Numbers Racket?

The argument in this piece is that over time, a series of "opportunistic" changes in the statistics describing the economy have led to a biased and overly rosy view of the health of our economic system. You can read it for yourself, there's a link to the article below along with a brief passage, but my take is a bit different.

There are answers to each of the claims in the article - none are new - but let me take a more general approach. The economists in the agencies that prepare our national economic statistics are dedicated individuals whose only goal is to make the statistics as accurate as possible. They are not political hacks willing to distort the picture of the economy to help the administration. Far from it. Their goal is to provide the most accurate statistics possible and there are always improvements they would like to see made. But due to factors such as data compatibility and the cost of redefining and recalculating a measurement, statistics are not generally adjusted until it is believed the improvement is large enough to justify the change (though sometimes the statistics can be adjusted backward or spliced together in a way that preserves data compatibility).

However, even when it is clear that the measurements could be improved, and the improvements are substantial, the political process may prevent definitional changes. When are we most likely to see desired changes implemented? At some points in time, the changes may work against the administration in power, in others the changes may be helpful, and it is unlikely that the administration in power would approve of a chance unless it was helpful.

This means that changes in the definitions of statistics will, looking back, generally appear to have been put into place to help the administration in power, i.e. it will appear that they were manipulated for political purposes since the changes almost always make the economy look better. But that's not the motivation - the reason for the change is to improve the measurements - and the politics surrounding such changes dictate that they will occur most often when the changes are politically favorable.

This does impart a bias, but it is not a bias in the measurements themselves, it is a bias in the types of improvements that can be made. All of the changes improve the measurements, they don't distort the picture of the economy they make it more accurate, at least that's the intent, but the changes do not generally occur unless they are politically favorable. (Let me acknowledge that there could be some bias toward a better than true picture if this is how it works. If the politics only allow definitional changes that make the economy look better, then there could be some changes that are known to be needed, but left undone because they would make the economy look worse. This would make the reported statistics biased toward a healthier looking economy, and this is why the agencies that measure the economy should be as free as possible from political influence, so they can adjust the measurements in either direction as needed.):

Numbers Racket: Why the economy is worse than we know, by Kevin Phillips, Harper's Magazine v.316, n.1896 1may2008: If Washington's harping on weapons of mass destruction was essential to buoy public support for the invasion of Iraq, the use of deceptive statistics has played its own vital role in convincing many Americans that the U.S. economy is stronger, fairer, more productive, more dominant, and richer with opportunity than it actually is.

The corruption has tainted the very measures that most shape public perception of the economy—the monthly Consumer Price Index (CPI), which serves as the chief bellwether of inflation; the quarterly Gross Domestic Product (GDP), which tracks the U.S. economy's overall growth; and the monthly unemployment figure, which for the general public is perhaps the most vivid indicator of economic health or infirmity. Not only do governments, businesses, and individuals use these yardsticks in their decision-making but minor revisions in the data can mean major changes in household circumstances—inflation measurements help determine interest rates, federal interest payments on the national debt, and cost-of-living increases for wages, pensions, and Social Security benefits. And, of course, our statistics have political consequences too. An administration is helped when it can mouth banalities about price levels being "anchored" as food and energy costs begin to soar.

The truth, though it would not exactly set Americans free, would at least open a window to wider economic and political understanding. Readers should ask themselves how much angrier the electorate might be if the media, over the past five years, had been citing 8 percent unemployment (instead of 5 percent), 5 percent inflation (instead of 2 percent), and average annual growth in the 1 percent range (instead of the 3–4 percent range). We might ponder as well who profits from a low-growth U.S. economy hidden under statistical camouflage. Might it be Washington politicos and affluent elites, anxious to mislead voters, coddle the financial markets, and tamp down expensive cost-of-living increases for wages and pensions?

Let me stipulate: the deception arose gradually, at no stage stemming from any concerted or cynical scheme. There was no grand conspiracy, just accumulating opportunisms. As we will see, the political blame for the slow, piecemeal distortion is bipartisan—both Democratic and Republican administrations had a hand in the abetting of political dishonesty, reckless debt, and a casino-like financial sector. To see how, we must revisit forty years of economic and statistical dissembling. ...

The U.S. Economy Ex-Distortion

The real numbers, to most economically minded Americans, would be a face full of cold water. Based on the criteria in place a quarter century ago, today's U.S. unemployment rate is somewhere between 9 percent and 12 percent; the inflation rate is as high as 7 or even 10 percent; economic growth since the recession of 2001 has been mediocre, despite a huge surge in the wealth and incomes of the superrich, and we are falling back into recession. ...

The question is which gives the more distorted picture, the old definitions from 25 or 30 years ago, or the newer definitions. I'm sticking with the newer definitions, though sometimes it doesn't matter much, e.g. take the unemployment rate figure. The figure above is simply the broadest definition that the BLS measures (U6 in Table A12), and this measure adds several percentage points to the standard measure (it's available every month - there's no secret here). But the difference is fairly constant over time so that while the number is higher, the relative picture doesn't change much in terms of when unemployment was lower or higher, i.e. when things were better or worse for labor market participants.

Or take core inflation, one of the things criticized in the article. As explained here many times, there are theoretical reasons for preferring the core measure, and some of the discussions criticizing the Fed's use of core inflation don't seem to realize that the "alpha-trimming" is symmetric, i.e. both large increases and large decreases in prices are excluded from the preferred inflation measures used for policy (that's not true of CPI les food and energy, but that's not what the Fed uses for policy). Nor do they realize that the Fed is not trying to measure the cost of living. (Just to say this once more, one use of a price index is to measure the cost of living, but another is to determine the course of monetary policy, and it turns out the best price index for the Fed to use for policy excludes highly volatile prices. The Fed is not trying to measure the cost of living, it is trying to find the best signal for the course of monetary policy - see here). The CPI less food and energy measures of prices reported in the news  aren't the best measures of "core inflation," and I'd prefer that the headline number be the symmetrically trimmed PCE figure, but the two measures aren't that far apart, generally, and making such a change would likely create as much confusion and suspicion (see above) as it would improvement in communication about economic conditions. As for how to measure prices, the old way which gives the higher inflation figure in the last paragraph (does the author really want the Fed to increase the federal funds rate that much?), or the new way, again, I prefer the new way.

Comments

[kio]

1. I agree that new data are of better quality and any possible improvement should be done where appropriate. On the other hand, an ultimate demand is to keep data as compatible over time as possible. In many cases higher precision is much better than accuracy. Imagine that there is some TRUE (for economic theory linking various macroeconomic variables) value of inflation, GDP, unemployment, etc. If you consistently measure the same portion of this true value you can easily carry out quantitative analysis by scaling the measured value to the true one. Same in physics - scaling and calibration is a common technique.

2. Some macroeconomic variables like real GDP and inflation are inherently virtual, i.e. there is no direct way to measure them because of new goods and services with unknown prices in past come to the market. (Only valuation is possible, which is subjective.) In such a situation severe discussions on evaluation methods are likely to arise. As we observe. Before start, one has to understand that there is no way to win the discussion in a quantitative way and people soon drop to personalities, i.e. to emotions involved in valuation. So, better not to start and just follow up the definition you like the best. If you think it provides an approximation which fits your understanding of true values.

3. One useful outcome of the revisions to old definitions consists in the opportunity to estimate the uncertainty of the data in possession. The BEA people in their technical papers admit that the uncertainty of annual real GDP is around 1 percentage point. This estimate comes directly from annual and comprehensive revisions to GDP and its components. This experience says that the future revisions still may be as large as 1%. So, when you see an advance quarterly estimate of +0.6, you may read it in the following way – 90% probability that it comes in the end (and there is no actual end to the revision process) somewhere between -0.4% and 1.6%.

[Ryberg]

The quality of our numbers matters, even if the systematic bias still allows certain kinds of analysis. The fact that reported inflation and "actual" inflation are highly correlated may allow a policy analyst to use either for marginal changes to policies. If the goal of the Fed is to reduce inflation, it can use either measure to gage success.

However, if the goal is to understand a phenomenon, then a biased measure will provide a biased understanding. A corollary to the old saw, "You manage what you measure" would be "You can't manage what you don't measure."

There used to be a common belief that the market rate of interest could be decomposed into premia for inflation, risk and return. If inflation is understated, then either risk or return must be overstated. Now I believe that risk is also understated which only leaves return as overstated. But policy makers (e.g., the Fed) may assume that the return premia is fully adjusted for inflation and risk when they set policy. How else can you explain interest rates at 60 year lows for an economy that is so unremarkable.

This understated inflation and even more understated risk will come back to haunt us. The systematic biases may not effect our marginal analyses but this economic frog is about cooked.

But I wonder why economists and the media are so willing to accept the administration's proferred statistics even when more appropriate ones are readily available? Is it that we don't really want to know?

[Gabriel Milhache]

>> "The economists in the agencies that prepare our national economic statistics are dedicated individuals whose only goal is to make the statistics as accurate as possible."

Let's not overdo it! -- There is a point to Public Choice too.

[hari]

There we go again...*How to Lie with Statistics*!!!

The alternative to current *de facto* inaccuracies may be a function of making the statutes (of reporting agencies) *legally* independent of official influence (one way or another) by WH budget controllers or whoever. In the final analysis, an Act of Congress establishing a single *Federal Office of Statistics* would be the way forward from this current CPI index and its abnormalities.

Australian Govs had a temerity to hold-up official Statistics from being *announced* during general election process (ie. undermining credibility of party in power) on inflation, etc. The subsequent establishment of an independent office of statistics, as a statutory body, made it independent of gov machinery/depts/etc.

[ken melvin]

Even if, ... things have changed so much , the numbers may have new or different meanings.

[bakho]

The fallacy in the article is the belief that people would be even angrier if the numbers were worse. People get angry about economic conditions when those conditions are affecting the daily lives of themselves and their family. Numbers cannot hide this truth. The elites think that the average American pays more attention to the media (and pundits like themselves) than they really do. We are not sheep.

[baileyman]

Is there a way to check some these stats using IRS data?

[save_the_rustbelt]

I'm not enough of a math geek to understand all of this, but what I do know is for the past few years the unemployment numbers in Ohio and Michigan have been pure crap.

The unemployment rate in Michigan is only a couple of points higher than the rate in California? Gimme a break.

[RegularFolk]

"This does impart a bias, but it is not a bias in the measurements themselves, it is a bias in the types of improvements that can be made."

Mark, it is difficult to see how this is a coherent argument. Consider another example, lets say that pharmaceutical industry has complete hold on what treatments can be given to patients and pharma will only approve those treatments that will clearly increase their profits. Would anyone then think that only the best treatments are given?

Another issue is that it is not possible to know whether bias is small or large and therefore the amount to which it distorts the real picture.

One can try to ignore the actual value and concentrate on the change only, but there again the problem is the underlying assumption. The assumption is that the relationship between the actual measurement and the biased measurement is linear. However if the relationship is nonlinear, concentrating on change alone is not sufficient without knowing the form of correlation function.

Let me give you a simple real life example. I bought a computer in 1998 in 1.5K, few years later I had to replace it with another one (again around in 1.5K in 2003) and earlier this year, I had to replace again (1.5k). I still do the same things with my computer (mainly using internet and word-processing) that I used to do in 1998. The price of computer has not changed for me but it has changed for BEA. They have continued to depreciate such things at around 17.5% per year, while I have not noticed any decrease in price and no one else has.

Lastly, no single profession has only "dedicated people". If the easy way to move up the career ladder is to compromise, you will find people compromising to any extent possible.

[hari]

The difference in CPI between US and EU may be developing into a serious (political) policy issue, I suspect. Because the ECB is mandated to control overall inflation (not just *core* inflation) whereas Fed is using core inflation (less food and energy prices). This type of statistical variance will eventually meet a bilateral roadblock, including how US unemployment is calculated. There's under-reporting, so it seems, from published figs.

Today, we've EuroStat (in Luxemburg) which is responsible for collecting and analysing ALL EU-27 statistics. They've problems, as I noted, on Eastern European countries stats - but these distortions are being discussed at the relevant level of EU dialogue with a view to streamlining the statistical evidence - esp. for comparative economic analysis/reporting + ECB rate setting machinery (although not all are presently in Euro).
Also a lot of TA has been budgeted for these new members to reform their (old) reporting system.

[me]

The US is the only one that excludes food and energy.

The Boskin Commission revised the CPI to limit Social Security cost of living increases.

When I started a graduate econ class I remember the first thing the Old prof said. The United States has the worst economic statistics in the world. decades later he was correct.

Does the rest of the world use hedonics? Does the rest of the wold say, well we can't find any new jobs so each month they make up about 75% of the jobs?

I suppose you could say the CIA are good, hard working bureaucrats, but when the politicians get done with it I would say it is slightly biased.

[odograph]

What a surprising piece. It starts with "good people" but then describes a mechanism for systematic bias, leading to somewhat biased answers. How biased? I think Mark gave us a "gut feel" for that, as we all have.

Someone else above asked what we could use to cross check. IMO it's the census, household budget, and household savings, numbers.

What happened when US savings declined?

The "cost of living" obviously rose faster than income.

[jamzo]

the utility of statistical descriptions of an economy depend upon the customer

bureaucrats, politicians, citizens, business executives, investors and economic scientists have need different measures to construe what is happening in an economy

much of the debate on statistical descriptions seems to be related to political conflicts

[cm]

RegularFolk: I think you can find machines for far less than $1500 (peripherals included) to do "word processing and internet". I would suspect back in the old days your $1500 machine was nowhere top of the line, and perhaps not even midrange. Today it will buy you quite a bit better.

(Whether today's computer equipment is cheaper more because of artificial price distortions -- all manufacturing offshored -- than improvements in productivity and technology is a separate question.)

[cm]

bakho: It may not make a difference in the level of anger, but in where the anger is directed ...

[RW]

Following previous comments to the effect that (a) regardless of how "dedicated" individual statisticians may be if ascension to upper management is contingent upon compromise then compromise is what the system will favor and (b) what Phillips describes is a mechanism for systematic bias that not only demands compromise but demands it along lines favorable to elite's management of public relations and perceptions of accountability; e.g., granting that elites may have liberal or conservative or, as in the current case, larcenous biases they will still all tend to favor numbers and policies that give them room to maneuver, favor constituencies (or disfavor them when necessary but without seeming to do so), etc.

In short I do not believe this is a mechanism that will lead to slight bias but rather must lead to ever widening gaps that, over the decades, result in a set of numbers that best serves the powerful independent of what the putative social policy functions of those numbers may be. In a way it's not an intentional lie it's another form of bullshit (pace Harry G. Frankfurt): Truth about the economy and the state of its citizens is not the primary goal the numbers serve as far as elites are concerned but if they do happen to serve that goal, well, that's perfectly all right.

The situation with those who are charged with creating those numbers is far more complex I think, not conducive to simple-minded conspiracy theory at all, which is why I prefer in-depth analyses such as Theodore Porter's.* The statisticians who develop these numbers are not just attempting to cater to the powerful or to the populace at large, I believe they are genuinely trying to be 'objective,' but above and beyond the problem I noted above -- the possibility that systematic biases are growing regardless -- the problem remains that no amount of high powered algebra and statistics will ever completely forge a connection between quantification and objectivity in public life. This is not really a result of being unable to approach a scientific ideal -- physics envy as it were -- it is a result of the practical realities that relatively insecure bureaucracies and research fields dealing with social phenomena confront when they are obliged to mediate strong and competing interests, interests to which they themselves are not immune.

*Porter, Theodore M (1995). Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. Princeton University Press.

[robertdfeinman]

Those with a real interest in accurate numbers (say banks and investment firms) have their own staffs to gather data and do analysis. Even if they have to rely on "official" numbers originally, they are free to adjust them as they see fit to compensate for any shortcomings or biases they feel are present.

Do these groups make such adjustments? If so, are their predictive abilities any better than the Fed or other government agencies? Soros made a lot of money, twice now, betting against government policy, but he seems to be part of a small minority even of "sophisticated" investors.

Governments may be blinded by ideology or political considerations, but is this also true of money men? Something doesn't seem to add up.

[Alex Tolley]

It is a little disingenuous to suggest that the CPI is used for Fed policy, and just "used" for cost of living adjustments - Phillips is more right - businesses use these numbers as a stick when granting pay increases etc.

Like many people, I don't buy the official CPI numbers either. A simple example, H/C insurance has had double digit price increases for at least 2 decades, it is becoming a major part of the family budget, businesses are dropping or passing on much of the increases to employees, yet amazingly, CPI is 2-4%.

This is a good example where PK's admonition to "do the arithmetic" might be helpful.

While statistics collections are important, I get the uneasy sense that the use of those statistics is not exactly benign, but being used for purposes other than running the economy for the benefit of the population.

[Bob]

>> But the difference is fairly constant over time so that while the number is higher, the relative picture doesn't change much in terms of when unemployment was lower or higher, i.e. when things were better or worse for labor market participants.

The political accountability changes significantly though.

Not to mention the appropriate politcal pressure which would be on legislators to address the problen with benefits or food stamp extension.

The cons just used the phony 5% unemployment rate as an excuse for not doing that

The mainstream media keeps providing cover for bad economic policy by spreading the lie that unemployment is only 5% as opposed to the 9 to 12% that Philips talks about.

Matter of fact it's reasonable to say that the real unemployment rate could have been the tipping point in the last prez election since it was close .

[Bruce Webb]

Well I would agree that the career professionals at BLS and BEA are doing the best they can. Then again the professionals at NASA, EPA, and Fish and Wildlife are doing the best THEY can. Which did not prevent actual political hacks from changing the data post facto for political ends or to protect industry interests.

I had a brief interchange with both Max Sawicky and Dean Baker on productivity. Each assured me that their colleagues insisted that tampering with this number series was impossible. Well those are two guys I defer to on such matters. But still it would be dangerously naive to conclude that official financial numbers are by nature protected from tampering by the political appointees in ways that scientific conclusions are not. It may be that it is impossible for Hank Paulson to influence GDP or Elaine Chao to move the labor numbers, but there is nothing in the record to suggest that they would resist political pressure from the White House to do so if they could get away with it.

[Alex Tolley]

regularFolk: "Let me give you a simple real life example. I bought a computer in 1998 in 1.5K, few years later I had to replace it with another one (again around in 1.5K in 2003) and earlier this year, I had to replace again (1.5k). I still do the same things with my computer (mainly using internet and word-processing) that I used to do in 1998. The price of computer has not changed for me but it has changed for BEA. They have continued to depreciate such things at around 17.5% per year, while I have not noticed any decrease in price and no one else has."

That's a pretty expensive machine you're buying each time.

While appreciate the hedonic approach, I think there is a fundamental problem of trying to measure economic growth as though everything is a commodity, like coal or wheat. There is also the reverse problem when automation makes services cheaper, e.g. ATM machines vs bank tellers, yet the economy has apparently shrunk with this transition with the closure of bank branches and fewer tellers.

[Alex Tolley]

One other point about CPI. It is an aggregate number for the economy as a whole. But individuals have their own personal CPI - what inflation is doing for their incomes and budgets. It is quite possible for a low income family to experience higher CPI than higher income families, yet their wage increases will not reflect that. For the super wealthy, does inflation even touch them at all?

[Bruce Wilder]

If an economist wants to argue that changes to the CPI have made it "more accurate" then he ought to have some definite idea of what "more accurate" means, up front in his argument.

The kinds of changes made by the Boskin Commission to the CPI, arguably, did make it a more "accurate" measure of changes in the cost-of-living. But, if so, those changes also made the CPI a less accurate measure of monetary inflation.

Let me repeat myself: "accuracy" in the definition of a statistic must mean accurate in relation to some particular purpose -- absolutely more accurate for every purpose is simply a logical impossibility. Making the CPI more accurate as a measure of changes in the cost of living made it less accurate as a measure of inflation.

[Bruce Wilder]

I am very disappointed in Mark Thoma's "Support the Troops" argument concerning the integrity of the thankless souls, who generate these statistics.

I agree with Bakho that the chief fallacy of Phillips' argument is that people would be "angrier" somehow, if the statistical numbers were different. Ordinary people experience economics first-hand, not thru the mediation of the Bureau of Labor Statistics.

The statistics matter. But, they matter to the behavior of those officials and leaders, who do experience the economy through statistical mediation. The statistics affect both official action and elite opinion.

The CPI is the most obvious example, as it has implications for inflation-indexing of all kinds of payments, income tax brackets, etc. The changes made by the Boskin Commission implemented a political agenda, affecting the distribution of income. There's no escaping that.

Do I think that unemployment statistics, which seriously underestimate unemployment, and inflation statistics, which seriously underestimate inflation, distort fiscal and monetary policy?

The simple answer is: Yes.

The extensive answer is more than I could condense into even one of my marathon blog comments. The gist of my answer has to do with how determined political interests interact with technocratic political impulses.

Fundamentally, I would say that there are powerful political interests, who feel that the New Deal was an abject "failure" precisely because of how the New Deal succeeded. There are powerful political interests, who want to bring on depression as a means of further redistributing income and wealth upward. They want to destroy Social Security as a means of further redistributing wealth and income upward. Etc.

Those plutocratic interests have to overcome the political mythology of technocracy, which supports and undergirds our economic institutions, particularly the economic institutions of fiscal and monetary policy. It is a whole worldview, which must be undermined and then replaced.

The systematic changes in economic statistics are the minutiae of a vast political movement that includes turning the Media into a monolithic propaganda apparatus, and the corruption of Academia.

To see what is going on, you have to connect the dots. And, if you do, you will be thoroughly frightened.

Consider the implications of the statistical changes in the context of the Wall St. Journal survey of economists, (mentioned by Paul Krugman) which showed a plurality attributing a sound fiscal policy to John "tax cuts" McCain. Then, recall that the Wall St. Journal is controlled by the same man, who brings us Fox News.

[donna]

I tend to see the problem as more basic - it is an aggregate of the economy, and not a consideration of what it is like for an individual within the economy. It doesn't offer any help to people to know what GDP is, or even GDP per capita. What counts is what are the opportunities and potentials for an individual, and that is not considered by any of these measures. The CPI does not reflect ANYTHING about my cost of living. Not at all.

[Goldilocksisableachblonde]

I suspect the vast majority of Enron accountants were "dedicated individuals whose only goal is to make the statistics as accurate as possible". The lone exceptional and brave Enron whistleblower who did come forward might have been a "black swan" , so we shouldn't count on similar truthtellers arising from government ranks anytime soon.

For most American households , the last 8 years have felt like the stagflation of the seventies , except that the pain has been eased by open access to cheap credit , created , we now know , out of (very) thin air.

Now that the credit has disappeared , all of the "malaise" indicators are setting off alarms , except the official one , the 'Misery Index' of inflation plus unemployment.

However , if you use the Shadowstats numbers , or something close to those , it all makes sense again.

I think Phillips is spot-on.

[kio]

Well, I still have three basic questions about economic measurements. Depending on answer economics exists or does not exist as a science.

1. Do "true" values of such macroeconomic variables as inflation (equivalent to real GDP) and [un-]employment exist?

2. Are they measurable?
If yes - then methods of measurements can be improved and estimates justified.
If no - discussions about the merits of various approaches are forever.

3. Is is possible to reveal some reliable links between these (and other) macrovariables to validate their values by cross-variable statistics?

[bob]

>> I am very disappointed in Mark Thoma's "Support the Troops" argument

Well what would you expect.

He's one of the troops.

says...

Also i think the "economists are doing the best they can" argument is almost irrelevant in some cases.

For example they do calulate the real unemployment rate (u6) but that rate is never presented to the public as the real rate. U3 is what they try to pass off.

So it's not a matter of a cooked or a real stst. It's a matter of what they are going to present as a real stat.

[rdaneel]

1. The CPI changes that have been made, to adjust for quality improvement and substitution, and to use owners' equivalent rent rather than owner asset prices, all have very good rationales. By that, I mean that all these changes make the CPI a better measure of changes in the cost of living, defined in terms of the costs of reaching a particular utility level. I think this is the best goal for the CPI.
2. The Shadow Stats website that Phillips promotes is very poorly documented, and some of the alternative inflation number provided by this site are hard to believe. Given the lack of documentation at this website, there is no reason to believe these data.
3. Data from BLS suggests that the adjustments for quality improvement and substitution generally do not have drastic effects in lowering the inflation rate.
4. Using owners equivalent rent rather than asset prices does have more of an effect in certain time periods. At present, using owners equivalent rent is probably increasing the measured inflation rate compared to what you would get using old BLS methodologies.

[odograph]

Remember, average people looked at the inflation rate, and the interest rate, and chose to go negative on their savings. They chose to pile on debt.

A rational choice, given negative real interest?

[Bob]

All one has to do is look at the divergence between the stats and the polls.

5% is thought of as a low unemployment rate.

Yet polls on the question of "is the country on the right or wrong track" are at a record high when it comes to wrong track.

And the presidents approval rating on the economy in specific resembles the depression.

That wouldn't be happening if the economy was good in the "reality based community" as Suskine said.

[RW]

"A rational choice, given negative real interest?"

The answer would seem to be yes but when it appears the debt was primarily used to acquire depreciating* assets and/or maintain consumption/lifestyle I'd guess the picture becomes much more cloudy.

Personally I looked at this as the last gasp of the middle class, an attempt to evade the terrifying possibility they were being abandoned. JMO

*Granted real estate appeared to be an exception, appreciating more quickly than inflation recently, but historically it mostly just keeps up and buildings are a depreciating asset regardless.

[Jim]

Inflation is how well you lived this year compared to last considering your change in income. Ignore official pronouncements, look to your own situation and decide.

[gordon]

A bit off-topic but Goldilocksisableachblonde's (any chance of shortening that?) remark about "black swans" made me check the Wikipedia article. It says "Nassim Nicholas Taleb's definition, a black swan is a large-impact, hard-to-predict, and rare event beyond the realm of normal expectations. Taleb regards many scientific discoveries as black swans—"undirected" and unpredicted. He gives the September 11, 2001 attacks as an example of a Black Swan event". The trouble with that example is that something like Sept. 11 '01 was predicted - by Nicholas Guyatt in his book published in 2000 "The Next American Century?" (Ch. 3). He wrote: "[US foreign policy] will ultimately encourage a small minority to express their grievances in horrific ways: by attacking American interests abroad and even targeting cities within the US" (my emphasis).

Very few events are totally unpredicted. There is a huge grey area between "accepted as a possibility (though remote) by the mainstream" and "predicted by at least one person". The subprime meltdown was quite widely predicted, for example, but not initially by mainstream commentators. Concern widened over time until some very respectable academics and businesspeople began to express worry about the "housing bubble".

I wonder whether this article and its exposure (though with heavy qualification) on a "respectable" blog indicates the spreading of concern about these big economic aggregates from a few outsiders (Like John Williams) to include more mainstream academic and policy commentators.

[gordon]

Sorry, the title of Guyatt's book I referred to in my previous comment is "Another American Century?" Profuse apologies.

[Lafayette]

Plain English

MT: The economists in the agencies that prepare our national economic statistics are dedicated individuals whose only goal is to make the statistics as accurate as possible.

This may be true, MT, and I've no reason to doubt it. Nonetheless ... I've a beef to make.

I was culling statistics at the BEA web-site (to make a comment in this forum). The statistic in question (GDP by Industry) is located here. They've put a glossary up on the site, which is a nice thing to have, particularly when presenting tables of data by industry.

So, why the hell is it that just about any “Industry Sector” in the GDP tables has no definition in the glossary? And, when you ask them by email about the apparent deficiency, some nerd tells you to link to a humongous explanatory document with the standard industry definitions – where the one you want is somewhere in the mess at the bottom. Have they never heard of definitional pop-up windows?

[I will remark that the BLS data is far simpler to access and understand.]

OK, I’m bitching. But, my message is simple: If we want ordinary people to understand economics and its impact upon our lives, then a public web-site containing base data should NOT require a BA in economics to read, obtain and (most importantly) understand information contained therein.

Academics and the supposed "professionals" in state agencies have an obligation (to their public) to explain their science in “plain English”. Otherwise, economics will remain an esoteric discipline only for the initiated.

It's impact on public policy will otherwise remain curtailed and therefore limited.

NB: Another beef is why some public agencies, funded from government revenues, think they have the right to publish information that comes from the public domain and sell it. That information should be free, gratis and for nothing. The OECD and the ILO are particularly guilty of this practice.

[Lafayette]

Devil in the details

Ryberg: How else can you explain interest rates at 60 year lows for an economy that is so unremarkable.

Easy, externalities.

Besides, definitional notions of risk and return have probably changed since the economy was last "remarkable" (e.g., the dot.com boom 'n bust).

That is the challenge academics face when trying to make sense historically of market data phenomenon. It is always a Movable Feast, meaning definitions set in the public domain are rarely common to all agents within that domain.

I have seen, for instance, four or five different ROI calculations all different because "investment" was accounted for quite differently in each circumstance. The calculation made sense only internally to the organization making it, and were completely incomparable amongst organizations.

So, pray tell, define risk -- and you will have the premium associated by that definition. Define it another way, and the risk premium is altogether different.

The devil is in the details, where even angels fear to tread.

[hari]

Establish a single office of *US Official Statistics* as a statutory body responsible only to US Congress. No one - repeat no one - must be allowed to influence their professional responsibility to put TRUTH in their *branded*(label) output. It should be legally allowed to collaborate across the universe of statistical evidence and publications in order to *standardize* global statistics.

In global trade negotiations (ie. GATT Rounds) we faced
unmitigated statistical *horrors* which constrained official dialogue in negotiations. Let's try and get it right; US statistics, in my opinion, is not adequately independent of
bureaucratic interference and whatnot.

[odograph]

This is getting off topic, but on the Black Swan thing, I think it is more about what society (broadly) entertains as a possibility.

There will always be prophets in the wilderness, and (as Taleb points out!) we tend to identify them in hindsight, in order to convince ourselves that the future is less random.

[Bob]

>> The economists in the agencies that prepare our national economic statistics are dedicated individuals whose only goal is to make the statistics as accurate as possible.

Well one needs to ask the most obvious question which is...

How does Mark know this to be a fact ?

Does he know all these people ?

Sounds fishy to me.

I've worked in a lot of different jobs and i never got the impression in any job that everybody without exception was only interested in putting out the truth.

[Lafayette]

odo: ... in order to convince ourselves that the future is less random.

Anyone ever seen a good regression of random events? Show me. I'm all eyes.

There a lies, damn lies and .... ???? (Three guesses, the first two don't count.)

[vimothy]

Ok Mark, thanks for addressing this piece. I've been thinking about John Williams quite a bit, recently.

I'm left with a question I can't resolve on my own. Perhaps one of your readers can help.

Surely the whole point of these measures is to have a stable measure of "inflation", i.e. to be able to pick a vantage point from which to observe changes in prices driven by chaneges in the money supply. The epistemological questions that government economists seemingly wrestle with are less important than picking one method and sticking with it. I'm not implying that they are driven by ideology or self-interest, but I can't see how it is helpful to change the definition yet continue to present CPI (or any measure) as continuous. Fact is, is that if you discount innovations to the CPI, inflation looks a lot higher; if you don't, inflation looks low.

Now, the later has got to be good for governments with a lot of debt. (Right)? Keeping inflation high has got to be good for governments with a lot of debt. (Right)?

I'm nopt saying that it's malicious, but I do question the need to alter the definitions, which amounts to moving the goal posts half way through the game.

Otherwise, why not retrospectively go back and apply todays CPI measure to the '70s? We might discover that there was no inflation at all: great news!

[Marbles Redux]

Even if the CPI is the perfect tool for inflating sticky prices at 2% per year, the method of currency expansion can still create inequalities. If the entire monetary expansion is distributed to a fraction of the population, the recipients will experience increased purchasing power. Non recipients will experience a decrease in purchasing power.

If you have 4 children, each with 10 marbles to trade for chores, 1 marble buys one chore. If you expand the supply of marbles by 2 each year, but give the extra 2 marbles to just one child every year, the one child will be able to "buy" more chores from the other children over time. The other children will systematically be able to buy fewer chores each year with their marbles.

In effect, the parents are forcing the 3 unloved children to work extra chores for the benefit of the 1 favored child. This, of course, is bound to create resentment among the 3 unloved children. The method of introducing monetary expansion into the economy makes a difference, and a more equitable method of introducing extra currency should be considered.

[kthomas]

A very wise man once siad that you can lie to some of the people all of the time, and you can lie to all of the people some of the time, but you cannot lie to all of the people all of the time.

[Blissex]

«The question is which gives the more distorted picture, the old definitions from 25 or 30 years ago, or the newer definitions.»

That's a very good question, one that I have asked a few times.

Well, if the new definitions are more accurate, then Volcker was one of the biggest morons in the history of the USA, causing a decade of poverty and hardship on the basis of rather flawed statistics.

Current statistics computed with 70s methods give the same numbers at then -- so either there is a stagflation now, or there was no stagflation in the 70s.

So which is which? Were people in the 70s completely fooled by their grossly flawed statistics, or are people in the 00s completely fooled by the their grossly flawed statistics?

[Lafayette]

blis: Were people in the 70s completely fooled by their grossly flawed statistics, or are people in the 00s completely fooled by the their grossly flawed statistics?

What grossly flawed statistics, pray tell?

The price of a loaf of bread then is calculated differently from now? A can of tuna fish? A hammer and a nail? A two-by-four. A gallon of gas? A standard four-door car, of a given cylinder capacity?

The salary of an office worker? The salary of a nurse? The salary of a machinist on the shop-floor? A systems analyst?

What has changed are things like the cost per unit of computing, the cost of a radio (FM/AM versus FM/AM plus CD plus SD plus two large woofers), that of a 737 at 130 seats then and 160 seats now. As technology changes products, there is a discontinuity in the base of comparison, yes, but these items can be finagled.

Are we presuming that the way the CPI was calculated then is radically different than the way it is calculated today in terms of the means of gathering the data, or the data gathered that constitute the calculation?

If it's the latter, there is not much to be done, except avoid comparisons that go beyond a specific period or time, say a decade. And, then, so what? What's important, in terms of policy-making, is quarterly changes and then year-on-year.

Comparisons beyond three or four years or ten years are interesting only to an historian, and not many of them.

[cm]

Lafayette: "Comparisons beyond three or four years or ten years are interesting only to an historian, and not many of them."

I disagree. People set, and are guided by, references, in all aspects of life. In that sense we are all "historians".

[Lafayette]

cm: People set, and are guided by, references, in all aspects of life. In that sense we are all "historians".

True enough.

But, I maintain, that those references have limited horizons. The individual tends to care more about a year-on-year CPI movement because their salary is typically the consequence of an annual appreciation by their employers.

The CPI gives them an idea of how their purchasing-power is reduced or enhanced. That consideration is an immediate one, meaning the time-frame is over only a year.

Deflating GDP over a ten year period seems to give historical consistency to the development of an aggregate value, but I wonder about the real relevance of that manipulation. Because, I question how much GDP could have changed in its composition over that extended period.

Don't you?

[reason]

Blissex is clearly correct. But there are extenuating circumstances. The problem is who is the CPI directed at in a world of increasing inequality. I think in general relative prices have moved to the detriment of families. The real answer is we should (in this day of wide spread and sophisticated computing power) be using a greater variety of measures to get a more nuanced view of the world, rather than pretending that a single number contains all the information we require.

[Lafayette]

reason: The real answer is we should (in this day of wide spread and sophisticated computing power) be using a greater variety of measures to get a more nuanced view of the world, rather than pretending that a single number contains all the information we require.

I think you may be right.

But, for economics to be comprehensible beyond the sphere of academia, it must remain "simple to sell", that is, readily understandable to the general public.

Any economist who starts trying to explain a complex, but accurate CPI will have the attention of the American public as measured in nanoseconds.

Unless, of course, we have a silver-tongued, one-handed economist somewhere to explain the complexity ... ;^)