Opportunity and Constraint:
Chain-to-Component Transfer Learning in Multiunit Chains of US Nursing Homes 1991-1997
Will Mitchell1, Joel A.C. Baum2, Jane Banaszak-Holl3, Whitney B. Berta4, Dilys Bowman3
1: University of Michigan Business School
2: University of Toronto Rotman School of Management
3: University of Michigan School of Public Health
4: University of Toronto Department of Health Administration
AbstractHow and when do multiunit chains affect the capabilities of their component units? Chains are collections of component units that produce similar goods and services, linked together into larger 'superorganizations' (Ingram & Baum 1997). This study explores how the level and similarity of capabilities of chains and their components - both existing and newly acquired - affects transfer learning across component units. Transfer learning occurs when one organization causes a change in the capabilities of another, either through sharing experience or by somehow stimulating innovation. Attention to transfer learning processes is critical to understanding organizational performance because it is one of the most important routes through which organizations develop competitive advantage (Capron & Mitchell 1998). Within chains, transfer learning leads to changes in the capabilities of component units and, in turn, to changes in component performance.
Our attention to transfer learning processes within chains is premised on the belief that how chains change and deploy their knowledge is key to their performance. A fundamental question, therefore, is what are the potential performance implications of the support for our model? Three observations seem relevant here. First, standardization should lead chains to prefer to operate similar components, and for those that do to outperform those that do not (Ingram 1996; Ingram & Baum 1997). Second, capabilities may not transfer easily between chains and components that emphasize different capabilities. Forcing chain-to-component transfer of dissimilar capabilities could be worse than useless; it could even be harmful if the chain's managers are unable to differentiate capabilities that apply from routines that do not (Mitchell 1992; Greve 1999; Ingram & Baum 1997). Third, chains are more knowledgeable about the nature of competition they face in their current service specialties than they are about competition in other service areas in which they would be exposed to a different and unfamiliar set of competitors. Performance often declines when chains develop new specialized services.
Taken together, these three observations suggest that the greater the variety of capabilities that chains transfer to components in chain-to-component transfer learning, the poorer the performance of the components will be. Given this conclusion, the damping effects of the similarity-capability interactions on capability transfer is consistent with the notion of absorptive capacity, the standardization benefits at the core of the chain strategy, and, in turn, with enhancing performance. However, notwithstanding the moderating effects of similarity, component capabilities still changed most when chain and component were most dissimilar. That is, high-capability chains transferred knowledge and resources to low-capability components, while low-capability chains required high-capability components to switch to capabilities with which the chain had more experience (but may not fit the component). Thus, although support for our transfer learning model is consistent with improved performance, it also points clearly to the boundary conditions for business change of such potential benefits, as well as the significance of the acquisition selection process to chain performance.
To appear in "Strategic Management of Intellectual Capital and Organizational Knowledge" edited by Nick Bontis & Chun Wei Choo (Oxford University Press).