Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.

Reasons for despair: The retarded brother of signal processing

by JakeS Fri May 21st, 2010 at 12:21:05 PM EST

Paul Krugman: Core Logic

Now, the measurement issue: we'd like to keep track of this sort of inflation inertia, both on the upside and on the downside -- because just as embedded inflation is hard to get rid of, so is embedded deflation (ask the Japanese). But in the real world, while some (many) goods behave like this, some don't: their prices rise quickly with supply and demand changes, and don't display inertia. So we need a measure that extracts the signal from the noise, getting at the inertial part of the story.

The standard measure tries to do this by excluding the obviously non-inertial prices: food and energy. But are they the whole story? Of course not -- and standard core measures have been behaving a bit erratically lately. Hence the growing preference among many economists for measures like medians and trimmed means, which exclude prices that move by a lot in any given month, presumably therefore isolating the prices that move sluggishly, which is what we want.

So let me get this right: The concept of "core inflation" is justified by the need to keep track of "embedded," or systemic inflation caused by oligopolistic price-setting.

Now, the best way to do that would be to study the actual economics of the issue and try to figure out which parts of the economy are dominated by oligopolies and monopolies, and which parts of the economy are dominated by competitive firms. Of course, that may or may not be practical. So the second best way to do it would be to develop some kind of heuristic for detecting oligopolistic pricing in the raw price data.1

But that's not what our dear econometricians are doing.

[editor's note, by Migeru] Part of the Reasons for Despair occasional series.


What they are doing is simply arbitrarily stripping out sectors that "everybody knows" "obviously" do not contain systemic inflation. And now, after 20-odd years of talking about "core inflation," they figure out that gee, maybe that isn't as meaningfully defined as we thought it was. Oh R'lyeh? I am shocked, shocked I tell you, to see gambling going on in my casino!

So what do our dear econometricians do to remedy this defect? According to Krugman, they go and make a heuristic and apply it to the raw data. Which is good. About 20 years late, but good. What's bad is that they're only looking at this iteration's data, which contains no information about inertia,2 which is supposedly the effect that they want the heuristic to look for. As far as I can tell from Krugman's description, they're killing the signal with the noise.

But the upshot of all this is that, since they are now essentially cherry-picking data that behaves like they want it to behave, they'll never have to be worried about the index not displaying inertia again. The resulting index is still nonsense, but now it's nonsense that matches the models of how it's supposed to behave.

- Jake

1Just off the top of my head, and based on Krugman's description of the phenomenon that they want to model, you could decompose the price changes of a given good by magnitude and frequency and call oligopolistic price-setting by looking at the shape of the resulting spectrum.

2That 10 % price hike you saw this month? Could be adjustment to changing market conditions, or it could be adjustment according to the seller's price adjustment schedule - without looking at the pattern this price shows over several previous iterations, you have precious little way of telling.

Display:
It may be just my current mobile browsing solution, but I can't see where the blockquote is from?
by Metatone (metatone [a|t] gmail (dot) com) on Fri May 21st, 2010 at 03:16:38 PM EST
Sorry about that - dog ate my link tag.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Fri May 21st, 2010 at 03:20:28 PM EST
[ Parent ]
"solution". LOL

By laying out pros and cons we risk inducing people to join the debate, and losing control of a process that only we fully understand. - Alan Greenspan
by Migeru (migeru at eurotrib dot com) on Sat May 22nd, 2010 at 02:05:05 AM EST
[ Parent ]
Great diary. Your footnote one is exactly the kind of signal analysis that I've often asked economists about - and they usually say - "there's some literature on that" but don't follow up. So I don't really know what has been done and further what is done on a regular basis.

I realise Drew may have access to some of the work on this stuff? So I guess I shouldn't be too rude about the state of the art in the field, maybe more exists than I know...

I really like the idea of spectrum analysis because there's clearly different kinds of inertia - and they should give clues to different issues.

I'll throw in my usual complaint about "core inflation" which is that "inflation" has very specific meanings... which always need considering before you go from the "inflation" or "core inflation" figure to a policy judgement.

The most obvious case is that core inflation was low throughout the finance and housing boom that preceded/caused the current financial crisis. So - there are important things (asset inflation?) that the core inflation (and indeed various other inflation measures) just don't measure...

So, in this case, core inflation shows we are heading towards deflation - which does indeed match some other potential measures of the "real economy" - but at the same time the financial economy appears to be inflating again, and with it some commodities that are pretty vital to the real economy (energy? food?)

What does all that mean? I don't know... but I'd be a lot happier if anyone out there can point me to a non-recursive/tautologous definition of inflation that explains what it should mean...

by Metatone (metatone [a|t] gmail (dot) com) on Fri May 21st, 2010 at 03:49:40 PM EST
I really like the idea of spectrum analysis because there's clearly different kinds of inertia - and they should give clues to different issues.

You would need to have a clear idea of how you'd expect a frequency spectrum to look like for different kinds of pricing mechanisms before you start mining the data. Ideally, you'd want to calibrate your a priori expectations against several industries that you know to be competitive/monopolistic/oligopolistic.

What you'd want to emphatically not do is chug a lot of data through a frequency decomposition and then mix and match the patterns you get out. That would probably give you turtles all the way down.

This is no minor task, and the calibration in particular will probably involve digging into historical price/volume data that is less complete than we might have wished. Although it is, of course, preferable to using a nonsense index to make serious policy decisions...

The most obvious case is that core inflation was low throughout the finance and housing boom that preceded/caused the current financial crisis. So - there are important things (asset inflation?) that the core inflation (and indeed various other inflation measures) just don't measure...

Yes. The ordinary inflation indices don't touch the balance sheet at all - they only care about the cash flow. What's worse is that it is less than obvious how to construct an inflation index for assets, or that it is even an appropriate thing to try to do.

As a practical matter, that's not a very big problem, because we have a number of other indicators that can tell us whether asset prices are out of line. The problem is the political will to use them (or rather the lack thereof).

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Fri May 21st, 2010 at 04:18:25 PM EST
[ Parent ]
Why can we not just use what the normal everyday expenses of the average household are?
  1. Housing. All costs. Homeless without it. Don't exclude it.
  2. Food. Starve without it. Groceries and restaurants.
  3. Health care. Total monthly cost of insurance, doctors and pharmaceuticals, both prescription and non prescription. Simpler in countries other than the US.
  4. Transportation. Vehicle ownership and operation costs.
  5. Personal services. From barbers & beauticians, to childcare, personal shoppers and chauffeurs.
  6. Entertainment.

This should be broken out by income deciles and tracked over time. Instead of excluding things that vary widely, such as housing costs, accommodate them. For each decile determine the proportion of households that own and rent, track the expenses for each and average the sums for each decile. Same with transportation. Break it down into automotive expenses, public transit expenses and other, track by decile and average.

We need some analysis as to what really is happening, rather than the cooked numbers prevalent in the USA.

"It is not necessary to have hope in order to persevere."

by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Fri May 21st, 2010 at 11:38:45 PM EST
[ Parent ]
That's pretty much what the consumer price index is already supposed to do. Now, whether the consumer price index does what it is supposed to do is certainly debatable, but that's a slightly different kind of story.

The point here is that there may well be other interesting information in the price/quantity data than the consumer price index.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Sat May 22nd, 2010 at 05:55:48 AM EST
[ Parent ]
The point here is that there may well be other interesting information in the price/quantity data than the consumer price index.

Yeah, the lower deciles are probably just too dumb to use available personal services such as chauffeurs and personal shopping assistants, foolishly forgo tax benefits from home ownership, spend most of their transportation budget on public transit, wasting all that time walking from home to the bus stop, etc.

"It is not necessary to have hope in order to persevere."
by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sat May 22nd, 2010 at 11:20:23 AM EST
[ Parent ]
Hey, I didn't say that the consumer price index isn't a useful measure.

But, just like GDP (which is also a useful measure) isn't the be-all-end-all measure of economic development, the CPI isn't the be-all-end-all of price indexing.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Sat May 22nd, 2010 at 07:39:22 PM EST
[ Parent ]
My comment was not intended as criticism, but more as commiseration. When we deal with economics we enter the land of the retarded, or perhaps of the autistic. But I am not sure. Do the autistic tend to be religious? I have trouble so imagining.

Were we to classify disciplines by the intellectual approaches employed, economics does not belong in the social sciences. What science starts off with axiomatic assumptions that are presented as self evident and are beyond questioning? Example: "Inflation is everywhere and always a monetary phenomenon." Then consider a situation where simultaneously the creation of all money tokens is stopped and foodstuffs along with energy products become exceedingly scarce, such as after a nuclear exchange or the impact of a large asteroid. One can see that the price of food and energy will go up, but it is not because of the money. Nor can the problem be solved by applying any monetary manipulation that does not produce more food or energy. So obviously the axiom on which monetarists base their analysis is falsifiable. Economics, as currently practiced, should be in the religion departments.

 

"It is not necessary to have hope in order to persevere."

by ARGeezer (ARGeezer a in a circle eurotrib daught com) on Sat May 22nd, 2010 at 10:28:53 PM EST
[ Parent ]
What about stopping analysing "deciles" when the real inequality happens at the top centile ?

Un roi sans divertissement est un homme plein de misères
by linca (antonin POINT lucas AROBASE gmail.com) on Mon May 24th, 2010 at 09:39:41 AM EST
[ Parent ]
So cut the data into tranches of AAA+ reliable indicators through to Z- pump and dump stocks, then sell people the AAA+ indicator as a measure of the economy...
by njh on Tue May 25th, 2010 at 03:52:13 AM EST
[ Parent ]
Just follow the Dow...

By laying out pros and cons we risk inducing people to join the debate, and losing control of a process that only we fully understand. - Alan Greenspan
by Migeru (migeru at eurotrib dot com) on Tue May 25th, 2010 at 04:14:47 AM EST
[ Parent ]
they usually say - "there's some literature on that" but don't follow up

That's because they don't get it.

By laying out pros and cons we risk inducing people to join the debate, and losing control of a process that only we fully understand. - Alan Greenspan

by Migeru (migeru at eurotrib dot com) on Sat May 22nd, 2010 at 02:03:53 AM EST
[ Parent ]
Don't get lost, folks.

I love to see patterns in events, emergence, tipping points, et al, but this is obfuscation.

Remember, the current purpose of the "profession" of economics is to justify capitalism and its parasite class of interest casino operators. They're better at bullshit than you are because it's their full-time job.

"A man can be awful stupid if his salary depends on it."

I'm less and less impressed with OK economics, or Obama/Krugman. Every once in a while one or the other emits some surprising burp of reality, but I strongly suspect it is all part of a carefully tuned plan that will not ever result in a significant societal change that removes the grievous excess of the plutocrats that is bleeding the world dry.

Align culture with our nature. Ot else!

by ormondotvos (ormond.otvosnospamgmialcon) on Fri May 21st, 2010 at 05:22:07 PM EST
Even if you jettison most of the orthodox economic theory surrounding them, you still need reliable econometric indicators if you're going to do industrial policy and economic planning and want it to work.

It is possible that you can't get reliable information on the structure and development of the economy by mining price data. But we won't know that until we've tried to do it - preferably in a way that wouldn't make Shannon turn over in his grave.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Fri May 21st, 2010 at 05:33:11 PM EST
[ Parent ]
... that Post Keynesians on the basis of study of the actual american Industrial State divided the economy into fixprice and flexprice sectors about fifty years ago now.

It was rejected because, while being consistent with reality, it was difficulty to reconcile with the utility maximizing general equilibrium model. The same general equilibrium model that was found in the 1970's to be intrinsically and irreparably flawed by the very same "rigourous" type of mathematical analysis which was so loudly applauded when the "existence" of a general equilibrium under certain highly restricted assumptions only loosely anchored to reality.

Of course, Post Keynesians were among those who "called" the housing bubble as unsustainable.

And what is inertia, anyway? Its not an explanation under the utility maximizing theory ... its just a label for the residual that is there when everything that the utility maximizing theory can find a rationale for is taken off the table. It is, in other words, like physical inertia before we had a successful cause and effect explanation for the phenomenon.

Now "fixprice" and "flexprice" is a useful heuristic, but we today have the computer processing power on our desks (in order to play youtube clips of the cat that collapses in frustration when the pigeon gets away) to do find which price series tend to move together.

So, yes, of course anyone looking to decompose inflation due to ordinary price spikes in flexprice markets and inflation due to the evolution of price setting strategies in fixprice markets would, first, go to the actual price series measured on commensurate intervals and, second, sort them according to the ratio of high frequency to low frequency volatility and, third, do a cluster analysis to find which price series are moving together.

However, that would merely extend our understanding of the economy. It would be very unlikely to act as an ideological justification for the ever greater empowerment of transnational corporations.

I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.

by BruceMcF (agila61 at netscape dot net) on Fri May 21st, 2010 at 11:16:17 PM EST
The advantage of using a simple heuristic to mine the price/quantity data directly is that you don't have to conduct a large, expensive survey every ten years or so to group sectors into fixprice and flexprice.

Incidentally, can you point me in the direction of some of the literature on those studies? And do you know whether price/quantity data exists in digital that far back? It would certainly help avoid the most common model-making traps if we could test the models on real historical data before letting it loose on contemporary data...

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Sat May 22nd, 2010 at 07:43:39 PM EST
[ Parent ]
I don't know where the full times series data are collated, but the components of the PPI and CPI indices would have a heck of a lot of disaggregated price level data in index form.

So for an example of coverage of monthly price indices, Table 5. Producer price indexes for the net output of selected industries and their products (pdf)

I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.

by BruceMcF (agila61 at netscape dot net) on Sat May 22nd, 2010 at 08:54:45 PM EST
[ Parent ]
People in financial economics know how to handle a price series and how to construct an index. See, for instance, the Case-Shiller Index. This means that asset-price inflation is well understood, and indices are well-known: they're all over the serious media and people bet their pensions on them!

Why similar techniques couldn't be applied to the less sexy prices of everyday goods, I don't know...

By laying out pros and cons we risk inducing people to join the debate, and losing control of a process that only we fully understand. - Alan Greenspan

by Migeru (migeru at eurotrib dot com) on Sat May 22nd, 2010 at 02:30:24 AM EST
So what's the idea here? That oligopolistic/regulated prices have inertia and free competition prices don't, and the "real inflation" is in the free competition prices? This is nonsense for a variety of reasons.

First of all, the more capital-intensive an industry is the more it tends to oligopolistic/regulated prices. So, if you're going to exclude them from the price index because they exhibit inertia even if they have a large impact in consumers' shopping baskets or in economic flows generally, you're measuring something useless.

Second of all, I have recently developed this intuition that the more volatile prices and volumes are, the more inflation there is [for the physicists: I'm thinking this is related to the fluctuation-dissipation theorem - if I understood the latter I'd be able to tell you for sure]. So, if an industry sector displays price inertia, it should also display lower inflation than average. And so, by excluding it from the analysis, you get a high inflation estimate.

So, what are we doing? We're excluding the capital-intensive sectors of the economy from the price index, and we're introducing an upwards bias in the inflation estimate. Coincidentally, Krugman just wrote:

And all of this has a real, damaging effect on policy. The econ team at Goldman Sachs (not online) makes the interesting point that FOMC inflation forecasts are pulled up by a small group that keeps forecasting much higher inflation than anyone else; this in turn helps limit the Fed's willingness to support the economy. And the deficit hawks have, of course, killed any hope of more stimulus.
FOMC is the Federal Open Market Committee at the Fed.

By laying out pros and cons we risk inducing people to join the debate, and losing control of a process that only we fully understand. - Alan Greenspan
by Migeru (migeru at eurotrib dot com) on Sun May 23rd, 2010 at 02:55:01 AM EST
So what's the idea here? That oligopolistic/regulated prices have inertia and free competition prices don't, and the "real inflation" is in the free competition prices?

No, the real inflation is whatever the inflation is.

The point is that fixed prices tell you something about what the price-fixers think the future inflation will be. And prices being what they are, that can become a self-fulfilling prophecy.

Pure flexible prices, OTOH, don't price in expectations. So if you want a handle on what people believe the future inflation will be, you will have to compare the inflation in the fixprice and flexprice sectors of the economy. In other words, "core inflation" is pretty uninteresting on its own - the interesting figures are total inflation and the difference between inflation in fixprice and flexprice sectors, because it is that difference which tells you whether the fixprice sector expects higher or lower inflation in the future.

Second of all, I have recently developed this intuition that the more volatile prices and volumes are, the more inflation there is

That's an interesting intuition, and it may very well be correct. In that case, figuring out the relevant expectations would get a bit more complicated...

So, what are we doing? We're excluding the capital-intensive sectors of the economy from the price index, and we're introducing an upwards bias in the inflation estimate.

Actually, no. "Core" inflation is an attempt to exclude the flexprice sector, giving a downward bias.

And the more likely explanation for the FOMC having a problem with the whole "making accurate predictions" thing is that they assume long-run money neutrality. Forecasters who actually have to make more-or-less reliable forecasts can't afford rigid adherence to such ideological drivel, and will have to (tacitly) cast it by the wayside... (That's the kind version. The unkind version is that the FOMC is a group of hacks who fit the analysis to a pre-ordained conclusion.)

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Sun May 23rd, 2010 at 07:06:45 AM EST
[ Parent ]
JakeS:
the real inflation is whatever the inflation is
That begs the question of that "inflation" is. And most definitions of inflation are circular or "fudge-factor" definitions, where all we have is a disparate collection of price and volume series and it's by no means clear that there is an index that captures it.

By laying out pros and cons we risk inducing people to join the debate, and losing control of a process that only we fully understand. - Alan Greenspan
by Migeru (migeru at eurotrib dot com) on Tue May 25th, 2010 at 04:16:41 AM EST
[ Parent ]
Yes, discounting - of which inflation is a sub-field - is inherently political.

But my point is that "core inflation" isn't an attempt to arrive at a meaningful discount rate: It is an attempt to divine the expected discount rate of fixprice market participants.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Tue May 25th, 2010 at 06:39:03 AM EST
[ Parent ]
The issue that Krugman is trying to explain is that inflation is not an objective, verifiable fact about the world, universal from all points of view.  Rather, inflation is something that individuals experience differently based on the kinds of things they want to consume or have to consume. How one measures inflation is thus dependent upon the question one is really asking and upon the political arguments one is trying to make. Do you want to know whether poor renters are facing higher prices relative to wages, or do you want to know whether more productive economic agents are facing higher prices? (Are you interested in welfare and equity, or are you interested in economic growth?)

What Krugman is explaining here that central bankers, whose job is to try to match the money supply to the amount of real things produced and consumed in the world, are going to be more interested in core inflation, without food and energy noise in the mix.  The reason is because food and energy are classic supply and demand commodities whose price movements reflect changes in real supply and demand conditions more than other things, so higher food and energy prices usually don't mean that too much money is being printed.  Higher oil prices, Krugman is arguing, is not due to monetary issues but rather due to perceived shortages of oil. Central bankers shouldn't worry about oil prices when determining whether to print more, or less, money, just the prices of things that don't fluctuate so much with supply and demand changes, such as wages.

by santiago on Tue May 25th, 2010 at 11:54:23 AM EST


Display:
Go to: [ European Tribune Homepage : Top of page : Top of comments ]

Top Diaries