Macroeconomic factors, liquidity issues and research and development investments: empirical evidence from the EU pharmaceutical industry

In recent decades, the business landscape is influenced by corruption, a pervasive phenomenon, faced by all countries, irrespective of their stage of development. The pharmaceutical industry is recognized as a “fertile ground” for corrupt practices. The paper aims to investigate the impact of corruption, economic freedom, and gross domestic product (GDP) growth on research and development (R&D) investment using a dataset of European Union (EU) pharmaceutical companies from 2011 to 2019. It also investigates the moderating effect of liquidity issues on the relationship between corruption and R&D investment. The study employs a quantitative approach using fixed effects models. Results show that corruption has a negative influence on pharmaceutical firms’ decision to undertake R&D activities, while economic freedom and GDP growth have a positive and significant impact on R&D investment. The findings are especially important given the deleterious effects of corruption and may be useful for both managers and policy-makers.


Introduction
Innovation is crucial for the pharmaceutical industry. At the heart of innovative activities lie research and development (R&D) investments. An unfavourable institutional environment may lead to inappropriate behaviour regarding R&D investment decisions, particularly in the case of pharmaceutical companies which have a high level of control over the R&D activity. In recent decades, R&D behaviour and its determinants have represented a concern of economists. The institutional theory is considered more adequate for explaining the significant trends of R&D expenditures than other firm-level theories (Alam et al., 2019). Institutions are considered a "heart determinant" of investment in general (Hashi & Stojcic, The article is structured as follows. A theoretical discussion regarding the link between R&D investment, corruption, economic freedom and GDP growth rate is presented in the next section. The research methodology is described in Section 3, Section 4 presents the results and robustness checks, while the last section concludes the study.

An institutional theory perspective on R&D investment
According to institutional theory, a country's investments are facilitated by its institutional frameworks as they provide incentives and support and create a stable business environment, decreasing risks and uncertainty. Waarden (2001) stated that institutional quality generates effects on R&D investment. Wang et al. (2015) consider that institutions can have a significant influence on firms' innovative activities. This influence can be realized through regulations, laws, and policies. Laeven (2003) stated that stronger institutional frameworks can improve knowledge accumulation in a country, prompting innovation in general. North (1990) associated institutional quality with economic and political governance, as well as societal interactions. He stated that low-quality institutional settings hinder society to collect productivity gains from innovation, technologies or the specialized division of labour and hence they fail to achieve economic growth. In this regard, their development trajectories can be modified by promoting principles aimed at combating corruption, developing human capital and promoting freedom (economic, political or religious) (Lee & Kim, 2009).
According to Wu et al. (2016), "Better institutional environment may stimulate R&D activity by providing enhanced collaborative capacity to the firms". Bringing several arguments, Alam et al. (2019), emphasize the importance of institutional factors for innovative activities. R&D investments are risky, bear fruit in the long run and can be affected by agency problems. Choi et al. (2014) document that high-quality institutional factors contribute to the reduction of agency problems, thereby leading to increased R&D investments. The high quality of institutional factors facilitates access to resources, is likely to attract foreign investments (Bénassy-Quéré et al., 2007), increases the transparency of information (Hillier et al., 2011), and facilitates the access of firms to external finance (La Porta et al., 1997).

Corruption and R&D expenditures in the pharma industry
The quality of the institutional environment can be affected by corruption. Widely known as a major deterrent to economic growth (Chen et al., 2022), corruption can have a corrosive impact on health outcomes (Dincer & Teoman, 2019). The pharmaceutical industry is prone to corruption for a number of reasons. The commercialization of medicines is lucrative, thereby creating a favourable context for embezzlement and misuse of power (Zirojević, 2020). Patients are often in a vulnerable position as their health comes first. Therefore, consumer price sensitivity tends to be lower as compared to other industries. Generally speaking, patients are less likely to discontinue treatment than are consumers to give up other types of products. Although the elasticity of medicines is influenced by the availability of substitutes and their prices, previous literature (Landsman et al., 2005;Yeung et al., 2016) shows that the demand for pharmaceutical products is relatively inelastic, especially for highvalue drugs. The high degree of asymmetry of information in the pharmaceutical industry and the large number of actors (medical suppliers, pharmaceutical companies, health care providers, payers and policy-makers) hinder the detection of conflicts of interest and smooth the path for different forms of corruption such as collusion, favouritism/nepotism, extortion, and fraud (Hussmann, 2010).
Although there are universally accepted forms of corruption (pharmaceutical drug counterfeiting, misuse of funds, sham contracts with doctors, bribery, illegal promotion), one has to bear in mind that there is a fine line between gifts, socially accepted favours and informal payments or bribes (Deliversky, 2016;Martinez et al., 2017;Stepurko et al., 2017).
Within this context, the pharmaceutical industry represents a good setting to scrutinize the impact of corruption on R&D decisions. The development of a drug portfolio is not possible without R&D activities. The pharmaceutical industry spends on R&D to find the cure to different diseases or to develop generic drugs and to stay competitive in a highly dynamic market (Nandy, 2022). It is widely acknowledged as the industry with the highest R&D intensity in the manufacturing sector (Mahajan, 2020). As far as these activities are concerned, Sismondo (2021) draws attention to the "ghost-management" of medical research. In other words, the medical literature is often biased because drug trials are funded, designed, and written up by pharmaceutical companies. As Egharevba and Atkinson (2016) emphasise, clinical trials are the mainstay in process of authorizing medicines. Apparently, the industrysponsored trials are conducted by independent researchers, but previous literature (Martinez et al., 2017;Sismondo, 2021) suggests that it is very likely that pharmaceutical companies silently control them. In this context, it comes as no surprise that the treatment schemes prescribed by physicians are often viewed with scepticism (Marmat et al., 2020). Unethical practices of the industry adversely influence the trust of consumers (Stepurko et al., 2017).

Hypotheses development
The pharmaceutical sector is vulnerable to corruption. Investments in innovative activities are discouraged by the low level of institutional trust caused by corruption (Anokhin & Schulze, 2009). International business literature recognizes that corruption shapes and affects the firms' behaviour (Rodriguez et al., 2005) and foreign direct investment (Hannafey, 2003). According to Daude and Stein (2007), a high level of corruption causes managers and investors to engage in acts of bribery to obtain licenses and permits, which translates into increases in the cost of the investment. Through these increased costs, corruption depresses the development of new products, services, and technology. Furthermore, a high level of corruption is associated with more uncertainty and less profitable investments, thereby making investors less motivated by such costly investments like R&D. Corruption also hinders entrepreneurship, productivity, and investment in R&D (Anokhin & Schulze, 2009). Moreover, a high level of corruption generates an increase in information asymmetries and the cost of doing business.
As this phenomenon remains a major problem that makes the actors of the pharmaceutical sector be distracted from promoting the well-being of the population, health, and innovation in the healthcare sector (Lexchin, 2019), the following hypothesis is formulated: H1: Corruption negatively influences R&D investment.
Economic freedom refers to all the rights of individuals, including labour and property rights. Nordin (2014) defines economic freedom from three points of view based on individuals, government, and the economy. Individuals have the right to choose where to live and work, their own properties and can control their productivity. The concept of freedom on government view refers to the transparency, visibility and openness of the decision-making process and the removal of discrimination. As for the economic view, every individual or firm has an equal chance of success because free and open competition permits an appropriate allocation of resources for consumption and production, and the power of economic decision-making is widespread. Countries with a high level of economic freedom accept diversity and promote creativity which encourages innovation.
Widely considered one of the main factors that generate economic prosperity, economic freedom influences the effectiveness and efficiency of resource use. There is a set of studies that shows that nations with a high level of economic freedom have evolved economically (Dawson, 1998;Easton & Walker, 1997;Hanson, 2000;Weede & Kamph, 2002). In a report conducted by McQuillan and Murphy (2009), it is stated that higher economic freedom encourages, especially in developing countries, both income growth and production growth. Depken and Sonara (2005) infer that the growth of trade flows is strongly correlated with a high level of economic freedom. Using Latin America as a research setting, Calvo and Robles (2003) show that a high level of economic freedom from the host country is conducive to increased foreign direct investment inflows. At the same time, researchers (Calvo & Robles, 2003;Pourshahabi et al., 2011) note that economic growth from the host country promotes FDI and economic freedom fosters economic growth (Nordin, 2014). Berggren (2003) shows that economic freedom has a complex character, as some components (in particular, the use of property rights, business freedom) of the economic freedom index cause economic growth, while other components (monetary freedom, government integrity, investment freedom, financial freedom, and trade freedom) come as a result of economic growth.
Turning to the pharmaceutical industry, whose products are designed to protect the health of individuals, economic freedom presupposes that patients' access to safe and effective treatments should take precedence over profit (Kreiner, 1995). Moreover, the profit obtained from the development of new drugs gives pharmaceutical companies the economic freedom to choose what drugs to develop (Hole et al., 2000). Economic freedom is also correlated with patent protection in the pharmaceutical industry (Qian, 2007) which stimulates innovation (Haley & Haley, 2011). Therefore, the following hypothesis is proposed: H2: Economic Freedom has a positive effect on R&D investment.
R&D investment and income growth mutually interact (Wang, 2010). On the one hand, Coe & Helpman (1995) stated that a high level of productivity is created by R&D outcome and spillover mechanism, and this leads to higher revenue growth. A high level of GDP growth rate makes firms consider that there are possibilities for greater use of existing production capacity, increased sales and rising profits. This is confirmed by the acceleration principle which shows that the GDP growth rate rates influence the variation in R&D investment (Schmookler, 1966). The effort involved in the R&D activity materializes in a diversified portfolio and superior quality products, which generates high returns. This increases productivity, value added, and further ensures income growth (Nordin, 2014).
On the other hand, Markusen (1986) confirmed that R&D-intensive products are more sought after by high-income consumers. Researchers have found a direct relationship between GDP growth rate and R&D investment. Cheung (2014) demonstrated that the higher its per capita real GDP, the more innovative a state is. Alam et al. (2019) observed that in developed markets GDP growth rate is correlated with the firms' R&D expenditure. The global pharmaceutical industry is changing landscape and moving towards mergers and acquisitions, contract manufacturing and R&D activities. In this case, the impact of GDP growth rate is paramount (Vaidya et al., 2018). Tran (2021) assumed that a high GDP growth rate indirectly favours R&D projects by providing companies with lower costs of external financing. A high level of GDP growth rate is expected to increase interest in innovation and technological progress in the pharmaceutical industry. Thus, following Wang (2010) who remarked that R&D investment can be stimulated by GDP growth rate, the following hypothesis is proposed: H3: Gross domestic product (GDP) growth has a positive effect on R&D investment.
The level of investments is affected by the firms' financial status (Alam et. al., 2019). External financing, and in particular debt, is considered inadequate for financing R&D investments (Hall, 1992). Obtaining funding for R&D activities can be hampered by a series of barriers, like uncertainty, risk and the absence of collateral that characterizes R&D projects (Hall & Lerner, 2010). R&D expenditures of multinational and domestic corporates are positively influenced by cash flows (Bae & Noh, 2011). According to Rafferty and Fund (2005), R&D projects are stimulated by cash flow increases. Results show that R&D projects are sustained by the success of new medicines that generate large cash flows. Firms' internal funds are considered the most important sources of R&D investment funding (Czarnitzki & Hottenrott, 2011). At the same time, companies are reluctant to use external funds to finance R&D projects in order to avoid revealing sensitive information regarding their R&D projects (Bhattacharya & Ritter, 1983). Thus, besides government funding, internal financial resources represent the safest source of financing. Himmelberg and Petersen (1994) estimate that R&D decisions depend on the constant trend of increasing cash flow. If there are no prospects for increasing cash flow, managers are reluctant to R&D investments (Brown et al. 2012;Krammer, 2015;Sasaki, 2016). However, the permanent character of the increase in the cash flow determines the firms' managers to be more willing to carry out R&D activities. According to Hubbard (1998), the firms' decisions to invest are affected by capital market imperfections. These arise as a result of the asymmetric information and agency problems between investors and management (Myers & Majluf, 1984). The firms can obtain certain public contracts in exchange for paying bribes (Liu et al., 2017). An agency problem may easily occur when corporate cash is used by managers for paying bribes, without shareholders' knowledge (Tran, 2019). Consequently, the firms' cash flow sensitivity increases. At the same time, firms may be more tempted to engage in corruption if they face certain liquidity issues. In light of these findings, it is expected that the relationship between corruption and R&D investment is influenced by liquidity issues: H4: Liquidity issues moderate the relationship between corruption and R&D investment. Figure 1 exhibits the conceptual framework of the research.

Sample and variables
The paper scrutinises the influence of corruption (captured through a composite indicator retrieved from the World Bank database), index of economic freedom (IEF), and gross domestic product (GDP) growth on R&D investment of pharmaceutical manufacturing firms listed in the European Union 1 . Furthermore, it is empirically assessed the moderating effect of liquidity issues on the corruption-R&D investment nexus. Firm-level variables are retrieved from ORBIS. Only those companies that reported R&D expenditures for at least 3 consecutive years were considered. Several observations were deleted due to a lack of data for the variables of interest. Following the use of the two filters, the final sample consists of 118 companies. The analysis is conducted on unbalanced panel data with 749 firm-year observations and spans 9 years (2011-2019).
The study uses R&D expenditures as a dependent variable. Consonant with earlier research (Chen et al., 2016;Lee & Hwang, 2003), R&D investment is quantified using the natural logarithm of R&D expenditures. The log transformation facilitates discussions and enables findings comparability.
The main independent variable is corruption. Corruption in all forms is more likely to occur in an environment characterised by uncertainty, divergent interests, and information gaps between pharmaceutical companies (insiders) and customers (outsiders) as emphasised by Global Corruption Report 2006 (Transparency International, 2006). Wang (2010) also assumed corruption as an explanatory variable of R&D investment. Secondary data concerning corruption are retrieved from the World Bank database. Therefore, corruption is quantified using a composite indicator, constructed using a plethora of individual variables that reflect the misuse of power (irregular payments, frequency of bribery, corruption in different sectors, etc). This aggregate indicator is computed using the unobserved components model (UCM) and it is based on a multitude of different data sources (such as survey institutions, think tanks, organizations and private companies) (Alam et al., 2019).
As explanatory variables, the models also include the index of economic freedom and GDP growth rate. The index of economic freedom (IEF) shows the extent to which the factors of production can move freely in a given economy. Data on the index of economic freedom comes from the World Heritage Foundation (WHF) database. According to WHF, the aggregate index is computed as the simple average of the following sub-indicators: "Property rights, Freedom from corruption, Fiscal Freedom, Government Spending, Business Freedom, Labor Freedom, Monetary Freedom, Trade Freedom, Investment Freedom, and Financial Freedom". A country's economic health is proxied by GDP growth rate, a country-level variable extracted from the World Bank database.
An indicator named liquidity issues was constructed, a dummy variable that takes value 1 if the company encounters liquidity issues and 0 otherwise. It is considered that a company faces liquidity issues if the liquidity ratio is less than 1 (Abusalah & Ng, 2012). Data regarding liquidity ratio, also known as quick ratio, is retrieved from the ORBIS database.
In addition, the models include control variables, three firm-specific factors, commonly known as good predictors of corporate innovative activity, namely: return on assets, the number of employees, which serves as a proxy for firm size, and leverage. Scott (1995) and Oliver (1997) inferred that R&D behaviour may be determined by both external and internal factors. The paper considers the abovementioned factors to minimize the omitted variable bias. Table 1 summarizes the main information about the variables included in the models.

Empirical models
Panel data analysis is based on two fixed effects models, consistent with the results of the Hausman test. The study adds cross-section fixed effects, a widely employed approach by studies that investigate the drivers of investment decisions (Sun et al., 2019). R&D expenditures are seen as a function of several macroeconomic factors and internal factors: LnR&D = β0 + β1Corruption + β2IEF + β3GDPG+ β4ROA + β5LEV + β6LnEMPL (1) LnR&D = β0 + β1Corruption + β2IEF + β3GDPG+ β4Corruption x LqI + β5ROA + β6LEV +β7LnEMPL (2) Equation (1) represents the basic model and allows the exploration of the influence of corruption, IEF, and GDP growth rate on the R&D activity of pharmaceutical companies, while equation (2) allows scrutinising the moderating role of liquidity issues on the impact of corruption on R&D investment. Data analysis is performed using EViews.

Descriptive statistics and correlation
The main indicators of descriptive statistics are presented in Table 2. The average value of R&D expenditures is 116,968 € and the standard deviation is 587.180 €. It can be inferred that pharmaceutical companies invest considerable amounts in R&D projects. Official statistics regarding the R&D landscape point out that the pharmaceutical industry is one of the top investors in R&D, right after ICT producers, owning together almost 43.5% of R&D investment in 2019 (European Commission, 2020). The average value of the corruption index is equal to 1.5417 and the standard deviation is 0.6749. The range is between -0.2673 (for Bulgaria) and 2.4049 (for Denmark). This indicates that Bulgaria suffers from a lack of control of corruption, while Denmark enjoys a strong control of corruption. As far as economic freedom is concerned, the mean value of IEF is 70.4727 which suggests that, overall, UE countries are mostly economically free. The minimum value (53.2) corresponds to Greece which indicates a mostly unfree economy, and the maximum value (80.90) corresponds to Ireland which implies that its economy is one of the freest in 2020. On average, 22,94% of the sampled companies encounter liquidity issues. The average GDP growth rate is 1.7622% and the standard deviation is relatively high (3.1278), which suggests that the countries in the sample have GDP growth rate rates that are not clustered around the mean. It is observed that the companies in the sample have, on average, negative ROA (-14.8748), which is not necessarily a sign of low efficiency of assets. Typically, firms that invest heavily in assets are associated with lower ROA due to the time-lagged impact on profit. The sampled companies are large, with an average number of employees of 3650.8. Debtto-assets ratio (leverage) mean value (0.4571) shows that, overall, the companies are not in a risky position.
To ensure there are no suspicions of multicollinearity, the matrix of correlation and the variance inflation factors (VIF) are computed and displayed in Table 1. The values of correlation coefficients are low and all the VIF values are below the reference value of 10, introduced by Kennedy (2008) and Hair et al. (2010), respectively 5, proposed by Studenmund (2016). Additionally, the condition number was examined and the results (less than 100) (Montgomery et al., 2001) confirm that the issue of multicollinearity does not arise in this study.  Notes: *significant to 10%; **significant to 5%; and ***significant to 1%. Source: Authors' representation

Regression results
As previously mentioned, this paper scrutinises the impact of some of the potential institutional determinants on research and development (R&D) investment of pharmaceutical manufacturing companies listed in the European Union. As it can be seen from Table 1, corruption adversely affects the R&D behaviour of European pharmaceutical companies. This result is congruent with the prediction and supports the institutional-based view and the "sanding-the-wheels" hypothesis. More specifically, an increase in corruption is associated with around 0.7321% fall in companies' R&D expenditure, other factors being fixed. It is evident that R&D activity is sensitive to malfeasance, bribery, irregular payments, and embezzlement. These practices have a detrimental effect on resource allocation and undermine the company's growth (Bukari & Anaman, 2021). The literature also postulates that corruption exacerbates the costs (Ivanovic-Djukic et al., 2019;Nguyen et al., 2016). Therefore, the high cost of obtaining licenses or loans acts as a deterrent to a firm's decision to undertake R&D activities (Sena et al., 2018). The paper corroborates the findings by Alam et al. (2019) who analysed 663 firms from emerging markets and concluded that corruption harms R&D activity, as it leads to higher investment costs and hinders foreign investments. Using a longitudinal dataset of 48 U.S. states and two measures of corruption, Dincer (2019) also demonstrates that corruption decelerates innovation. The inverse relationship between corruption and innovation is also empirically evidenced by other analyses (Lee et al., 2020;Mahagaonkar, 2008).
Concerning the second hypothesis, table 3 clearly shows that the index of economic freedom has a positive impact on R&D investment, consonant with conventional wisdom. Economic freedom is widely recognized as a catalyst for economic development (Rapsikevicius et al., 2021). The report of McQuillan & Murphy (2009) lists many benefits of greater economic freedom, inter alia, increased investments, and more innovation. The view that economic freedom enhances innovation is also shared by other scholars. Using panel data of 5809 companies, spanning 22 years, Zhu & Zhu (2017) provide evidence that economic freedom is a key driver of corporate innovation. Similarly, Erkan (2015) finds a positive association between innovation and IEF and emphasizes the importance of economic freedom in devising policy recommendations. Asandului et al. (2016) find a positive correlation between IEF and GDP per capita and between IEF and social progress index. The hypothesised direct relationship between GDP growth rate and R&D investment is supported by the empirical analysis. A rise in GDP growth rate leads to increased R&D investments. The result is hardly surprising and merely confirms that firms have the confidence to invest more when economic growth is strong. Galindo & Méndez (2014) demonstrate that the direction of causality runs both ways as economic growth positively influences innovation and vice-versa. Hypothesis 3 is thus supported. Table 3 also conveys the results of the regression analysis examining the moderating effect of liquidity issues on the corruption -R&D investment relationship. The findings highlight that liquidity issues moderate the influence of corruption on R&D activity (β=-0.1079). In other words, the adverse effect of corruption on the R&D activity of the manufacturing pharmaceutical industry is more pronounced when firms deal with liquidity issues. One possible explanation found in the literature is that debt is not suitable for financing R&D activities (Ughetto, 2008;Xu & Yano, 2017), and internal equity should be employed instead. It is concluded that corruption and economic freedom play a relevant role in explaining the R&D behaviour of pharmaceutical companies.

Robustness checks
Several additional tests are performed to check the robustness of the results. To reinforce the results, the sample is further divided into subsamples based on the median of the variables GDP growth rate and government effectiveness. Sub-sample analysis running the abovementioned specifications is performed (equation 1 and equation 2). The use of alternative sub-samples in running supplementary analyses to ensure the generalizability of the findings is a recommended practice (Hair et al., 2010). The country-level variable 'government effectiveness' was retrieved from the World Bank database and refers to the competency and the capacity of the government. Third robustness test considers an additional variable (i.e. ECenforcing contracts) that measures the time and cost for resolving a commercial dispute through a local first-instance court, and the quality of judicial processes index. The results presented in Tables 4-6 lend strong support to the baseline findings and the hypotheses. There are no noteworthy changes in the values or statistical significance of the coefficients. The results are quite robust.
As a final exercise, we performed an analysis to eliminate endogeneity concerns. The independent variables, suspected to be endogenous, were tested. First, independent variables were regressed. The resulting residuals and Wald Test were used to ensure that there are no problems of endogeneity.

Conclusions
To bring the paper to a close, the main points are summarised here: corruption, economic freedom and GDP growth rate play a meaningful role in explaining the R&D behaviour of European pharmaceutical companies and the adverse effect of corruption on the R&D activity is more evident when firms encounter liquidity issues.
Corruption in the pharmaceutical sector is a fervent topic. Raising awareness about this issue by studying the nexus between corruption and R&D behaviour of pharmaceutical manufacturing companies is a paramount step in developing effective policies to address this issue and stimulate a collective effort of mitigating the effects of this nefarious phenomenon. Understanding corruption and its consequences are not solely of academic importance but are also relevant in designing anti-corruption policies. The results yield important implications for managers and policy-makers and may be of interest to both civil society and scholars concerned with the issue of corruption in the pharmaceutical industry and the determinants of R&D activity. By providing anecdotal evidence on the impact of corruption and economic freedom on R&D, the study can contribute additional pieces to the knowledge of factors susceptible to influence R&D expenditures. Most importantly, the paper emphasizes the moderating role of liquidity issues corruption -R&D investment relationship. It is recommended that pharmaceutical firms take notice of the corruption phenomenon in all its forms owing to its deleterious effect on R&D activity and take several measures such as adopting new technologies throughout the pharmaceutical value chain to raise transparency and establishing a management oriented towards promoting anti-corruption principles. It is paramount that all actors from the health sector understand that patients are more important than profits. The problem of corruption exacerbates when firms face liquidity issues. Therefore, managers may take into account to access a flexible line of credit or negotiate favourable credit terms with their suppliers, as needed. Governments are expected to create a climate conducive to R&D investment through fiscal facilities and discourage corrupt practices through unequivocally enforced regulations and severe sanctions.
The research has some limitations due to the lack of firm-level evidence regarding corruption, a relatively small-time span of analysis, and the inclusion of all EU countries, disregarding the development level. Notwithstanding that the results are related to the EU context, they could have international relevance as well. The shortcomings of this paper offer opportunities for future research. Therefore, it would be useful to include firm-level data regarding corruption, extend the period of analysis, conduct the same study on different industries and develop the analysis by investigating different subsamples such as euro area and non-euro area states.