Lobbying in the Shadows: A Comparative Analysis of Government Lobbyist Registers

Abstract

Policy Points.

• Our research provides a starting point to benchmark government transparency measures to improve the quality and accessibility of information about lobbying.

• Policymakers and regulators can use our framework to develop or strengthen their own lobbying registers.

• Countries like Canada are a good example of how to design an accessible database about lobbying, and Chile, Ireland, and Scotland provide examples of comprehensive contact logs.

Context

Information about lobbying is crucial to alert the public about undue influence in government decision making. Yet, government disclosures of lobbying activities are rare internationally and vary considerably in their completeness and accessibility. Building on a framework to measure lobbying transparency, this study benchmarked national government disclosures to understand what information was shared and to develop recommendations to strengthen political transparency.

Methods

To identify lobbyist registers, we reviewed four international surveys that assessed lobbying transparency in 128 countries. For each country with an online register, we assessed the 50 indicators in the Framework for Comprehensive and Accessible Lobbying (FOCAL), generating an overall score for each government. To highlight the importance of transparency for public health, we compared the visibility of tobacco industry lobbying across all registers by documenting the information provided about two prominent tobacco companies: Philip Morris International and British American Tobacco.

Findings

We identified 28 countries with online lobbyist registers, all from upper or upper‐middle income countries. No country fulfilled all 50 indicators in the FOCAL. The category of “scope” had the highest scores across countries, whereas the “revolving door” and “financial” categories had the lowest scores. We found evidence of lobbying by Philip Morris International, British American Tobacco, or one of their subsidiaries in 14 of the 28 countries with online lobbyist registers.

Conclusions

Our study empirically demonstrates the hidden nature of lobbying internationally. In the case of industries whose interests conflict with public health, poor lobbying transparency presents a risk that vested interests may undermine public health policymaking without anyone knowing. To ensure that health harming industry interests cannot escape scrutiny, public health advocates should support efforts to strengthen political transparency measures.

Open access publishing facilitated by The University of Melbourne, as part of the Wiley ‐ The University of Melbourne agreement via the Council of Australian University Librarians.

Many of the most egregious examples of commercial conflicts of interest and undue influence in politics come from ad hoc or unconventional sources. It was litigation against major tobacco companies such as Philip Morris International (PMI) and R. J. Reynolds Tobacco Company that led to the trove of documents released during discovery. In these papers, we can see internal documents detailing company strategies to influence governments and block public health policies. Examples include developing weaker voluntary regulation as policy alternatives, cultivating relationships with influential stakeholders to lobby on their behalf, and threatening legal action. ¹ Litigation against pharmaceutical and chemical companies have led to document disclosures revealing similar practices in other industries. ² , ³ , ⁴

The Panama and Paradise papers published by the International Consortium of Investigative Journalists not only revealed a global network of tax minimization (and evasion) but also highlighted the politically well‐connected world of the very wealthy (e.g., high–net worth individuals). Investigative journalists uncovered how Purdue Pharma provided political donations and campaign contributions to US congresspeople to influence regulations regarding prescription drugs such as OxyContin. ⁵ Whistleblowers revealed how the fossil fuel industry lobbied US agencies to cast doubt and delay action on climate change. ⁶

These examples highlight that corporate political influence is extensive and a barrier to public health policymaking. They also highlight that current regulations around political transparency are woefully inadequate. In the absence of strong transparency and disclosure laws for lobbying, political donations, and other forms of political influence, commercial actors can cultivate and hide extensive access to political elites and influence government policymaking—lobbying in the shadows so to speak. Of course, the importance of some commercial sectors to the economy (e.g., banking, telecommunication, energy) predisposes governments to make probusiness decisions even in the absence of lobbying. ⁷ Nonetheless, we focus here on the explicitly political strategies used by commercial actors to influence government decision making as an important manifestation of corporate power and influence that should be interrogated.

We note that lobbying itself is not inherently problematic. Indeed, civic participation in government is crucial to healthy democracies. Citizens, community groups, nongovernmental organizations (NGOs), and other organizations engage with policymakers and make their voices and perspectives heard through lobbying, albeit this lobbying is often labeled “advocacy” to conceptually separate it from more business‐driven lobbying agendas. The problem is that access to governments tends to be imbalanced in favor of powerful interest groups. Political donations, for instance, have been shown to facilitate access to politicians. ⁸ Studies have shown that business interests continually dominate lobbying efforts. ⁹

Better transparency about who is meeting with policymakers, the interests they represent, and the purpose of the meeting can provide crucial evidence about the extent and nature of corporate political influence. Although transparency is not a panacea for undue influence, it is an important first step toward generating awareness about the extent to which some commercial actors may have an outsized influence on government decision making. Transparency can help to level the playing field so that all actors understand who is lobbying governments. For public health actors, better transparency could provide insights into the strategies used by different industries to lobby governments. ¹⁰ With this intelligence, health advocates could be better prepared to counter harmful industry interference in public health policymaking. Better lobbying intelligence can also alert public health actors about potential conflicts of interest—for instance, if health organizations are hiring tobacco lobbyists (a common scenario in the United States). ¹¹ Evidence of undue influence could attract media attention, which could in turn help draw attention to specific advocacy issues and aid advocates to pressure governments to reform their rules around lobbying. Ultimately, transparency is one of the core foundations of accountability and trust in decision making that is essential to a healthy democracy. ¹²

Lobbyist registers are one of the ways that governments share information with the public about lobbying. Registers vary in their scope and sophistication, with some more akin to a directory of lobby firms and others providing linked data sets and detailed records of lobbyist meetings. ¹³ Registers likewise vary in their purpose, with Holman and Luneburg ¹⁴ arguing that some registers were designed to foster transparency and public trust in government, whereas others were designed to provide businesses with better access to policymakers, noting that the latter tended to be weaker. There have been a few attempts to analyze the range of lobbying transparency measures internationally. These have tended to focus more on the regulations and less on the actual registers. ¹⁵ , ¹⁶ One key exception is the Global Data Barometer, which, in 2021, surveyed 109 countries about data governance and data ecosystems, with political integrity as one of its key domains. ¹⁷ They found 19 countries provided lobbying data online.

Our project builds on these initial data sets, expanding the number of countries analyzed and providing a deeper analysis of lobbyist registers by analyzing them directly. We were interested to know the following: What do they disclose? How do they provide this information? What can governments learn from this to strengthen or build their own transparency regimes? These questions complement analyses of lobbying regulations that consider their compliance and enforcement (often lacking).

To measure and benchmark government lobbying disclosures, we analyzed existing frameworks used to benchmark lobbying regulations ¹⁸ , ¹⁹ to develop a Framework for Comprehensive and Accessible Lobbying (FOCAL) comprising 50 indicators of best practice lobbying disclosure. ²⁰ Here, we apply the FOCAL to measure actual government disclosure practices. In addition to benchmarking governments, we also sought to test how easy it was to implement and use the FOCAL. To highlight the public health importance of transparency, we looked at a case study of tobacco industry lobbying. For each country with a lobbyist register, we explored whether it provided information about tobacco industry lobbying and, if so, the nature of those data. This gives us insights into the extent of tobacco industry lobbying in different countries and highlights that some countries may be missing out on valuable intelligence because of weak disclosure regulations.

Methods

This study is the second stage of a project to measure lobbying transparency internationally. Here, we applied a framework (FOCAL) we designed to benchmark government lobbying disclosure practices. We have focused on lobbyist registers, as they are the commonest form of lobbying disclosure. ¹⁷ Lobbyist registers document the lobbyists and organizations involved in lobbying as well as (depending on the jurisdiction's requirements) their activities. Other similar forms of disclosure include open diaries or agendas (such as those in Argentina, the United Kingdom, or some Australian states). ¹⁵ , ²¹ These diaries provide a record of meetings between high‐level government officials and lobbyists (or other individuals and organizations) and provide complementary information about the extent and nature of lobbying. Some aspects of the FOCAL apply to open agendas/diaries, and we discuss opportunities for future research on this topic.

In the first stage of this study, we evaluated the transparency of government lobbyist registers. In the second stage, we compared the visibility of tobacco industry lobbying across each register.

• Stage 1. Benchmarking Government Lobbying Disclosures

Our first step was to identify governments with online lobbyist registers. We used four existing international surveys of lobbying transparency: the Global Data Barometer, ¹⁷ the 2020 and 2023 surveys of lobbying regimes created by the French High Authority for transparency in public life, ²² and the Organisation for Economic Co‐operation and Development report Lobbying in the 21st Century: Transparency, Integrity and Access. ¹⁵ Data for these sources were collected between 2019 and 2023 (Israel's data were collected in 2025, as the website was inaccessible prior to that). Collectively, they assessed 128 countries. From these three data sets, we found only 28 countries with online lobbyist registers (Supplementary File 1 Table 1). For each country, we sourced the policy, legislation, act, code, or other legal instrument that underpinned the register (Supplementary File 1 Table 2). We also collected information about each country's World Bank region and income category. We chose to exclude subnational registers (such as the Catalonia register or the Queensland register in Australia), territories, and international organizations (e.g., the European Union).

We used Google Translate to convert policies and websites to English if no English version was available. In some cases, an entire website could be converted to English, whereas in others, text had to be copied from individual web pages. We note that the language conversions may have some errors, and we flag this as a limitation in our discussion.

In building our template to review lobbyist registers, we reviewed the 50 indicators in the FOCAL, creating a set of notes to assign a value of yes, no, or partly for each of the indicators, weighted two, one, and zero, respectively (Supplementary File 1 Table 3). Drawing on our analysis from the first phase of this project in which we developed the FOCAL, we reviewed the three frameworks with weighted indicators and mapped them against our 50 indicators (see Supplementary File 1 Table 4 for details).

To assess the transparency of each register, we first reviewed each country's lobbying law (or other relevant legislation/act) to answer the scope and timeliness categories. At this stage, we combined two of the timeliness indicators (around updating changes in general versus activities), as the answers were the same across all registers.

To evaluate the other categories, we assessed the website hosting each lobbyist register. Depending on the structure of the register, we reviewed the first three pages (or ten entries) of the website or document. When relevant, we deliberately reviewed different types of entries (e.g., a professional vs. in‐house lobbyist), as different information was provided depending on the nature of the lobbyist. We created an additional indicator for listing the lobbyists employed by a company or lobby firm, as we found this was inconsistently disclosed. We note for some indictors, such as the descriptors, there may be variability in how well all firms or lobbyists comply. Our assessment is based on a sample of each register. Once the benchmarking was complete, we tallied the points for each country (out of 100 possible) and ranked them (Figure 3).

Figure 3.

Benchmarking Lobbying Transparency [Colour figure can be viewed at wileyonlinelibrary.com]

0, red; 1, pink; 2, orange; 3, yellow; 4, light green; 6, dark green.

CSV, comma‐separated value; FTE, full‐time equivalent; NGO, nongovernmental organization; PDF, portable document format.

Throughout our assessment, we documented challenges and anomalies we found in lobbyist registers that made it easier or more challenging to implement the FOCAL and benchmark the transparency of each register. We reflect on these learnings in the discussion.

• Stage 2. Comparing the Visibility of Tobacco Industry Lobbying

In the second stage of this study, we sought to demonstrate the importance of lobbying transparency for public health by comparing the relative visibility of tobacco industry lobbying. For each register, we performed a keyword search of the two largest transnational tobacco companies (PMI and British American Tobacco [BAT]) as well as variations of their names or affiliated and subsidiary companies (e.g., PMI, Philip Morris, Altria, BAT, Benson, Rothman and Hedges, JP Carroll, Coltabaco). We note that there are many other tobacco companies and names that could be searched (including industry associations); however, for feasibility reasons, we limited our search to two prominent and politically active companies. We note that in some countries, they may not be the dominant tobacco companies, and we discuss this and other limitations in our discussion.

For each register, we documented 1) if either company (or a subsidiary) was listed, 2) the information provided in the register about the companies (e.g., names of lobby firms hired, names or number of lobbyists), 3) the number of reports or meetings listed and the names of officials or departments contacted if the information was provided for the registers with contact logs, and 4) finally, we extracted the information provided in contact logs to provide illustrative examples of the details provided in some disclosures for the countries with the most detailed “topic” and “purpose” disclosures (Chile, Ireland, and Scotland). Supplementary File 2 details these searches.

Results

Part 1. Benchmarking the Accessibility of Government Lobbyist Registers

Although a few countries were early adopters of lobbyist registers, the past 10 years have seen a rapid expansion in their use. Some countries have also been proactive in redesigning elements of the registers. Germany, for instance, recently redesigned its register to be interactive and searchable (previously it was a portable document format [PDF] file). Figure 1 shows when countries first implemented a register (the year on the left end of the bars) and the year when they last redesigned their register (the year on the right end of the bar).

Figure 1.

Timeline of Laws Requiring Lobbyist Registers UK, United Kingdom; USA, United States of America. [Colour figure can be viewed at wileyonlinelibrary.com]

Countries with lobbyist registers are heavily concentrated in the European and Central Asian World Bank regions. There were no countries from the Sub‐Saharan Africa or South Asia regions and only one country (Australia) from the East Asia and Pacific region. Most countries with lobbyist registers belonged to the World Bank's group of high‐income economies, with none from low or lower‐middle income economies and only six from upper‐middle income economies (Colombia, Georgia, Mexico, Montenegro, Peru, and Serbia). Nonetheless, we identified only 22 of 83 possible high‐income countries that had publicly accessible lobbyist registers.

Lobbyist registers predominantly took one of two forms. Most common was a directory format (often as a table) that listed the names of lobby firms, organizations, and lobbyists, with varying degrees of detail provided against each entry (such as contact information or topics they lobbied on). A smaller number of countries structured their lobbyist registers like a record of meetings, including Canada, Chile, and Estonia, whereas Scotland and Slovenia provided both formats (albeit Slovenia's record was across two separate websites).

Most countries placed the onus of responsibility for making disclosures on lobbyists (even if the register was administered by a government department). However, in three countries (Chile, Estonia, and Israel), government representatives were required to make disclosures about their engagement with lobbyists. Of these, both Chile and Estonia structured their lobbyist registers more as contact logs.

Some lobbyist registers provided interactive charts to share statistics about lobbying (such as number of registrants or meetings, types of organizations registers, or the topics of meetings). From these, we were able to collect information about the number of registrants (both companies and individuals), lobbyists, and clients (companies or individuals that hired professional lobbyists) from most of the registers (Figure 2). When countries did not provide aggregate counts, in the case of very small registers, we were able to manually count the entries; however, this was not possible for large registers with thousands of entries (Chile, Estonia, and Peru). It was immediately apparent that there was huge variability in the number of registered lobbyists/organizations across countries, ranging from more than 26,000 lobbyists in Germany to just 14 lobbyists in Montenegro.

Figure 2.

Count of Registered Lobbyists and Organizations Chile, Estonia, and Peru are excluded, as it was not possible to collect this information. AUS, Australia; AUT, Austria; BEL, Belgium; CAN, Canada; COL, Colombia; DEU, Germany; FIN, Finland; FRA, France; GBR, United Kingdom; GER, Georgia; GRC, Greece; IRL, Ireland; ISR, Israel; ITA, Italy; LTU, Lithuania; LVA, Latvia; MEX, Mexico; MNE, Montenegro; NLD, the Netherlands; POL, Poland; ROU, Romania; SCT, Scotland; SRB, Serbia; SVN, Slovenia; USA, United States of America. [Colour figure can be viewed at wileyonlinelibrary.com]

Of the 28 countries fully assessed, none fulfilled all 50 indicators in the FOCAL (Figure 3; see also Supplementary File 1 Table 5, which shows the unweighted benchmarking). The category of “scope” had the highest scores across countries, whereas the “revolving door” and “financial” categories had the lowest scores. For some countries, it was not possible to assess all indicators, indicated by en dashes in the table. Below, we discuss each category, highlighting some of the unusual cases we identified.

Scope. Overall, countries performed best in this category, with most having relatively inclusive definitions of lobbyists, though this was less the case for government targets and lobbying activities. Several countries limited registration requirements to consultant lobbyists (Australia, Georgia, Greece, Lithuania, Poland, and the United Kingdom). In other cases, only paid lobbying (i.e., excluding lobbying by volunteers or private citizens) was included (in the case of Canada, Chile, Greece, Poland, Scotland, and the United Kingdom). The United States was the only country to impose specific time and financial qualifiers on lobbyists (requiring registration only if more than 20% of their time was spent on lobbying in a 3‐month period and if expenses exceeded $3,000 USD for individuals and $13,000 for organizations). Some countries excluded particular groups (e.g., Lithuania excluded NGOs and religious groups). Other countries took an inclusive approach; for example, Estonia defined lobbyists as follows ²³ :

A lobbyist (including an advocate) is a natural or legal person who is engaged in lobbying for interests, whether it is private, public or some social group's interests, for a fee or free of charge. For example, they following can be lobbyists: lawyers, consultants, public relations specialists, etc; chambers of commerce and industry; professional and business associations; trade unions; non‐profit and interest associations; other organizations, associations and persons representing interests affecting public life.

Several countries limited disclosure requirements to meetings or contact with specific government representatives. For example, the United States excluded the executive branch, Estonia excluded members of parliament, and Chile excluded meetings with government staff.

Most countries had inclusive definitions of lobbying, with Scotland and France as prominent outliers. Scotland required disclosures only for face‐to‐face meetings (either in person or via a digital medium), whereas France only required disclosures for meetings that were initiated by lobbyists. Most countries defined lobbying as some form of direct contact (whether in person, over email, or at an event) with a government representative or staff member. Ireland, however, also included “grassroots” lobbying, such as media campaigns to influence the general public.

We identified four countries with public ministerial diaries in which government representatives documented their meetings with lobbyists, members of government, or other individuals. In the case of Chile, diaries were embedded as part of the lobbyist register, whereas in Estonia, Romania, and the United Kingdom, they were published separately. For most countries, it was not possible to confirm whether diaries were published publicly.

Timeliness. No country provided real‐time disclosures, although Latvia had the most robust timeliness requirements (2 days) for updating the register. Six countries required updates within 2 weeks (Australia, Chile, Colombia, Georgia, Lithuania, and Poland).

Openness. All the countries that we analyzed had public lobbyist registers. Although most countries had one lobbyist register, some had two or more, such as Estonia (one for each department), Mexico (one each for the senate and chamber of deputies), Poland (one for the upper and lower chambers of parliament as well as for the government), and the United States (the House and Senate each hosted a register, albeit drawing on the same data). Although information was available online (our primary inclusion criteria), the design of lobbyist registers often prevented a straightforward search for a particular client or to see connections across clients or lobby firms (for instance, a list of all registered lobbyists or clients for an organization). In some cases, search functionality was limited or nonexistent. In Peru, for example, the register could only be searched in 1‐month blocks for specific names, making it challenging to assess patterns of lobbying over time or to get a sense of the overall extent of lobbying. We reflect on this further in our analysis of tobacco industry lobbying below. More than half of the registers allowed users to download data, although this often excluded some data, such as historical lobbying. Only Canada, Finland, France, Montenegro, Scotland, and the United States allowed users to download the full register in machine‐readable format (such as a comma‐separated value [CSV] or JavaScript Object Notation [JSON] file) to support further analysis or data linkage (although Ireland, Peru, and the United Kingdom also allowed users to download a limited data set). Three countries (Canada, Slovenia, and the United States) provided unique identifiers (IDs) for lobbyists and lobby firms/clients.

Descriptors. Beyond the full names of the registered organizations and/or lobbyists, little information was provided to identify lobbyists. Poland was unique in that the register provided for its lower chamber of parliament included photos of each lobbyist.

Revolving Door. Only four countries provided information about the revolving door (Australia, Canada, Ireland, and the United States). Information varied across countries and individual disclosures. Ireland provided information about whether a lobbyist was a former designated public official but did not provide dates of when they exited office. The Australian register included information about the last position held and the cessation date. Both Canada and the United States provided information about multiple former government positions (albeit these data were not part of the United States’ downloadable data set).

Relationships. Because of the organization of lobbyist registers, it was often challenging to get a sense of how different individuals or organizations were connected. In Canada, for example, the focus on contact logs means that it is not possible to see a lobby firm's full list of lobbyists or clients (although this could be achieved through linking and analyzing the downloadable data, this step requires a level of data literacy likely beyond that of the public). Only four countries provided aggregate lists of clients (Australia, France, Germany, and the United Kingdom). In five registers, companies disclosed a list of their membership or sponsorship of industry associations, though it was beyond the scope of this study to assess whether these lists were complete (Austria, Finland, France, Germany, and Romania). We found no evidence of the reverse, in which associations that lobby would list the members on whose behalf they lobbied. Canada was the only register to disclose information about board seats held by lobbyists (e.g., in associations or companies).

Financials. Only two countries (France and the United States) provided specific information about the costs of lobbying, and in France, this was limited to total lobbying expenditure. The United States performed better than all other countries in this category, providing a breakdown of income from specific clients as well as the total lobbying expenses for organizations engaging in in‐house lobbying. Canada and Germany noted if companies or lobbying firms had received government funding (although the amounts were undisclosed). Six countries (Australia, Austria, Belgium, Canada, Colombia, and Germany) provided counts of the lobbyists employed, though not their full‐time equivalent (i.e., an indication of their lobbying workload). Chile and the United States provided information about campaign contributions made by lobbyists. In some cases, a country's lobbying legislation required disclosure of financial information, but we could not find it. For instance, Georgia's legislation requires lobbyists to disclose money they received for lobbying, but we could not find evidence of this in the register. We note that from 2026, Finland has committed to providing information about the number of people engaged in lobbying, costs of outsourced lobbying services, and other marketing costs.

Contact Log. Only eight countries provided contact logs with specific meeting dates (Canada, Chile, Estonia, Greece Lithuania, Peru, Scotland, and Slovenia). Chile, Ireland, and Scotland provided the most information about the purpose of the meeting (or other form of contact). In other cases, registers provided summary reports of contact in aggregate form. In the United States, for example, organizations filed quarterly reports listing the departments contacted and the bills lobbied on. Although Colombia did not have a contact log, it had the additional requirement that government representatives must provide information about trips funded by lobbyists including the destination, cost, the organization funding it, and its purpose. We note that Finland's legislation commits to a detailed contact log; however, because of recent implementation, it is too recent to view. More detailed examples of the level of detail provided in each contact log are provided in the following section in which we analyze tobacco industry lobbying.

Part 2. Comparing the Visibility of Tobacco Industry Lobbying

We found evidence of tobacco industry lobbying (determined through the naming of either PMI, BAT, or one of their subsidiaries in a lobbyist register) in 14 of the 28 countries with online lobbyist registers (Table 1). In Australia, PMI and BAT were listed in the archives as previously hiring professional lobby firms but did not have active lobbyists. Likewise, in Ireland, PMI was not actively lobbying (though contact logs show they lobbied between 2016 and 2018). We note that we did not look for tobacco trade associations or other tobacco companies, so the absence of tobacco lobbying in our sample does not mean there was no tobacco lobbying through other groups.

Table 1.

Visibility of Tobacco Industry Lobbying in 28 Countries (Data Circa 2024 Unless Otherwise Indicated)

	PMI			BAT
Country	Lobbyists	Meetings	Expenses	Lobbyists	Meetings	Expenses
United States a	80	–	$3.17M USD	1	–	$40K USD
Germany a	14 (37 b )	–	1.24‐1.25M € (2022)	9 (17 b )	–	790‐800K € (2022)
Lithuania c	14	14 (2021‐2024)	–	0	0	–
Canada a	7	141 (2009‐2024)	–	7	244 (2008‐2024)	–
France a	6	–	400‐500K €	3	–	200‐300K €
Austria c	4	–	N	0	–	N
Belgium a	3	–	–	2	–	–
Italy a	1	–	–	2	–	–
Scotland a	1	20 (2018‐2023)	–	2	7 (2019‐2023)	–
Chile a	0	21 (2018‐2022)	–	3	96 (2015‐2024)	–
Australia a	0	–	–	0	–	–
Ireland a	0	12 (2016‐2018)	–	1	37 (2017‐2024)	–
Slovenia a	–	5 (2023)	–	–	15 (2023)	–
Finland a	–	–	–	–	–	–
Colombia	–	–	–	–	–	–
Estonia	0	0	–	0	0	–
Georgia	–	–	–	–	–	–
Greece	–	0	–	–	0	–
Latvia	–	–	–	–	–	–
Mexico	0	–	–	0	–	–
Montenegro	0	–	–	0	–	–
The Netherlands	–	–	–	–	–	–
Peru	0	0	–	0	0	–
Poland	–	–	–	–	–	–
Romania	0	0	–	0	0	–
Serbia	0	–	–	0	–	–
Israel	0	–	–	0	–	–
United Kingdom	–	–	–	–	–	–

BAT, British American Tobacco; K, thousand; M, million; N, none; PMI, Philip Morris International.

Data that were not available for a category are indicated by an en dash.

^a Countries that have listed PMI and BAT in the lobbyist register.

^b Lobbyists may not all support PMI/BAT, as they were listed against multiple clients.

^c Countries that have listed PMI only in the lobbyist register.

Depending on the country, different information was made available. Twelve of the 14 countries with evidence of tobacco lobbying disclosed how many lobbyists were working for the companies, with the United States listing the most lobbyists (80 for PMI in 2023, including both in‐house and consultant). As noted in the previous section, some countries, such as Australia and the United Kingdom, only required disclosure of consultant lobbyists, meaning that there may also be in‐house tobacco lobbyists that we could not identify. In Austria, France, Germany, and the United States, we could see annual lobbying expenses, with PMI disclosing the highest amount ($3.17 million USD) in the United States in 2023. There was no information about how much money was spent by PMI or BAT in the ten other countries they were registered in. Six countries provided information about meetings, with Canada recording the most meetings (244 for BAT between 2008 and 2024).

No country provided information for all three categories that we considered, and some provided no information for any of the categories (such as Finland). Considering the evidence we found in other countries, this raises questions about the nature of lobbying that takes place in countries without records of meetings or expenses.

Chile, Ireland, and Scotland provided more detailed information about the purpose of meetings (or other forms of communication) as well as the government representative they met with. From this, we were able to get a sense of each company's regulatory priorities and the frequency of its lobbying activities.

Of the 117 meetings held by BAT and PMI in Chile, 86 (74%) focused on the topic of smuggling and the illicit cigarette market. However, beginning in October 2018, PMI began to hold meetings to discuss “innovative” and “reduced‐risk tobacco products.” In contrast, all 20 meetings in Scotland focused on the companies’ “smoke‐free” products and the development of regulations targeting vaping. Lobbying in Ireland likewise focused strongly on vaping. However, it also addressed issues of tax, plain packaging, and smuggling. Table 2 provides illustrative examples of the information disclosed in the three registers.

Table 2.

Examples of Tobacco Industry Lobbying Disclosed in Chile's, Scotland's, and Ireland's Lobbyist Registers

Register	Illustrative Meeting Record
Chile	“Regarding the upcoming visit to the country of a leader of the research area of innovative tobacco products of our company, based in Switzerland, we request the possibility of a meeting so that the expert can present the scientific evidence developed to date, which demonstrates that reduced risk products have the potential to produce a substantially lower risk to the health of smokers, compared to the consumption of conventional cigarettes. This presentation does not seek in any way to exert any type of undue influence regarding possible future policies that the Chilean Ministry of Health could adopt on the matter. It simply seeks to create the space to provide scientific evidence in this regard that can subsequently be evaluated independently and impartially by the authorities and experts of the ministry. The visit would take place during the first days of November.” (BAT, November 2018, legal advisor cabinet minister; undersecretary of public health)
	“Explanation of the general framework of APEC. ‐ Philip Morris will be a sponsor of the CEO Summit APEC Chile 2019. ‐ Explained a new device that seeks to revolutionize the consumption of non‐combusted tobacco and that will be launched in Chile on a date close to the CEO Summit. ‐ This is the IQOS, a technological device that heats tobacco instead of burning it, thereby eliminating combustion and smoke and producing 90% lower levels of chemicals compared to traditional cigarette smoke. ‐ He said that the company's objective is to stop selling cigarettes and dedicate itself to this. ‐ He also expressed his interest in the Women, SMEs and Inclusive Growth priority.” (PMI, January 2019 APEC department head, general directorate of international economic relations [Direcon])
Scotland	“To introduce [BAT] UK to the Member and outline the company's position on UK and Scottish vape product regulations. The Member was briefed on the vape product category and the importance of the category to the Scottish independent retail sector. We asked the Member for his views on the proposals laid out by the Scottish Government to further restrict vape product advertising and promotion. It was discussed with the Member the available evidence and scientific research on vaping. We highlighted to the Member the public health views which suggests vape products are less harmful than smoking tobacco.” (BAT, June 2020)
	“We discussed the Scottish Government's consultation on electronic cigarette advertising and promotion setting out Philip Morris Limited's position on the matter as well as our general approach towards tobacco harm reduction in light of the publication of the new Tobacco Control Strategy next year. In addition we briefed the Member on the transformation of our business away from traditional combustible cigarettes towards the development and sale of smoke‐free alternatives. Finally we highlighted our portfolio of smoke‐free products including heated tobacco and explained the harm reduction potential of such devices including for progressing towards 2034 smoking cessation targets in Scotland.” (PMI, June 2022)
	“The Member was briefed on the vaping product category and a discussion was held on Scotland's ambitious target to create a tobacco‐free generation by 2034 and the role we believe vaping can play in offering smokers a reduced risk alternative. We raised two issues facing the category firstly the rise in non‐compliant and underage appealing vaping products and secondly the recent developments concerning the environmental impact of disposable vaping devices. Our position and policy solutions to both was shared with the Member including affirming our support for measures that require vaping products to be marketed responsibly to adult consumers only.” (BAT, April 2023)
Ireland	“Provide a briefing on [PMI]’s work in the area of potentially Reduced Risk Products ([RRPs]) was provided. [RRPs] is the term we use to refer to products with the potential to reduce individual risk and population harm in comparison to smoking cigarettes.” (PMI, September 2016, one email, one meeting)
	“Present PJC pre‐budget 2017 submission, in which the Company calls on Government to introduce a predictable and sustainable excise plan that minimizes shocks to the legitimate market and stops the illegal trade of tobacco products from growing even further. The plan should be based on objective economic indicators such as average household income, inflation, market decline and price differentials compared to other EU Member States. The submission also explained how holding the current policy of repeated excise hikes is not the solution to achieve Government's objectives in terms of revenues and public health. The submission was also aimed at offering [PJC]’s technical expertise to design the plan, but also to study solutions aimed at better tackling the illicit trade of tobacco products, including the implementation of articles 15 and 16 of the revised Tobacco Products Directive (Track & Trace).” (BAT, January 2017, two to five letters/emails)

APEC, Asia‐Pacific Economic Cooperation; BAT, British American Tobacco; CEO, chief executive officer; EU, European Union; PJC, P. J. Carroll; PMI, Philip Morris International; RRP, reduced‐risk product; SMEs, small and medium enterprises.

Discussion

Of the 28 countries with publicly available, online lobbyist registers, none provided complete transparency. Coupled with our analysis of tobacco industry lobbying by PMI and BAT, this suggests that commercial actors who seek to undermine or block the development of public health policies may be able to exert influence and hide it. This reinforces our core concern and argument: public health advocates must pay more attention to corporate political activity and support efforts to increase political transparency to make it easier to see when, how, and why commercial actors (especially those linked to harmful industries) engage in politics.

Our study builds on previous comparative analyses of lobbying regulation (see, for instance, the work from Chari and colleagues comparing national regulations). ¹⁶ , ²⁴ In contrast to previous studies that sought to evaluate lobbying regulations, we focused on the practical question of disclosure (as opposed to enforcement mechanisms or codes of conduct). Although the overall conclusion of our study is one of limited and fragmented transparency (the sheer number of “0” scores in our framework indicates that important features of transparency are frequently absent), we also found several examples of good practices that countries seeking to reform and improve their lobbying disclosures could emulate. For contact logs, Chile, Ireland, and Scotland provided detailed information including of the topics and purpose of meetings, as well as the specific government representatives that were contacted. Although Canada lacked this level of detail, the design of its lobbyist register made it possible to analyze meetings (and attendees) over time. This is important to get a sense of an industry or company's lobbying strategy over an extended period (such as who they met with, which departments they focused on, or how frequently they sought to contact the government).

The most consistent pattern across all countries we evaluated was the omission of information. It is immensely telling that our highest scoring country (Canada) scored less than 50%. In particular, most lobbyist registers had extremely limited information about relationships between lobbyists (or companies engaged in lobbying) and their government targets—a dynamic that raises concerns for conflicts of interest or even corruption. ²⁵ , ²⁶ , ²⁷ Only four countries provided information about the revolving door, and separate analyses of some of these disclosures suggest that they substantially underrepresent the true extent of the revolving door. ²⁸ , ²⁹ We found no countries that required disclosures about direct business relationships between lobbyists or companies that lobby and public officials or their close family members (such as shareholdings or employment). It may be that these data are available in other data sets (such as gift registers or conflict of interest disclosures) that we did not analyze in this study. On the one hand, the availability of data across multiple data sources highlights the benefits that data linkage can offer in the realm of political transparency, in which different, incomplete data sets can be triangulated to present a more complete picture of the extent and nature of corporate political activity. On the other hand, the separation of the data sets in the first place highlights the fundamental limitations with current political transparency mechanisms.

This leads us to a particular element of disclosure that has received limited attention in the literature to date: how the information is presented and made accessible. The “openness” category of the FOCAL provides a number of indicators including whether the data are machine readable (e.g., in structured format like an Excel table vs. a list of lobbyists in a PDF) and whether the register is searchable (usually through the aid of metadata, which categorizes and tags data [e.g., tagging whether a lobbyist has a revolving door profile or not, classifying the industry of companies, and, for those with contact logs, classifying the government departments or levels in which meetings occur]). Although seemingly quite technical, these elements are crucial for ensuring that the data are easy to analyze and reuse. Although this was a challenge within the 28 countries we evaluated, this issue is of course vastly compounded in low‐income countries, where not only is information about lobbying not available online, but the data that do exist (e.g., about committee hearings or executive meetings for implementing guidelines and regulations) are often disconnected and difficult to search. ³⁰ The issue of disconnected data sets is especially obvious in countries with multiple registers (e.g., Estonia, Mexico, United States). Although, on the one hand, it is beneficial to have transparency over multiple levels and branches of government, on the other hand, the fragmentation of data is an enormous barrier to easily making sense of it.

If we compare our analysis of the Canadian register (with downloadable CSV files that can be linked through unique IDs) to the UK register (with no downloadable data), there was a stark difference in the speed with which we could explore the data and identify patterns (for instance, whether companies are linked via the lobby firms or lobbyists they employ). As a more general reflection, the task of applying the FOCAL to each country's lobbying register was immensely time‐consuming, requiring in some cases searching and comparing multiple pages and lobbying entries to ensure that variations in lobbying data presentation were documented as thoroughly as possible. For countries looking to improve the accessibility of their lobbyist registers, we suggest the following two priorities: first, ensure data are downloadable in machine‐readable format (e.g., CVS), and second, use metadata to make the data searchable and categorized. The Canadian register provides an excellent model for how this can be done.

Comparing the performance of 28 countries helped us to identity outliers. For instance, most countries used an inclusive definition of lobbyist that captured both consultant and in‐house lobbyists, as well as peak bodies, unions, and nonprofits in other cases. However, six countries limited their registers to professional lobbyists only, rendering the remainder of their lobbying transparency effectively opaque. Some jurisdictions (such as Australia) are seeking to reform this deficit in their legislation, arguing that this not only allows many lobbyists to escape scrutiny but also creates loopholes in other aspects of lobbying regulation. ³¹ For instance, rules around cooling‐off periods to prevent conflicts of interest when government representatives move into the private sector are based on definitions of lobbying. In Australia, a limited definition of “lobbyist” means that many former members of government have moved directly into roles that other jurisdictions would define as lobbying without any cooling‐off period imposed. ³² Reforms to expand the “scope” of lobbying legislation would thus have important flow‐on effects for other rules to protect against conflicts of interest or corruption. ³³

Two other prominent outliers in our data were the United States and France for their provisions of financial information. Despite many frameworks recommending that lobbyist registers provide information about how much money was spent (or received) on lobbying, this was extremely rare in practice. ¹⁶ , ¹⁹ We also note that the interest in financial information in frameworks evaluating lobbying regulations reflects a US bias in the frameworks in which many were designed specifically to evaluate US state disclosures (for which financial data are quite common). ¹⁸ , ³⁴ Our study did not seek to understand or explain why some countries provided information whereas others did not—in our section on future research, we suggest how this might be explored.

A final crucial lacuna in lobbyist registers was information about meetings (or other types of communication). Few countries provided contact logs, a major omission as we view contact logs as one of the most imperative features of disclosure. Contact logs allow us to see which members of government (or their staff) are being contacted. They tell us about the volume of lobbying—does a group organize meetings only a few times a year, or is there a dedicated campaign of frequent interactions? We can see which specific individuals are involved in the communication, and coupled with information about the revolving door, we can then see if there are any connections among the participants. Finally, and perhaps most importantly, a contact log can provide details about the purpose of the meeting, as our analysis of 193 tobacco industry meetings in Chile, Scotland, and Ireland showed. Analysis of contact logs can provide insights about the issues companies or industries are concerned about as well as the arguments they make. ³⁵ For instance, we were able to see a stark difference in strategies across the countries, as lobbying in Scotland had a singular focus on “reduced‐risk tobacco products” (e.g., vapes), whereas Chile had a much greater focus on the topic of smuggling. These are common tobacco industry strategies to oppose electronic cigarette and tobacco standardized packaging proposals, which were proposed in Scotland and Chile, respectively. ³⁶ , ³⁷

Some of the null findings in our analysis of tobacco lobbying serve to highlight shortcomings in lobbyist registers. For instance, we did not find any evidence in the Colombian lobbyist register of tobacco industry lobbying. However, researchers have previously identified multiple examples of extensive lobbying by tobacco companies and third‐party allies in Colombia by analyzing documents, news articles, and interviews with health authorities and tobacco control advocates. ³⁸ The gap between what is set out in legislation and what is implemented in practice is illustrated in Figure 2 of this study, which shows that some countries have a suspiciously small number of registered lobbyists, suggesting (as others have documented elsewhere) that lobbyist registers are not always a true reflection of the extant of lobbying. ³⁹ , ⁴⁰

There are several limitations to our study. We looked only at one form of lobbying disclosures (online lobbyist registers), whereas there are other data sets that provide similar or complimentary information about lobbying. Argentina, for example, has open diaries for the presidency but no lobbyist register. Other legal instruments that provide transparency include Sunshine Principles, legislative footprints, conflict of interest policies, public procurement registers, public agendas of policymakers, and political donation registers. ³⁹ Likewise, our analysis is a snapshot in time. Countries are continually updating and reforming their registers (such as Finland, which effectively overhauled its lobbying regulations during the course of this project). In their analysis of political transparency, the Global Data Barometer flagged that some countries had pledged to develop lobbyist registers (e.g., Taiwan and the Philippines), though these had not progressed at the time of our study. ¹⁷ Future iterations of our project are likely to capture a larger set of countries or more robust disclosure mechanisms as countries continue to strengthen regulations.

A second limitation is that lobbyist registers were sometimes in languages other than English, and we relied on Google Translate in most cases (E.C. is fluent in Spanish and was able to review the Chilean lobbyist registers). Although we are confident the translations are robust, there may be nuances that are missed. It was also challenging to evaluate some indicators in the FOCAL because of the variations and discrepancies within lobbying registers. In some cases in which registers had thousands of entries, it was not feasible to assess the whole register, and the random sample we reviewed may not accurately reflect all entries. Likewise, it was sometimes impossible to assess the completeness of client lists, and we may have missed some ambiguous entries, such as those listed only as abbreviations, an issue previously flagged by advocacy group Alliance for Lobbying Transparency and Ethics Regulation in the European Union (ALTER‐EU). ⁴¹ Future studies could review a larger sample of entries or apply data science tools to extract the complete data set for analysis, which could help to answer questions about the level of compliance with lobbying laws.

A third limitation concerns our analysis of tobacco industry lobbying. Our findings of no lobbying in some countries may be a consequence of the political climate at the time (e.g., a nonelection year). Further, PMI and BAT may not be the dominant tobacco companies in some of the countries we analyzed, or they may lobby through their trade associations or third parties instead of directly.

Lastly, an important caveat for transparency in general is that better disclosure practices do not necessarily lead to lower levels of corruption or less corporate political activity. ⁴² The United States is a clear example where relatively strong political transparency mechanisms exist comfortably (or uncomfortably, as the case may be) alongside an environment of extensive corporate political activity and influence. ⁴³ , ⁴⁴ To protect against undue influence, enforcement, sanctions, and a broader culture of integrity are crucial elements that must sit alongside transparency mechanisms. ³³ The revolving door in particular highlights the increasing blurriness between government and business interests and the importance of other measures to protect against conflicts of interest, such as cooling‐off periods or divestment of interests for public office holders.

We would like to conclude by suggesting several avenues for further research arising from this study. Many lobby firms are part of larger corporate affairs companies, in which public‐facing, reputation management strategies are a core part of their repertoire. ⁴⁵ A first avenue would be to explore the less direct forms of lobbying that are captured in the Irish lobbying register. Although “grassroots” or “outside” lobbying (i.e., mobilizing broad public pressure on governments) is often an advocacy strategy for NGOs and nonprofits, it is also an important astroturf campaign strategy (e.g., corporate advocacy masquerading as genuine grassroots). ⁴⁶ , ⁴⁷ To what extent do companies or lobby firms employ these strategies? A better understanding of the range of influence strategies used by corporations can help public health advocates anticipate and challenge undue political influence from harmful industries.

A further question for future research is understanding why lobbying regulation has emerged in some jurisdictions as opposed to others. For instance, scholars have used policy process theories (e.g., Multiple Streams Framework ⁴⁷ and Punctuated Equilibrium Theory ⁴⁸ ) to analyze and explain why Germany reformed and strengthened its lobbying laws. ⁴⁹ Similar case studies could be applied in other countries to deepen our understanding of the circumstances and opportunities that enable robust transparency laws to be passed and enforced (including at the state or city level). Similarly, we could explore whether lobbyist registers are predictive of other transparency measures. Along similar lines, researchers could apply the FOCAL to state governments in the United States, as many have more detailed disclosures than those required by the federal government.

Lobbying data could also be mapped against policy timelines to get a sense for how lobbying plays out against different policy or regulatory debates. What political influence strategies are used (e.g., are particular members of government lobbied, or do some commercial actors play a prominent role)? Our comparison of the different arguments made in Chile, Scotland, and Ireland suggests that there were strategic reasons for those tactics that could help us to understand (and better anticipate) future lobbying activities. Previous studies of the tobacco industry have found a common strategy is to lobby key leaders in government, such as presidents of committees, as they are gatekeepers to not allow legislative bills to be further discussed or leave committees. ⁵⁰ This is also when more robust data sets such as Canada's (with unique IDs facilitating data linkages) could be used to compare patterns of lobbying over time and across industries or perhaps link lobbying data with other political activities (such as campaign finance).

For us, one of the next steps is to examine the practical utility of this framework with policymakers and bureaucrats responsible for managing the registers to understand the opportunities and barriers for implementing the level of detail and specifics set out in the FOCAL. Do the weightings we assigned to categories resonate with policymakers? Do they agree with our priority areas? And perhaps most importantly, what do they see as politically feasible to implement, and how might that be achieved?

Conclusions

When lobbying occurs in the shadows, it prevents scrutiny, potentially giving rise to conflicts of interest, undue influence, or even corruption. Transparency, although not a panacea for these issues, is an important tool to minimize the risk of undue influence. Our analysis of the strengths and weaknesses of existing national lobbyist registers provides a starting point to benchmark government transparency measures. Likewise, our snapshot of the visibility (or invisibility, as the case may be) of tobacco industry lobbying highlights the urgency of strengthening existing lobbying regulations. Beyond the lobbying registers analyzed in this study, we would like to point readers to other avenues to increase the visibility of political influence, such as investigative journalism, freedom of information requests, whistle blower protections, and litigation (in which the discovery process can reveal lobbying practices). ¹⁹ We hope that our work provides a useful tool for policymakers and advocates seeking to improve lobbying transparency, as we strongly believe that political transparency is a fundamental prerequisite for successful public health policymaking.

Funding/Support

This work was supported by an Early Career Researcher grant from the University of Melbourne. The funder had no role in the design or conduct of this project.

Conflict of Interest Disclosures

J.L.‐N. is the recipient of a fellowship from the Victorian Health Promotion Foundation. J. L.‐N.’s partner works for a technology company whose clients are pharmacies and pharmacy suppliers.

Supporting information

Supporting Information

Supporting Information