ABSTRACT
The Manifesto Group data series has only one equivalent of equal weight in the whole of political science, namely the time-series of national election surveys in most democracies. The world needs both series if it is to test even the most commonplace generalisations about the interaction between voters and politicians. However, many observers are sceptical of the value of Manifesto Group data, unlike that of the national election studies. This section considers the criticism that content analysis, however sophisticated, offers only garbage in, garbage out. It defends the method by reference not only to the MRG/CMP project but also to the other most important content analysis in political science, namely W.H. Riker’s analysis of arguments for and against ratifying the Constitution of the United States. The second section compares the Budgean analysts of the CMP with other work in neighbouring fields of political science. The final section considers how many political dimensions there truly are in a large democracy.
