In the second quarter of 1997, the production and sale of new homes was down from the first quarter when favorable weather encouraged buyers and builders to act early. But the ups and downs between quarters should not obscure the remarkable strength of the housing sector. Housing production and sales continue at high levels; interest rates remain under 8 percent; housing affordability hovers at a high level; and the homeownership rate is at a 17-year high.

Housing production is high, although it was flat or off slightly in the second quarter compared with the first quarter.

• Permits, at 1,423,000 on a seasonally adjusted annual rate (SAAR), are above 3 out of the 4 quarters in 1996 and above 6 of the past 9 quarters. The 1997 second-quarter rate is the same as the actual rate for 1996, which was the highest rate in 8 years.

• Starts, at 1,440,000 (SAAR), are within 3 percent of the actual 1996 level, which was the best in the past 9 years.

• Shipments of manufactured homes, at 349,000, are only 4 percent off 1996's all-time record level.

Sales of new and existing single-family homes also continued at high levels, with sales of existing homes slightly ahead of last quarter and sales of new homes off from last quarter.

• Sales of existing homes were at 4,150,000 (SAAR), which is 1 percent above last quarter but down 2 percent from last year's second quarter. However, the May 1996 sales rate was the highest ever monthly rate.

• Sales of new homes, at 785,000 (SAAR), were off 5 percent* from last quarter but were up 7 percent from last year's second quarter.

• The new home median price, at $144,700, was unchanged from last quarter and up 3 percent* from last year's second quarter. The median price of an existing home was $123,700, up 4 percent from both last quarter and last year's second quarter.

• Inventories of new homes for sale at the end of the quarter were down a slight 2 percent* from last quarter but down 21 percent from 1996's second quarter. Inventories of existing homes for sale were down 12 percent from last quarter and down 7 percent from last year's second quarter.

• Placements of new manufactured homes continue at high levels in the first quarter with 330,000 homes placed, down a slight 1 percent* from the fourth quarter of 1996.

• Builders' views of housing market activity are mixed. They are more upbeat about current sales than they are about future sales and prospective buyer traffic.

Low interest rates and favorable affordability have led to a near-record high homeownership rate. Interest rates at 7.93 percent for 30-year, fixed-rate mortgages rose 14 basis points from the first quarter but are 18 basis points below last year's second-quarter rate. Affordability -- which is determined by combining house prices, interest rates, and consumer incomes -- remains at a high level, although home prices have had their impact. The family earning the median income and purchasing the median-priced existing home had 125.2 percent of the needed income, down 5.2 percentage points from last quarter but equal to the second quarter of last year. The U.S. homeownership rate, the focus of a national partnership, rose 0.3 percentage point* to 65.7 percent in the second quarter of 1997 -- the highest quarterly rate in nearly 17 years.

Multifamily housing (5+ units) performance was also somewhat mixed in the second quarter. Starts and market absorptions were up, but permits were down and vacancies were up.

• Permits, at 302,000 (SAAR) units, were down 3 percent from last quarter but up 7 percent from the second quarter of last year.

• Starts, at 293,000 (SAAR), were up 14 percent from last quarter and up 15 percent from last year's second quarter.

• Seventy-four percent of the 37,600 apartments completed in the first quarter of 1997 were rented within 3 months compared with 71 percent* of apartments completed in the previous quarter.

• Rental vacancy rates increased to 7.9 percent in the second quarter from 7.5 percent in the first quarter, but the change is not statistically significant.

*This change is not statistically significant.

Regional Perspective

While single-family permits were down 9 percent in the Southwest, Dallas-Fort Worth and Houston-Galveston continued to report strong homebuilding and sales activity. In the Denver area new homes priced between $100,000 and $125,000 move fast, but builders have difficulty producing in that range. Affordability has become an issue with businesses seeking to locate in the Denver area. In California the improving economy resulted in a 9-percent increase in single-family permits during the first 6 months of 1997. Increased activity was reported in the San Francisco Bay Area and in Southern California. In Arizona and Nevada, the markets remained strong, and construction and sales for all of 1997 are expected to equal last year's strong performance. In the first half of 1997, single-family permit activity in the Northwest's strong Portland-Salem, Oregon, and Puget Sound markets was close to the very strong 1996 levels for the same period.

Rental market conditions remain strong, and many major markets report balanced or tight conditions. New England rental markets continue to tighten, and multifamily permit activity was up in the first half of 1997. In the Mid-Atlantic region, the Philadelphia suburbs have seen a dramatic increase in apartment and condominium activity; in the Washington suburbs activity in the seniors' housing market is booming. Most California rental markets continued to tighten, and multifamily construction has increased in Southern California.

However, in some other areas market conditions are softening and local sources are reporting concerns about future market conditions. Multifamily activity in a number of previously hot markets has begun to decline. In Atlanta the large number of new apartments that entered the market in the second half of 1996 has weakened the rental market, and concessions have become deeper and more widespread. In Memphis, Knoxville, and Nashville, Tennessee, multifamily permit activity in the first half of 1997 dropped as builders have cut back in response to increasingly competitive conditions.

In the Southwest region, the Dallas-Fort Worth and Houston-Galveston areas are currently balanced and new units are being absorbed within the expected time. However, an increasing number of people in the housing industry have expressed concerns about overbuilding in both market areas, particularly in Class A and AA rentals, given the increases in construction over the past 2 1/2 years.

In the Rocky Mountain region, where the number of multifamily units permitted fell 30 percent, rental market conditions are balanced in the Denver, Colorado Springs, and Salt Lake City markets, but rent increases have slowed as concessions have become more widespread. The Phoenix and Las Vegas rental markets remain balanced, but vacancy rates have begun to drift upward. In both markets the continued expansion of the supply of new luxury apartments has increased competition and the use of concessions. In the Portland, Oregon, area the supply of new apartments in the past 2 years has effectively eased demand pressures, and multifamily permit activity in the first half of 1997 was down 25 percent from the same period last year.

JOB LOSS AND RELATED CHANGES IN CENTRAL
CITIES IN THE EARLY 1990s

The 1990-91 recession and the initially slow recovery weakened the economic base of cities, particularly large cities across the country and cities in the Northeast. Because job and related data are generally not available at the city level, these consequences have not been well documented.

Jobs vs. Employment

Employment refers to whether a person has a job, regardless of where the job is located. A resident of a central city will be counted as employed even if he or she works in the suburbs. Employment and unemployment data tell how well city residents are doing in the regional labor market.

Jobs are employment opportunities at a particular location. A job in a central city may be filled by either a central-city resident or a suburbanite. Job data (and the related establishment data) measure the economic base of a city.

The Bureau of Labor Statistics (BLS) publishes monthly and annual data on employment and unemployment for cities and metropolitan areas. Because these data are readily available, employment data are most often used to describe economic conditions in cities. But as noted these data tell about a city's residents, not about its businesses.

The Census Bureau collects data annually on jobs and establishments according to their locations.¹ This series, known as the Standard Statistical Establishment List (SSEL), contains data on all establishments that file Federal tax returns and have employees on their payroll.² The SSEL data do not include public-sector employment, many nonprofits, and sole proprietorships or partnerships with no employees.

Normally, the Census Bureau reports these data by county, but for the recent The State of the Cities report, HUD requested that the Census Bureau recode the SSEL data so that jobs and establishments could be distinguished in 77 central cities of 74 metropolitan areas, presented in table 1. The sample includes the Nation's 50 largest cities and 27 other cities chosen to ensure inclusion of all States. HUD obtained data for 3 years: 1991, 1993, and 1994 (the most recent years for which data are available).

This article uses the special Census Bureau tabulations to study how the recession affected cities in terms of job growth and economic function.

Table 1. 77 Central Cities by Population in 1994

Table 2. Changes in Central-City Jobs

Cities and Jobs

From 1991 through 1993, the number of jobs in the 77 central cities declined by 310,000. Between 1993 and 1994, central cities gained 276,000 jobs, resulting in a net loss of 34,000 jobs over both time periods. Large central cities lost jobs while middle-sized (250,000 to 900,000) and small central cities gained jobs.

Table 2 shows that the 10 largest cities (those with populations greater than 900,000) accounted for almost all of the job losses. Because of their size, the losses in the 10 largest central cities offset the gains in the 67 smaller central cities. The smallest cities among the 77 grew the fastest throughout the period. All classes of cities -- even the 10 largest -- had positive job growth between 1993 and 1994.

Although the 34,000 decline is trivial compared with the 23.3 million jobs in these central cities in 1991, the fortunes of central cities stand in stark contrast with that of their suburbs.

Suburbs recorded strong job gains during this period. Overall the suburbs of the 77 central cities created 3 million new jobs, an 11.2-percent increase, as shown in table 3.³ At every size level, jobs grew at a substantially high rate in the suburbs. But not all suburbs had an expanding number of jobs during this period. Among the 74 metropolitan areas studied, suburbs in Hartford, Los Angeles-Long Beach, Sacramento, and Washington, D.C., experienced job losses.

Table 3. Change in Suburban Jobs

Important regional differences also exist. Only central cities in the South experienced positive job growth from 1991 to 1994. Suburban job growth far outstripped central-city job growth in both the Northeast and the Midwest (see table 4). Suburban growth was modest in the West, reflecting defense industry cutbacks in California.

Table 4. Changes in Central-City and Suburban Jobs by Region

SSEL data are not yet available for the period of rapid job growth between 1994 and 1997. Information in The State of the Cities for 19 central cities that were tracked through 1995 indicates that the improvement in 1994 continued into 1995. But even with the improvement after 1993, 7 of the 19 central cities experienced job losses in the first 5 years of the decade. The losses were concentrated among the 10 largest cities and in the Northeast and Midwest.⁴

Cities and New Businesses

Table 5. Changes in Establishments in 77 Cities

In 1991 there were 1,256,000 establishments in the 77 central cities studied. By 1994 this number had grown by 14,000 to 1,270,000. The number of establishments increased by 7,000 between 1991 and 1993 and by another 7,000 between 1993 and 1994.

Once again the aggregate numbers hide important differences. The 10 largest central cities basically held their own with respect to growth in the number of establishments, while small and middle-sized central cities had modest gains. The number of establishments in the 10 largest cities increased by only 0.1 percent during the 1991-94 period.

Suburbs experienced strong growth in the number of establishments. The number of establishments in suburbs of all four size classes of central cities grew by more than 10 percent from 1991 to 1994.

Central cities in the Northeast lost establishments, while central cities in other regions gained establishments. Once again suburbs in the West failed to keep pace with suburbs in other regions because of the prolonged recession in California.

Types of Jobs

The Census Bureau data suggest that, in most industries, cities are increasingly the workplace of the most productive workers. At the same time, while jobs in cities are declining in many industries, they are increasing in low-wage industries that provide opportunities for entry-level workers.

Table 6. Changes in the Number of Jobs by Industry

In the 77 central cities studied, on average only service sector jobs were increasing (see table 6). Manufacturing jobs continued to decline, especially in the 10 largest central cities. The decrease observed in the 10 largest cities from 1991 to 1994 represented an annual average decline of 4 percent. The increase in service jobs was welcomed because, at 41 percent of all city jobs, service is by far the largest sector. Job growth generally was strongest in business services, health services, education services, social services, and eating and drinking places. These jobs pay below-average compensation and provide opportunities for low-skilled residents to enter the labor force.

Census Bureau data on average compensation show that the better paying jobs, and therefore the more productive jobs, were located in central cities (see table 7). Average compensation was higher in central cities and grew at a faster rate.

Table 7. Real Average Annual Compensation in 1991 and 1994.

Central-city jobs typically pay approximately 10 percent more than suburban jobs, and the difference widened between 1991 and 1994. The ratio was somewhat lower for the 10 largest cities -- about 7 percent.

These differences reveal the changing nature of jobs in central cities. The highest paying jobs were in sectors such as wholesale trade, depository institutions, and insurance carriers where the number of jobs in central cities was declining. Even though these high-paying sectors had a declining number of jobs, average compensation rose in the cities. This apparent anomaly is explained by the fact that within these declining sectors, average compensation increased faster than the number of jobs decreased. Firms in these sectors appeared to be keeping only their most productive jobs in central cities. Less productive jobs were either eliminated or moved to suburban locations. This trend worked to keep average annual real pay growing in central cities despite declining numbers of jobs in high-paying sectors and growing numbers of jobs in low-paying sectors.

Conclusion

Traditionally, cities have been viewed as the economic hubs of metropolitan areas, with suburbanites commuting to jobs in the central cities from bedroom communities. Over the past 40 years, this picture has been changing. Retail and service establishments have grown in the suburbs to accommodate rapid suburban population growth. Manufacturers have moved to the suburbs, and suburban business centers have attracted other firms.

The 1990-91 recession and its aftermath contributed to these long-run trends. Between 1990 and 1994:

• Cities experienced job losses while suburbs experienced job gains, although job loss appears to have turned around after 1993.

• New establishments were created much more rapidly in suburbs than in cities.

• Cities experienced job losses in all categories except services.

• Cities maintained their high-wage character mainly through the elimination or suburbanization of lower and middle-level jobs.

Still, cities typically have more jobs than residents in the labor force and, therefore, continue to attract commuters. In addition, the concentration of high-paying jobs in cities serves as evidence that firms continue to take advantage of economies of proximity and agglomeration.

Notes

4 The State of the Cities, page 31. The 19 cities, in size order, were New York, Philadelphia, San Antonio, Indianapolis, San Francisco, Baltimore, Jacksonville, Washington, Austin, Nashville, Denver, New Orleans, Norfolk-Virginia Beach, St. Louis, Wichita, Corpus Christi, Anchorage, Lexington (Kentucky), and Richmond-Petersburg.