An Empirical Framework for Testing Theories About Complimentarity in Organizational Design

摘要: This paper studies alternative empirical strategies for estimating the effects of organization design practices on performance, as well factors which determine organizational design, in a cross-section firms. Our economic model is based firm where multiple are en endogenously determined, and these affect output through an 'organizational production function.' The econometric includes unobserved exogenous variation costs returns to each individual practices. used evaluate how different testing theories about complementarity can be interpreted under assumptions statistical environment. We identify plausible hypotheses joint distribution unobservables several approaches from existing literature will yield biased inconsistent estimates. show that sign bias depends two factors: whether organzational complements, correlation between practice. find sets conditions we provide interpretations. analysis shows particular set hypotheses, variety procedures may all qualitatively similar biases, presenting challenge identification complementarity. then propose structural approach, system simultaneous equations describing productivity demand As long variables observed uncorrelated with practices, parameters identified, yielding consistent tests cross-equation restrictions implied by static optimization organizatin's profit function.