EconSouth (First Quarter 2003)
Conventional wisdom holds that service-producing industries are better able to weather economic slumps than are goods-producing industries and can be expected to provide stability during downturns. But the recession that began in March 2001 may show that this assumption requires a second look.
hat a difference a decade has made in the composition of the Southeast’s economy. While traditional manufacturing activities such as apparel, paper and textile production were shutting down and moving operations offshore over the past 10 years, the service sector, especially activities related to communication services, transportation and general business services, was rapidly expanding in the region. In 1992 the Southeast’s service industries accounted for some 79 percent of the region’s total nonfarm payroll employment. Today that share is closer to 83 percent. In Florida, which exceeds the national average because of its heavy focus on tourism and health care, the service sector’s share of employment increased from 86 percent of employment in 1992 to 88 percent in 2002.
Conventional wisdom holds that service industries, especially those outside of retail and wholesale trade, are less affected by changes in aggregate supply-and-demand conditions than are goods-producing industries, such as manufacturing and construction. Performance during past recessions upheld this belief, and analysts have often argued that growth in these types of service industries adds stability to an economy. But the most recent recession has put this assumption to the test.
The national service sector: Similarities and differences over time
In some ways, the national recession that began in March 2001 was similar to the recession that began in July 1990 and ended in March 1991. For instance, 2001 saw a net decline of a little less than 1 percent in payroll employment in the United States; by comparison, during the 1991 recession, employment declined just over 1 percent.
As in the earlier recession, manufacturing accounts for the bulk of the decline in national employment this time around too, subtracting about 1 percent from total U.S. employment rolls since March 2001. As is typical during a downturn, the government sector — comprising federal, state and local governments — has added modestly to payrolls. The drag caused by the loss of retail and wholesale trade jobs has also been similar between the two recessions. But that’s where the similarity ends.
The construction industry provides an example of the differences between the two recessions. The cut in payrolls in this industry since March 2001 is not anywhere close to the cuts during the 1990–91 recession. This difference reflects the relative strength of the demand for new housing during the current period, tempered only by weakness in commercial construction.
Another important difference between conditions today and during the recession of a decade ago is the relatively poor performance of transportation and communication service industries (see the table). The overcapacity problems plaguing the telecommunications industry and Sept. 11’s adverse consequences on the transportation industry explain why these industries have not fared as well as they did in the 1990–91 recession.
Other service industries play a role
The relative performance of a category referred to as “other” services over the past year and a half also differed markedly from the early 1990s. This broad service industry category — which includes hotels, temporary staffing firms, health care providers, computer support services and data processors, auto repair businesses, and engineering and management service companies — has been growing in size and relative importance over time.
For the nation as a whole, employment in these industries as a proportion of total employment increased from less than 27 percent in 1992 to more than 31 percent in 2002. This increased employment share reflects the rapid growth in these service industries relative to manufacturing. In 1992 manufacturing accounted for 17 percent of total employment in the nonfarm sector; this share declined to slightly over 13 percent in 2002.
The overall contribution of other services to employment during the most recent downturn has been considerably less than during the prior downturn, and this difference can be traced to a number of factors. For one thing, temporary staffing firms saw a sharp decline in demand for their services starting in late 2000 and persisting throughout 2001. Temporary workers or those on the temporary staffing firms’ payrolls are counted as part of the service sector regardless of what industry they may actually be working in, so part of this decline reflects the coincident weakness in manufacturing. But office clerical staff reductions also accounted for a significant part of the most recent decline. During the 1990–91 recession, in contrast, temporary staffing employment declined for only six months before rising again.
Demand for additional types of business services, such as computer and data processing, has also taken a more significant hit during this downturn. This development is partly a result of the dot-com implosion in 2000 and partly a consequence of generally lower business spending on information technology. The hotel industry and amusement and recreational services make up another group of service providers that have experienced a distinct slump in employment during the most recent downturn, especially following Sept. 11. Employment at hotels and amusement and recreational facilities declined by some 4 percent in 2002, more than twice the next-largest percentage decline, which was experienced in 1991.
Of course, not all of the other service industries have performed poorly over the past couple of years. The health care industry has been rapidly expanding its employment base to keep pace with U.S. demographic changes that are increasing the demand for health-related services. But even in this industry, the rate of employment increase is only half that of the early 1990s boom.
The Southeast service sector
Like the nation, the Southeast experienced a different kind of recession in 2001 than in 1992. Indeed, some of the differences are even more pronounced in the Southeast than at the national level.
In Alabama, overall employment was essentially flat during the 1990–91 recession. In contrast, total state employment is estimated to have declined by close to 1 percent in 2001 and 2002. The biggest contributor to employment growth in Alabama has been the health care industry, but this positive effect is being swamped by weakness in some service industries, such as retail trade and communication services, as well as the ongoing weakness in manufacturing.
In its manufacturing sector, however, Alabama has developed a shield over the past decade that protected the state from an even worse employment outcome during 2001–02: its auto industry. Mercedes and Honda have expanded their plants in the state, and other assemblers and parts suppliers are establishing facilities there.
Unlike Alabama, which has a relatively small service sector compared to most other states, Florida has a service sector that dominates the state’s economy, with tourism and health care as its mainstays. In 1991 total employment in Florida declined by a bit less than 2 percent, in large part because of a 14 percent decline in construction employment following the speculative construction boom of the late 1980s and a 3 percent decline in retail trade employment. The service sector, supported by 8 percent growth in health care employment, had a positive overall impact on employment in 1991.
During 2001 and 2002 the service sector continued to contribute to employment stability in Florida. Modest gains in business services, especially back-office operations and call centers, as well as an increase of approximately 3 percent in employment at health care service providers more than offset the weakness in hotel and amusement services. Outside of services, construction employment did not decline in Florida overall, a fact that helped ensure that total employment remained essentially unchanged in the state during 2002.
Although Florida has the largest service sector in the Southeast, the impact of the service sector contraction was felt most significantly in Georgia. Employment in the state declined by around 2 percent in 2002 — about the same percentage decline as was experienced in 1991. But the composition of the 2002 decline was quite different.
In 1991 employment in the health care industry grew by some 6 percent, and in 2002 health care employment increased by 4 percent. Essentially all other sectors of the state’s economy were drags on employment in 2002, however. One example, the construction industry, was neutral for the nation as a whole in 2002 but was down in Georgia as the pullback in commercial construction caused by sharply rising vacancy rates in Atlanta more than offset the strength of the residential housing market. The state has also witnessed considerable downsizing in Atlanta’s telecommunications industry, a business that created a significant number of jobs during the 1990s.
In addition, Georgia’s air transportation segment was hit hard by the Sept. 11 attacks, and Atlanta-based Delta Airlines subsequently laid off thousands of workers. On net, the transportation and communications industries reduced employment by almost 6 percent in the state in 2002, compared to the 2.5 percent decline in 1991. Moreover, Georgia is especially reliant on business travel. As a consequence, hotel employment in the state experienced the largest percentage decline in the Southeast during 2002.
In Louisiana, Mississippi and Tennessee, as in the nation, health care has again been a much less significant employment cushion during 2001–02 than in the early 1990s. For example, while Louisiana’s health care industry is still expanding moderately, it is growing at less than a quarter of the pace seen in the early 1990s.
The overall employment situation in Louisiana was flat in 2001 and 2002, but the state enjoyed significant contributions from the oil and gas industry, which provided stability to employment throughout the year. Also, Louisiana is one region where the hotel industry actually performed reasonably well during the most recent recession, reflecting the state’s lower reliance on visitors traveling by air. Some reports suggest, though, that the hotel industry in Louisiana may be experiencing slowing in the first quarter of 2003.
For Tennessee, several service industries have remained solid contributors to the employment picture during 2001 and 2002. The state’s hotel industry performed especially well compared to the early 1990s, but the overall contribution from services is much less than in the early 1990s. Again, the relative weakness here is related to the poor demand for business services combined with a reduced contribution to growth from the health care industry as compared to during the 1990–91 recession.
Overall employment levels in Mississippi were generally stable during 2002, but this stability followed a sharp decline in payrolls in 2001. Unlike the situation in the early 1990s, employment growth in Mississippi over the past couple of years has mainly come from state and local government employment rather than from services. Also, unlike Louisiana and Tennessee, Mississippi reduced staffing levels in the hotel industry in 2001 and 2002, perhaps reflecting a leveling off of growth in the state’s gaming industry.
Not immune to a downturn
The latest recession has several lessons to teach about the service sector of both the U.S. and regional economies.
One lesson from the 2001–02 experience is that many service industries are intimately linked in a way that means a shock to any component has repercussions for all other components. The dot-com bust of 2000 and its subsequent fallout is one example; reduced spending on technology has direct consequences for other service providers, including those providing information technology services and telecommunications services.
The latest recession also demonstrates that considerable regional concentration or specialization in the provision of different types of services exists — so a specific shock can affect some areas of the region more significantly than others. Comparing two Georgia cities illustrates this point. Because of its reliance on business and air travel, Atlanta was much more severely affected by the pullback in business-travel spending in the aftermath of Sept. 11 than was Savannah — a more traditional drive-to tourist destination. The fact that Atlanta’s employment declined by over 2 percent in 2002 while Savannah eked out employment growth of over 1 percent is directly related to this difference in specialization.
Opinions on how the service industry will fare as the economy continues to recover are varied. Undoubtedly, the various service industries will pick up momentum as the economic recovery unfolds, though the road to recovery in some segments, such as telecommunications and transportation, will likely be rocky. But one thing is certain — any claim that the service sector is somehow immune from negative economic shocks has been dispelled by the events of the last two years.
This article was written by Navnita Sarma and John Robertson of the regional research group in the Atlanta Fed’s research department.