Housing Prices and the Consumer Price Index

From www.mcgeepost.com .Copyright © 2013 Michael H. McGee. All rights reserved. Please feel free to share or re-post all or part non-commercially, hopefully with attribution.

The annual rate of inflation in the United States and in most other countries is measured by the Consumer Price Index (CPI). The CPI shows that the United States has had a very low and manageable rate of inflation, with prices increasing an average of less than three per cent a year since the 1980s. It’s all good. People can still buy what they need at a reasonable price. The dollar is still sound in the marketplace.

So why are we weeping? Why has the owner-occupied residential housing market crashed? I offer here one suggestion which may have had more impact on the recent recession than most people have realized.

The Consumer Price Index is used almost exclusively as the measure of inflation in the United States. It is based entirely on a “basket of goods and services” most commonly purchased by retail buyers for personal use. Eight major groups of consumer products are measured. The US Bureau of Labor Statistics, within the Department of Labor, has compiled the CPI since 1913.

The CPI is compiled by looking at price changes in eight major categories of consumer spending. One of these major categories is defined as follows: “HOUSING (rent of primary residence, owners’ equivalent rent, fuel oil, bedroom furniture)” [emphasis supplied]

Changes in the costs of items on this list have over time actually done a reasonably good job of measuring the rate of inflation in our economy as a whole. The only problem is that the contents of this list have not remained the same over time, and therein lies the problem.

You will note in the items included in the CPI under HOUSING: “owners’ equivalent rent.” In this seemingly innocuous phrase lies what may be one the root causes of the housing bubble and the lending crisis which hit the US in 2007 and following years.

In footnotes to the list of items included in the CPI, the Bureau of Labor Statistics makes a full disclosure that, and I quote: “The CPI does not include investment items, such as stocks, bonds, real estate, and life insurance. (These items relate to savings and not to day-to-day consumption expenses.)”

From 1913 to 1982 the actual increase in the value of owner-occupied housing had been included in the Consumer Price Index. This was the “asset-price approach,” which treated the purchase of a home as a consumer good.

In 1983, however, the increase in value of owner-occupied housing was removed from the CPI, and replaced by the term “owners’ equivalent rent.” The new measure relates to a statistical estimate of how much a homeowner would have to pay to rent a home.

The statistical estimate is not a “real” number, but a mathematically derived construct based on certain theoretical assumptions. The use of these assumptions since 1983 means that in practice the “owners’ equivalent rent” tends to artificially track the “rent of primary residence” function, and bears little relation to what consumers actually have to pay to live in a home they have purchased.

The rationale for the 1983 change was that (1) it was too difficult to measure the increases in the prices of owner-occupied homes, and (2) the increase in the value of a home was an “investment item” and therefore did not belong in the Consumer Price Index. The details and timing of the change are given in “Changing the Treatment of Homeownership in the CPI,” by R. Gillingham and W. Lane, http://www.bls.gov/opub/mlr/1982/06/contents.htm .

According to the chart below, house prices remained in line with the “owners’ equivalent rent” element of the CPI until 1999. The purple line on the chart, labeled “Owner’s Equivalent Rent,” is taken directly from Consumer Price Index figures, and therefore accurately states how the CPI measured the increase in prices of owner-occupied housing after 1983.

The dark line on the chart, labeled “House Price Index,” is not an accurate statement of the rise in value over time of owner-occupied housing. The numbers are taken from the S&P/Case-Shiller Index, and probably significantly understate the rise in value of owner-occupied housing.

The S&P/Case Shiller index measures all existing single-family housing stock. Rental units occupied by tenants are included. New construction, condominiums, co-ops, and multi-family dwellings are excluded. You cannot have a clear measure for the CPI unless all types of owner-occupied dwellings are included, and rental properties are excluded. I borrowed this chart only to illustrate the trend, not as an exact statement of fact.

At least from 1999, and possibly earlier, the price of housing went up like a moon rocket, while the measure of “owners’ equivalent rent” used in the CPI increased at only a rate of less than four per cent a year. This meant that the CPI rate of inflation stayed within limits tolerable to the Federal Reserve Board, while at the same time owner-occupied home prices rose spectacularly.

The removal of the value of owner-occupied housing from the Consumer Price Index in 1983 in fact created a squeeze on homeowners and homebuyers after 1999. Wages, benefits and the amounts earned by entrepreneurs tended to rise slowly, based roughly on the increases in the Consumer Price Index. At the same time the cost of purchasing owner-occupied homes went up through the roof, as it were, without being in any way tethered to the cost of living or to the rate of inflation.

So when it came time for a consumer to buy a house, the cost of the house far outstripped the income available to pay the mortgage. Lenders had to come up with all kinds of special types of loans to enable consumers to buy homes on incomes which were largely tied to the CPI. Even starter homes began to cost more per month than the average person could pay.

Lenders were forced to abandon the “non-speculative ten per cent down thirty year fixed rate mortgage with a total monthly payment, including taxes and insurance, of no more than a fourth of their income.” At the same time huge amounts of new money became available to lenders, along with instructions to lend it whatever the circumstances.

The natural result was that lenders gave easier terms to for the most part good people, and created new types of loans. Worried by their own efforts to help their customers, they passed on their “hot potato” risk by selling their more conforming loans to Fannie Mae and Freddie Mac; and by selling tranches of their riskier loans to investors as mortgage-backed securities. The collapse was inevitable. The only question was when. Now we know.

Now I suspect that if you asked most consumers, they would say that owning a house was more about giving them a place to live, and any increase in value during the time they owned it was gravy. Most owners who bought or already owned homes during the period after 1999 took the gravy. They treated their homes as additional sources of income, through refinancing and second mortgages or equity lines of credit. Most of those who sold their homes during the same period used the profits as a down payment on a more expensive home.

During all the time both before and after the 1983 change in the CPI standard regarding home ownership, purchasing and owning a home has always been a part of what each family spends on normal expenses. Since at least 1970 more than sixty per cent of consumers have owned the homes they lived in (See the chart below). These homeowners are also consumers, and one of the things they consume on a daily basis is housing. Owner-occupant homeowners are for the most part purchasing as retail buyers for personal use.

The purchase price of owner-occupied homes should never have been removed from the Consumer Price Index. I say this with twenty-twenty hindsight. It was probably a good idea when it was done in 1983. This change, however, created a ticking time bomb starting no later than 1999 when conditions in the lending market changed.

There’s a fair consensus that the collapse of the owner-occupied housing bubble, along with the associated collapse of mortgage-backed securities and other housing debt, is what caused the recent recession. It doesn’t get much simpler than that. Could the housing bubble from 1999 to 2007 have been averted? Not with the information the Fed had at the time. Was their information faulty? As I have just explained, yes, their information was faulty.

The toxic and nearly fatal recession which began in late 2007 had its genesis in a bureaucratic policy decision made in 1983, as we have seen. The decision was benign when it was made, and remained benign until 1999 (or earlier), when conditions in the home lending market changed. From that point it was only eight years until the biggest financial downturn since the Great Depression. No one saw it coming until it was too late. The collapse of the housing bubble beginning in 2007 is what caused the recession.

The Federal Reserve Board (Fed) largely relies on the Consumer Price Index (CPI) in setting its Federal Funds Rate, which is its primary tool for fighting inflation. Since the increase in the purchase price and value of owner-occupied homes was not a part of the CPI in 1999, the Fed did not see it as a problem when the prices of owner-occupied homes began to skyrocket, even during the recession of 2001-2002.

In fact, Fed Chairman Alan Greenspan, who I assume was speaking for the whole Board, seemed blissfully unaware of any problems in the owner-occupied housing segment. In testimony before the Congressional Joint Economic Committee on November 13, 2002, he graciously took a bow when congratulated for keeping inflation at under three per cent for the last few years, and under four per cent for his whole tenure. He noted that the Fed was reducing the Federal Funds Rate even further, thereby making mortgages more readily available to consumers.

I’m not saying Mr. Greenspan did anything wrong or made any error in this analysis! It’s too easy to blame others using twenty-twenty hindsight. He was working within the established definitions of the Consumer Price Index. So was everyone else. The increase in the “investment value” of owner-occupied homes was at the time seen as a great benefit for the average person, adding to the net worth of individuals as well as giving them greater purchasing power through such vehicles as home equity loans.

As an “investment vehicle,” the owner-occupied home became more and more speculative and untethered to economic reality. And since the owner-occupied home was not a part of the CPI, no one was minding the store when home prices took flight. Consumers have paid a high price for the 1983 “theoretical” change in the CPI. And the crisis in mortgage loans took down the rest of the economy as well.

The solution, of course, is to restore the cost of owner-occupied housing to the Consumer Price Index. Right now the prices of owner-occupied homes have declined to a point where a switch at this time would not cause too much of a change in the CPI rate of inflation.

The real stumbling block to change in the CPI is that the Federal Reserve Board has no authority whatsoever over how the CPI is compiled. Thus a turf battle could slow down the ability to make such changes as may be necessary. The Consumer Price Index has been compiled by the Bureau of Labor Statistics, within the Department of Labor, for the last hundred years.

According to Janet Norwood, former Commissioner of the BLS, the CPI was “originally developed for wage adjustment,” which was well within the purview of the Department of Labor. As we have seen, over time its use has expanded to where it is used as the primary measure of monetary decisions affecting the whole economy.

Yet the Consumer Price Index still sits in the Department of Labor, and we see the effects of the unintended consequences of a seemingly benign 1983 decision. It’s likely that institutionally the Fed has never even questioned the elements used in compiling the Consumer Price Index.

It would make better sense at the present time in history for the Consumer Price Index to be compiled by the Treasury Department in coordination with the Federal Reserve Board. Until such a change is made, though, the Fed should immediately go to the Bureau of Labor Statistics and request the necessary changes needed to compile an inflation rate which more accurately reflects the experience of consumers in all areas of the economy, including owner occupied housing.

According to the Gillingham and Lane study cited above, the major problem the Bureau of Labor Statistics had before 1983 in calculating the House Price Index was that it was difficult to establish realistic price changes in owner-occupied housing over time. They did all their computations in-house and there were too many variables, they said.

The problem of difficulty in establishing realistic price changes in owner-occupied housing resulted from trying to take into account too many irrelevant factors, which necessitated dozens of surveys of various elements of housing cost, including mortgage interest rates. If we focused solely on the increase in median owner-occupied home prices from year to year, the statistical difficulties would evaporate.

It is highly probable that we could come up with a composite index of increases in value of owner-occupied housing over time, using the current level of computer sophistication which was not available in 1983. For example, I expect that the national Board of Realtors could generate a computer-generated report of the actual median sales prices for most owner-occupied housing in the country.

Making the recommended changes will have a huge long-term impact on homeowners. Their homes will no longer rise in value like a moon rocket. It’s likely that homeowners are feeling so battered right now that the change will be palatable. If we let the prices of owner-occupied housing rise again, without these price increases triggering anti-inflation measures such as raising the federal funds rate, we’ll be setting ourselves up for another economic crisis.