My objective in this chapter is to develop a general model for comparative analysis of the way in which the context of a positive discrimination (PD) policy influences its likely consequences-i.e. its benefits and costs. I will then utilize this model in the following two chapters to undertake a qualitative comparative analysis of PD policies in the US and India.
The success or failure of a PD policy favoring an under-represented ethnic group (UREG) will clearly depend upon the context in which it is applied. In analyzing any given PD policy of this kind, we would like to know the extent to which the potential benefits and costs listed in the previous chapter will actually be realized. In some cases, under some conditions, there may be many benefits and few costs; under other conditions, in other cases, the opposite may be true. In order to analyze how variations in the context of a PD policy affect its chances of success or failure, it will be helpful to formulate a qualitative causal model linking various kinds of conditions to different kinds of potential benefits and potential costs.
What kinds of conditions are likely to affect the extent to which a given PD policy generates beneficial as opposed to costly consequences? First of all, the characteristics of the PD policy itself-its procedures, how they are implemented, etc.-will surely be important. As well, certain characteristics of the under-represented ethnic group favored by the PD policy are bound to matter. Furthermore, some characteristics of the societal environment within which the PD policy is applied are likely to make a difference. These three kinds of characteristics I will label "primary factors"; they are primary in the sense that they define the PD policy and the relevant context in which it is being carried out. Some or all of these primary factors may differ as between different PD policies and, as I will show below, variations in the characteristics of each primary factor have predictable effects on the consequences of a PD policy.
Two other kinds of circumstances associated with a PD policy seem highly likely to affect the benefits and/or costs it generates. The first and most