Public-sector innovation fuels corruption, waste, lack of accountability and

A first perverse effect is the “lack of stability” that results from the emphasis on continuous innovations. Although change may often be desirable, innovations may disrupt things that actually work well (academic respondent 1). The lack of stability can have negative consequences for people working in an organization in the form of stress but also for customers and external relations since they do not know what to expect from the organization. For China, it has been argued that policy innovation may create uncertainties and threaten social stability, economic growth, and ultimately the legitimacy of the state (Zang & Musheno, 2017). This lack of stability may not only be a problem for China but also for the Western governments.

A second perverse effect is that innovation may result in “illegal practices”: if there is no control the emphasis on innovation can result in(semi-)illegal actions (Fuglsang & Sundbo, 2016). Rules have been developed to regulate behavior and for this reason they may inhibit new approaches. A recent example from the Netherlands illustrates this risk. The tax department had opened up an innovation lab to experiment with the use of data science to improve its regulatory processes. The application of data science, however, resulted in a variety of privacy breaches which were only exposed at a later stage (Zembla, 2017).

A third perverse effect is that public innovation may result in “corruption.” In innovation projects with public and private parties, regular procedures for collaboration generally do not apply and there may be a risk of corruption (Borins, 2002). There may be too much freedom for individuals and perhaps people are not speaking up when integrity may beat risk. Corruption can occur if there is only attention for stimulating individual initiative as a requirement for innovation and little attention for control. Measures against corruption tend to put much emphasis on rules and procedures and these are generally regarded as barriers to innovation. In addition, a relation of trust instead of strict control may be beneficial to learning processes around public innovation but the intimacy that is created to support innovation also may result in patterns of corruption.

A fourth perverse effect is that public innovation results in a waste of public money. Public innovation projects often fail to deliver on their promise (academic respondent 1) and this can mean that public money is spilled. Jordan (2014, p. 75) refers to “pork projects” that are of doubtful utility but represent flashy changes which may play a part in a public relations campaign to justify increases in budgets or other resources. Osborne and Brown (2005, p. 190) stress that a risk of innovation is that it will cost more than was intended. This perverse effect is very strongly present when innovation is seen as a goal in itself and the outcomes of innovation are not measured (academic respondent 2) and also when the innovation is ICT-driven. Academic respondent 3 mentioned the public service broker and e-voting in Ireland as examples of a waste of money on digital innovation.

A fifth perverse effect of public innovation is “absence of democratic control.” Public innovation entails that new solutions are developed and tried out at the work floor. One of our key informers referred to this as a process of de-politicization (academic respondent 4). This also means that these solutions have not yet been approved by democratic representatives. In technological innovation, politics are more or less driven out by the idea that making decisions is just a rational process of having the “right” information (Bekkers et al., 2006). This may not be a problem when the innovation relates to operational issues, but it may be more dangerous when the innovation entails more substantial issues.

A sixth perverse effect of public innovation is “damage to local initiatives.” Broersma et al. (2016) indicate that national approaches to citizen participation may force diverse approach to follow one fixed method. The focus on rolling out new generic solutions may be damaging to local approaches when funding is provided for generic solutions and local solutions are not supported (academic respondent 4). An example mentioned by the academic respondent was the compulsory nature of certain projects for neighborhood participation in the Netherlands. One can also think of centrally enforced standardization of processes or information systems that limits the possibilities for local innovation.

A seventh perverse effect of public innovation is the “disruption of a power balance.” A disruption in itself is not necessarily a perverse effect, but it is when public innovation results in a further concentration of power in the hands of the powerful (academic respondent 1). Larsson and Brandsen (2016) stress that the politics of social innovation are often “denied”: innovations are presented as purely instrumental changes. Jordan (2014, p. 79) concludes that innovation may entail uses of power that need to be checked against the possibility of abuses. The political aspects and effects on power distributions often remain in the dark. This insight is in line with the work of King and Kraemer (2012) and more specifically with their “reinforcement” theory with regard toward ICT. They state that ICT systems are unlikely to diminish the power of interests who can stop such effects (e.g. through legislation) or who control the development and deployment of ICT itself. Furthermore they conclude that “it is naïve to plan for ICT developments with the expectation that such power effects will result” (King & Kraemer, 2012, p. 293).

An eighth perverse effect of public innovation is undesirability: practices or processes may be developed that produce “negative outcomes for stakeholders.” Hartley (2005) refers to innovation that do not constitute an improvement since, for example, increased choices are not desired by service users. In addition, Osborne & Brown (2005, p. 190) indicate that the innovation may be pursued for ideological reasons rather than for its contribution to the efficiency or effectiveness of public services. Similarly, Brandsen et al. (2016, p. 307) stress that a significant part of social innovations represent cultural, economic and social aims and practices that are highly controversial or even seen by many as threatening rather than promising. These innovations may stress the perspective of only a small group of actors—i.e. they are one-sided—or they may conflict with human values. An example that is often highlighted—and that has reemerged around the discussion about robotization—is that innovation results in the loss of jobs. The most extreme example is this respect are concentration camps. These camps qualify as public innovations in many definition but they clearly represent a highly undesirable change. The undesirability in our days often relates to problems of exclusion of certain groups of citizens. Academic respondent 5 mentioned the exclusion of citizens due to legally enforced digitization as an example of a practice that is regarded as undesirable by many actors. Similarly, academic respondent 4 mentioned a case in Ireland where a person was denied e13because she refused to get a public service card.

A ninth perverse effect that we have identified is “the technocratic dominance in public processes.” Larsson and Brandsen (2016, p. 295) highlight the denial of politics. This perverse effect especially occurs if the technology used is rather new or complex and requires specific knowledge which many people lack. The lack of knowledge can play a role at the operational level, but also at a more political level. Often politicians do not have enough knowledge to accurately steer, judge or control the way in which public sector organizations use specific technology. This perverse effect comes into play when algorithms are used within a public sector context as O’Neil (2016) points out. More generally, the perverse effect associated with ICT-driven innovations is that they possible lead to a dominance of what is seen as “objective” and “hard” information (Bekkers et al., 2006). Especially when Big Data are concerned there often is a tendency to use the available data to work toward more “evidence based” decision making.