Does the Government Understate Inflation? — Bill Gross Says Yes
Kurt Brouwer June 18th, 2008
In the first post in this series (Does the Government Understate Inflation?), we focused on the Consumer Price Index (CPI) and what it measures. We also covered some of the problems faced by government statisticians as they try to measure what a representative family would buy and how much inflation they are actually experiencing in housing, food, clothing, transportation, medical care and all the other purchases they make. If you have not read it yet, I’d recommend you read that post first.
Today, many pundits, portfolio managers and others think the CPI understates the true cost of inflation. And, as we face soaring energy and food bills, most of us would agree that the cost of living seems to have gone up more than 4% in the past year. Yet, the argument that the CPI miscalculates inflation is not of recent vintage. In fact, it has been raging quietly for years. There are many more adherents of the ‘CPI understates inflation’ crowd than there are of the ‘CPI overstates inflation’ crowd. And, those who believe the government understates inflation are far more passionate about their views than those who think the government overstates inflation. However, both points of view exist and both sides have interesting arguments in favor of their position.
The ‘Understates’ Point of View:
Well-known bond manager, Bill Gross, believes the CPI understates inflation. Let’s take a look at his most recent thoughts on this topic [emphasis in the original]:
Hmmmm? (PIMCO Investment Outlook, June 2008, Bill Gross)
…Let me reacquaint you with the debate about the authenticity of U.S. inflation calculations by presenting two ten-year graphs – one showing the ups and downs of year-over-year price changes for 24 representative foreign countries, and the other, the same time period for the U.S. An observer’s immediate take is that there are glaring differences, first in terms of trend and second in the actual mean or average of the 2 calculations. These representative countries, chosen and graphed by Ed Hyman and ISI, have averaged nearly 7% inflation for the past decade, while the U.S. has measured 2.6%. The most recent 12 months produces that same 7% number for the world but a closer 4% in the U.S.
Source: Pimco / ISI Group
A big part of Gross’s argument rests on this chart, which is a foreign CPI composite. One oddity I noticed is that this composite does not include India or China, which together represent 40% of the global population. I believe they should have been included even if that meant dropping Estonia, Latvia or Vietnam. The BBC recently reported, for example, China’s inflation hit a record:
…China’s inflation hit 8.7% in February, the highest rate in over 11 years, the National Statistics Bureau said.
I don’t think Chinese inflation at nearly 9% hurts his argument, but results for China and India certainly should be included in any summary of global inflation in my opinion.
Gross continues:
…This, dear reader, looks a mite suspicious. Sure, inflation was legitimately much higher in selected hot spots such as Brazil and Vietnam in the late 90s and the U.S. productivity “miracle” may have helped reduce ours a touch compared to some of the rest, but the U.S. dollar over the same period has declined by 30% against a currency basket of its major competitors which should have had an opposite effect, everything else being equal. I ask you: does it make sense that we have a 3% – 4% lower rate of inflation than the rest of the world? Can economists really explain this with their contorted Phillips curve, output gap, multifactor productivity theorizing in an increasingly globalized “one price fits all” commodity driven global economy? I suspect not. Somebody’s been foolin’, perhaps foolin’ themselves – I don’t know. This isn’t a conspiracy blog and there are too many statisticians and analysts at the Bureau of Labor Statistics (BLS) and Treasury with rapid turnover to even think of it. I’m just concerned that some of the people are being fooled all of the time and that as an investor, an accurate measure of inflation makes a huge difference.
Source: Pimco / Bloomberg
Let’s pause for a moment because the next section may be rough going. It’s important though because it spells out the key elements in the argument made by Gross and others who believe the CPI understates inflation. He contends that changes made to CPI over the past couple of decades have resulted in a long-term understatement of inflation. Essentially, he does not trust the way the government calculates the index, largely because other countries have not adopted these methodologies. He points to three main adjustments that were made, some of them dating back decades:
- Quality adjustments: If a car has more features or is more reliable than a previous version of the same automobile, should that ‘quality’ difference be measured and, if so, how?
- Housing adjustment: The current method surveys home owners (owners’ equivalent rent) to determine what the owner would have to pay if he were renter in the same home.
- Geometric weighting / product substitution: We’ll ignore geometric weighting for now because that is a math wonk question. Product substitution has to do with buying a similar though not identical product. For example, if steak prices go up, then a shopper might buy hamburger instead.
Gross continues:
…The U.S. seems to differ from the rest of the world in how it computes its inflation rate in three primary ways: 1) hedonic quality adjustments, 2) calculations of housing costs via owners’ equivalent rent, and 3) geometric weighting/product substitution. The changes in all three areas have favored lower U.S. inflation and have taken place over the past 25 years, the first occurring in 1983 with the BLS decision to modify the cost of housing. It was claimed that a measure based on what an owner might get for renting his house would more accurately reflect the real world - a dubious assumption belied by the experience of the past 10 years during which the average cost of homes has appreciated at 3x the annual pace of the substituted owners’ equivalent rent (OER), and which would have raised the total CPI by approximately 1% annually if the switch had not been made.
In the 1990s the U.S. CPI was subjected to three additional changes that have not been adopted to the same degree (or at all) by other countries, each of which resulted in downward adjustments to our annual inflation rate. Product substitution and geometric weighting both presumed that more expensive goods and services would be used less and substituted with their less costly alternatives: more hamburger/less filet mignon when beef prices were rising, for example. In turn, hedonic quality adjustments accelerated in the late 1990s paving the way for huge price declines in the cost of computers and other durables…
In 2004, I claimed that these revised methodologies were understating CPI by perhaps 1% annually and therefore overstating real GDP growth by close to the same amount.
This is the heart of Gross’s claim: he believes that changes made to the CPI have resulted in a protracted understatement of inflation by 1% per year. According to Gross, this understatement has persisted for decades.
The Importance of the CPI Calculation
Moving beyond the nuts and bolts of the CPI calculation, Gross points out why CPI is so important. GDP growth is generally calculated after inflation (real GDP), therefore the impact of understated inflation would be to overstate GDP growth. In addition, Social Security and other government programs often have cost of living calculations that are based on CPI. If the CPI actually should have been higher, then those payments would have been higher also. In addition, Treasury Inflation-Protected Securities (TIPS) use the CPI to adjust for inflation. So, we need to use the best information we have to calculate CPI correctly.
And, quite frankly, the Consumer Price Index (CPI) calculation — like many government statistics — is convoluted and complex. It is calculated by the Bureau of Labor Statistics (BLS) and, for decades, it was a quiet bureaucratic backwater. Now, it is required to be the ‘measure’ of inflation and this subjects both the CPI itself and the government statisticians who create it to plenty of criticism. Further, like Coca-Cola, we now have umpteen versions. CPI, Core CPI (that is, CPI excluding food & energy), Personal Consumption Expenditures (PCE) and more.
I can see that the changes Gross outlined could have understated inflation. However, I would have liked a lot more meat in his analysis in these areas:
First, does he think we should make adjustments for quality at all? As I pointed out in the first post, cars are significantly safer and more reliable now than they were in the 1980s. Should CPI ignore quality improvements? Second, I’d like to understand how he arrived at the 1% understatement figure. Is that a precise measurement or just a ballpark? Third, Gross criticizes the method for measuring inflation in housing because it understated inflation when home prices were soaring,. Yet, if the owners’ equivalent rent system led to an understatement of inflation when home prices were rising, is it overstating inflation now that home prices are falling? Finally, he criticized the concept of product substitutions. In real life, we all adjust our buying when one item seems a bit high. Does he think we should just rule out product substitutions and have a fixed basket of goods we measure?
As you can see, Gross’s argument raised more questions in my mind than he answered. I’m sympathetic to the argument he makes because I have so much respect for him. One additional problem with his argument is the concept that the government has been understating or hiding the true inflation rate for decades and the financial markets just have not figured it out. That seems unlikely to me.
As you can imagine, I’m not ready to join the ‘understates’ crowd, but I recognize the importance of the discussion and I hope the Bureau of Labor Statistics and other government agencies are taking a hard look at this issue. Also, it may be that there is an opportunity for a private entity to step in and provide a better measurement of inflation. Who knows?
In our next episode, we will look at the viewpoint of some who believe the CPI actually overstates inflation. Stay tuned…
See also Part One: Does the Government Understate Inflation?
- Business , Economy , Geopolitics , Money , Personal Finance
- Comments(5)
Did you enjoy this article?
What is the goal of a CPI number? Once that is agreed upon, it becomes easier to make choices about baskets of goods and quality improvements. If the goal of a CPI number is to give politicians some insight into the present quality of life of their constituents (so as to avoid the bread to cake substitution problem that historically leads to decapitation) then equivalent rent, substitution, and quality improvements make a certain amount of sense. If, on the other hand, the goal is to provide decision makers with good data to plug into their models so as to predict the future (ostensibly what they are trying to do with their soft landings and whatnot) I’m not so sure.
Here’s a riddle for ya: A certain farmer needs a trench dug and employs a bunch of men with shovels to dig it. He pencils in a ledger that it took 365 man-days to dig the ditch. A number of years later he again needs a similar ditch dug. Progress during this time has invented the mechanical shovel (backhoe) and the farmer employs a man with his backhoe to dig the ditch. It takes this man one day to dig the ditch and the billed cost is the equivalent of 50 man-days when compared to the first digging. 20 of these man-days went to the operator, and 30 man-days represented the cost of providing the backhoe.
What effect would this capitalization of ditch digging have on the CPI?
Both Owner Equivalent Rent and the Case-Schiller index are flawed measures of housing inflation. Neither actually measure how much families are paying for housing.
The Case-Schiller index tells you how much a family pays for new housing, but very few houses are new. Most families are making payments on houses that they purchased in prior months or years. As such, the Case-Schiller index over marks for current housing price changes.
Owner Equivalent Rent isn’t any better. It is a more stable measure, but it doesn’t include any information on the cost of buying a house. Essentially what it does is increase the weight for rental price changes.
What the CPI really needs is an index that tracks actual housing payments made. Such an index would have been increasing during the housing bubble but would not be falling as sharply as the Case-Schiller index now. It would reflect the reality that two houses with identical assessments might have widely different costs, because the mortgage snapshots a house’s value at a particular point in time.
Such a housing payments index would have measured inflation higher for the last five years (and presumably led to a tighter monetary policy). It might show inflation slightly lower now, but not by too much until a number of mortgages have a chance to be financed for the new house prices.
[…] Part Two: Does the Government Understate Inflation? — Bill Gross Says Yes […]
[…] Part Two: Does the Government Understate Inflation? — Bill Gross Says Yes […]