EXISTING RESEARCH ON CONTINGENT WORK
Contingent work research at the organizational level has focused on what environmental conditions predicts the level of contingent use. These studies primarily take an Internal Labor Market or Transaction Cost Economics approach in explaining why firms may or may not use this form of employment. The findings generally support hypotheses that variability in the need for employees and cost containment efforts are correlated with more external workers.
In a study drawing on the Department of Labor's 1982 Employment Opportunity Pilot Project, the effect of employment costs, external environment, organizational size and bureaucratization, and skill requirements on the use of temporary workers and contract employees was examined (DavisBlake & Uzzi, 1993). Variation in employment needs was correlated with higher levels of both contractor and temporary work use. Level of bureaucratization and establishment size was correlated with high levels of contractor use, but low levels of temporary worker use. Firm specific training necessary to perform a job, level of government oversight and level of information or technical skills all contributed to lower use of temporary workers, but had no significant effect on the use of contractors. This study measured use of contingent workers as a binary variable; the level of use and the capacity in which these workers are used was not examined. This study is interesting, though, in that it separates temporary workers from independent contractors. While not universal, temporary work generally consists of categories of labor with less specialized skills that fit the description of an open market better than into the description of a craft / professional labor market, compared to the skills found amongst independent contractors.
In a follow up study, Uzzi and Barsness (1995) used an internal labor market framework to examine cost, power and bureaucratic control perspectives across numerous industries in Great Britain in 1984 and again in 1990. They found, interestingly, that wage costs as a percentage of revenue did not lead to increased use of external workers. The use of external workers was positively correlated with the percentage of managers in a firm, the presence of a formal evaluation system and high levels of worker conflict. The percentage of union membership was negatively correlated with the use of external workers. This study distinguished between two types of contingent workers: freelance workers and fixedterm contract workers. While the exact definition of these two forms of contingent workers is not clear, it is interesting to note that at the second time point (1990), the two forms of contingent work do not have the same relationship (found at time one) with any of the independent variables tested. Further, the only control variable on which both forms of contingent work share a statistically significant relationship was that of the percentage of technical/professional jobs in an organization. The results for time 1 (1984) were not as disparate; both forms of contingent work share statistically significant relationships with 4 of the 9 independent variables tested (forprofit enterprise, formal performance evaluation system, percentage of managers, job reorganizations in the last 3 years), and on the control variable for single site employer. This suggests that researchers should be cautious about aggregating different forms of contingent work into one category.
Another study focused on the food service industry in Japan, looking at organizational factors that influenced whether traditional or nontraditional employment arrangements were used. This study found firms that reward long service, require internal training, and pay above market wages to permanent employees employ more parttime and temporary workers (Morishima, 1995). Variability in employment needs did not significantly impact the use of contingent workers, contradicting the findings of DavisBlake and Uzzi(1993). Also, most of the growth over the past three years has been in the contingent, not full time workforce. While this study does include an analysis of the level of use in each organization, it does not examine the capacities in which these workers are used. Also noteworthy is its analysis of food service workers, who fit the description of open market labor. This may partially explain the contradiction with the DavisBlake and Uzzi study, and also lend more support the suggestion that not all forms of contingent work are equal. Further, the results are based on a response rate of 5% from a survey administered; it is not clear whether the nonrespondents were qualitatively different than the respondents.
Pfeffer (1994) attributes the growth of the trend towards employing more contingent workers to several organizational factors. First, bringing in a contingent worker is much quicker than going through the recruitment process for a full time employee. Second, a unit may not have the authorization to hire a full time employee, but may be able to circumvent the system by bringing someone in as a contractor or a temporary worker. Third, the contingent worker can absorb demand fluctuations. And fourth, these employees presumably cost less than full timers do. The downsides include lower levels of loyalty, dedication, and the willingness to expend extra effort. This trend, the author notes, has been found to lead to lower levels of quality service provided by some hospital nursing staffs, and more accidents in the petrochemicals industry, for example. The author concludes that companies should be cautious about using these kinds of employees.
Some research at the microlevel has also been completed. These studies focus primarily on contingent workers hired through agencies for positions where the supply generally exceeds the demand for these specific skills. It concentrates on two areas: whether traditional and nontraditional workers are fundamentally different from each other and how the presence of nontraditional workers influences the traditional workers with whom they are in contact. These studies indicate that contingent workers do not exhibit attitudes and behaviors that are significantly different from traditional employees (Kidder, 1995; Porter, 1995). These findings are interesting; they contradict popular business press articles which depict contingent workers as less serious about work responsibilities. They are also interesting to consider in conjunction with some of the organizational level hypothesizing in this area. While nontraditional and traditional workers are not significantly different, the presence of nontraditional workers may negatively impact traditional workers with certain job characteristics. For example, Pfeffer (1994) suggests that quality suffers when contingent workers are used. If empirical studies show that traditional and nontraditional workers are more similar than different in attitudes and behaviors, the source of lost quality must lie elsewhere, perhaps in how the organization uses different types of workers.
The above studies suggest several environmental and organizational factors that may contribute to the use of contingent workers. The organizational level studies find that levels of bureaucratization, organizational size, percentage of managers, the presence of formalized evaluation systems, levels of internal training, and rewards for longterm service may play a role in determining the level of contingent work use in firms. Interestingly, this list of items could be used to define many of the differences between entrepreneurial and nonentrepreneurial firms. The salience of these items, then, suggests that entrepreneurial firms may, indeed, have dramatically different use patterns than nonentrepreneurial firms. Also, these studies produce mixed results on what the contributing factors are, suggesting that aggregation across different types of contingent workers may be problematic, that there are mixed motives on the side of organizations for using such workers, and that this is a complex phenomenon that does not lend itself well to simple operationalizations.
With the exception of the work by Pfeffer which uses popular business press articles as its main information source, the organizational level studies all look at use of contingent workers as the dependent variable; they do not consider the impact of such use on organizational processes or outcomes. Perhaps one of the most intriguing issues surrounds the gulf between micro and organizational level phenomena; microlevel research finds that traditional and nontraditional workers are not substantially different from one another, but organizational level research contends that the use of contingent resources negatively affects organizational processes and outcomes. This begs the question of why.
In sum, existing research leaves unaddressed several key issues. First, entrepreneurial and nonentrepreneurial firms may use contingent work in fundamentally different ways. Second, confounding results due, in part, to aggregating many different kinds of contingent work use across different industries suggest that a more finetuned examination of what the mixed motives and use patterns are within certain industry groups is warranted. Third, the question of how this fundamental reorganization in the world of work affects organizational processes and outcomes remains unaddressed.
© 1997 Babson College All Rights Reserved
Last Updated 03/21/98