Homeownership in a Bubble: The Fast Path to Poverty?

Dean Baker and Simone Baribeau[1]

 August 13, 2003

Executive Summary

            Homeownership is widely promoted in the United States as both an end in itself and a good investment. The tax code has been deliberately tilted to favor homeowners over renters, through items like the mortgage interest deduction. In addition, there is a wide range of initiatives at all levels of government, and within the non-profit sector, which are intended to promote homeownership among low and moderate income families who might find it difficult or impossible to purchase a home without some assistance.

            While homeownership may be indeed be desirable in normal times, it is not clear that encouraging moderate income families to buy homes at present is a good strategy. There is good reason to believe that the nation is experiencing a housing bubble very similar to the stock bubble of the late nineties. Nationwide, the rise in home prices has exceeded the overall rate of inflation by more than 30 percentage points since 1995. This sort of run-up in home prices has no precedent in the post-war period. No economist has been able to put forward a plausible explanation for such a sudden run-up in home prices, apart from a speculative bubble.

            In metropolitan areas that have been especially affected by the housing bubble, the rate of price appreciation has been even greater. In the most affected areas (largely in California and the East Coast north of Washington, DC), home prices have risen in real terms by more than 70 percent over the last five years. These areas are likely to see especially large price declines when the bubble bursts.

            There is evidence that the bubble has affected home prices at all segments of the market. In Los Angeles, one of the most affected bubble areas, the greatest home price appreciation over the last four years has occurred in the zip codes with mid-range home prices.  Further, the increase in housing prices in zip codes with the lowest median home price is comparable to the increase in prices in zip codes with the highest median home price.   This suggests that the price of many relatively low cost homes has also been inflated by the bubble.

            This situation implies that the price of many of the homes that moderate income families may seek to buy at present are likely to tumble when the bubble collapses. It is entirely reasonable to believe that the price of some of these homes could fall 30 percent or more, when the housing market returns to a more sustainable path. This sort of price decline will leave many new homebuyers with negative equity and could imply enormous losses on the sale of a house. Such losses will be especially devastating for families who see homeownership as a key step in escaping poverty.

Introduction

            Most American families view homeownership as an important goal. In general, ownership provides a large element of security that is not available to renters. Tax laws also strongly encourage homeownership, providing tax deductions for mortgage interest and property taxes. In addition, owning a home has generally proved a sound investment, as home prices usually rise in step with the overall rate of inflation and capital gains on selling a home are tax free for middle and lower income families.

            However, homeownership is not always better than renting. There can be times, when homes are selling at abnormally high prices, in which a home purchase will virtually guarantee a loss to homebuyers. In this case, the conventional wisdom – that buying a home is solid investment – may not hold true, just as it was not true that stocks were the best long-run investment for retirement savings, when the stock market was hitting its bubble peaks of 1998-2000.

            With home prices nationwide having outpaced the overall inflation rate by more than thirty percentage points since 1995, it is likely that the housing market is now in a housing bubble similar to the recent stock bubble. While individuals who sell before the bubble deflates will do very well, families who buy their first home at bubble-inflated prices are likely to experience substantial losses. In many cases, especially where homes have been bought with very low down payments, the losses could be several times the size of the homeowner’s equity.

            For all but the richest families, a home is  the family's most important asset, and a large loss on its sale will be painful. However, for moderate income families struggling to escape poverty, bubble related losses on the sale of a home could prove devastating. In many cases, these families will be far better off delaying home purchases until after the real estate bubble has deflated.

The Basic Arithmetic of the Housing Bubble

            Through most of the post-war period, housing prices have increased at approximately the same rate as other prices. Table 1 shows the rental price component of the consumer price index and the overall index at approximately ten year intervals.[1] As can be seen, rental prices shown in column 2 of the table have generally risen at approximately the same pace as the overall consumer price index (CPI) shown in the third column. For the whole period from 1951 to 1995, the overall price index actually rose slightly more rapidly than the rental price index, although the difference is sufficiently small that it could well be attributable to measurement error. There were some periods in which there was some divergence in inflation rates, for example in the seventies, when rental prices grew somewhat less rapidly than other prices, but this followed a twenty year period in which they had grown somewhat more rapidly.[2]

Housing Share of Consumption and Relative Price Changes

Source: Bureau of Labor Statistics and Baker 2002.[3]

            The left column shows the share of household consumption spending that goes to pay rent or owner’s equivalent rent – the implicit value of the rent for an owner-occupied home. The share of consumption going to rent did rise sharply over the post-war period, nearly doubling between 1951 and the 1982-84 period. However, the growth in the share of consumption going to rent has slowed sharply since the early eighties, increasing by just 1.6 percentage points over the following decade. This point is important in assessing the evidence for a housing bubble, because it shows that very large increases in the demand for housing in the seventies, associated with baby boomers forming their own households, did not lead to a substantial run-up in housing prices Therefore, the much smaller increases in the relative demand for housing of the last two decades should not be expected to have had a substantial impact on housing prices.

            Figure 1 shows the path of real home prices and rental prices from 1975 through the first quarter of 2003.[4] It is important to note that these are real prices – the price of homes and rent has been adjusted for the rate of inflation in other items. This means that if rent and home prices had just increased in step with the overall rate of inflation, the indexes would remain flat at 100.

Figure 1

            Source: OFHEO, BLS, and author’s calculations, see Baker 2002.

            As can be seen, both home and rent prices somewhat outpaced the overall rate of inflation through the eighties. However, the most striking feature in this figure is the sharp run-up in real home prices since 1995. In the period from 1995 through the first quarter of 2003 (the most recent quarter for which the home price index [HPI] is available), the rise in the home price index exceeded the increase in the CPI-non-shelter index by more than 33 percentage points. This means that if home prices were to again have the same relationship to the price of other goods and services as they did in 1995, they would have to fall on average by 25 percent from their current level.

            There is one obvious explanation for this extraordinary run-up in home prices – a real estate bubble that was triggered by the stock bubble of the same period. The logic is straightforward. Many families that had large capital gains from the stock market began moving into bigger and better homes. The supply of housing is relatively fixed in the short-run, so the increased demand pushes up the price of homes. After a period of time, people become willing to pay more for homes because they see prices rising, and believe that they will continue to rise.[5]

            The comparison of the movements of the HPI and the rental index is informative. The rise in the HPI outpaces the rental price index throughout the period, but the rise in both indexes substantially exceeds the overall rate of inflation. The explanation for the gap in the two indexes is that homes are sold in a spot market – the buyer and seller have a one-time transaction. These prices will reflect the supply and demand conditions in the market at that point in time. However, many renters have long-term leases and stay in the same rental unit for long periods of time. A landlord cannot raise rental prices to adjust to the spot market if a lease is in effect, and even if a lease has expired, a landlord may be reluctant to raise rents as much as may be justified by real estate prices, for fear of losing a good tenant. For this reason, it may be expected that rental prices will follow sale prices upward, albeit at a somewhat slower pace.

            However, the figure shows a somewhat different pattern. In the last year and a half, the rate of rental inflation has slowed sharply, with rental prices actually falling in real terms in the last year. This decline in rental prices can be readily explained by the fact that the nationwide rental vacancy rate is currently at a record high.[6] This is entirely consistent with the housing bubble scenario. Soaring home prices led to a rapid growth in the supply of housing units. While people are still buying ownership units at prices that are near bubble peaks, an enormous glut of rental units has developed. This is putting downward pressure on rental prices. Eventually, declining rental prices will feed back into the ownership market as potential owners opt to take advantage of low rents and landlords with high vacancy rates convert their rental units to condominiums. In time, this process will burst the bubble, unless other events, like rising mortgage interest rates, bring the bubble to an end sooner."

The Implications of a Collapsing Housing Bubble

            Real estate markets are local, not national. The HPI gives a nationwide average for home prices, but there are many areas, mostly in the south and Midwest, in which there has not been a substantial run-up in home prices. However, the fact that some areas have not seen substantial run-ups in home prices means that the price increases in the areas affected by the housing bubble have been considerably larger than indicated by the HPI. The areas that have seen the largest run-up in home prices – primarily cities in California and in the areas surrounding Boston and New York City -- have experienced price increases of more than 80 percent over the last five years. Table 2 shows the ten metropolitan areas with the largest increase in home prices over the last five years (these are nominal increases – the inflation rate has been approximately 12 percent over this period).

Table 2

Areas With Most Rapid 5-Year Housing Price Increases 

Source: Office of Federal Housing Enterprise Oversight, 2003.

While some of the extraordinary run-up in prices in these areas may be real, reflecting the fact that they have become more desirable places to live in the last five years, it is likely that much of this price increase simply reflects the psychology of the housing bubble. In many of these bubble markets, most of the price appreciation of the recent past is likely to disappear when the bubble bursts.

            It is also worth noting that the housing bubble appears to have affected all segments of the market. There have been many accounts from realtors suggesting that the most expensive houses have experienced the largest run-up in prices (e.g. “Housing ‘Bubble’ May Not Exist, but the Market Could Still Teeter,” Washington Post, September 20, 2002, Page E1).[7] Figure 2 shows the movement in home prices in Los Angeles over the last eight years. The figure shows the change in home price by zip codes, with zip codes with the ten lowest, ten median, and ten highest median home prices grouped together.[8] As can be seen, the zip codes with home prices at the median have experienced the sharpest increase in prices over this period. Prices in the zip codes with the least and most expensive homes have seen price increases that are comparable to each other, with both considerable less than medium-price homes. This suggests that even poorer parts of the Los Angeles metropolitan area have been affected by the housing bubble. This can be good news for long-time homeowners who are in a position to sell their home at present; however, it can be very bad news for first time homebuyers who buy into this bubble-inflated market.

Figure 2

        

Source: Authors’ calculations, see footnote.

Low Income Homebuyers in a Bubble Market

            In keeping with a widely shared goal of promoting home ownership, there are a variety of programs at all levels of government, in addition to initiatives from the non-profit sector, which are designed to facilitate home ownership among low and moderate income families. These programs offer assistance in a variety of forms – such as low- interest mortgages, mortgage guarantees for less than standard down payments, and matching support for down payments.  They all have the intention of assisting moderate income families in moving from being renters to homeowners.

            However, in a bubble-driven housing market, it is not clear that this is a desirable outcome. In some of these markets, home prices could very easily decline by 30 percent or more. Something like this happened in the Los Angeles housing market between a bubble peak in 1990 and a subsequent trough in 1996. Housing prices fell by 20.5 percent over this period – before adjusting for inflation[9] If a family had bought a home for $150,000 at the bubble peak, this price decline implies that it would have been worth just $119,250 at the trough six years later. If this family had placed a down payment of just 5 percent – as is allowed in some cases – it would have been in a situation where it owed nearly $132,500 on a home that was worth $119,250. In this case, the mortgage payments would be a huge albatross dragging down this family. The temptation to walk away from such a mortgage – turning over the home to the mortgage holder -- would be enormous, although the implications for its access to credit in the future could be severe. Since transaction costs on home sales are typically in the neighborhood of 7 percent, the implied loss in this situation is just under $50,000, an enormous blow to a family with a moderate income.

            This situation is outlined in table 3. It shows total home ownership costs including some additional assumptions. The table assumes that the mortgage payment is based on a 30-year, 6 percent mortgage; that annual property taxes are 1.0 percent of the originally assessed value, and that maintenance costs average $40 per month or $480 per year. In this case, the total cost over six years for a family who bought a home at the peak of the earlier Los Angeles real estate bubble in 1990, and was forced to sell at the trough in 1996, was $111,573, approximately $18,600 per year. In other words, if this family could have found comparable rental housing for less than $1,450 per month, it would have been better off renting over this period.

Table 3

Costs of Homeownership in a Bubble

Source: Authors’ calculations, see text.

            While this situation is highly stylized, many families would have likely ended up as losers if they purchased homes at the peak of the 1990 bubble, even if they did not end up selling at the very trough. Prices had already fallen sharply by 1994, so even people who had sold before the absolute bottom would have suffered large losses. It is also important to note that current bubble is considerable larger in many metropolitan areas than the 1990 bubble was in Los Angeles, possibly setting the stage for considerably larger losses.

            The Los Angeles housing market did bounce back from the 1996-97 trough and homes prices are now higher, adjusted for inflation, than they were at the 1990 peak. However, this should be viewed as grounds for caution rather than seen as a basis for reassurance. The one-day stock market crash in 1987 could be contained and eventually reversed by effective action by the Federal Reserve Board. However, the Fed could not sustain the record price to earnings ratios that the stock market hit at the peak of the bubble in 2000. In the same vein, the fact that a late eighties housing bubble was succeeded by a late nineties housing bubble is not a guarantee that a third bubble will come along to rescue victims of the second bubble. There is no reason to expect that bubble-driven prices will ever return to their bubble peaks – except over a very long period as inflation gradually drives prices upward. Homebuyers in a bubble – especially low and moderate income home buyers-- are putting their economic security at serious risk.

Conclusion

            There has been an extraordinary run-up in home sales prices in the years since 1995, with the rate of increase in home prices exceeding the overall rate of inflation by more than 30 percentage points over this period. There is no obvious explanation for this unprecedented run-up in housing prices, except that the wealth created by the stock market bubble spilled over into a real estate bubble, as happened in Japan in the eighties.

            The bubble has not been even across the country. Many areas in the mid-west and south have seen little real increase in home prices. By contrast, the areas most affected by the bubble, which are largely on the west coast and the east coast north of Washington, have seen home price increases that exceed the overall rate of inflation by more than sixty percentage points. In at least some of these bubble areas, the bubble appears to have affected all segments of the housing market. For example, in Los Angeles, zip codes with average home sale prices have actually experienced the largest run-up in home prices over the last four years, and the increase in prices in zip codes with the lowest home prices has been comparable to the increase in prices in zip codes with the highest home prices.

            The fact that many homes in bubble areas -- including those that could potentially be purchased by moderate income families – are selling at prices that are far higher than the levels that will likely be sustainable in the long-run, means that many families could lose large amounts of money by buying a home in the bubble-infected markets. Efforts by the government or non-profit organizations to promote homeownership in this environment could prove counter-productive, as families may see wealth accumulated through years of sacrifice disappear quickly with the bursting of a housing bubble. While homeownership may still be a desirable goal for moderate income families, in many areas they may be best served by delaying their plans until real estate prices have returned to more normal levels.