Professional Friendship, Resource Competition, and Collaboration in a Homeless Service Delivery Network

ABSTRACT

While existing literature on human services interorganizational network has focused on collaboration, networked service providers do compete for the same public and private funding sources. Interorganizational competition for scarce resources could be detrimental to collaboration. Yet organizational leaders’ external professional friendship ties with peers – an understudied type of informal relationship – may mitigate the effects of resource competition. Using data collected from a homeless service delivery network, we investigate such relational dynamics. Quantitative network analysis shows that professional friendship ties facilitate collaboration and moderate the negative impact of competition on collaboration. Both funders and managers need to invest in developing professional friendship networks.

KEYWORDS:

• Collaboration
• competition
• professional friendship
• homeless service network

Notes

¹ The lead author helped facilitate multiple strategic planning sessions for both the regional homeless service networks and individual homeless service organizations. During the strategic planning process, the four areas were identified by the participating organizations as crucial collaboration items. When the authors designed the survey, they reached out to local leaders of homeless service organizations and elicited their feedback on the key areas for collaboration in homeless housing service delivery.

² Habitat for Humanity usually does not provide shelter services directly. The study area is prone to natural disasters, especially hurricanes and tornados. The local Habitat for Humanity organization runs an active disaster response program that provides shelter assistance services, help repair the damaged housing, and help the homeless transition to permanent housing.

³ This national directory provides a comprehensive list of homeless shelters, transitional housing, and other homeless service organizations in all 50 states. The website was created and maintained by volunteers who aim to provide quality online resources for the homeless population. This website has been one of the top search sites according to Google statistics. Many homeless service organizations or networks have embraced this directory. For instance, the National Coalition for the Homeless has included this directory on its website.

⁴ We treated the unanswered as missing data in the network. A response rate of 79.4 % is acceptable for this study for several reasons: First, seven organizations did not respond to the survey for random reasons. They may have declined to respond because they did not feel comfortable sharing their information about their relationships with others or they did not enjoy collaborating with other service providers. Second, we checked the data and found that these organizations were marginal to the networks of our study, as few other responding organizations had ties with these organizations. Finally, the quantitative network analysis in this study, MRQAP, is relatively robust to missing data because of its permutation techniques (Borgatti et al., 2013).

⁵ We used a 0–4 scale (0 = no collaboration, 4 = very frequent collaboration) instead of asking about whether the organizations have collaboration on a weekly, monthly, or quarterly basis, because recalling the exact frequency may put heavy response burden on the respondents.

⁶ The correlations between competition variables and collaboration are positive. This mixed result might come from the intertwined impacts of multiplex interorganizational relationship on collaboration. When we hold constant the effects of other variables in the following multiple regression model, the relationships become negative, as expected.

⁷ We averaged the scores for collaboration items to make it have similar scale as competition. The other key variables, such as professional friendship and information sharing, were measured in general terms. In addition, we separated the items and conducted analysis to test the idea of separating the collaboration items; however, our analysis could not produce consistent findings to interpret. This might be because organizations collaborate more often in case management than in grant writing. Therefore, there are big variations in the number of collaboration ties and the strengths of ties across different collaboration areas.

Additional information

Funding

Qian Hu would like to acknowledge funding support from the Center for Public and Management in the School of Public Administration at the University of Central Florida.