Farewell. The Flying Pig Has Left The Building.

Steve Hynd, August 16, 2012

After four years on the Typepad site, eight years total blogging, Newshoggers is closing it's doors today. We've been coasting the last year or so, with many of us moving on to bigger projects (Hey, Eric!) or simply running out of blogging enthusiasm, and it's time to give the old flying pig a rest.

We've done okay over those eight years, although never being quite PC enough to gain wider acceptance from the partisan "party right or wrong" crowds. We like to think we moved political conversations a little, on the ever-present wish to rush to war with Iran, on the need for a real Left that isn't licking corporatist Dem boots every cycle, on America's foreign misadventures in Afghanistan and Iraq. We like to think we made a small difference while writing under that flying pig banner. We did pretty good for a bunch with no ties to big-party apparatuses or think tanks.

Those eight years of blogging will still exist. Because we're ending this typepad account, we've been archiving the typepad blog here. And the original blogger archive is still here. There will still be new content from the old 'hoggers crew too. Ron writes for The Moderate Voice, I post at The Agonist and Eric Martin's lucid foreign policy thoughts can be read at Democracy Arsenal.

I'd like to thank all our regular commenters, readers and the other bloggers who regularly linked to our posts over the years to agree or disagree. You all made writing for 'hoggers an amazingly fun and stimulating experience.

Thank you very much.

Note: This is an archive copy of Newshoggers. Most of the pictures are gone but the words are all here. There may be some occasional new content, John may do some posts and Ron will cross post some of his contributions to The Moderate Voice so check back.


Thursday, July 21, 2011

Esoteric Macroeconomics, Substitution Bias and a Variety of Metrics

By John Ballard

This is a post I hesitate to write because it will make me look like I'm supporting the president and the Gang of Six and as an old-fashioned Liberal I don't relish being outcast from the Left.  I already live in the Bible Belt among Christian Zionists, the same voters who put xenophobic, arch-Conservative, anti-choice candidates in office for the last forty or fifty years, so when I can't find myself in agreement with today's progressives I am forced to stand alone.

That said, I must confront some uncomfortable realities we all know and hate.

  • Government is done by politicians, bureaucrats and technocrats, in that order.

  • Today's econometrics are more precise than those of the past

  • Both parties in our two-party system play the same games

  • Like it or not, policies are determined by money, not principles.

I could write an essay about each of these topics, but in the end my rage and frustration, no matter how eloquent or angry, would not alter any of those four realities. They are in philosophical terms a priori.

Here are a few links that will shed more light than heat on the budget negotiations. That's a polite way of saying the reader may find them dull or sneaky, depending on a pre-formed political judgment, but I read them as good faith explanations with no hidden agenda. That's why I mentioned technocrats in the first bullet point. As politicians come and go, parties and coalitions shift policies and change places -- but the bureaucrats remain the same (except for political appointments) and the technocrats are the professionals who really make government function.

Something tells me the Gang of Six, like the Catfood Commission before them, have been listening to the technocrats in the same way that a patient with a serious diagnosis (think heart disease, macular degeneration or some kind of cancer) often becomes a more informed expert than their primary care physician...at least about their own disease.

In fact, the Gang of Six seems to be a Congressional manifestation of the deficit commission which has become everybody's whipping boy.  When we call it the Catfood Commission we are applying one of the principles the technocrats at the Bureau of Labor Statistics calls substitution bias. They speak of how buying habits change when prices go up, which is exactly the point when the proverbial frugal old person substitutes catfood for canned seafood.

?The Gang of Six Simplified at The Committee for a Responsible Federal Budget.
New America Foundation board members include a few names I respect, including James Fallows, Fareed Zakaria and Atul Gawande. If that organization publishes something I'm not afraid to read what it says.

Since 2003, the Committee for a Responsible Federal Budget has been housed at the New America Foundation. New America is an independent, non-partisan, non-profit public policy institute that brings exceptionally promising new voices and new ideas to the fore of our nation's public discourse. Relying on a venture capital approach, the Foundation invests in outstanding individuals and policy ideas that transcend the conventional political spectrum. New America sponsors a wide range of research, published writing, conferences and events on the most important issues of our time.

?Anderson Sees `Good Things' in Gang of Six Debt Plan
  Anderson Take a moment to watch this video which I cannot embed.
("disabled by request" really pisses me off, especially when all I have to do is send out a Twitter message which makes it universally available. But even then it has to be watched at You Tube. How stupid is that?)

July 21 (Bloomberg) -- Barry Anderson, a former deputy director of the Congressional Budget Office, talks about U.S. budget negotiations and the so-called Gang of Six senators' deficit-reduction proposal. Anderson speaks with Erik Schatzker on Bloomberg Television's "InsideTrack." (Source: Bloomberg)

?Observation Post � New CPI Might Be Used to Set COLAs
Pay attention to the date on this one, last December.
I know nothing about The Military Officers Association of America but this Q&A is very even-handed. Somebody was paying attention last December and asking the right questions. Lots to learn here.

Recent debt-reduction studies, including the National Commission on Fiscal Responsibility and Reform in December, recommend adopting the �chain-weighted� CPI for Urban Consumers (C-CPI-U) to set COLAs for federal programs that now rely on the CPI-U or CPI-W. What is a chain-weighted CPI? It�s an index first introduced in 2002 based on a different methodology to address one of the criticisms of the other CPIs, which is substitution bias.

When prices change, people can change spending behavior. They can substitute, or move away from more expensive items to buy less expensive items. CPI-W and CPI-U don�t account for consumer substitution in any way.

In 1996, a Senate-appointed panel, the Boskin Commission, assessed the current CPI [in their report Toward a More Accurate Measure of the Cost of Living] and said one of the weaknesses of current indices is they essentially assume people buy the same stuff regardless of [how] prices [changed]. There�s a certain consensus that was a valid point, although you could find people who disagree.

So the idea is that, in measuring cost-of-living changes, if you don�t account for consumers changing their behavior, substituting cheaper goods for more expensive goods, then the index is biased?

Can you give an example of substitution bias? There are two kinds. One is within an item category. Let�s say I go to a store to buy cheese. I usually buy cheddar, but the store has doubled the price, so I buy Swiss instead. That is what�s call lower-level substitution bias.

In 1999, at the BLS, we started using a geometric means formula within [an] item category to account for that sort of bias. But there is also substitution across item categories. For example, when the price of beef goes up, I am more likely to buy chicken. [Or catfood. JB]

Why assume consumers� buying habits change that way? Economic theory certainly suggests that demand for a particular good is related to price. As price goes up, compared to other goods, we tend to demand less of it. That�s not always true, but, in general, if things become more expensive, relative to other things, you buy less.

Substitution is much easier in some things than in others. For medical care, there essentially is no substitute. For housing, there is very little [substitution]; we don�t go out frequently to find another apartment. For something like apparel or food, there is quite a lot of substitution.

So how does the chain-weighted CPI address substitution by consumers? What you would really like to do in an index is not make assumptions about how people behave but actually get expenditure data, note how it really does change, and base your CPI on actual expenditure as it occurs. That is the basis of the chain-weighted CPI.

With CPI-U and CPI-W, we measure price changes for a market basket from base period. We estimate total expenditure based on those weights.

The chain-weighted CPI incorporates not only the base weights but the updated weights, too. So we�re not assuming what people do based on reaction to price change; we are actually measuring what their new expenditures are. The weight arguably could be said to be more accurate because it is mostly free of substitution bias.

More at the link. This is just the juicy part.
That makes you real excited to see the rest, right?

1996 was several administrations ago. Just as every Congress has kicked the can down the road with a "Doc Fix" ever since the Medicare reimbursement formula proved unworkable, every Congress since then has been aware of the flawed metric for COLA, but as every schoolboy knows that third rail will kill you if you touch it.

?Measuring Up: The Case for the Chained CPI
Last but not least.
This is the sock puppet of the Deficit Commission.
It was written in May but anyone who has read the other three links will be able to follow along with better understanding.
Get ready to discern shades of difference among

  • CPI

  • CPI-W

  • CPI-U

  • CPI-E (not mentioned here. See here and here)

  • C-CPI-U


Most arguments I am reading want to simplify this adjustment for partisan reasons. When I look at the details I cannot go along with them. The sentence that jumps off the page is this. Both parties really want this to change. They have seen the need for a long time but each party wants the other one to take the blame. Why do you think the discussion is behind closed doors and the details are being left for later?

Unlike the methodological changes in the calculation of CPI-U and CPI-W that are automatically reflected in the published measures used for indexing programs under current law, using the more accurate chained CPI for indexation instead of the CPI-U or CPI-W requires a statutory change in law.

When I moved from the private, for-profit sector to the so-called not-for-profit segment I got to see how budgets trump operational realities in a big way. When you're working to squeeze a nickel profit out of every dollar your "budget" is nothing more than a guideline. In the private sector if a disaster strikes or a competitor opens up across the street whatever the budget said better be in for a sudden adjustment to tighten up.

But in the not-for-profit sector -- and this includes government operations -- the budget is set in stone. In fact, if there is "extra money" in the budget, the norm is to find a way to spend it so next year's budget won't be reduced. The system is designed to move only in one direction. No one but no one is stupid enough to suggest not spending all that is in the budget.

From what I can tell (since zero-based budgeting is not politically feasible) this method of designing future budgets will be as much a fiscal improvement as the catalytic converter was for the internal combustion engine. Unless somebody convinces me otherwise, I'm sold.


  1. "when the price of beef goes up, I am more likely to buy chicken"
    What about people who were buying chicken to begin with? What do they do when the price goes up? Or people who could not afford meat at all?
    The point is that the index is supposed to reflect what PRICES are doing, since it is a PRICE index, not what shoppers are doing. That whole argument is smoke and mirrors to manipulate data.

  2. I do understand. But the problem is bigger than the Social Security COLA formula. It has to do with how budgets and regulations are formulated for the entire government. In the same way that many union contracts are coupled with the official federal minimum wage rate (base multiple...when the official minimum increases it triggers an automatic union increase way up the wage tree), this "chained CPI" metric is apparently outside the legal parameters for budget planning. Using this tighter formula will have a long-range squeezing effect on federal budgets having nothing to do with Social Security. The entire federal spending budget will be impacted. So who says we aren't spending enough already?
    The present formula, remember, has resulted in zero COLA for 2010 and 2011. I think both parties sense a moment when SS beneficiaries are frustrated enough that any change will be perceived as something better. I'm watching and listening to what position the AARP takes.
    This is not a popular idea, I know. But neither is deficit spending. Something has to break the hammerlock the crazies have on Congress and the GOP. I tried to find out if this new metric would have resulted in anything better than zero for the last two years but it appears the answer would still be "no". This two-year-old Fact Check piece is the best I could find, and the whiplash effect of the oil price spike seems to have been the problem. (Hardly anything to do with the prices of chicken, beef or catfood.)

  3. You need to watch the second segment of Friday's Democracy Now.

  4. Indexes, I haven't thought about them for a while. I suspect the main issue isn't about "chaining", so-called, which, I think, really is trying, from a statistical POV, to get around the problems of usual CPI indexes using Laspeyres' formula that tends to over-estimate the index, in an inflation framework, by geometrically averaging it, and the results of a Paasche formula approach, [Paasche under estimates inflation] Doing this, in effect, results in the [Irving] Fisher ideal index.
    The real problem, as may have been noted, is picking the basic basket weights for goods & services consumed. In other words, these baskets vary over income levels, heavens, eh. And, in other words again, CPI under any formulation isn't a point estimate it's a, wait for it, curve. Most Stat Bureaus pick a level of income and use other surveys to determine the basket of goods etc. consumed by the picked income level to feed into any CPI index calculation. Hey you can pick the top, bottom, middle, etc. etc. etc. . Always useful to know what actually is being used, I think & it isn't always mentioned in any "sources and uses document".
    As an indication of the different results one may get fiddling around a quick peruse of this paper (pdf) might provide some indication: http://j.mp/nwb0X1 . Syd Afrait, noted as the "guru of the price index", in the paper always made a point of telling his grad classes that CPI was only a single point for Stat agencies & this was wrong but practical, before embroiling them in symbolic logic.

  5. Fer crying out loud, geoff, you have posted in another language. Makes me feel like Prissy in GWTW. "I don't know nothin' 'bout birthin' babies."
    In lay terms I need to find out if (a) the pros and cons of the proposed CPI index, and (b) am I correct to assume that if approved it will have a far-reaching impact on much more than just what's in the spotlight?
    As I said, zero-based budgeting would be my preference (with incentives for spending less instead of more) but that's in the ponies and unicorns category. Federal programs are as close to immortality as mankind has reached, so any indexing feature slowing growth is a move in the right direction, which is why I'm attracted to this "chained" index.
    It's an odd term. Makes me imagine a moving average but as you seem to have said, the effect is multi-dimensional.