Business as usual? How the pharmaceutical industry protected its long-term interests during and after eurozone bailouts (2011–2020)

ABSTRACT In this article, we start from the assumption that the pharmaceutical industry accumulates material, instrumental, ideational and institutional power. Considering this, we expect that it would be very difficult for governments to reduce the rents of pharmaceutical companies in the long-term, even when the former are externally constrained to reduce their spending. We test our argument by carrying out a qualitative analysis of a paradigmatic case study, the Portuguese economic adjustment programme (2011–2014). We demonstrate that the conditionality associated with the international bailout was welcomed by Portuguese reformist ministers as an opportunity to decrease public pharmaceutical spending. However, the representatives of the industry took several initiatives to make sure that the decrease of its profits would be limited in scope and time. As a result, as soon as the bailout terminated, the industry was able to regain its rents.


Introduction
A few years ago, an OECD report calculated that 20% of resources in the health sector are 'pure waste', i.e. 'the equivalent of burning money' (OECD 2017).One important source of this waste is related to pharmaceuticals.Often governments and/or public hospitals reimburse drugs that are inefficient, not needed, or for which a cheaper alternative exists; they purchase a quantity of medicines or vaccines which is larger than the demand; or they pay too much for a given medicine.An indication of the latter is the variation of prices for the same drug that exists within and across countries (OECD 2017).Spanish hospitals, for example, pay on average ten times more than their Swedish counterparts for the same generic anticancer medicine (Godman et al. 2019).
What is 'waste' for governments constitutes excessive profit ('rents') for othersnotably the pharmaceutical industry.While it would be clearly in the public interest to reduce those rents (profits above a normal return on investment, including the R&D effort), the government has few incentives to do so, given that the benefits of such a reform would be diffuse and the costs concentrated in a powerful industry.Governments' willingness and capacity to decrease the rents of the pharmaceutical industry, however, might change under new circumstances.
For example, when a country is bailed out, governments must commit to a series of policies and objectives, allegedly to make their economies more performant in the long term and cut public sector spending in the short term.As the provision of loan instalments are conditional to the implementation of those conditions, governments' incentives to reform grow, while the industry is less able to protect the status quo (Moury et al. 2021b;Vreeland 2005).We also know, however, that general interest reforms hurting powerful actors are often reversed a few years after their adoption (Patashnik 2014).
In this article we look at the willingness and capacity of governmental actors to decrease the (unjustified) rents of the pharmaceutical industry during a bailout, and at what happens when the reform programme associated with the bailout terminates.We argue that, bailouts enable Government executives to reduce public spending on pharmaceuticals, and given the power of the pharmaceutical industry, those policies are not likely to be permanent and the political status quo will remain largely unchanged in the long term.
To illustrate our argument, we focus on a paradigmatic case: how Portugal managed to decrease the rents of the pharmaceutical industry during its economic adjustment programme (2011)(2012)(2013)(2014); but how this policy was reversed after the termination of the programme when yield decreased (starting in 2015).
Theoretical framework: the pharmaceutical industry, the state and the Troika Our theoretical premise is that the pharmaceutical industry is a very powerful actor.To make that assumption, we base our examination on the typology made in the literature regarding different sources of business power: structural, instrumental, institutional and ideational, and show how the pharmaceutical industry power scores high for each of those types.
First, business power might be structural, that is associated with the groups' position in the economy.State dependence on business financial liquidity, investment, employment, or tax revenue predisposes governments to adopt policies that benefit firms, without the latter having to do anything actively (Lindblom 1977;Woll 2019).The pharmaceutical industry is no exception, as it is a key employer (Mossialos and Oliver 2005, 292) and invests considerably (Permanand and Mossialos 2005, 688) in a crucial economic sector (Baena del Alcazar 2002).
Business power can also be instrumental.This form of power arises from firms' lobbying and campaign donation capacity, and hence is positively correlated with the latter financial, organizational or human resources, as well as access to decision-makers (Fuchs 2005).Many authors have argued that the instrumental power of firms has grown exponentially with the increase in number, size and lobbying resources of transnational corporations and the growing reliance of policy makers on business donations for their rising campaign costs (Fuchs and Lederer 2007;Wilks 2013;Woll 2019).
In the pharmaceutical sector, the costs of pharmaceuticals are diffuse to citizens (often paid by the Government) and the benefits (profits) are concentrated, and that is why it is rational for the industry to invest in lobbying (Permanand and Mossialos 2005).They also have the financial means to do so, as the sector is growing and increasingly profitable.Moreover, the industry has increasingly been concentrating, through mergers and acquisitions by major companies.This allows the industry to hire professional lobbyists, to donate to campaigns and/or to hire past decision-makers (revolving doors) to gain access to government officials (Ladi, Moury, and Stolfi 2022).
In the United States, for example, the largest lobby is the one of the pharmaceutical industry, and it gives copiously to electoral campaigns (Angell 2004).Similarly, in the European Union, the health lobby is dominated by the industry (Judge and Earnshaw 2002), which employs at least 290 lobbyists in Brussels and spends at least 36 million Euros a year to lobby the EU (Corporate Europe 2021).A well-known example of revolving doors is the former director of the European Medicine Agency, Thomas Longrenn, who was hired by the pharmaceutical industry three days after leaving the agency (Goldacre 2014, 125).
The pharmaceutical industry also influences actors outside of the government.For example, in 2005 the British House of Commons health committee found that the industry also 'buys influence over doctors, charities, patient groups, researchers and journalists' (House of Commons 2005, 855).In the United States Randall (2007) provided considerable evidence that the activities of the pharmaceutical industry included covert influence over citizens testifying at congressional hearings and costly gifts to those in the medical profession.In 2009, a report by the European Commission signalled the use by the innovative pharmaceutical industry to quasi-systematically go to Court to challenge the granting of a patentthereby delaying its introduction during the ruling (European Commission 2009).
A third important source of business power is institutional (Busemeyer and Thelen 2020).It exists when the Government delegates crucial public tasks to private actors, and consequently depends on firms' existence (and hence profitability).Here again, the power of the pharmaceutical firms is undoubtedly strong: the state needs the industry to produce medicines and vaccines.Taking decisions that endangers the profitability and existence of the industry runs the risk that crucial goods are no longer produced.It is a credible threat: some countries, particularly low-and middle-income ones, face problems of availability because some 'off-patents' cheap medicines are not produced anymore (Vogler et al. 2016).Moreover, innovative pharmaceuticals are introduced later in markets in which price regulation is stronger (Danzon, Wang, and Wang 2005).
Finally, ideational power relates to business capacity to influence other actors' normative and cognitive beliefs (Carstensen and Schmidt 2016;Lukes 2005).From such a perspective, firm power is reflected in discourse and cultural values, and is fuelled by their capacity to frame the problem, make compelling arguments (Fuchs 2007), and present 'scientific evidence' to their advantage (Miller and Harkins 2010).Because it provides public goods that can save lives, the pharmaceutical industry has a potentially strong capacity and legitimacy to use ideas for their purposes.In that line, scholars have shown how 'big pharma' in the United States have constructed narratives that generics are dangerous (Facchinetti and Dickson 1982) and that innovation will decrease if governments seek to lower drug prices (Kapczynski 2022;Randall 2007).A strategy often used by the industry is to vehiculate its messages by funding patient groups (Khabsa et al. 2020;Permanand and Mossialos 2005).Such strategies help the industry to 'capture policy', e.g. to secretly create or fund organizations to influence the public health discourse in order to advance their interests (Miller and Harkins 2010).
Business power also varies according to the circumstances.Notably, research has shown that business power decreases when the industry lacks ideological affinity with policy-makers (Klüver, Braun, and Beyers 2015) or with the economic elite (Gilens 2012); but also when their claims are perceived as illegitimate (Keller 2018;Ladi, Moury, and Stolfi 2022), when an issue is salient for public opinion (Culpepper 2010), and when firms are divided (Maloney, Jordan, and McLaughlin 1994).Here again, the balance is weighted in favour of the pharmaceutical industry.As discussed above, the fact that the industry produces medicines or vaccines that can save lives allows them to speak with a certain legitimacy, and few decision-makers would contest the principle that such production is necessary.
Moreover, the issue of the rents of the pharmaceutical industry is generally not most important for the general publicespecially when most medicines are co-paid by the Government or by private insurance.The fact that as a rule pharmaceutical studies and agreements with public authorities are confidential also keeps rents outside of public scrutiny, thereby leaving room for greater business influence (Carone, Schwierz, and Xavier 2012;Fierlbeck, Graham, and Herder 2021;Persson and Jönsson 2016).
Finally, the pharmaceutical industry is often highly united (Young and Pagliari 2017).While in the 1990s the research and development (R&D) companies had interests that differed from those producing generic drugs, and were thus divided, things have changed over the last 15 years.As (Roemer-Mahler 2013) shows, R&D firms have been expanding into the market for generic drugs to profit from the emerging new markets.They have done so by buying generics companies, building generic businesses, and partnering with generics companies (Mazzoleni and Nelson 1998).As a result, business unity has grown.
Given this strength, governments are little willing or able to hurt the interests of the pharmaceutical industry, allowing for the existence and resilience of excessive profits or rents.Journalists and civil society often report on the profit and influence of the industry.To start with, pharmaceutical firms are not required to disclose their research and development and production costs, how much public money they receive to help develop medicines, or what prices they charge in different countriesallowing them to charge higher prices than would occur in a free market (Mancini 2020).
The Financial Times, for example, reports the case of Lipitor, an anti-cholesterol drug, which generates for the pharmaceutical firm Pfizer a profit of 3.3 million USD a day (Jack 2011).A group of journalists recently reported how the material and instrumental power of the industry prevented the adoption of a proposal made at the World Health Organization that was meant to make Covid vaccines widely available in poorer countries (Furlong, Aarup, and Horti 2022).
More generally, a study by the General Accountability Office showed that the average profitability of pharmaceutical firms is twice that of the world's Top 500 companies (Florio and Gamba 2021).Prices of medicines, moreover, are not related to the costs of its investment in R&D (Wouters et al. 2022).The latter, especially in Europe, is often pursued by 'tax-payer funded scientists' during the early stage, yet all benefits are reaped by the industry (Florio and Gamba 2021;Sarpatwari and Kesselheim 2021).We also know that patents last much longer than justified by the return-to-investment (Chu 2008) (20 years on average, Pereira 2018); and that prices and profits are higher than what they would be in a free-market economy (Baker 2021).
Governments, however, are not helpless, they have at their disposal a series of opportunities and strategies to defend the public interest against powerful (pharmaceutical) industry.For example, politicians may divide the industry to weaken its resistance to change (the 'divide and conquer' strategy) (Klein 1995).Various studies also suggest that creating an independent commission, engaging public debate, or presenting innovative reforms during the electoral campaign may be successful (Cardoso et al. 2021;König and Wenzelburger 2014;Tompson and Price 2009;Williamson 1994, 601).
Another strategy to tackle the general interest reforms that hurt firms' interest is delegating competences to the international level (i.e.Dyson and Featherstone 1996;Grande 1996;Moravcsik 1994).International negotiations usually take place behind closed doors.This grants to executives privileged information about the constraints they face (Moravcsik 1994) and hence the possibility to use 'false dilemmas' (Gago and Moury 2017) that is, a choice between two options (for example the policy change or bankruptcy) despite the existence of other solutions.
Related to this, countries that asked for a bailout are subject to explicit conditionality as the progressive payments of the loans granted to them were conditional upon a series of reforms detailed in the Memorandum of Understanding (MoU), signed by the Portuguese Government with the European Commission, the European Central Bank and the International Monetary Fund in May 2011.As demonstrated recently (Moury et al. 2021a), those MoU are not imposed 'from above'.Executives have some leeway in the design and implementation of the MoU.This allows governments to pass the reforms hurting business interests that they would not have been willing or able to pass in normal times.During a bailout, indeed, opponents to policy changes take the risks of losing the loan under consideration when choosing their strategies (Dreher, Stum, and Vreeland 2015;Vreeland 2005).Bailouts have thus proven to reduce the traditional veto power of actors such as civil servants (Hick 2018;MacCarthaigh and Hardiman 2020), trade unions (Cioffi and Dubin 2016;Dukelow 2015;Moury and Standring 2017), parliaments, opposition parties (Lütz, Hilgers, and Schneider 2019) oras it interests us hereinpowerful businesses.
Based on this literature we posit that: H1.During a bailout, governments are willing and able to pursue general-interest reforms that reduce the rents of powerful businesses such as the pharmaceutical industry.
However, if external circumstances strengthen executives during a bailout vis-à-vis the pharmaceutical industry, what happens once those external circumstances change?In that context, the conditions that allowed reformist ministers to push for their favourite policies are no longer met; and the current executive, even if there is continuity in office, might be tempted in this new context to surrender to firms' pressure (Moury and Afonso 2019).Researchers looking at reversals after a bailout have indeed shown that if most of the policies adopted during a bailout are kept, governments would POLITICAL RESEARCH EXCHANGE revert policies when reversals would bring concentrated benefits and diffuse costs (Branco et al. 2019;Moury, Cardoso, and Gago 2019;Rickard and Caraway 2019).
Related to this, general-interest reforms that hurt firms' interests are often later dismantled (Patashnik 2014), as firms use all their power to regain their rents once the public attention on the topic has shifted and the issue becomes less salient to the public.According to the same research, reversals of general interest reforms are less likely in the following circumstances: (1) when the reform created influential, well organized actors who benefited from the change and would mobilize against a reversal (see also Mandelkern and Koreh 2018), (2) when the firm's reputation had been tarnished in the reform battle or (3) when the reformists lock in the reform (e.g. by creating sunk costs or by introducing the need for a super-majority to reverse the reform).
In the case of the pharmaceutical industry, the benefits of a reduction of the rents are diffused amongst taxpayers, and hence it is not likely that they would mobilize if those rents were progressively re-introduced.Similarly, and because the issue is not salient for the general public, the industry reputation is not likely to be strongly endangered during the bailout.Consequently, the only possibility for governments that are concerned about the resilience of their reform is to lock it in by creating institutional mechanisms that make the change permanent or difficult to reverse.
The pharmaceutical industry, however, does not need to read political science literature to anticipate that such lock-in mechanisms would be prejudicial for their interests.Consequently, we expect it to bring all its weight to avoid them.We expect those efforts to be successful, because of what the IMF has called 'reform fatigue' (Bowen et al. 2016).The concept relates to the fact that policy makers who have to implement many reforms need to make a trade-off between the political costs of implementing the reforms and the perceived benefit.Given the power of the pharmaceutical industry (and hence the associated costs of hurting its interests), and that the bailout (and therefore the domestic empowerment of the executives) is temporary, we expect that governments will not be willing to fight the industry to lock-in the decrease of their rents.As a consequence, the rents of the industry would be regained after the bailout.
We thus posit: H2: The government will not lock-in the reform that reduces the industry rents.
H3: The rents of the pharmaceutical industry will be regained after the bailout.

Methodology
In our view, Portugal in the period just before, during and after the bailout (2011-2020) provides a paradigmatic case study to test our arguments (Flyvbjerg 2006, 232).The Portuguese case is, indeed, a case that can be used to construct an ideal example of incapacity of bailout governments to structurally decrease the rents of a powerful industry.
The Portuguese system is generally quite vulnerable to business power.The legacy of the 'Carnation Revolution' has given parliamentary committees centrality in the policymaking process and made them particularly open to interest groups (Lisi, Oliveira, and Loureiro 2021).Additionally, due to the relatively small size of Portugal, its government has traditionally been business-friendly in order to attract foreign direct investment (da Silva 2016).
Within this 'friendly' environment, the pharmaceutical industry is particularly powerful.First, it is very well organized.The major pharmaceutical companies operating in Portugal are affiliated with APIFARMA, which coordinates their public positions regarding pharmaceutical policies.Additionally, the incentives to lobby in Portugal are more than proportional to the size of the market, because Portuguese medicines prices are part of the basket of reference countries that contribute to fixing the price of medicines in larger markets, such as Turkey and Brazil (Informant 6 2017;Kanavos et al. 2020).
Moreover, the system of revolving doors between the health ministry, the Parliament's health commission, public and private hospitals, and the pharmaceutical industry has paved the way for conflicts of interests and traffic of influence in Portugal (Ladi, Moury, and Stolfi 2022).Finally, the power of the pharmaceutical industry is magnified by the fact that it serves as an important creditor to the Portuguese state, as pharmaceutical companies are the major takers of public hospitals' debt (Informant 3 2017; see also Pereira 2018, chapter 7).Consequently, the Government depends on the industry to indirectly fund its hospitals (through existence of arrears), giving the latter institutional power.Hospital managers are allied with the industry to protect that system, which allows the former to circumvent the Government's chronic underinvestment of hospitals and the second to sell expensive pharmaceuticals with few limitations (Campos 2017 Not surprisingly, then, in Portugal the profitability of the pharmaceutical industry is well above the average (Bank of Portugal 2016, 25, 2017).Compared to other countries of the European Union, a larger share of the health spending is dedicated to pharmaceuticals.For example, before the bailout, in 2009 22.3% of public health spending in Portugal was dedicated to pharmaceuticals and other medical non-durables, while the EU average was 13.9%.As a percentage of GDP, in 2009 1.25% was spent by the Government on out-patient pharmaceuticals while the EU average was 0.79%.
Validity is ensured by triangulating different types of data.First, we rely on a qualitative analysis of official documents from public and private entities from the Ministry of Health, the European Commission, INFARMED (the Portuguese National Authority of Medicines and Health Products) and APIFARMA (the Portuguese Pharmaceutical Industry Association).Second, we examine two books by investigative journalists (Dinis and Coelho 2012;Pires and Martins 2015) regarding the bailout period, and refer to the detailed description of the events that occurred in Eurozone bailed-out countries by Moury et al. (2021a).Third, the first author of this article conducted face-to-face interviews with major health policy making actors (from the Health Ministry and the pharmaceutical industry) who had responsibilities during the period.In total, 35 individuals were contacted (30 of whom answered positively).After 30 interviews, data saturation was reached (Buthe et al. 2015, 10).
Interviews were conducted from November 2017 to May 2018.The average duration was 87 min, and some interviews were recorded with the authorization of the informant.Promising anonymity, we asked interviewees open questions such as: 1. How were agreements negotiated before the bailout between APIFARMA and the government?2. What changed with the bailout regarding public expenditures for pharmaceuticals?3. How did the negotiations take place between the governments and APIFARMA?4. Which were the positions of each negotiating actor?What happened at the end? 5. Were the new provisions efficient in containing public expenditures for pharmaceuticals in the short and long term?Why so?
In case of divergence in the information given, we contacted a third source (official documents or another informant).

Setting the scene: pharmaceutical prices in Portugal
In Portugal, as in many other countries, the retail price of each medicine is set using a system of international reference pricesa method in which the retail price of a medicine is calculated in comparison to the average price charged for that drug in other EU countries.In 2009, for example, Italy, France, Greece and Spain constituted the reference for Portugal (Pereira 2018).The reference countries are changed every year by the Portuguese Government.
On top of this, since 1997 APIFARMA and the Ministry of health have celebrated agreements every year, establishing ceilings for the overall expenses, the total price for each medicine sold to the national health system (NHS) (which is kept confidential), and the pharmaceutical expenditures of hospitals (Informant 11 2017; Informant 13 2018; Informant 14 2018).According to those agreements, all values that exceeded a certain amount were returned by pharmaceutical companies directly to hospitals in credit notes (Fernandes et al. 2015).
This type of agreement is common in Europe and is generally preferred by the pharmaceutical industry to the negotiations over retail prices, which affects the international reference pricing system, thereby reducing the profits in the other (larger) markets where the companies operate (Carone, Schwierz, and Xavier 2012).Moreover, it is an alternative to more permanent price reductions of pharmaceutical products.As noted by Carone, Schwierz, and Xavier (2012), such payback policies lower incentives for structural reforms of the health care sector.They also reduce the transparency in the actual prices of pharmaceuticals and consequently the effectiveness of external reference pricing.).These were motivated by budgetary reasons, when the government needed to make adjustments in the public finances (Pereira 2018).

Before the bailout: strategic negotiations by the industry
In 2006 the same government also appointed a special commission, the 'Commission for the financial sustainability of the National Health Service', composed of healthcare experts with the task of presenting proposals that would ensure the sustainability and efficiency of the healthcare system (Asensio and Popic 2019).With the purpose of controlling the rising price of pharmaceuticals, the Commission proposed a series of measures such as the revision of the existing reference price system, the encouragement and limitation of the retail price of generics and the extension of the medicines that are not reimbursed and/or for which prior authorization is required.
In 2008 the budgetary situation worsened dramatically following the fall of the global financial service firm Lehman Brothers in the United States.After the short-lived fiscal expansion promoted by the European Union, early in 2010 the EU started to worry about the credibility of national bonds and the European Council urged Portugal to rapidly engage in policies aimed at medium-term fiscal consolidation (Moury et al. 2021b)including the reduction of the price of medicines (Informant 15 2018).In closed-door conversations between the government and APIFARMA, the latter suggested a 6% overall discount to the bill paid by the (Government-run) National Health Service, on the condition of no changing the official price of medicines.The Government agreed.API-FARMA thus took the initiative to propose to the Government a deal that would solve the governments' financial problems in the short term, but without inducing long-term changes that would damage its own international financial interests (Informant 20 2018).This short episode illustrates well how the government used external constraints to reduce the expenditures in pharmaceutical spending, and also how the industry actively took the initiative to frame the deal so that the losses were only momentary and the least possible.

The drafting of the memorandum of understanding as a window of opportunity
On 9 March 2011, in an attempt to avoid asking for a loan, the Prime Minister, José Sócrates, negotiated what was known as a 'shadow program' with the EU (Dinis and Coelho 2012;Moury and Standring 2017).This plan included several measures to reduce the price of pharmaceuticals that had been proposed by the 2006 Commission including a change in the reference price system.
As this package was not approved by Parliament, Sócrates immediately resigned and the Socialist caretaker government requested the bailout and called in the European Commission, the European Central Bank, and the International Monetary Fund (a group which has become known as the Troika) on 6 April 2011.In the following month the Portuguese authorities reached an agreement with the European Union and the International Monetary Fund on a €78 billion Financial Assistance Programme.The resulting Memorandum of Understanding (MoU) 1 foresaw actions on deficit reduction (notably by cutting spending in education, pensions and health care; lowering financial transfers to regional and local authorities; and reducing public sector employment); on competitiveness (reduction of severance payments, labour reforms and liberalization of public sector); on deleveraging of banks and on reforms to the judicial system.
Regarding the pharmaceutical industry, the MoU included the commitment to 'reduce the public spending on pharmaceuticals to 1.25% of GDP by end 2012 and to about 1% of GDP in 2013'.Given the above-average weight of the public expenditure for pharmaceuticals in Portugal (Informant 2 2017), reducing expenditure in pharmaceuticals was one of the main goals of the Troika (Informant 1 2017; Informant 20 2018; Informant 4 2017).To reach this objective, several measures were included such as setting a maximum price on the first generic introduced in the market to 60% of the branded product with similar active substance, revising the existing reference-pricing system, monitoring of prescriptions by doctors, and imposing a maximum profit margin for pharmacies.
In line with our first hypothesis, informants from both government and the troika testified of a 'common vision' with the troika regarding the need to reduce pharmaceutical expenditures.They also reported that while the macro-objectives of pharmaceutical spending reduction had been imposed externally, the concrete measures to fulfil them had been proposed by the Portuguese ministers (Informant 20 2018;Informant 28 2018;Informant 29 2018;Moury et al. 2021a).In fact, as noted by Asensio and Popic (2019), all the measures in the MoU regarding health expenditures were the ones already recommended by the health commission established in 2006.
A good example of this congruence of views is the obligation inserted in the MoU for physicians to use the name of the active substance (and not the name of the brand) when they make a prescription.This obligation already existed in hospitals but not in the realm of community care.Both the European Commission (European Commission 2009, 61;Informant 20 2018) and the members of governments had been campaigning for this change.Before the bailout, however, medical doctors and the industry successfully allied to oppose the reform (Informant 6 2017; Informant 18 2018).The MoU thus offered a window of opportunity to introduce a long-wished-for change (Informant 6 2017).
The following quote from a minister (Moury et al. 2021a, 95) illustrates the logic regarding pharmaceuticals: Portugal had the highest per capita spending on medicines in Europesomething wrong in a country with a relatively low GDP (…) prescriptions for generic drugs were minimal in Portugal, we all knew something had to be done; but we could not pass reforms, doctors opposed them, the industry was blocking them … We asked to have this measure inserted in the memorandum.
In this second section, we thus find support for our first hypothesis, whereby the bailout opens a window of opportunity for reformist ministers in the health sector.
Implementation of the memorandum: the essential 'machinery' to protect the status quo APIFARMA quickly manifested its concerns about the reduction of the price of medicines to policymakers.It insisted, first, that the new system of reference countries for price determination should not push prices down too far in Portugal, mainly for the sake of the larger markets such as Brazil in which Portugal served as a reference.It also asked for guarantees against parallel exports, which occur when a stockist buys a drug at a lower price in one country and sells it at a higher price in another (Pereira 2018(Pereira , 2019;;Treacy and Watson-Doig 2016).The industry finally expressed its concerns about the price of innovative medicines sold to hospitals, which contributed massively to its profits (Pereira 2018(Pereira , 2019)).
In the June 2011 elections, the right-wing PSD/CDS-PP coalition obtained an absolute majority of seats and immediately started to implement the MoU.One of the first measures of the government was to calculate and re-negotiate hospitals' debts to the pharmaceutical industry (Pereira 2018, 132).For this purpose, representatives to the troika were asked to speak directly to the pharmaceutical industry (Informant 19 2018).
Another strategy of the government to negotiate those debts was to 'divide-andconquer', by committing to reimburse with priority the firms accepting a greater reduction of their debt (Informant 1 2017;Informant 12 2018;Informant 19 2018;Informant 30 2018).Another decision of the government was to change the basket of countries associated with the international reference pricing system.In November 2011 a decree law replaced France and Greece with Slovenia (APIFARMA 2018).
Those changes, however, would not be sufficient to reach the commitments made with the international lenders, and the government announced that if no alternative deal was agreed with APIFARMA, it would unilaterally reduce the price of all drugs by 12% (Informant 15 2018; Protocolo Com a Apifarma Que Define Redução Na Despesa Do SNS Com Medicamentos Hospitalares e de Ambulatório 2012).This was a credible threat given the existing external constraints (Informant 10 2017; Informant 13 2018; Informant 14 2018; Informant 15 2018).In response, and to sway public opinion to its side, APIFARMA made the public claim in Portugal that the Government was asking it (APIFARMA) to make many sacrifices and endangering the health of the population (Informant 10 2017; Informant 13 2018; Informant 14 2018; Informant 15 2018).Industry representatives also started on a series of visits to Brussels to convince the Portuguese permanent representation and the Portuguese Members of the European Parliament to intervene on their behalf with the European Commission.They also met with the Directors Generals and the commissioners themselves and obtained that troika representatives would talk to them during their visits to Lisbon (Informant 22 2018).
Given the issue at stake, APIFARMA also pressured the European Federation of Pharmaceutical Industry and companies (EFPIA) to defend their interests (Informant 9 2017; Informant 30 2018); and indeed lobbyists from EFPIA met with the international lenders and with Portuguese negotiators (Informant 9 2017; Informant 11 2017; Informant 22 2018).In May 2012 EFPIA also declared publicly that 'the industry could not be an easy target of measures of austerity' (APIFARMA 2012a).They warned that changes in the reference pricing would threaten the supply of medicines (APIFARMA 2012b; thepharmaletter 2012).Also, the federation strategically used the (empirically correct) idea according to which austerity and budgetary cuts would damage (people's) health (EFPIA 2013), citing scientific papers demonstrating the deterioration of people's health in countries where health spending had been most cut.
On 16 May 2012 an agreement was reached between the pharmaceutical industry and the Portuguese government.The reference system changed again, returning France to the basket (instead of Italy) and replacing Slovenia with Slovakiathus potentially raising the price of medicines to earlier (higher) levels.To compensate, and in accordance with past practices, this agreement included a payback system to implement the international commitments regarding the reduction of expenditures (both hospitals' and ambulatory) in line with MoU objectives (Informant 15 2018;Informant 26 2018;Informant 30 2018;Pereira 2018).According to our informants, this type of system allowed the pharmaceutical industry to protect the status quo regarding pharmaceutical prices in the long term (Informant 15 2018;Informant 26 2018;Informant 30 2018).In exchange, the government committed to quickly pay all hospital debts to the industry and to take faster decisions regarding the reimbursement of innovative medicines (Informant 16 2018;Informant 19 2018).The deal was to be monitored by a special commission (Acordo Bienal 2012/2013 Entre o Estado Português e a Indústria Farmacêutica, 2012, 11).
The implementation of the deal was controversial (Relator Fernando Bento 2015).A first issue concerned the total amount of money due from hospitals to the industry, as it was calculated differently by the Government and APIFARMA (the former proposing a lower amount in order to reduce the payment).The Government's figure was finally accepted after the Ministry of Health threatened to introduce an administrative reduction of pharmaceutical prices in the event of prolonged contestation (Informant 19 2018).A second issue regarded the contribution of the companies not represented in APIFARMA.Administrative price reductions were first applied to the non-signatories of the agreement (Relator Fernando Bento 2015) and later an extraordinary tax was applied to the latter (Orçamento Do Estado Para 2015, 2014).
The agreement had a positive effect on pharmaceutical expenditures, which were reduced by 668 million euros between 2011 and 2012 (Morais Nunes, Cunha Ferreira, and Campos Fernandes 2019).This amounted to 1.29% of GDP and thus fell short of the planned 1.25% (Pereira 2018, 88).The objective for 2013, however, was more ambitious ('about 1%').To keep control of the process, APIFARMA proposed a payback that would amount to a reduction of 1.12% for 2013a proposal that was initially rejected by the then secretary of state for health for not respecting the agreed targets (Relator Fernando Bento 2015).Following this, representatives of APIFARMA met with the Portuguese president of the European Commission, José Manuel Barroso, (Informant 13 2018; Informant 14 2018; Informant 22 2018) with whom it was agreed (informally) that the 1% goal of public spending on pharmaceuticals was 'flexible' (Informant 1 2017; Informant 4 2017; Informant 13 2018; Informant 14 2018; Pereira 2018, 104).After this, the government lacked the possibility to use the external constraints to its advantage, and it was finally agreed that 1.25% of public expenditure would be acceptable for both sides (Aditamento de 2013 Ao Acordo Bienal 2012/2013 Entre o Estado Português e a Indústria Farmacêutica, 2013; Informant 1 2017; Informant 13 2018; Informant 14 2018).According to our informants, an important reason for finally accepting this proposal was the fear that the Portuguese market would no longer be profitable for the industry and that the industry stop selling medicines in the country (Informant 30 2018).
In 2013 the electronic prescription of medicines (using the substance name) by physicians working either in the public or the private sectors was made compulsory, and a monitoring system with periodic detailed reports per prescribing doctor was created.This report contains information about each doctor's prescription pattern (Nunes et al. 2020).
One year later, in 2014, the agreement between the Government and APIFARMA was renewed, maintaining the principal elements of the first agreement (Acordo de 2014 Entre o Estado Português e a Indústria Farmacêutica, 2014).This time it was clarified that the contribution would be made in proportion to the market share of each signatory company.The contribution of APIFARMA members and adherents to the agreement had a target of €120 million (Pereira 2018).
Such agreements had clear effects on pharmaceutical expenditure, as reductions in public health spending for the period 2010-2014 were estimated to be worth almost €2 billion (41% of which are linked to reductions in medicine costs).The average price of drugs supplied in ambulatory services fell from €15.19 to €12.21 between 2011 and 2015.Furthermore, generic drugs increased their market share from 36.2 to 47.0% in the same period (Morais Nunes, Cunha Ferreira, and Campos Fernandes 2019).As a matter of fact, Portugal was the country with the greatest decrease in retail pharmaceutical spending in the period 2009-2013 compared to the average of prices in the period 2005-2009(Belloni, Morgan, and Paris 2016).It must be noted, however, that the target of 'about 1%' was not reached; by the end of 2014 1.25% of GDP was still spent by the Government on pharmaceuticals.
The period of the bailout thus supports our first hypothesis according to which governments exploited the bailout to reduce the rents of the pharmaceutical industry.Those rents, as we have seen, were indeed substantially cut during the period.The fact that the government could credibly threaten the industry with a unilateral decision to reduce pharmaceutical prices, which would have had an impact on profits in other countries, also illustrates well the mechanisms at play.
However, we have also seen that thanks to the industry lobbies at both the national and international levels, those rents had not fallen as much as was planned in the MoU.Moreover, and in line with our second hypothesis, the pharmaceutical industry strategically and successfully proposed to the government solutions to decrease spending (the already-used 'paybacks') without making this reduction permanent and structural.

The post-bailout agreements: inertia is back
In June 2014, Portugal exited the economic adjustment programme.Until the repayment of 75% of the financial assistance, Portugal was put under post-programme surveillance and received six-monthly visits and recommendations from the EC and the IMF.Moreover, Portugal was still under an EU excessive deficit procedure, which had started in December 2009 when the Council issued a recommendation to bring the deficit to below 3% by 2013.This deadline was postponed twice (first to 2015 and then to 2016).Under this procedure, there is closer scrutiny of the country's actions and failure to meet the goals and deadlines can lead to sanctions.During this period, however, the strongest constraint on ministers came from the markets.Moury et al. (2021a) quoted many government members claiming that their efforts to reduce the public budget deficit was driven more by their desire to reassure investors (and hence keep the yields and associated expenditures low) than by the fear of European sanctions.
During this time the pharmaceutical industry and the Government negotiated two agreements, one for 2015 and another for 2016-2018 (Acordo de 2015 Entre o Estado Português e a Indústria Farmacêutica, 2014; Acordo Trienal 2016-2018 Entre o Estado Português e a Indústria Farmacêutica, 2016), which had the same contents as the one celebrated during the troika (paybacks if spendings exceed €2 billion, timely payment by the Government to public hospitals, and speedy authorization on reimbursement of new pharmaceuticals).However, a 'confidential annex' to the agreement allowed the industry to negotiate the paybacks in case of 'force majeure or changing circumstances' (Anexo II Ao Acordo de 2015 Entre o Estado Português e a Indústria Farmacêutica, 2014, 2-3).This clause clearly illustrates how, after the bailout the industry was able to regain its bargaining power and to introduce ambiguous (and confidential) annexes that would allow for re-negotiations of the paybacks (Pereira 2018).
By the end of 2016 Portugal's economic situation was improving as growth started to pick up and yields to decline.Thanks to this the public deficit came down and Portugal exited the EU excessive deficit procedure in June 2017, moving from the corrective to the preventive arm of the Stability and Growth Pact (Moury et al. 2021b 2017).Another interesting development in the post-troika period pertains to the state authorization for reimbursement of innovative medicines.In 2016, for example, the number of authorizations tripled compared to 2015, and continued to increase thereafter (INFARMED 2016).Those events (several informants told us) were the successful result of the attempt of an industry under pressure by shareholders to regain their lost rents (Informant 6 2017; Informant 17 2018; Informant 18 2018).
As seen in Figure 1, public expenditures with medicines, especially in the ambulatory setting, fell only during the period of the economic adjustment programme, and then rose again as soon as the lenders 'left town'.The reduction of the rents of the pharmaceutical industry was thus very short-lived and was reversed after the termination of the bailout, as we hypothesized.

Concluding remarks
We analyze the relationships between the Portuguese government and the pharmaceutical industry, and the agreements celebrated during the financial crisis and thereafter to contain public expenditures with pharmaceutical products.We first developed our theoretical premise that the pharmaceutical industry is a powerful actor.We show how the industry accumulates material, instrumental, ideational and institutional power.Notably, this industry has the financial means to lobby the Government and non-public actors, and consequently to influence decision-making directly and indirectly through discourses and ideas.The industry also produces a crucial public goodmedicinesthat no country can do without.The growing cohesion of the industry, and the fact that most deals it makes with governments are confidential, boosts its influence in decisionmaking even further.
Following our assumption, we hypothesized that during a bailout the Portuguese government would be both willing and able to exploit external constraints to reduce the rents of the industry.We also expected that the pharmaceutical industry would prevent the locking in of the agreement, and that consequently the rents would be reintroduced after the bailout.
Our research confirmed initial expectations.Our qualitative analysis of this case study showed that the Portuguese government had been trying with limited success to reduce the rents of the industry before the bailout, and that the conditionality associated with the economic adjustment programme was welcomed by reformist ministers as an opportunity to reduce public pharmaceutical spending.However, the representative of the industry, APIFARMA, undertook several initiatives to ensure that the reduction of its profits would be limited in scope and time.For this, it 'moved first' and proposed using the existing systems of paybacks to reach the objectives of the Memorandum of Understanding without substantially reducing the prices of pharmaceuticals.This proposal would solve the accountancy issues of the government, but without introducing structural and permanent changes.
Relying on European associations, the pharmaceutical industry lobbied at both the national and European levels to ensure that its proposals would be accepted and that the fall in its profits would be limited.Both national and European executives were sensitive to this lobbying effort, amongst other reasons because they were afraid of a possible exit of the industry from the Portuguese markets.Moreover, the direct contacts of API-FARMA with the international lenders, in particular with the president of the European Commission, destroyed the possibility of governments to use to their advantage the privileged information about the external constraints they were facing.
As a result, even though the government managed to reduce the rents of the industry during the programme, this success was limited in time and scope.Given that the changes had not been locked in, after the bailout (and especially when the financial pressure on governments relaxed) the industry regained its earlier profits.Our analysis shows how difficult it is for governments to reduce the rents of a powerful industry, even when they are externally constrained.
This research offers several lessons to policymakers willing to rein in the excessive profits of the pharmaceutical industry.A first advice to reformists subjected to conditionality would be to have specific long-term reforms and mechanisms to lock them in specified in the Memorandum of Understanding.
More generally, it might be opportune to discourage confidentiality and payback agreements; for example, by creating international mechanisms so that countries openly share data about those agreements and prices.A good example of such cooperation is the European price medicine database (EURIPID), a voluntary cooperation to produce a data base with information on national prices of medicines; but that should be enlarged to include pay-back agreements.
In a similar way, more transparency would also allow for monitoring the profitability of each medicine, so that governments could determine whether a threat by the industry to stop producing a given pharmaceutical for lack of profitability is credible or not.A good example of this is Italy's Pricing and Reimbursement decree (Criteri e Modalita' Con Cui l'Agenzia Italiana Del Farmaco Determina, Mediante Negoziazione, i Prezzi Dei Farmaci Rimborsati Dal Servizio Sanitario Nazionale 2019), which obliges pharmaceutical firms to disclose the company's R&D costs, the public funding they receive, the prices that they are charging in other countries, and the profits they are making on the drugs.
More ambitious proposals to give a greater role to the public sector in developing and producing medicines (in co-existence with the private sector), or to request pharmaceutical companies to return investments when R&D had been publicly funded (Florio and Gamba 2021;Mazzucato and Li 2021), are, in our view, also worth pursuing.

In 2005 a
Socialist government was elected with a majority of seats and a reformist agenda (Stella et al. 2021).At the outset the government decreed unilaterally administrative reductions in outpatient drug prices in 2005 with Correia de Campos and in 2007 with Ana Jorge as Ministers of Health (Portaria n.o 30-B/2007, de 5 de Janeiro 2007; Portaria n.o 618-A/2005, de 27 de Julho 2005

Figure 1 .
Figure 1.Public expenditure with medicines (M€).Source: Chart by the authors with data from INFARMED.
).At that time external constraints were at the lowest for the period under study herein.While negotiating the 2016/2018 agreement (Acordo Trienal 2016-2018 Entre o Estado Português e a Indústria Farmacêutica, 2016) APIFARMA lobbied to review the basket of reference countries, replacing Slovakia by Italya plea that the Government initially resisted but finally granted (Informant 6 2017; Informant 27 2018; Portaria n.o 290-B/2016 2016).Additionally, given that some medicines were cheaper in Italy than in Slovakia, a 'brake mechanism' was inserted in the agreement at the request of APIFARMA.This mechanism ensured that pharmaceutical prices would not fall by more than 10% (Acordo Trienal 2016-2018 Entre o Estado Português e a Indústria Farmacêutica, 2016; Aditamento de 2017 Ao Acordo Trienal 2016-2018 Entre o Estado Português e a Indústria Farmacêutica,