined by need, not microeconomic policy. Perhaps we decide we need national health insurance—okay, that’s fine, but the level of spending should be determined by the smallest amount we can spend and do the job.

    Okay, you can say, “Okay, but we can cut taxes.” That’s a good plank, but then we should also cut spending, or else we keep piling up deficits.

    I won’t even get into the messy zone of giving extremely political Congressman and Senators an excuse to spend more money in their districts.

    I would prefer the Fed conduct QE, and keep on pressuring, until more money is being spent in the private sector.

    As a last resort (and here I venture into terra incognito, and perhaps the netherworld) I would favor government outlays financed not by borrowing, but simply printing more money. There I said it.

    But if that doesn’t cause first economic growth and then inflation, I don’t know what will. If not overdone, I have never understood why boosting the economy this way is considered so bad. Why is new money worse than borrowed money (and extra debts for our children).

    Okay, throw the creme pies at me now….


  30. 31 Benjamin Cole January 13, 2012 at 5:31 pm

    PS–My use of the word “microeconomic” should be “macroeconomic.”


  31. 32 Lorenzo from Oz January 13, 2012 at 7:20 pm

    Is income and expenditure in this discussion nominal or “real”? It is not entirely clear to me, since money hoarding is purely monetary phenomena but does mean that (nominal) income need not equal expenditure. Indeed, if it is nominal income and expenditure we are talking about, expenditure can also exceed income (via external borrowing).


  32. 33 TravisA January 14, 2012 at 10:34 am

    @David: I am curious that you and Nick/Andy still don’t agree about expenditure equaling income. Is it as simple as you are making the point that planned expenditure doesn’t necessarily equal planned income? Or are you saying that actual expenditure doesn’t equal actual income, even in the accounting sense?

    Getting back to your definition of the terms in a comment you made above:

    “Income measures payments made to factors of production. Expenditure measures payments for final goods and services.”

    Are you talking about actual or planned? If actual, how is ‘expenditure’ here different from ‘purchases’?

    I can see how expenditure is different from income after taxes. But I am not sure whether it’s materially different whether the taxes are sales (or other indirect) taxes or income taxes. Regardless, expenditure is equal to income before taxes, since income before taxes is sales apportioned to the factors of production.

    Depreciation shouldn’t make a difference in determining whether expenditure equals income, as depreciation is not a monetary payment.

    I agree that if some of the income goes to foreigners, then *domestic income* doesn’t equal *domestic expenditure*. But in a closed economy, that’s not relevant.

    Finally, expenditure (payments) from retained earnings becomes someone’s income, so again, actual expenditure equals actual income.

    Thus the only exception to expenditure equaling *net* income is that taxes are paid to the government out of a portion of the expenditure. But it is true that actual income (before all taxes) equals actual expenditure, simply as a matter of accounting since all sales becomes someone’s income and then a portion of the sales is paid to the government.

    Thoughts? How is Lipsey using the terms expenditure and income?


  33. 34 Nick Rowe January 14, 2012 at 11:51 am

    David: that sounds right to me. Suppose we start in one equilibrium, and then the Expenditure Shedule shifts, then unless we explicitly introduce some behavioural or expectational lags, all we can say is that the model jumps straight to the new equilibrium. (It’s exactly like the assumption that the players in Cournot oligopoly jump straight to the Nash equilibrium).

    Yep. All that stuff about unplanned inventories is a red herring. I prefer to think of a service economy. The lag i like to introduce is an expectational lag. Each individual decides to increase his expenditure, but doesn’t know that all other individuals are planning to do the same thing. So each individual is surprised to see his income is higher than he expected it would be. Do they then revise their expectations? Or do they think this was just a lucky one-off increase in income? If they revise their expectations so that Ye=Y(t-1), for example, we get the slow multiplier process.


  34. 35 David Glasner January 14, 2012 at 7:38 pm

    Peter, Better in this context means (I think) is more effective in moving NGDP in the desired direction. Bernanke’s public statements are not necessarily to be taken at face value. I don’t say that fiscal policy has no effect when deployed in a depression or deep recession, and I am not against using it in such circumstances, but I do think that it is monetary policy that does most of the heavy lifting. The problem with WWII as evidence for fiscal policy is that there was also a huge monetary expansion and inflation (partially suppressed, or disguised, by price controls) taking place the same time, so it is hard to tease out the effect unique to either one alone.

    Andy, I did not phrase my question precisely enough. I should have written: “If households are financing consumption out of accumulated cash or selling securities does expenditure necessarily equal income at the moment that they increase their spending/” The answer is no if the increased spending is to business firms which do not immediately distribute the proceeds of the increased spending to hired factors of production or to the owners of the firm. You are right that the increase expenditure leads to an increase in income, but the two are not identical. Chickens cause eggs and eggs cause chickens, but chickens and eggs are not identical.

    Benjamin, One reason for spending more in a depression is that that is when the government gets the biggest bang for the buck. Building roads and bridges and schools is cheaper then, because workers are more readily available and there aren’t so many private employers to bid against in seeking workers. Borrowing costs are also low. It’s also a good idea for businesses to invest in low demand periods except that they are more concerned about whether there will be sufficient demand to purchase the extra output, a consideration that doesn’t inhibit the government, at least not as much. Also the government is more likely than private businesses to take into account that its own increase in spending will increase the income and expenditure of the private sector; private businesses are too small to take that effect into account.

    Lorenzo, In principle it can be either, but the original Keynesian model was specified in terms of an exogenous price level, so that income and expenditure were in real terms. I agree that external borrowing is one way of accounting for a difference between income and expenditure.

    TravisA, Nick talked about planned expenditure and I said that there was no reason to introduce that concept into the model. In the simplest version of the model (with no lags), all you get is an equilibrium solution. In equilibrium, income equals expenditure, because that is what defines the equilibrium. So in equilibrium, income and expenditure are indeed identical in equilibrium. The issue is are income and expenditure always identical, i.e., out of equilibrium. The answer is that depends on how we define income expenditure. But according to most common sense definitions, the answer is no they do not have to be equal.

    I am talking about actual expenditure, which is spending on final goods and purchases by households and businesses. Income is payments made businesses to their owners and to factors of production. It does not include the undistributed profits of business (retained earnings). You are right that retained earnings eventually become income when they are distributed to the owners. That is just like saying that eggs eventually become chickens. It doesn’t establish that eggs are chickens or that undistributed profits are income.

    Nick, I am glad that it seems I have made myself clear (to you at least). I think that I agree with you about the alternative lag structures.


  35. 36 Mitch January 14, 2012 at 9:30 pm

    I have to say that I have been finding Krugman’s Keynesian arguments much more persuasive than those of his opponents. I am not an economist, but I have sufficient technical training to evaluate quantitative arguments, and I find the level of rebuttals against him quite lacking.

    What is so disappointing is that the arguments against fiscal policy are generally so trivially wrong. They inevitably depend on an assumption that markets clear and supply is equal to demand. This may be true in normal circumstances, but *during a depression* they manifestly are not. (Clearly the labor market is not clearing, or we wouldn’t have 8+% unemployment.)

    So, for example, consider the back and forth between Krugman and John Cochrane (which Krugman pointed out on his blog).

    Cochrane’s argument basically comes down to: if the government borrows money and spends it, someone else won’t. This “crowding out” renders fiscal stimulus ineffective.

    Again, this is silly: *in a depression* the market for lending may not be clearing *either*, and in fact no one may be borrowing the money. The “market” rate for loans may be below 0%, or the banks may be reluctant to lower their interest rates for some other reason to enable the market to clear.

    So, if the government borrows money during a depression it probably isn’t “crowding out” anyone else’s borrowing, and if it hires unemployed people it won’t be driving up wages. *That* is the essence of the argument, and these other arguments don’t really even engage with it.


  36. 37 TravisA January 14, 2012 at 10:52 pm

    @David: This horse is getting thoroughly beaten, but I wonder what the economic significance is to saying that income and expenditure aren’t in equilibrium because retained earnings are with the firm and not with the shareholders. It seems rather insignificant. Much more economically significant would be borrowing from abroad. But other than that and taxes, I can’t think of any example where this disequilibrium is economically significant. More interesting seems to be to focus on planned expenditure and income and unexpected changes in the two. Even Nick’s latest response seems to be focusing more on planned expenditure changing as a result of changes in actual income. In other words, I don’t know what difference it makes economically to say that expenditure/income are always in equilibrium but that doesn’t mean this equilibrium point is changing all of the time.


  37. 38 TravisA January 14, 2012 at 11:01 pm

    Whoops. Last sentence should read, “…but that doesn’t mean this equilibrium point *isn’t* changing all of the time.” Bad double negative 😉


  38. 39 David Glasner January 15, 2012 at 7:20 am

    Mitch, I agree that fiscal policy can be effective in a depression. At one time, the issue was whether monetary policy could also be effective. Cochrane’s position is a repetition of what used to be called the Treasury view, but that view never commanded the approval of most non-Keynesians. If the Treasury view defines what it means to be a non-Keynesian, then I am a Keynesian, though I have never considered myself to be one. But I am not concerned about labels.

    Travis, This particular discussion started when I criticized Scott for asserting that savings is really no different from investment. I pointed out that that is not a correct interpretation of the Keynesian model, because expenditure consists of household spending on consumption and business spending on investment while income is the earnings of households which is either spent on consumption or saved. Since expenditure (spending) is not the same as income (earnings), savings is not the same as investment. You can do almost all the calculations associated with the Keynesian model even if you don’t bother with this distinction between an accounting identity and an equilibrium condition, but every once in a while failure to recognize the distinction leads people into fallacies. So I am not claiming that the (largely philosophical) distinction that I am making makes a huge practical difference; it just clarifies how to think clearly about what the model really means. You can work with the model (i.e., find numerical solutions) quite well even if you don’t really understand what the model means.

    In the simple Keynesian model there is only one equilibrium solution; to say anything about how equilibrium changes over time, you have to introduce some dynamic elements and that opens up all kinds of interesting theoretical options.


  39. 40 Andy Harless January 15, 2012 at 2:55 pm

    “If households are financing consumption out of accumulated cash or selling securities does expenditure necessarily equal income at the moment that they increase their spending/” The answer is no if the increased spending is to business firms which do not immediately distribute the proceeds of the increased spending to hired factors of production or to the owners of the firm.

    OK if your definition of income doesn’t include undistributed profits, but that strikes me as a perverse and not particularly useful definition. If you happen to spend the money at a business that happens to be organized as a direct proprietorship, the spending increases income; if you spend it at an otherwise identical business organized as a corporation, then it doesn’t. I don’t see that there’s much practical macroeconomic substance to that distinction.


  40. 41 Mitch January 15, 2012 at 6:56 pm

    David –

    Just to be clear: I know that the position I was criticizing was not yours – you have made that clear on a number of occasions. What is so surprising is that in economics such obviously wrong ideas still persist.


  41. 42 David Glasner January 17, 2012 at 9:08 am

    Andy, Payments to factors of production are not made continuously. You probably get paid every two weeks or every four weeks. The receipts of business firms are not immediately converted into payments to factors of production. I agree there is a circular flow of income so the receipts of business firms eventually flow back to factors of production, but the process is not continuous and the receipts of business firms are simply not identical to payments to factors of production. There are injections into the circular flow, and there are leakages from the circular flow, and when injections into the circular flow (expenditures not previously generated by income) exceed the leakages from the circular flow, the flow tends to expand until the injections again just balance the leakages and there is a new equilibrium with expenditure equal to income. Again, just think of expenditures as eggs and income as chickens.

    Mitch, Economics is complicated and it’s hard to come up with evidence that clearly rules out one theory or another. Just about any fact or observation can be reinterpreted in such a way as to be consistent with someone’s prior beliefs. There’s no use complaining about it; it’s just the way of the world.


  1. 1 TheMoneyIllusion » Saving isn’t “setting money aside,” it’s BUILDING CAPITAL GOODS Trackback on January 12, 2012 at 6:40 am
  2. 2 I Figured Out What Scott Sumner Is Talking About « Uneasy Money Trackback on January 16, 2012 at 6:01 pm
  3. 3 Advice to Scott: Avoid Accounting Identities at ALL Costs « Uneasy Money Trackback on January 22, 2012 at 10:33 am
  4. 4 Krugman on Mistaken Identities « Uneasy Money Trackback on January 31, 2012 at 1:18 pm
  5. 5 Thompson’s Reformulation of Macroeconomic Theory, Part IV « Uneasy Money Trackback on October 19, 2012 at 12:36 pm
  6. 6 Hawtrey’s Good and Bad Trade: Part II | Uneasy Money Trackback on September 24, 2013 at 8:41 pm
  7. 7 CAUTION Accounting Identity Handle with Care | Uneasy Money Trackback on February 11, 2015 at 5:44 pm

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 )

Facebook photo

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

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

About Me

David Glasner
Washington, DC

I am an economist in the Washington DC area. My research and writing has been mostly on monetary economics and policy and the history of economics. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey’s unduly neglected contributions to the attention of a wider audience.

My new book Studies in the History of Monetary Theory: Controversies and Clarifications has been published by Palgrave Macmillan

Follow me on Twitter @david_glasner


Enter your email address to follow this blog and receive notifications of new posts by email.

Join 3,263 other subscribers
Follow Uneasy Money on

%d bloggers like this: