28 Jan 2011

Sumner Provides an Answer…Or Does He?

Economics 3 Comments

Apparently there are a fixed number of comments to be posted on this blog per week, and so if I don’t keep the posts flowing, you guys pile up 58 comments in a single one.

Because of this, many readers may not have seen Scott Sumner’s reply to my queries. (If you need to refresh your memory, here ya go.)

This was Scott’s response (I fixed a typo), and then I reproduce my reply below it:

1. NGDP growth slowed steadily. It was lower in the first half of 2008, then in 2006 and 2007. That slowdown in demand helped slow RGDP growth. Because of the slow growth in overall demand, the fall in housing construction (and autos due to high oil prices) was not fully offset by gains elsewhere. Hence RGDP was flat in the first half of 2008. Housing and autos declined, other output rose, but only about the same amount. Flat RGDP is a very mild recession.

2. I don’t understand why you think the rise in productivity solves the sticky-wage problem. The rise in productivity does help boost output, but it doesn’t put people back to work. If national income is barely higher than two years ago, and people who are employed make often make more than two years ago, and profits have risen due to productivity, then as a matter of sheer arithmatic won’t there be fewer people employed?

3. Follow up to previous example. NGDP is $10, and each of ten workers ge s a dollar in wages. Now they negotiate a 10% wage increase, in anticipation of 10% more NGDP. So wages rise to $1.10. Now there is only enough NGDP to emply nine workers, because NGDP unexpectedly stayed flat at $10. And that is true no matter how high productivity rises. Obviously the real world was more complex, but that’s what I see as the flaw in your logic.

And my reply:


Thanks for the replies. Quick responses:

On (1): So housing construction has nothing to do with it, right? If NGDP had grown at the same rate through early 2008, then RGDP wouldn’t have gone flat? So rather than having two explanations–a real one for the mild recession and a nominal one for the sharp recession–you have one explanation, I think. Specifically, NGDP slowed in early 2008, so there was a mild recession, and then when NGDP crashed, there was a bad recession. Right?

On (2) – (3): You are misunderstanding me. I said right in the beginning that rising productivity would hurt things, if one thought people needed to spend enough to “buy all the products” to keep everybody employed. But my point was, in that case, you should be rooting for CPI to fall, and yet you don’t.

Last point–not to Scott, but to people who thought I was an idiot for saying Scott would think rising productivity would hurt things–I told you so.

3 Responses to “Sumner Provides an Answer…Or Does He?”

  1. Captain_Freedom says:

    Now there is only enough NGDP to emply nine workers

    Although I am repeating myself, as I said this in the other post, Sumner here is displaying a very common and very wrong economic fallacy regarding employment and money expenditures. In this comment, he is implying that the total money spending for the economy’s output is the money that pays wages and thus employs workers. No no no no no! Wages are paid out of SAVINGS and CAPITAL, not consumer goods or capital goods spending.

    It is very easy to see this basic fact. Suppose that starting tomorrow, every single individual in the entire economic system decided to take every last dollar that they earned, and from then on, they spend all earnings on their own consumption. The wage earners would take their entire paychecks and spend it on their own consumption. Consumer goods company owners would take 100% of their sales revenues, declare the entirety of it a dividend, and then take that money and consume it all as well. This money that is spent on consumption will then be earned by other consumer goods companies, and then those consumer goods company owners would take 100% of those revenues, convert them to a dividend, and then turn around and spend that money on their own consumption as well. And so on throughout the entire economic system.

    What would happen?

    NGDP would skyrocket. If every individual in the economy took the entirety of their earnings and spent it on their own consumption, then the demand for capital goods would completely collapse, and, most relevantly, the demand for labor would collapse to ZERO. This is the case because if all company owners took 100% of their sales revenues and used the entirety of it on their own consumption, they would have no money left over to pay their workers. They would have to lay them all off. Each of the company owners throughout the economy would experience a gigantic consumption binge (just think of 100% of Wal-Mart’s revenues going to Wal-Mart owner’s consumption), but then as soon as they spent their money, they would have no way of maintaining their business.

    Every company owner would cease paying for electricity, land rent, materials replacement, factory repairs, machinery, equipment, anything and everything that requires making expenditures the purpose of which are not for consumption, but capital investment, would eventually wear out and rot. Transport trucks would break down. Factories would crumble. Fuel and energy would become so scarce as to be non-existent.

    So real productivity would drastically decline, unemployment would be so high as to be almost universal, save those who are able to live for a time off their savings and employ non-productive servants, the division of labor would collapse, civilization would collapse, a substantial percentage of the world’s population would die, and those who survived would support themselves as self-sufficient farmers and artisans.

    But, and this is what will probably baffle quasi-monetarists, NGDP would be massively high. Total spending for the crudest and most basic of goods that could be produced without the aid of capital goods spending and labor investment would be roughly in line with whatever quantity of money exists.

    So here’s the million dollar question. Since total NGDP is positive, indeed it would tend to be in the trillions of dollars if there existed trillions of dollars, WHERE is the employment? If NGDP “paid” wages, then why aren’t there any wages being paid here?

    The answer of course is that wage payments do not arise from consumer spending, or capital goods spending. Wages arise ONLY from a demand for labor and only labor. Wages come into existence with the onset of individuals who save and productivity expend their sales revenues rather than using the entirety of it on their own consumption, which is what almost everyone did in pre-capitalistic times. Before the onset of capitalism, there were product sales revenues, as the producers were self-sufficient producers, but there were little to no wage payments.

    Those product sales revenues earned the sellers 100% profit, because they had no money outlays and hence no money costs to deduct from those sales revenues.

    As productivity grew on account of respect for private property rights and economic freedom, more and more individuals found that they could save a larger portion of their product sales revenues. Those individuals who became more productive than others earned enough money and purchasing power to hire other individuals to help them. It was here that wages came into existence. And, since money costs of production now existed alongside product sales revenues, profits declined from 100% of sales revenues down to a smaller percentage, say 75% of revenues, since wage costs had to now start being deducted from sales revenues to calculate profits.

    Instead of the self-sufficient farmer consuming his $1000 in revenues, he only consumes say $750 and pays a worker $250 to help him, and that worker then spends that $250 on his own consumption. There was $1000 in sales before, and there are $1000 in sales after, but in the latter case there is $250 in wages paid and thus $250 in money costs to be deducted from $1000 in sales revenues. Profits go from $1000 to $750. But real productivity grew because there is a higher degree of division of labor and specialization.

    If there is one myth that needs to be put to rest in economics, it is the myth that aggregate “spending”, like NGDP, pay wages. NGDP does not pay wages. Wages come into existence due to savings and individuals making available money to demand labor.

    “The demand for commodities is not demand for labor” – John Stuart Mill.

  2. Scott Sumner says:

    Bob, I am saying the housing slump would reduce GDP, but only by a small amount. The slowdown in the first half of 2008 produced flat RGDP despite three negatives; a housing slump, an auto slump, and a slowdown in NGDP. With NGDP growing normally, the other two factors still would have produced a slowdown (perhaps1% annualized RGDP instead of zero) but not as much as actually happened. You seem to often want to pin me down to simple “it matters a lot or not at all” choices, which I reject.

    I understand the CPI can fall with productivity gains, but if NGDP is inadequate then because of sticky wages you will likely have more unemployment regardless of the change in prices and productivity. Of course that’s not a tautology, as your commenter points out, but it is very likely.

  3. JP Koning says:

    Bob, more generally, if you are interested in grappling with the monetary disequilibrium theorists, I’d suggest you read (or reread) Hutt’s “Rehabilitation of Say’s Law”. I view Hutt’s work as a theoretical effort to meet Clower, Yeager, Leijonhufvud and their progeny head on.

    Says Hutt: “To most economists trained in the Keynesian or neo-Keynesian tradition, which today means the overwhelming majority of the younger economists in the United States and Britain, such a proposition may at first appear almost intuitively as preposterously wrong. They will feel that, in the actual world, where there is indirect exchange and the use of money, “effective demand” may be insufficient for reasons other than pricing. Only under hypothetical barter, they may protest, can Say’s law hold. I have written this essay to refute such a view.”


    “Of more recent years Harry G. Johnson, R. W. Clower, Axel Leijonhufvud and Leland Yeager have continued the process of challenging the very foundations of the Keynesian system, in the process indirectly, if unintendedly, re-asserting Say’s law. Yet paradoxically these eminent theorists, and other economists of repute who accept their criticisms of, or changing attitudes towards, the Keynesian system, still seem to leave the impression that, after all, Say’s law does not work—at least not in the manner in which the old general equilibrium analysis suggested it did; and they suggest that in some way the world must feel grateful to Keynes, not so much for his contributions to economic method as for the beneficial policy consequences of his General Theory up to some unspecified turning-point some years ago… Their objections to the notion that supplies constitute the source of demands are always traceable, directly or indirectly (as I suggested before) to the surviving notion that the use of money somehow frustrates the operation of the market-clearing process. Under theoretical barter, the implication (or explicit assertion) is, Say’s law would apply. In a money economy it does not.”