Rising Tide of Temps

The number of workers in Washington's temporary help supply industry more than quintupled from 5,496 in 1982 to 28,222 by 1995, translating into annual growth of 13.4 percent. State employment, by comparison, grew at a healthy but clearly lesser annual rate of 3.2 percent over the same period.

At $16,658 in 1995, average covered wages in Washington's temporary help supply industry were well below the overall state average $27,446. This wage is consistent with what is generally understood about the industry: that it's traditionally a broker for lower-skilled, lower-paying assignments which provide fewer hours per worker.

The rise of the temporary help supply industry has been a peculiar feature of the U.S. economy during the 1980s and 1990s. This is not to say that the industry did not exist prior to the 1980s. Indeed, the National Association of Temporary Staffing Services (NATSS) notes that the modern-day industry traces its origins to Chicago in the 1920s when clerical workers with an expensive capital good—an electric adding machine—moved from assignment to assignment fulfilling the needs of the service-based economy. Despite its history, it has been only of late that the industry has reached a critical mass in employment large enough to warrant investigation, particularly here in Washington.

Over the past three years, the Bureau of Labor Statistics, the Federal Reserve, and other agencies have stepped up their analysis of the temporary help industry. The Journal of Labor Research published an entire edition (Fall 1996) edicated to flexible employment with particular emphasis on temporary help workers. Similarly, the U.S. Department of Labor's Monthly Labor Review dedicated an entire issue (October 1996) to contingent workers and alternative workers which highlighted the temporary help work force and its characteristics.

Why the Interest?

Businesses, unions, and labor force participants all have an interest in the rise of the temporary help industry because it serves as a symbol of both the new business climate and the new rules of the labor market. Two perspectives, driven by the same phenomena, have given symbolic meaning to the temporary help industry.

From a business perspective, the use of temporary workers is one response to the push for increased productivity and increased flexibility that grew out of the 1980s and continues today. At that time, manufacturing in the U.S. was considered bloated and sluggish. But the drive to correct the U.S. economy went beyond manufacturing. Business is now bound to a code of lean competitiveness crucial to competing with emerging and industrial economies. Greater productivity has been the mantra guiding many decisions, including the use of a more flexible temporary help work force.

From a worker perspective, the temporary help supply industry symbolizes the new employer-employee relationship. Gone are the days when securing a job meant remaining with an employer through one's working life. Recent writings tell of employees attaching loyalty to their occupation rather than to their employer. While the magnitude of change in the data has not paralleled the mountain of anecdotal evidence nor general perceptions, the movement toward a new labor market is evident (see Employee Tenure in the Mid-1990s, page 17).

Temporary Help, Permanent Presence

The number of workers in the state's temporary help supply sector more than quintupled over the past 14 years from 6,471 in 1981 to 28,222 in 1995 (see Figure 8). This is a net gain of 336 percent, or a compounded annual gain of 11.1. This is, to say the least, phenomenal growth; Washington's overall covered employment base grew 44 percent from 1981-95 or at an equivalent compounded rate of 2.6 percent. As such, temporary help saw its share of Washington State's total covered employment climb steadily from 0.4 percent in 1981 to 1.2 percent in 1995.

The U.S. temporary help supply sector saw a gain of nearly 400 percent over the 1981-95 period or 11.7 percent in annual terms. That was slightly above the rates posted in Washington (see Figure 9) .

It is clear from the annual rates that growth in the Washington and U.S. temporary help supply sectors was extraordinary. However dynamic it may have been, it was not a straight-line trend. As a matter of fact it was quite volatile, with some years reporting employment declines (see Figure 10).

Evident in both the Washington and U.S. trendlines was the cyclical nature of temporary help supply employment. This is not a new revelation. Much has been written about temporary help supply as a leading indicator of the economy since firms tend to let go of temporary workers first as the economy weakens and hire temporary workers first as they cautiously test a strengthening economy. Presumably, the more cyclical the economy, the greater the impact on the temporary help supply industry. There is some speculation, though, that the industry's leading indicator status may not hold up over time as the nature of temporary help supply arrangements increasingly change. Temporary workers are not as temporary as they used to be, and may in the long run not be distinguishable to any great extent from permanent workers.

To put into perspective the national scope of this phenomenon, the National Association of Temporary and Staffing Services (NATSS) reported that from 1994-95, temporary help receipts grew 12.9 percent to $39.2 billion. Temporary help payrolls also rose 12.9 percent to $27.9 billion, and the number of U.S. temporary workers climbed 9.6 percent to more than 2 million or 1.8 percent of total U.S. employment. NATSS also noted that there were currently from 5,000 to 6,000 temporary help companies in the U.S.—twice the number ten years ago. Viewed this way, the temporary help supply sector in Washington is but one component of what is clearly a nationwide trend.

What are some of the reasons behind this phenomenal growth? According to NATSS, the growth in the temporary help supply sector can be attributed to several key factors:

• Temporary employment is viewed by workers as a source of job security and job opportunity in an evolving and changing workplace.

• Temporary employment is viewed by business as a way to increase flexibility, responsiveness, and adaptability to changes in the global economy.

• Temporary work provides a reliable pool of labor during a period where businesses find it difficult to attract, evaluate, and recruit qualified employees.

• Temporary work is more inclusive of more types of workers, especially professional and technical, in a period of more diverse workflow and staffing arrangements.

• The U.S. economy is in its fifth consecutive year of growth.

Another View of Growth

Another avenue for detailing the rise in use of temporary help employees in Washington is provided by the Department of Labor and Industries (L&I). L&I requires employers to pay a fee based on the hours worked by each employee. The fee is placed in the state fund to pay for employee disability claims. Some large temporary help agencies provide their own disability insurance and are exempt from these hourly fees. Employers, however, must still report the hours worked for each staff and temp.

Translating the hours data reported to L&I into full-time equivalents (FTEs) shows an enormous gain in the number of workers in temporary help. In the first quarter of 1988 there were an estimated 9,576 temporary help workers; by the third quarter of 1996 that had grown to 29,234 (see Figure 11).

The data also portray a strong seasonal trend. For the period examined, average hours fell 9.0 percent in the first quarter followed by an 8.9 percent increase in the second quarter, bringing the total hours worked in the quarter nearly back to the level of the previous year's end. Hours rebounded 11.4 percent in the third quarter followed by minimal average growth of 1.6 percent in the fourth quarter. The cycle repeats itself each year starting from a higher base.

Occupational Shifts

Of particular interest to most of those who study the temporary help supply sector is the occupational composition of the employment. Some of this information can be obtained from the Bureau of Labor Statistics (BLS) publication entitled Occupational Compensation Survey: Temporary Help Supply Services that was conducted in both 1989 and 1994. The Seattle PMSA (King and Snohomish counties) was among the major labor market areas surveyed by BLS. Data from the survey were analyzed in terms of annual percent change and in terms of shifting share.

Annual Percent Change

Between 1989 and 1994, temporary employment in the Seattle PMSA rose at 7.6 percent compounded annual rate, twice as fast as total employment in the Seattle PMSA (see Figure 12) . The same relationship held for each major occupational category surveyed in the Seattle PMSA—with temporary employment rates outpacing total employment rates by several-fold. For example, the number of temporary workers in executive, managerial, and administrative occupations soared nearly 50 percent compared to 8.3 percent for the category generally. With the exception of clerical and administrative support occupations, the balance of categories all had double-digit growth in temporary employment while total employment remained in the single digits.

Nationally, the growth rates for temporary employment and total employment displayed relationships similar to those in the Seattle PMSA. However, the actual rate of growth in total employment was less vigorous nationally than in the Seattle PMSA. This should not come as a surprise given the stronger overall performance of the Seattle PMSA in the 1990s (for example, the national economic recession impacted the U.S. economy while barely registering in the Seattle PMSA and Washington).

Shifting Occupational Share

With respect to shares by major occupational groups, temporary workers make up a slightly higher share of total Seattle PMSA employment than they do total U.S. employment in both 1989 and 1994 (see Figure 13 ). What distinguished Seattle from the U.S. were professional specialty occupations and technical and related support occupations; Seattle posted higher shares of both groups in 1989 and then proceeded to boost those shares by 1994—all while the comparable U.S. shares remained fixed. This is significant because it adds a quantitative grounding to the anecdotal evidence that temporary help supply services were used to a greater extent in the Seattle PMSA than was the case nationally and that one of the drivers of that higher than average use was high tech assignments.

Geographic Comparisons

The temporary help supply industry in Seattle was compared against other presumably technology-oriented metropolitan areas. Though all saw gains in temporary help supply workers, a clear distinction emerged between what might be called an emergent economy versus a <M>mature economy. As the annual percent changes show, Seattle and Denver fall into the former, Los Angeles-Long Beach and Boston into the latter (see Figure 14 ). This is in large measure related to the bases from which each area had to start. Seattle and Denver had relatively small temporary worker bases in 1989 which, given gains by 1994, translated into large percentage increases. Los Angeles-Long Beach and Boston, meanwhile, had a relatively large base of temporary workers in 1989, so a healthy increase in number reaped but a modest rate increase. Keying specifically on temporary professional specialty occupations and technical and related support occupations, Seattle was the only area to post positive percent change in both.

Another View of Occupations

Labor and Industries (L&I) data provide another view of the changes over time in the use of temporary help workers in Washington. Hours worked by temporary help employees are separated into 18 classifications. Each staff and temp is assigned a risk class based on the type of work performed. The data are reported such that workers can be placed in varying risk classes if they perform different types of work.

Among other things, the L&I data show that internal temporary help agency staff, which are classified under administrative offices , has not grown in recent years despite the increased number of employees placed with outside firms (see Figure 15 ). Also, since 1994, the driving force behind growth has been on the clerical and white-collar end of the occupational spectrum.

Unfortunately, the data for office support services are not broken out into specific categories. This classification includes accountants, attorneys, clerks, computer programmers, data entry, dispatchers, receptionists, secretaries, typists, telemarketers, and any other persons engaged wholly in office work. While the educational and experience backgrounds vary markedly, these workers can all be grouped in the white- or pink-collar classifications. From first quarter 1994 to second quarter 1996, the FTE equivalent of hours worked by office support service workers has risen 66 percent from 7,945 to 13,149.

While the office support sector exhibited little seasonality, many of the 16 other L&I classifications had highly seasonal components. An explosion in use of "other temps" occurred in 1993 as a result of increases in packaging of dry goods and electronic and scientific equipment assembly. In 1994-95, it was the office support sector that expanded rapidly.

SURVEY OF WASHINGTON'S
TEMP HELP FIRMS

Who uses temps?

In assessing employment growth in the temporary help industry, a central question is whether it reflects growth in services or a transfer from other sectors of the economy. In other words, are the temporary workers actually working on assembly lines and in retail stores while being counted as service sector employees? Quantifying which industries use temporary workers difficult. A recent analysis conducted by the Employment Security Department took a first step toward quantifying the use of temps by major industry division in Washington State.

How to measure?

To measure the use of temporary help employees in Washington, temporary help supply agencies were selected and surveyed by phone and fax, and their respective employment figures from June 1995 and June 1996 were extracted from the Employment Security Department's covered employment and wage (ES-202) database. The data (SIC 7363) covered temporary help agencies and employee leasing firms. Looking solely at the firms that responded to the survey:

• Employment for June 1996 was 20,072 or 61 percent of the state total for SIC 7363.

• The year-over-year employment gain was 4,061, which accounted for 76 percent of the over-the-year growth in SIC 7363.

How many staff are there?

The initial survey question asked firms for the percentage of internal staff workers. The survey shows that approximately 96 percent of the payroll were temps with the remaining 4 percent working as staff. This is in line with two other sources of data. First, the 1994 Occupational Compensation Survey from the U.S. Bureau of Labor Statistics estimated staff in Seattle area establishments at approximately 4.9 percent. Secondly, L&I data for 2nd Quarter 1996 showed that staff and temps working in temp help agency offices accounted for 5.5 percent of the total hours worked. Assuming that some temps hours were included and that staff likely worked more hours than the average temp, the L&I data provide an upper bound on the number of staff.

How long are assignments?

The second question asked for the share of employees in assignments (leased or otherwise) lasting two months or more. The responses varied dramatically. A large portion of respondents (33 percent) had at most 10 percent of their placements in long-term assignments. Another large block of respondents (40 percent) had at least 40 percent of their temporaries in long-term assignments. The largest response was a firm with 75 percent of its temps in long-term assignments.

Many of the firms surveyed could not document their responses. Others said they had no way to track this information or had only the average length of assignments. Regardless, the consensus was that the average length of assignments was growing. Moreover, one firm volunteered that it used temporary-to-permanent assignments. That firm regularly sends temps out on 90-day trial assignments after which the contracting firm has the option of hiring the temps for full-time positions. Another firm stated that 30 percent of its temporaries eventually found full-time work through a temporary assignment.

Which industries are served?

The survey asked for the percentage of temporary workers assigned to various major industry divisions. These industry breakouts were then multiplied by the number of employees (less a firm's internal staff) reported on the firm's June 1996 unemployment insurance report. This provided an approximate share of temporary help payroll workers engaged in each major industry division in June 1996. While 41 percent of the workers were performing work for firms within the services industry, the other 59 percent were serving other industry divisions (see Figure 16 ).

The next step was to assess which areas saw the most growth over the June 1995 to June 1996 period. By assuming the temporary help agencies did not change the type of business clients they served, each firm's over-the-year growth was divided proportionally among the industry divisions it served. For example, if a firm had 70 percent of its employment in finance and increased by 100 employees over the year, then a gain of 70 employees was noted in the finance division.

The use of temporary help workers has increased in all industries. On the whole, the use of temps increased by 19.6 percent over the year (see Figure 17 ). Any decline in the composition of temps has been due to smaller relative gains rather than a decrease in use. Wholesale trade, for example, experienced a declining relative share of temps despite increasing its use of temps by 9.0 percent.

National data concerning the breakout of industries serving temporary help workers were derived from a special supplement of the February 1995 Current Population Survey (CPS). The major difference between Washington and the U.S. is the decreased use of temps in Washington's manufacturing sector (see Figure 18).

Temporary Wages and Benefits

This analysis of wages focuses on two data sources: (1) the Occupational Compensation Survey:Temporary Help Supply Services published by the U.S. Bureau of Labor Statistics and (2) average covered wages from the Washington State Employment Security Department.

Average Hourly Earnings

Nationally, the Bureau of Labor Statistics occupational compensation survey revealed that the average U.S. temporary worker earned $7.74 per hour in 1994, a mere 15 cents or 2 percent per year more than in 1989. BLS cites as one likely cause a shift in temporary workers into blue-collar occupations that tend to be concentrated at the lower end of the pay spectrum between $5 to $7 per hour. The data do reveal that temporary workers went from 30 percent of blue-collar workers in 1989 to 40 percent in 1994. Despite the overall average, hourly rates for temporary workers varied widely by occupation in 1994 (see Figure 19 ).

In 1994, the average hourly wage for all temporary help supply occupations in the Seattle PMSA was $9.76 compared to $7.74 for the U.S. In fact, looking at the major occupational categories of temporary help supply workers in 1994, one can see that the Seattle PMSA posted higher average hourly wages than did the U.S. in each major category. This was not the case in 1989; at that time, average hourly wages for sales and marketing and clerical and administrative support workers were lower in the Seattle PMSA than in the U.S.

The real change in average hourly wages for temporary help supply workers in the Seattle PMSA from 1989-94 was hardly glowing (see Figure 20 ). Clerical and administrative support, sales and marketing, and services essentially managed to hold constant over the period. However, the higher-end occupations—executive and managerial, professional specialty, and technical and related support—saw real wages plunge from 5 percent to 9 percent on an annualized basis. It was this trend that pulled the Seattle PMSA down 1.7 percent per annum from 1989-94. At the national level, most of the losses were from 1 percent to 3 percent per year, with technical and related support occupations pushing the range at -6.2 percent.

Average Covered Wages

It is clear from the data that average covered wages in Washington's temporary help supply sector were considerably below the average for all industries over the 1981-1995 period (see Figure 21). In 1995 the state's temporary help supply sector had an average covered wage of $16,658 compared to $27,446 for all industries. That was only 60 percent of the state average. The wage disparity is consistent with what is generally known about the temporary help supply sector—that it has often been a broker for lower-skilled, lower-paying assignments that provide fewer hours per worker, i.e. part-time jobs.

Employee Benefits

According to the BLS temporary help supply worker compensation survey the share of temporary workers in the Seattle PMSA receiving paid holidays, paid vacations, and health care benefits was markedly lower than that for the U.S. as a whole in 1994 (see Figure 22 Real Annual Percent Change in Hourly Earnings). This may be indicative of a higher degree of industry maturity at the national level. The same survey did show, though, that a greater share of temporary workers in the Seattle PMSA received training than did temporary workers nationally. Another positive note with respect to the Seattle PMSA is that the share of temporary workers getting paid holidays and vacations rose over the 1989-94 period (much more noticeably in the former category than in the latter category).

Other Key Findings

• Department of Labor & Industries data show an enormous rise in temporary help FTEs from 9,576 in 1st Quarter 1988 to 29,234 in 3rd Quarter 1996. The data also revealed a strong seasonal trend over the four quarters as the seasonal swings repeated each year starting from a higher base.

• What distinguished the Seattle PMSA from the U.S. was its significantly greater shares of both professional specialty occupations and technical and related support occupations . The Seattle PMSA posted higher than average shares in the 1989 BLS survey and then boosted those shares even higher by 1994—all while the comparable U.S. shares remained fixed.

• Temporary worker usage in emergent economies (Seattle and Denver) grew at significantly higher rates than in mature economies (Boston and Los Angeles-Long Beach).

• Hours data from the Department of Labor & Industries show that internal temporary help agency staff have not increased as a share of total temporary help employment in recent years despite the surging number of workers placed with client firms. The data also show that while growth in the temporary help sector has taken place in virtually every risk class (occupation), the major driver of that growth since 1994 has been office support and white-collar occupations.

• An Employment Security Department survey of temporary help firms in Washington revealed that a third of the respondents had no more than 10 percent of their placements in long-term assignments (two months or longer). Two out of five respondents had at least 40 percent of their temps in long-term assignments; one respondent had 75 percent of its temps in long-term assignments.

• The survey of temporary help firms also found that 41 percent of the assignments were in services, while the remaining 59 percent were in the balance of industry divisions. Although the temporary assignments were dominated by the service industry, temporary help employment had increased in all industries.

• In 1994, the average hourly wage for all temporary help supply workers in the Seattle PMSA was $9.76 per hour compared to $7.74 per hour for the U.S.

• The share of temporary workers receiving paid holidays and vacations rose between 1989 and 1994 for the Seattle PMSA, though that ratio was lower than that for the U.S. The survey did show, however, that a greater share of temps in the Seattle PMSA received training than was the case nationally.

Gary Kamimura,
Economic Analyst