ees with experience for what have been seen as entry-level management jobs and away from hiring new entrants straight from college (Rynes et al., 1997).
The operation of firms through informal supplier networks, formal joint ventures, and similar arrangements has become the norm in some high-tech industries, especially those associated with Silicon Valley and similar industrial communities like Austin, Texas. Descriptions of the operations of these companies suggest that the flow of employees across companies is a distinctive characteristic of their operations, providing the crucial source for the transfer of technology and ideas, learning by employees, and flexibility in terms of a company's ability to restructure and change. For employees, this model of how a firm is organized makes it almost necessary to move across employers in order to progress in a career, given that the relatively flat structure of firms provides little prospect for internal promotion and that training and internal development have been replaced by hiring from outside as a means for meeting skill needs (Saxenion, 1994). The growth of outsourcing and greater reliance on the supply chain outside the firm (Lewis, 1995) in other industries (Gereffi, 1994) creates similar opportunities for employees to get exposure to information and opportunities outside their firm.
Especially for managerial, professional, and technical employees, career paths across firms may become more common as employees move across organizations to perform similar work. If so, it is important to consider their attachment to occupations. Rose (1995), presenting limited information on the stability of employment within occupations in the 1970s and 1980s, found some evidence of modest increases in the length of time that individuals are remaining in the same occupation. Ultimately, this question calls for an analysis as detailed as that applied to job stability and job security.
Although an analysis based on retention rates is required to draw firm conclusions, Table 3.2 nonetheless reports some preliminary calculations based simply on the distributions of incomplete spells of occupational tenure in the 1983 and 1991 Current