Corrected for Inflation

by Russ Roberts on March 24, 2006

in Prices

Studies suggest that measured inflation is overstated because of a failure to control for quality differences. (Here’s one example from Mark Bils at the University of Rochester.) Michael Boskin of Stanford chaired a commission a few years back that found (if memory serves) that the annual error is something on the order of 1% per year. That’s make a huge difference over time when trying to adjust wages or income for inflation to measure changes in the standard of living. If inflation is overstated, then growth in our standard of living gets understated.

USA Today reports on the 25th anniversary of the mountain bike and illustrates the fundamental problem of quality and measuring inflation. When the mountain bike was first introduced, it cost $750. Now, a mountain bike costs $1400. So if mountain bikes are in the CPI, you would say they have almost doubled in price. But as the USA Today article points out, that is less than the average price increase over the same time:

You can still buy a Stumpjumper today — at less than the original cost.
The 1981 price was $750, or $1,537 when adjusted for 25 years of

I’m not sure which inflation series the reporter was using. Any calculation I make finds that if the price of mountain bikes had risen at the rate of inflation, they’d cost over $1600.  But the main point is that when the economy-wide inflation rate is calculated, it includes items like the mountain bike. Some prices have risen even more than the price of the mountain bike. Some less. Some have even gone down. The CPI and the rate of inflation are an attempt to calculate an overall average of all goods in the economy.

Suppose the price of mountain bike today had gone up at exactly the same rate as the average of all other prices. Then you’d conclude that the mountain bike had gotten no cheaper and no more expensive in real terms over the last 25 years.

But then there’s the quality problem. As USA Today reports:

Nobody knows Stumpjumpers like Ned Overend. He
won the first-ever mountain bike world championship in 1990 on a
carbon-fiber prototype. In 2004, at age 48, he placed seventh at the
U.S. national championships on a Stumpjumper.

"The bike has changed a lot over the years,"
says Overend, who also won the 1998 and 1999 Xterra Mountain Bike
Triathlon world championships. "In 1990, we couldn’t have imagined the
bikes we ride today."

The first "superlight" Stumpjumper weighed just under 30 pounds, was made of steel and had 15 gears and no shock absorbers.

Most of today’s mountain bikes are 10 pounds
lighter and have a front shock and 27 gears. A FSR Stumpjumper, with
disc brakes and "smart" dual shocks that analyze the terrain, weighs 26

So the mountain bike is actually a lot cheaper than it once was, controlling for quality. As I understand it, the Bureau of Labor Statistics has tried in recent years to correct for these improvements. They use what is called "hedonic analysis." In this case that would involve trying to measure how much bike riders value the lighter bike and the shock absorbers and any other improvements and then deducting those values from the price to take the quality improvements out of the measure of inflation.

Obviously, those calculations are difficult to do accurately.

And the $1400 bike today is the low end bike. You can spend a lot more and get even more quality:

Several top-end bikes, such as the $7,100
Specialized S-Works Stumpjumper FSR Carbon, feature "smart" suspension
technology that reads the terrain and adjusts the shock-absorbers to
stiffen for smooth trails or climbing, then soften to suck up bumps and
holes. That bike is the heir to the original Stumpjumper, considered
the first mass-market mountain bike.

Unlike the rigid and finicky early models, the
new bikes allow less-experienced riders to tackle technical terrain
with less risk of a crash.

Other models, such as the $5,500 Trek Fuel EX
9.9, the $3,000 Kona The King and the $7,500 Scott USA Ranson Limited,
give experienced riders several on-the-fly shock settings to squeeze
extra energy from the terrain by compressing the shocks into berm turns
or trail depressions, then releasing that energy for a speed burst.

You can see the challenge for measuring inflation accurately. As we get wealthier, choices proliferate. But as we choose higher and higher quality variations as a result of that wealth, it gets harder and harder to measure how much wealthier we are in real terms. Throwing some of those more expensive bikes into the CPI category called "mountain bike" is really going to distort the estimate of the cost of living.


Be Sociable, Share!



76 comments    Share Share    Print    Email


Mcwop March 24, 2006 at 2:55 pm

Just exclude food and energy – nobody uses that stuff.

caveatBettor March 24, 2006 at 3:18 pm

Inflation is tough to nail down precisely, but mountain bikes are just like food, clothing, housing, telecomm, media, electronics … they all demonstrate we are wealthier across the board than we were decades ago. Even isolating by strata (e.g. lowest quintile of income or wealth, poverty) people in this country are way ahead in the goods and services consumed and enjoyed.

Robert Cote March 24, 2006 at 3:31 pm

What's the deflation factor for one's catching bovine spongiform encephalitis because of the modern "advances" in beef production?

johngaltline March 24, 2006 at 3:34 pm

…yet, somehow, the poor keep getting poorer.

dj superflat March 24, 2006 at 4:01 pm

ummmmmm, doesn't this post prove the poor are by no means getting poorer? and doesn't every study show that? poor people in the US have more space, TVs, cell phones, cars, etc., than ever. i hope you were kidding, but assume you're some fairly misguided lefty person bemoaning how the poor can't afford $7K mountain bikes (but now ride cheapo bikes that are way better than even a bike costing twice as much 20 years agao).

Don Lloyd March 24, 2006 at 4:26 pm


It is logically impossible to simultaneously try to track both subjective living standards and the purchasing power of the monetary unit with a single measure.

When the CPI is used to adjust government entitlement payments, this is an attempt to compensate for the loss in the purchasing power of the dollar primarily caused by the FED-induced expansion of the money supply.

The use of hedonic, quality adjustments in this application is generally a fundamental logical error. If you can imagine that all goods and services were to have their quality increased by just the appropriate amounts to leave all relative prices unchamged, and no monetary expansion existed, then it should be clear that absolute prices would also be largely unchanged and no adjustment for the purchasing power of the dollar would be appropriate.

A dollar would buy the same number of widgets, but the subjective standard of living would increase as widgets, along with every other good and service, would be improved in quality.

Hedonic adjustments are thus fundamentally flawed when used to attempt to adjust for the purchasing power of money. This is in addition to all the specific hedonic adjustment errors made where the 'quality improvements' do not align with their actual economic value in use. For example, multiplying both the speed and memory of a desktop computer by 100X would have little or no economic advantage in running a word processing program. You'd still be waitng for the typist and the printer. Nor would you expect the market price, primarily based on the subjective use-values of consumers, to increase as a result.

Regards, Don

liberty March 24, 2006 at 4:28 pm

Given the name/link, I think it was sarcastic.

Patrick March 24, 2006 at 4:28 pm

To move a little further out from the specifics of which type of mountain bike compares with a similar type today, the cost of mountain bikes in general is much less than in 1990. For around $500 you can get a mountain bike that, in my estimation, is far superior to the 1990 $750 model. And, if you are willing to accept the low end models (which will still carry you nicely over hills, through brushy trails and in and out of muddy potholes) you can pay as little as $100. The point of my comment is, beyond simple economic inflationary formulas, technology, innovation, lower manufacturing costs and the like have completely obliterated the meaning and purpose of much inflationary analysis.

Jay Godse March 24, 2006 at 4:28 pm

The Economist magazine did an interesting survey a few years ago comparing the relative price of goods and services from the year 1900 versus the year 2000. Interestingly, there were a number of goods and services that went up in price (in relative terms such as a percentage of the median or mean incomes), while another number of goods and services went down in price.

The common factor in all of this is that any good or service that could be electrified, motorized or computerized went down in price drastically, while labour-intensive products and service went up. That explains why telephone calls and banking transactions are much cheaper now, while health-care and education costs grew (and continue to grow) much faster than the rate of inflation. If my memory serves me well, potatoes went up in price, while ice-cream went down…draw your own conclusions.

This maxim suggests that health-care and education will continue to get more and more expensive as long as they remain very labour-intensive.

johngaltline March 24, 2006 at 4:40 pm

dj_superflat — I was *definitely* kidding.

The poor are getting richer.

Half Sigma March 24, 2006 at 4:59 pm

Going up in price: housing, healthcare, education, the very basic stuff you need.

Going down in price: mountain bikes, TV sets, frivolous consumer goods you don't need.

Stefan Karlsson March 24, 2006 at 5:05 pm

Correcting for quality changes is not problematic in principle, but difficult in practice, of course, as it is difficult to know just how the actual consumers value these changes.

Still, though, the most fundamental problem with your post is that it does not offer any evidence that the "hedonic adjustment" actually performed understates the value of quality improvements . Some people believe that they overestimate it, and that inflation is therefore underestimated. Until evidence of systematic underestimation of quality improvements have been presented, both claims must be considered equally plausible.

johngaltline March 24, 2006 at 5:25 pm

"Going up in price: housing, healthcare, education, the very basic stuff you need."

Not true; you missed the *whole point* of the post.

In fact, it's against the law to build a house like they built them in the 70s — it would fall short of code in more ways than I could count. The law has effectively changed the definition of "housing" to the point that you can't compare its price to the inferior product that was known as "housing" twenty or thirty years ago.

Similarly, it's against the law to give you 1970s-style medical care. Your doctor would get sued unless he gave you state-of-the-art treatment, and a lot of extra procedures that have become standard as a result of litigation. Government has effectively changed the definition of "healthcare" as well.

And like healthcare, education has basically turned into a third-party-payer system. In such a model, rising prices are not the same as "higher costs."

liberty March 24, 2006 at 6:58 pm

>Going up in price: housing, healthcare, education, the very basic stuff you need.

I will only add that although housing has gone up in price in many areas, I know that the median house cost is generally only higher because people spend a much smaller portion of income on food, appliances and other necessities (because they are cheaper) and so choose to spend more on – much nicer, larger – houses.

In most small cities you can still buy a house for $40k 2006 dollars. This price, inflation adjusted or compared against median income (avoiding the inflaton debate) is very very cheap. I challenge anyone to compare the house that you could get in 1970 with the house you could get today for the same percent of the median income.

Robert Cote March 24, 2006 at 7:15 pm

In most small cities you can still buy a house for $40k 2006 dollars.

OKAY. I pick my small city. Camarillo, CA pop 68,000. Please draw me concentric circles until you find a 2br, 1ba 900sf house on a 6000sf lot anywhere for $40k. I'm thinkin' you might get a outlier somewhere in Montana.

Cheapest SFR on today;

Even Taft is $80,000 and up.

Morgan March 24, 2006 at 11:46 pm


I've found exactly what you're looking for, only it's twice the square feet, twice as many bedrooms (make one a home office?), and $7,100 cheaper than your budget.

And it's not in Montana – it's right here in sunny Moberly, Missouri! Now granted, the night life isn't much…

quadrupole March 25, 2006 at 1:50 am

It's important to note that the health care and education are becoming tremendously more expensive because the government has basically broken the market. If you look at the sectors of medicine that are subject to market pressures (like laser eye surgery) you see exactly what you see in other areas of the economy: prices decline, quality increases. If you look at some of the movement going on around online charter schools, you are seeing rapid productivity growth in education beginning to come up.

Housing is an interesting question. I think there are basically two things going on. First, as Elizabeth Warren noted, there is a huge bidding war on for good schools. Since this has been pushed into the housing market (because our government has broken the education market), resulting in strange, inefficient pricing. Second, since land is finite, and some areas to live are more desirable than others, the cost in the highly desirable areas is rising fast. A previous poster did note that if you were willing to live in less desirable areas, you could find affordable housing. But many folks (including me) would prefer to pay more (if they are able) to live in a nicer neighborhood. That's a quality choice, just like choosing a nicer mountain bike.

Patrick March 25, 2006 at 9:46 am

I agree with quadrupole here. While I mentioned earlier that the mountain was not just cheaper than in 1990 but much cheaper the "healthcare and education is higher" mantra. But of course, with government involved in education there isn't the same incentive to make the system more competitive. In healthcare both the government and poorly run (a la public) institutions erode the efficiency. Case in point, my wife went in to a local hospital for outpatient surgery. Before the surgery she met with her doctor who completed a standard pre-op medical form (allergic to any meds, have any of the following diseases, meds you are taking, ever experienced reaction to anaesthegia,etc.) When we got to the hospital and checked in they ask her the same questions again and filled out the form, then the pre-op nursing team got her prepped and ask her the same questions on an identical form, the anaesthia guy did the same followed by the surgeon-that's the same set of forms 5 times. They were all paper forms so each of them had to be typed into different computer systems, separate bills and reports sent out to the insurance company, etc. Instead of a single automated system that passed the patient info from doctor to facility to surgical team they did it the hard way. That alone probably costs in excess of $500 for a $2,000 operation. On the back office side post-op God only knows what other inefficiencies were happening. The HIPPA rules seem to cause barriers in passing info on this seamlessly but password protection and data encryption would make compensation easy. Healthcare itself is locked in a overly complicated and bizarrely uncompetitve cycle. Less government and more patient control would help but no eliminate the problem. Still, healthcare is not a good example of inlfationary measure-it's just not in the system with the rest of economy.

Aaron Krowne March 25, 2006 at 9:48 am

Don Lloyd has it. Hedonic adjustments are a way to sneak in a "cost of survival" CPI, rather than a cost/standard of living.

There are two components to this problem. As I understand it, the two effects one wants to account for are:

1) quality/capability increases at the same price point (e.g., bikes with more gears and shocks)
2) down-substitutions to inferior goods (e.g. steak to hamburgers)

I agree with Don that if you account for #1, you are measuring something, but it really has nothing to do with the price index. Capabilities add to quality life in a very hazy way, and we are probably better off just assuming that technology will cause progress and leave it at that.

To be more explicit, it doesn't help me much that mountain bikes are advancing, because I bought my last bike nine years ago (ack!) and I can't purchase something like that on a routine basis. However, the cost of more routinely-purchased goods *does* matter to my bank balance. Further, if all bikes now include technological advances that have a nonzero cost, I do not even have the option of buying what would be a cheaper replacement for the older bike I have now. So there is a sort of conflation here that results in making the CPI look better than it is.

Trying to account for the downward-substitution, however, is more insidious. Taking the hamburger example, if you can no longer afford your weekly steak and instead must buy hamburger, how in the world can this be considered comparable? Clearly some loss of value has occurred here, and the clear proximate cause is rising prieces for the same set of goods.

Such debates are not the realm of abstract pontification. With energy pushing prices to rise across the board, I now find myself having to drink orange juice from concentrate because "real" orange juice is too expensive.

Somewhere under all the layers of manipulation, there's a real CPI moving up sharply, and you can taste the difference.

save_the_rustbelt March 25, 2006 at 10:24 am

Those poor people should just shut up, crawl into the attic and prepare to drown. They can take their mountain bikes with them.

We have plenty of real economic problems, so why spend so much time proving that poor old ladies aren't really poor?

spencer March 25, 2006 at 10:52 am

While I agree there are problems with measuring inflation, not all the problems are in one direction. For example, the Boskin report found that the main reason the CPI overstated inflation was housing.
So they suggested that the previous measure of home prices and interest rates be replaced by the home owners equivalent rent.
This measure acounts for 23% of the CPI as compared for 22% for goods excluding food and energy — the segment where most quality improvements occur.

Probably the best single measure of home prices is The House Price Index (HPI) published by the Office of Federal Housing Enterprise Oversight.

so if you compare what it reports for home prices compared to the CPI home owners equivalent rent measure you get interesting results.

Annual % change
2000 …….6.9………3.0
2001 …….7.9………3.8
2002 …….6.9………4.1
2003 …….6.8………2.4
2004 ……10.8………2.2
2005 …..13.2…….. 2.3

This certainly suggests that the corrections to the CPI following from the
Boskin Commission have moved in the other direction of causing the CPI to understate

My point is that you are taking all the arguments on one side of the argument that in and of their selves are valid arguments
and ignoring all the arguments on the other side of the argument that are just as valid.

For example one of the things that happens in healthcare is the redefinition of procedures. Medicare and the insurance companies will say that they are going to pay $x for a given procedure. So what do the doctors do. They split the exam or procedure into two steps so they can get paid the $x for the original procedure and bill for another sum for the new procedure that use to be part of the old procedure. But the inflation data does not capture this drop in quality.

liberty March 25, 2006 at 12:18 pm

>OKAY. I pick my small city. Camarillo, CA pop 68,000. Please draw me concentric circles until you find a 2br, 1ba 900sf house on a 6000sf lot anywhere for $40k. I'm thinkin' you might get a outlier somewhere in Montana.

You didn't exactly take me up on my challenge. You picked one small city – I said most, not all. California is notoriously expensive. Get some housing prices in a few small cities from 1970 and compare to the median income in 1970. I will do the same for 2006. If I cannot find houses as cheap in 2006 compared to 2006 median income, of equivalent quality then you win.

BTW, if you really think that your city is a typical example of housing prices today then you need to get out more. I live in NM and I know for a fact that I can find you bigger nicer houses here and in the midwest and the south as well as places like Montana for under 40k.

Max Born March 25, 2006 at 3:35 pm

I believe what is actually measured is the median 'spending' by people on a particular good or service. In other words inflation tells you how much more you need to make to keep up with your neighbours. And in that sense, it is a useful metric.

However, it really is a bad metric to compare prices of goods and services over the years. In fact, the best way to measure the price of goods and services is to go to India or Mexico or China- especially for goods. You can find bikes that cost $40. Now that is deflation. The same bike that costed $200 in the USA 20 years ago, and costed $200 in India- now costs $40 in India. Whereas in USA, we are willing to still pay $200 for the 'average' bike so the quality shifts upwards.

Ditto for shirts, pants, skirts.

Deflation is happening all around us. And at Wal-Mart, and in India and China. We just choose, for good reason, to move up the quality ladder.

My 2c. Inflation adjsuted. :)

Aaron Krowne March 25, 2006 at 4:37 pm


Because if CPI was still being calculated like it was in 70s and 80s, those poor old ladies' social security checks would have to be 70% higher. You know, like, somewhere around what they need to live.

And we couldn't have that.

quadrupole March 25, 2006 at 6:58 pm


It's not just the government spending that distorts medicine, it's also the government tax policy that moves most of us into third party employer provided health insurance, and the government discount requirements that inflate the retail price of health care.

Have you ever looked at your insurance statements for a medical visit/proceedure? I read mine. It shows the retail price, and then what was paid by the insurance company to discharge the bill. Typically the retail price is about 3 times higher. What this tells me is that the retail price is at least three times the market price for medical care, even at the current level of inefficiency.

The problem in both medicine and education is the third party payer problem. In both cases, it dramatically inflates costs and reduces quality. In both cases it's present because of misguided government policies. Fix those, and you'll see a race up in quality and a race down in price, just like in every other industry subject to health market forces.

Chris Meisenzahl March 25, 2006 at 10:41 pm

I've felt for a while that the entire concept of keeping track of the CPI with any remote level of accuracy is simply not doable. WAY too many variables and volume of goods/services.

nate March 26, 2006 at 7:01 pm

What about "spring break" trips and inflation, and related trends over time?

Half Sigma March 27, 2006 at 12:19 pm

"The problem in both medicine and education is the third party payer problem."

This is even worse than that in education, becaues the "free money" to pay for the education is often not actually free but a student loan that can't even be discharged in bankruptcy.

quadrupole March 27, 2006 at 9:09 pm

Half Sigma

I am deeply torn about how we do financial aid for college students today.

On the one hand, I see it leading to run away college spending and declining higher educaton productivity. I also see it encouraging young people to take on huge debts, often for degrees that don't even begin to repay those debts (does that English degree really pay back the $120k in debt plus the 4 years of lost earnings, plus the interest on them?).

On the other hand, I know a *lot* of folks who suffered from crappy parent syndrome, and made a lot of their lives because there was financial aid to allow them the opportunity.

I desparately want to preserve the accessibility of higher education, but without the productivity declines, and the deadweight loss of overproduction of underdemanded majors. I just don't know how to do it.

Nathan T. Freeman March 28, 2006 at 8:14 am

How the hell can any of you say that education is more expensive now? Everyone of you is getting a lecture series from the chairman of a major economics department for the cost of your computer and a connection to the internet — probably less than $75/month for those of you in the US. And you probably read the other GMU-related sites like EconLog and Marginal Revolution. You could probably spend an hour a day reading through each of these and get an undergraduate understanding of microeconomics — certainly one that's more useful than your typical state university's BS in Econ.

Is there anything you can't learn from the 'net these days? World-class thinkers in just about any subject offer up knowledge for free to anyone willing to type a few words into Google. There's tons of commentary to go with it, and it's easy to find a discussion group able to educate you on just about any topic imaginable.

Google, Wikipedia, blogs and Usenet have DRAMATICALLY lowered the cost of education over the past two decades, to the point where the most expensive component for just about anyone is the opportunity cost of the time you spend reading!

Okay, maybe it's more expensive to get CERTIFIED in some area of education now. I won't judge that. But surely readers of Cafe Hayek are smart enough to understand that "education" isn't whatever gets defined by some university administrator.

Really, argue all you want about housing and healthcare — but the cost of education has never been lower.

bbartlog March 28, 2006 at 8:47 am

Nathan -
the cost of educating yourself is indeed low now (nothing more than your time really), but then public libraries have provided similar albeit less efficient services for years. You seem to be deliberately ignoring important benefits of an actual university education above and beyond knowledge alone. First of all, the degree is a certification of knowledge which may provide a lot of value in the job market. Second, the university provides an environment for meeting a variety of people and developing a valuable social network. Lastly, it provides a venue for research and publication (try and get a grant or get published without university affiliation – not so easy).

For the rest, I would point out that as certain types of goods become cheap (manufactured widgets, information and entertainment, shipping, travel, perhaps food) there will be more money chasing the rest, so inflation in some categories should be expected – especially for goods in relatively finite supply, such as apartments in desirable urban areas…

Half Sigma March 28, 2006 at 10:29 am

"the cost of education has never been lower."

The value that society places on a certification has NEVER BEEN HIGHER.

Abraham Lincoln was considered a great lawyer, and he never even graduated high school. Today, one needs to have four years of college plus three years of law school in order to be legally allowed to practice law, and unless you attended a top law school, good luck convincing a major law firm to hire you.

I think it's a sad indictment of society that self-learning has so little value, yet that's the case.

The cost of getting certified has never been higher, partially because formal education requires high wage labor and the cost of high wage labor is increasing faster than the rate of inflation.

Third party financing of education contributes becuase there is no incentive for a college to lower costs.

And student loans hurt the average college student more than they help, because people who have no business going to college in the first place are being saddled with debts.

18 year old kids, who society perceives as not having sufficient wisdom to drink a beer, are nevertheless given the opportunity to shackle themslves with debt that can't even be discharged in bankruptcy.

Nathan T. Freeman March 29, 2006 at 3:32 am

bbartlog – "You seem to be deliberately ignoring important benefits of an actual university education above and beyond knowledge alone."

me – "Okay, maybe it's more expensive to get CERTIFIED in some area of education now."

Clearly you spent too much on your own certification.

Half Sigma, you've done a wonderful job of completely ignoring my point.

Half Sigma March 29, 2006 at 9:18 am

Nathan, your point was that people can learn stuff for free, and I didn't ignore it, I agree 100%, and explained that society only values the knowledge gained from expensive formal education.

TC March 29, 2006 at 2:56 pm

Prices go up and down for reasons (supply and demand) that have nothing to do with the purchasing power of the dollar. CPI is bogus. Some believe it understates true inflation by about one-half. The Fed is the culprit, not prices.


Nathan T. Freeman March 30, 2006 at 6:18 am

"explained that society only values the knowledge gained from expensive formal education."

Half Sigma, I don't know what society you live in. It's certainly not the same as mine, which has consistently valued and rewarded my knowledge, in spite of the fact that almost none of it is from formal education.

Nathan T. Freeman March 30, 2006 at 6:21 am

Ah… "Big Mike has a BS in Economics from the Wharton School of the University of Pennsylvania, an MBA and JD from Arizona State University."

No wonder you think formal education is so important.

Half Sigma March 30, 2006 at 10:35 am

I said that formal education is important to "society." Part of it is because laws and political correctness make uncustomary and maybe even ILLEGAL to use objective tests as a means to hire people, so college and graduate degrees are heavily relied on instead. This is a tremendously inefficient use of resources to make people spend 4 years in college and more years in graduate school just to demonstrate they have a high IQ.

Previous post:

Next post: