Thursday, August 22, 2019

Inflation, and history

Phil Gramm and John Early have an excellent WSJ oped on inflation measurement.

Many conventional inflation measures are  overstated by around 1% per year, because they don't capture quality change well. When the iPhone 22 comes out, and costs (say) the same as the iPhone 21, that looks like there is no inflation. But the iPhone 22 has dozens of new features, so you're really getting "more phone," and there is deflation. Think of all the things like mapping that are now free. It's not just tech. Houses are bigger and better, cars are much better and safer, and lots of services are now free
No price index takes account of the fact that each month more than twice as many people get medical advice from WebMD as visit the doctor’s office in America. 
The people who make price indices try hard to account for quality, but the result is still biased.

Now for many uses we don't get too upset about this, just as a clock that's always 5 minutes fast is still useful. In particular, for spotting trends in inflation, is it rising or falling, a steady 1% bias doesn't make much difference.

Where it does make a big difference is when you cumulate inflation over a long period of time. We  do that in the politically charged question, is the average American better off now than his or her parents were in 1975, and if so how much? Now whether 1% per year times 44 years is more inflation or better stuff matters a lot.
When corrected for documented price overstatements, real average hourly earnings from 1975 to 2017 are shown to have risen some 52%, not 6%—an additional $6.77 an hour. Real median household income increased 68%, not 21%—$17,060 more annually. 
And published poverty incidence fell by almost half. Combined with the 67% drop in poverty that comes from accounting for all government transfers, poverty incidence sank from 12.3% to about 2%. 
In the political debate over the fate of the "middle class" these are huge differences. The mantra-repeated narrative that the "middle class" hasn't advanced at  all  since the 1970s is simply not true. (Quotes because I reject the idea that we are a class society,  and hence the use of this word.)



Gramm and Early emphasize the other long-run issue: when one uses 1% cost of living adjustments in government payments, that too cumulates to dramatic numbers over long time periods
By overstating inflation, CPI indexes inflate real spending in Social Security, federal employee retirement, military retirement and poverty programs like Supplemental Security Income. Overstating inflation also overstates the minimum-income threshold to qualify for Medicaid, food stamps and other subsidies.  
If the tax code and transfer payments had been indexed to an even more accurate index embodying all known price-index overstatements, revenues since 2000 would have been more than $1 trillion higher, transfer payments $2.6 trillion lower, and the national debt 23% lower
Gramm and Early's main point: better indices exist. Why don't we start using them?
Since PCEPI and C-CPI-U overstate inflation less than other official indexes, it would seem reasonable to adopt one of them immediately for all historical comparisons, official statistical estimates, and inflation adjustments for taxes and spending
You know the answer, right? Because a lot of voters' checks depend on using an inflated number.

The other issue not mentioned is the great puzzle of our current monetary policy, inflation, and interest rates. Why are interest rates so persistently low? Well, if inflation were really two percent overstated, the world would make a lot more sense. Then we would really have positive two percent real interest rates, because inflation is really zero. I don't see evidence that it's that large, but it sure would make a lot more sense.

The next issue, that I think is much under-studied: the huge local variation in prices, and, via huge variation in what we consume, the inflation experienced by people in different parts of the country. Living in California and especially the Bay Area is like living in a different country with a different currency. Even gas costs twice what it does in the rest of the country. A lot of what appears to be income inequality is just different prices, and especially land prices. (Much of the productivity of tech workers went in to the pockets of existing land owners.) If you get paid 100 yen in Japan, you're not 100 times wealthier than someone who gets paid 1 dollar in the US.  Being paid $100,000 per year in San Francisco is something like that -- and being paid $20 per hour in much of the US is nowhere near the disaster than being paid that much in San Francisco would be.

How much local variation in prices is really variation in quality, and how much is just higher prices? Like the slow 1% bias in national inflation rates, local variation likely doesn't matter for short-run issues, but cumulates over time, so local  variation in the price level is large. And local variation in typical consumption baskets even larger. In my last look at this I could find local inflation rates, but not good price level  comparisons. But I haven't looked that hard.

(Gramm and Early write
Two recent studies, one by Bruce Meyer and James Sullivan and another by Brent Moulton, combine more than 50 credible studies documenting specific overstatements in consumer price indexes.
but the WSJ does not have links.  I can't quite figure out which of the many excellent Meyer and Sullivan papers they have in mind. Be prepared like me to lose some time browsing Meyer's website and Sullivan's website. I think this is Brent Moulton's article.)

14 comments:

  1. "real average hourly earnings from 1975 to 2017 are shown to have risen some 52%, not 6%"

    Which would be consistent with the fact that people now seem to be buying bigger houses and better cars, than they were in 1975.

    Apart from the problem of measuring inflation there is probably a problem with defining "inflation". If hypothetically the price of oil went up because the world was running of oil would that be "inflation" so that the Fed should tighten monetary conditions. If the price/rent for land goes up because the population has risen, is that "inflation" justifying higher interest rates and COLAs? It is possible that the definition of "inflation" should depend on the use we want to make of the answer.

    ReplyDelete
  2. Three big challenges to that point. 1. How do you even measure that? There is an incredible amount of subjectivity when it comes to "it's not inflation it's improvement". 2. For every item you show which has improved, I can give you an example of an item which deteriorated in quality. Think about clothes, they might cost less today but fast fashion means they are incredibly inferior. Think about food, compared with 30 years ago how much more cheap soy and corn make up your calories? 3. Why not look at measures which are less open to interpretation? Comparing 1975 to 2017, how many average salaries did a degree cost? How many average salaries did a house cost?

    ReplyDelete
  3. My son hated school and so after high school took a job as a laborer working for $9 or $10/hour a plumber. He lived with us 2 years and saved his money plus some college money he got from my father and bought a really nice condo for $44,000 cash (it would now go for about $75,000 (https://search.bosshardtrealty.com/results-gallery/?countylike=Alachua&proptype=C&status=A) but it would probably go for $500,000 in San Fransisco.) He is doing great.
    I think all these people complaining about inequality need to be looking at residential building restrictions and not much else. It's not capitalism it is your neighbors using Government to limit the right to build that are causing the squeeze. You should have a right to build pretty much any number residences on your land unless there is a very good reason.

    ReplyDelete
  4. It would be nice if a team of statistician-economists would prepare, publish, and maintain a conversion table showing the "dollar diminishment" factor: Starting with a base at some local place, such as Wright County, MO, the mean center of US population.
    The factor for San Francisco might be something like .55 and for Phoenix, .85 and perhaps for Houston something like .75 (my guesswork numbers).
    But absent a "separate currency" to measure the local "price level," this diminishment factor would be useful for job-changers.
    (I had to insist on a 100% salary increase in moving from Chicago to Washington, DC, in 1979 due to much higher local taxes - since I had been living in "rent controlled" Hyde Park, where the UChicago maintained a rent-fixing scheme among neighborhood realtors to discriminate against non-U apartment seekers [personal research, 1971-78]).

    ReplyDelete
    Replies
    1. I found an index State-by-State, but it is only for the tax hit. The map has it at the county or SMSA level.
      From the Tax Foundation: https://taxfoundation.org/real-value-100-metro-2019/

      Delete
  5. A recent QJE paper by Jaravel (https://docs.wixstatic.com/ugd/bacd2d_b55c96a87ae8431790eb037a963cc4ef.pdf) shows that the consumption basket for the bottom quintile has a 0.7-0.9 percentage points higher inflation rate than the one for the top quintile the last 15 years or so.

    Potentially, though the referenced study does not go into this, product quality upgrades disproportionately benefit top income earners too.

    Both is to say that its unclear whether the fact you discuss should lead us to question the "political mantra".

    ReplyDelete
  6. Has the possibility of inflation understatement because of quantity changes been considered? The old joke about a Hersey bar now being the size of a razor blade makes the point. Many producers seem to hold price constant but reduce product size/quantity.

    ReplyDelete
  7. There will always be measurement error. A price index that perfectly anticipates 'inflation' ('deflation') would be ideal, but it would not be practicable. Gramm and Early contend that 'inflation' is overstated by 100 basis points. When the 'true' (unobserved) rate of inflation is 8 percent per year, a 100 basis point measurement error is 12.5% of the true value. Not all that remarkable, even if as Gramm and Early argue improvement (i.e, reduction in the error rate) is possible. When the (erroneous) measured rate of inflation is on the order of 1 to 2 percent per year, an error of 100 basis points in the measured rate is more significant, on a relative basis--is the true rate of inflation 0 or 3 percent per annum?

    From a practical perspective, inflation running at 8%/year is a larger problem than inflation running at somewhere between 0%/year and 3%/year. Monetary policy responses will vary accordingly. If the policy target is 1% to 3% per year, the monetary authority can be expected to act strongly when inflation is 8%/year, and act complacently when inflation is between 0% and 3%/year.

    But this is not the point of the argument by Gramm and Early. Their position has two aspects: government welfare expenditures are too high when inflation is mismeasured, whatever the true rate of inflation is; and, welfare economics being of particular concern, apparently, to Democrats, mismeasurement of inflation and income distribution leads to greater political pressure favouring redistribution via government tax and spending legislation favoured by Democrats compared to lower government policy intervention favoured by Republicans. Correcting the perceptions arising from the effects of mismeasurement of economic variables, Gramm and Early contend, would improve the picture of the American economy and raise the spirits of the American people, almost surely.

    Monetary policy making might also become more effective and efficient, to boot!

    ReplyDelete
    Replies
    1. I think I have seen old comments (by Yellen perhaps) that one of the reasons for choosing a 2% inflation target was that with measurement errors and uncertainty, anything less than 2% ran the risk of tipping into actual deflation.

      Delete
  8. Hedonics is tough.

    How to measure housing costs along the West Coast, New York City, or Boston markets?

    Given that inflation is an amorphous concept and not possible to measure precisely, shouldn't we be happy with roughly good results on this score? That is, if inflation is under 3% that is good enough.

    ReplyDelete
  9. Regarding variation in local price level across the US, one source of data is the Association of Chamber of Commerce Executives cost of living index. It has been used by, for example, Crucini, et al. (2015 JME vol 74) to study LOP dynamics across US cities.

    ReplyDelete
  10. The troubles with "quality" are that it is often almost hopelessly subjective, and that it is almost useless when inflation is being judged in terms of "cost of living". The giant-screen TV illustrates both issues.

    Giant screens are now common, and have gotten (somewhat) cheap. On the other hand, who really seems to care? Sports broadcasts, for example, often feature a fairly small image in the middle, not much beyond NTSC quality, showing the game or contest, surrounded by a vast peripheral field of nattering rubbish occasionally including ads. And this "quality improvement" is no help at all to, say, millennials forced to live in unaffordable cities because billionaire bosses absolutely insist on locating jobs at prestigious, i.e. outrageously expensive, addresses. The giant screen is of little help in either case, no matter how impressive it might be to tech nerds.

    (And the array of nattering, dangerously distracting gizmos that has helped make new cars somewhat unaffordable is much like the giant screen. And the "cultural/arts" argument in favor of those expensive addresses falls flat since most megacity jobholders, falling ever further behind, have absolutely no time left over for such things. Life is just plain around 10x more expensive now in unadjusted dollars than some decades ago.)

    ReplyDelete
  11. Different price levels across cities: not full picture, but groceries in US cities.
    https://voxeu.org/article/systemic-price-mis-measurement-rethinking-ppp

    ReplyDelete
  12. These quality adjustment arguments fail to take into account the actual constraints in the market. The average consumer is forced to accept the “higher quality” of these products along with their higher prices. Official statistics then correct for inflation by informing consumers of all the great benefits that (i) no one asked for, and CRITICALLY (ii) there is no viable market alternative to.

    I always laugh when they report quality adjustments in cars, like air conditioning or shatter-proof windshields or airbags, to justify higher prices. Ask yourself if there are any cars available on any dealer lot in the US without features such as these. Ask yourself if our parents’ generation (maybe 20% of the population) really benefits from the iphone, or perhaps they would be just as happy with the 2005 Moto Razr. Then go back and tell them why they should pay for smart phones they can't use and don’t want.

    In some cases, inflation statistics are correcting for “quality improvements” that Federal agencies mandate, like airbags. You have no choice bu to pay the higher price for these! You may argue quality if life is higher, but that’s your opinion. The higher price is an observable fact, and these “quality adjustments” are nothing but instruments to hide inflation, they can make a Hyundai more expensive than a Mercedes. It’s absurd.

    More broadly, ask yourself why quality adjustments coincide so much with the sectors that have the highest inflation, like medicine and education. Have you looked at USA PISA scores lately? You think the kids in the US is getting higher quality education than 50 years ago? How’s our engineering? What about health? How are we doing there in terms of quality, cost, value for money, etc., etc.?

    The bottom line is that inflation is at its heart an index of the cost of living. While there should be some basic controls around what is being measured, you have it exactly wrong. The quality adjustments have done nothing but hide inflation.

    Oh and by the way, since you mentioned housing, so has rental equivalents (and substitution and other such “adjustments”). Check out US inflation in 2004 - at the height of the housing bubble. As it turns out, the housing component of CPI was lower than general CPI. Thank goodness that in 2004 housing (which is 35% of the basket) was pulling down overall inflation according to the official numbers. Awesome. I guess the housing bubble, student loans and medical bankruptcy are all just illusions. We should believe the reported statistics and not our lying eyes, and just enjoy the higher quality of life as we enter our 16 hour workday to pay these low prices.

    Rafael Romeu
    rromeu@devtechsys.com

    ReplyDelete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.