The EU's Footprint in Swiss Policy Change: A Quantitative Assessment of Primary and Secondary Legislation (1999‐2012)

Abstract

In this brief contribution to the debate on Europeanization, we further expand previous work on the proportion of Swiss domestic legislation that is influenced by the EU (Gava and Varone 2012).11 This contribution is an output of the project Agenda Setting in Switzerland funded by the Swiss National Science Foundation (ref.105511-119245/1). In so doing, we focus on questions of considerable normative implications: To what extent are policy changes traceable in legal reforms “infiltrated” by the EU? Is the EU's footprint higher in primary legislation (enacted by Parliament) or in secondary legislation (enacted essentially22 “Ordinances” are the main form of secondary legislation in Switzerland and are generally adopted by the Federal Council. Exceptionally, the Parliament may also enact them (see http://www.parlament.ch/f/wissen/parlamentswoerterbuch/pages/verordnung.aspx). For the period under study, less than 1% of the total ordinances were adopted by Parliament. by the Executive)? Do these linkages result from bilateral agreements (i.e. direct Europeanization) or indirect Europeanization processes? How do policy sectors differ in this regard? The proportion of national legislation influenced by the EU has recurrently been an object of political and scientific debates in EU member-states (Brouard et al. 2012, Töller 2012). This also holds true for Switzerland (see Jenni's contribution to this debate). At the political level, both Europhiles and Eurosceptics generally agree on two points. First, even if they mostly do so without resorting to empirically-based figures, both estimate the proportion of Europeanized policies as being relatively high and increasing over time. Second, as a result of both direct and indirect Europeanization processes, both assume that Switzerland would be adopting EU rules at a high pace, particularly when it comes to commercial regulations and technical standards. Such points of agreement between Europhiles and Eurosceptics quickly fade away as they move onto prognosis. On the one hand, Europhiles point out that since much of the legislative action takes place in Brussels, Switzerland should develop closer institutional ties to the EU (e.g. full EU or EEA accession, à la carte membership). By doing so, the country would participate in decision-making instances and have a voice in shaping the rules that it incorporates all the same. On the other hand, Eurosceptics denounce a loss of national sovereignty and a lack of democratic control over EU policy-making. Europeanization is thus often depicted as taking place “behind the scenes”, through bureaucracies and technocrats rather than elected policy-makers and popular votes. It is not by chance that politicians and scholars argue about the scope of Europeanized legislation. In Switzerland, the question is less about “does the EU matter for domestic policies?” and more about “what is the extent and what are the domains where it exerts more or less influence?”. Despite its limitations (Gava and Varone 2012: 217), the quantitative content analysis of legislation can be a useful tool to investigate such influences. Policies are embedded in written rules that find their way into legislation. As long as references to the EU are manifest in domestic legal texts rather than being latent, they may be traceable.33 For example, we are not able to capture the “copy-pasting” or “paraphrasing” of entire paragraphs or sections from EU rules into Swiss legislation lacking explicit references to the EU or to bilateral agreements (Jenni 2014: 10-11, Linder 2011: 48). As anyone dealing in some way or another with legislation may agree, the entirety of a country's legislative production represents big data for political science standards. With the ever-increasing availability of electronic archives, computer-assisted content analysis allows the assessment of large amounts of legislation and the undertaking of reproducible comparisons across policy domains, time and countries. In the framework of a cross-national project, we previously explored the extent to which manifest references to the EU were identifiable in Swiss legislative acts that are subject to mandatory or optional referendum (Gava and Varone 2012: 213-214). This subset of primary legislation consisted of constitutional amendments, federal laws, federal decrees and the ratification of international treaties. Between 1987 and 2007, 6.4% of the respective legal reforms had an explicit EU footprint. Furthermore the EU's footprint increased across time (from 0.8% in the 1987-91 legislative term up to 13.4% for 2003-07), and varied sharply across policy domains (i.e. 34.8% in immigration versus no references in defense policy). Relying on similar techniques, we here discuss data for the period 1999-2012. This time, we attempt to provide a more complete picture by including secondary legislation (i.e. Verordnungen/ordonnances). “Ordinances” are norms that are hierarchically lower than federal laws, from which they draw their legal basis. In contrast to federal laws, ordinances are protected from the threat of optional referendum. For theoretical and empirical reasons, an extension of analysis to secondary legislation is crucial. On the one hand, Parliament often sets general policy frameworks while leaving enough room for maneuver for implementation rules to be set by the competent executive agencies. The completion of public policies and minor adjustments are thus left to specialized actors and are enacted through a more flexible decision-making process. On the other hand, ignoring secondary legislation would mean that much of the action, in terms of overall Swiss domestic legislative production, would be left aside. In our empirical sample, no less than 85% of the observations correspond to secondary legislation. Our interest in policy change, as one possible outcome of Europeanization processes, leads us to focus on the “flow” of federal legislation.44 Legislative “flows” make reference to legislative activity (creation, abrogation, modification) during a given time-frame. In contrast, the legislative “stock” refers to the total legislation existing at a given point in time (Linder, Schwager, & Comandini, 1985). Legal reforms (i.e. adoptions and amendments to laws, decrees and ordinances) between 1999 and 2012 represent our units of observation.55 We resorted to the internet version of the Official Compilation of Federal Legislation (http://www.admin.ch/bundesrecht/00567/index.html). The full text of the weekly edition of the Official Compilation is available online only from September 1998 onwards. Primary legislation refers in this case to legal reforms in the form of federal laws or federal decrees (arrêtés fédéraux). The term secondary legislation is here used for federal “ordinances”. Since our analysis is limited to the capturing of textual references rather than resorting to more resource-intense qualitative coding, we talk about the presence or absence of an “EU footprint” in the legislative flow. First, we searched a series of EU-related keywords to capture references to the EU in the full text of legal reforms.66 We relied on a computer keyword search based on regular expressions. The list of EU-related keywords comes from the cross-national “Delors' Myth project” (Brouard et al. 2012: 10). Terms include: European Union, European Community, European Economic Community, Common Market, Single Market, European Market, European Coal and Steel Community, European Atomic Energy Community, European Monetary Union and European Monetary System, European Directive, Community law, European law, and its acronyms. As done in the past, we relied on the French version of the texts and included “Schengen” and “European Economic Area” for Switzerland (Gava and Varone 2012: 207). Second, among reforms identified as having an EU footprint, we looked for explicit references to agreements signed between Switzerland and the EU.77 We identified references to international agreements between Switzerland and the EU, but also exchanges of letters and Mixed Committees' decisions. In this way, we proceeded to a (very) rough distinction between direct and indirect Europeanization.88 Note that the “indirect” category includes but is not limited to instances of “autonomous adaptations” to EU rules. This category additionally captures, for example, rules setting differential treatment for EU and non-EU citizens without making an explicit reference to bilateral agreements between Switzerland and the EU. Due to restrained data availability our time series begins in 1999, just before the signature of the first set of bilateral agreements between Switzerland and the EU. This is unfortunate since we have good reasons to believe that Europeanization, in its indirect form, was relatively strong in the 1990s (Fischer 2007, Gava and Varone 2012: 212). Although the lack of a cross-time comparison with the 1990s is highly regrettable, the period 1999-2012 offers an account of policy impacts since Switzerland institutionalized its “third-way” stand towards the EU. Such a stand has consisted in coming closer to the EU legal framework by means of both bilateral agreements and autonomous adaptations. By taking into consideration primary (i.e. new or amended federal laws and decrees) and secondary legislation (i.e. new or amended federal ordinances), Figure 1 provides an overview of the proportion of legal reforms bearing an EU footprint. First, throughout the period, the share of Europeanized legal reforms per year is systematically higher for secondary than for primary legislation. Policy change emanating from secondary legislation is thus Europeanized at a higher pace. For the thirteen years included in the analysis, reforms having a link to the EU are 76% more frequent in secondary legislation than in primary legislation (16.4% and 9.3% respectively). Second, the difference between Europeanized primary and secondary legislation is not stable over time. Although both evolved in a seemingly synchronized fashion during the 2000s, differences in the direction and proportion of Europeanization are higher towards the end of the period. This suggests that assuming Europeanization trends for policy change by exclusively looking at primary legislation may hold water for some periods, but prove misleading for others.99 Several questions remain open in relation to the dynamics between primary and secondary legislation. To what extent do changes in one type of legislation complement, substitute, prepare or result from changes in the other type of legislation? See Jenni's contribution to this debate on similar questions regarding the dynamics between direct and indirect Europeanization. Last but not least, the EU's footprint found in secondary legislation more often takes on an “indirect” form. For primary legislation, manifest references to the EU are equally shared between direct (4.5%) and indirect (4.8%) Europeanization. In contrast, the EU's indirect footprint is roughly four times more important in the case of secondary legislation (13.3% versus 3.1%). This short overview highlights the importance of taking into consideration secondary legislation and indirect Europeanization when studying policy changes from a quantitative stand-point. Policy change linked to the EU takes place at a higher pace in the executive than in the parliamentary arena. At the same time, reforms are almost six times more frequent in secondary than in primary legislation. In a nutshell, a non-negligible proportion of the Europeanization of Swiss policies emanates from the Executive, and takes on an indirect form. In contrast to parliamentary-induced policy change, this is an area that to a large extent remains uncharted territory for most legal and policy scholars.1010 There are notable exceptions, particularly in the work of Wolf Linder and collaborators (Linder et al. 2009, Linder et al. 1985), who have studied both primary and secondary legislation. An additional motivation to investigate this area is the fact that merely looking at the proportions and trends of Europeanization in primary legislation is not sufficient to assess the impact of the EU on secondary legislation. How does this extension to federal ordinances influence our understanding of the differential impact of Europeanization across policy domains? Figure 2 provides an overview of the share of the EU's direct and indirect footprint on primary and secondary legislation across policy areas (N=7'459). As is the case with other contributions to this debate (Sciarini, Tresch and Fischer; Fischer and Sciarini), the widely used topics of the “Comparative Agendas Project” (Baumgartner et al. 2011) are used to classify legal reforms into policy domains.1111 The classification of legal reforms following the Comparative Agendas Project's topics scheme (http://www.comparativeagendas.info) was performed automatically through a conversion from the (sub)sections of the Classified Compilation of Federal Legislation (Recueil systématique de droit fédéral, available online at: http://www.admin.ch/bundesrecht/00566/index.html?lang=en). The dotted lines indicate the overall proportion of “direct” (3.4%) and “indirect” (12%) references. By retaining these average scores it is possible to contrast the traces of Europeanization across policy domains and the relative importance of a direct and an indirect EU footprint among them. The “Immigration” policy domain detaches itself as an outlier with around 40% of its legal reforms carrying a textual reference to the EU. It is worth noting that half of these references are of a direct nature. This is a result of the legal modifications prompted by the bilateral agreements on the free movement of persons and the inclusion of Switzerland in the Schengen and Dublin areas. Besides this outlier position for “Immigration”, that was already observed when we focused exclusively on a subset of primary legislation (Gava and Varone 2012: 213-214), “Foreign Trade” (25.1%) and “Transportation” (21%) unsurprisingly belong to a first group of policy sectors characterized by relatively high shares of EU references of both direct and indirect nature. Given the country's geographical situation, lying at the heart of a traffic intense North-South axis, the development of Swiss transport policy is intimately related to that of the EU. Legislation with an EU footprint in this policy domain can thus be linked to measures regulating the transportation of goods through the Alps, the development of related infrastructure projects, as well as the harmonization with EU standards regarding road, rail and air traffic. The comparable high ratio reported for “Foreign Trade” is also predictable, as the EU is by far Switzerland's largest trading partner. A noteworthy absentee in this first group of highly Europeanized policy domains is “Domestic Commerce” (13%). We would expect a higher EU footprint in this domain due to efforts deployed by Swiss authorities in promoting a regulatory framework compatible with the EU Single Market. This domain is nevertheless close to a second group of policy sectors that have also been considerably influenced through bilateral agreements, such as “Law & Crime” (12.4%) and “Labor” (9.7%). In contrast, “Environment” (25%), “Agriculture” (22.5%), “Health” (14.8%) and “Energy” (14.3%) form a third group of relatively high Europeanized areas influenced essentially by indirect processes. Last but not least, among the “least” Europeanized domains located on the “south-west” of the scatterplot, we find policy sectors that remain quasi immune to EU references such as “Defense”, “Social Affairs” or “Public Lands”. At the same time, we also note unexpected low scores for “Science and Communications” (10%) and “Education” (3%). Intuitively, we expected these areas to get higher scores, particularly when it comes to the indirect EU footprint. For instance, the gradual liberalization of the telecommunications sector has been frequently identified as an emblematic case of the indirect Europeanization of Swiss policies (Fischer et al. 2003, Mach et al. 2003), similarly to what occurs in the energy or transportation sectors. Our empirical study does not pretend to capture all Europeanization processes and outputs. It simply equates Europeanization with policy change, captured here through legal reforms. This neglects the fact that continuity and stability also belong to the policy outcomes of Europeanization. For example, we have not assessed deliberate choices of the Parliament or the Executive not to adopt a legislative act or regulation that would be incompatible with EU norms. In addition, despite the identification of an EU footprint in legal texts, we have no indication of the relative importance of these references to the EU, nor of the substantial scope of induced policy changes. However, if we take into consideration the hierarchy of Swiss legal norms, we could imagine that reforms in secondary legislation generally correspond to minor policy (implementation) changes. In contrast, revisions of primary legislation may hint at intermediate or major policy (formulation) changes. Under such a heroic assumption, our exploration suggests that the Europeanization of Swiss domestic policies is characterized by a considerable proportion of frequent, but predominantly minor policy changes introduced by the Executive. As shown above, reforms with a manifest link to EU policies are far more frequent in secondary than in primary legislation. Moreover, the EU footprint present in secondary legislation takes on the indirect form more often. This observation is non-trivial: Europeanized policy-making processes could follow a somewhat similar pattern as purely domestic policy-making processes, namely a succession of incremental policy changes. If this preliminary observation is confirmed by upcoming studies, it puts into question the general wisdom according to which Switzerland is characterized by a two-speed decision-making process: a slower pace with small innovations for domestic policies versus a faster pace with paradigmatic changes for Europeanized policies (Linder 2013: 4). It goes without saying that only a more in-depth comparison between Europeanized and non-Europeanized policy changes could confirm or invalidate such a hypothesis. Roy Gava is a researcher at the Department of Political Science and International Relations, University of Geneva, Switzerland. He recently finished his PhD dissertation on the regulation of the Swiss banking sector. His research interests include comparative policy analysis, internationalization and financial regulation. Address for correspondence: Department of Political Science and International Relations, University of Geneva, Uni Mail, Bd du Pont d'Arve 40, 1211 Geneva 4, Switzerland. Phone: +41 22 379 83 78; E-mail: roy.gava@unige.ch Frédéric Varone is professor of political science at the University of Geneva, Switzerland. His current research focuses on comparative policy analysis, program evaluation, public sector reforms and interest groups. Address for correspondence: Department of Political Science and International Relations, University of Geneva, Uni Mail, Bd du Pont d'Arve 40, 1211 Geneva 4, Switzerland. Phone: +41 22 379 83 82; E-mail: frederic.varone@unige.ch