Who generated the loopholes? A case study of corporate tax advisors’ regulatory capture over anti-tax avoidance legislation in Finland


 Recently welfare economists and international political economy scholars have increasingly discussed how the corporations seek profits by corroding policies that tackle tax avoidance and undermine public interest. This article contributes to these discussions on so-called regulatory captures in the global wealth chains by providing a comprehensive case study on anti-tax avoidance legislative processes in Finland. The author analyzes the statements that various stakeholders provided during several phases of enacting the interest deduction limitation rule that targets so-called thin capitalization arrangements. Because of this specific research material, the author is able to undertake a nuanced analysis in describing how and whose statements made the difference from the draft version of legislation to the final wording of the law. The evidence suggests corporate interest groups and tax advisory firms influenced the content of the rule as notable tax base eroding loopholes have been included in the Finnish anti-tax avoidance rule. The author also assesses remedies of regulatory captures.


Introduction
The last decade has seen a change in the global agenda to tackle tax avoidance 2 by multinational corporations (MNCs). The international community and academics have broadly shared the view that tax avoidance and related harmful tax competition not only undermine national budgets and create inequality, but also harm the global economy and welfare (OECD 2013; IMF 2019). Consequently, anti-tax avoidance measures that protect national tax bases from cross-border profit shifting have been a key part of corporate tax reforms globally. The Organisation for Economic Co-operation and Development (OECD) has been at the forefront of this shift by coordinating the development in its Base Erosion and Profit Shifting (BEPS) project initiated in 2013 by the G20 (OECD 2021). Nevertheless, this turn has not been without opposition, which has generated growing academic discussion on how tax professionals are able to corrode policies that tackle tax avoidance (see section 2.2). However, much of the evidence on their influence has been anecdotal, not least because of the difficulties in accessing relevant research material. In this article, I contribute to these discussions over regulatory capture (Stigler 1971;Carpenter and Moss 2014) by providing a comprehensive case study on the legislative processes in Finland. According to the theory, regulatory capture occurs when corporate lobbying extract rents or minimizes regulatory costs in a way that it harms the public interest 3 . Based on the evidence, I argue that corporations and tax advisory firms not only benefit from tax base loopholes 4 in law, but also actively generate and aim to preserve them.
adviser to the minister of local government in the Finnish Ministry of Finance. 2 I discuss the concepts in section 3. 3 I discuss the concept and previous literature more thoroughly in section 2. 4 In this article, a loophole stands for an ambiguity or inadequacy in the law facilitating base erosion and profit shifting that abuses public In the case study, I analyzed the statements provided by several stakeholders during several phases of enacting the interest deduction limitation rule 5 (IDLR) in Finland. These rules target specifically so-called thin capitalization arrangements that are perhaps the best-known form of MNCs' profit shifting (OECD 2017b;Ruf and Schindler 2015). Written statements that were provided to the public consultations during the legislative processes are publicly available in Finland. The first statements date back to 2009 when the Ministry of Finance had published its initial report on the various options for introducing IDLRs. The second set of statements are from the two consultation rounds in 2012 before adopting the initial IDLR. I also analyzed the statements submitted before amending the IDLR after adopting the EU Anti-Tax Avoidance Directive (ATAD) in 2016. The latest statements included in my research are from the parliamentary consultation held in 2018 during the ATAD implementation in Finland.
Drawing from the case study, I analyzed how different types of stakeholders were able to affect the final content of the IDLR that aimed to tackle the wide-spread tax avoidance through intra-group finance structures. These stakeholders include corporate interest groups, law firms as well as audit and accounting firms, civil society organizations, academics and authorities.
The previous empirical research on regulatory capture has relied much on statistical correlation between political connections of firm executives and board members and the economic performance of the firms (Amore and Bennedsen 2013;Cingano and Pinotti 2013;Coviello and Gagliarducci, 2017;Deniz and Mishra 2014;Faccio, Mazulis, and McConnell 2006;Goldman, Rocholl, and So 2013;Khwaja and Mian 2005). Some researchers have also assessed different mechanisms of regulatory capture or lobbying and analyzed the roles of different stakeholder groups and their objectives (Botzem 2008;Hertel-Fernandez, Mildenberger and Stokes 2018;Korkea-aho and Leino 2017;Kramnaimuang King, and Hayes 2017;Vesa, Kantola, and Binderkrantz 2018;Young 2012).
However, these previous studies failed to show the causal connection between a specific lobbying event and its consequence. My case study method has been drawn from Christensen's (2020a) study with similar systematic qualitative content analysis approach (Drisko and Maschi 2015). The specific research setting allows me to find eviinterest. The OECD (2021a) has also used the concept gaps and mismatches in similar meaning. 5 Rules of this kind are also referred to as thin capitalization rules (OECD 2017b). dence of these causalities and thus answer the call from Carpenter and Moss (2014,): "Perhaps the deepest problem with much of the research on regulatory capture is not merely its tendency to overstate the evidence for capture, but its lack of nuance in describing how and to what degree capture works in particular settings." My specific research material allows me to undertake nuanced analysis in which I can describe whose statements made the difference from the draft version of legislation to the final wording of the law. The specific context makes this analysis more relevant since Finland is known for transparency and high trust in public institutions (OECD 2021b). Open public consultations play a significant role in the Finnish legislative drafting process as they represent the various political and expert views at the issue in question (Vesa, Kantola, and Binderkrantz 2018). Thus, they have real impact on the final legislation (Interview 2020). This approach allows me to contribute to previous research by identifying systematically the causal connection between lobbying and its consequence instead of macro-level correlation or merely anecdotal evidence.
The legislative amendments based on the consultations suggest corporate interest groups and firms offering tax advisory services to corporations influenced the content of the rules (see section 5). I also found evidence that notable loopholes have been included in the Finnish IDLR because of this influence. For example, the equity escape rule (EER) of the IDLR has been considered adverse since it has facilitated tax avoidance that has caused tax losses of at least tens of millions of euros each year (see sections 3 and 5; Ministry of Finance 2018). This understanding of how corporations and their advisors are able to capture legal processes also provides a way to amend the structures that undermine public interest they should serve.
The influence of corporate lobbyist groups has been emphasized by the lack of stakeholders representing the public interest and other interests in the context of the tax legislation processes (see section 5). Consequently, most of the stakeholders involved in the legislation processes have an interest to block more effective anti-tax avoidance measures that could increase tax burden of their beneficiaries or clients. Even many legal scholars in Finland offer legal advice to MNCs and some are employed as tax consultants or advocates, which might have an impact on their statements regarding their clients' interest (Raitasuo 2019).
This imbalance of interest could have been emphasized by the epistemic communities of tax specialists, in which views are shared and certain policy viewpoints such as social fairness seem to be sidelined (Forstater and Christensen 2017;Christensen 2020b;Shapiro 2012). The tax professionals' communities globally are similar to those in Finland and the professionals are also involved in influencing tax policies in other countries as well as within international organizations such as the EU and the OECD (Christensen 2020a;Sikka 2008). Therefore, my findings will be helpful in analyzing the influence of different stakeholders on tax reforms in other countries and within international organizations.
The remainder of this article progresses as follows. In the next section, I review the key literature on regulatory capture and lobbying. In section 3, I briefly describe the mechanisms of tax avoidance as well as how and why it is tackled. Section 4 introduces my research methodology and section 5 is devoted to the case study. Finally, in section 6, I conclude by discussing my findings in the context of previous research and assessing remedies of regulatory captures.

The evolution of regulatory capture theory
According to the public interest theories of regulation, governments should regulate in order to maximize the social welfare of a society as a whole (Boehm 2007). In this sense, regulation could be considered economic policy intervention that should aim to correct market failures that threaten social welfare (Pigou 1932). However, as the corporations strive to maximize short-term profits, their interest is often in conflict with the welfare of the whole society. The conflict is evident in the given context of corporate taxation, because income tax means direct regulatory cost for a corporation, but the corporate benefits of taxation emerge indirectly and within a longer time frame. For this reason, the corporations have an incentive to influence regulation to help themselves. This might result in government failures if a regulatory agency such as a ministry or parliament advances the interests of some corporations over the public interest leading to a net welfare loss for the society. This phenomenon described as the concept of regulatory capture but also often called policy capture, was introduced by Nobel Memorial Prize in Economic Sciences laureate George J. Stigler (1971) in an article entitled The Theory of Economic Regulation. Regulatory capture occurs when all these three criteria are met within a policy process (Carpenter and Moss 2014, OECD 2017a): 1. There is an explicitly defined policy goal that either helps or harms the public interest.
2. The corporations consciously intervene in the policy process to extract rents or minimize their regulatory costs. 3. The public interest is abused due to the intervention.
Regulatory capture is often driven by corporate lobbying, but these two concepts must be distinguished as regulatory capture is not necessarily a goal or a consequence of lobbying, because not all influence on policy harms the public interest (Boehm 2007). This harm derives from a process through which regulatory capture not only leads to a misallocation of resources, but also creates economic and social inequality that reduces trust in democratic society and thus deprives it of its legitimacy (OECD 2017a, see section 3).
Stigler focused specifically on the corporations' role in policy processes, but later theoretical and empirical contributions have presented a more nuanced picture by also discussing remedies of regulatory capture (Shapiro 2012) as well as the influence of other interest groups and civil society organizations that lobby things such as human rights (Becker 1983;Peltzman;1976;Posner 1975;Forstater and Christensen 2017). However, such organizations do not target regulatory capture as they do not extract rents for their beneficiaries and usually support public interest based on their bylaws.
The decomposition between strong capture and weak capture introduced by Carpenter and Moss (2014) clarifies the different circumstances in which regulatory capture occurs. A strong capture means the corporations have been able to set the agenda, i.e., the initiative to enact or amend legislation comes from them. The case study in section 5 describes the context of a weak or corrosive capture, when other groups such as civil society organizations, individual whistleblowers and the media, have set the regulatory agenda of anti-tax avoidance. In this case, the regulatory capture is restricted merely to reducing or limiting regulative costs for the industry.

Previous evidence on regulatory capture and lobbying
Several empirical approaches have been used to study regulatory captures. A stream of studies has exploited econometric methods in the search for evidence on lobbying.
Several studies have pointed out the statistical correlation between the political connections of firm executives and board members and the economic performance of the firms (Amore and Bennedsen 2013;Cingano and Pinotti 2013;Coviello and Gagliarducci, 2010;Deniz and Mishra 2014;Faccio, Mazulis, and McConnell 2006;Goldman, Rocholl, and So 2013;Khwaja and Mian 2005). However, these statistical methods fail to address the clear causality between actual lobbying and its results since they do not identify any lobbying event or its direct outcome. The outcome of these studies also lacks generalizability since each is tied to their specific empirical context. Hertel-Fernandez, Mildenberger, and Stokes (2018) have used observational analysis and surveys to find evidence of a mechanism by which corporate interest groups are able to influence US Members of Congress through their staffers. Botzem (2008) draws from an analysis of developments in the organizational body that sets accountancy standards and argues that private sector accountants play the main role in rule-setting. However, Young (2012) was not able to find straight-forward evidence on how the private sector was able to influence Basel II rules on banking in his study based on archive material and interviews. These studies considered the institutional setting of regulatory capture, but do not identify the capture itself as I do in my case study.
The research on Finnish lobbying has focused on the role of interest groups on legislative processes at a general level as well as the election funding of political parties. Vesa, Kantola, and Binderkrantz (2018) have used surveys of interest groups and civil servants to study the involvement of interest groups in policy-making. They found corporate interest groups and trade unions have important role in Finnish policy making and use their influence in working groups and public administration consultation processes rather than in parliament. Blom (2018) arrives at similar results through interviews of policymakers from the perspective of political history. However, a recent report (Hirvola et al. 2021) based on various methods indicates the corporate interest groups and institutional organizations have emerged as more significant lobbyists compared to the trade unions. This study also highlighted the importance of early influence in legislative drafting processes as well as different channels of lobbying where also informal contacts play important role. Kuusela and Kantola (2019) show how the ultra-wealthy have considerable control over the Finnish corporate interest groups. These contributions provide evidence of the influence of certain groups at a general level, but the approach does not systematically focus on results or the means of lobbying in individual cases.
Regulatory capture is not limited to legislation processes, but public agencies' functions could also be captured when they apply laws and treaties. Korkea-aho and Leino (2017) analyzed the decisions of the EU courts and the European Ombudsman and found evidence show-ing that third parties have an influence on EU agencies' decisions on the public disclosure of documents. Kramnaimuang King and Hayes (2017) arrived at similar results on the corporate influence on risk regulation based on analysis of relationships between regulators and the Australian pipeline industry.
Regulatory captures specifically in the context of corporate taxation and tax avoidance have been discussed within the discipline of international political economy. These studies have identified the technicality of tax legislation as a key factor (Surrey 1957;Peters 1991;Seabrooke and Wigan 2016). The current international corporate tax regime was created in the 1920s in the League of Nations and its deficiencies undermining welfare were highlighted in academic research already in the 1960s when crossborder trade started to grow exponentially (OECD 2013;Ylönen 2016). However, this critical discussion decelerated with very little impact by the end of 1970s and the regime remained more or less the same until the beginning of the 2010s. The highly technical nature of international tax legislation, and the relatively closed communities of officials, practitioners and researchers that mastered the laws and the practices, contributed to protecting the regime from political interventions (Forstater and Christensen 2017). Finally, the end of the first decade of 2000s saw an increase in public interest about tax avoidance caused by several data leak scandals that coincided with the financial crisis of 2007-2008. Consequently, many countries have adopted anti-tax avoidance rules in their recent tax reforms to answer the public call and patch national budgets. This development has been led by the OECD (2021) BEPS project, initiated in 2013. However, the corporate lawyers' technical expertise and superior corporate lobbying resources have played their part in corroding the resulting OECD recommendations and subsequent national anti-tax avoidance legislation (Christensen 2020a).
Earlier research on the global wealth chains has also highlighted the specific role of tax advisors and accountants in facilitating profit shifting Wigan 2016, 2017;Baden and Wigan 2017;Finér and Ylönen 2017;Finnwatch 2017). The global wealth chain theory is also useful in analyzing the role of advisors and accountants who serve multinational corporations, influencing legislation in their clients' favor (Christensen 2020a). This influence is bolstered by the epistemic nature of transnational tax specialist communities that has an impact on how civil servants and tax scholars take part in political processes such as legislative drafting (Carpenter 2007;Genschel and Rixen 2015;Finnemore and Sikkink 1998;Keck and Sikkink 1998).
There have also been a few recent empirical contributions specifically related to tax avoidance regulation. A case study by Sikka (2008) shows how the Big Four accounting firms 6 have assisted Jersey in designing tax legislation that harms other states as well as exploited this legislation to pressure legislators in the United Kingdom. In their article, Karlsson and Matthiasson (2015) describe how special interest groups have been able to influence tax legislation to facilitate their tax avoidance in Iceland since the 1930s. Raitasuo (2018 and has used empirical court material to show that many -or perhaps most -publishing tax scholars in Finland have their clients' interest at stake in tax avoidance cases related to their research. This might impact their interpretation of law and political involvement (Eisenberg 1993). For this reason, Raitasuo (2019) calls tax law research the science of tax avoidance. Christensen (2020a) used qualitative content analysis (Schreier 2012) of two public consultations and interviews of informants involved in the policy process to analyze how private sector tax specialists and civil society organizations influenced the OECD country-by-country reporting (CBCR) standard 7 . Christensen's analysis shows clearly that the vast majority of private sector tax experts were against increased tax transparency that would have been consequence of extensive public CBCR. He concludes that they successfully used technicization of policy discussions as a tool to re-frame public CBCR as excessive tax transparency. This eventually "moved CBCR from a revolutionary public lever for systemic change to a confidential risk assessment tool" (see also Murphy, Janský, and Shah 2019).

Tax avoidance and anti-tax avoidance regulation in Finland
Tax avoidance by MNCs, that is also called aggressive tax planning, involves optimization of corporate income tax (CIT) that is paid out of corporate profits. Tax avoidance takes advantage of legal means such as incentives or loopholes in tax regimes and should be separated from tax evasion that refers to illegal arrangements such as underreporting income (Kirchler, Maciejovsky, and Schneider 6 The term Big Four refers to KPMG, EY, Deloitte and PwC that are global networks offering audit and tax consulting services. 7 The CBCR standard defines how multinational corporations must report their business activities and taxes accrued in different countries.

2003)
. Aggressive tax planning -as opposed to regular tax planning -is understood to be literally legal, but ethically questionable or against the intended purpose of the law (Knuutinen 2014). The OECD more accurately describes the phenomenon in a cross-border setting as profit shifting that consists of two components (OECD 2015). First, a MNC makes arrangements whereby profits are shifted from the jurisdiction in which the activities creating the profits take place. Second, these shifted profits are taxed at a lower rate. Profit shifting is facilitated by harmful tax competition whereby countries often called tax havens use tax incentives to attract profits from other countries (European Commission 2021). The thin capitalization or debt shifting that is specifically discussed in this article is generally understood as a form of tax avoidance whereby MNCs or international investors use debt arrangements to shift taxable profits to countries with substantially lower CIT rates (OECD 2017b). This kind of profit shifting is possible as interest costs are tax-deductible but the corresponding interest income in another country is tax exempt or subject to a low tax rate. Thin capitalization is one of the most studied methods of tax avoidance and there is wide empirical evidence on its application (OECD 2017b; Ruf and Schindler 2015;Finér and Ylönen 2017). In Finland, thin capitalization has caused annual tax losses of hundreds of millions of euros, which was the motivation behind enacting the antitax avoidance rule discussed in this article (see section 5 and Finnwatch 2016; Lumme and Ropponen 2020). Also, recent international studies have estimated the Finnish tax losses because of profit shifting at somewhere between hundreds of millions and a billion euros, i.e., 12% of corporate revenue (Tørsløv, Wier, and Zucman 2020).
Profit shifting is facilitated by the two pillars of the international tax system: the arm's length principle and the separate entity doctrine (Finér and Ylönen 2017;OECD 2013). According to these principles, individual entities belonging to the same MNC use arm's length prices in their mutual transactions and are separately liable for their taxes. These principles facilitate tax avoidance as MNCs have an incentive to report profits in group entities registered in low tax jurisdictions. However, most countries have different anti-tax avoidance rules meant to tackle these kinds of tax arrangements, such as general anti-tax avoidance rules (GAAR) and IDLRs (OECD 2017b).
The corporate tax regime in Finland is similar to that in most developed countries. Since 2014, the CIT rate in Finland has been below the global average at 20%, but the tax base is broad and therefore Finland is not involved in harmful tax competition that would attract profit shifting from abroad (European Commission 2017a). Interest costs have generally been fully tax-deductible from corporate income and intra-border group contributions are allowed between group companies to offset taxable profits and losses 8 . This has facilitated base erosion as the Finnish GAAR 9 , transfer pricing adjustment rule 10 have not been adequate in tackling thin capitalization, which was the reason for adoption of the IDLR in 2014. However, the IDLR has not been able to tackle all forms of thin capitalization due to its loopholes (see later in this section).
Currently, there is consensus in the international community led by the OECD, the International Monetary Fund (IMF), the United Nations (UN) and the European Union (EU) that profit shifting and the tax competition facilitating it are harmful for the societies and cause inequality (see e.g., European Commission 2015; OECD 2013; IMF 2019). This is because ownership of MNCs is in the hands of the rich and their tax avoidance is ultimately of benefit to these owners (Cronin et al. 2013;Riihelä and Tuomala 2020). Progressive income taxation is not possible if some of the highest incomes, such as the profits of MNCs, are not sufficiently taxed (Rixen 2016) These failures in the current corporate income tax system undermine social welfare as profit shifting prevents countries from financing budgets equitably (Reisch 2020). The consequences are not equal for different countries. A few smaller countries sometimes referred to as corporate tax havens benefit from the phenomenon at the expense of the rest of the world whose fiscal sovereignty is undermined by the harmful tax practices of these havens (Clausing 2021; IMF 2019; OECD 2013; Tørsløv, Wier, and Zucman 2020) 11 . Specific concern has been raised for developing countries as they are more dependent on CIT revenues and inability to grow their budgets harms their chances of generating growth and welfare (Gaspar, Jaramillo, and Wingender 2016; OECD 2020a). Funding healthcare, education and infrastructure publicly are all necessary for 8 Contributions between Affiliated Companies Act, 21.11.1986/825. 9 Act on Assessment Procedure, 1558/1995, §28. 10 Act on Assessment Procedure, §31. 11 A special tax committee of the European Parliament (2019) has specifically named seven EU countries that facilitate aggressive tax planning by displaying traits of tax haven: Belgium, Cyprus, Hungary, Ireland, Luxembourg, Malta and the Netherlands. These countries as well as several jurisdictions outside the EU such as Singapore and Switzerland are also highlighted in academic studies and statistics concerning corporate profits shifting (Clausing 2021;IMF 2019;Tørsløv, Wier, and Zucman 2020). These corporate tax havens that are often relatively transparent and contribute to administrative cooperation in tax matters should be separated from tax havens attracting primarily tax evasion, which are sometimes called secrecy jurisdictions (Beard 1985; Tax Justice Network 2021a). their development. The loopholes in the current tax system could also be considered to cause market failures as profit shifting is a luxury to the few MNCs (European Commission 2017b). They gain competitive advantage as their local competitors cannot benefit from similar arrangements. Therefore, it is generally understood that a broad tax base with fewer loopholes for tax avoidance support fair competition and thus economic growth (European Commission 2015; Rixen 2016).
There is also wide consensus in the international community that effective national anti-tax avoidance measures are necessary to tackle abuse of public interest due to profit shifting (OECD 2020b). Specifically effective IDLRs are considered necessary to prevent profit shifting since "it is well known that groups can easily multiply the level of debt at the level of individual group entities via intragroup financing" (OECD 2017b) 12 . Previous empirical research has not either observed any significant harmful consequences of the IDLRs (Harju, Kauppinen, and Ropponen 2017; Schindler and Ruf 2015). Therefore, I consider that effective IDLRs serve public interest. However, it is not possible to determine unequivocally, which kind of IDLRs should be considered sufficiently effective to tackle tax avoidance that is harmful to the public interest 13 . Nevertheless, the OECD (2017b) best practice approach could be deemed as a consensus benchmark since it was developed within the OECD Inclusive Framework on BEPS that involves over hundred jurisdictions globally. The OECD best practice is based on a fixed ratio rule and is designed in a way that it should not generally limit the deduction of interest that is not related to debt shifting 14 . The best practice approach includes some level of discretion, but the OECD recommendation includes criteria that could be used to assess the effectiveness of the rules 15 .
12 The OECD BEPS Action 4 is dedicated to Limitation on Interest Deductions (OECD 2021c). In 2019, 67 of the OECD Inclusive Framework jurisdictions had at least one specific interest limitation rule in place. 13 In his article, Shapiro (2012) has discussed the complexity of defining public interest. 14 The fixed ratio rule could apply to loss-making entities even if there is no debt shifting involved. However, in a loss-year entities generally have no profits to tax and the OECD approach allows interest from the loss-years to be deducted on a later year when the entity is profitable. The OECD best practice approach includes a fixed ratio rule with a benchmark net interest to profit (EBITDA) ratio within a corridor of 10-30% and several optional components such as the group ratio rule. The OECD approach also allows applying stricter rules to debt shifting. 15 E.g., the fixed ratio should be set closer to the lower boundary of the 10-30% corridor in a country such as Finland that has a group ratio rule or EER, allows carry-forward of unused interests and does not The Finnish IDLR is based on a fixed ratio rule that follows the OECD best practice approach that also allows an EER 16 . However, it has included several unique features and exceptions that are not explicitly discussed by the OECD. For instance, the equity escape rule (EER) of the Finnish IDLR has been unique (Tell 2017). Most countries with IDLRs have not had such exceptions or had in place the OECD best practice first choice group ratio rule (OECD 2017b). The EER is more generous for many MNCs compared to the group ratio rule as it allows also a lossmaking entity to deduct all its interest expense when applied (OECD 2017b). In consequence, the interest expense could be utilized by another profit-making entity of the MNC. Germany has had a similar but significantly less generous EER since it could only be applied, if "the entity can demonstrate that interest payments on related-party loans [25% threshold] from shareholders outside the group do not exceed 10% of the group's total net interest expense" (OECD 2017 b). Diverging from the OECD (2017b) best practice, the Finnish EER has not had such a limitation or any other targeted rule to address the risk that a group ratio rule or an EER "could be inflated using interest paid to a related party outside the group".
Due to these unique features, the Finnish IDLR has facilitated profit shifting that has caused tax losses of at least tens of millions of euros each year (see e.g., Ministry of Finance 2018; Finnwatch 2018 and 2020a). Based on the Ministry of Finance evaluations these tax advantages benefit only a few dozen MNCs. Therefore, I deem that the gaps in the IDLR undermine the public interest. These deficiencies of the IDLR and EER have also been highlighted in previous literature as well as in several statements discussed in section 5 (Viherkenttä 2020;Väisänen 2020) 17 . Also, the current government has acknowledged the profit shifting have other anti-tax avoidance rules that target thin capitalization. The Tax Justice Network (2021a) Corporate Tax Haven Index includes more unambiguous criteria for assessment. Generally, all opt-ins as well as low thresholds compromise the effectiveness of the IDLRs. There is evidence that the Finnish threshold allowing the total of all intra-group interest below = C500 000 of an entity to be deducted has been used in tax planning by dividing loan arrangements between group entities (Finnwatch 2020b). 16 The case study in section 5 discusses the rule and its amendments in detail. 17 In the 2018 consultations, e.g., Professors Niskakangas and Viitala, as well as the Finnish Tax Administration, the Ministry of Finance and several civil society organizations highlighted the problems of the EER. Primarily the corporate interest groups and the tax advisors supported the EER (see section 5.4, Figure 8 and Table 7). The Tax Justice Network (2021b) Corporate Tax Haven Index also notes the deficiencies of the Finnish IDLR. facilitated by the EER specifically in venture capital investment structures and has announced that the IDLR is to be revised from the beginning of 2022 (Ministry of Economic Affairs and Employment 2021).

Research methodology and selection of the case public consultations
The case study in section 5 covers various stages of legislative drafting process of the IDLR ranging from 2009 until 2018. The Finnish legislative drafting processes can be outlined in nine stages (Finlex 2019) The consultation documents from the various stages of the legislative drafting process of the Finnish IDLR form the main research material of this article. I used qualitative content analysis (Drisko and Maschi 2015) to analyze these documents in the context of regulatory capture discussed in section 2. This approach allowed for the systematic analysis of the statements and other documents in their specific context. The method also facilitated the creation of numerical and graphical presentation of findings. I also used selected quotations from the 148 stakeholders' statements that were included in the research material. I have chosen and translated these quotes from Finnish as they illustrate the contents of the statements and are meaningful from perspective of my research questions. I began the analysis by looking at the requests and other material such as legislation drafts that were submitted to the stakeholders for consultation 18 . Then I looked to see if there were any major legislative changes abusing 18 The consultations in Finland are generally open to all individuals and entities. However, the consultation requests are generally submitted to a limited group of stakeholder groups that provide most of the statements, as other groups are not aware of the process or do not have direct interest regarding the proposed law. the public interest to the draft indicating a possible regulatory capture. Then I analyzed the actual consultation documents to answer specifically these questions: 1. Which stakeholder groups were involved? 2. Were the stakeholders for or against the suggestions? 3. What amendments did they request? E.g., did they request more/less effective tax avoidance legislation? 4. How did different stakeholder groups respond to the consultations? How did their background and interests appear in their views? 5. Whose comments were taken into account if the consultation influenced the corresponding rule?
My case study included all seven consultation rounds related to the Finnish IDLR in the course of nine years. This allowed me to assess if there were any changes during the various consultations. To check and improve validity and reliability of my analysis, I also conducted two semi-structured interviews (Hirsjärvi and Hurme 2000) of key people responsible for drafting and political negotiations of the initial IDLR before 2012 (Interview 2019 and 2020).

The 2009 consultation on the options for the interest deduction limitation rule
Tax avoidance that exploits interest deductions has been discussed in Finland within ministry working groups since the 1970s when Canada had already adopted an IDLR (Ruf and Schindler 2015). However, this discussion remained mostly private between tax specialists until 2009, when the Ministry of Finance prepared a preliminary report on options to introduce IDLRs in 2009. The report studied IDLRs in ten reference countries and included a statistical study that showed the increase in interest deductions in Finland 19 . According to a former government official, the initiative to draft the report came from civil servants (Inter-

19
The author received a copy of all the consultation materials from the Ministry of Finance as the documents are public information in Finland.
view 2020). The Ministry of Finance submitted the report to 13 stakeholders and presented them with two questions: 1. Should Finland adopt an IDLR? 2. If an IDLR were to be adopted, what kind of rule should it be?
Eleven of the stakeholders responded to the request and two other groups submitted a statement (see Figures  1 and 2 and Table 1). Most of the respondents represented private interests as the consultation involved six corporate interest groups and a tax advisory interest group. The consultation also involved an economic research institution, two tax scholars and three public entities. However, the consultation did not involve any civil society organization or other entity representing general public interests 20 .
The views on adoption of IDLR varied according to the interests of different groups. The corporate interest groups and the tax advisors were generally against any IDLR (see Figure 1). Two of the corporate interest groups did not present any clear view for or against the rule whatsoever. In contrast, all the tax professors and public entities considered Finland should adopt an IDLR. View (1-5): 1 = Should not be adopted 2 = Should not be adopted with reservation 3 = No opinion 4 = Should be adopted with reservation 5 = Should be adopted Also, the views on the facts regarding tax avoidance varied a lot. The Confederation of Finnish Industries EK 20 The public entities in Finland are generally considered not to represent any systematic political views as their responsibilities are defined by law and other means of political steering. However, e.g., tax authorities pay attention to aspects such as governance of tax system and equal tax control that are meaningful from the perspective of their responsibilities defined in law. Tax professors and researchers do not represent any specific perspective as their personal political views and other interests vary. They are generally considered to represent the views of their academic discipline. View (1-5): 1 = Should not be adopted 2 = Should not be adopted with reservation 3 = No opinion 4 = Should be adopted with reservation 5 = Should be adopted Type (1-3): 1 = Should apply merely to artificial arrangements 2 = No opinion on type 3 = Fixed ratio rule (Elinkeinoelämän Keskusliitto) 21 presented the other extreme as they claimed that tax avoidance is merely a theoretical phenomenon that does not have significant impact on Finnish CIT returns as the GAAR and transfer pricing rules allow possible abuse to be tackled. Inconsistently, they claimed the lack of IDLR is more of a competitive advantage for Finland as it could attract foreign corporations.
EK cited no empirical evidence to confirm these claims that were in contrast with the data referred to in the statements of tax authorities and tax professors that showed debt shifting erodes the Finnish tax base. E.g., Professor Helminen from the University of Helsinki stated 22 : The Finnish tax regime has been exceptionally favorable for taxpayers compared to the relevant reference countries. This situation has been well-known for a long period and the contrast to the other countries has increased as they have adopted new thin capitalization rules. It is clear that taxpayers have actively exploited this loophole with the help from tax advisors by making specifically cross-border tax arrangements. This has naturally meant losses of tax revenue for Finland.
The Large Taxpayers' Office responsible for tax control of all major MNCs presented detailed case examples of debt shifting as well as statistical information of possible tax losses in their statement. They studied more closely 23 companies with high interest expenses that had paid out intra-group interest expense of total = C350 million in 2006 21 EK is considered to be the most powerful corporate interest group representing 15 300 member companies including the largest corporations operating in Finland. 22 Professor Helminen is the supervisor of the author's doctoral thesis. and = C494.6 million in 2007. In most of these 23 cases, a company had received a group contribution that more or less matched the interest paid abroad. This consequently eroded the Finnish tax base by over = C100 million 23 .
Some of the corporate interest groups admitted the base erosion in their statements: The Finnish Chamber of Commerce considers that the Finnish legislation allows certain tax avoidance arrangements that could be considered leading to results that are not in line with general principles of international taxation and cause inequality between taxpayers. 24 Some of the corporate interest groups proposed that Finland should facilitate base erosion instead of protecting its tax base with an IDLR. The Association of Finnish Tax Advisors suggested "Finland should also consider rules that favor equity contributions that are available e.g., in Belgium and in the Netherlands". Also the Family Business Network Finland proposed the adoption of notional interest deduction (NID) for equity contributions. However, the European Commission (2015) has considered deductions of this kind to be to harmful tax competition as they facilitate profit shifting.
The views on the content of the possible IDLR also divided between the different groups ( Figure 2). The corporate interest groups and an association representing private sector tax advisors claimed that the application of the possible rule should be limited to merely artificial arrangements and only to major cross-border intra-group transactions. These targeted rules could be considered to be ineffective in tackling anti-tax avoidance as their application requires an ambiguous case-by-case assessment of taxpayers' motives similar to the application of GAAR (OECD 2017b). Because of this, the Large Taxpayers' Office proposed an explicit IDLR, whose "application should not be based on possible motives of taxpayers as it is extremely difficult for authorities to prove if there are no business reasons for such arrangements". Also, the Finnish Tax Administration concluded such fixed ratio rule would be better as it would increase legal predictability by being of benefit to the taxpayers as well.

The 2012 consultations on the draft initial legislation
The 2009 consultation did not materialize into the legislative drafting process of the IDLR before the change of rightwing coalition government after the 2011 parliamentary elections. Therefore, it is possible that the corporate interest groups were able to delay the application of the rule with their statements, indicating a weak or corrosive regulatory capture. The new six-party coalition government with parties from both left and right in the parliament was led by Prime Minister Jyrki Katainen of the center-right National Coalition Party. The Finance Minister Jutta Urpilainen responsible for tax legislation drafting was from the center-left Social Democratic Party. The adoption of the IDLR was not included in the government program (Prime Minister's Office, 2011) that merely contained a general statement that "Finland will be active in tackling international tax flight". However, the situation changed on November 17, 2011, when the Keskisuomalainen newspaper had a scoop about the thin capitalization of Mehiläinen, a major health care corporation (Martinmäki 2011). This caused much upset as nearly all of Mehiläinen's turnover came from public sector health service procurements. Just four days after the scoop, Finance Minister Urpilainen published a press release on November 21 in which she stated that the rules on interest deductions would be amended to tackle debt shifting. The Ministry of Finance published a similar press release the next day (Ministry of Finance 2011). Due to this, the IDLR is also referred as the Mehiläinen law (Bernitz 2013).
Consequently, the regulatory drafting of the IDLR began in the Ministry of Finance and it launched public con-sultation of the first IDLR draft in April 2012. The proposed rule was much against the will of the corporate interest groups in the 2009 consultation 25 ; i.e., the rule much drafted based on the German IDLR was not targeted. It could be concluded the proposed draft rule was an explicit fixed ratio rule that only limited deductibility of intragroup interest costs. The key elements of the draft rule included: 1. Net interest costs exceeding 30% of the annual earnings before interest, taxes, depreciation and amortization (EBITDA) are not tax-deductible. However, these costs are deductible if the total net interest costs of an entity are less than = C500 000. Also, all interest costs paid to unrelated parties are taxdeductible independent of any thresholds. 2. The non-deductible interest costs may be deducted on a later year according to the same formula with no time limit. This interest is deductible even after mergers and acquisitions. 3. Parties are considered related, if other party holds an ownership of more than 50% or other corresponding command over the other party (or a thirdparty commands both parties). Also, interest costs paid to third parties are considered to have been paid to a related party if a related party has provided collateral for the concerning debt. 4. The draft rule included no other exceptions (such as an equity escape rule). It covered limited companies as well as cooperatives and also covered income from real estate investment that is generally not considered to be business income in Finland.
The Ministry of Finance made estimates about the impact of the rule based on 2007-2010 corporate tax returns. It estimated the rule would be applied to around 300 companies and consequently annual CIT income would have increased between = C111-149 million 26 . The long run full tax revenue increase potential was estimated at = C152-246 million annually 27 . 25 All the consultation documents are available at https://vm.fi/hanke?tunnus=VM044:00/2012. 26 Approximately 3% of total Finnish CIT income of = C4,443 million in 2010. 27 The estimates were cautious as they did not include most real estate business. Neither do the estimates represent the full impact of thin capitalization tax losses as the thresholds of the proposed rule still allowed a part of interest costs created in tax avoidance arrangements to be deducted in taxation. E.g., all the interest costs below the 30% EBITDA threshold are tax deductible and the interests could be deducted on a later year if the business turns out more profitable. The OECD has suggested a threshold at 10-30% to EBITDA, since a lower In total, 25 parties took part in the first consultation round (Table 2). Again, majority of the statements (14) were from corporate interest groups and tax advisory groups. Two tax researchers and the Finnish Tax Administration also provided statements similar to 2009.
In contrast to the 2009 consultation, this time civil society was involved as one non-governmental organization and a trade union submitted statements. Both supported the aim of anti-tax avoidance. Their statements differed extensively from others as they were on a general level with no technical proposals or comments. Many of the other statements included detailed technical comments, specifically the ones by the tax advisors.
Another significant change compared to 2009 was the fact that six parties representing a specific public interest took part in the consultation. These public entities included municipalities' institutions, publicly owned entities involved with public infrastructure projects and an alliance presenting public employee pension funds. All these parties brought attention to the risk that the proposed rule might affect the financing of their projects or investments. These groups considered the IDLR to be justified, but argued their specific situation deserves an exemption from the rule as their activities promote the public interest. This highlighted a phenomenon, by which public entities representing a narrow perspective resist a rule such as the IDLR that strives to promote the welfare of the society. In some circumstances, this could result in harmful sub-optimization within the public sector as from its viewpoint tax payments and other monetary transfers between public entities could be considered a zero-sum game. This highlights the need for coordination within the public sector when seeking to achieve cross-sectoral policy targets.
Like in 2009, the views presented in the statements varied greatly according to the interests of the various groups. The corporate interest groups and tax advisors were generally strongly against the proposed IDLR (see Table 2 and Figure 3). In contrast, both the tax professors, the Finnish Tax Administration, and civil society considered that Finland should adopt the IDLR. These parties also generally considered the proposed fixed ratio rule to be well functioning (see Figure 4). However, all the corporate interest and tax advisory groups proposed amendments to the rule that would have made it less extensive in threshold is more effective in tackling tax avoidance (OECD 2017b, see footnote 15). It is also possible that the estimates included interest not related to debt shifting due to the fixed ratio rule. However, the design of the rule minimizes this opportunity (see footnote 15). tackling tax avoidance. Over half of the corporate interest and tax advisory groups demanded three moderations to the rule: (1) an equity escape rule (EER); (2) an exemption for the real estate business; and (3) the delay of effective date of the rule by a year from 2013 to 2014 (see Table 2). For example, the Finance Finland additionally requested the rule should not cover finance companies.

Figure 3: Should the interest deduction limitation rule be adopted?
View 1 = Should not be adopted 2 = Should not be adopted with reservation 3 = No opinion 4 = Should be adopted with reservation 5 = Should be adopted Figure 4: What kind of rule should be adopted? Type 1 = Should be a lot let extensive 2 = Should be little less extensive 3 = No opinion 4 = As proposed 5 = Should be more extensive This time the corporate interest groups based their argumentation on the claim that such a rule would harm the competitiveness of the Finnish tax system and consequently investments. They also claimed that the proposed fixed ratio rule would apply to finance not related to any tax avoidance. EK was again clear with its claims: Interest costs should be as widely deductible as they are currently and the proposed IDLR should not be put forward. -- Table 2: Statements in the 2012 April-June consultation.
View 1 = Should not be adopted 2 = Should not be adopted with reservation 3 = No opinion 4 = Should be adopted with reservation 5 = Should be adopted Type 1 = Should be lot let extensive 2 = Should be little less extensive 3 = No opinion 4 = As proposed 5 = Should be more extensive

Equity escape rule / Delay until 2014 / Real estate exemption / Specific public interest 1 = Yes 2 = No
The proposed general limitation of interest deduction will be applied to all Finnish groups independent of if there is any tax avoidance. This will harm finance of nearly all Finnish corporate groups.

--The IDLR will increase investment costs and decrease the desire to invest in Finland.
Similar claims about negative consequences for the economy were presented by all the corporate interest and tax advisory groups. However, these statements did not refer to any supporting empirical evidence 28 . Many of the corpo-28 An exception to this was the statement by the Finnish Venture Capital Association FVCA that included a reference to the working paper that had assessed the likely impacts of IDLR on tax avoidance and investment (Büttner et al. 2008). The paper used a crosscountry panel data approach and indicated that thin capitalization rules would be effective in curbing tax planning and have an adverse effect on foreign investment (see also Büttner, Overesch, and Wamser 2018). Other studies have not found any significant effect on investment due to thin-capitalization rules (Weichenrieder and Windischbauer 2008;Ruf andSchindler 2015, Harju, Kauppinen, andRopponen 2017). However, none of the studies referred to above have rate interest groups also argued the rule would not suit the specific group contribution system applied in Finland. A couple of statements including those of Finnish Bar Association and PwC included example calculations to support the argument. These calculations were an attempt to show the proposed rule might lead to unequal treatment of taxpayers belonging to corporate groups. However, these calculations did not assess the alternative ways to organize intra-group finance that would have led to different result. That is, they did not present a realistic consequence of the legislation, which is emphasized by the fact that these arguments have not been supported by later experience (see e.g. (Harju, Kauppinen, and Ropponen 2017).
The statements by other parties were contrary to those of the corporate interest groups. The Finnish Tax Administration did find the proposed rule would be an effecassessed the impact on total investments or in the long run. The venture capital investors are the most prone to thin capitalization arrangements, as the great majority of publicly-discussed tax avoidance arrangements in Finland have been made by them (Finnwatch 2016). tive way to tackle thin capitalization. It also claimed that thin capitalization arrangements had become more popular not only among foreign corporations investing in Finland, but also Finnish MNCs had shifted profits abroad with debt arrangements. Additionally, the Corporate Income Tax Specialist Group, whose members included Ministry of Finance officials and tax researchers favored the proposed IDLR in its memo (Ministry of Finance 2013).
Corporate lobbying against the IDLR had begun after the government announced its plans to adopt the rule in November 2011 (Interview 2019). However, according to the two interviews (2019 and 2020) conducted for this article, the details of the first legislation draft described above were designed by civil servants without significant political influence. The high technicality of international tax avoidance and IDLR also limited the direct influence of the political parties in the legislative details (Interview 2020). The idea was that the political decisions on the details could be based on the consultation round of the draft. Both interviewees said that the corporate lobbyists' proposals and arguments had some impact on the second draft of the IDLR in September 2012. Corporations were directly in contact with major political parties but also some academics opposed the proposed rule with corporate interest groups at the meeting of the International Fiscal Association Finnish branch in 2012 (Interview 2019 and 2020). The goal of this lobbying could be perceived in the statements of the corporate lobbyist groups (see Table 2).
The lobbying efforts of the corporate interest groups and tax advisors had some impact on the second draft. According to an interview (2019), some of these compromises were made specifically because of requests by Prime Minister's the center-right National Coalition Party (Kansallinen Kokoomus) that has close connections with corporate interest groups (Koiranen et al. 2017). Moreover, the civil servants endorsed some of the amendments as they considered prompt adoption of the rule and wide consent of the details among the government parties was important for the legislation to pass in the parliament (Interview 2019 and 2020). The precautionary principle also played a role in the decisions, as such a rule had not been used before in Finland.
Nevertheless, the explicit fixed ratio rule approach was preserved also in the second draft September 2012, but several major exceptions were included based on the corporations' and their advisors' statements that made rule more generous: 1. The equity escape rule (EER) was included. It meant that the IDLR would not be applied to a group company at all if the equity ratio of the whole group was the same or below the equity ratio of the given group company (OECD 2017b). The justification was based on a claim that such ratio would indicate that interest costs have not been created in tax avoidance structures 29 . However, the IDLR did not include a targeted rule to address the risk that the EER could facilitate tax avoidance using interest paid to related parties outside the formal group structure (see section 3). 2. An exemption for finance, insurance and pension companies was included, which meant the rule would not be applied to these industries. 3. An exemption for real estate companies and other non-business income enterprises was added to the rule 30 . 4. The effective date of the rule was delayed one year from the 2013 tax year until 2014.
According to the Ministry of Finance's estimates, these amendments reduced the impact of the IDLR by roughly half compared to the previous version. The number of companies affected decreased from around 300 to 185. The decrease in real estate business is not included in the estimate, which means the actual decrease was higher (see footnote 27). The Ministry estimated the annual CIT income would increase by = C70 million compared to the = C111-149 million in the previous draft.
It could be considered these four amendments promoted by the corporations and their tax advisors facilitated profit shifting, which compromised public interest and thus can be deemed evidence of a weak regulatory capture 31 . Additionally, pension funds and certain public real estate companies representing specific public interest were granted an exemption from the IDLR. It is possible that their requests for the exemptions lowered the threshold to include the other exemptions such as the EER rule that favored private interest.
In the second consultation round, 11 stakeholders submitted their statements (see Table 3). Six of them were from corporate interest groups. These groups were still critical of the adoption of the IDLR, but also acknowledged the 29 However, no evidence backing this claim has been presented. Professor Viitala argued in 2018 that this claim on EER lies on merely hypothetical assumptions (see section 5.4). I discuss the specific features of Finnish EER in section 3. 30 It is a Finnish legal peculiarity that income from real estate investment is generally not considered business income. 31 I discussed the impact of EER in section 3, the impact of the exemption for real estate business was discussed in the parliamentary consultation. There is also later evidence on thin capitalization in the real estate business (Finnwatch 2020b). amendments were positive from their perspective. Therefore, their criticism was less pronounced 32 . The other five stakeholders were more or less satisfied with the amendments to the rule.
To conclude, the second consultation round did not include any significant new argumentation. Neither did it have a major impact on the final government bill presented to the parliament in November 2012 33 , which was all but identical to the Ministry of Finance second draft in September. The parliament held another public consultation with 20 stakeholder groups attending (see Table 4). These statements did not have an impact on the enacted law accepted in December 2012. The MPs of the centrist opposition Centre Party and conservative Finns Party criticized the loopholes left in the law in the parliamentary Finance Committee 34 . The Finns Party MPs considered that the IDLR should apply to real estate businesses and did not see any justification for the EER. The Centre Party MPs also raised concerns over continuing real estate companies' tax avoidance, the EER, exemption for finance, insurance and pension companies and the delayed effective date of the rule.
The IDLR took effect from the beginning of the 2014 tax year. However, it was amended before the effective date as the EBITDA rule was amended to the EBITD 35 rule and the threshold in proportion to EBITDA was decreased from 30% to 25%, which made the rule slightly more effective on tax avoidance. This amendment was done with respect to larger tax reform including CIT rate deduction from 24.5% to 20% and did not raise any public debate 36 . 32 However, EK and the Finnish Chamber of Commerce requested in their statements the lowering of the CIT rate from 24.5% to 20% due to the IDLR. This is notable as the government made the decision to reduce CIT correspondingly the following year. This was a significant tax reform that reduced tax revenue by nearly = C1 billion, which benefited the corporations and their owners.

The 2016 consultation on anti-tax avoidance directive proposal
The European Commission made a proposal on the Anti-Tax Avoidance Directive (ATAD) in January 2016 as part of its Anti-Tax Avoidance Package (European Commission 2016). The previous work in the OECD BEPS project played a significant role in enabling the proposal as the EU was committed to the BEPS project and neither the Commission nor the member states had paid much attention to harmonizing thin capitalization rules prior to BEPS (Interview 2020). This is evident as the recommendations prepared in the OECD (2017b) formed the basis of the Commission's proposal. The proposed ATAD 37 included a minimum standard of six anti-tax avoidance measures to be applied by the EU member countries. These measures included IDLR similar to the Finnish rule with the notable difference that the minimum standard would limit deductions of all net interest costs including those paid out to third parties (e.g., banks).
Before forming their view on the Commission's directive proposal, the committees of the Finnish Parliament consulted twelve stakeholders on a few occasions from March to May 2016 (see Table 5) 38 . This time the corporate interest groups and tax advisors submitted just below half of the statements. A major difference compared to the previous consultation was that civil society was better represented as two human rights-based non-governmental organizations (NGO) and two trade unions took part along with two tax professors and the Finnish Tax Administration.
The corporate interest groups and tax advisors were against the proposed directive (see Figure 5). Some of them argued that the directive should not be passed without a more thorough assessment of its economic impact, which was one of the shortcomings of the directive proposal (Interview 2020) 39 . The corporate interest groups also demanded a lower minimum standard in case the ATAD would be adopted (see Figure 6). The Confederation of 37 COM/2016/026 final. 38 All the documents are available at https://www. eduskunta.fi/FI/vaski/KasittelytiedotValtiopaivaasia/Sivut/U_7+2016 .aspx. 39 It must be noted that the exact fiscal implications of tax avoidance and its tackling are often impossible to measure (Finér and Ylönen 2017). This was also shown in the BEPS assessment made in Finland the following year, that could only provide numeral estimates regarding thin capitalization but not on other means of aggressive tax planning (Ministry of Finance 2017). View 1 = Should not be adopted 2 = Should not be adopted with reservation 3 = No opinion 4 = Should be adopted with reservation 5 = Should be adopted Type 1 = Should be lot let extensive 2 = Should be little less extensive 3 = No opinion 4 = As proposed 5 = Should be more extensive The tax professors, tax authorities and the civil society supported the ATAD. This time, some of civil society orga- View (1-5) 1 = Should not be adopted 2 = Should not be adopted with reservation 3 = No opinion 4 = Should be adopted with reservation 5 = Should be adopted Type (1-3) 1 = Should be less extensive 2 = No opinion 3 = Should be more extensive nizations also made technical comments as human rights based NGOs Finnwatch and Kepa (predecessor of Fingo) provided detailed proposals to make the rules more effective against tax avoidance 40 . Finnwatch argued that the ATAD would increase the actual tax sovereignty of Finland and other EU member states as it would limit harmful tax competition that erodes their tax bases 41 . They also highlighted the possible dynamic effect on growth supported by better functioning markets with less tax avoid-40 The author worked for Finnwatch at the time. 41 Also, the OECD (2013) have had similar conception on the impact of BEPS on national sovereignty. ance and harmful tax competition. Professor Helminen also presented arguments in favor of the directive: The current international CIT regime offers excessively attractive chances to avoid tax. In this situation, it is clear that rules governing international taxation must be amended on national, EU and bilateral treaty level. These reforms are not only necessary to collect tax revenue, but also to preserve taxpayers' sense of justice and tax morale.
--It is clear that EU level regulation always limits national sovereignty on given regulations. Concerning the proposed rules, it cannot be considered to be a major problem, since the proposal covers merely a coordinated minimum standard of regulation that is for the most part already part of Finnish law. 1 = Should not be adopted 2 = Should not be adopted with reservation 3 = No opinion 4 = Should be adopted with reservation 5 = Should be adopted Figure 6: How should possible ATAD be adopted?
Type (1-3) 1 = Should be less extensive 2 = No opinion 3 = Should be more extensive Agreement about the ATAD in the EU process was agreed exceptionally quickly in June 2016, less than six months after the Commission's proposal 42 . The public pressure caused by the Panama Papers data leak 43 that broke in April 2016 might have been one reason behind the quick outcome. Also significant was the rather surprising will of the Netherlands -often considered to be a corporate tax haven -to push ATAD forward during their EU presidency in the first half of 2016 (Council of the European Union 2016; Interview 2020). Moreover, the EU and many of its largest member states had pledged their political commitment to the OECD BEPS project thus tackling tax avoidance and harmful tax competition (Interview 2020). The quick adoption of BEPS measures not only raised the credibility of the project but also increased the political weight of the EU member states in global tax policy-making within the OECD (Interview 2020). However, the final ATAD was somewhat less ambitious than the Commission's proposal, as one of the six anti-tax avoidance rules -namely the switchover rulewas dropped. Also, the interest deduction minimum standard was compromised as the minimum threshold of deductible net interest costs was increased from = C1 000 000 to = C3 000 000, the member states were provided with the chance to use more liberal EER than the Commission had proposed 44 and they were given a chance to use several other exclusionary criteria 45 . Despite these compromises indicating a possible corrosive regulatory capture in favor of the corporate interests, the ATAD can be considered to have been a major change in the EU tax policy as a groundbreaking measure of income tax base harmonization and anti-tax avoidance. The previous direct taxation directives had merely covered tax assessment and exchange of information or limited the taxing rights of member states (Interview 2020).
42 Council Directive (EU) 2016/1164. 43 The Panama Papers covered mostly private international investment and related tax evasion in contrast to 2014 Lux Leaks that provided information on MNC tax avoidance facilitated by the Luxembourg authorities (see e.g. Obermaier et al. 2016). 44 In the Commission's proposal, EER could only be applied to companies whose interest payments to associated enterprises do not exceed 10% of the group's total net interest expense. The Commission's EER was close to the German version and might have forced Finland to amend its more generous EER (see section 3). 45 E.g., so-called standalone entities not part of corporate group and public interest serving infrastructure projects could be excluded from the application of the rule.

The 2018 consultations on draft legislation amendments
The ATAD minimum standard meant that Finland had to revise its IDLR before the beginning of 2019 to meet the standard. Changes had to be made as the ATAD IDLR covered third party interest costs in contrast to Finnish rule that merely covered intra-group interest. Also, the ATAD did not allow an exemption to real estate business. Consequently, the Ministry of Finance drafted amendments to the law and launched a public consultation for comments in January 2018 46 . In addition to mandatory changes based on the ATAD, the ministry proposed several other significant changes to the IDLR: 1. The equity escape rule (EER) should be abolished.
2. An exemption for finance, insurance and pension companies should be abolished, which meant the amended rule would be applied on these industries. 3. Parties are considered to be related if another party or a third party owns more than 25% or other corresponding command over the other party. The lower = C500 000 fixed net interest cost threshold would apply to interest costs between related parties instead of the higher = C3 000 000 threshold on arm's length interest costs. 4. The rule would not be applied to standalone entities not part of a corporate group.
The Ministry of Finance estimated the proposed amendments to the IDLR would increase CIT income by = C46 million annually compared to the effective rule. The long run full annual tax increase potential was estimated at = C106 million. The new rule would apply to 290 companies compared to the 185 affected by the effective rule. The estimates were cautious as they did not include most real estate business (see footnote 27).
The exclusion of the EER was the most significant proposal. The Ministry of Finance had two separate justifications for the exclusion. First, exemption had been used in base eroding tax avoidance. The impact assessment showed the exemption had been applied to 67 companies and the excluded interest costs at = C280 million were onethird of the otherwise non-deductible intra-group interest costs. Second, the EER had harmed taxpayers' legal security and equality as it had been open to different interpretations.
These proposals attracted unparalleled interest from stakeholders compared to the previous consultations, as the Ministry received 45 statements (see Table 6). The corporate interest groups, tax advisors and private companies submitted 19 statements. As in 2012, parties representing a specific public interest formed another major stakeholder group with 16 statements. In addition to municipalities and other public entities that had also contributed in 2012, this time, several non-profit housing associations as well as non-profit energy companies expressed their view. Tax professors (5), civil society organizations (4) 47 and the Finnish Tax Administration filed the other statements.
Like all previous consultations on IDLRs, the views presented in the statements varied greatly according to the interests of the different groups. The corporate interest groups and tax advisors argued for rules that should generally apply the sheer ATAD minimum standard (see Figure  7). Most of these stakeholders (16/19) representing private interests also demanded that the EER should be preserved and none of them argued for its abolition (see Figure 8). This was much in contrast with the other 26 statements, in which only two stakeholders argued reservedly for preservation of the EER and six stakeholders favored the Ministry's proposition of abolishing the rule. These six parties included civil society as well as tax administration and the Association of Finnish Municipalities. The four civil society organizations supporting human rights and other public interests also insisted on a generally more extensive rule compared to the Ministry's proposal 48 .
The preference of public entities presenting specific public interest generally was that the IDLR should exclude their specific activities including non-profit housing, nonprofit energy production and public infrastructure investment.
The corporate interest groups argued in favor of less extensive rules that were mostly based on a claim that extensive rules harm the competitiveness of Finnish enterprises. Diverging from previous consultations, a couple of tax scholars also gave their support to these claims 49 . However, as in previous consultations these arguments 47 The author worked at the time for the Social Democratic Parliamentary Group that submitted one of the statements. 48 They e.g. argued the interest to EBITDA ratio of the rule should be lowered to 10%. The current 25% ratio could be considered ineffective to tackle debt shifting based on the OECD best practice approach (see footnote 15). 49 This corporate view was most visibly shared in the statement of Professor Penttilä, who compared choosing rules exceeding the ATAD minimum standard to "shooting oneself in the foot". Penttilä has connections with private interests as he has worked for PWC and provided were not supported by related evidence. All the corporate interest groups and the tax advisors also failed to address any empirical research in their statements. This is notable as VATT, the renowned state-funded Institute for Economic Research had published an empirical study on the Finnish IDLR the previous year (Harju, Kauppinen, and Ropponen 2017). The VATT researchers concluded in their press release (2017): "The IDLR seems to be an effective measure against profit shifting. We did not observe any harmful consequences from the reform." Previous research from abroad has also not found evidence on significant harmful impact on investment (Schindler and Ruf 2015). Only four of the stakeholders referred to the relevant empirical research, including Professor Lindgren and three civil society organizations.  The Confederation of Finnish Industries EK was again unambiguous in its argument in favor of a less extensive anti-tax avoidance rule: Unfortunately, the draft proposal would limit interest deductions more than the directive requires even though these measures cannot be justified based on tackling tax avoidance.
--The proposal should be amended, and Finland should apply all the possible options in the ATAD to make the rule less extensive in order to secure the competitiveness of Finnish enterprises.
--If the proposed rule is put forward against our conception, the enterprises should be compensated for increases in tax costs by reducing the CIT rate by at least two percentage points [ = C600 million].
Also, the Finnish Chamber of Commerce argued that the EER would not facilitate tax avoidance by referring to legal praxis: The current EER can be considered to be well-functioning based on legal praxis. The Supreme Administrative Court has ruled its application in decisions KHO:2018:2 andKHO T 65 10.1.2019 in a way that e.g. artificial ownership restructurings cannot be used to apply the exemption. If these situations are still considered problematic, clarifying the definition of concept would effectively limit interest deductions.
However, the argument could be criticized as generalizations could not be drawn based on the court cases as they only covered very specific arrangements (Vesikansa and Stellato 2018) 50 . In reality, the EER could be used in tax avoidance as indicated in the Ministry of Finance (2018) draft as well as several statements in the consultation 51 .
The extensive lobbying by the corporate interest groups had an influence as several amendments were included in the final government bill after the consultation 52 . At the time, Finland had a right-wing coalition government with close connections to corporations, which might have had an influence on these amendments as well (Koiranen et al. 2017).
1. The EER was re-introduced. 50 I have discussed the issue more thoroughly in my blog (Finér 2018a and2018b). 51 This profit shifting has also been later highlighted in case studies on Caruna, the electricity distribution company, that specifically used the EER (Finnwatch 2018 and2020a). Finnwatch estimated the tax avoidance was worth = C12 million annually. 52 See the government bill HE 150/2018 available at https:// www.eduskunta.fi/FI/vaski/KasittelytiedotValtiopaivaasia/Sivut/HE _150+2018.aspx. View (1-6) 1 = Should be lot less extensive (ATAD minimum standard) 2 = Should be less extensive 3 = Exemption for specific group 4 = No opinion 5 = As proposed 6 = Should be more extensive Equity escape rule (1-5) 1 = Should remain 2 = Should remain with reservation 3 = No opinion 4 = Should be abandoned with reservation 5 = Should be abandoned 2. An exemption for finance, insurance and pension companies was re-introduced, which meant the rule would not be applied to these industries. 3. An exemption for interest paid out to debts granted before June 17, 2016. 4. The previous definition of related party was included in the rule (ownership of more than 50%). 5. Non-profit housing is not subject to the IDLR.
The Ministry of Finance estimated that after these amendments, the impact of the IDLR was = C10 million more than the then effective rule, which was less than a quarter compared to the = C46 million impact estimate of the previous draft version. Nearly all of the decrease was due to the inclusion of the EER. Because of these changes, the number of affected companies decreased from 290 to 221. The impact on real estate business has not been included in the estimates (see footnote 27).
These amendments indicate a similar weak regulatory capture compared to 2012 as it could be considered they harmed public interest in favor of a few dozen corporations (see section 5.2). Also, this time, the amendments requested by several entities representing specific public interest might have supported the claims to amend the legislative draft.
The final parliamentary consultation in fall 2018 did not lead to any notable amendments to the IDLR (see Table 7). However, some opposition left-wing and center parties demanded abolishing the EER and introducing other more extensive measures to tackle tax avoidance such as reducing the EBITD threshold to 10% in line with the OECD recommendation 53 .The drafting process was also covered in the media as it was revealed during the parliamentary hearing that Caruna, an electricity transfer corporation, had exploited the equity escape loophole (Finnwatch 2018). The theory of regulatory capture is drawn from the notion that corporations' profit-seeking is not limited to the given regulatory framework. The corporations are also active political actors that extract rents and minimize their regulatory costs by influencing the regulatory framework in their favor (see also Karlsson and Matthiasson 2015). In this article, I presented a thorough case study that allows nuanced analysis of the mechanisms of regulatory capture based on several public consultations covering 148 stakeholder statements on interest deduction limitation rules (IDLR) in Finland over a nine-year period. Thus, I answered the call from Carpenter and Moss (2014) by describing how and to what degree regulatory capture works in a particular setting. These findings show how corporate interest groups were able to affect the anti-tax avoidance regulation in favor of some corporations (see also Viherkenttä 2020; Väisänen 2020). This weak regulatory capture was supported by tax advisory interest groups. Tax advisors benefit from their corporate clients' tax avoidance, since it is facilitated by tax advisory services (Finnwatch 2017;Seabrooke and Wigan 2017;Sikka 2018). These findings present unique evidence on weak regulatory capture as I was able to analyze how specific interest groups' statements influenced the wording of the law, thus abusing the public interest. The inclusion of the equity escape rule (EER) in the IDLR after both 2012 and 2018 consultations was the most evident proof of the weak regulatory capture, but the case studies also provide additional evidence (see section 5). The adoption of EER alone has caused tax losses of at least tens of millions of euros each year. The Ministry of Finance assessments show the EER benefited only several large MNCs. These MNCs gained a competitive advantage compared to over 100,000 other Finnish firms that do not have similar opportunities to avoid tax (see section 3).
The case study shows how corporate lobbyist groups and tax advisory firms not only benefit from tax law loopholes but also actively generate and preserve them, allowing them to increase their share in the global wealth chains (see section 2.2). This interaction of regulatory captures must be understood when analyzing reactions to regulation as the political profit-seeking could have an impact on the market behavior as well.
However, the corporate interest groups and their advisors were not able to set the agenda as my research material shows they have consistently opposed all measures tackling thin capitalization over the nine-year period. Nevertheless, they were able to corrode the effectiveness of the enacted anti-tax avoidance measures indicating a weak regulatory capture (Carpenter and Moss 2014). This resulted from tens of statements in which corporate interest groups and tax advisors consistently argued for less extensive anti-tax avoidance measures (see section 5). These views are highlighted by the contrast to all the other groups involved in the consultations over the nine years that generally supported more effective IDLRs or were merely neutral. The evident division of views based on the self-interest of different groups supports the regulatory capture theory that assumes corporations extract rents in regulatory processes.
The initial anti-tax avoidance agenda came from the civil society as well as international organizations, specifically the OECD (Interview 2019; OECD 2013). Additionally, the role of investigative journalism was highlighted in several data leaks that made the public more aware of the consequences of tax avoidance (Finnwatch 2016). The initial Finnish IDLR has been called the Mehiläinen law after a tax-avoiding medical corporation inspired the legislation process (Bernitz 2013).
Perhaps surprisingly, my findings show the role of Finnish tax scholars was relatively remote in setting the anti-tax avoidance agenda as well as in the legislative drafting process of the IDLR. In the first consultations in 2009 and 2012, the tax scholars generally supported action to tackle the wide-spread thin capitalization but provided little detailed commentary on how this should be done. I.e. tax scholars did not propose amendments to the IDLR draft. Consequently, their influence in the details of the final IDLR was nominal in contrast to the significant effect by the corporate interest groups and tax advisors. The lack of policy-relevant profit shifting research in Finland by the early 2010s might have been a major factor behind this. However, the tax scholars also failed to refer to empirical studies conducted abroad. A reason for this might be that the Finnish tax scholars have mostly relied on legaldogmatic research method that merely describes the application of current legal framework, but generally does not assess its implications and alternatives from the social perspective (Smits 2017).
The epistemic nature of tax specialist communities might also have restricted tax scholars from expressing proposals that were in conflict with the corporate interest (Anesa et al. 2019). The scholars and officials are closely connected to the corporate tax advisors in these commu-nities, as many of them also work for the private sector at some stage of their careers. Some tax scholars are also regularly involved in the corporate tax advisory business (Raitasuo 2018 and. Previous research has shown that this epistemic nature of tax specialist communities has an impact on how tax officials and scholars take part in political processes (see section 2.2). Raitasuo (2019) previously showed in his case study that the scholars' involvement in the tax advisory business could "introduce bias into legal interpretation for purposes of advantage to private interest groups". My findings raise the question whether this phenomenon that Koskenniemi (2006) has named structural bias could as well extend to tax policy drafting processes as scholars' involvement in tax advisory business might result in scholarship functioning as a vehicle of private interest. Specifically, in the later IDLR consultations of 2016 and 2018, a couple of the tax scholars were notably critical to more extensive anti-tax avoidance regulation without citing previous literature such as recent Finnish empirical evidence that was much in conflict with their critic (Harju, Kauppinen and Ropponen 2017). Perhaps further research could address this question more thoroughly.
The role of civil society has been notable in setting the anti-tax avoidance policy agenda in the 2000s. The human rights and other civil society organizations could be considered to represent the public interest as public finances and tax justice are necessary to advance human rights (OECD 2011; Knuutinen and Pietiläinen 2017; Quentin 2019; Reisch 2020). However, my case studies show that the role of NGOs in the actual drafting process of anti-tax avoidance legislation has been limited. In the initial consultation in 2009, the human rights-based NGOs were not represented at all. Since then, the role of civil society has increased, but even in the 2018 consultation round, the statements representing corporate interests outnumbered statements presenting wider public interests such the human right perspective by 19 to 3. This ratio is a picture of the superiority of private resources in tax lobbying that produces imbalances in the interests in the legislative drafting processes that could have an impact on the outcome (Shapiro 2012). The resource superiority is also highlighted by the fact that in 2018 the three NGOs employed only one full-time tax professional in contrast to tens of tax professionals representing the corporate interest groups and private tax advisors 54 . In the latest consul-tations, the human rights organizations presented technical comments to enhance the IDLR. This is significant as specifically the detailed comments are relevant regarding the possible amendments in the legislative drafting processes. Previously, the lack of relevant expertise has prevented civil society from having a greater role in these processes related to taxation, which was also brought out in the interviews (2020). Specifically, in the Finnish context, the minuscule role of trade unions in my case study is notable. Previous studies have highlighted the significant role of both employer and employee organizations in the Finnish political culture (Vesa, Kantola, and Binderkrantz 2018). However, according to my findings it seems that corporate tax policy drafting is dominated by the corporate interest with the labor perspective mostly absent. This could represent a shift in the Finnish political power structure, as a recent study arrived at a similar conclusion (Hirvola et al. 2021).
In my case studies, early involvement was essential for successful lobbying as all the significant amendments in the legislation were made in the early phase of regulatory drafting before the government bill was submitted to the parliament. This is in line with earlier lobbying research that highlights the importance of timing (Hirvola et al. 2021). In the Finnish political system, the view of the parliament is generally transmitted through the government, whose parties have controlled the majority in the parliament for several decades. It is likely that the composition of the government has some impact on regulatory captures, but it is beyond the scope of this article to discuss this in more detail.
The global wealth chains theory has particularly emphasized the role of high technicality of international tax regulation that allows tax professionals to affect where the wealth is accrued (see section 2.2). My findings supplement this picture by showing specifically the technical comments regarding EER influenced the legislative drafting process of the Finnish IDLR. In line with the previous research, I also show corporate tax advisors facilitated a corrosive regulatory capture that protected the regime from political intervention (Forstater and Christensen 2017). The technicality of anti-tax avoidance legislation seems to have resulted in the absence of other disciplines' scholars from the legislative processes. The economists are often involved in tax policy discussions but their role in the consultations in my case study was nominal (see also Ylönen et al. 2020). On the other hand, this shows that transparency and simplicity of regulation is necessary to support public interest in a democracy. Technicality and the epistemic communities of tax professionals have been effective in keeping public debate on anti-tax avoidance regulation at bay. The consequences of the technicality of the IDLR could have been emphasized by the lack of policy-relevant research regarding the rule. Further research could address the possible impact by using a similar method to conduct a study in a context of a more researched rule that facilitates tax avoidance.
The domination of corporate lobbying and the limited role of tax scholars in the IDLR legislative drafting process emphasized the role of officials. The role of the civil servants in the Ministry of Finance and the Finnish Tax Administration was decisive in drafting the initial Finnish IDLR in 2012 (Interview 2019 and 2020). This highlights that policy goals of democratically elected politicians might not be accomplished if the civil servants do not spontaneously identify possible legislative shortcomings and solutions to them. This is relevant specifically in the context of tax avoidance that requires technical knowledge on tax law and accounting.
The case study in section 5 also represents some lessflattering aspects of corporate lobbying. Some corporate interest groups continuously diminished the significance of profit shifting and claimed the phenomenon is nonexistent even though this was in conflict with empirical research and officials' observations (see section 5.1). They also argued that the EER does not facilitate tax avoidance even though there was empirical proof about such tax avoidance structures (Ministry of Finance 2018; Finnwatch 2018 and 2020a). The corporate interest groups argued against IDLR mostly based on a claim that extensive antitax avoidance rules harm investment and competitiveness of Finnish enterprises. However, they did not support the argument with empirical evidence or research. This shows that limited knowledge on the impact of tax avoidance structures allows unfounded argumentation. This emphasized the need for a comprehensive impact assessment of legislation (Interview 2020). It could also be concluded that tackling regulatory capture requires acknowledging interests of lobbyist groups as well as basing legislative decisions on empirical research and other verifiable evidence (Anesa et al. 2019;Shapiro 2012).

Tackling the regulatory capture
In contrast to ordinary lobbying, the regulatory capture is harmful to welfare by definition (see section 2.1). My case study is an analysis of the context and mechanisms of regulatory capture that allows assessing how it could be tackled. The case study highlights specifically two elements that facilitated the regulatory capture. First, the high technicality of thin capitalization and anti-tax avoidance legis-lation. Second, the significant imbalance of resources between parties representing the corporate and public interest. I conclude my article by briefly discussing remedies of these elements.
To be realistic, it must be taken for granted that the international corporate tax regime will always be complex. It is also obvious that the corporations will always have the need and resources to employ great numbers of tax advisors to advance their interest. Therefore, it is necessary to provide adequate counterforce to represent the public interest (Levine and Forrence 1990). Otherwise, the lack of recognition of different perspectives and groups undermines social justice (Fraser 1998).
The governments could invest in independent research organizations whose objective should be to preserve the public interest such as tackling harmful tax practices. The Norwegian Centre for Taxation is an example of such a publicly funded institution. It is also necessary that the tax authorities and other public entities share their policy-relevant data on tax avoidance. Another option would be to enhance the independence of scholars in universities and other research institutions. A first step could be the transparency of scholars' economic interests and secondary occupations. Decreasing the dependency of universities from private project funding could improve academic independence and thus the validity and reliability of their research (Eisenberg 1993).
Additionally, the autonomous NGOs representing e.g. human rights or other groups such as the trade unions could serve the public interest in certain tax matters. In this sense, the government could provide a favorable operational environment for the civil society. The civil society organizations and trade unions could pay attention to hiring enough technical tax expertise to enhance their influence.