May 28, 2021 Reading Time: 10 minutes

Understanding “hedonic” adjustments

The currently reported high rate of consumer price inflation is grabbing considerable attention. We are not surprised by this development, as we observed in an earlier note. In that article, we argued that the inflation rate was likely to rise to the 3.5-4% range for the foreseeable future due to predicted increases in the intensity of the use of money and credit, also known as increased velocity of money.

A month later we observe that the current CPI increased by 4.2%, year-over-year. This surprising increase is being analyzed and decomposed from every possible angle to argue that it is transitory; that is, mostly one-off in nature. Which means the data do not augur a “persistent” upward trend.

The Fed is tasked with the “dual mandate” of minimizing the rates of inflation and unemployment. For inflation, the Fed’s stated objective is to adjust monetary policy so as to target an average inflation rate of 2% over time. The index of price inflation used by the Fed is referred to as the “core” inflation rate of personal consumption expenditures (PCE – a component of the GDP accounts). By “core,” they mean all prices except energy and food prices, which are considered too volatile and mean reverting, for the most part. It is further believed that there is a close measurement similarity between the “core” CPI and the “core” consumer expenditure price index.

Though the Fed targets the inflation rate, there is much more to price inflation measurement than just collecting and averaging price data. There are many adjustments and conventions used, of which quality adjustments is a major one. For those concerned with having a deeper understanding of what the Fed is targeting, in this article we explore the effect of “hedonic” (a code word for quality) price adjustments the government makes to the raw price data it collects in the field.

Other adjustments that can influence the reported inflation rate, which we do not discuss here, include differing weights assigned to the price components, the scope of expenditure categories, and the treatment of rent. The CPI uses fixed weights that are revised infrequently. The PCE inflation index measures adjustments in the weights that change from quarter to quarter based on shifts in consumption patterns. The PCE index covers a wider scope of expenditure categories. Both indices introduce a controversial measurement of the imputed rent paid for residences owned by the occupant.

Hedonic (Quality) Price Adjustments

The purpose of this economic commentary is educational and to draw attention to aspects of the measurement of price inflation that are confusing, logical, illogical, and infuriating all at the same time. Right below the surface of the raw data reported for prices paid by consumers there are all sorts of adjustments made by government statisticians. Among them, adjustments are made to normalize for quality improvement in the aspects and features of a product. To do so they use a statistical technique called ‘Hedonic’ price adjustments. Those adjustments reduce the apparent rate of inflation that is seen in posted prices. To what end are these adjustments made? Do they make sense from a consumer’s point of view versus the government’s point of view?

In competitive markets, all aspects of goods offered in commerce are freely evaluated and prices are freely contracted. Too big, too small, too round, too flat are examples of features that might enhance or detract from the price. In free markets, it is easy to determine whether various feature differences have value.

Putting a value on each of the basic features of a base-model car is largely nonnegotiable, because it is a package deal, but the US government tries to do that job for us in ways that will surprise you.

Hypothetical Examples

To help understand the consequences and significance of the issue of government estimates of value, two questions are posed. These questions involve hypothetical examples that illustrate the commonsense problems associated with government adjustments to price data.

When is a price increase not a price increase?

If a butcher increased the price of a pound of steak by 10% and told you it had less fat and 10% more protein per pound, would you think that was a price increase or no increase at all because he was selling you a 10% quality improvement? The answer depends on whether you want and value that quality increase – an increase in protein. 

The perplexed consumer might say, “I don’t want an increase in protein content. I like my marbleized steak just the way it was. If you can’t sell it to me then I’ll go next door and buy the old-fashioned style steak there.”

In this case, from the consumer’s perspective of having free choice, the qualitative change in the nutritional composition of the steak was considered a PRICE INCREASE! He valued taste over protein.

But what if he had no choice and the government mandated that all beef have less fat and a 10% higher protein content because that was better for your health and better for the climate? In the absence of choice, our hypothetical consumer would be forced to buy it. Or maybe he would buy only nine-tenths of a pound so as to hold his outlay to the same dollar amount as before. After all, if it were only proteins he wanted, he could make do with a smaller piece of meat.

What would the government report for the steak component of the CPI? The government’s perspective would be that there was a quality improvement to the consumer’s health and the health of the planet. Therefore, it would report that the price of steak did not go up at all – THERE WAS NO PRICE INCREASE. Hence no inflation!

Can you be impoverished by “quality” improvements?

Everyone has a different sense of “value” based on their own subjective preferences. If consumers find less value in the quality improvements to the CPI assigned by the government, then these changes can, in fact, make them poorer.

We know that most people live on a budget based on their paycheck. Money will only go so far. Let’s assume, for a basic example, that an individual has a monthly budget of $1,000 to spend on only two items: food (meat), and housing (rent). Meat takes up $250 of the budget, and the rest is for rent, or $750.

Now assume rent goes up by 10% to $825 due to government-mandated quality improvements (e.g. the landlord installed a new energy-efficient air conditioner, which the local rent control authority said was a quality improvement and justified a 10% rent increase). Since one can’t fractionally reduce the size of a home, and moving can be quite expensive, the person’s rent outlay goes up by 10% or $75. The amount remaining in the budget for meat drops by $75 which results in a 30% drop in money available for meat.

Meanwhile, the price of meat was increased by 10% due to the supposed “quality” improvement mentioned earlier. This poor consumer would be impoverished by quality improvements over which he had no choice. He would have to reduce his meat consumption by $75 due to the rent increase and then by another 10% due to the protein-related price increase. His meat consumption would be driven down by 37%.

So, what is the true state of our hypothetical consumer? What happened to his standard of living?

According to the government statistical method of adjusting prices for quality improvements, the consumer experienced NO INFLATION. In fact, he was 10% better off. After six months of this statistical betterment our hypothetical consumer would be skin and bones and quite certain that his standard of living had fallen.

If Mr. Hypothetical had a 10% pay increase, perhaps then all would be well. By government standards, he would be eating healthier meat and living in the luxury of efficient air conditioning. Government statistics would actually show that he experienced a REAL INCREASE in his standard of living – a 10% pay increase and ZERO inflation. 

Of course, in this example the consumer had no choice. His spending went up by the same amount as his pay increase. He doesn’t feel better off.

But the reality is even worse. It’s really bad news for Mr. Hypothetical because his pay raise is tied to a CPI-based COLA. He won’t get a cost-of-living adjustment because the CPI didn’t go up. NO PAY INCREASE.

Does this sort of thing actually happen with U.S. Government CPI data? Yes, it does. Just ask the retirees living on Social Security if, over the past few years, their benefits have risen to match the cost of living increases they faced.

Reality – Some Quality Adjustments are Sensible, Others Not 

When it comes to quality adjustments to prices, there is some room for appreciation. Here are a few clear-cut examples. First is an example where Hedonic Price adjustments for quality improvements have wiped out decades of price increases but the adjustments never work the other way.

Well-known standard car models have more than doubled in price over the last 25-30 years. For example, a basic Ford Mustang has risen by 200% and a basic Honda Accord is higher by slightly more than 100%. Yet the component of the CPI for new cars has barely changed – NO AUTO PRICE INFLATION. The reason given is that it is all quality improvements, not price increases. The quality improvements over time include such things as:

  • Longer power-train life and warranties
  • More horsepower
  • Catalytic converters (good for the environment but less horsepower)
  • Improved mileage efficiency
  • More cargo space
  • All-wheel/4-wheel drive
  • Anti-skid locking brakes
  • Air bags everywhere
  • 3-point safety belts, front and back seats
  • Power brakes
  • Rear view camera
  • Lane departure warnings
  • Blind side warning lights
  • High intensity/LED lights
  • … etc.

Because of these many improvements, the CPI reports that the average price of new cars has risen by much less than other prices in the economy. Over the past 25 years (1995-2020, the period when hedonic adjustments were introduced), the CPI for new vehicles has risen cumulatively by 4.9% compared with 70% for all items in the CPI. Even though the price of a Honda Accord, with all its new features, more than doubled in price, its quality adjusted price has risen by far less than other items in the economy.

Most of these improvements are appreciated and desired and have improved the quality of the car in ways, more than just safety wise, that mitigate much of the price rise. But how can we tell if the bureaucratic quality adjustments are in any way related to the “qualities” consumers want?  Try and buy a new vehicle with only the old 1990 technology. Without being able to do that there is no market-based foundation to tell where real inflation ends, and quality improvements begin.

Many consumer goods have improved in quality in ways that are difficult to measure. If one thinks about computers, cellphones and TVs, all of which have gone down dramatically in price according to the CPI, there may be less disagreement. For many people cellphones have simplified life and combined the functions of separately purchased devices such as a computer, a camera, a portable video player, and a Walkman; all that and more are now available on a basic smartphone. But are these improvements accurately categorized in CPI adjustments? If not, the CPI may be largely overstating inflation due to the much better products consumers can buy at relatively lower prices. The quality adjusted components of the CPI says telephone hardware and calculators have fallen by 87% since 1997 and computer prices have declined by 97% after all quality adjustments are made.

There are some reverse examples where quality has worsened yet an adverse quality adjustment to price doesn’t result in a higher CPI price. For example, downsizing airplane seats and the addition of water-saving regulators in showers, toilets and dishwashers. Greater discomfort doesn’t seem to count as a price increase due to the reduction in quality. As for water regulators, ostensibly water is saved but consumers complain they have to shower longer to fully rinse, flush toilets twice, and rewash some dishes.

Where Does Truth Lie?

Overlooking the technological improvements in computers, cellphones, and automobiles would cause the CPI to overstate the true degree of price inflation. On the other hand, excessive quality adjustments could understate inflation if these “improvements” are not judged by consumers to be true measures of quality. But how can we know if the government’s quality ratings are accurate?

Fortunately there is a critic of the government inflation data who believes that quality and other factors in the CPI result in greatly understating the true rates of consumer price inflation. John Williams publishes Shadow Government Statistics. His classic chart shows what the CPI inflation rate would be without all the adjustments the government makes to the data and compares it with the official inflation data.

consumer inflation official vs alternate
Courtesy of ShadowStats.com

Two points stand out from the chart above: (1) Unadjusted prices are rising much more rapidly than what the official data report; (2) the growing wedge of effects due to implicit quality improvements and other adjustments, which occurred mostly in the 1990s. In 1990 the difference in inflation rates was about 2%; by 2000 it had widened to nearly 7%. In 2020, it was close to 8%, so the difference has remained fairly stable over the past two decades. ShadowStats.com reports that the April 2021 CPI, which the government reported at 4.2% (year-over-year), was actually 12.2% stripped of all adjustments.

Even if the ShadowStats numbers are correct, they seem to represent a different type of inflation than what concerns most Federal Reserve Bank economists.

Is the Fed Targeting Prices or Quality Improvements?

These adjustments to prices create much more uncertainty about inflation than people realize. So how can one disentangle all the effects of statistical adjustment so as to know what to expect – how to read the official data on the rate of inflation? What is the Fed actually targeting?

Of course, the Fed is not explicitly targeting quality improvements. To do so would imply that the Fed expects quality improvements to continue at a rate sufficient to offset all but 2% of posted price increases. That seems a stretch. What if the pace of quality improvements faltered? Would that justify raising the benchmark interest rate?

Even if we can assume that the Fed was able to adjust monetary policy so as to achieve its 2% inflation target, how seriously can we accept it as relevant to the average American’s cost of living since it is a crafted number, which is the outcome of narrow definitions and statistical manipulations?

Are We Better Off By Having a 2% Target?

Given this measure of inflation, what are the actual effects on US consumers from achieving the target? Do the quality adjustments really make us better off?

There is an old expression in the investment industry: ‘You can’t eat relative performance; you can only eat if there is positive performance.’ In the world of price inflation measurement, the folks adjusting the data for quality improvements are telling us that those quality improvements mean our money is going further because our costs aren’t rising rapidly. Thus, our budget is better off. We can eat more steak.

But that is not true for all goods or all consumers.

The quality adjustments to the CPI may not accurately measure the true cost of inflation faced by many consumers. It does not measure changes in the actual cost of living. This may be especially true for those whose budgets favor goods that are less technological such as food, rent, and healthcare.

Understanding government data for what they are and what they are not is vital to understanding the impossibility of targeting the cost of living, even though the Fed is targeting a crafted measure of price inflation.

Gregory van Kipnis

Gregory van Kipnis

Gregory van Kipnis is Chairman of the Board of the American Institute for Economic Research. He was President and CEO of Invictus Partners, a statistical arbitrage hedge fund manager from 1997-2007, prior to that he was EVP at Jefferies & Co., in charge of proprietary trading from 1993-1997; Managing Director of NatWest Financial Products (London) and Executive Director of County NatWest (London) responsible for derivatives issuance and proprietary trading from 1990-1992; and Principal at Morgan Stanley responsible for proprietary statistical arbitrage trading, 1985-1990. His earlier career was as an economist and research director at Donaldson Lufkin & Jenrette (1973-1985) and IBM Corporation 1966-1973. He studied with Ludwig von Mises at New York University where he obtained his MBA in economics and finance.

Get notified of new articles from Gregory van Kipnis and AIER.