Innovation in the public sector: a systematic review and future research agenda

TL;DR: In this paper, the authors investigate 181 articles and books on public sector innovation, published between 1990 and 2014, and develop an empirically based framework of potentially important antecedents and effects of public-sector innovation.

Abstract: This article brings together empirical academic research on public sector innovation. Via a systematic literature review, we investigate 181 articles and books on public sector innovation, published between 1990 and 2014. These studies are analysed based on the following themes: (1) the definitions of innovation, (2) innovation types, (3) goals of innovation, (4) antecedents of innovation and (5) outcomes of innovation. Based upon this analysis, we develop an empirically based framework of potentially important antecedents and effects of public sector innovation. We put forward three future research suggestions: (1) more variety in methods: moving from a qualitative dominance to using other methods, such as surveys, experiments and multi-method approaches; (2) emphasize theory development and testing as studies are often theory-poor; and (3) conduct more cross-national and cross-sectoral studies, linking for instance different governance and state traditions to the development and effects of public sector innovation.

Summary

1 Introduction

• In addressing this topic, the authors embed their research questions in the open innovation debate that stresses the content, course and outcome of the innovation process as the result of complex interactions between intra-organizational antecedents, resources and actors and external, environmental antecedents, resources and actors.
• Then, Section 3, the ‘Results of systematic review’, presents the characteristics of the eligible studies found and provides answers to the research questions listed above.

2.1 Literature search

• Four strategies were used to identify eligible studies (Cooper 2010).
• He defines entrepreneurship as ‘Die Durchsetzung neuer Kombinationen’: that is, the will and ability to achieve new combinations that can compete with established combinations.
• These journals were Governance, Journal of Public Administration Research and Theory, Policy Sciences, Public Administration and Public Administration Review.
• This search was last conducted in April 2014 and generated 89 possible studies for inclusion.
• The authors must acknowledge a potential limitation caused by the search criterion of seeking the terms innovation and entrepreneurship.

2.2 Eligibility criteria

• In reporting the systematic review, the authors adhere to the widely used ‘Preferred Reporting Items for Systematic Reviews and Meta-Analyses’ .
• Topic – Studies should contain the words innovat* or entrepreneur* in their title and/or abstract in order to prevent confusion with related concepts.
• The word ‘public’ had to be included in the title or abstract as their review was focused on innovation in the public sector.the authors.
• All research designs were allowable (e.g. questionnaire, case study, experiment) but case studies that were purely illustrative in nature were excluded.
• Year of publication - Studies were retrieved that were published in the period from January 1990 to March 2014.

2.3 Study selection

• In total, the authors screened around 10,000 studies.
• First, the authors screened the studies by scanning the abstracts and titles.
• The authors then inductively divided the primary studies' findings on the antecedents into four broad categories that refer to four levels: (1) the environmental level, (2) the organizational level, (3) the innovation itself and (4) the individual level.
• New labels for antecedents, goals or outcomes were introduced and others deleted.
• The main aim of this tool is to study the development of a research field over time as ‘by showing the most important publications in a field, ordered by the year in which they appeared, and the citation relations between these publications, one obtains a picture of the development of a field over time’ (p. 803).

3.1 Journals and countries

• The articles included in the systematic review were published in 90 different journals.
• Besides these public administration journals, articles were also found in very specific and dedicated journals such as Health Care Management Review.
• When looking to the various book publishers, most of the books included were published by well-established publishers such as Palgrave Macmillan.
• Many of the studies were conducted in the USA and in the UK (25%, and 19% respectively).
• This suggests that the American - Anglo-Saxon perspective is central when studying innovation, which could have important implications as there might be an institutional bias present.

3.2 Research methods

• Most of the studies analysed were qualitative in nature (101; 56%), mainly adopting a multiple (50) or single case (21) study approach.
• Only a small group of studies (24; 13%) were based on data that were both quantitative and qualitative in nature (for instance, Nählinder 2010).
• Given this approach, the context of innovation and the antecedents within this context have received substantial attention.

3.3 Policy fields and government layers

• Given the broad sweep of their review of public sector innovation studies in general, the authors were interested in identifying the specific policy fields in which the innovations took place as well as the dominant layer of government.
• The largest group of innovation studies were conducted on the local government level (58; 27%, some studies included more than one policy field or government layer), followed by central government (39; 18%) and healthcare (30; 14%), with many of the latter being carried out in the UK (e.g. Turner et al. 2011).
• Some studies also referred to the public sector in general terms without identifying subsectors (e.g. Kumar and Rose 2012).
• In the following sections, the authors provide the answers to their research questions: the definitions of innovation used (RQ1, Section 3.4), innovation types (RQ2, Section 3.5), goals (RQ3, Section 3.6), antecedents in the innovation process (RQ4, Section 3.7 for general and Section 3.8 for adoption/diffusion) and outcomes (RQ5, Section 3.9).

3.4 Definitions used

• The most remarkable finding is that most articles do not provide a definition of innovation (137; 76%).
• When innovation was defined, the definition was often quite general (44 of their sample (24%) used a general definition).
• Also based on Rogers, various authors defined innovation as ‘the adoption of an existing idea for the first time by a given organization’ (e.g. Borins 2000).
• Seventeen studies included both elements (e.g. Salge and Vera 2012).
• Osborne and Brown (2013, p. 3) argue how the distinctive nature and challenges of innovation, as opposed to ‘continuous’ change, can otherwise become lost as innovation can be considered a specific discontinuous form of change.

3.5 Innovation types

• As the definition of innovation in the public sector is often quite broad, innovation types are often specified (Moore and Hartley 2008).
• The authors consider dimensions and types to both refer to the same phenomenon and indeed the terms are often used interchangeably.
• When analysing the studies, each innovation identified was allocated to one of the abovementioned categories depending on its main goal (as identified in the publication studied).
• In the literature, much less attention has been paid to technological process innovations (a subset of process innovations, often related to e-government and redesign), governance innovations and conceptual innovations.
• In summing up, the authors can say that the literature seems to lean towards intra-organizational process innovations, which are often closely related to two major reform movements in public administration, namely NPM and e-government.

3.6 Innovation goals

• Table 3 shows, based on the studies analysed, the goals that public sector innovations sought to achieve.
• The first striking observation is that 35% of the articles studied failed to mention any goals.
• This goal was quite closely followed by goals related to participation and cooperation (on 68 occasions), for instance through involving citizens (e.g. Carter and Bélanger 2005).
• These findings can be related to the two logics of action put forward by March and Olsen (1989) when trying to understand the functioning of the public sector: the logic of consequence and the logic of appropriateness.
• That is, public sector innovation is not only about efficiency but also focused on acquiring trust and legitimacy (e.g. Bekkers et al. 2011).

3.7 Antecedents in the innovation process

• The authors analyse antecedents that were identified as influential in the innovation process.
• Antecedents related to the environmental level Table 4 presents an overview of the antecedents related to the environmental level.
• Table 4 also shows that on eight occasions the number of compatible organizations adopting an innovation was addressed and this, at least partially, fits the notion of normative isomorphism.
• The degree of risk aversion is also mentioned in various studies, including in the description of an administrative culture that hampers innovation (e.g. Borins 2001).
• Having identified these various antecedents, it is also interesting to see whether they are present in both the generation and the diffusion/adoption stages of the innovation process.

3.8 Antecedents related to the two stages of the innovation process

• This section looks at antecedents that are distinctly related to either the generation or the diffusion/adoption stage of the innovation process.
• Almost half of the studies identified (73; 40%) dealt with diffusion and/or adoption, indicating that the diffusion and adoption process is rather well covered, although some authors disagree (for instance, Hartley 2005).
• Whereas the characteristics of an innovation were only considered in studies on the diffusion and adoption stages (e.g. Carter and Bélanger 2005), environmental, organizational and individual antecedents were seen as present in both the generation and the adoption stages.
• When looking at these common antecedents, similar patterns can be found.
• Studies related to the individual level similarly include autonomy and skills in both stages (e.g. Bartlett and Dibben 2002).

3.9 Innovation outcomes

• The authors last research question concerns the outcomes of innovation.
• Studies often mentioned some objective of the innovation in their introduction, such as improving effectiveness and efficiency, but failed to report whether these goals had been realized (e.g. Bartlett and Dibben 2002).
• Where outcomes are reported, studies often record, in line with the goals, increased effectiveness and efficiency (e.g. Dias and Escoval 2013).
• Other outcomes, such as achieving citizen satisfaction, were less often reported.
• Only a few studies describe the pursuit of traditional public sector values such as safety and equality in schooling (Maranto and Wolf 2013).

3.10 Relationship between innovation types, outcomes and antecedents

• After having described the main antecedents and outcomes, the authors analyse whether some innovation types are more closely related to certain antecedents and outcomes than to others.
• This is in line with their previous findings in this section, reflecting a strong emphasis on internal-oriented organizational antecedents.
• Second, governance innovations are frequently connected to environmental antecedents, including the resources of private partners.
• Related to this, the authors examine whether innovation types differ in the way they are connected with certain outcomes (see Table 10).
• Further, Table 10 also highlights the failure of many of the studies to address outcomes.

4 Conclusions

• Quantitative studies, and especially mixed-method studies, were less common.
• The authors found that little attention had been paid to innovation outcomes such as legitimacy, and also that conceptual innovations had scarcely been researched.

Figures

TABLE 4 Environmental antecedents

TABLE 10 Relationship between innovation types and outcomes of public sector innovation

TABLE 8 Outcomes of public sector innovation

TABLE 9 Relationship between innovation types and antecedents in the public sector innovation process

TABLE 5 Organizational antecedents

TABLE 6 Innovation characteristics

TABLE 2 Types of public sector innovation

TABLE 7 Individual antecedents

FIGURE 2 Heuristic framework of public sector innovation

TABLE 1 Public sector innovation types applied

FIGURE 1 PRISMA flow diagram

TABLE 3 Public sector innovation goals

Full text

Innovation in the Public Sector: A Systematic Review and
Future Research Agenda
Hanna de Vries
1
, Victor Bekkers
2
, Lars (L.G.) Tummers
3
Final version for Public Administration
Address correspondence to the author at Devries@fsw.eur.nl
To be cited as: De Vries, H.A., Bekkers, V.J.J.M., L.G. Tummers (forthcoming).
Innovation in the Public Sector: A Systematic Review and Future Research Agenda. Public
Administration
1
Department of Public Administration, Erasmus University Rotterdam, The Netherlands,
Devries@fsw.eur.nl
2
Department of Public Administration, Erasmus University Rotterdam, The Netherlands,
Bekkers@fsw.eur.nl
3
Department of Public Administration, Erasmus University Rotterdam, The Netherlands & Center
for the study of Law and Society, University of California, Berkeley, The United States,
Tummers@fsw.eur.nl

2
Abstract
This article brings together empirical academic research on public sector innovation. Via a
systematic literature review we investigate 181 articles and books on public sector innovation,
published between 1990 and 2014. These studies are analysed based on the following themes: (1)
the definitions of innovation, (2) innovation types, (3) goals of innovation, (4) antecedents of
innovation and (5) outcomes of innovation. Based upon this analysis, we develop an empirically-
based framework of potentially important antecedents and effects of public sector innovation. We
propose three future research suggestions: (1) more variety in methods: moving from a
qualitative dominance to using other methods, such as surveys, experiments and multi-method
approaches; (2) emphasize theory development and testing as studies are often theory-poor; and
(3) conduct more cross-national and cross-sectoral studies, linking for instance different
governance and state traditions to the development and effects of public sector innovation.
Keywords:
 Systematic review
 Innovation
 Public sector
 Public entrepreneurship

3
1 Introduction
Scholars and practitioners have become increasingly interested in innovation in the public sector
(Osborne and Brown 2011; Walker 2014). Many embrace the idea that innovation can contribute
to improving the quality of public services as well as to enhancing the problem-solving capacity
of governmental organizations in dealing with societal challenges (Damanpour and Schneider
2009). Frequently, public sector innovation is linked to reform movements such as New Public
Management (Pollitt and Bouckaert 2011), electronic government (Bekkers and Homburg 2005),
the change from government to governance (Rhodes 1996) and, most recently, to the discussions
on the retreating role of government in a ‘Big Society’ (Lowndes and Pratchett 2012).
In the private sector, innovation is an established field of study that tries to explain why
and how innovation takes place (Fagerberg et al. 2005). General literature reviews and systematic
reviews have been carried out to assess the state-of-the-art in this field as well as to generate new
avenues for theory building and research (Perks and Roberts 2013). There are even some meta-
analyses, such as that of Damanpour (1991), that pull together the results of empirical research on
the relationships between organizational variables, such as slack resources, and innovation.
However, what is known about innovation in the public sector? What topics have been
addressed in the innovation studies to date, and what are the possible avenues for future research?
Moreover, what can be added to the current methodological state-of-the art when it comes to
public innovation research?
The first contribution of this article is methodological in that we have elected to conduct
a systematic review (Moher et al. 2009). These differ from traditional literature reviews in that
they are replicable and transparent, involving several explicit steps such as using a standardized
way to identify all the likely relevant publications. In public administration, such systematic
reviews have become increasingly popular (e.g. Tummers et al. forthcoming). Nevertheless, a
comprehensive systematic overview of public sector innovation is still lacking.
Second, most of the literature reviews on public innovation that have been conducted in
recent years aim to conceptually, rather than empirically (for example, based on explicit data
such as in case studies and surveys), grasp the meaning and importance of public sector

4
innovation (examples are Osborne and Brown 2011; Sørensen and Torfing 2011). Others address
this challenge through a normative approach (for instance, Bason 2010). This can be seen as a
substantial shortcoming as systematic overviews of empirical evidence are essential to
summarize the existing, evidence-based body of knowledge and to establish a future research
agenda (e.g. Greenhalgh et al. 2004). As such, our investigation is able to identify areas where
substantial progress has been made, and point to areas where future studies could best be
directed.
A third related contribution concerns the antecedents in the innovation process. Given the
predominance of conceptual or normative overviews, the question can be raised as to how much
we currently know about the underling process of public sector innovation as mapped in the
innovation studies. Do we really know the impeding and the stimulating antecedents?
In addressing this topic, we embed our research questions in the open innovation debate
that stresses the content, course and outcome of the innovation process as the result of complex
interactions between intra-organizational antecedents, resources and actors and external,
environmental antecedents, resources and actors. This interaction presupposes rather open
boundaries between an organization and the environmental context in which it operates, and can
be understood in terms of drivers and barriers (Chesbrough 2003). Recently, such approaches can
also be seen in research into public sector innovation (Osborne and Brown 2013, p. 7).
As a result of these porous boundaries, antecedents that need to be further explored in
public innovation research include both the environmental and the organizational contexts in
which innovations take place, their nature, and also the enabling antecedents and their underlying
contingencies. Moreover, there is a need to look deeper into the goals and effects of the
innovation process since, whilst innovation and improvement have often been assumed
synonymous, this is by no means always the case (Osborne and Brown 2013, p. 4; see also
Hartley 2005).
In response to these questions, this article provides a comprehensive overview of how
public innovation has been studied by addressing (1) the definitions of innovation, (2) innovation
types, (3) goals of innovation, (4) antecedents in the innovation process and (5) outcomes. This

5
research design is aligned with other systematic reviews in the social science field such as that of
Greenhalgh et al. (2004).
Based on this, our overall guiding research questions can be phrased as follows:
1. What definitions of public sector innovation are being used?
2. What public sector innovation types can be distinguished?
3. What are the goals of public sector innovation?
4. Which antecedents influence the public sector innovation process?
5. What are the outcomes of the public sector innovation process?
This brings us to the outline of this article. The next section describes the methodology used
to conduct the review. When reporting, we will follow the ‘Preferred Reporting Items for
Systematic Reviews and Meta-Analyses’ (PRISMA) approach (Moher et al. 2009, see
Appendix). Then, Section 3, the ‘Results of systematic review’, presents the characteristics of the
eligible studies found and provides answers to the research questions listed above. Based on these
results, we draw conclusions in Section 4 and develop a future research agenda on innovation in
the public sector in Section 5.
2 Methodology
2.1 Literature search
Four strategies were used to identify eligible studies (Cooper 2010). We selected the period from
January 1990 to March 2014 to include two important publications published in the early 1990s,
namely Hood (1991) and Osborne and Gaebler (1992). These provided strong inputs to the NPM
debate, which in turn stimulated new ways of working in governmental organizations and
resulted in growing attention being given to public sector innovation.
First, we carried out an electronic search in two databases, ISI Web of Knowledge and
Scopus, to ensure we included a broad range of scientific output. We started with the search term