View article

Recent empirical studies of firm-level performance have tested complementarity in the case of multiple practices. These papers have drawn conclusions using potentially biased estimates of pair-wise interaction effects. We develop a consistent and simple testing framework and test it against alternatives.