Public procurement is considered a tool for both purchasing services and triggering innovation. However, public procurement for innovations (PPI) is considered to be a challenging process. Studies concerning PPI have been conducted from perspectives of e.g. risk management (Kalvet & Lember, 2010); demand-side policies (Edler & Georghiou, 2007; Edquist & Zabala-Iturriagagoitia, 2012); policy instruments (Vecchiato & Roveda, 2014; Georghiou et al., 2014); and practices that support or induce barriers to innovations (Rolfstam, 2012; Edler et al., 2011; Uyarra et al., 2014). However, very few studies explore PPI in health and social services (Yeow & Edler, 2012; Pelkonen & Valovirta, 2015). 

According to the extant literature, innovations can be supported by using certain types of procurement practices e.g. outcome-based specifications, early interactions with suppliers, and requirements for innovation in tenders (Edler et al., 2011; Uyarra et al., 2014; Georghiou et al., 2014). By contrast, barriers that are said to prohibit PPI include e.g., over-specified tenders, the overall complexity of the procurement process, and the lack of early interaction between purchasers and providers (Uyarra et al., 2014). Furthermore, also context-related characteristics create challenges within PPI. For instance, in health and social services, manifold laws guide service production, which, in turn, influence the process of procurement (Pelkonen & Valovirta, 2015). Given the financial significance of public procurement of health services in the European Union, it is surprising how sparse academic research about PPI in health and social services has been. This paper explored public procurement in health and social services and the barriers to PPI. Following research questions were comprised: what innovations supportive procurement practices are applied in health and social services and what barriers prohibit their application? 

The empirical data were collected with 18 interviews from four purchasing organizations in Finnish municipalities. These organizations used different strategies in organizing and in procuring health and social services. Analysis revealed that innovations supportive purchasing practices were emerging in this field. However, added to the barriers already mentioned in the extant literature, also several context-related barriers preventing PPI were found. Context-related barriers included uncertainty of demand, rigid regulation of context, an ethically sensitive context, multiple suppliers, and complex environments, as well as challenges in outcome definition and measurement. For example, according to the purchasers, uncertainty of demand prevented long contracts. Despite the fact that outcome based procurement would allow providers to adjust to demand, purchasers were unwilling to use it because of problems in defining the outcome in health and social services.

Purchasers described dreadful stories about the risks related to innovation supportive procurement practices and therefore used more traditional and “safe” practices. Purchasers have been blamed for avoiding risks too extensively and active risk management has been highlighted (Kalvet & Lember, 2010; Elder et al., 2015). However, in the context of health and social services, there are excessive financial and ethical risks that can result from unwanted service outcomes. Perhaps general recommendations should not be provided either in PPI or in risk management, instead more studies presenting context-specific good practices could be needed.