Tuesday, October 2, 2012

Clonal Antibody — Corrected post on US industrial production

I need to make some corrections to the previous post - as pointed out by Edmund in the comments. This gives a new interpretation to the graphs. My mistake in not digging deeper.



my mistake. The Industrial production index is a real index. So my conclusion of declining Industrial might was wrong.

However, what it does show is that real per capita production increased linearly, dropping only in the recessions, with the 2008-20010 drop being the steepest. See

Further, dividing by the price index, it shows that price rises contributed smaller and smaller amounts to increasing real per capita production.

In other words, prices are increasing exponentially while real production is increasing more or less linearly See

You can clearly see the inflationary 70's. So in summing, the graphs in the post show periods of inflation when the sharp declines occur - and of course they coincide with the oil shocks.

But the different growth rates of production and prices points to the need of increasing quantities of money. This can come endogenously through bank credit, or exogenously through government deficit spending. Endogenous money, because of the need to repay debt and interest, will cause instabilities in the system as shown by Minsky and Keen. Government deficit spending avoids that problem, and is much less likely to cause instability in the system.


Anonymous said...


Further, dividing by the price index, it shows that price rises contributed smaller and smaller amounts to increasing real per capita production.

Is this sentence correct? The graph seems to the series for production divided by the population, not the price index.

I'm also a bit puzzled by the second half--if the series is a real measure of production, then prices are being held constant, so what do you mean by, "it shows that price rises contributed smaller and smaller amounts"?

Clonal said...


What I was attempting to say, was that even though real production was increasing, each dollar was buying less of that real production. This implies a need for a larger money supply, or an increasing velocity of money. The increasing money can come from two sources increasing private debt, or from government deficits. Increased money velocity comes via reduced savings.

We see all of these three factors coming into play.

Further, we have seen that from 1980, real hourly wage has remained stagnant, while returns to the FIRE sector have increased. The implication of this is that most of the increases in real production have gone to the beneficiaries of the FIRE sector, and not to those who work for a living.

So another way to look at the original graphs is that a decline in the real per capita production to price ratio is a shift from the worker to the rentier.

If real wages had kept up with increase in real production, this would not be an issue. But that has not been the case.

The other issue which is very hard to glean from the data series, is the issue of transfer pricing. The real production index is based on a "value added" basis. In other words, if I in the US were to import raw material very cheaply, whose price was low because of very low wages, and my marginal costs being much higher in the US, most of the value added would be accruing to my manufacturing sector, and not to the manufacturing sector of the exporting country, event though the bulk of the effort in manufacturing the product has been done spent in the exporting country. This again, I believe is a result of the financialization of the economy.

Tom Hickey said...

There seem to be several factors involved. First, rate of profit on industrial investment began to fall below that of financial investment, so finance began to replace industry as the economic driver.

Secondly, labor share fell relative to top managers' and owners' share for a number of reasons related to decreasing labor bargaining power, as well as the expansion of consumer credit to fill the gap that was opening up so that labor was able to be co-opted into acquiescence with cheap imports as their jobs were exported, and newly abundant credit.

This began the run up in private debt that culminated in the crisis, as the financial sector overreached and the late-stage financial cycle of Ponzi finance took hold, finally imploding. That's where we are now and no one in charge knows how to put Humpty Dumpty back together again.

Bob Roddis said...

The transfer of wealth from average people to the financial elite comes about because when fiat funny money is created, there is a transfer of purchasing power to the first receiver of such money (generally the elite and wealthy) from those holding or being paid in the existing money (the working class and poor). Like all things "progressive", the cause of society's problems are "progressive" solutions to "problems" that do not otherwise exist but for the "progressive" intrusion.

Anonymous said...


Perhaps I’m still misunderstanding you. The idea behind time series of real variables is that prices are held constant for the duration (or removed), so that we don’t need to worry about changes in the value of the dollar. In other words, if a dollar buys you X amount of stuff in 1970, it buys you X amount of the same stuff in 1990 as well.

Over the same period, you can see from your CPI series that the general level of prices is also rising. What this means essentially is that it takes more dollars to buy the same amount of everything, or the price of everything is going up over time. Naturally this includes industrial output.

So, since the price of everything is going up, it seems reasonable to ask whether it’s getting harder and harder for consumers (or households or workers or whatever) to buy what it is they need. Wages are also a price and so might be expected to rise with the price level. In addition, wages might rise as the economy grows, becomes more productive, and so on. You can create series for real wages (i.e., wages in constant dollars) to test this out. If nominal wages are rising slower than the price level, then the real wage will be falling.

Does the trend in the price level mean that a larger money supply is needed to purchase all of the output for sale? From a strict quantity theory point of view, the larger money supply (or perhaps, as you note, greater velocity) caused the rise in the price level. If you epand the supply of money in response to inflation, you will continuously chase the price level further and further upwards.

In any case, it is not money per se that consumers or workers need, but income. If my income increases, then I can afford to buy more stuff with it. Most importantly, as we’ve seen, my real income needs to increase, because if my income increases less than prices in general, I’m not getting better off.

With all that said, I think this is a cool post. There’s really no substitute for digging through the data to find out what’s really going on. It would be great if there was a substitute, but unfortunately, there just ain’t.

Clonal said...


My own analysis of the data, looking at the money supply seems to indicate that money chases prices, and not the other way around - at least over the last 70 or so years. The sudden burst of price rises appear to be related to supply shocks (e.g. OPEC oil embargo, Iran oil crisis, or war related supply shocks)

Money supply changes seem to occur in response to price changes. That would make sense because of Federally mandated COLA adjustments to incomes as well as the creation of endogenous money through loans to cover increasing cost of raw materials. If prices increase, I take on new loans, and if they decrease, I pay off old loans.

Anonymous said...


That’s an interesting idea. It’s very counter-intuitive, though, from the point of view that people usually take when they look at this problem.

It’s counter-intuitive for two reasons: (1), it leaves open the question of why the price level is rising in the first place, and (2), it’s hard to see why money should be chasing prices rather than vice-versa.

You provide answers for both of these in your comment, but I’m not sure how plausible they are.

It’s easy to see why a supply shock would raise input prices and get passed through to the price level. But these are one-off shocks, so they can’t explain the general trend in the price level over time, which is to rise.

Then there’s money. If we get a certain amount of inflation, it doesn’t seem obvious why it would cause consumers or workers to take out more loans. Since prices are rising everywhere, the ability of workers and firms to service whatever their increased costs are is also increasing.

Your story also makes the actions of the Fed look rather strange. The Fed has a particular inflation target. If inflation is above this target, the Fed tightens monetary policy by raising interest rates. If inflation is below, it tightens by powering rates. But if your story is correct, then the Fed should look and see inflation above target and in response lower rates in order to try to stimulate bank lending.

This means that the Fed would need to have a very different model in mind to the one most people think it has, because most people think that it operates according to a generic story whereby, if the economy overheats (i.e., is above capacity or potential), inflation increases and the Fed needs to raise rates in order to cool it off. If the Fed is going to be lowering rates when inflation is higher and raising rates when inflation is lower, then it must have some alternative theory that helps it to make sense of its own actions.

I do find the idea interesting, though, and it would be cool to see more. Do you have a blog? How did you test it—some kind of VAR plus policy shock?

Anonymous said...

Does the trend in the price level mean that a larger money supply is needed to purchase all of the output for sale? From a strict quantity theory point of view, the larger money supply (or perhaps, as you note, greater velocity) caused the rise in the price level.

BTW, I meant to add: It’s not necessary to buy into the QTM to arrive at a similar conclusion. The point is that inflation means that every price is rising, not just my costs. I shouldn’t need to take out a loan or increase my borrowing. (On the other hand, it might be the case that my real—i.e., inflation adjusted—wage is falling or rising slowly, for whatever reason, which seems like it could cause me to increase my borrowing.)

paul meli said...
This comment has been removed by the author.
Clonal said...
This comment has been removed by the author.
Clonal said...

One thing that is not clear to the people espousing the "Quantity Theory of Money" is that it is a shortage of any item that leads to a price rise, and not an excess of money. I do not know of anybody with. an excess of money who would pay more for that item just because they had a jingle or a jangle in their pocket. They may not bargain as hard, but they will not pay more than the asking price. They will not even pay the asking price if another similar item was available at arms length more cheaply. Excess money will always get saved. It will not be spent gratuitously.

QE1, 2, 3 have shown that well. An excess of money does not lead to a shortage of goods and hence inflation. If there is more demand because people have more income, the producers will increase their output, and not their price (unless there is a shortage of raw materials and labor - in other words, the economy is using all available resources and/or there are bottleneck resources.) An excess of money leads to the excess money being saved.

All inflationary episodes over the last fifty years are linked to a shortage of a key bottleneck resource - namely energy, and the inability to find a quick substitute.

Anonymous said...


Shortages of particular items lead to changes in their relative prices (i.e., real prices), but not to changes in all prices generally. That is, except for one good, which is the numéraire. When the relative price of money changes, the nominal (not relative) price of all other goods must also change. That’s why you’ve been fiddling about with time series in real prices here. If the value of money changes over time, then the price of everything is changing as well.

The issue with excess money is that it is only excess money relative to other assets that are imperfect substitutes for it. It is not excess relative to your previous level of wealth or income. If you save the excess money, you do not have greater wealth. You have the same amount of wealth, but now with a greater amount of it in liquid assets.

The problem with QE is that the friction between imperfect substitutes is no longer there. If monetary policy is about the split between bonds and money in the public’s risk free portfolio, at the zero lower bound, this split becomes trivial and uninteresting. No one cares any longer about the difference between the two—they have become perfect substitutes.

It’s easy to say that inflation is caused by shortages in the supply of energy. It’s less easy to prove it, or even to come up with a plausible story as to why it should be so. Inflation is the (proportionate) growth rate in the price level. What energy shock is causing this rate to be where it is currently? What shocks can explain its evolution over the last 50 years in major industrial countries? A one-off shock to a widely used input (like oil) might cause a one-off rise in the price level, but why would it cause the price level to grow at a faster rate for a decade?

Roger Erickson said...

A friend's comments seem relevant.

why use the cpi instead of the ppi for inflation?

IP is a non-monetary index; how do you relate it to inflation?