Introduction
During the New Public Management (NPM) era, the shift from a focus on input to a focus on results has led to an ‘increase in the amount of information and change in the type of information generated and used for budgeting and management purposes’ (Kristensen, Groszyk, and Buhler 2002, 10). It has renewed the attention by both practitioners and academics to the budgetary use of performance information (OECD 2007), which this work refers to as performance-based budgeting (PBB). PBB is designed to develop information about results and use it in both the budgeting process and the allocation of resources.
Despite the relevant body of literature on PBB and the wide variety of research findings, surprisingly there has not been any systematization of the knowledge regarding PBB. Previous academic reviews have broadly investigated the research patterns in accounting (Broadbent and Guthrie 2008; Goddard 2010), focused on the more general stream of literature on performance measurement and management (Modell 2009; Van Helden, Johnsen, and Vakkuri 2008; Van Helden and Reichard 2013), or alternatively, on the evolution of the public budget (Anessi- Pessina et al. 2016). In these works, PBB has played only a marginal role. Robinson and Brumby (2005) conducted a review of the topic, but focusing on the empirical literature on output-based hospital funding systems. Therefore, the claim of the paper is that not enough attention has been given to reviewing and systematizing empirical and theoretical knowledge regarding PBB, and this work is thus an attempt to fill this gap through a systematic review of the literature (Tranfield, Denyer, and Smart 2003).
Systematic review may be defined as ‘a review of the literature according to an explicit, rigorous, and transparent methodology’ (Greenhalgh et al. 2004, 582). Widely used in the medical sciences, it has been applied more recently to management research to counteract the bias of traditional literature reviews and enhance the ‘legitimacy and authority of the resultant evidence’ (Tranfield, Denyer, and Smart 2003, 208). Therefore, a contribution of this work is methodological since a systematic review of the literature in the field of public management/administration and accounting is conducted to guarantee the reliability of the evidence produced (De Vries, Bekkers, and Tummers 2015).
Similar to Broadbent and Guthrie, who reviewed 20 years of accounting research, the current work is designed to answer two questions: ‘What has been done?’ and ‘What could be done?’ (Broadbent and Guthrie 2008). Indeed, the review has a twofold purpose: (a) to understand PBB by systematizing the results achieved until now; and (b) to map relevant previous contributions to highlight overstated and overlooked areas and thus address a possible future agenda. Accordingly, the review is designed to contribute to the current body of knowledge by describing how previous works have addressed the topic under investigation (Descriptive analysis), synthesizing the main themes investigated until now and determining new research questions (Thematic analysis).
To fulfil these purposes, the research process underwent different phases. First, the research protocol was defined and the search process carried out (Review design). This process resulted in a set of papers that were analysed and codified (Descriptive analysis). Then, the content of these papers was subjected to further investigation in an attempt to provide an answer to the first research question, ‘What has been done?´, as discussed in the section with the same name. According to the analysis, relevant issues were suggested for further analysis, thus addressing the second research question, ‘What could be done?’ (Discussion: What could be done?). In the last working phase, the main findings were summarized and conclusions were drawn (Conclusion).
Review design
A systematic review designed to be both methodologically rigorous and theoretically sound was carried out through a well-documented process (Tranfield, Denyer, and Smart 2003). Each of the following subsections illustrates a stage of this process.
Search strategy
The systematic search began with an identification of the keywords. This was a dynamic process aimed at finding search terms suitable for reflecting the variety of labels coined and used by practitioners and scholars. As demonstrated by the scoping study (Tranfield, Denyer, and Smart 2003), performance, output, outcome and results are all concepts at the heart of the current debate on the topic and partly explain the different expressions used (e.g., Diamond 2005). Our search process used these terms and some acronyms (Appendix 1).
We were interested in discovering the different contributions from the fields of public management/administration and accounting, which are the most relevant disciplines in studies of public sector performance measurement and management (Modell 2009; Van Helden, Johnsen, and Vakkuri 2008). Thus, a variety of databases, powered by Google (Google Scholar), EBSCOHost (Business Source Complete), Jstor (Jstor: Arts and Science I and II collection – Business collection) and Thomson Reuters (Web of Science), were selected to frame the search. They were used in combination, or alternatively, by previous literature reviews on adjacent topics (e.g., Boyne 2003; Kroll 2015).
In addition, four search strings guided the selection of papers:
• First parameter (language): To avoid translation problems, only papers written in English were selected.
• Second parameter (time frame): The review included ...