Advocates of the high-commitment or high-performance model of HR practices (HPHR) argue that this model is universally 'best', as it exhibits the potential to outperform all other models of HPHR practice (Appelbaum and Batt, 1994; Pfeifer, 1994). Some studies indeed support this claim, indicating a positive effect of these practices on performance indicators such as productivity and quality (for example Arthur, 1994; Ichniowski, Shaw and Prennushi, 1995; Kalleberg et al., 1996). If these practices are even partially as effective as they are claimed to be, then they should be of great appeal to a wide array of companies. Yet rather than a rapid diffusion, an increasingly routine finding is that the take-up of these practices appears to remain low (Wood and De Meneze, 1998; Roche, 1999; Osterman, 2000; Cappelli and Neumark, 2001). To explain this apparent reluctance in HPHR adoption, two prominent issues are considered: first, the contingency literature highlights the role of strategic integration, which sees one area of company activity, say HR management, being influenced by and reflecting the goals of another, for example, the company's business and/or production strategies (Schuler and Jackson, 1987; Arthur, 1992; MacDuffie, 1995, Wood, 1999); second, a principal assertion of the neo-institutionalist perspective is that under conditions of high uncertainty the assumption that companies are predominantly guided by efficiency concerns can be misplaced (Meyer and Rowan, 1977; Haunschild and Miner, 1997; Shaw and Epstein, 2000). For example, in the early stages of diffusion, the effectiveness of new practices is as yet, for many companies, unclear. Under such circumstances, these companies are more likely to be reluctant to adopt HPHRs. [ABSTRACT FROM AUTHOR]