As analysts continue to digest the data being produced by the U.S. Census, what happened to the United States housing markets in the past decade begins to look all the more remarkable. For the first time since the Baby Boom hit the housing markets in the late 1970s, regional and local housing market differences are more important to consider than the overall level of demand. Although many analysts believe that it will take another five years in most urban markets to house all of the starter households left over from the Baby Boom, nevertheless, market to market differences have grown so profoundly that one wonders it it is even logical to talk about a "national" housing market: national housing statistics, yes, national housing market, no. In fact, if there is anything that the Savings and Loan debacle has taught the capital markets, it is that credit support of real estate projects is nice, especially if the guarantor is Uncle Sam, but that local market demand for what is being financed is just as important. Investing in housing in the 1990s will he characterized by radically different demand-supply equations that operate in individual markets. A comparative analysis of the housing characteristics of six major states should serve to highlight the investment challenge of the coming decade.
U.S Housing Characteristics
There was a slight drop in the nation's owner-occupancy rate during the 1980s. Apparently most of the drop in this key rate occurred before 1985. It should be added however, that the household formation rate dropped after 1985, so it is likely that the two factors were interrelated. In other words, if the household formation rate had been higher throughout the 1980s, the owner-occupancy rate would have dropped more than it did. The drop during the 1980s in the owner-occupancy rate resulted in the equivalence of approximately 200,000 households renting rather than owning by the time of the 1990 census. Ironically there were over 300,000 more vacant owner units in 1990 than in 1980. Apparently developers knew of the demand for owner units; it is just that developer prices (and locations) were wrong. As disappointing as these statistics are, the rental numbers are even worse.
During the 1980s, the nation's capital markets produced so many unneeded rental units, that by 1990, there were 833,000 more vacant rental units than in 1980. Overall, the nation ended up with 1.6 million more empty housing units for the decade. Of this number, 30% were second or seasonal units, 20% were owner-occupied units, and most of the rest were rental units. Clearly, the rental supply system was not paying attention to the constraints put on demand by household income. No one can argue that there are plenty of Americans who might want to live in all these empty units, but just as with empty owner-occupied units, housing quality and income go hand in hand.
A relentless population boom, the anti-growth spin-off of Prop 13, and scarcity of developable land, have combined to give California the most severe housing situation in the nation today. Although the owner-occupancy rate in the state held its own during the decade, household size in California went up, while the national trend was going down. In other words, people were doubling up to make ends meet. By now, it would take 900,000 additional owners units to give California the same owner occupancy rate as the rest of the Country.
While California was short approximately 31,000 owner-occupied households for the decade, at the same time the state ended up with 5,500 more vacant owner units in 1990 than in 1980. Significantly, vacant rental units increased by 78,000 units. And of course, investors in California rental projects are beginning to recognize that there is a significant oversupply of expensive rental units there. Even with this oversupply of rental units however, the increase in vacant units in California amounted to just 1.1% of the total 1990 housing stock, well below the national rate of 1.57%.
The most seriously negative aspect of the supply-demand equation in California is the cost of housing units. Owner-occupied prices went up nine times faster than the national rate. In 1990, the median home value was 230% of the U.S. value; rents were 150% of the national median These numbers indicate that most of California had joined the ranks of those select urban markets in the U.S. where cost has become its own housing problem.
It is interesting to see how rapid growth and the scarcity of land has put California in the ranks of the really expensive housing markets. This means that California is now in the same category as New York. Interestingly enough, New York showed a surprising gain in its owner-occupancy rate. Perhaps for the first time in history, the state's owner-occupancy rate exceeded fifty percent. In fact, New York gained 239,000 more owner units than would have been the case had its owner-occupied rate not increased. It should be pointed out, however, that it would take another 800,000 owner units to put New York even with the national rate.
The negatives in the New York equation include an oversupply of 22,000 owner units in the past decade, as well as an increase in vacant rental units of 31,000. In addition, prices for owner units increased much too quickly, equivalent to twelve times the U.S. rate. Rental rates increased faster than the national average as well. Overall, however, housing remains in relatively short supply, with the excess of vacant units produced in the 1980s totalling only 0.6% of the total housing inventory.
Florida had a good owner-occupancy rate in 1990. At 67.2%, it compares very favorably with the U.S. rate of 64.2%. However, during the 1980s, the state's owner-occupancy rate declined by 1.1 percentage points, equal to a decline of 56,000 units. However, the decline in the owner-occupancy rate did not necessarily mean that there was a shortfall in production. In fact, there were 57,000 more vacant owner units in 1990 than in 1980. Apparently, the mismatch once again has to do with price, since Florida prices increased somewhat faster than the U.S. rate and by 1990 were almost comparable to the national median price, even though Florida incomes are lower.
The increase in vacant rental units was more significant, with an additional 100,000 units added during the decade. Rents in Florida were raised faster than the U.S. rate and are now 8% higher than the national median. Overall, nearly 200,000 additional vacant units were added to the Florida housing stock during the past decade, an amount equal to 3.1% of the total housing stock. This is a vacancy rate nearly double the U.s.average.
As with Florida and California, there was significant overbuilding in the Texas rental markets. At a time when the Texas owner-occupancy rate was dropping nearly three and one-half percentage points, 127,000 additional vacant rental units were added to the state's housing stock. Even though owner-occupancy was off by 200,000 households, there were 276,000 additional vacant units produced during the decade. This is as serious a mismatch between demand and supply as exists among the fifty states. The additional increase In vacancies was equal to 4% of the Texas housing stock. Unlike in other states, Texas housing prices and rents both dropped during the past decade, with housing prices well below the national average and rents significantly lower than the U.S. figure. Nevertheless, it can be assumed that prices and rents didn't drop far enough to match the lower income levels in the state.
During the 1980s, the owner-occupancy rate went up in Pennsylvania to 70.6%. The state gained another 32,000 units for ownership, which meant that by 1990 the state was well ahead of the national norm, to the tune of nearly 300,000 households. It should be noticed, however, that there was some significant increase in vacant units in the state during the 1980s, with 43,000 vacant units added.
Equally significant is the fact that housing prices and rents want up faster than the U.S. rate, although both prices and rents ended up well below the U.S. 1990 medians. It appears that without the pressure caused by population growth, Pennsylvania managed to improve its housing condition, without pushing housing costs too far beyond affordable levels.
Like Pennsylvania, Ohio housing prices and rents remained well below the U.S. medians during the 1980s. Significantly, owner prices actually dropped 11% during the decade, finishing at 80% of the U.S. median. Rents in Ohio rose at about two-thirds the national rate, also finishing up at 80% of the U.S. median.
Unlike Pennsylvania, the owner-occupancy rate fell slightly in Ohio during the 1980s, down to 67.5%. And approximately 37,000 households ended up renting rather than owning. Nevertheless, Ohio's owner-occupancy rate was still three percentage points above the U.S. average, equal to having 134,000 more owner households than the national average would allow.
Most interesting about the Ohio numbers is the fact that there was a net decline of 3,500 owner vacancies during the 1980s. Because there was an additional 7,100 vacant rental units produced, overall the state experienced a very modest increase in vacant units, equivalent to only 0.2% of the 1990 housing stock. The increase in vacancies in Ohio was only 12% of the national increase in vacancies. It can be assumed that underproduction was probably the rule in the state, a fact which bodes well for future housing production, providing that the units produced are owner-occupied and carefully matched to market demand.
All in all, the decade of the 1980s produced a dazzling array of market differences among the states, a trend well worth studying in every state. State housing agencies and institutional investors should heed what these census numbers are telling us, and that is that housing is a preeminently local enterprise. The tax-exempt market has a unique opportunity to take advantage of these local differences. Unlike the other capital markets in this country, the municipal market has kept the following maxim of investing foremost in its collective mind: when it comes to housing finance, credit quality and demand are always interrelated.
[TABULAR DATA OMITTED]