December 2009
Article
Administrative Science Quarterly

Hiding the Evidence of Valid Theories: How Coupled Search Processes Obscure Performance Differences Among Organizations

By: Nicolaj Siggelkow and Jan Rivkin

Format:Print
| Pages:33

Abstract

Theorists argue that an organization's high-level choices, such as its organizational design or the attributes of its top management team, should influence its performance, yet empirical researchers have struggled to detect such influence. The impact of high-level choices may appear weak, we theorize, because the choices are embedded in coupled search processes. A coupled search process exists in an organization when managers search for high-level choices that shape the search for low-level, operational choices, which in turn determine performance. Using a simulation model, we show that coupled search processes obscure the performance impact of high-level choices through two mechanisms: (1) a survivor effect, arising because firms that persist with poor high-level choices are those that luckily happened on good low-level choices despite their poor high-level choices and (2) a wanderer effect, arising when firms use good high-level choices to find good low-level choices and achieve strong performance, but then wander toward poor high-level choices. We show that these effects are particularly strong in stable environments, and we identify empirical strategies that can tease out the true performance impact of high-level choices.