Skip to main content

As we have seen from this brief review of research findings, the explanatory power of our preliminary model suffers from being too linear. Although it incorporates several influential relationships among dimensions, we learned from the case data that more powerful feedback loops are at work that explain more about the key dynamics of the projects. Although the preliminary model hypothesizes the unfolding of projects in stages, it does not account well for the dynamics of these efforts over time. In most cases we could not discern the discrete and predictable stages that the model contemplates. While the model does prompt us to look for differences in goals and behavior from earlier phases to later ones, it does not lead us to expect the constant learning, iteration, and adaptation that we heard in the interviews about how and why these changes took place.

As one example, the preliminary model views service and collaboration performance as the final outcomes of the collaboration effort. However, the evidence suggests that performance is an ongoing factor and one that is strongly related to other dimensions in the model. For example, the case data revealed that early performance influenced the objectives, motivations, and contributions of different stakeholders at different points in time. Many interviewees said that small, early successes encouraged sponsors to persevere and more stakeholders to participate. However, in the case of Bremen Online, early success may not be enough for smaller private players to sustain their participation into the future. The initiative is funded initially by the federal government and large telecommunications and financial institutions. When the government funding ends, the program must become self-sufficient. The large companies can factor the public funding-to-private revenue transition into their financial plans; but smaller ones are likely to quit the partnership, despite its substantive success, if substantial revenue cannot be generated immediately. On the other hand, in at least three cases, (NYS GIS, Cadastre Quebec, and Access Indiana) early failures motivated the participants to rethink, revise, and renew their relationships and methods.

Another problem with the preliminary model is it combines collaboration performance and service performance into a single construct. However, most interviewees considered collaboration performance to be separate from the performance of the project in terms of its service delivery goals. Even projects that had early failures or took a long time to produce service results, were seen as successful collaborations as long as they contributed positively to shared knowledge, professional networks, and communities of practice. Conversely, as with the New Brunswick case, the service goal was achieved, personal professionalism and trust were built, but one of key organizational partners came away from the experience with a clear decision to avoid such endeavors in the future. Other projects, such as Service Canada put so much attention into successfully delivering the service project on time, that it failed to maintain the political support needed to sustain service integration across multiple agencies. After several years, the idea was abandoned and services were disaggregated back into separate agencies. In sum, a budgetary and technical success was also policy failure.

All of these findings suggest a need to reconsider both the specifications of and the relationships among the dimensions in the preliminary model. One possible alternative is presented in Figure 2 below.

In this revised model, we acknowledge the pervasive influence of the political, social, economic and cultural environments, as well as the institutional, business and technical environments, by nesting these layers of environment and embedding the collaboration initiative within them. This model better conveys the idea that these environments exert both obvious and subtle influences on collaboration projects, participants, and performance.

Figure 2. Revised model of service delivery collaborations

Figure 2. Revised model of service delivery collaborations

This new version of the model also illustrates the importance of the dynamic influences among the dimensions. It suggests that the collaboration process influences and is influenced by the players and their expectations, and by the modes and methods of collaboration they choose to use. The collaboration process leads to performance outcomes in terms of both the collaboration itself and the service goals it seeks to meet. These performance results continually influence the players’ motivations, the tools they choose, and the way they interact. In addition, the feedback arrows shown in this revised model better represent the double-loop learning that appears to take place – the participants not only learned better ways to manage these particular projects, they also learned how to approach collaborative working relationships more generally. This iterative cycle of influences better describes the full range of experiences documented in the case studies and suggests the key factors and dynamics that shape new models of collaboration across programmatic and national boundaries.