Contingent Employment, Workforce Health, and Citizenship By Marc ...

2001; Rebitzer, 1998). In their review of 93 studies on the health effects of contingent work, Quinlan et al. (2001) found that approximately 80% of these studies showed a clear adverse association between contingent work and health. A subsequent and more extensive review (of 141 studies) confirmed these initial findings (Quinlan, 2003).

Another, more modest body of literature suggests that contingent work might be beneficial for workers’ health (Artazcoz, Benach, Borrell, & Cortès, 2005; Benach et al., 2002; Benavides, Benach, Diez-Roux, & Roman, 2000; Sverke, Gallagher, & Hellgren, 2000; Van Dyne & Ang, 1998; Virtanen, Kivimki, Elovainio, Vahtera, & Ferrie, 2003; Virtanen, Vahtera, Kivimäki, Pentti, & Ferrie, 2002). In these studies, such work was found to be associated with less stress and fewer mental health problems (Benavides et al., 2000), lower levels of somatic complaints (Sverke et al., 2000), and better self-rated health (Virtanen et al., 2002). Notably, though, some of these studies also report inconclusive or contradictory findings (Aronsson et al., 2002; Artazcoz et al., 2005; Bardasi & Francesconi, 2004; Bernhard-Oettel, Sverke, & De Witte, 2005; Martens et al., 1999).

Analysis of the literature reveals a number of possibilities for making sense of the conflicting health findings. Definitional and conceptual disparities characterise the literature on contingent work. The term ‘contingent work’, for instance, is used to refer to a myriad of work arrangements, and embedded within it are numerous concepts and assumptions about the nature of contingent work and employment contracts. Analysts have proposed the use of classification schemes for the purposes of research, but there is no agreement among researchers on a coherent schema (Benavides et al., 2000; De Cuyper, De Witte, & Isaksson, 2005). Consequently, the conceptions of contingent work that are used in research studies tend to vary depending