« SRAS/SRPC; levels vs rates of change | Main | Bryan Caplan and Joseph Stiglitz »


Feed You can follow this conversation by subscribing to the comment feed for this post.

I would suggest starting with the unit elastic case, and then deal with the inelastic one.

Also, this is a case of "good deflation," and there is no need for an increase in nominal expenditure. Oddly enough, in the inelastic demand case, there is a possible argument for reducing aggregate demand. The demand for nongizmos is rising, and might outstrip productive capacity. Of course, as an advocate of nominal GDP targeting, I would say--don't worry about it.

When I read Stiglitz, I couldn't help but wonder if this isn't special interest propaganda. Is the goal maintaining the real wages of blue collar workers? If the UAW has to negotiate givebacks, who will buy life insurance, get haircuts, and the like?

If the farmers have less income, who will buy our cars? Seems like a great argument for the "farmer's party."

Sounds like a "lump-of-labor fallacy" to me. The catch is the problem has EVERYTHING to do with money, not "nothing". The central bank could fix it by adjusting interest rates IF that was all there was to it. But there's more to it than in your scenario. The more to it have to do with social costs, external economies and transaction costs. I'll have to have a look at Stiglitz later on and see if he really makes the mistake you and Ryan Avent suggest.

Bill: the unit elastic demand curve is simpler. But Stiglitz's "gizmo" is food, which is probably inelastic.

Whether you want to shift the AD curve does depend on which prices and/or wages are sticky.

Sandwichman. It does sound a little lump of labour fallacyish. But the demand for food probably is price-inelastic, and probably has a low income elasticity too, so it's very reasonable to assume that employment in agriculture would fall as a consequence of improved productivity in agriculture. And historically, that seems to have happened.

Yep, obviously I simplified Stiglitz's story, and left a lot of stuff out, but I think I have captured the key mechanism he outlines. See what you think.

It sounds very different from Arnold Kling's PSST story, for example.

"Sure, an improvement in productivity may cause deficient demand, but it does this by increasing Aggregate Supply, not by reducing Aggregate Demand. And the appropriate response by the monetary authorities would be to loosen monetary policy to increase AD by the same amount that AS has increased, in order to prevent deflation."

Remember what I said about a positive aggregate supply shock not being handled properly?

Represent AD by an rectangular hyperbola. Pronto, unit elastic. If AS shifts to the right from the increase in productivity, nothing has to be done to AD, just keep it growing as was. RGDP growth will rise and inflation fall. If the CB is targeting inflation it will cause instability...
Nobel winners feel they can talk and write about any "field"...A pity!

Nobel winners computer geeks on blog feel they can talk and write about any "field"...A pity!

A little self deprecating humour. I'll be quiet now.

I'm ambivalent about the Stiglitz piece. On the negative side, he credits government spending with "solving the economy's underlying problem." Christina Romer argued that it was monetary expansion generated by capital flight from Europe. I find Romer's explanation credible enough to have to be at least acknowledged by somebody peddling the "government spending (on armaments) ended the depression" line. Also, I think RA makes some valid points about the differences between consequences of the 1930s Fed response and the 21 century response. That said...

On the positive side, Stiglitz is telling a story that, in principle, cannot be modeled by the kind of highly abstract story. A structural change is not the same as a short period reallocation of incomes due to changes in productivity. This is what Marshall knew that economists today seem to not know. See his "Distribution and Exchange." And that is a lesson that Keynes very clearly learned from Marshall (but Pigou didn't) but that Keynesians seem to have not learned from Keynes.

Ultimately, I disagree with Stiglitz's prescription because I think his "roads, bridges and schools" notion of infrastructure is nostalgic and anachronistic. As Arlene Goldbard pointed out in an essay three months ago "infrastructure has another meaning, too." But I also disagree with Ryan Avent even more fundamentally. Macroeconomics hasn't "effectively solved the problem" of falling prices because falling prices were only a symptom. The pretense to "solving the problem" with rectangles and unit elasticity is even more wrong-headed. That's treating structural evolution a short period effect. Again, see Marshall and what Keynes learned from Marshall. Whatever it was that Keynes said about furiously bubbling ceteris paribus clauses, it echoes the following declaration by Marshall:

"But violence is required for keeping broad forces in the pound of Ceteris Paribus during, say, a whole generation, on the ground that they have only an indirect bearing on the question in hand. For even indirect influences may produce great effects in the course of a generation, if they happen to act cumulatively; and it is not safe to ignore them even provisionally in a practical problem without special study. Thus the uses of the statical method in problems relating to very long periods are dangerous; care and forethought and self restraint are needed at every step."

On Stiglitz and monetary economics: it's interesting that the criterion of easy money that Stiglitz uses (the monetary base) also suggests that monetary policy was loosened dramatically in the Great Depression. So why does he assume this is some revolutionary news? I shall apply the principle of charity and assume ignorance. He is, after all, a long way from home on this topic.

Stiglitz is mildly interesting for the fact he raises the issue of the restructuring of the North American economy from one focused on Big Enterprise who act like institutions to one focused on small businesses who believe themselves to be transient. I raised this on a previous thread and it is the same idea that Stiglitz has. GM is the poster child victim of this change and its management has been chronically unable to deal with it.

There is an underlying structural shift away from manufacturing to services that has changed the income-earning ability of households. There is also a shift away from "institutional" enterprise that attempts to provide a comprehensive welfare system as part of the employment package.

Change "technological change" to "business structure reform and bargaining power redistribution" and then you have something. Start with stagnant wages and then go from there.

The farm example was unilluminating. He also ignores the fact that the Great Depression struck Great Britain, France, Germany, Canada and other places. It got there by monetary links and was exacerbated by a global financial system still reeling from World War One.

Could someone please tell Mr. Stiglitz to look beyond American shores the next time he examines the Great Depression?

Stiglitz has done some important macro work (e.g. on the role of credit constraints and internal financing in the business cycle), so I wouldn't say that he's just a great microeconomist. But one thing that has become clear to me over the past few years is that very smart economists can sometimes say very foolish things, even about topics in which they're supposed to have expertise. This would appear to be one of those cases.

He is, of course, subject to the same defense as Fama -- that he had a reasonable model in mind but was simplifying it for non-economists in a way that made it seem nonsensical to other economists. It's plausible, perhaps, that, as a result of some shift in technological parameters, the natural real interest rate is so negative that no sensible monetary policy is consistent with equilibrium unless accompanied by the type of fiscal stimulus that Stiglitz advocates.

Nick, It's even worse than you suggest--you're being much too kind to Stiglitz. It's rare to see an article by a major economist who is wrong about almost EVERYTHING.

This is nothing new. See how Rogoff demolished Stiglitz's analysis of the East Asia crisis.

I did a post too.

Andy, The real interest rate was about 6% in 1929.


There is also some reason to think that part of the explanation of the US's problems lies with French monetary policy in 1928-1929.

I suppose explaining anything in the US economy from 2001-2011 using external factors is as unacceptable as thinking of monetary policy in terms of the money supply, NGDP and PT rather than the monetary base or interest rates.

Sorry to jump off topic for a moment, but I just read this article: http://www.theglobeandmail.com/globe-investor/personal-finance/household-finances/record-high-household-debt-in-canada-triggers-alarm/article2269210/

So a quick question for the Canadians in the crowd:

The article values the "debt to disposable income" ratio at around 150%, and says this is frighteningly high.

What I don't get is how this is measured. Is this a stock/flow calculation? I have a $50,000 NDI and a $100,000 mortgage, giving me a 200% D/DI number?

Or is it some derivative of a flow/flow number, like monthly debt service/income costs?

If it's the former, I don't understand why it's so bad. The case cited, for example, seems very manageable by most people. (given that debt service on a $100k mortgage is around $650 a month in the current rate environment.

Why the panic? What am I missing?

Paul: it's a stock divided by a flow. It has the units years. Why the panic? Well, it's an average. The average itself isn't so scary. It's what might be happening out at the tail end that might be scary.

All: my brain is running out of useful responses, and feeling guilty about exams that need marked. But thanks for the good comments.

Sandwichman said: "I'm ambivalent about the Stiglitz piece. On the negative side, he credits government spending with "solving the economy's underlying problem." Christina Romer argued that it was monetary expansion generated by capital flight from Europe ..."

Out in the real economy, what happened to supply during WWII?

Thanks Nick, that was my feeling exactly when I read the article. Seems like a bit of a silly metric.

Second point, it really, really grates to see good economists take the Great Depression in vain. Saying that "capital flows" caused the US to exit the Great Depression is wrong and ahistorical. Christina Romer needs to look beyond American shores (sigh).

Britain began to rearm in 1938 in a serious way; by 1940 a full command economy was implemented, the only time in memory a capitalist country has so greatly gone over to central control so completely. By 1940 the Great Depression was over in the UK, some argue for earlier but there was still widespread poverty.

Similarly Canada declared war on Germany on Sept. 9th 1939. The next item of business that Parliament considered was to pass a war budget. Before then the gross expenditures of the Government of Canada had been slightly less than $1 Billion. A $5 Billion program of military spending was passed that session. The federal government activated its reserve war powers under the BNA Act and took over the entire income tax system, placing provinces on an allowance. The entire tax system in Canada changed completely that autumn.

By 1940 the Great Depression was over in Canada, there was general expansion though it did not reach what we think of as full war production until 1941.

Crediting the end of the Great Depression to anything other than WWII is nonsense. I would also remind our American readers that the United States passed the Selective Service Act in 1940, the National Guard was placed on active duty and the US Army and Navy were on a war footing, if not yet what they would become, by Spring 1941.

Trying to explain capital flows in 1938-1941 in anything other than military terms is ahistorical. It was military keynesianism, no bones about it. I would also remind economists that you don't just flick a switch and get full war production. It took time in every country and that time was staggered due to events and politics.


I left out the other, more radical argument that the thing that brought an end to unemployment was not capital flows or arms spending but manpower mobilization. Without going into too much detail on the wages, working conditions and life expectancy of the enlisted personnel, it is clear that the elimination of unemployment resulted not from the creation of jobs but from the withdrawal of millions from the civilian labour force for an extended period of time. Meanwhile rationing and victory bond programs were enforcing a deferred consumption regime. People talk about government spending during the war as if that was the only thing happening. Conscription, wage and price controls, rationing, war bonds... those were just ceteris paribus details. Well... no....


"By 1940 the Great Depression was over in the UK"

If you're looking at GDP levels, the Great Depression was over by 1934. Crediting the end of the UK Great Depression to anything other than leaving the Gold Standard is nonsense; the most amusing example I've seen in print was a highschool textbook that attributed the recovery entirely to rearmanent, despite the fact that the UK saw near-record GDP growth in 1934 (BEFORE the 1935 election!).

Unemployment is a different story and cannot be sensibly discussed without reference to other factors during the 1930s e.g. the growth in the real value of unemployment benefits.

Incidentally, the World War II + aftermath was a period of stagnation in UK real GDP (about 1% growth PA) especially in comparison to the 1931-1939 period of recovery after the 1920-1931 period. Most of the NGDP growth during that period was due to rising prices.

(The period is full of interesting stuff: the Dalton experiment of 1946-1948; the hopeless mismanagement of the nationalised coal industry in the winter of 1947; the expansionary fiscal contraction under Stafford Cripps's excellent chancellorship; the dramatic rise in monetary velocity after its suppression in WWII etc. Britain was a horrible place to live in that time, but a lovely object of study for our time.)

Nick, to your credit at least you recognize that food may be a little 'special' because demand may be quite inelastic wrt price as well as income.

But this sentence here :

"If the price of gizmos falls by $1 each, and one million gizmos get sold each year, gizmo producers have $1 million less income, but non-gizmo producers now have an extra $1 million to spare which they can spend on something else."

What is that something else? What if that something else doesn't exist or is not easily accessible, in the aggregate? If you rule out that possibility, then haven't you just assumed your conclusion?

Just picking up on Ritwik's point, the assumption you make is that there is something else to spend on. In the long run, productivity growth causes consumption to shift to new products, not simply increased quantities of existing products. These new products do not create themselves. They need to be found via exploratory investment in product innovation.

Stiglitz's argument is simply a technological variant of Post-Keynesian theory where employment falls due to too much process innovation and too little product innovation. If you want to read an incredibly detailed version of this thesis as applied to the Great Depression, try Rick Szostak's book 'Technological innovation and the Great Depression' where he analyses empirical evidence that goes much beyond agriculture.

He also makes the point that a fall in nominal wages will make things worse in such a framework, again a point that many have made including Keynes and Tobin. It is also his emphasis on an investment-led recovery that puts him in firmly Post-Keynesian territory. I don't agree with his assertion that govt investment is the way forward and his thesis that we are making a transition from manufacturing to services is too simplistic. But the thesis that process innovation has accelerated and product innovation has stalled is entirely consistent with the 'Great Stagnation' thesis. In a Post-Keynesian reading, the increased household leverage of the last 30 years has simply postponed the inevitable structural problem.

I have written a much longer-form version of this argument applied to post-WW2 economic history from a more neo-Schumpeterian perspective here http://www.macroresilience.com/2011/11/02/innovation-stagnation-and-unemployment/

Ritwik: "What is that something else? What if that something else doesn't exist or is not easily accessible, in the aggregate? If you rule out that possibility, then haven't you just assumed your conclusion?"

No. See where I said: "Sure, if gizmo producers are very different from non-gizmo producers, and have different marginal propensities to hoard their income, the change in the distribution of income from the first to the second group might have aggregate effects on Aggregate Demand. But my theory made no mention of that."

If you want to assume that Joe Stiglitz was implicitly assuming that the marginal propensity to spend of non-gizmo producers is zero, then you can rescue his "theory". But Stiglitz doesn't say that. And it would be a bizarre assumption.

Ashwin: "Stiglitz's argument is simply a technological variant of Post-Keynesian theory where employment falls due to too much process innovation and too little product innovation."

Where does Stiglitz say that? Look, if that was his theory, why didn't he just say: "The improved technology in food production could have created an increase in real income, but people couldn't think of anything they wanted to buy, so they just saved their income instead, hoping that someone would invent some neat stuff they could buy in future".

It would then have been clear, and logical, and unbelievable.


Start with the unit elastic demand case, and then go to inelastic (or elastic.) I think it helps.

I saw that you initially assumed inelastic demand, which is plausible enough for food, and also necessary for Stiglitz's story.


Agiculture made up 20% of the economy. There is the other 80% of the economy. If no one has no use for any more of any of those goods, then you are assuming that their demands are all perfectly inelastic. Not very plausible.

While it is possible that rising real incomes could result in an increase in demand for leisure/decrease in the supply supply of labor, or an increase in the supply of saving, perhaps saving on the chance that new goods will be developed, this is not a likely explanation of what actually happened either in the thirties or today.

The reduced supply of labor and increased demand for leisure isn't a problem, though claims that this must be what is happening when employment falls and lower nominal expenditure on output is necessarily just an epiphenomenon is a political problem in my view. Those damn RBC theorists!

As for the too much saving argument--secular stagnation--either the monetary authority takes risk and so risky and long term interest rates fall or short and safe nominal interest rates go negative which means existing forms of hand-to-hand currency have to go, or higher trend inflation makes the real interest rate on currency more negative.

or if there is outside money, a sufficient low price level so that real balances are high enough that saving is low enough, that all of output is consumed.

or government can borrow from savers and spend on public goods. Coming up with excuses for that last possibility is the whole point of the argument, right?

debt to disposable income

This is a stupid metric, as it measures your ability to repay your debts by looking at your income less your debt repayments. If you own your home, like lots of Canadians, by definition your biggest outgoing will be mortgage repayment, and you're committed to paying that as a first charge on your net income so it's hardly disposable.

Incidentally, the World War II + aftermath was a period of stagnation in UK real GDP (about 1% growth PA) especially in comparison to the 1931-1939 period of recovery after the 1920-1931 period

Given the constraints - shortages of key inputs, huge foreign exchange problems, 3 million men in the army, Germans physically blowing stuff up - hardly surprising.

We got essentially a no-growth full employment equilibrium with very fast growth in certain industries (Robert Watson-Watt reckoned that something like 1 in 20 people were working in electronics, radio, or telecomms by 1945 - 1945! - and AJP Taylor that 20% of industrial production went into RAF Bomber Command alone) balanced by squashing the consumer sector, rolling up massive depreciation on the fabric, and repressing the shit out of finance.

Nick, how is it necessary in my argument or Ashwin's to have marginal propensity as zero? Simply lesser than marginal propensity for the food-producers, as you yourself remark.

But either way, this is still only the demand side of the analysis. What about the supply of desired goods? If I was buying 100 dollars of wheat last month, and next month I only need to buy 80 dollars of wheat because wheat prices have crashed, this doesn't mean that I will automatically buy something else from the other 20 dollars. I may not.

To a monetary theorist, this can perhaps be subsumed under the broader category of 'increase in money demand'. But it isn't *really* an increase in my *intrinsic* demand for money, whatever that is.

I don't know what the implication is - could it be that the problem does not go away by the central bank 'meeting' my money demand, but will only go away when the next-best-thing-since-wheat comes along?

Nick - I'm obviously giving Stiglitz the benefit of the doubt but given that he is being accused of literally not having a coherent model, I'd say that is fair. I'm assuming his views are Post-Keynesian simple because of his emphasis on public investment. This is classic Keynes, Minsky etc. - socialisation of investment and all that. And it's what distinguishes this viewpoint from say New Keynesianism.

You said "people couldn't think of anything they wanted to buy". That is not how new product innovation usually happens - entrepreneurs (like Steve Jobs) experiment with new product ideas that they think people will buy and some work while others don't. When you have a dynamically uncompetitive corporate sector without new firm entry and entrenched incumbents, then instead of disruptive product innovation we get the 'corporate savings glut'. Monetary policy cannot make a crony capitalist invest in risky product ideas.

Bill - I agree with you that an explanation that only relies on agriculture is very suspect. Again look up Rick Szostak's book which has a detailed analysis of how process innovation with no product innovation was the norm in much of manufacturing as well.

For what its worth, I think a significant proportion of the Great Depression was monetary/financial (gold standard +bank failures) but that doesn't mean other explanations are either incoherent or have no explanatory value. And I think this product/process innovation dichotomy has a significant explanatory role in our current problem of Great Stagnation + Automation/AI = Unemployment.

I should also point out that you can accept almost all the descriptive arguments made by Stiglitz and still oppose his call for more public investment. Where Post-Keynesians see a need for socialised investment, I see a need for creative destruction and allowing incumbent corporates to fail rather than propping them up (Just as Schumpeter did).

Alex: "This is a stupid metric, as it measures your ability to repay your debts by looking at your income less your debt repayments."

I'm not going to defend the metric, because I have my own serious doubts about it too. But "disposable income" normally means (and I think that's what it means here too) is "income minus taxes plus transfer payments from the government". It doesn't (normally, in macro) mean "income net of taxes and transfers minus debt service costs".

(But we are all a bit off-topic on this.)

Ashwin: I basically agree with what Bill said in response to you.

"Nick - I'm obviously giving Stiglitz the benefit of the doubt..."

OK. Understood. I think you are being very generous in your interpretation.

Empirically though, wasn't the 1930's a time of great product innovation in the US? (I'm not sure about that, and only vaguely remember reading it somewhere, perhaps by Tyler Cowen.)

If I wanted to defend Stiglitz, I would say: the productivity improvements in agriculture were a really big deal, and should be looked at anyway, regardless of whether they caused the Depression; when you get a real process like that happening, it's going to have implications for structural unemployment, etc., regardless of what is happening to money and finance and AD.

Ritwik: "Nick, how is it necessary in my argument or Ashwin's to have marginal propensity as zero? Simply lesser than marginal propensity for the food-producers, as you yourself remark."

OK. But Stiglitz completely ignores (100% ignores) the effect of the lower food prices on manufacturers incomes. Which you can only do if their marginal propensity to spend is zero.

You lost me a bit on the rest of your comment.

Scott, but the real interest rate now is negative. If one takes the view that a ridiculously low real interest rate is required to cure a depression, the early 1930's don't provide evidence one way or the other, because the solution wasn't tried. (Arguably it was tried later in the 30's and it's debatable whether it worked.) This time around we brought real interest rates down and it was insufficient to cure the depression. So, in this view, we now know that monetary policy (if one views it as cutting real interest rates and considers it to be limited to non-ridiculous real interest rates) wouldn't have worked in the 30's either. (I'm not saying I agree with this view, but I think it's tenable, and with a stretch it could be an interpretation of Stiglitz.)


You're being very generous to Stiglitz here. His article clearly suggesting that a further loosening of monetary policy would have zero beneficial impact. He also implies that aggressive monetary loosening in 1931 would have also had no positive impact.
Then later in the article he outlines the destructive process of a deflationary spiral. But he doesn’t make the connection between the aggregate price level and the policy of the central bank. I’m forced to conclude that he believes that the central bank can’t influence the price level. Or that he simply hasn’t thought it through.

This is the worst article on macroeconomics written by a major economist that I have ever read.

Nick - On productivity growth and innovation in the 30s, the definitive work is that by Alexander Field who, as you say, argues that innovation of all sorts was high during the 30s. I'm not really qualified to pass judgement on this topic but I find it reasonably convincing and this is the reason why I am not convinced by Stiglitz's simplistic argument or Rick Szostak's more nuanced argument.
So I don't even agree with Stiglitz on the empirics but all I am arguing is that there is a coherent macroeconomic model where his chain of thought makes sense. For what its worth, I think its an argument that has relevance to our current situation though again Stiglitz's argument of a manufacturing to services transition is way too simplistic.

And just chiming in on Andy Harless' point, it is a big part of why I think such an explanation has more legs today than in 1931. Real rates are negative today whereas FDR's devaluation was done in an environment of high positive real rates when clearly there was a significant monetary free lunch available.

"...really bad macroeconomics. God it's depressing."

Didn't you say something similar about Frank Hahn? Just accept it: there are people who don't agree with you.

W. Peden:

Canada didn't have conscription for overseas service and we still exited the Depression by 1940.

Full GDP recovery without full employment recovery is a red herring recovery. It isn't a recovery.

Kevin: I actually did a search, to see what I said about Hahn.

I disagreed with Hahn over some very esoteric stuff concerning Say's Law. I think Hahn was wrong. Most macroeconomists would agree with Hahn on that point, not me. I think they are all wrong! But they are not *obviously* wrong.

No, it's not like that with Stiglitz. Stiglitz is Intro Macro wrong. Income = output = expenditure wrong. Like watching someone get muddled over the Keynesian Cross. He just doesn't get macro.

Nick this is the comment I wrote at Diary of a Republican Hater:

"None of the above makes any sense. You just can't do macro like that. It all goes wrong from the very beginning. The fall in the price of gizmos, for a given quantity of gizmos sold, and for given prices of non-gizmos, reduces the real incomes of gizmo producers, but increases the real incomes of non-gizmo producers by an equal amount. If the price of gizmos falls by $1 each, and one million gizmos get sold each year, gizmo producers have $1 million less income, but non-gizmo producers now have an extra $1 million to spare which they can spend on something else."

As I understand Stiglitz, his point is that due to improved productivity of "gizmos" if you will, the gizmo producers no longer need to hire anywhere near the level of previous labor in the gizmo producing industries.

As there is nowhere for all these displaced gizmo laboreres-let's say there are millions of such workers-their loss of income depresses demand. By this depressed demand, other producers in non-gizmo industries see their sales depressed, leading them to lay off millions from their industries, continuing the viscious circle as this further depresses demand.

Mass unemployment depresses demand as I would understand it and this is his point.

The real point is that increased productiivty reduces the need for labor. This is where the depressed demand comes from.

Incidentally Nick I'm still in my happy just to be here mode-I'm stoked that you have checked in again, as I was when Selgin dropped by. Sumner till now pretends that he's too cool to come here.

If my above analysis in this comment is wrong let me know where please. Preferrably in time for my next post about your comment-LOL.

For the post Nick commented on please see http://diaryofarepublicanhater.blogspot.com/2011/12/scott-sumner-faces-off-with-joe.html?showComment=1323909635872#c7805825661205385219

I can be honest enough to admit that I'm not wholly sure who is right but I'd like to figure it out.


If 100 gizmos are produced and sold each year, that creates 100 gizmo's worth of income for those who produced and sold those gizmos. It doesn't make any difference if they were produced by 100 workers, or by 1 very productive worker, with the other 99 being unemployed.

An increase in productivity makes no difference to the income earned from producing gizmos, *measured in gizmos*, provided the total quantity of gizmos produced and sold stays the same.

Now, an increase in productivity will probably reduce the price of gizmos, relative to non-gizmos. If so, the income from producing gizmos, *measured in terms of the non-gizmos it will buy*, will fall. But the income from producing non-gizmos, *measured in terms of the gizmos it will buy*, will rise by the same amount.

One of the first things we explain in Intro Macroeconomics is that the production and sale of goods creates income equal to the value of the goods produced and sold. It's basic National income Accounting. And Joseph Stiglitz totally blew it, Nobel and all.

If Stiglitz were a right wing economist, arguing against fiscal policy, making a basic mistake like this, Brad DeLong would have crucified him by now.

Ok thanks for the answer Nick. I'm gonna chew it over a bit. I'm also going to import your answer to the coment section of Diary hope that's all right. LOL

Nick: "One of the first things we explain in Intro Macroeconomics is that the production and sale of goods creates income equal to the value of the goods produced and sold."

and sale... and sold... so, assuming that markets clear... in a depression, not all goods produced can be sold. That's what a depression is about.

Sandwichman: I'm not assuming markets clear.

Take a simple Keynesian Cross model, for example. Y=C+I, C=a+bY, so Y=(1/(1-b))(a+I). Now assume productivity doubles, but a and I stay the same. Y stays the same, but employment halves. AS doubles, there's excess supply, and non-market clearing, but AD and Y stay the same. Stiglitz is saying that an increase in productivity causes Y to fall. Why? Because it causes unemployment which causes Y to fall. Which is just wrong.

Ashwin and Ritwik above recognised this point, and were trying to figure out some other model that Stiglitz might have had at the back of his mind, which is fair enough.

The only other possibility I could see would be if gizmos were half the economy and fell by half, a really big sharp shock in other words. The non-gizmo economy would try to expand but face such a quick and large drop that instead of basking in its gizmo richness becomes fearful they will be next and begin hoarding making it self fulfilling. An excess money demand story but one the central bank has difficulty in countering, perhaps even intensifying the fear with dramatic action. Even as fear subsides the wrenching change makes recovery slow and difficult.


"Full GDP recovery without full employment recovery is a red herring recovery. It isn't a recovery."

Without an observable definition of "full employment" that allows for structural changes, I don't see how such a proposition is even remotely plausible.

Holy crap I need to go back to intro econ.

Why is this wrong?

Assume a world of just gizmos. Gizmo boss hires gizmo worker to help him make food (okay so gizmos in my world are food), but it is produced in quantities beyond which Gizmo boss needs. Gizmo boss pays gizmo worker for his help. Gizmo worker pays Gizmo boss for food. Gizmo boss then increases his productivity by discovering a way to make enough food for himself without hiring gizmo worker. Gizmo boss now generates less money income (actually no income, he is no longer receiving money from gizmo worker, and it's under his pillow) and produces less product.

Isn't there a story about Henry Ford paying his workers enough to buy his cars. Is this a flawed macro story?

Here's another probably incredibly stupid anecdote.

A hires B. C hires D.

A and B need one each of what C and D make. A and B pay $2 total to C and D to get 2 of what C and D make. C and D pay that $2 back to A and B to get their stuff, one each as well.

A no longer hires B, but C and D send over the $2 for 2 of what A now makes. But A only needs one of what C and D make. He pays them $1. C now no longer hires D. C sends that $1 back to A for A's stuff. Then A to C, etc etc. A sleeps under $1 permanently. Y down and income down?

wiwsd: "Isn't there a story about Henry Ford paying his workers enough to buy his cars. Is this a flawed macro story?"

Yes, it's an extremely flawed macro story.

1. Ford could have paid his workers $1 less, which would have given him $1 more income, which he could have spent himself.

2. Ford is small relative to the US economy as a whole. The mpc of Ford workers to spend on Ford cars is small.

3. If Ford had really wanted his workers to spend their high wages on Ford cars, he would have cut their wages and paid them partly in cars. (You get partly the same thing when Ford workers get discounts on Ford cars, which we actually observe).

4. It's historically false. The reason Ford said he paid high wages was an efficiency wage story. He needed high quality workers who were scared of losing their high wage jobs. Because a screw up by one worker halted the whole production line. (I read this on an Austrian blog somewhere about a year ago, maybe Bob Murphy's?)

And yet, I seem to hear that totally flawed story about Ford wanting to pay high wages to increase AD about every couple of months.

wiwsd: Food story:

1. Remember, if I grow tomatoes in my garden, and eat them myself, they should strictly be included as part of my income. Real income = goods produced, even if i sell them to myself.

2. Let me re-tell your story. Barter economy. Food is the only good. Output of food = national income. Food needs land and labour. There are landowners and workers. Landowners pay workers their Marginal Product in food. Then there's an improvement in technology that increases output for any given quantity of land and labour, but also reduces the Marginal Product of Labour. (Draw the production function, with Y on the vertical, Labour on the horizontal, shift it up, but make it flatter).

2a. Non-satiation. Full employment, wage income falls, but total income (Y) rises).

2b. Assume landowner is satiated. There is now free land and free labour. But the landlord loses nothing by letting the worker use the free land. And if he does this, Y rises. Unless the landlord likes to go hunting on the free land. But now there are two goods, not one.

2c. Nevertheless, real wages can fall below subsistence. This is a model of the highland clearances.

2d. This model is not a Keynesian deficient demand model. It's a model of technological unemployment. It's like my old post about men and horses:


wiwsd: let me tell you a simpler version: suppose I own all the land in the world, and I'm satiated. Then, on a whim, I can either let other people use my land to produce goods, or keep some of my land idle. I don't care one way or the other. But total output is lower if I keep some of my land idle.

(I wish everyone were as "dumb" as you, in wanting to think this stuff through.)

Revised: suppose I just own *some* land, and am satiated. I can choose to keep my land idle, or let other people use it, on a whim. (I don't need to own all the land in the world for my model to work). But then if I have a tiny ounce of charity, or just like a nice smile from the grateful crofters, I will let them use it.

Even simpler. Forget land and trade and anything. Assume I'm satiated. I can either produce lots of goods and give them away, or just produce enough for my own needs. So, on my whim, GDP either rises or falls.

Nick: You make a powerful case for a land value tax. Good idea!

K: it's an equally powerful case for a labour value tax. All depends on what "L" stands for!

You mean assume Labour is satiated? Good one, Nick!

K: Yes. Think about it. Ever heard of people earning very high incomes from selling their labour? Watch TV sometime.

I am intentionally being very generous to Stiglitz (in a devil's advocate sort of way), just as some were very generous to Fama when he said things that similarly seemed to show a lack of knowledge of macroeconomics. I think you overstate the case against him, however. If the monetary policy necessary to fix a depression is so extreme as to be ridiculous (for example, if you have to set a 30% annual inflation target and make massive quantities of loans to entities that are know to be likely to default), then, for practical purposes, one can simply say that monetary policy won't work. The range of policies that are considered even vaguely reasonable may not include anything that will have a material impact on output.

It seems to me that Stiglitz and his critics are employing fundamentally different kinds of models and the critics are assuming that Stiglitz is or should be using the same kind of model as theirs. Stiglitz is trying to conceive of an explanation for things that have occurred. This is what in Marshallian terms would be called a "realistic investigation." His critics are countering with a projective model of possibility that contains the implicit assumption that if A is better than B and can happen it will happen.

Stiglitz errs in assuming that the same retrospective model can be used to explain the 1930s and the 2000s. I think he also needed to have gathered a lot more facts than he did. The critics make the same errors to an ever greater extent. They have both missed the "third dimension."

The third dimension? From Sydney Chapman's autobiography:

I had no notion of writing just a descriptive book: I wanted to be discovering something. And soon I found myself turned to reading again. This arose in the following way. In the course of my talks I would ask "why this and why that?" and frequently in his reply my informant would say, "when my father started fifty years ago it used to be so and so," or, "my grandfather used to tell me that in his early days this and that happened." So it became clear to me that my study must be given a third dimension. At first I had thought of taking the industry as it was and tracing the laws running across its surface, as it were. But evidently this would not do. Forces which could not be ignored were also operating from the back. The industry had to be viewed as a solid in time and as a changing one. So its history had to be gone into. Not history as Unwin liked to do it, when be got drawn back and back, even to the middle ages, and lost all thought of the present in tracing a trade or its forerunner from country to country under the guidance of St. Nicholaw. My history was to be much more modest; merely enough to keep the object of my study a self-contained whole and not a section artificially cut from the body of which it was a part.

Nick: yes Ford's story is partly about efficiency wage, a totally micro concern. But it also was about fear of concentrating wealth-income in such a way that that the mass demand needed for his mass production could't be realised, a partly macro worry.

Ford's story was an analogy and public relations spin. To criticize its literal "falsehood" is to miss the point, which highlights A truth despite its deviation from The Truth.

Stiglitz is Intro Macro wrong. Income = output = expenditure wrong. Like watching someone get muddled over the Keynesian Cross. He just doesn't get macro.

I think this is too harsh. I think Stiglitz understands Intro Macro - he just thinks that those kinds of high-level accounting identities obscure the important factors in explaining something like the Great Depression. There's a longer paper he wrote that spells out his thinking in more detail: "Rethinking Macroeconomics: What Failed and How to Repair It," Journal of the European Economic Association, 9(4), pp. 591-645.


Here's a relevant quote:

"But more generally, distribution matters: if prices of agricultural goods fall rapidly, farmers reduce their spending by more than urban workers and rentiers increase their spending. Aggregate demand thus falls. More generally, with both supply and demand concave functions of firm equity, there are real, and potentially large, consequences to such redistributions."

o. nate:
In other words, Stiglitz doesn't think macro arises onlt because of money but because of frictions an sdelayed reactions?

Nick: "K: Yes. Think about it. Ever heard of people earning very high incomes from selling their labour? Watch TV sometime."

Ok, sure. I'm quite familiar with that kind of labour. I guess I lumped it more together with Land (rent collecting), though of course it's not 100% like that (I suppose Steve Jobs is the canonical counterexample).

Oh, maybe all true. BUT "non-gizmo producers now have an extra $1 million to spare which they can spend on something else:" people aren't spending that money because they are being hit in so many other ways and are frightened.

"You just can't do macro like that. . . The fall in the price of gizmos, for a given quantity of gizmos sold, and for given prices of non-gizmos, reduces the real incomes of gizmo producers, but increases the real incomes of non-gizmo producers by an equal amount."

Um....suppose you have an open economy? one that exports lots of gizmos? so that the fall in P_Gizmo is a bad terms of trade shock? or put another way, suppose the benefits of cheaper gizmos mostly accrue to foreigners?

What you're describing as a nonsense sounds like a textbook case of immiserizing growth; Harry Johnson and Bhagwati worked it out in general equilibrium before you or I were born, Nick. Now, you can argue that it is very unlikely or unrealistic, if you like. But it is very, very logical.

(Exercise for the grad students: suppose the gizmo-exporting country now has a large foreign debt denominated in foreign currency. Explain whether the change in the country's wealth is now larger or smaller.)

Simon: but the standard terms of trade effect says that the manufacturing sector will have higher real income. Stiglitz says the manufacturing sector has lower real income. He doesn't mention exports or foreign debt. His is a closed economy analysis, as far as I can see.

All of this blackboard economics is making me very sad. Do any of these ideas make any sense if we do back of the envelope calculations over their magnitudes?

Chris: well, my back of the envelope calculation would give me a rough magnitude of 0!

Let's work backwards. Suppose you wanted a 30% decline in GDP. Divide it by the Keynesian multiplier, say 3, to get a required 10% decline in autonomous expenditure. Now suppose farm income is say 30% of GDP, and it declines by 50%, and the difference between farmers' and non-farmers' mpc is (say) 0.666. That should do it. No, I don't believe those numbers.


Damn. Now you're forcing me to read Stiglitz's article. I'm still working my way through, but I have a hard time thinking of the early 2000s experience that Stiglitz highlights in a closed economy setting.

"The American standard of living was sustained only by rising debt—debt so large that the U.S. savings rate had dropped to near zero."
"The fact is the economy in the years before the current crisis was fundamentally weak, with the bubble, and the unsustainable consumption to which it gave rise, acting as life support."

He's talking about a zero savings rate but not a lack of investment. We both know the reason for that; persistent and large capita inflows a.k.a. current account deficits and mounting foreign indebtedness.

When you say Stiglitz is making a closed-economy argument, is that because of what Stiglitz says? or is that your assumption?

Simon: Make sure to read the JEEA paper if you really want to know what he is talking about. It's full of ideas and far more clearly laid out (to an economist) than the Vanity Fair piece.

Simon: "When you say Stiglitz is making a closed-economy argument, is that because of what Stiglitz says? or is that your assumption?"

Because, IIRC, he never mentions anything to do with foreigners, exports, imports, capital flows, etc.

K is right.

But the link I gave doesn't work. here it is.

In other words, Stiglitz doesn't think macro arises onlt because of money but because of frictions an sdelayed reactions?

I don't want to put words in Stiglitz's mouth. He can explain himself better than I can, so I'd recommend reading the full paper. But if you want my very short summary, he basically highlights the importance of things like credit, expectations, and firm equity in amplifying shocks.

Do you think you might be taking Stiglitz too literally, like he doesn't literally mean monetary measures can't help at all. He just means that the big root of this is a huge structural shift, and if you don't address it, don't speed the transition, then you'll just keep getting this drip, drip, drip, over decades.

It's like monetary policy helped in the 2000's, but you had to keep doing it for years and years. You didn't really address the cause of the constant drip, drip, drip. You didn't actually re-surface your roof, you just kept bailing with buckets. Sure it had an effect, it kept your house from molding and rotting, but it would be a lot better to really get the roof in good shape.

Certainly I strongly prefer Stiglitz's solutions. These are really high return investments of the kind the pure free market will grossly underprovide due to externalities, asymmetric information, coordination problems,…

Growth economists Charles I. Jones of Stanford and John C. Williams of the San Francisco Fed wrote in a 1998 QJE paper:

Is there too much or too little research and development (R&D? In this paper we bridge the gap between the recent growth literature and the empirical productivity literature. We derive in a growth model the relationship between the social rate of return to R&D and the coefficient estimates of the empirical literature and show that these estimates represent a lower bound. Furthermore, our analytic framework provides a direct mapping from the rate of return to the degree of underinvestment in research. Conservative estimates suggest that optimal R&D investment is at least two to four times actual investment [emphasis added].

– Vol. 113, No. 4 (Nov., 1998), pp. 1119-1135

the monks chase each others assumed tails/tales

two words
open and closed

two nominal quantities

output value debt service value

final ponder

credit is rationed like spots in the columbia college freshman class

not like oranges at the korean fruit and veg stand on broadway and 111 st

price of orange <=> price of money
nope interest is not the price of credit
like orange prices
unless the fuit and veg guy decides if you are qualified to buy her oranges

stig is trying to tell the chattering literate class a story

a vague underlying parallel between great D and great R
its a just so story
that is not a logic boxed story like our dear nick
sunny abbot of this monastic establishment

TV stats. I hope they are correct, and Nick finds them useful.


TV Prices Fall, Squeezing Most Makers and Sellers

"There were roughly 32 million television sets sold in North America in 2004, for an average cost of $400, Mr. Gagnon said. The average size of a television was 27 inches. Today, 44 million sets are sold a year in North America, with an average cost of $460 and an average size of 38 inches.

Consumers buy a new television set every seven years or so, and an average household owns 2.8 TVs, he said. While those numbers would suggest a bonanza for television manufacturers, Mr. Gagnon said the larger, more sophisticated sets were expensive to manufacture and cut into manufacturers' profit margins.

To help reduce costs, manufacturers invested heavily in sophisticated new factories or retrofitted old ones that were capable of cranking out more televisions at lower cost. The problem is that the factories became operational about the time the recession hit, creating a glut of televisions and forcing prices down.

A strong yen, relative to the dollar, has further hurt Japanese manufacturers like Sony and Panasonic, while Korean manufacturers like Samsung have benefited from a weak won."

And, ""Everybody is fighting for a limited amount of consumer dollars," said Gregg Richard, president of PC Richard and Son, which has 66 electronics and appliance stores. "We are selling more TVs, more units, at lower retail prices."

It does not help that consumers are reluctant to pay much more for the latest features, like 3-D and Internet connectivity. Instead, they are likely to wait patiently for a few months until the price inevitably comes down.

"People used to pay additional to get a Sony Trinitron," said Riddhi Patel, director of television systems at IHS iSuppli, a market research firm. "But the industry has trained the consumer that any time there is a new technology, if they wait six months the price will come down."

Paul Gagnon, director of North America TV research for DisplaySearch, which tracks the market, noted that a 60-inch LCD television by Sharp was now selling for as little as $799 -- about half of what it was selling just a year ago. "Absolutely amazing," he said.

The slump is a hangover of sorts for an industry that binged on years of double-digit growth, as consumers rushed to replace old television sets with flashy new models with new features like high definition and flat screens."

And, "It's a great time to buy a television, and Ram Lall, a television salesman, isn't happy about it. In a basement showroom of J&R, the huge electronics store in Lower Manhattan, Mr. Lall says the days of making big money from televisions are in the past. Pointing to a top-of-the line, 55-inch Sony television, Mr. Lall said it would have sold for $6,000 a few years ago. The current price? $2,599.

"We are making less money because the company is forcing us to slash prices," Mr. Lall said, standing amid rows of flickering television sets.

Televisions have become so inexpensive that the profits have largely been squeezed out of them, a result of a huge increase in manufacturing capacity that has led to an oversupply and continued downward pressure on prices from low-cost manufacturers and online retailers.

The near fire-sale prices are great for consumers, who can now buy a television for a fraction of what one cost just a few years ago.

But what is good news for consumers has been a nightmare for manufacturers of TVs and retailers that sell them. The earnings of mainstay television manufacturers like Panasonic, Toshiba and Sony have been hammered. Sony, for instance, is overhauling its television operations because of what one executive said recently was a "grave sense of crisis that we have continued to post losses in TVs." Even newer and more nimble competitors like Samsung and LG have struggled to make much money on TVs, if any.

Seeking to stanch its losses, Sony on Monday said it would end its flat-panel joint venture with Samsung, which was set up in 2004 to capture the boom in televisions with liquid-crystal displays. Samsung, based in South Korea, will pay about $940 million for Tokyo-based Sony's 50 percent stake; Sony aims to save on manufacturing costs while still buying panels from Samsung.

For retailers, the picture is not much better. This month, Best Buy reported a 29 percent drop in net income for the third quarter, in part because the retail chain had slashed prices on televisions and other electronics.

Perhaps even more ominously for the long term, the future of televisions appears to be more about what content they can provide, like Netflix and iTunes, than new hardware features like flat screens or 3-D technology. It is an area where television manufacturers have struggled with little success to get an edge, even as Apple and Google vow to upend the industry."


spam filter, someone?

TMF: sorry. Found it in spam filter.

Thanks, Nick!

I'm trying to figure out if the average cost of $400 to average cost of $460 is costs or ASP's (average selling prices).

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad