How do good governance and democratic quality affect policy performance?

This special issue of European Policy Analysis aims at combining the increasingly comprehensive comparative democracy research with public policy analysis. In comparative democracy research, various datasets have been developed and regularly collected in recent decades to describe and assess institutional features of democratic and non-democratic political systems. Some of these datasets also include at least some variables to cover public policies while others do not (Apaza, 2009; Coppedge et al., 2021; Bertelsmann Stiftung, 2020). At the same time, policy research always seeks to integrate institutionalist factors, in particular in the context of international comparisons (Béland, 2019; Hornung, 2022; Zohlnhöfer et al., 2016). Moreover, the perspective of policy research is increasingly broadening beyond Anglo-Saxon countries as the original scope of application and addresses policy processes and outcomes in a variety of states and political systems which makes the systematic study of the relationship between the characteristics of political systems on the one hand and their policy performance on the other particularly important (Bandelow et al., 2022).

The specific focus of this special issue is on a discussion of the relationship between democratic qualities, good governance (executive capacities and executive accountability), and policy performances (economic, social, environmental, and pandemic policies) in OECD and EU states. Data for the analysis of this relationship is provided by the Sustainable Governance Indicators (SGI) project of the Bertelsmann Foundation since 2009 (Jäckle & Bauschke, 2009; Schraad-Tischler & Seelkopf, 2016; Bertelsmann Stiftung, 2020). SGI can serve as a central data basis because it includes data on all of the above-named pillars. The SGI project provides a European perspective as it includes many European (mostly German) scholars even though it also collects data for non-European democracies. This issue is thus also intended to contribute to one of EPA's central goals, namely the discussion of European perspectives on policy research. In selecting authors for the contributions, a balance had to be struck between relevant knowledge of the data set on the one hand and the challenge of possible biases in assessing the SGI’s strengths and weaknesses on the other. While Bertelsmann Foundation officials were involved in discussions during the planning phase of this special issue to some extent, they had no influence on the composition of the contributions, their content, or the review process. The concept of this issue was developed independently of the Bertelsmann Foundation. There were neither financial nor content-related or other influences. Many of the methods and results presented here lend themselves to applications to other data sets. It is important for us to emphasize this here, as some of the contributions critically examine the theoretical underpinnings, methods, and data of the SGI project.

Nonetheless some authors of this special issue are partly involved in the project as country reviewers, regional coordinators, and board members of SGI. We will disclose this when presenting the papers in the following. We have ensured, among other things, that no bias arises from the involvement in the project for any of the contributions. This involvement in the SGI project also concerns the editors of the special issue. Nils C. Bandelow is regional coordinator for Northwestern Europe and a member of the SGI Board. Johanna Hornung, together with Nils C. Bandelow, was involved in the design of the special survey on COVID-19 in 2020 (Schiller & Hellmann, 2021), but she has no permanent role in the SGI project.

While the SGI's data have unique selling points, they are also inferior to competing surveys in some respects, as the first contribution shows. This first paper is authored by the democracy researchers Croissant and Pelke (2022). Aurel Croissant has been involved in the SGI project since 2009 as regional coordinator for Asia and Oceania and as a board member. He has also participated as regional coordinator and country rapporteur for the same region in the Bertelsmann Transformation Index (BTI) project since 2000, for which he has also served as an advisory board member since 2002. Lars Pelke has no role in the SGI project. The paper compares the theoretical concepts and measurement methods of SGI with the BTI, the Worldwide Governance Indicators (WGI), and, what is currently probably the academically most popular index, Varieties of Democracy (V-Dem). Croissant and Pelke present the three pillars of SGI's standard survey (in contrast to the one-time special survey on COVID). The unique feature of the SGI’s measurement is the pillar on Policy Performance, which includes Economic Policies, Social Policies, and Environmental Policies. A special methodological feature of the SGI’s Policy Performance Index is the combination of expert assessments and standardized external indicators. Using factor analysis, Croissant and Pelke show that the distinction between the three performance areas (economic, social, environmental) is supported by the empirical data. Both theoretically and methodologically, the other two pillars of SGI are less interpretable as measurement of specific latent variables. Additionally, the democracy pillar is less differentiated than the respective data of V-Dem. Finally, the governance pillar has several advantages and disadvantages compared to WGI and its sister project BTI. In terms of coverage, the SGI's regional coverage is rather small with 41 SGI and EU countries, especially compared to WGI with 214 countries. Despite remaining weaknesses, especially in the democracy index, the paper shows a high degree of agreement in the measures of the different methods and is thus suitable for a discussion between political institutions and public policies.

The second contribution to this special issue is by Tosun and Howlett (2022) who are public policy scholars with no involvement in the SGI project. Their contribution starts with a comprehensive historical, theoretical, and methodological overview of research on policy styles (Howlett & Tosun, 2019; Richardson, 1982). The concept is theoretically promising, but so far it faces the challenge of a still relatively thin international comparative data base. The paper then discusses some foundations, strengths, and weaknesses of the SGI project. The challenge of the SGI is almost the opposite to the concept of policy styles: SGI provides comprehensive data, but itself offers only limited starting points for a theoretically justified focus on the use of selected data within scientific comparative policy research. Tosun and Howlett argue that the SGI data on strategic planning and public consultation (which belong to the executive capacity part withing the good governance pillar) can be used for the operationalization of policy styles. To empirically underpin this claim, they discuss the data for these two indicators from 2014 to 2020 and compare scores and developments of scores to respective findings of the policy styles research. A particular strength of the contribution is that it not only uses the pure data of the SGI survey, but also takes into account the content behind it. The SGI's country reports each contain up-to-date justifications by the country experts for the respective scoring, which have hardly been used by international comparative research so far (an exception is another recent publication involving the authors of the paper Tosun et al., 2022).

Bazzan et al. (2022) contribute with the application of an innovative method, fuzzy-set Qualitative Comparative Analysis (fsQCA), to the SGI data to identify diverse configurations leading to sustainable policy performances of the 41 EU and OECD states covered by the SGI project. Giulia Bazzan is a comparative public policy scholar and Priscilla Álamos-Concha is a political methodologist with no involvement in the SGI project. Benoît Rihoux is a comparativist and serves as country expert for Belgium and is associated with the SGI project. The paper highlights the strengths of SGI data to be exploited with qualitative methods (e.g., data transformation to fuzzy-set membership values) and therefore studied with QCA (number of cases, outcomes, and conditions). Concretely, SGI data are conceptualized as set relations; the configurations are associated to a respective outcome, and each case is built up with case-based knowledge. The number of 41 countries as possible cases provides a typical basis for QCA research. The SGI policy performance data cover all sustainability dimensions and therefore may provide a suitable outcome required for the application of this research method. At the same time, SGI provides data for different conditions that can be used by QCA research. Empirically, and based on the data on executive capacity and executive accountability, the analysis shows the central role that executive accountability plays for successful economic and social policy performance. The authors also discuss the importance of high-quality media for successful environmental policy performance. Methodologically, the paper shows the potential of the SGI dataset for qualitative research and how QCA results can inform the data analysis.

Jahn and Suda (2022) also use SGI data to analyze the relationship between specific patterns of democracy and policy performance, but using statistical methods. The authors have a scientific background in comparative public policy and have special insight into the SGI data through Jahn's involvement as a regional coordinator for the Nordic countries and as SGI board member. The contribution uses original SGI data from 2013 to 2019 to operationalize the three dimensions of sustainability and make them visible in a cross-country comparison, both as an overall concept and in their development over time. It continues with theoretical considerations to specify the perspective of patterns of democracy (Lijphart, 2012). First, the authors suggest that comparisons of different types of democracy need not (only) be considered as types of society as a whole, but could focus on concrete government structures. This enables the second theoretical assumption, namely that consensus and majority need not be opposites. Rather, executive efficiency and consensus capacity can be seen as distinct concepts, each of which has its own—ideally positive—influence on policy performance. In its governance pillar, the SGI offers eight indicators respectively for each of the two concepts. Although the original data collection lacks an elaborate theoretical foundation and the data are not systematically theoretically processed, Jahn and Suda apply a factor analysis to show that the data set is well suited for operationalizing this innovative concept. As a first result, the paper identifies four types of governance in countries based on the strength of executive power on the one hand and the strength of consensus on the other. The classification of countries may change over time. The contribution discusses and illustrates this with reference to the examples of the United States, France, Italy, and Germany. Thus, the article provides another starting point for the policy styles perspective discussed by Tosun and Howlett. However, the research interest of the article goes beyond the construction of types but aims to identify relations to sustainable policy performance. A first regression confirms the assumption that both executive efficiency and consensus capacity are strongly significantly and positively related to sustainability policy performance. This result alone is very relevant, as it overcomes the previous perspective of an opposition between government efficiency and consensus capacity and shows the possibility of using both effects simultaneously as good governance. Jahn and Suda even go one step further, calculating the correlations of specific constellations of the two governance dimensions and introducing the political orientation of the government (right vs. left) as an additional factor. The results indicate that especially left-wing governments perform better in terms of sustainable policy when they can use efficient government structures.

Uwe Wagschal, who is a member of the SGI board and was involved in the construction of the special survey on COVID-19 of the project, uses SGI data among several other datasets to analyze determinants of mortality during the Corona pandemic (Wagschal, 2022). The dependent variable is the COVID-19 mortality rate until early December 2021. Many common assumptions surprisingly are not visible in the data: High values in democracy measurements tend to be negatively instead of positively associated with good performance on pandemic policy (at least within the group of established democracies considered here). Similarly, pandemic containment policies, as measured by the Oxford Stringency Index, are not related in the expected way with COVID-19 mortality rates. From a political science perspective, it is also interesting to note that despite the partial politicization of pandemic policies in party competition, no clear effect is found here either: Neither a left-wing nor a right-wing orientation of the respective national government significantly resulted in better policy performance. Regarding contributions of political science and policy analysis, long-term conditions rather than the short-term policies appear to have been successful: a good health care system, a good welfare system, and a high executive capacity appear to have contributed at least somewhat to a low COVID-19 death rate. Like other contributions to this special issue, these findings invite controversy. The results partly contradict widespread assumptions and certainly require further studies with additional data. However, it is an important merit of the contribution to have brought the possible perspectives of political science and policy analysis to this current and highly relevant topic and to have worked here with concrete data.

Overall, this special issue shows that the datasets of comparative democracy, governance, and policy performance measurement are very well suited to formulating and testing hypotheses about multiple contexts. Much of this treasure trove of data has yet to be mined. This concerns, for example, long-term correlations between the effects of current governance forms on time-lagged policy outcomes. The datasets also offer potential with regard to multiple sub-correlations between single institutions and policy outcomes. It should be noted, however, that neither the SGI data nor similar datasets are always fully valid. Nevertheless, it will be important to bridge the gap between indicators for good governance as conditions and policy performance as a phenomenon of interest. This can be done methodologically by theoretical contributions, empirical case studies, QCA, and statistical analysis. Above all, policy research must continue to work on identifying differentiated theoretical relationships between governance types and policy outcomes under concrete conditions. The approaches discussed here, especially the concept of policy styles, offer a first possible point of reference for this. In addition, the existing actor-based approaches of policy research must be used to deepen the discussion between theories and data (a similar suggestion is made by Wenzelburger & Jensen, 2022). Ideally, actor-based public policy research should also be increasingly integrated into the construction of surveys in the future.