Greater media choice increases inequality in political involvement. The last chapter demonstrated that the growing reach of cable television and the Internet has widened gaps in news exposure, political knowledge, and turnout between those who like news and those who prefer entertainment. But the Conditional Political Learning model and the empirical analysis had little to say about average political involvement. Increases in news exposure, knowledge, and turnout among news-seekers offset declines among entertainment fans. My theory does not predict whether or not one change exceeds the other to produce changes in average political involvement. Yet, even in the absence of firm theoretical expectations, the effects of new media on the size of the news audience and the level of electoral participation are important and have recently been subject to wide speculation.

In this chapter, I examine the effects of greater media choice on average news exposure and turnout. As far as turnout is concerned, assessing the relevant average is straightforward — it is simply the turnout rate. The following analysis examines if increasing cable penetration in the 1970s and 1980s affected average turnout in presidential and congressional elections. It relates changes in cable penetration in different media markets to changes in turnout rates in those markets. This time series research design provides rigorous estimates of cable television's causal effect on average turnout. It also allows an additional test of the Conditional Political Learning model because cable offered very little news before CNN began operation.