The Numbers are Broken

by Russ Roberts on September 13, 2005

in Data, Standard of Living

NOTE: I’m no longer convinced about the claim I make at the bottom of this piece about the CPI incorrectly accounting for increases in housing costs.  When I get the logic straight, I’ll post again.

Since 1997, real GDP has increased by over 20%.  Yet the average person seems to be standing still.  From today’s Wall Street Journal (rr):

When viewed over a longer time period, 1997 to
2004, wages beat inflation, rising 19.9%, compared with a gain of 17.7%
for the CPI.

That’s pitiful.  In that eight year period, the average worker gained 2% while the economy was going nuts.  Where’s the rest of the 20% growth in GDP going?  According to some, the richest eleven people in America, Bill Gates and ten of his friends just keep all the gains for themselves.  They have a system figured out that’s so secretive no one else knows about it.

But average people, even poor people, seem to be doing much better than they once did.  What’s going on?

Here’s an alternative theory: the numbers are broken.  Not in the usual way.  The usual way is to talk about fringe benefits or composition effects that I’ve written about before.  But there’s a bigger problem, a more basic problem.  And here I have to give a hat tip to my Dad who has been pushing this theme for a long time.

The CPI does not measure the cost of living.

Everyone in economics knows the standard problems with the CPI.  It tries to measure the increase in the cost of buying a particular bundle of goods.  It’s hard to deal with quality improvements and inevitably, the CPI understates inflation because it ignores people’s ability to substitute away from goods that have gotten more expensive.  The BLS tries to fix both of these problems periodically.  They try and adjust for quality and every once in a while, they change the weights on the goods in the bundle to take account of people making different choices than they did in the past.

But the biggest problem and one that I suspect is responsible for making real wages appear stagnant when people’s living standards are rising, is how to handle housing.

When housing prices are rising, are you better off or worse off, all things held equal?

Answer: it depends who you are.  As a first approximation, prospective home buyers are worse off.  Home owners, about 70% of the population, are better off.  But the CPI treats the gain to home owners as a cost.  Economics teaches that even if you own your house, there is an "opportunity cost" to living in it.  This idea argues correctly that you can always sell your house, so as your house gets more valuable, the cost of living in it rather than selling it rises.  This is true.  But there is a paradox.  The cost of living in it has gone up, but unlike most price rises, this one makes you better off, not worse off. 

Before I moved to Washington, DC, the salary that would induce me to move here was affected by the high cost of housing in DC.  But if housing prices double over the next five years (as they have in the previous five), will my standard of living fall if my salary fails to keep pace?  No.

About 22% of the CPI uses an implicit rent for homeowners.  How the BLS does it is extremely complex and is by its nature, something of a wild guess.  What they try and do is estimate how much your house would cost you to rent it.  So when rental rates of similar homes of equal quality go up, they use that percentage increase as a measure of housing inflation for homeowners.  But that’s wrong.  An increase in rental rates (usually reflecting an increase in housing prices) actually indicates a higher standard of living for the homeowner.

Some people (here and here for example) actually argue the exact opposite.  They argue that the CPI understates the cost of living because instead of using housing prices, it uses the cost of renting, and rental costs haven’t kept up with housing prices:


But it would be absolutely absurd to use changes in housing prices as representative of changes in the cost of living because 70% of Americans own their own house.  For those folks, rising housing prices improves your standard of living in the form of higher wealth.  But even using rental rates is wrong because a 3% rise in the rental rate of houses is actually a benefit to the homeowner who doesn’t pay that rent out of pocket.  It merely represents a higher opportunity cost of owning vs. renting.

I know, there are lots of complications to this argument, lots of caveats and lots of footnotes.   But the bottom line is that to correctly account for the impact of housing prices on my well-being you would have to take account of depreciation and taxes and maintenance and capital gains and expected capital gains.  Too complicated.  You can’t leave housing out of the index.  That would make the index meaningless.  But the index as currently estimated is a poor measure for deflating my salary and particularly poor when housing prices and rental rates are rising steadily.

Be Sociable, Share!



Add a Comment    Share Share    Print    Email

Previous post:

Next post: