The innovators scalpel: A knowledge-based view of corporate divestitures & technological performance
研究了企业通过剥离业务来重组知识结构的影响。基于919家高技术企业的数据发现，剥离通常降低创新产出，但在提升知识连贯性、降低分散度或协调负担时可能改善绩效。适合关注企业边界与创新的读者。
Extensive research examines how firms expand their knowledge bases through alliances and acquisitions. Considerably less attention has been paid to how firms reshape their knowledge bases through divestitures. Drawing on the Knowledge-Based View (KBV), we conceptualize divestitures as mechanisms through which firms reconfigure the structure and governance of their knowledge bases. We argue that, on average, divestitures are associated with weaker technological performance because they reduce the size of firms’ knowledge bases, disrupt innovation routines, and create organizational uncertainty. However, we also argue that this negative association is weakened, and in some cases reversed, when divestitures increase knowledge coherence, reduce knowledge dispersion, or lower coordination burden. We test our hypotheses using a sample of 919 unique firms in four high-tech industries that announced 1794 divestitures and filed 414,128 patent applications over a 15-year period. Our findings show that while divestitures are, on average, associated with lower innovation output, this association depends on the form of knowledge complexity divestitures remove. Knowledge Coherence conditions patent quality and breakthrough inventions, but only at relatively high levels of coherence, observed in 12.2% and 20.2% of divesting firm-years, respectively. Knowledge Dispersion conditions patent quantity and patent quality at more moderate levels, observed in roughly one-quarter of divesting firm-years. Coordination Burden shows the broadest pattern, conditioning all three innovation outcomes and exceeding the relevant thresholds in more than 60% of divesting firm-years. In doing so, our study contributes to the corporate strategy and innovation literatures by showing that divestitures are not merely acts of scale reduction, but can function as knowledge-based boundary choices that reshape technological performance by selectively cutting away excess complexity.