A recent comparative analysis of the content of lobby regulation schemes published in this journal casts measures developed at EU level as the vanguard of a ‘new wave’ of strong lobby regulation across Europe (Holman and Luneburg, 2012, p. 91). Other assessments have characterised EU lobby regulation as weak, primarily based upon the voluntary status of successive registers of lobbyists, as well as the quality of data within them (Chari et al, 2011; ALTER-EU, 2012a, 2012b). We examine these alternative perspectives through a focus upon the EU Transparency Register (TR), commencing with an examination of its features and ways in which the detail of the registration process might affect the disclosure and presentation of public information. We continue by contextualising it within path dependencies and examine the extent of differentiation from predecessor instruments. We then assess the extent of coverage of its core targets through a comparison of the entries in two of its categories (business related and NGOs) with other information sources. We follow this with an assessment of the potential and limitations provided by a substantial public database of 5513 entries in the Register at the start of 2013. We identify the boundary points from which the data can be put to public use, and after removing duplicated entries from the Register, we present the first such data drawn from it. Finally, we place the TR within traditions of comparative lobby regulation. Our analysis comes ahead of a scheduled review of the scheme by the European Commission and European Parliament (EP), originally due by mid 2013.Footnote 1 We do so at a time when most EU member states have established regulatory instruments or currently have them under active consideration (Chari et al, 2011; Holman and Luneburg, 2012; JTRS, 2012a).

EU Lobbying and the Emergence of the Registers

Although most developed political systems have substantial interactions with stakeholders for a variety of purposes (Halpin and Jordan, 2012), the EU is remarkable in a high degree of dependence upon organised interests to achieve its goals. This systemic dependence can be traced back to the lack of popular engagement with the EU, the need for consensus from decision making involving (soon to be) 28 member states, the relative degree of isolation of the European Commission from other potential constituencies of support, and a chronic lack of resources in EU institutions relative to the substantive functions performed. In a quest for a variety of different types of legitimacies, an infrastructure has arisen to formalise exchanges with ‘interested parties’Footnote 2 using devices commonly found elsewhere, centred on pluralist mechanisms of checks and balances and principles of transparency. Organised interests are also expected to have a systemic role of an ‘unofficial opposition’ in an otherwise consensually bound system, supported by a funding regime for (primarily) NGOs, which often accounts for a high proportion of their income. Various interpretations place the role of organised interests in the EU system somewhere between participatory governance and attempts to stimulate a European public sphere (Heidbreder, 2012). A lobby regulation scheme lies embedded within such a context.

The order of establishment for different instruments that regulate lobbying at EU level conform to a pattern often found elsewhere. Measures covering the behaviour of the lobbied usually precede the establishment of instruments aimed at the regulation of lobbyists because the target constituency is easier to define and regulate. EU venues are covered by a mixture of staff rules as well as those for political appointees, mainly involving disclosure requirements and measures to avoid a conflict of interests. The latter are still evolving in a process of incremental, though lumpy, development, often following the interjection of civil society watchdogs, sometimes with the support of a European Ombudsman. At the time of writing, this alliance has resulted in an investigation by the Ombudsman of the handling of complaints by the European Commission to avoid conflicts of interests caused by ‘revolving doors’ of staff crossovers; there is already an 18 month ‘cooling-off’ period for Commissioners, and measures for staff have been under discussion with trade union representatives for some time. The European Court of Auditors (2012) has also recently scrutinised arrangements within some specialised EU regulatory agencies and made a number of recommendations for improvement. Regulatory measures upon staff and politicians contribute important perspectives to analysis of lobby regulation, and we recognise these measures may well change in the lag between our analysis and its publication. Our focus of analysis falls upon the principal regulation component, which regulates lobbyists, the TR, ahead of its scheduled review.

When agendas first emerge for the regulation of lobbyists, a common pattern is for commercial practitioners to establish self-regulatory measures (OECD, 2009b). These arose at EU level in 1994 in the form of a code of conduct, which was to become incorporated within a new Accredited Lobbyist (AL) Scheme operated by the EP dating from 1997. In the same year the Society of European Affairs Practitioners (SEAP) was established, extending self-regulatory provisions to practitioners in (mainly) companies and trade associations. An amalgamated extension of self-regulatory codes was incorporated within a new regulatory instrument established in 2008 by the European Commission, the Register of Interest Representatives (ROIR). One indication of its impact is that the use of self-regulatory complaint mechanisms progressively declined afterwards. In mid 2011, a new joint TR of the European Commission and EP replaced the AL and ROIR Schemes as the main device aimed at the regulation of lobbyists, thus covering much of the principal EU institutional venues of lobbying. The Inter-Institutional Agreement (IIA) establishing the ‘Transparency Register’ identified a full-scale review of the scheme by no later than 2 years following its entry into operation, reflecting a will within the EP that its status as ‘voluntary’ would be re-visited at that point. In deference to this, the responsible European Commissioner (Šefčovič) identified at the time of commencement of the TR that a mandatory scheme was ‘a possibility if necessary’ (Moss, 2011, p. 25), despite a view within the legal services of the institutions that this might require a Treaty change.

The Features of the TR and its Registration Process

The TR involves the incentivised (but nonetheless voluntary) registration of ‘all organisations and self-employed individuals engaged in activities carried out with the objective of directly or indirectly influencing the formulation or implementation of policy and decision-making processes of the EU institutions’ (JTRS, 2012b). The most distinctive feature of the scheme is its wide scope of embrace, with an explicit focus upon indirect forms of lobbying. Accompanying guidelines identify informal networks as falling within its scope, as well as indirect means of exerting influence such as events and conferences. There are a mixture of incentives to join the Register, reviewed in further detail later, but in sum they include: the (in-theory) possibility to exclude non-registered organisations from selective consultation meetings, where there are other consultation opportunities (such as public consultations) in place; instructions to Commission staff to issue invitations to register in meetings; 12 months accreditation for a 1-day access pass to the EP; naming and shaming non-registered organisations; and the option in the Register to sign up to consultation alerts for nominated policy fields. On 5 January 2013, our census date, over 97 per cent of the entries in the Register were from organisations with the remainder comprising self-employed individuals working in a consulting capacity. Entries are spread across six categories and 13 subcategories covering the ‘usual suspects’ and beyond, drawing as far as think tanks, research and academic institutions, as well as some representations of public authorities. A registration can be made in any of the 22 official languages of the EU, which in itself raises transparency-related issues. We looked to see if there was evidence of ‘strategic’ use of languages (that is, of non-native and less popular languages) in the registration process, but analysis of the relatively small number of entries in Cyrillic and Polish,Footnote 3 at least, provided no evidence of this.

All who register enter responses, via a web-based interface, to compulsory information fields about: organisational details, contact information and responsible individuals; spending on lobbying (and for some categories, income received by type of source); ‘goals and remit’; the ‘number of persons engaged in activities falling under the scope of the Transparency Register’; the names of any individuals accredited for access to EP premises; and to confirm abidance with the code of conduct. A potentially important juncture at an early stage in the registration process is where a choice is made by the registrant of the category in which to appear in the Register, because this determines subsequent data entry requirements. A general rule is that consultancies, business-related interests and trade unions (that is, producer interests, comprising Categories I and II of the Register) are required to disclose spending on lobbying (consultancies are required to disclose their clients with turnover-related data), whereas for others (mostlyFootnote 4 non-producer interests in Categories III–VI of the Register, populated mainly by NGOs) it is optional. Conversely, non-producer interests are required to provide information about their territorial membership spread, whereas most producer interests are not.Footnote 5 These differences are reinforced by the presentation of registration categories in the scheme website by use of a ‘black bold line’ visibly separating producer interests from others (de Castro Asarta, 2011). There are options during the process to provide supplementary information in some fields, but a registration can be completed without answering questions about lobbying activities during the preceding year, or about membership of networks. There is an option to receive consultation alerts for nominated fields of interest. On completion of the registration process, an entry is published on the web automatically, that is, without prior review by EU institutions. A search engine permits a limited number of data fields to be examined and cross-referenced, with the selection of enquiry fields seemingly driven mainly by the needs of EU institutions to identify consultees.

The scheme is overseen and administered by an inter-institutional (Commission/Parliament) Joint Transparency Register Secretariat (JTRS, 2012a), in which the equivalent of four full-time staff is deployed at any one time. Its work includes a limited extent of data checks, administering complaints procedures and the imposition of sanctions ranging from withdrawal of access to an EP building pass through to those with reputational consequences. The JTRS does not undertake routine checks of the registration entries because of a European Commission preference to avoid overtones of accreditation, as well as a lack of resources to do so.Footnote 6 The JTRS does undertake random quality checks, as well as ‘targeted’ data fields, which are identified on the TR web pages as subject to particular scrutiny. One field flagged in this way is the category of entry chosen by registrants, because producer-related interests can avoid disclosing financial information by selecting other categories. There are some design faults, which provide opportunities for mis-categorisation. A question that asks ‘in which section do you want to register?’ seems to invite deliberation and choice. Accidents of nomenclature might seem to justify the selection by producer-related organisations of an alternative subcategory ‘non-governmental organisations, platforms and networks and similar’ or another for ‘other mixed entities’. Although the latter category did not seem to contain any such dubious entries, we identified 235 entries of a total of 1553 in the NGO category (15 per cent), which we judged would more accurately be reassigned to other categories, mostly (n=150) to the subcategory ‘trade, business and professional associations’. Among the many examples at the turn of 2013 were Qantas Airways Limited, the Federation of European Private Port Operators, the International Organisation of Aluminium Aerosol Container Manufacturers, the Architects’ Council of Europe, the German Association of Political Consultants, and the Malta Stock Exchange plc. There are also impressionistic consequences, in that the presentation of category head counts is one of the first entry points to the tabular information on the TR web pages.

Another area flagged by the JTRS as a target for checks are entries for lobbying expenditure, which lie at the extremes of ranges. Extremely high outliers have a transient life on the Register, suggesting some monitoring is in place. The first Annual Review of the TR by the JTRS in November 2012 identifies an average of 15 checks undertaken each week, with 40 per cent of these revealing problematic entries. Data quality has been somewhat reliant upon scrutiny of entries undertaken by outside campaigners, and its improvement is one of the goals identified in the Review. One practical means to achieve this involves the progressive introduction of IT tools by the JTRS (2012a, p. 10) aimed at improving the efficiency of its random checks. The European Commission (2007) had at one time considered an external monitoring agency, but rejected the idea on the grounds that to do so would blur the lines of accountability for relations with interest groups.

The Path Dependency of the TR

Although the TR resulted from a merger of registration schemes operated by the European Commission and EP, it most closely resembles the Commission’s preceding ROIR Scheme in concept and in operationalisation, symbolised by the Commission’s leadership of the JTRS. The TR followed each of its predecessors in incentivising registration, but leans more towards the legacy of the ROIR in providing the public with information about lobbying organisations. The Parliament’s AL Scheme had been based upon the registration of individuals in exchange for signing a code and an access pass to EP buildings, and provided no public information from registration returns beyond (from 2003) listing on its website the names of passholders and their employing organisations (de Castro Asarta, 2011). Contrary to the nomenclature of the AL Scheme, the TR web pages follow the ROIR in making explicit that registration should not be used to imply accreditation. Rather, the TR orientation lies with generating useful public information, instead of creating access arrangements to political institutions for elites, helping to place the scheme in a tradition concerned with legitimacy.

The TR lies embedded within a family of procedures with a twin track heritage, the European Commission’s White Paper on Governance (WPG), and the high-level Mandelkern 2001 Report on Better Regulation, each of 2001. The Mandelkern Report had been prompted by criticisms about the technical quality of policy outputs from the European Commission (Radaelli, 2004), while the WPG was notable for its focus upon inputs from civic participation as a pathway towards legitimacy. The result was a set of procedures formalising interaction between the European Commission and civil society in which legitimacy is sought from the process of involvement in shaping public policies, whether by technical improvement of their quality or via the act of involvement itself. In this drive towards ‘participatory governance’ (Heidbreder, 2012), minimum standards for consultation and for the use of expertise were articulated in 2002, and embedded within a regime of impact assessment in the years immediately following. The process of establishing an ‘Access to Documents’ instrument of 2001 had involved articulating an underlying discourse about transparency as a necessary pre-condition for legitimate governance, which was used explicitly in justification for a new ‘European Transparency Initiative’ (ETI) by Commissioner Kallas, first announced in 2005. Article 11 of the 2009 Treaty on EU now establishes and summarises the underlying concepts of participatory governance (but not the instruments themselves) as principles with the status of law. The embeddedness of the TR within a set of wider governance arrangements orientated towards legitimacy meets core criteria for lobby regulation schemes recommended by the Organisation for Economic Co-operation and Development (OECD, 2009a).

The main concern of Commissioner Kallas (2005) had been to establish the ROIR in the first instance, rather than to be overly concerned with points of operational detail with it. Once the ROIR was launched, Kallas was driven by achieving a ‘headline figure’ for the number of registrations, orientated towards a target of ‘2600 interest groups with a permanent office in the capital of Europe’ he had prominently mentioned in his ETI launch speech to justify the need for regulation (p. 5). His team declined to provide guidance as to how to compile disclosure information about lobbying spending and head counts, other than advising organisations to make a ‘good faith estimate’ (Hood, 2009). Organisations were able to skip questions on these factors entirely during the ROIR online registration process. With a ‘light touch’ in administrative oversight, and the equivalent of only one full-time individual (in the main, an intern) providing operational duties on it, large numbers of organisations in the ROIR often entered skimpy or implausible data, or none at all on some key disclosure factors (ALTER-EU, 2009; 2011). The ROIR launched in 2008 therefore had something of the feel of an experiment about it, but left a problematic culture of minimalism in data returns for its successor to seek to combat.

Both the ROIR and the EP AL Scheme had been designated a ‘lowly regulated’ category in a comparative assessment of lobby regulation measures undertaken on behalf of the Council of Europe, published just before the TR came into force (Chari et al, 2011). These authors had assigned schemes to their lowest regulated category, which did not recognise executive branch lobbyists, nor involve the filing of spending reports, where there was no ‘cooling-off’ period on leaving political institutions and where there was little by way of enforcement mechanisms (Chari et al, 2011). Although these failings applied to the EP AL Scheme, the ROIR had elements of each of these regulatory features (or separately, in the case of requirements placed upon Commissioners when leaving office).Footnote 7 The ROIR also involved disclosure of spending in categories and income.

The reputational based enforcement mechanisms (suspension or exclusion from the Register) of the TR continue a tradition from the ROIR. They are particularly significant for commercial affairs consultancies; in 2009 GPlus felt the effect on its client base after being suspended from the ROIR for 4 weeks after including in its data return an explanatory note that three clients had not been identified at their request (EurActiv, 2009). Since the establishment of the ROIR, leading members of ALTER-EU have had an active role in monitoring registers, presenting complaints against some of the most prominent business organisations active at EU level. Where its complaints have been upheld, the reputational damage has contributed towards a ‘trickle down’ effect of a general improvement in the quality of data about organisations, which regularly lobby the EU. Thus, the lobby spending declarations for two subjects of complaints, the European Chemicals Industry Association (CEFIC) and Business Europe, has increased from an implausible €50 000 entry for each in the original ROIR, to €6 million and €4 million, respectively in the current TR. Nonetheless, neither disclosure requirements nor enforcement mechanisms that the TR inherited from the ROIR compare with the extensive arrangements of schemes in North America.

The Data Covered by the TR

A change of name to the TR extended the scope of coverage by making it easier to join for organisations, which objected to being included on a ROIR. The reinforcement of this by the ‘black bold line’ separating producer interests from others was also a new feature of the TR. There is also a change of justificatory discourse. Although the ETI Green Paper of 2006 had articulated the ‘legitimacy of lobbying’ (de Castro Asarta, 2011), the words introducing the TR on its website centre upon the citizen alongside ambitious discourse about ‘avoiding privileged access to information and to decision makers’.

The ultimate test for registers is the extent to which they cover the population regularly lobbying political institutions. The ROIR had attracted more than 4000 registrations before the switch to the TR. The strengthened incentives to register in the TR led its new custodians, Commission Vice-President Šefčovič and EP Vice-President Wallis, to describe it as ‘de-facto mandatory’ (Moss, 2011, p. 25). In dispute of this claim, ALTER-EU released a report entitled ‘Dodgy Data’ in mid 2012 to coincide with the first anniversary of the TR. The report listed 120 companies and 68 lobby consulting firms not on the Register but with a Brussels address identified in either of the two most popular commercially available directories of EU lobby organisationsFootnote 8 (ALTER-EU, 2012a). Of those listed, nine companies and twelve consultancies have since registered. Although the remaining total of 167 absentees represents a relatively small proportion of the 2095 entries in the TR with an office in Brussels, the proportion they represent relative to their categories is much larger. When the 111 absentee companies are added to the population of companies in the TR with a base in Brussels (303), the absentees constitute 27 per cent (111 of 414) of the total. The equivalent figure for consultancies (self-employed or otherwise) is 24 per cent (56 of 233). The missing constituency thus constitutes 25.8 per cent. We then compared the contents of the ‘NGO’ category of the Register with a list of 122 member organisations we identified from the websites of eight ‘families’ of EU NGOsFootnote 9 as constituted at EU level or with an EU policy office in Brussels. This represents a sample size of 35 per cent of all NGOs representing a European interest with an office in Belgium, a size large enough to make generalisations about the population of NGOs representing a European interest. Forty-one per cent of our sample (50) did not have an entry in the TR, which we offer as an indicative proportion of NGO coverage, that is, around 60 per cent. From this analysis, it seems that the label ‘de-facto mandatory’ is not as yet justified. It is possible that some organisations may (incorrectly) consider they had ‘indirect’ registrations through their membership of another organisation on the Register. The budget airline Ryanair made such a claim in respect of its membership of the Association of European Airlines (Crosbie, 2012), just before the establishment of an explicit guideline to the contrary on the TR web site.Footnote 10

The consulting firms on the ALTER-EU list range from those with a single practitioner through to multi-service global law practices. Law practices base a refusal to register on client confidentiality norms, and because the JTRS acts as judge and jury. These arguments are undermined by the presence of five international law firms offering consultancy services on the Register. For the remainder, the advantage of being able to recruit clients who do not wish their business to be disclosed is used as an argument in favour of a mandatory scheme (ALTER-EU, 2012b). However, they are a relatively small segment of the EU lobbying population as a whole. One indicator is that 15 law firms held a EP access pass when the scheme was not conditional upon signing up to the TR (Corporate Europe Observatory, 2009), compared with 1179 organisations on the TR, which had at least one person with access accreditation to the EP at the start of 2013. Commissioner Šefčovič (2012) recently announced his willingness to meet law firms (and consultancies) part way with consideration of an exceptional formula where there are client confidentiality concerns.

The EP pass is one among a number of incentives to join the TR. A potentially strong incentive is that EU institutions reserve the right to restrict invitations to meetings to registered organisations, in that discussions with key stakeholder organisations over the details of specific policy are held in a variety of consultative forums. In practice, the European Commission does not exclude non-registered organisations, but at most prominently invites them to join the TR, whether in bilateral or multilateral meeting settings. Moreover, the web pages of Commission consultative fora vary; those for DG Trade’s Dialogue with Civil Society prominently reinforce the importance of being on the TR, whereas web pages about DG Sanco’s EU Health Policy Forum make no mention of it and its membership list includes a number of non-registered organisations. However, a ‘naming and shaming device’ lists respondents to consultations that are not on the TR separately from others in the follow-up reports available on the European Commission’s consultation portal, Your Voice in Europe.

The expansion of operational guidelines from the ROIR to TR to identify a list of activities and structures covered by, and exempt from, the Register seems to have had the effect of capturing the majority of its intended targets. However, without active administrative intervention there is nothing to deter registrations motivated solely by an opportunity for free publicity. Around two-thirds of entries on the Register (34.9 per cent; n=1917) do not identify ‘European’ among the territorial levels of interests they represent. A number of bizarre, and often amusing, examples can be found in the Register, with little or no apparent connection to EU lobbying. Although their presence seems to some to discredit the Register, they are ‘distracting noise’ when set against the progress made with improving the quality of entries for the main players active on the Brussels scene. NGOs demanding transparency from others have understood the need to set an example, with ALTER-EU circulating a methodological template amongst its 200+ members and beyond. Some companies have used this methodology to calculate their own data entries, while others follow guidelines from Society of European Affairs Practitioners (SEAP) and the European Public Affairs Consultancies Association (EPACA). The result has been a gradual drift upwards in data quality since the start of the TR, with the most complete entries from those campaigning for higher standards, and from those in their line of fire.

Using the TR Data: A Case Study

Although the TR database is freely downloadable, converting it into a usable format remains a task beyond most lay IT users. Having undertaken this with help from specialists, we then sought to see how useable the data might be by working with it at a level of detail.

Our first task was to remove duplicated entries (18Footnote 11). We then re-assigned the 240 mis-placed entries in the NGO category to other categories; although this produces a more accurate picture of the categories, it carries the consequence of including among producer categories a relatively small number of entries, which had provided information to a slightly different structure of questions, and in particular optional (though mostly provided) reporting of spending data. Table 1 details the newly established population of the TR by category and subcategory.

Table 1 TR: Population by category

Thus, just under one-third of the Register are ‘trade, business and professional associations’, slightly under one-quarter are NGOs, around one in seven entries are companies and one in nine are consultancies. Territorially, 85.6 per cent have a headquarters in the EU15 and 91.8 per cent in the EU28. There were 455 registrations with headquarters outside the EU, of which around one-quarter (121) were from the United States, and almost one-fifth (84) from Switzerland.

We assessed the extent to which the data in the Register would allow us to establish whether it is populated mainly by organisations primarily orientated to engaging the EU. We found it could not do so with any precision, although using combinations of responses from 2–3 questions does allow a core of such organisations to be identified. When registering, organisations have the option to select up to four territorial levels of interest they represent: subnational, national, European and global. Almost two-thirds of all entries (65.1 per cent; n=3577) identified that the interests they represent were European, raising some doubt as to the relevance of other entries. There was relatively minor variation between Categories I–V, ranging from three-quarters of professional consultancies (74.5 per cent (306)) to (56.5 per cent (437)) of companies in the territorial level of interest represented. Another question asks for the location of a headquarters, supplemented by a non-compulsory invitation ‘here you can enter more contact details for your organisation that you think useful (for example, contact details of its Brussels office)’. 1360 organisations identified a headquarters in Belgium, of which 1207 were in Brussels and a further 888 organisations outside of Belgium identify a supplementary address in Brussels. This brings the total known to have a Brussels address to 2095. This is clearly the core of the population engaging the Commission and the Parliament, but there are other organisations in the TR that do so and that have no Brussels address. These additional organisations will primarily be from locations within around 3h door-to-door commuting distance of the EU institutions. The best available, though imprecise, indication of the size of this population from TR data is provided by the number of organisations (145) with a headquarters in neighbouring countries, which have at least one person with access accreditation to the EP (France 52; Germany 35; the United Kingdom 34; the Netherlands 20; and Luxembourg 4). Around the time we captured the data at the start of 2013 the population of organisations with at least one pass to the EP (1179) was still growing at a rate of around 30 per week (JTRS, 2012a). Nonetheless, the core of organisations with a significant level of engagement with EU institutions seems more complete with than without them. When we added them to organisations on the TR, which indicate a European interest with an address in Brussels, we then assessed the population distribution across subcategories. We found that the distribution does not differ markedly from the population of the Register as a whole, and may be accounted for by differences in trends among categories in joining the Register. The results suggest that each interest segment seems able to establish itself close to the seat of EU decision making in approximate proportion to their size of the lobbying population covered by the TR.

The number of organisations on the Register that had at least one person with access accreditation to the EP (1179) was lower than might be expected, accounting for just over one-fifth of all TR registrations. The growth in this population of the Register follows a long period of stagnation after a high profile ‘Cash for Amendments’ scandal in the EP in the spring of 2011, described in more detail in the next section. Although we were looking to evaluate the extent to which the EP access pass is an incentive to join the TR, the current population growth rate makes it too early to draw any firm conclusions about this. Eighty-one per cent (953) of organisations with at least one person with EP access accreditation have an address in Brussels, comprising 45 per cent of all organisations with an address in Brussels (Table 3). The important message from Table 3 is that variation between the main categories is not great; 58.1 per cent (176) of companies with a Brussels office, compared with 54.9 per cent (124) of Brussels address consultancies and 47.1 per cent (193) of NGOs with a Brussels address. Organisations with an address in Brussels held an average of 2.5 individuals with accreditation, ranging from 3.03 for consultancies to 2.17 for NGOs. Consultancies hold the first five places for the number of accredited individuals, including one outlier (Burson–Marsteller) with 44; Greenpeace is unusual among NGOs with 13 passes, placing it in eighth position among all organisations (Table 2).

Table 2 European parliament access pass holding organisations with Brussels address by TR category

Spending and lobbying personnel data: Producer organisations and NGOs compared

The calculation of EU lobbying spending (for consultancies, by share of turnover) and head counts of those involved in EU lobbying activities are the most demanding data components for registrants, requiring the application of guidelines with scope for wide differences of interpretation. The JTRS guidelines on these points are more detailed than for any other indicator, though supplemented by further advice from practitioner self-help guides. Most registrants seem to have engaged with guidelines sufficient to enter plausible data, but even the mainstream EU organisations range widely in interpretation when asked to ‘estimate the costs related to the direct representation of interests to EU institutions by your organisation in the current year’, from a minor portion of their total expenditure through to all of it.

Beyond interpretative issues, the TR and its predecessors have suffered a continuing supply of implausible entries at the high extremes of both spending and head count ranges. Some are no more than simple data entry errors; during the time we were writing this article, the entry for the European Women’s Lobby (with a budget of €1 111 853) changed from €7.5m–7.75m to €800–900k. There are a larger number of implausible head count reports. The origin of this may lie in a question, which requires systemic knowledge in seeking the ‘number of persons engaged in activities falling under the scope of the Transparency Register’. There are a number of cases where estimates result in the entry of an entire workforce rather than the number of EU lobbyists. Thus, each of the 104 organisations in the NGO category claim to have more than 50 people engaged in EU lobbying activities; by way of comparison the best staffed EU business association, the CEFIC, is at the high end in a declaration of 72 such staff.

The use of ranges for spending reporting, rather than the need to calculate a precise figure, help to drive up the number of returns from (mostly non-producer) categories for which disclosure is not compulsory. Almost three-quarters (74.2 per cent – 968) of all NGOs provided information (rising to 80.5 per cent – 277 when restricted to NGOs based in Belgium with a European interest) in the ranges data field. A number that entered no information in the data field stated an exact figure in the accompanying comments field. Producer-related interests (Categories I and II) have a choice of spending format (absolute or in ranges), and other categories report in ranges; 82.6 per cent of the former chose to report in ranges, assisting the task of aggregating comparable data. Where we assessed the reporting of absolute figures, we excluded from our analysis a small number of reports of €2 or less from Category II and less than €100 for Category I.

The most striking picture that emerges from the data is that the differences between Categories I (consultancies) and II (business related and trade union interests), and Category III (NGOs), are not as great as might be expected. For head counts, 90.3 per cent (560) of consultancies (Category I) and 86.3 per cent (2433) of Category II report between 1 and 10 people engaged in EU lobbying activities, whereas 75 per cent (978) of NGOs (Category III) report 1–10. When we restricted the comparison of head counts to Belgian based organisations with a European interest the differences were narrower, ranging from 88 per cent (Category II) to 82 per cent (282) (Category III). One quartile of European NGOs with a Brussels address have 1–2 lobbying personnel (25.9 per cent – 89), another quartile between 6 and 15 (27.6 per cent – 95) and around one-third (35.5 per cent – 122) have 3–5.

For spending in ranges (Table 3), 67 per cent of Categories I and II; 67 per cent–1823) and 79.1 per cent (765) of NGOs (Category III) report less than €100 000. When comparing Belgian based organisations with a European interest the differences were similar, with 51.2 per cent Categories I and II (together), and 63.6 per cent of Category III (176) reporting less than €100 000.

Table 3 EU lobbying spending ranges by category

After discounting unreliable looking cases, we assessed spending reports from ranges as well as absolute figures. There were five entries exceeding €4.25m in Category I (Table 4) and 13 in Category II (Table 5). Ten Brussels based EU NGOs organisations reported spending above €800 000, and five over €1m (Table 6).

Table 4 Top 5 lobbying spending EU political consultancies
Table 5 Business-related organisations EU lobbying spending over €4.25 m
Table 6 NGOs with EU lobby spending above €1million

Table 7 presents the largest spending declarations among EU business associations; of note is that four German associations are among the top 20 EU lobbying spenders, among other spenders organised at EU level.

Table 7 Business associations declaring EU lobbying spending above €2 million

As the extent of funding of interest organisations is a remarkable feature of the EU system of interest representation, we examined the constituency of organisations with a Brussels address and a European interest field, which held a grant from one of the EU institutionsFootnote 12 (Table 8). Almost one fifth (19.3 per cent – 405) of this population had received a grant during the 12 months preceding their registration. In all, 41.2 per cent (169) of NGOs and 31 per cent (31) of the small sample of trade unions had received a grant during the preceding 12 months. One NGO with a Brussels office, Food and Water (Watch) Europe, receives a grant of €6.4m, comprising 96.7 per cent of its total budget. EU grants contribute an average of 42.9 per cent of the total budget among this population of NGOs, which provided budget information (330). In addition, 17.5 per cent of companies (53) and 10.7 per cent (84) trade/business/professional associations had received an EU grant in the preceding 12 months, although these mostly relate to participation in the EU’s research and technology development programmes, often alongside university partners.

Table 8 EU grant holders by category and Brussels address

The TR in Comparative Perspective

Six EU member states currently have schemes regulating interest representation in some form, nine have measures under consideration, and in a further two countries measures have been either formally considered and rejected (Italy) or adopted and then abandoned (Hungary, 2006–2011). Schemes commenced in Lithuania in 2001, Poland in 2005,Footnote 13 France in 2009 and in Austria and Slovenia in 2011. The earlier schemes among these have a small number of entries (Kalnins, 2011). Longer standing instruments in Germany are more linked to access arrangements and an era where corporatist practices were at their peak (Pross, 2007; JTRS, 2012a). There have been active recent debates about establishing new lobby regulation instruments in Bulgaria, Croatia, the Czech Republic, Denmark, Estonia, Ireland, Italy Latvia, Romania and the United Kingdom. Schemes elsewhere on the European continent exist in Georgia and Macedonia, and are under consideration in Bosnia, Macedonia, Switzerland, Turkey and the Ukraine (Chari et al, 2011; Holman and Luneburg 2012; JTRS, 2012a).

Debates in European territories have been influenced by the discourse surrounding comparative lobby regulation instruments, in which the TR (and predecessors), as well as longer standing schemes in North America, have featured prominently. Schemes in these latter territories are regarded by a wide range of commentators as those with the highest standards (Pross, 2007; OECD, 2009a Chari et al, 2010 and 2011; Kalnins, 2011; Holman and Luneburg, 2012). These reports share a counsel against the simple transposition of regulatory devices across territories, and particularly across continents, citing arrangements in some European countries designed to institutionalise business associations and trade unions within corporatist governance arrangements (Pross, 2007; Holman and Luneburg, 2012). In consequence, OECD’s undated ‘10 principles for transparency and integrity in lobbying’ counsel an appraisal of lessons learned in other systems and the extent to which these can be applied to their own context. Instruments aimed at the regulation of lobbyists at EU level first took appraisal of these lessons at an early stage.

Despite differences of context, lobby regulation schemes tend to have a core set of linked drivers (Pross, 2007). A starting point often involves a case of corruption, sometimes so high profile that it becomes cited in discourse well beyond the territory where the events occurred. Thus, the ‘Abramoff scandal’ in the United States is widely quoted in justificatory discourse elsewhere; Kallas 2005 explicitly cited the wish to avoid a ‘European Abramoff’ in his speech announcing the ETI. Journalists posing as lobbyists tried this out by offering ‘cash for amendments’ to Members of the EP (MEP) in early spring 2011. There were three takers, one of whom was an Austrian MEP (and former domestic Minister) subsequently jailed for the offence. The scandal was a strong factor in changing the EP access element from a 12 month pass under the AL Scheme, to the current arrangement of 12 months accreditation for a 1 day access pass issued by a special desk in the EP entrance foyer. The new (2012) scheme in Austria that followed the scandal, as well as one in Slovenia that preceded it by a year, each involve mandatory registration of lobbyists, and used as supporting evidence for the claim of ‘Europe’s new wave of strong lobbying regulation’ (Holman and Luneburg, 2012, p. 91).


Does the TR component of EU lobby regulation place it in the vanguard of a ‘new wave’ of strong lobby regulation? If the criterion is the extent to which it places information in the public domain, the answer must be a qualified ‘yes.’ There are now more than 5500 individual entries, of which one-third did not select ‘European’ among the territorial level of interests they represent. We estimate that around three-quarters of business-related organisations active in engaging EU political institutions are in the Register and around 60 per cent of NGOs with a European interest are in the Register. Not quite ‘de-facto mandatory’, but substantial when compared with the rather small number of entries in some of the national registers to have emerged in Europe in recent years. The quality of the data in the TR has progressively improved considering the starting point was its predecessor schemes. Although some data quality problems remain, with a steady supply of quirky (but nonetheless transient) entries, the reliance upon those in the Register to monitor it has driven up standards of data entry among the main lobbying players. Nonetheless, there are faults of design and nomenclature, resulting in 15 per cent of entries in the NGO category being somewhat mis-placed, influencing the extent of public disclosure from this constituency. All of these issues will doubtless be considered in the upcoming review of the scheme, including the key question as to its voluntary nature. For research use, the identification of a ‘European interest’ represented and use of a Brussels address makes the data less prone to quirky outliers. Nonetheless, there are limits in aggregating data on some indicators (particularly head counts of those engaged in EU lobbying activities) because of the extent of outliers. However, some clear pictures emerge from the data, including a key finding that differences in reported resources are less than might be expected between business-related organisations and NGOs. Moreover, we can now know much more about EU lobbying from the available data, including about a key feature, which differentiates it from other political systems, the extent of public funding of civil society organisations and their relative dependence upon it. This helps give rise to other research questions, such as the extent to which the reliance by recipients upon EU funding creates dependencies. The available data informs analysis that can contribute to reviewing measures in place at EU level, and at a time when most EU member states have established regulatory instruments or currently have them under active consideration.