This thesis analyzes the extent to which governments are able to use collaborative partnerships between educational institutions and businesses in order to improve the education of the (future) labor force and address issues such as the ‘skills gap’. The aim of the research is to discover whether there is a causal link between the governance regimes in place for partnerships and problem-solving behavior within these partnerships. A large-scale quantitative analysis combined with in-depth case studies revealed that the network features of the governance regimes were stronger than the principal-agent features, and that this approach was generally successful in guiding partnerships towards a set of mixed activities that were all relevant to the overall policy aims, with a relatively high self-reported success rate of 40%. However, the framework has been unable to overcome the fact that the educational institutions have become ever more dominant within the partnerships, prioritizing the area of initial education and directing the majority of funding towards this area. This dominance was reinforced because the principal-agent features of the regime echoed familiar routines in which the government and the educational institution are in a bureaucratic relationship. Implementing the principles of experimentalist governance could provide an opportunity to combine accountability with learning, and enable a stronger focus on more challenging outcomes and prevent one particular (type of) actor becoming too dominant, while also avoiding heavy-handed, top-down governance methods.