The ISSP 2016 Role of Government Module: Content, Coverage, and History

The Role of Government (ROG) module of the International Social Survey Programme (ISSP) is a unique high-quality data source for comparative research on political attitudes and orientations. This article describes the content, coverage, and history of the ISSP 2016 ROG module, which was fielded in 35 countries. The module has been fielded five times since its inception in 1985, and a majority of the items in the 2016 module are replicated from previous waves to facilitate comparisons over time. In addition, a substantial number of new items are included to cover pertinent issues not previously addressed by the ISSP. Topics include (but are not restricted to) civil liberties; national security and challenges; state intervention in the economy; government taxation, spending, redistribution, and responsibilities; political trust and efficacy; corruption and institutional trust; and government responsiveness. This new wave of the module gauges political opinion at a moment in history characterized by substantial political turmoil and change in many countries. At the same time, this fifth wave strengthens the analytical capacity of the module for charting longitudinal developments both within and across countries. Overall, this makes the ISSP ROG module an attractive platform for asking new questions that can further the mutual development of theory and empirical analysis in comparative research.

analysis principally had to rely on national surveys which differed in question wordings, question order, and survey context.
One truly ambitious example of doing cross-national research in such a scientific environment is Coughlin's (1980) book: Ideology, Public Opinion & Welfare Policy. In this pioneering comparative study of eight rich countries, Coughlin traveled from the United States to Western Europe for eight months of fieldwork, analyzing country-specific data sets with the assistance of national experts. Looking back on the situation for doing research in that era, Coughlin remarks: Judged by today's standards my methodology clearly left much to be desired. However, having observed a large gap in cross-national research on public opinion and social policy, I set out to do the best I could at a time when the technology we take for granted today was non-existent. No personal computers, no [I]nternet, no fast, cheap international communication. Looking back, I am amazed by the sheer naivety with which I undertook the research (Richard Coughlin, e-mail to author, November 29, 2018).
Reading the book now, some 40 years later, it is striking how the opportunities for performing cross-country survey analysis have improved. The ISSP has played a key role in this transformation. As well as constructing modules that are comparable across countries, 1 the ISSP provides a comprehensive infrastructure for data storage with free access for the research community: two key improvements from the times when Coughlin executed his innovate research. 2 This article briefly describes the content, coverage, and history of the fifth wave of the ROG 2016 module. The module has previously been fielded in 1985, 1990, 1996, and 2006. As shown in Table 1, the expansion of ISSP member countries means that the number of countries that have fielded the ROG module has increased over time, although this growth has flattened out in recent years. The data for most countries cover at least two decades and, for a few, impressive time-series covering more than three decades are now available. Appendix A lists all the countries that have ever fielded the ROG module, divided by wave.

TOPICS COVERED IN THE ROG 2016 MODULE
For each ISSP module, a drafting group (DG) is selected by election within the ISSP general assembly (for more information about the organization of the ISSP, see Smith 2009;Br echon 2009). The DG of the 2016 module consisted of seven countries: Sweden (convenor), Great Britain, France, Japan, Lithuania, Mexico, and Spain. The DG's general idea was that the fourth replication of the ROG module in 2016 should ensure that opportunities to compare  Table 2 lists the individual topics that were replicated from previous waves of the module, as well as the new topics introduced in the 2016 module. The same question numbering system is used as in the 2016 ROG source questionnaire (ISSP 2015). For readers interested in the exact wordings of the questions, it is helpful to read the questionnaire alongside this article. Items replicated from previous waves are labeled with "Q" and the new items introduced in the 2016 module are labeled with "N." In the next section, we will briefly describe the DG's rationale for the replicated topics. This is followed by a presentation of the new topics in the 2016 module. Table 3 offers an overview of all 2016 ROG items that were replicated. It starts with the questions on civil liberties, which have a long history going back to 1985. Although as yet little-used, these questions are valid indicators for monitoring changes in this area, which are central to debates about the public legitimacy of liberal democracy (Levitsky and Ziblatt 2018;Mounk 2018;Inglehart 2018;Norris and Inglehart 2019). Six of the seven items from the 2006 module were replicated. The topic of civil liberties is theoretically associated with the topic of security challenges (Q14a-c, described below). In addition, the DG wanted to further strengthen the coverage of this broader issue area by introducing a closely related new topic: national security versus citizens' privacy and rights to information (see the section below on new topics).  The batteries on state intervention in the economy, government spending, and government responsibility were also introduced in 1985. The first of these topics, state intervention in the economy, has a medium usage but was considered a core topic by the DG, and so no items from the 2006 module were dropped from this battery. The items on government spending, on the other hand, represent one of the most used batteries. The DG considered it important that this battery was kept intact, since comparability with previous waves may be affected if individual items were dropped or added. The items cover three dimensions of social spending: law and order, welfare state, and post-materialism. The battery on government responsibilities is the most widely used from the ROG module, and so it was considered essential to keep all items. It was decided that this battery should be complemented by adding an item measuring government responsibility for promoting equality between women and men (N7k; see the section below on new topics).

REPLICATED TOPICS AND ITEMS IN THE ROG 2016 MODULE
The three-item battery on security challenges was introduced in 2006 as a response to the changes in the political climate after 9/11 (Baker 2003;Brooks and Manza 2013;Davis and Silver 2004;Hetherington and Suray 2011). Despite low usage so far, this topic is considered core and may be of considerable interest to researchers monitoring potential social change within this domain. As previously mentioned, it is related to the topic of civil liberties, and the DG reinforced the centrality of this issue by adding a closely related topic on government security versus citizens' rights and privacy (see the section below on new topics).
The topic of political interest, trust, and efficacy includes five items; one was introduced in 1990, and the other four were introduced in 1996. This subject is theoretically related to the topics of government responsiveness and institutional trust (see the section below on new topics). Some of these items may also be useful for research on populism (Norris and Inglehart 2019).
The three items on taxation were introduced in the 1996 ROG module, but had previously appeared in the 1987 and 1992 Social Inequality modules. Taxation is a core topic to cover in relation to the role of government. Answer combinations on these three items correspond to two basic preference dimensions. The first captures attitudes along the progressive/regressive taxation continuum, while the second is related to the overall level of income taxes (Bechert and Edlund 2015;Barnes 2015).
The final topic replicated in the 2016 module was corruption. Beliefs and experiences of corruption constitute an important subject for the ROG module (Rothstein and Varraich 2017). From a general standpoint, it seems that observed country differences in corruption using public opinion survey data correspond very well with measures based on expert judgments (Bechert and Quandt 2010; see also Svallfors 2013). The DG considered that the corruption battery could be trimmed without causing any serious damage to its measurement properties, and so only three of the five items on corruption in the 2006 module were replicated in 2016.

NEW TOPICS AND ITEMS IN THE ROG 2016 MODULE
At the annual meeting in May 2014, the DG proposed several new topics to be included in the 2016 module. The general assembly preferred the DG to focus principally on three of these topics-(1) national security versus citizens' privacy and information rights; (2) institutional trust in the state and market; and (3)

government responsiveness and constraints on
The ISSP 2016 Role of Government Module government actions-that fit well with the overall content and history of the module (see Table 4 for the items included). Apart from these topics, the DG proposed an item on gender equality (N7k) to be part of the existing government responsibilities battery. Although the ISSP Family and Changing Gender Roles module covers the topics of work-life balance and women's participation in society, the DG felt it was important to add an item on whether it is the responsibility of government to promote equality between men and women.
The topic of national security versus citizens' privacy and information rights is theoretically related to the existing topics of civil liberties and security challenges. The general research question refers to the conflict between national security and civil liberties, a question of increasing relevance due to contemporary societal developments (Baker 2003;Brooks and Manza 2013;Davis and Silver 2004;Hetherington and Suray 2011). This topic is a more generalized form of the security challenges topic introduced in the 2006 module, asking whether the authorities have the right to detain people, stop and search people at random, and tap people's telephone conversations in cases when a terrorist act is suspected (Q14a-c). The DG felt that it was important to strengthen these ideas by adding questions covering the rights of government versus citizens' liberties in more general situations (N11-13).
The second new topic, institutional trust in the state and market, broadens the topic of trust in the government by including trust in market actors (N8, N18, 19). The decision to expand the topic was built on two observations on the trust literature. First, previous research places too much emphasis on the state per se and too little on institutional configurations perceived as conceivable alternatives to the state for allocating and administering social security 104 and services. In other words, when analyzing citizens' beliefs regarding institutional capability-the extent to which an institution is trusted to be capable of managing and ultimately providing solutions to specific social problems-it is necessary to move beyond public institutions (Edlund and Lindh 2013). In most countries, the institution outside the family that represents the most consistent counterpart to the state in providing for citizens' welfare is the market. To advance our understanding of the role institutional trust plays in citizens' political preferences, we need to incorporate trust in market institutions within the analytical framework (Edlund and Lindh 2013). On this note, it should be underlined that while some of the batteries focusing on advanced welfare statism in the ROG module are perhaps more relevant for the rich countries in the Organisation for Economic Co-operation and Development, the concepts of institutional trust in the state and the market are clearly relevant for all countries within the ISSP. Second, it is important to update measurements of institutional trust to keep up with ongoing theoretical developments in the field. In the current arguably leading theory on institutional trust, the Quality of Government approach (Rothstein 2011), trust in institutions is anchored in procedural impartiality and efficiency. To put trust in an institution implies a belief that this institution will handle the tasks and responsibilities assigned to it in an impartial and efficient way. In the search for empirical indicators following Quality of Government theory, trust in public institutions is indicated by public perceptions of whether tax authorities work efficiently while providing citizens with services in an impartial way (N18 ;Rothstein 2011;Svallfors 2013). Similarly, having trust in the market and its actors is indicated by public perceptions that major companies comply with law and regulations such as paying their taxes (N19; Uslaner 2010; Edlund and Lindh 2013).
In addition to institutional trust in state and the market, the DG received several suggestions from other ISSP members to widen the topic by bringing in third-sector actors such as the family and nonprofit organizations. In many countries, it is likely that these actors may be an important and realistic alternative to the state and the market (Daly and Lewis 2003). Against this background, the DG designed three items (N8) on this subject focusing on which type of actor is best suited to provide social services, such as health care, education, and elderly care (Lindh 2015;Taylor-Gooby et al. 2019).
The third new topic is government responsiveness and constraints on government actions. In the ideal-typical model of representative democracy, elected representatives should implement policies that correspond to the will of the people (Dahl 1989). This topic approaches government responsiveness from a citizens' perspective in two ways. First, different theories offer competing predictions about which groups and actors have substantial influence over public policy (Gilens 2012;Carnes 2013;Carnes and Lupu 2015;Gilens and Page 2014), and so two new items have been added for examining citizens' beliefs and perceptions about which actors/groups have the strongest influence on government decisions from a crossnational perspective (N9). Second, specific attention is given to public perceptions on the effects of economic globalization on state actions. The last two decades have seen a growing literature analyzing the extent to which increased economic globalization/internationalization affects government behavior (Berger 2000;Scharpf and Schmidt 2000;Crouch 2018). A central issue that might have repercussions for democratic politics concerns the extent to which the forces of economic globalization eradicate the substantive agency and capacity of The ISSP 2016 Role of Government Module national governments to devise their own economic and social policies (Mosley 2003;Crouch 2018;Norris and Inglehart 2019). The DG designed one item that asks respondents about the extent to which government action is constrained by the situation in the world economy (N10).

CONCLUSION
A top priority of the DG of the fifth wave of the ROG module was to ensure that the opportunities to compare with previous modules were as large as possible. Hence, no fewer than 44 items, of which 38 date back to 1996, were replicated from previous waves. The second priority was to review contemporary scholarly debates and identify issues where ISSP data can contribute useful information. In doing so, an important discriminator was that each new topic should be relevant for theories where the explanatory value of national context forms an integral part of the analytical framework.
Returning to the story told in the introduction to this article, we recall that although Coughlin's (1980) cross-national research was clearly theory-driven, relevant data were scarce. Today, the situation seems more the opposite. While access to high-quality crossnational survey data has greatly improved in recent decades, we share Svallfors' (2010) view that carefully designed and theoretically informed comparative analyses still lag behind.
In addition to the fact that quantitative analysis of large-scale survey data may be useful for testing established hypotheses, such data analysis can also be important for theory development (Goldthorpe 2000: chapter 5). Here, the ROG module provides ample opportunities for making progress in comparative research on political attitudes. The fifth wave of the module gauges political opinion at a moment in history characterized by substantial political turmoil and change in many countries. At the same time, this new wave strengthens the analytical capacity of the ISSP ROG module for charting longitudinal developments both within and across countries. Overall, this makes the ISSP ROG module an attractive platform for asking new questions that can further the mutual development of theory and empirical analysis in comparative research.