Logistic Regression, Segmentation Modeling and Governance Choice in the Waste Management Industry

To acquire technologies in the waste management industry, firms can elect vertical integration, alliances or contracts as governance modes. Using a two-tier questionnaire to build a 434 transactions data base, this research tested hypotheses based on transaction costs and resource-based constructs. The same hypotheses were examined with two different statistical models : multinomial logistic regression and segmentation modeling. The transaction cost rationale is respected. Vertical integration is chosen under conditions of high competence and high redeployment costs. High external uncertainty and low competence induce waste management firms to prefer collaboration over other governance choice. Multinomial logistic regression and segmentation modeling (SM) give complementary and robust results while SM is more straightforward to use and interpret.