Theprevious chapter showed that the preferences of Scheduled Caste voters across parties correspond directly to their representational profile. In this chapter, I argue that for substantial numbers of Scheduled Caste voters, preferences do not automatically translate into votes. Even though they prefer the BSP, these voters are likely to vote for the BSP only when they expect their vote to install the BSP in government, or to affect the victory or defeat of one of its competitors. Where they do not expect to be able to affect the outcome in one of these two ways, they are unlikely to vote for the BSP. These Scheduled Caste voters, in other words, are strategic rather than expressive actors: They treat the vote as an instrument through which to obtain the best possible outcome rather than as an opportunity to declare their preferences. This chapter does not offer a precise estimate of the proportion of strategic voters among Scheduled Castes. However, it does provide evidence to indicate that this proportion is substantial, and that without the support of such voters, the BSP could not obtain majority support among Scheduled Castes.

Expectations of efficacy, in turn, depend upon the distance between the numerical strength of the Scheduled Castes in the electorate and the threshold of winning or leverage imposed by the competitive configuration. Scheduled Castes know that others from their category also prefer the BSP to other parties. They also know that those groups not represented by the BSP (or its partner, where it contests elections in alliance) are not likely to prefer the BSP. Simply by counting heads in their constituency and in their state as a whole, therefore, they have the capacity to make some prediction about the efficacy of their vote.