Carpe Diem

Addressing the myths and misconceptions about the CPI

In 2008, the Bureau of Labor Statistics addressed some of the myths and misconceptions about its Consumer Price Index that have circulated over the years on the Internet, especially at several non-academic fringe websites that have attracted many, conspiracy-theory driven, cult-like followers.  Here are some excerpts from the BLS article “Addressing Misconceptions about the Consumer Price Index“:

From the introduction:

A number of longstanding myths regarding the Consumer Price Index and its methods of construction continue to circulate; this article attempts to address some of the misconceptions, with an eye toward increasing public understanding of this key economic indicator.

Within the past several years, commentary on the CPI has extended well beyond the circle of economists, statisticians, and public officials. The strongest criticism of BLS methodology has not been concentrated in a single profession, academic discipline, or political group, but comes instead from an array of investment advisers, bloggers, magazine writers, and others in the popular press. Also, whereas in the past the CPI frequently was held to be overstating inflation, recent criticism has focused on supposed downward biases.

Appearing as they do in national media and in the age of the Internet, these criticisms probably have been more widely quoted and circulated than most academic journal articles and panel reports on CPI issues. Although the BLS welcomes comments and regularly discusses and debates measurement issues with its advisory committees and at professional meetings of researchers in economic measurement, the recent criticisms of the CPI have been directed not so much to the BLS itself as to the public at large. This article is an attempt to correct some of the misunderstandings underlying those criticisms.

From the conclusion:

It is hoped that this article has put to rest some of the misconceptions and myths about the CPI:

1. It is a myth that the BLS reduced the growth rate of the CPI by assuming that hamburger is substituted for steak.

2. It is a myth that the use of hedonic quality adjustment has substantially reduced the growth rate of the CPI.

3. It is a myth that the 1983 adoption of owner’s equivalent rent systematically reduced the growth rate of the CPI shelter index.

4. It is a myth that Social Security payments are updated by a CPI that does not include food or energy.

Conclusions:

1. The sizes and effects of the changes implemented by the BLS have been overestimated by critics. The introduction of the geometric mean formula to account for product substitution has decreased the rate of change of the CPI by less than 0.3 percentage points annually, not by 3 percentage points annually as some have claimed.

2. The changes implemented by the BLS that some critics construe to be a response to short-term political pressure were, in fact, the result of analysis and recommendations made over a period of decades, and those changes are consistent with international standards for statistics.

Finally, the CPI is not, and can never be, a perfect index. Moreover, all of the topics raised in the recent commentary on the CPI — including the methods for dealing with consumer substitution, quality change, and owner-occupied housing—are critically important to the accuracy of the index. The very existence of the CPI methodological changes discussed here attests to the fact that the CPI must always be working to enhance the index. The CPI benefits from the work of academics and others who identify ways in which the CPI can be improved. The BLS also benefits when the public understands how the CPI is constructed and what the index’s strengths and limitations are. It is hoped that this article will help increase that public understanding.

MP: As the BLS points out, the CPI can never be a perfect index, and it faces many challenges constructing an index that accurately measures consumer prices, subject to the difficult issues of quality changes, product substitution, the introduction of new products, calculating owner’s equivalent rent, etc.  But I’m confident that the career, Ph.D.-trained economists and statisticians at the BLS are giving us a reasonably reliable measure of consumer prices that is consistent with the international standards of the OECD, and is free from any politically motivated influence. Maybe I’m just too much of a mainstream economist and BLS “shill” but I see no conspiratorial evidence of intentional data manipulation, just hundreds of highly-trained and dedicated economists, analysts and statisticians at the BLS who take their jobs and research on price indexes seriously, and do the best they can to create the best possible “imperfect” price index – the CPI.

HT: Steven Hales

48 thoughts on “Addressing the myths and misconceptions about the CPI

    • The CPI uses hedonic price adjustments, so when you substitute lower priced items for higher priced items due to higher prices, the CPI actually decreases. It’s crazy, I know. The high cost of orange juice has caused us to substitute to lower priced fruit juices. You can thank us for keeping the price index down.

      • “The CPI uses hedonic price adjustments, so when you substitute lower priced items for higher priced items due to higher prices, the CPI actually decreases.”

        Does it? BLS is claiming that that is not, ahem, “substantially” true. (Does that mean it’s just a little bit true? “True but we’re not doing it as industriously/vigorously as we really want to?”)

        Or maybe they’re weasel-wording so much that they won’t or can’t say straight out that the “hedonic” adjustments properly allow for quality difference… or the reverse, that they actively hide lowering quality of living due to decreasing quality of products. Why do they seem to refuse to be clearer on this? Perhaps because they’re afraid that people would object even more vociferously.

        One would hope that they would assess a higher value to steak than soy-extended hamburger, for instance. But then we could open up a vicious flame-war between the vegans and the omnivores over the relative value of various fake meat-like products vs. real meat and fowl and fish products. There are some vegans who turn up their noses at the fake products because they don’t like that they have some of the appearance of actual, you know, meat.

        And there are (insane IMO) people who loooove the frangible plastic garbage in cars and foil instead of real leather, hard-woods, and sheet and bar steel. They consider such scams and smaller cargo and passenger room and lower horse-power, instead of improved higher-efficiency, more power producing engine designs, as the one key path to what they consider to be “higher efficiency”. They’d rather focus on bizarre (and government-mandated) gimmicks to counter the increased dangerousness of the smaller, lighter, i.e. IMO far lower quality, vehicles.

        And don’t get me started on the garbage they push at big box stores of every kind, but especially the home deterioration stores.

    • I stock-up on Skippy or Jif peanut butter when it goes on sale, e.g. 50% off.

      One or the other seems to go on sale at least once every three months.

    • will-

      1. you inadvertently argue against your own point. if you switched to a brand you preferred less, that implies a drop in quality. if you thought the other one was better, you would have been buying it before. this means there should be a negative quality adjustment that offsets at least part of the price drop from switching brands. the fact that such an adjustment is not applied DOES mean CPI is misleading.

      think of it this way: teddies peanut butter goes up in price from $1 to $1.50. stop and shop goes up from 99c to $1.29. you switch to the cheaper brand. how can that possible mean that the price level dropped? the price of both brands went up, thus the price level went up.

      2, not all price moves are supply driven. in many cases prices rise because demand rose. thus, to say that consumers will use less is untrue and the weighting adjustment winds up having the wrong sign.

      • I bought the Teddy’s because it came in no salt added. The Stop and Shop brand has salt and I’ve gotten used to it to the point that I like it just as much. And it was a whole dollar difference ($3.99 to $2.99).

  1. It should be noted, health insurance, paid by employer, and Medicare, paid by government, are compensation, not consumption, paid by consumer.

    • Consumers are the *only* ones who pay on an ultimate basis, especially on all medical items.

      Bingo!

      Individuals as end users ultimately pay all costs, as well as all taxes.

    • “Medicare is paid via taxes.”

      Yes, it’s paid via extortion, and by increasing federal government debt… and thus by inflation.

      Producers produce the value that is being drained and diverted from other parts of the economy. Who are the “producers”. Well, they’re production workers and managers and executives and janitors who produce whatever is of value that others want.

      Of course, not all production workers and managers and executives and janitors are net producers, but it’s impossible to pin-point which ones are net producers and which ones are net consumers because of the costs of accounting and the ambiguities both artificial and natural. Sometimes, one is a net consumer and at other times that same individual is a net producer, or maybe he’s doing productive work that doesn’t show up on the balance sheets for a while — sometimes until long after he’s left the firm. So, it’s very hard to tell.

      And most economists don’t even try, relying on gross aggregates to tell the tale while hiding the details, instead.

  2. Also, it should be noted, many Americans pay fixed monthly housing payments, or their houses are paid-off, and pay fixed monthly car payments, with low interest rates, until their cars are paid-off.

    So, there’s no inflation there.

    Moreover, it seems, many retired Americans are well-off, e.g. receiving Social Security and retirement income (capital gains on Treasury bonds increased substantially).

    • peak-

      those arguments do not make any sense.

      so, if your employer pays for healthcare, that does not affect price?

      if foreigners pay, that does not effect price?

      so, if you want to rent a hotel room, but the town is full of foreigners who are filling them up, the higher price you pay is not really higher?

      how does that follow?

      price level for consumption level goods is price level for consumption level goods. who pays/buys them is irrelevant.

    • that’s also a bogus argument.

      if i buy a coffee maker in 2010 for $100 and am still using it in 2012, that says NOTHING about the price level.

      if they now cost $120, then that is the price level.

      by your logic peak, the price of cars should be the weighted average of the cost of every car on the road, not what a dealer asks if you want one.

      if i buy 10 years worth of soup in cans, that does not mean the price of soup cannot rise for a decade.

      you are mistaking monthly outlay for price level.

      this is because, as ever, you do not understand the difference between price level and cost of living.

      inflation is a change in price level, not in cost of living.

      the price of life preservers changes when your boat gets hit by an iceberg. the fact that you bought one before does NOT mean there has been no change in the price of life preservers, only that you are not in the market for one.

      by your logic, there can never be inflation in brussels sprouts as far as i am concerned because i never buy them. they can go to a million dollars a pound, and it has zero effect on me.

      do you really think that is the same thing as no price level change?

      • The fact of the matter is that by their own admission, they have no idea how to measure inflation or the price level. They have developed a methodology and they are motivated to defend it even as they admit that it is ‘imperfect’. A centralized standard for measuring the price level is a non-market based solution that will do as well as any centralized non-market agency that looks to set prices, that is, not well and it will very likely be catastrophic. Keep in mind that inflation measures the decline in the purchasing power of the currency and nothing else. I maintain that the best measure of this is against unchanging assets such as gold and other commodities that are used as alternatives to fiat currency. The fact that one ounce of gold required only $24 in 1920 and $32 in 1970 but almost $2000 today provides much more information than the CPI about the purchasing power of our currency than the CPI.

    • “many Americans pay fixed monthly housing payments, or their houses are paid-off, and pay fixed monthly car payments”

      Yes, people lock into contracts, but people still move around, though not as much these days (in part because employers are not providing relocation assistance*). OTOH, the county extortionists/assessors have tended to keep on cranking up the “values” despite actual market prices.

      Some tech firms used to buy homes and land, help the employee find a new place, and take on the risk of selling the old one… but that was before the elimination of tax breaks for interview, relocation assistance and employee training, and before the flood of cheap, young, pliant foreign labor with flexible ethics.

    • True. I believe that the role of the CPI should be to measure the erosion of our purchasing power related to increases in the money supply over extended periods of time. Economic impacts that are unrelated to the money supply (like droughts and innovation) should be removed. As a proxy, I use the price of gold as an index. I divide the price of a product by the price of an ounce of gold at that time to get an estimate of the gold adjusted price of the product. A car today for example offers much more than an equivalent vehicle of a century ago, albeit at a higher dollar cost. Rather than using the imperfect CPI to compare the prices, I gold adjust them to determine how many ounces of gold it takes to buy each vehicle. I find this to be very informative. Likewise, in the 1970′s we can see that the decoupling of the last links to a gold standard and the breakdown of the BW agreememt led to a spike in gold prices (in dollar terms). This caused the gold adjusted price of oil to collapse by up to 90%. This led of course to outrage by the oil exporters, the formation of OPEC and to the oil embargo. I admit, this approach may also be imperfect, but I find it more valuable than the CPI because it is market driven,

      • Gold is a ridiculous measure of the CPI. Gold was $250 an ounce in 2000.

        Purchasing power increased, because nominal income rose faster than inflation.

    • Many goods are just better, just cheaper, or better and cheaper. So, there’s also deflation.

      That is not deflation. You are misusing the term.

  3. Some people make lots of false assumptions about the CPI:

    Common Misconceptions about the Consumer Price Index
    BLS

    “When the cost of food rises, does the CPI assume that consumers switch to less desired foods, such as substituting hamburger for steak?

    No. In January 1999, the BLS began using a geometric mean formula in the CPI that reflects the fact that consumers shift their purchases toward products that have fallen in relative price.

    Some critics charge that by reflecting consumer substitution the BLS is subtracting from the CPI a certain amount of inflation that consumers can “live with” by reducing their standard of living.

    This is incorrect: the CPI’s objective is to calculate the change in the amount consumers need to spend to maintain a constant level of satisfaction.

    The BLS is not assuming that consumers substitute hamburgers for steak.

    Substitution is only assumed to occur within basic CPI index categories, such as among types of ground beef in Chicago.

    Hamburger and steak are in different CPI item categories, so no substitution between them is built into the CPI-U or CPI-W.

    Furthermore, the CPI doesn’t implicitly assume that consumers always substitute toward the less desirable good.

    Within the beef steaks item category, for example, the assumption is that consumers on average would move up from flank steak to filet mignon if the price of flank steak rose by a greater amount (or fell by less) than filet mignon prices.

    If both types of beef steak rose in price by the same amount, the geometric mean would assume no substitution.

    In using the geometric mean the BLS is following a recognized best practice for statistical agencies. The formula is widely used by statistical agencies around the world.

    Critics often incorrectly assume that BLS only adjusts for quality increases, not for decreases, and that hedonic adjustments have a large downward impact on the CPI.

    On the contrary, BLS has used hedonic models in the CPI shelter and apparel components for roughly two decades, and on average hedonic adjustments usually increase the rate of change of those indexes.

    A recent article by BLS economists estimated that the hedonic models currently used in the CPI outside of the shelter and apparel areas have increased the annual rate of change of the All Items CPI, but by only about 0.005 percent per year.”

  4. “2. It is a myth that the use of hedonic quality adjustment has substantially reduced the growth rate of the CPI.”

    this seems like an absurd statement. if it has so little effect, then why do it? it’s time consuming, arbitrary, and a potential source of distortion and bias. so why include such adjustments if they mean so little?

    the comment about steak and hamburger is also terribly disingenuous.

    while strictly true that hamburger is not being substituted for steak, this comment masks far more truth than it reveals.

    hamburger is not being substituted for steak.

    what is being substituted is cheaper steak. prime becomes choice. rib eye becomes flank steak.

    all the best cuts of meat have disappeared from cpi. but there is no offsetting quality adjustment. if flank steak is just as valued as ribeye, then why does rib eye cost so much more?

    the bls is being far too cute with that comment. it’s a glib rhetorical dodge to miss the (pardon the pun) meat of the issue by misdirecting readers.

    • Agreed. The entire article smacks of the kind of glib professorial elitism rampant in people that are disassociated from reality,

  5. i also echo bart’s point:

    if the bls really wants to “clear up misconceptions” they why don’t they make their work transparent?

    why not give us the actual data and show its effects for all their quality adjustments as opposed to keeping them like some sort of state secret?

    this whole piece from them seems outrageously disingenuous. if they really wish to achieve their stated goal, it could be very easily done. show us the math, not glib misdirection about steak and hamburger to mask the fact that it’s really flank steak for t bones and choice for prime and why not show us the math on the quality adjustments as opposed to making unsubstantiated claims?

    this stuff would make Goebbels blush.

    • Bart, the article reflects how the BLS works to provide unbiased results on price changes, whether they’re seasonal, structural, or cyclical.

      • agreed bart.

        the actual info on how weighting is determined, specific shifts, and the magnitude of the effects is simply not provided. there is even less information on quality adjustments.

        if this is such a great system and works so well, then why not share this masterpiece with the market?

        let us check your work guys. that’s how science works, even the “dismal science”.

        the bls acts like the AGW fraudsters like mann who refuse to let anyone see their raw data or the mathematical formulas they use. they just claim “trust is” and tell us the results but do not let outsiders check their work.

        that’s a foundational issue in any scientific discipline.

        i could claim to have produce cold fusion in my garage over the weekend.

        i’ll send you the results of my experiment and show more energy coming out than going in. but i will not tell you how i did it not even show you what calculations i used, just the final answers.

        will you trust me? ready to invest in this world changing new technology? probably not.

        if the BLS has nothing to hide, then it ought to share this info.

        i wonder if we could serve them with a FOIA request and demand this?

        they are, after all, a government agency.

        of course, such requests have accomplished little with such noted bastions of non transparency as the GISS run by hansen at nasa, so i doubt the bls would be much different, which is ironic as it is their job and stated purpose to provide economic data.

        is it really so much to ask that they actually do so?

    • It’s easy to assume the CPI understates inflation when you’re looking at little information.

      When I moved from Colorado to California, one of the first things I noticed was high-quality, and low quality, steak was much more expensive.

      Also, I recall being surprised fish (salmon, in particular) was much more expensive in California than Colorado.

  6. The myth of the CPI is that it is a farsical measurement used by supposed scientists to act like they are somehow saying something meaningful about the economy. The fact that so many “Ph.D-trained economists and statisticians” continue to emphasize something that measures nothing says something very profound about the economics profession.

    If a geologist measured the rates of erosion of sand dunes, granite mountains and asphalt parking lots, and tried to sell it as a meaningful ‘erosion index’, nobody would think that index meant anything. It would be rightly ignored as meaning nothing.

    Yet, large groups of economists regularly aggregate completely unrelated prices, based on the subjective valuations of individuals, and state manipulated money, and proclaim the all-wonderful CPI to the world.

    “Imperfect” indeed, and not scientific at all.

    • Do you think the CPI would have more credibility if it were run by a non-government agency? Some people will not believe anything the government has to say (some people don’t even think we landed on the moon). I think the other problem with the CPI is that people try to use it for their own personal inflation rate when it is clearly not.

      I personally find fault with that same United States department (BLS) that decides to highly publicize the U-3 unemployment rate (currently 7.9%) while sitting on the U-6 unemployment rate (currently 14.6%). Is there a good reason to do that if you want to maintain credibility as an organization?

      • A meaningless measurement is still meaningless, no matter who does the measuring.

        If you want to measure inflation, then measure the money supply, but economists can’t even agree on what that is.

        • With good reason, as the money supply is not the only cause of inflation, check out the velocity charts here. Money supply is never a good indicator of inflation, only potential inflation, otherwise we’d have had a ton of it already, with the explosion of reserves at the Fed since the financial crisis, now up to three times the pre-crisis total.

          All that said, I agree with your initial statement that an aggregate like the CPI is not a very useful measure to begin with. Perhaps it was worthwhile a century ago when there were much fewer goods and they didn’t change very fast, I don’t know, but that’s certainly not the time we live in today, when I can buy hundreds of different kinds of cereal and my ultrabook is outdated within a year.

          What I find interesting is how bart will rail against how the CPI is measured- I knew when this post was put up that it would soon be festooned with bart’s comments :) – yet still argue for the index, just measured the way he wants. I wonder what utility he finds in CPI at all. Also, what does he think motivates the BLS to mismeasure CPI: what do they gain from their conspiracy? That would be much more interesting than going over hedonics and all the other useless minutiae that dominate this argument, every time Mark posts a defense of the CPI.

          • I don’t think the CPI can be argued as meaningless because many adjustments are made using the index (COLAs and the cap on property taxable values in Michigan for a couple of examples). It can be argued that the CPI is inaccurate, but what complex aggregate measurement cannot be improved?

  7. Mark: “But I’m confident that the career, Ph.D.-trained economists and statisticians at the BLS are giving us a reasonably reliable measure of consumer prices that is consistent with the international standards of the OECD, and is free from any politically motivated influence.”

    I am as well. It is such a waste of time to have to wade through all the comments of those at Carpe Diem who continuously fault every economic analysis that depends on BLS figures. Just a damned waste of my time and everyone else’s time.

      • mark-

        that seems like a really questionable comment esp[ecially coming from one, such as yourself, who has so little to say about the substantive issues.

        volcker, bill gross, david einhorn etc are all fringe kooks?

        there is a strong and sizable set of folks on wall st that believe that cpi understates inflation.

        there are significant and valid reasons for this.

        you sound like the “the science is settled” crowd trying to declare victory without having a meaningful debate by declaring those who do not agree with you to be fringe kooks.

        i have a great deal of respect for you and you blog, but i think you are going way too far here in your claims of tin foil hattery.

        and hey, if the BLS really wants to clear this up and the data is really as they claim, well, then it’s easy to do it: release the numbers. show us the calculations.

        if their desire is really to “clear up misconceptions” that would seem like the best way to do it, no? so why don’t they?

    • Well, Bart, I’ll actually respond to this comment because it seems to me a legitimate concern.

      The CPI does include all taxes which are directly associated with the purchase of goods and services. those taxes would include sales taxes, excise taxes, and government user fees such as toll charges and parking fees.

      The CPI does not directly include such taxes and corporate income taxes, business property taxes, and taxes paid by business proprietors for a very good reason: such taxes should be reflected in the final prices paid by consumers. For that matter, the portion of wages which employees pay to the government are also reflected in the final price of the goods and services received by the buyer.

      If one believes in the concept of equilibrium price – which I do – then one understands that the long term equilibrium price for a good or service will include all the costs which are incurred in producing that good or service. Those costs would include:

      - coporate income taxes;
      - business property taxes;
      - all wages, including the portion of wages withheld as income tax and payroll taxes;
      - excise taxes;
      - air travel taxes included in air fares;
      etc,

      In other words, Bart, taxes are either directly or indirectly included in the CPI.

      • Bart,

        All of the expenses incurred by a producer are covered by the price paid for that producer’s good or service. It is true that some producers can sell below cost for limited periods of time. But such practices cannot be sustained.

        There is no need to “prove” that the prices the BLS includes in its compilations include the taxes paid by the producer. A basic understanding of the concepts of supply, demand, and equilibrium price is sufficient to know that taxes have been implicitly taken into account.

  8. Bart: “then why doesn’t the BLS actually break them out, even as a memo item?”

    Becuase, Bart, the BLS does not collect data on taxes. Instead, they collect data on end user prices. Those prices – which are set by the producers to cover all expenses, including taxes – do include the amount of income and every other tax paid by the producer. Those prices also include an amount for the business insurance paid by the producer. And an amount paid for whatever raw materials are consumed in the production of the goods and services. And amounts for all other expenses the producer incurs. But you won’t see separate line items for raw materials and for business insurance and for all those other expenses incurred, either.

    • Barr: “It doesn’t bother or concern you that taxes don’t appear anywhere in BLS data?”

      No. i know that the BLS includes end user prices. Because I understand economics – because I understand that prices will necessarily include coverage for all the taxes paid by producers (including the payroll and income taxes of its employees) – I know that all taxes are included in the values the BLS uses to derive CPI.

  9. Geoih says: “Large groups of economists regularly aggregate completely unrelated prices, based on the subjective valuations of individuals, and state manipulated money, and proclaim the all-wonderful CPI to the world.”‘

    Basically, you can say the same thing about the “State highway/transportation agencies” that built the Interstate Highway System.

    • “State highway/transportation agencies actually publish raw details.”

      Where? Show me. That sounds like it might be economically interesting.

  10. Bart, State highway/transportation agencies don’t publish their accounting worksheets, and neither does the accounting profession.

    Products and prices aren’t hidden. No credible source has been able to prove the BLS understates inflation.

    I’m sure there are lots of people who believe we never landed on the moon or Obama has done a great job on the economy.

    What does that prove?

  11. Those charts say nothing about an unbiased consumer price index or the general price level.

    You’d get more complete and more accurate information following the BLS methodology, including taking statistically significant surveys.

  12. “2. It is a myth that the use of hedonic quality adjustment has substantially reduced the growth rate of the CPI.”

    See! There’s some weaseling right there.
    “Substantially”, i.e. nothing to see here; move along.
    “Growth rate of the CPI”, i.e. the acceleration in the increase in prices. “Yes, prices are rising for the little people who make and purchase goods and services, but we’re working hard to keep the rate of increase within our vastly excessive target range, so shut your traps!”

  13. “you need to look at all prices”

    No. I only need to look at the prices of goods and services I “need” or “want” to buy. The others are irrelevant, to me, as is any price index which is heavily biased by them.

    If BLS weighs those other things too much, their index is distorted and non-representative of my range of options. it does not properly reflect the prices and trade-offs I face.

    Then again, a price index based only on things I buy, or would like to buy, would not reflect the market basket some/many others face.

    This connects with Hayek’s recognition that the multitude of individuals know more than any bureaubums could possibly know about what is significant in any economy at any time.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>