Establishments employing two workers or more in an occupation oftenpay these workers at different rates. How frequent is such payvariation? How wide is the resulting spread in rates? Does the degreeof pay dispersion differ by occupation? This article explores theseissues using data collected in the Bureau of Labor Statistics’ 1983Area Wage Survey program.
Where an establishment had two workers ormore in a job, the percent by which the salary of the highest paidincumbent exceeded that of the lowest paid incumbent was calculated.Percentage differences for individual establishments were then averagedover all establishments providing such comparisons. Rate structures were clearly different for white- and blue-collarworkers. More than three-fourths of the workers employed in the 40white-collar occupations studied were in establishment paying more thanone rate for their job. Fewer than half of the workers in the 28blue-collar occupations studied were employed in multi-rate situations.The remainder were either the only incumbents in the job or were paid atthe same rate as the other incumbents of the job.
Among workers employed in establishments paying more than one ratefor a job, the pattern was again different for white- and blue-collaroccupations. Average wage spreads between highest and lowest paidworkers in the white-collar occupations studied ranged from 17 percentfor industrial nurses to 42 percent for intermediate electronicstechnicians. For the 12 skilled maintenance occupations, average wagespreads for all but two were betwee 7 and 14 percent. Among unskilledplant occupations, ranges were as small as 13 percent for power-truckoperators (other than forklift) and as large as 45 percent for lowerlevel guards. These differing structures reflect differences in pay systems inU.S. industry.
Employers commonly adopt formal pay systems,establishing either a single rate for a job classification or a range ofrates in which the minimum, maximum, or both of these rates arespecified. Pay of individual workers within a specified range dependson performance (merit), length of service, or both. In the absence of aformal pay systems, rates in a given job are determined largely by theemployer’s appraisal of individual workers. Data are not availablefrom the Area Wage Survey program to distinguish between the effects offormal and informal systems.
Pay spreads among workers in the same job and establishment cannotbe determined from the pay variations typically published inoccupational wage survey reports. Because of differences in pay levelsamong employers, industries, and localities, these reports showconsiderably wider ranges of pay rates in a job than would be expectedin a single establishment. It is not unusual for BLS area wage surveyscovering a variety of industries to find the highest paid worker in anoccupation earning twice as much as the lowest paid. In nationwidestudies, the highest paid worker may earn more than three times as muchas the lowest paid. In contrast, the average pay spreads found in thisstudy ranged from 7 to 45 percent. Information on pay spreads within establishment can be used for avariety of purposes.
For example, it is important to those establishingand administering rate-range pay plans. It is also useful in analyzingwage structures in that it helps to explain overall patterns of paydifferentials. In addition, it indicates the extent to which pay may beincreased without promotion to another job. Computing wage differences Information for this review of pay spreads within the same job andestablishment comes from data collected in more than 11,200establishments located in 70 metropolitan areas throughout the country.For Each of 68 BLS occupational classifications surveyed in 1983, thepercent by which the highest rate paid exceeded the lowest rate wascalculated where an establishment employed two workers or more atdifferent rates. These percentage differences were averaged, afterweighting the pay spread for each establishment by the number of workersit employed in the occupation. Establishments paying the same rate toall workers, as well as those with single incumbents in a job, wereexcluded from the calculations.
A standard set of occupational descriptions was used in allestablishments. In some cases, a single BLS occupation or level coveredmore than one company job. For example, the wide average pay spreadshown in table 1 for level I accounting clerks in unionizedestablishments is partly explained by the existence in sometransportation and utilities companies of two pay grades which fit theBLS description for this occupational classification. In other cases,the company job was barely broad enough to fit within the BLSdescription. This narrow span of duties could restrict any related payrange. This study of wage rate dispersion is limited to spreads betweenhighest and lowest rates actually paid to incumbents by individualemployers. It does not measure the full spread of formal rate rannges.This topic, however, was covered in a recent study by Martin Personick.
In a review of formal pay systems for white-collar workers in medium andlarge firms, Personick noted that “…
differences between thehighest and lowest rates actually paid are generally much smaller thandifferences between the maximum and minimum rates specified for arange.” Personick also found that workers tended to be clustered in the lower half of the rate range. Single and multiple pay rates In establishments with two workers or more in a job, the relativeimportance of single and multiple pay rates varied by occupationalgroup.
(See table 1.) The generally lower incidence of blue-collaremployment in multiple-rate establishments partly mirrors the greaterextent of collective bargaining among these workers than amongwhite-collar workers. Negotiated pay structures are more likely tocontain single rates than are non-negotiated structures. As explained byDavid Belcher. “Unions often favor the single-rate principlebecause it eliminates judgment-based differentials in individualpay.” Among plant jobs, guards and janitorial workers were most likely tobe in multiple-rate establishments, partly because many of thesurvey’s guards and janitors worked in protective or janitorialservice firms. In these firms, almost all workers are employed in thesame occupation.
While most may be paid at or near the minimum wage, atleast some receive higher pay in recognition of length of service orproficiency in the job. Often, higher rates are also required becauseof a customer’s special needs. About 10 percent of the workers whose wages were surveyed in the1983 Area Wage Survey program were the only incumbents in their job.This percentage, however, varied considerably by jobclassification–from 1 percent for millwrights to 83 percent forswitchboard operator-receptionists.
Differences in pay Among establishments paying multiple rates to workers in the samejob, average spreads between highest and lowest rates varied byoccupation, industry, and establishment size category. These factorsand their relationship to union status are considered in turn. Occupation. White-collar jobs, which commonly include a broadrange of duties, provide an opportunity to demonstrate superiorperformance. Where promotion to a higher grade is inappropriate, arange of pay rates can be used to reward superior performance within ajob or pay grade.
Conversely, it has been argued that the workingenvironment of employees in certain blue-collar jobs–for example, thosein assembly line operations–offers limited opportunity to deviate fromestablished performance standards; under such conditions, a single rateor narrow rate-range system may be more appropriate. Furthermore,single rates or narrow rate ranges are generally favored by laborunions, whose current strength is in the blue-collar area. These differences between white- and blue-collar jobs are reflectedin the results of this study. In establishments with more than one ratefor a job, the percent by which the highest rate exceeded the lowest wasgenerally larger in white-than blue-collar jobs. The roughly 30 percentaverage wage spread in white-collar jobs was more than twice as wide asthe average spread in skilled blue-collar jobs, but only moderatelywider than the average for material movement and custodial jobs. Amongthe blue-collar occupations studied, the potential for performancevariation is smallest in the skilled maintenance jobs, which arerestricted by definition to workers who have achieved journeyman status.Also, skilled maintenance workers, on average, are more concentrated inunionized establishments than are material movement and custodialworkers.
Among the individual white-collar jobs studied, mid-levelelectronics technicians had the widest average wage spread (42 percent),followed by entry-level electronics technicians (39 percent). Thesespreads are affected by the fairly broad range of duties andresponsibilities in the BLS job descriptions for the various levels ofelectronics technicians. Nurses and switchboardoperator-receptionists–jobs that often have few incumbents within anestablishment–had the narrowest average wage spreads in thewhite-collar group, 17 percent for the former and 19 percent for thelatter. Excluding these extremes, spreads ranged from 22 to 35 percentamong office clerical job classifications and from 24 to 33 percentamong the professional and technical jobs. Except for guards and janitors, average wage spreads forblue-collar jobs ranged from 7 percent for maintenance pipefitters andmillwrights to 30 percent for material handling laborers.
Maintenancetrades helper–a more broadly defined job–was the only maintenance,toolroom, or powerplant occupation studied with a spread of more than 20percent. Conversely, 12 of 16 material movement and custodial jobclassifications had spreads exceeding 20 percent. The 45 percent average spread for lower level guards was thebroadest among all of the jobs studied; janitors followed closely withan average wage spread of 42 percent. Many workers in these twoclassifications had earnings at or near the Federal minimum wage and, aswith other relatively low paid workers, a modest dollar spread in theirpay produced a relatively large percentage spread. An establishment, forexample, with one janitor at $3.35 an hour and another at $4.
35 recordsa 30-percent spread; the same dollar difference between two electricianswho earn $13 and $14 an hour produces an 8-percent spread. Moreimportantly, perhaps, was the employment of many of the survey’sguards and janitors in protective or janitorial service firms. Asnoted, these firms often pay workers different rates based on thespecific contract under which the service is performed. Eleven of the white-collar occupations in the study are dividedinto two work levels or more, based on duties and responsibilities. Ingeneral, the average percentage spreads of wages were similar for allwork levels of an occupation. Major exceptions were file clerks (a28-percent spread in level II compared with 22 percent in levels I andIII) and drafters (24 percent in level IV and 30 percent in level I).
In the blue-collar area, average spreads were similar for forkliftoperators and other power-truck operators. Tractor-trailer drivers,however, had considerably narrower spreads (14 percent) than the threeother truckdriver categories (25 to 26 percent). The two levels ofguards surveyed had a 7-percentage point spread. Industry.
Wage spreads between the highest and lowest paid workerin a job were generally narrower in manufacturing than innonmanufacturing industries. This pattern applied to both white- andbllue-collar occupations. Among the 59 occupations for whichcomparisons between the manufacturing and nonmanufacturing sectors couldbe made, the average spread in manufacturing was narrower in 52 jobs, ofequal size in three, and wider in four.
For three jobs–stationaryengineer, lower level guard, and janitor–the average spread was 23 to27 percentage points narrower in manufacturing. Among blue-collar jobs, the generally narrower average pay spreadsin manufacturing industries are partly traced to the greater degree ofunionization in this industrial sector. For 1982–the most recent yearfor which unionization data are available from the Area Wage Surveyprogram–63 percent of hte manufacturing production workers were inunionized establishments, compared with 43 percent of the blue-collarworkers in nonmanufacturing industries. Table 1 shows, for allinudustries combined, that blue-collar jobs generally had narroweraverage spreads in unionized establishments.
Maintenance trades helpersand janitors–jobs with relatively wide average wage spreads for themaintenance and custodial occupational categories–were the onlyexceptions to this pattern. In 33 of the 37 white-collar comparisons that could be made betweenindustry sectors, average pay spreads were narrower in manufacturing,but unionization is not a major explanation. Unionized establishments inthe 1982 Area Wage Survey program employed 9 percent of thenonsupervisory office clerical workers in manufacturing and 15 percentin nonmanufacturing. Considering this limited degree of unionization,collective bargaining could not produce significant white-collar paystructure differences between these two industry sectors. Moreover,considering all industries combined, average pay spreads were wider inunion than in nonunion establishments in 15 of 23 white-collaroccupational classifications that were compared. Differences in thenature of the job and prevailing pay systems are reflected in the widerrate ranges that unions have negotiated for white-collar than forblue-collar workers.
Establishment size. The average wage spread was wider inestablishments employing 500 workers or more than in smaller units inall but three of the white-collar classifications compared–order clerk II, accounting clerk IV, and electronics technician III. This mayresult from the relatively greater use of formal rate-range pay systemsin large establishments. It may, however, also reflect increaseddiversity in pay because of greater numbers of job incumbents in thelarger establishments; that is, the more workers an employer has in ajob, the greater the likelihood of having incumbents at or near thebottom and top of the rate range for the job. Among blue-collar occupations, the pattern was mixed.
Fewestablishments with fewer than 500 workers paid more than one rate, orhad more than one employee, in maintenance, toolroom, and powerplantoccupations. Consequently, establishment-size comparisons were possiblein only 5 of 12 of these skilled worker jobs. In 4 of the 5 jobs,average pay spreads were wider in the smaller establishments. Thisresult–which is contrary to the general findings for white-collaroccupations–may reflect a greater incidence of skilled maintenanceworker unionization in the larger establishments.
Among the lessskilled material movement and custodial jobs, however, average wagespreads generally were wider in the larger establishments, although thereverse occurred in 3 of the 14 jobs studied (tractor-trailertruckdriver, shipping packer, and forklift operator). Other factors. Data collected in the Area Wage Survey programpermit analysis of variations in pay spreads by type of occupation,industry, union status, and size of employer. Several other factors,however, may influence the spread of pay rates within individualoccupations in an establishment. Although this study cannot measure theextent of their influence, some of these factors can be noted. Forexample, a company in a low wage industry, and with a formal rate-rangepay system, may be located in a high wage area.
As a result, hiringrates may be near the top of the range to attract employees, forcing anarrow spread in rates paid. The rate of worker turnover and the degreeof difficulty in recruiting new employees also affect the location ofhiring rates within established rate-range pay systems. Variations among establishments Averages of establishment pay spreads for individual jobclassifications conceal significant variations among the establishments.Table 1 sheds some light on establishment variations by presenting therange of pay spreads for the middle half of the employees in multiplerate situations. (The boundaries of this range are defined by the firstand third quartiles.
) For example, the middle half of the stenographersI were employed in establishments with pay spreads between 12 and 51percent. For white-collar occupational classifications, considerablevariation is evident among the pay spreads within individualestablishments. In all but four classification, the difference betweenthe first and third quartiles–the interquartile range–was 20percentage points or more. The narrowest interquartile range applied toregistered industrial nurses (16 percentage points) and the widest todrafters I (48 percentage points). Among material movement andcustodial jobs, establishment variations in wage spreads were similar tothose in white-collar classifications, but the variations were not aspronounced among skilled maintenance, toolroom, and powerplant jobs. For white- and blue-collar jobs combined, variations amongestablishments tended to be greater in occupational classifications withrelatively wide average wage spreads.
To account for this relationship,the interquartile range for each classification was standardized bydividing it by the median pay spread for that job, producing an index ofrelative dispersion. The indexes–which are not shown in table 1 butcan be calculated from quartile data presented–were generally higherfor blue- than for white-collar jobs.