This study aims to answer the following research question: compared to solo programming, is pair programming a more cost effective method for developing software? This paper proposes a research model based on theories and previous empirical studies. It discusses a multi-study approach to address the question in hand. The first study is a survey of practitioners in regards to their experience and perception of the cost of pair programming. Information acquired from the survey are then fed into simulation models as input parameters with the purpose to identify situations where pair programming is or is not more cost effective than solo programming.