vides some support for this notion. Meyer and Rowan (1977) write that myths explain organiza¬tional similarity, and that institutionalization of a set of rules within an organization is an attempt to attain legitimacy, to enhance resources and to enhance survival prospects. DiMaggio and Powell (1982) also stressed the fact that there are three reasons that can lead organizations to become similar to one another. First, organiza¬tions can be coerced from above to make changes to their form. Second, organizations can decide to mimic the form of another organization that is perceived as successful. Third, in some cases, professional norms can lead managers to reshape organizations in 'a pattern that leads to high levels of similarity between organizations.
Both sets of authors leave students of perfor¬mance pay systems with much to consider. As we noted earlier, the simplicity of the theoretical underpinning of performance pay systems is very attractive to managers and policy makers. However, the ideas offered by Meyer and Rowan (1977) and DiMaggio and Powell (1982) indicate that the adoption of performance pay plans may have as much to do with the way ideas spread across organizations as with the attractiveness of the rational underpinnings of expectancy theory.
A logical extension of this line of thought on performance pay systems suggests that the adoption and implementation question may not have a neat answer at all. As Ingraham (1993) notes, the implementation of performance pay schemes is often the result of a policy diffusion process that takes place without consultation. Instead, Ingraham notes that the popularity of performance pay systems in the private sector has led to large-scale diffusion in the public sectors of the OECD countries. In short, while higher degrees of pay decentralization may allow governments more leeway in experimenting with performance pay, there are still no plainly identi¬fiable macro-level variables that predict when performance pay schemes will be adopted or implemented.
What about the predictability of different types of performance pay schemes? As Milkovich and Wigdor (1991) point out, policy makers and managers have a variety of choices when deciding which performance pay plan to adopt. Within the OECD countries, most perfor¬mance pay plans consist of individual merit increments and bonuses (OECD, 1993, 1997). While there is little direct evidence to suggest why these two types of performance pay are the most popular, some conjecture is possible. For example, from an employee's standpoint, pay