Why do firms invest
in Research and
Development?
A country-level study
Virginia Salas Rodríguez
S3293823
Supervisor: Dr. P. Rao Sahib Co-assessor: Dr. R.K.J. Maseland
[1] 1. Introduction ... 3 2. Literature Review ... 7 Economic Situation ... 7 Government support ... 9 Education... 10 3. Hypotheses ... 11 Economic Situation ... 11 Government support ... 12 Education... 13
4. Data collection, model, and methodology ... 13
5. Results ... 16
Economic Situation ... 18
Government Support ... 19
Education... 20
Does the GDP growth have a retarded effect on BERD? ... 21
6. The Great Recession and R&D... 23
7. Conclusions ... 25
8. Annexes ... 28
1. Comparison of GOVERD, HERD and BERD between countries (2001 and 2017). ... 28
2. GDPppp vs BERDppp for OECD countries (2017). Source: OECD MSTI ... 29
3. Table of correlations (Stata output) ... 30
[2]
Abstract
This thesis examines the drivers influencing business decisions to invest in Research and Development (R&D). Using a panel data set for 25 OECD countries over 17 years (2001-2017), I investigate how different variables at the country level, such as GDP, government expenditure in R&D, education, and trade openness, determine business expenditures in R&D.
The findings are in line with the expected outcome, as the results show that firms invest more in R&D the bigger the GDP per capita of the country. They also decide to invest more when public organizations such as government or higher education institutions increase their investments in R&D. Education also plays a significant role, as R&D is usually carried out by highly skilled people. However, this thesis reveals that GDP growth and openness does not affect firms’ decisions regarding R&D expenditure.
Keywords: R&D determinants; Government expenditure in R&D; Business expenditure in R&D
Resumen
Este trabajo analiza cuales con los factores que influyen en las decisiones empresariales para invertir en Investigación y Desarrollo (I+D). He utilizado una base de datos de panel de 25 países pertenecientes a la OCDE y 17 años (2001-2017) para investigar como diferentes variables a nivel nacional, como el PIB, gasto del gobierno en I+D, educación y apertura comercial determinan los gastos de las empresas en I+D.
Los resultados se corresponden con lo previsto, ya que muestran cómo las empresas deciden invertir más en I+D en países con mayores niveles de PIB per cápita. También deciden aumentar su inversión cuando organizaciones públicas como el gobierno y las instituciones de enseñanza superior también aumentan sus propios gastos. El nivel educativo del país también está positivamente relacionado, ya que la investigación y el desarrollo normalmente se lleva a cabo por trabajadores altamente cualificados. Sin embargo, he encontrado que el crecimiento en el PIB y la apertura de un país no afecta las decisiones privadas en el gasto en I+D.
[3]
1. Introduction
It is widely known that Research and Development (R&D) is “key to generating new knowledge, solving problems, fostering productivity, and remaining internationally competitive in an increasingly globalized world” (OECD, 2018). R&D is defined as the activity “novel, creative, uncertain, systematic, transferable, and/or reproducible” (Frascati Manual 2015, 2015). It is “undertaken for the purpose of discovering or developing new products, including improved versions or qualities of existing products, or discovering or developing new or more efficient processes of production” (OECD, 2001).
After recovering from the crisis, “Europe’s economy is growing at its fastest rate in a decade” (Eurostat, 2018). The current agenda, the Europe 2020 Strategy, is focused on jobs and growth. This strategy was proposed and launched by the European Commission (EC) in 2010 to get out of the crisis and prepare the EU economy for the next decade (European Commission, 2010). It is used as a reference framework for policies and activities in the EU and at national and regional levels. Within this agenda, one of the targets is to have 3% of the EU’s GDP invested in R&D by 2020 (European Commission, n.d.-a). However, investment in R&D can be done both at the public and private levels, and the EC has not specified the relative efforts of each sector necessary to reach that objective (European Commission, 2013).
Graph 1 shows the gross domestic expenditure on R&D (GERD)1 as a percentage of the GDP. GERD is the principal aggregate measure used to describe a country’s R&D activities and a “primary indicator for international comparisons of R&D activity” (Frascati Manual 2015, 2015). The EU-28 GERD has stagnated around 2% since 2012, and it is growing at a slow rate when compared with countries such as Japan, the US, Korea, and China.
[4]
Graph 1. Gross domestic expenditure on R&D (GERD) as a percentage of the GDP. Prepared by the author based on data from the OECD (OECD, n.d.-a).
As a country develops, the governmental expenditure on R&D tends to decrease, and businesses assume on a larger role to fill the gap by increasing private expenditure on R&D, which increases local innovation (Raghupathi & Raghupathi, 2019).
For most industrialized countries, the business enterprise sector, which includes “all firms, organizations and institutions whose primary activity is the market production of goods or services for sale to the general public at an economically significant price, and the private non-profit institutes mainly serving them” (Eurostat, n.d.),is responsible for the largest share of R&D and human resources investments (Frascati Manual 2015, 2015). In fact, as seen in annex 1, we can see how the business enterprise expenditure on R&D (BERD) is larger than the higher education expenditure on R&D (HERD) and the government intramural expenditure on R&D (GOVERD) for almost every country in the analysis. In certain cases, such as Estonia, Latvia, Lithuania, and Turkey, high education institutions invest more money in R&D than businesses.
Between 2000 and 2007 the business expenditure on R&D was increasing at a rate of 4% average annual growth. However, in the period between 2007 and 2011 this increasing rate was attenuated compared to the pre-crisis period, as there was a 2.3% average annual growth. Smaller Eastern European countries have particularly sought to catch up, both in terms of nominal growth and business R&D intensity. On average in the EU, business R&D intensity went from 1.18% of the GDP in 2007 to 1.26% of the GDP in 2011. This level is, however, notably inferior to that of “South Korea (2.99%, 2010), Japan (2.54%,
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7 Japan Korea United States
[5]
2009) and the United States (2.02%, 2009). It is also below the level of China, which was more than 1.4% in 2011” (European Commission, 2013).
Graph 2 highlights some of the major differences in the business enterprise expenditure on R&D within the countries included in the analysis. Certain countries stand out as having grown significantly, such as Israel and Korea spending more than 3% of the GDP in BERD in the last 6 years. Estonia also has an increasing BERD, reaching its peak in 2011, and then decreasing significantly again. Finland also stands out because its BERD has shrunk since 2009.
In this thesis, I build on Falk’s (2006) early work through investigating the underlying causes of variations in business R&D intensity among countries and, most significantly, what factors influence business sector expenditure in R&D. Falk uses a panel of OECD countries for studying the “potential determinants of business sector R&D intensity” for the period 1975-2002 (Falk, 2006). Falk obtains averages every five years for creating his dataset, but omitting data impedes to analyze the impact of certain shocks, as it was, for example, the early 1990s recession. I instead use new data from 2001 to 2017 to assess the sign and magnitude of each variable described in section 3. This is important for policy makers (Falk, 2006) to understand the best instruments to stimulate R&D expenditure. Moreover, it is important to analyze this data yearly, as certain countries experience significant variations each year. Estonia, for example, spent 1.46% of the GDP in BERD in 2011 and then less than half of this (0.62% of the GDP) in 2014.
R&D’s relevance has increased during the 21st century due to the growing social, environmental, and economic challenges the world is facing. As such, this period is important to analyze. This updates the existing results and ensures the relevance of certain variables for firms’ investment decisions. In addition, the Great Recession of the late 2000s unexpectedly affected the firms’ performance, as economic activity decreased significantly. The main research question is therefore:
Graph 2. BERD as a percentage of the GDP 2001-2017. Prepared by the author based on data from the OECD (OECD, n.d.-a). 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 2 0 1 7
Belgium Canada Czech Republic Denmark Estonia Finland France
Germany Hungary Ireland Israel Japan Korea Latvia
[7]
This first section provides an introduction to the topic, which, as previously mentioned, is incredibly relevant for the development of a country. The remainder of the thesis is organized as follows: In the next section, the studies and results from numerous researchers are summarized to provide an overview of the current conversations in the literature. The third and fourth sections present the hypothesis and the data collection and models. Section 5 discusses and interprets the results obtained from running the regressions. For the sake of clarity, these sections are classified in three main themes: economic situation, referring to the GDP and the international trade in the country; government support, which refers how the public sector stimulates R&D; and education, in which the focus is on the R&D investment by higher education institutions and the education level of the countries.
Following this is a section on the Great Recession and its implications for firms’ investments in R&D. Lastly, I provide conclusions to the study, policy implications, and recommendations.
2. Literature Review
Falk (2006) has investigated the “potential determinants of business sector R&D intensity” to find that the two main policy tools (fiscal incentives and direct subsidies) are significant and positively related with business R&D spending. Furthermore, he finds that investments on R&D made by higher education institutions have a positive impact on business enterprise sector expenditures on R&D. This indicates that private sector R&D and public sector R&D are complements (Falk, 2006). In this section, however, the results of other studies estimating the impact of economic variables on firms’ R&D expenditure are summarized, classified in the three themes mentioned above.
Economic Situation
R&D is “statistically and economically important for both technological catch-up and innovation” (Griffith, Redding, & Reenen, 2004). Social returns on R&D seem to increase with distance from the technological frontier, reflecting the gains to catch-up (Griffith et al., 2004). For this reason, less developed countries are expected to have higher levels of investment in R&D. Available evidence, however, shows the opposite2, as private R&D
[8]
efforts increase as the GDP rises (Lederman & Maloney, 2003). This perhaps refers to how governments often reduce their expenditures when the nation starts developing, and it becomes the turn for companies to increase their investments (Raghupathi & Raghupathi, 2019). Also, there are no incentives to invest high amounts of money in R&D, as less advanced countries, which are farthest from the world technology frontier, usually take advantage of R&D by imitating existing technologies and adapting them to local conditions.
Conversely, economic growth can be both a cause and a consequence of R&D investment. Following the Solow Model, increasing investments in R&D are expected to boost the country’s economic growth. One study, for example, has found that “ceteris paribus, an increase in R&D expenditure as a percentage of GDP by 1% would cause an increase of real GDP growth rate by 2.2%” (Sokolov-Mladenović, Cvetanović, & Mladenović, 2016). In this thesis, however, I examine the influence of GDP growth on R&D investment. There are some theories that explain this relationship. On the one hand, the acceleration principle, which posits that increases in consumption increase investments, and therefore “rising GDP implies that businesses in general see rising profits, increased sales and cash flow, and greater use of existing capacity” (Sameti et al., 2010). On the other hand, the R&D-driven growth model, which suggests that “larger markets have stronger incentives to invest in R&D, which consequently leads to a faster growth” (Sameti et al., 2010).
Furthermore, there seems to be consensus regarding the positive relationship between economic growth and FDI inflows, which assumes that the country receiving the investment has reached a “minimum level of educational, technological and/or infrastructure development” (Hansen & Rand, 2006). One might expect that FDI inflows would cause an increase in the receiving nation’s R&D and innovation activities. However, while there are studies indicating that international trade and FDI are positive, high innovation determinants (Erdal & Göçer, 2015), others might argue that they are not significant (Damijan, Knell, Majcen, & Rojec, 2003; Haddad & Harrison, 1993).
[9]
to assimilate knowledge created by its trade partners. The higher the willingness of a country to be competitive the higher its spending on R&D.” (Varsakelis, 2001). However, while the question of whether an open economy stimulates technological progress is still unanswered (Parameswaran, 2010), international trade can certainly alter private R&D investment through different channels. Parameswaran (2010), using firm-level data, has found that “exports encourage investment in innovation”, while technology imports have a “positive effect only on sectors that have relatively lower level of engineering and technological capability.” If we focus on the impact of import competition, this depends on domestic market structure: “It promotes investment in R&D only when the market structure is highly concentrated otherwise it has negative effect.”
A country’s comparative advantage in high-technology products can be also measured by its exports. Moreover, “high technology exports are considered an important factor for a country’s sustainable economic growth” (Turen & Gökmen, 2013). According to Bebczuk (2002), “international trade in intermediate or final technology-intensive goods avoids the duplication of R&D efforts which, while improving the allocation of world resources, tends to deprive some countries from a national R&D industry.”
Government support
Private R&D intensity can be stimulated by governments in various ways, for example through policies, loans, subsidies, procurement, and tax incentives. All these instruments can, directly or indirectly, encourage firms’ R&D.
One of the incentives for stimulating innovation is intellectual property rights (IPR) (Kanwar & Evenson, 2003), as “sound property rights protection, a just legal system, and a clean government willing to help, are crucial to encouraging corporate R&D and long-term economic development” (Lin, Lin, & Song, 2010). However, stronger protection eventually impedes technological progress, especially when it comes to fundamental research (Falk, 2006).
[10]
and Trade stimulated long-run company-financed R&D expenditures.” Public funding is, therefore, believed to result in increased private investments that would not occur without governmental aid (Carboni, 2017).
However, the participation of the public administration in the R&D market reinforces the lack of private R&D (Bebczuk, 2002). This is mainly because public programs are usually aimed at supporting projects with important expected benefits for the society but low expected private returns. As such, they are efficient in supporting marginal projects that would have been abandoned without a subsidy (Carboni, 2017). It is also possible that the government supports the business R&D even when the results will belong to the R&D performer (the firm that accomplishes the R&D) and thus the return will be private (not necessarily have social returns). However, the funder usually chooses the specific goals. One example is the EU Framework Programme for R&D, Horizon 2020, whose thematic work programmes describe the overall objectives, the calls for proposals in order to reach those objectives and the topics within each call (European Commission, n.d.-b). The funder, in this case the European Commission, sets the actual goals, the funding that will be allocated for the project and the requirements for the proposals. They normally require the firm to cooperate with other firms or universities for the research (Guellec & Van Pottelsberghe De La Potterie, 2004).
Tax incentives are a unique form of public support and they have a direct positive effect on private investment in R&D as they are not discriminatory. This means that firms are free to use “public money”, which they have saved from taxes, for any project, no matter the social rate of return (Guellec & Van Pottelsberghe De La Potterie, 2004).
Education
While certain factors, such as policies and infrastructure that enhance R&D are definitely necessary for accomplishing the innovation goals, they are not enough if they are not coupled with financial and human capital investments in innovations (Furman & Hayes, 2004). According to Bebczuk (2002), “due to its scientific nature, formal R&D is intensive in human capital.” This means that there is a positive correlation between the number of researchers in a country and R&D expenditures (Bebczuk, 2002).
[11]
profitability and the industrial fields in which the companies operated” (Scherer & Huh, 1992).
Other factors, such as corporate governance drivers, are relevant for a firm’s innovation capacity. These include executive incentives, the presence of independent non-executive directors (Bobillo, Rodríguez-Sanz, & Tejerina-Gaite, 2018), marketing abilities, number of workers with university degrees and number of engineers, previous experience of employees, proportion of staff with leadership roles and responsibilities, and incentives offered to the employees to contribute to innovation (Souitaris, 2002).
In regards to education, Van Pottelsberghe de la Potterie explains that academic research also stimulates business R&D because “universities generate new ideas which are then transferred to the private sector. The transformation of these ideas into products or processes requires further applied research activity and development” (van Pottelsberghe de la Potterie, 2008).
The purpose of this review is to gather and summarize the most relevant studies that have been recently conducted. As the relevance of R&D continues to increase, it consequently continues to be researched and discussed. As I have based my hypotheses and variables on the literature discussed in this section, I expect the results to align with their findings.
3. Hypotheses
Given the research question established in the introduction and the literature review, six different hypotheses are outlined in this section. The hypotheses are classified following the same structure of the three main themes.
Economic Situation
[12]
H1a: The level of business enterprise expenditure on R&D is positively related to
the level of the GDP.
However, not only the current development of a country might influence firms’ investments, but the growth it is experiencing. Therefore, the variable GDPgrowth is included as a measure for economic growth. This is because it is expected that BERD increases as a nation develops.
H1b: The level of business enterprise expenditure on R&D is positively related to
the growth of the GDP.
In addition, as described in the literature review, the trade balance and the openness of the country can also impact its BERD. This is measured by using two different variables: “openness” and “high-technology exports.”
The OECD was first created to pave the way for a “new era of cooperation that was to change the face of Europe.” Currently, the OECD “brings around its table 39 countries that account for 80% of world trade and investment” (OECD, n.d.-b). Therefore, all the countries included in the analysis trade internationally. Thus, the variable openness, reflects the intensity of the international trade, not whether they trade or not. The
high-technology exports variable is also added, because it is expected that countries with a
favorable position in high technology products invest more in R&D, as those products have a high R&D intensity.
H2: The level of business enterprise expenditure on R&D is positively related to
the country’s level of international trade.
Government support
It is expected that BERD increases when the government supports R&D through both direct and indirect measures. In line with the Frascati Manual (2015), this thesis uses the government intramural3 expenditure on R&D as an indicator of government support.
H3: The level of business enterprise expenditure on R&D is positively related to
the government expenditure in R&D.
[13]
Education
Lastly, countries with high levels of education are expected to spend more in R&D, as it is necessary to have the knowledge and skills to conduct the R&D activities financed by companies or the government.
H4: The level of business enterprise expenditure on R&D is positively related to
the education levels in the country.
Additionally, consistent with the prior literature, this thesis considers academic research another driver for private R&D. Businesses often only invest in R&D when they can get a service or product with practical applications that can be sold, thereby receiving a benefit from the investment. To carry out this applied research, it is necessary that some basic research has already been conducted. Basic research often takes place in universities, which is why the variable higher education expenditure in R&D is include in the analysis. It is expected that businesses in countries with higher levels of higher education expenditure in R&D invest more in R&D.
H5: The level of business enterprise expenditure on R&D is positively related to
the higher education expenditure in R&D.
4. Data collection, model, and methodology
Building on the determinants mentioned in the preceding section, as well as the data collection conducted by Falk (2006), data has been gathered on several variables. This is summarized in Table 1.
[14]
head, at current purchasing power parity (ppp), given that the countries to be compared have different currencies. The GDP growth has been calculated4 using this data.
The data on high technology exports as a percentage of the manufactured exports is obtained from the World Bank open data (The World Bank, n.d.). In addition, the
education index is a measure from the United Nations Development Program that is
calculated using mean years of schooling and expected years of schooling (United Nations, n.d.).
Lastly, the openness variable. There is not a fixed measure to calculate it, as it is often described very differently among empirical studies (Yanikkaya, 2003). In this thesis, however, it is calculated5 based on Falk (2006) as the sum of exports and imports of goods and services, as a ratio of GDP at nominal market prices. The data for calculating it is obtained from the OECD economic outlook data (OECD, n.d.-c).
Table 1. Variable description.
Variable Description Source
Dependent variable:
BERD Business enterprise expenditure on R&D as a
percentage of the GDP MSTI – OECD
Independent variables:
HERD Higher education expenditure on R&D as a
percentage of the GDP MSTI – OECD
GOVERD Government intramural expenditure on R&D as
a percentage of the GDP MSTI – OECD
GDPpc GDP, per head, US$, constant PPPs, reference
year 2010 OECD
GDPgrowth GDP vs previous year OECD HTECHEX High-technology exports (% of manufactured
exports) World Bank
EDUINDEX Education index calculated using mean years of
schooling and expected years of schooling.
United Nations
Development Programme OPENNESS Exports and imports as a ratio of the GDP at
nominal market prices
OECD Economic Outlook Data
4 GDP growth rate at year t has been calculated as follows: 𝐺𝐷𝑃𝑔𝑟𝑜𝑤𝑡ℎ
𝑡=𝐺𝐷𝑃𝑝𝑝𝑝𝐺𝐷𝑃𝑝𝑝𝑝𝑡−𝐺𝐷𝑃𝑝𝑝𝑝𝑡−1
𝑡−1
5 Openness has been calculated using the following equation: 𝑂𝑝𝑒𝑛𝑛𝑒𝑠𝑠 =𝐸𝑥𝑝𝑜𝑟𝑡𝑠 + 𝐼𝑚𝑝𝑜𝑟𝑡𝑠
[15]
In regard to the choice of countries, the full dataset of the OECD offers information from a total of 47 countries, including some non-OECD nations, such as Argentina, China, and Russia and aggregates for the EU-28, EU-15, and the total of the OECD. However, not all the variables analyzed in this study are available for all these countries. For that reason, some of them have been excluded from the analysis, cutting down the sample to 25 countries6.
This compilation of data, after removing countries and years with missing variables, results in a “short and wide” balanced panel data set of 17 years (2001-2017) and 25 countries. In total there are 425 observations. As analyzing the relationships among these variables for only one country is insufficient, using this kind of panel data offers greater data variation, less collinearity, and more degrees of freedom. Moreover, it allows us to account for unobserved individual differences, or heterogeneity (Carter Hill, Griffiths, & Lim, 2012).
Given the literature review, the data collection, and previous analyses on the determinants of R&D intensity in the business sector, the following economic model entails the basis of this thesis:
𝐵𝐸𝑅𝐷𝑖𝑡 = 𝑓(𝐻𝐸𝑅D𝑖𝑡, 𝐺𝑂𝑉𝐸𝑅𝐷𝑖𝑡, 𝐻𝑇𝐸𝐶𝐻𝐸𝑋𝑖𝑡, 𝐺𝐷𝑃𝑖𝑡, 𝐸𝐷𝑈𝐼𝑁𝐷𝐸𝑋𝑖𝑡, 𝑂𝑃𝐸𝑁𝐸𝑆𝑆𝑖𝑡) where i indicates the country (i = 1, …, 25) and t indicates the period (t = 1, …, 17) for all years of data from 2001 to 2017.
The descriptive statistics of the study variables are represented in Table 2. As explained in the previous section, all those variables are expected to be significant and have a positive relation with the business expenditure on R&D.
[16]
Table 2. Descriptive Statistics
VARIABLES mean sd min max
BERD 1.193 0.863 0.108 3.913 HERD 0.419 0.172 0.0513 1.037 GOVERD 0.213 0.100 0.0236 0.495 GDPgrowth 4.987 4.310 -14.60 35.99 GDPpc 33,008 10,989 11,591 66,334 HTECHEX 13.88 7.438 1.474 60.66 EDUINDEX 0.835 0.0729 0.510 0.941 OPENNESS 0.833 0.729 0.000482 2.674
Multicollinearity arises when many of the variables are correlated to one another (Dohoo, Ducrot, Fourichon, Donald, & Hurnik, 1997). Consequently, this can be a problem when interpreting the results as the coefficient estimates would be unstable. Dohoo (1997) explains that multicollinearity is a problem when the correlation coefficients are over 0.9. The variables mentioned above have correlation coefficients all lower than 0.67, and it is therefore assumed that there is no correlation problem.
5. Results
Two equations have been estimated according to the previous sections. I have been estimated two equations because there are two variables that represent GDP, GDPpc and
GDPgrowth, and therefore need to be estimated separately.
(1) 𝐵𝐸𝑅𝐷𝑖𝑡 = 𝛽1+ 𝛽2𝐻𝐸𝑅𝐷𝑖𝑡+ 𝛽3𝐺𝑂𝑉𝐸𝑅𝐷𝑖𝑡+ 𝛽4𝐻𝑇𝐸𝐶𝐻𝐸𝑋𝑖𝑡+ 𝛽5𝐺𝐷𝑃𝑝𝑐𝑖𝑡+ 𝛽6𝐸𝐷𝑈𝐼𝑁𝐷𝐸𝑋𝑖𝑡+ 𝛽7𝑂𝑃𝐸𝑁𝐸𝑆𝑆𝑖𝑡+ 𝑒𝑖𝑡
(2) 𝐵𝐸𝑅𝐷𝑖𝑡 = 𝛽1+ 𝛽2𝐻𝐸𝑅𝐷𝑖𝑡+ 𝛽3𝐺𝑂𝑉𝐸𝑅𝐷𝑖𝑡+ 𝛽4𝐻𝑇𝐸𝐶𝐻𝐸𝑋𝑖𝑡+ 𝛽5𝐺𝐷𝑃𝑔𝑟𝑜𝑤𝑡ℎ𝑖𝑡+ 𝛽6𝐸𝐷𝑈𝐼𝑁𝐷𝐸𝑋𝑖𝑡+ 𝛽7𝑂𝑃𝐸𝑁𝐸𝑆𝑆𝑖𝑡+ 𝑒𝑖𝑡
The Hausman test is used to compare the coefficient estimates from the fixed effects (FE) model to those from the random effects (RE) model. The idea is that both RE and FE
[17]
estimators are consistent if there is no correlation between the explanatory variables and the error term (Carter Hill et al., 2012). After performing Hausman’s (1978) specification test for both models8, the null hypotheses cannot be rejected and therefore random effects regressions should be used. This is confirmed after performing the LM test (Breusch-Pagan Lagrange Multiplier) which is used to test for the presence of random effects.
The random effects model assumes that all individual differences are captured by the intercept parameter and that all individuals in the sample are randomly selected, and thus the individual differences must be treated as random rather than fixed (Carter Hill et al., 2012).
The minimum variance estimator for the random effects model is a generalized least squares (GLS) estimator (Carter Hill et al., 2012). The results obtained after running the GLS regressions in Stata are summarized in Table 3, which shows the determinants of
BERD intensity using random effects. The probability values of the Wald Chi-square
statistic for both models are less than 5% and therefore, it is possible to conclude that at least one of the regression coefficients is not zero.
[18]
Table 3. Determinants of BERD
(1) (2) HERD 0.748 0.650 (5.30)** (4.41)** GOVERD 1.057 1.130 (4.45)** (4.61)** GDPpc 0.000 (4.67)** GDPgrowth -0.001 (0.51) HTECHEX 0.008 0.007 (3.35)** (2.89)** EDUINDEX 0.927 2.187 (2.03)* (5.77)** OPENNESS -0.047 0.012 (0.96) (0.24) _cons -0.842 -1.248 (2.51)* (3.77)** R2 0.324 0.319 N 425 425 * p<0.05; ** p<0.01
As the R-squared shows, a bit more than 32% of the variation in BERD in model (1) is explained by the variables included, and approximately 32% is explained by in model (2). Nearly all variables are significant at the 5% level, with the exception of openness, GDP
growth, and the education index, which is significant at the 1% level in the first model.
The results are also explained through the three major themes with which the independent variables are classified.
Economic Situation
In regards to the economic situation, the included variables are GDPpc, GDPgrowth,
HTECHEX and openness. GDPpc is significant and it has a minimal but positive impact.
[19]
An important note is that all these variables are assumed to have an immediate impact on
BERD. This might explain why GDPgrowth is not significant, as it is reasonable to expect
that there is a delayed impact in the dependent variable. This is checked later this section.
The high technology exports variable is also significant and positively related to BERD in both models, as expected. Therefore, if the exports of high technology products increase, the BERD also increases. Analyzing the high technology exports measures the country’s specialization in high-technology industries. It is, therefore, understandable that the more a country specializes in high technology products, the more businesses decide to invest in R&D. Nevertheless, there are studies showing that a positive inverse relationship is also present: “by increasing high technology manufactured goods’ exports, countries could increase their GDP per capita which also requires increased R&D that translates itself as high technology manufactured exports” (Prof & Ustabaş, 2016).
The third variable in this group, openness, is measured as the sum of exports and imports as a ratio of the GDP. This is the only variable in the model that is not significant, and the sign in model 1 is not consistent with the expected. The results show that liberalization does not spur business enterprise expenditure in R&D, which is in line with Bebzcuk’s robust findings. He explains that “international trade in intermediate or final technology-intensive goods avoids the duplication of R&D efforts which, while improving the allocation of world resources, tends to deprive some countries from a national R&D industry” (Bebczuk, 2002).
These results, therefore, support the acceptance of H1a, given that the GDP per capita is positively related to BERD. Nonetheless, H1b cannot be accepted nor rejected as
GDPgrowth is not significant. H2 is rejected because based on the data collected, an increase in the international trade of a country does not necessarily increase its firms’ expenditures in R&D.
Government Support
[20]
also increases. This means that for the OECD countries and in line with many other studies, public R&D expenditures stimulate private R&D.
Education
It is well-known that “human capital is seen as an enabling feature in cost-effective innovation” (Leiponen, 2005). Corporate abilities and technological innovation are therefore related (Piva & Vivarelli, 2009), meaning that R&D is usually carried out by highly-skilled workers. Skills are determinant for firms’ R&D since “labor skills are crucial in enhancing innovation in terms of individual creativity, organizational change and acquisition of external knowledge” (Piva & Vivarelli, 2009).
Skills can be cognitive and non-cognitive, and certain skills are “determined by parental environments and investments at diverse phases of childhood” (Cunha, Heckman, & Schennach, 2010). This is, however, difficult to measure at the country level. The level of education shows us how well-educated the people are in the country, as many skills are learned during schooling years. Therefore, the first variable in this group is the education index. As shown in Table 4, it is significant in model (2) at the 5% level and in model (1) at the 1% level, and it is positively related to BERD. An increase in the education index of the country increases business enterprise expenditure.
A second variable included in this group is HERD, which in both models is significant at the 5% level and positively related to BERD. An increase in HERD increases BERD.
Academic research plays an significant role in stimulating BERD, as the higher education institutions generate new technological knowledge that reinforce the applied R&D done by businesses and their innovation processes (Coccia, 2011).
[21]
Table 4. Summary of results.
Variable Results Bibliography
Economic situation
GDP per capita GDPpc + + Falk (2006)
GDP growth GRPgrowth No
significance + Lederman & Maloney (2003)
International trade Openness No significance + Falk (2006) + Varsakelis (2001) - Bebczuk (2002) HTECHEX + + Falk (2006) - Bebczuk (2002) Government Support Government intramural expenditure on R&D GOVERD + + Carboni (2017) - Bebczuk (2002) Education Higher education expenditure on R&D HERD + + Falk (2006)
+ van Pottelsberghe de la Potterie (2008)
Education index EDUINDEX + + Bebczuk (2002)
Does the GDP growth have a retarded effect on BERD?
As previously discussed, all these variables are assumed to have an immediate effect on
BERD. Nevertheless, it might take a significant amount of time for the companies to
adjust to changes in the national GDP. It has previously been shown that GDPgrowth is not significant for BERD, but here I intend to check if the effect is delayed. In Table 5 the different regressions are shown, and the first column is the same one as that of Table 4, with fixed effects and no lags. In columns 2, 3, and 4, I add 1, 2, and 3 lags respectively to the GDPgrowth. GDPgrowth is still not significant with the addition of more lags, and the growth of the total GDP therefore does not have an impact on BERD.
[22]
Table 5. BERD determinants
(1) (2) 1 lag (3) 2 lags (4) 3 lags HERD 0.650 0.646 0.603 0.637 (4.41)** (4.19)** (3.83)** (3.94)** GOVERD 1.130 1.097 1.141 1.098 (4.61)** (4.20)** (4.19)** (3.87)** GDPgrowth -0.001 (0.51) GDPgrowth1 -0.002 (0.86) GDPgrowth2 -0.005 (1.95) GDPgrowth3 -0.004 (1.79) HTECHEX 0.007 0.007 0.007 0.007 (2.89)** (2.60)** (2.59)** (2.65)** EDUINDEX 2.187 2.289 2.359 2.400 (5.77)** (5.62)** (5.48)** (5.22)** OPENNESS 0.012 0.043 0.061 0.085 (0.24) (0.81) (1.02) (1.27) _cons -1.248 -1.343 -1.397 -1.461 (3.77)** (3.76)** (3.69)** (3.59)** N 425 400 375 350 * p<0.05; ** p<0.01
[23]
6. The Great Recession and R&D
The fact that the growth in GDP does not affect firms’ investment decisions in R&D raises another question regarding whether the Great Recession positively or negatively influenced BERD.
It is well-known that the economic and financial crisis that started in the US in mid-2007 marked a turned point in global economics. Few could have known that it would be the worst collapse since the Great Depression. Even today, 12 years later, some countries are still experiencing the consequences of that economic and financial downturn. According to the European Commission, “historical data shows that private R&D investments follow economic downturns to some extent due to liquidity pressure, difficulties in finding appropriate financing, credit constraint, falls in sales and available cash-flows, and difficulties facing shorter-term payments” (European Commission, 2011). For many countries, the annual increase in GERD slowed down in the 2009-2010 period, but they rapidly caught up with the previous rates after 2010. As shown in Graph 2, most countries appear to maintain the same (or even higher) levels of BERD, as a percentage of the GDP when compared with the period prior to the Great Recession. However, countries such as Finland and Estonia, show an inflection point in their BERD in 2009.
The Estonian economy, like the Baltic economies, was heavily affected by the 2009 recession (Ruttas-Küttim & Stamenov, 2016). They faced this recession after years of economic growth, but with increasing exports, the economic growth became positive again in 2010 and 2011 (Ruttas-Küttim & Stamenov, 2016). The overall level of R&D investments in Estonia during the crisis in 2008-2009 demonstrates growth as a percentage of the GDP. They were spending 0.62% of the GDP in BERD in 2009. Surprisingly, these numbers doubled in 2011, with them spending 1.46%, and then decreased significantly again to 0.62% in 2014. This peak in BERD in 2011-2012 was due to a one-off investment boom in oil shale research and technology by a single company, Eesti Energia9. This indicates that Estonian BERD is concentrated to a limited number of companies (Ruttas-Küttim & Stamenov, 2016). The huge decrease in Estonia’s
[24]
BERD after 2011, as seen in Graph 2, was therefore due to an adjustment after the large
investment of that single company rather than an effect of the crisis.
Investing in R&D is the best way for companies to remain competitive and survive in today’s business environment. A decline in sales during a single year can hold back the results for the following years. As such, being competitive is important. Furthermore, it has been shown that companies spending more on R&D perform better during a recession than those that spend less (Dugal & Morbey, 1995).
However, US and Japanese companies suffered the impact more than the European companies. There was a slight decrease in private investment in R&D in Japan, that experienced a slow recovery, and in the US. According to the World Bank, the US is the most innovative country in the world, having half of the 50 most innovative companies in the world. It therefore dominates most R&D-intensive industries (IWULSKA, 2012). Fortunately, the impact of the crisis did not last long, as large R&D investors in the US rapidly recovered back, and even increased, their previous sales rates. As a result they augmented their R&D investments soon after the shock (OECD, 2012). They also implemented the Recovery and Reinvestment Act of 2009, which provided a “short-term economic stimulus for research and research infrastructure to strengthen the knowledge base for future economic growth in different areas” (OECD, 2012).
Studies have been conducted regarding the behavior of corporate R&D in periods of crisis. The EC have found that there are three types of companies: those that maintained their innovation activities, those that significantly reduced them, and those that have increased their R&D activities (Cincera, Cozza, Tubke, & Voigt, 2012). According to Arrow (1962), investment in R&D and innovation activities is a fixed factor of production due to their strategic and longer-term nature (Arrow, 1962). Therefore, R&D investments tend not to be subject to financial constraints. However, it is important to distinguish between companies that have a constant investment in R&D, whose decisions will not be affected by external shocks like the crisis, and companies that want to engage in R&D, whose decisions to start new R&D investments might be altered by financial constraints (Cincera et al., 2012).
[25]
industrial mutation, that constantly revolutionizes the economic structure from within by destroying the old one and creating a new one” (Schumpeter, 1942). Consequently, the negative impact of a crisis on profitability forces the firms to search for higher productivity. Furthermore, crises bring new opportunities that can be taken advantage of by re-organizing and up-skilling R&D activities. In addition, if firms decide to cut R&D investment to reduce costs during a recession period, they increase the risk of falling behind competitors that continue innovating (Cincera et al., 2012).
In terms of public support, it seems that R&D was considered a priority in European countries in times of crisis, as both nominal and percentage rates were maintained for most of the Member States.
7. Conclusions
The main goal of this thesis is to discover the drivers, at the country level, that influence firms’ decisions to invest in R&D. The most unexpected result is that growth in GDP does not affect BERD, which means that businesses do not modify their R&D investments due to declines or increases in economic growth. This result raises the question regarding whether the Great Recession affected firms’ decisions to invest in R&D, as most countries experienced recessions in their economy in 2009. Because of the long- term nature of R&D investments, the yearly level of spending on ongoing projects is not usually affected by economic crisis. Conversely, there are no incentives for involvement in new projects when the economic growth rate improves.
[26]
research activities” (van Pottelsberghe de la Potterie, 2008), as it has been shown that
HERD also stimulates private R&D.
Regarding other variables, high technology exports also have an influence in BERD. However, the results show that trade openness does not influence the firms’ decisions.
Innovation is a major driver of employment, competitiveness, and more importantly of economic growth. Consequently, Research and Development must be a priority for policy makers. The positive relationship between R&D and growth is generally conducted by business R&D, and this is mainly because “public R&D is more focused on fundamental research than business R&D” (European Commission, 2011). Current European and national policies have set the targets for the average EU level of R&D expenditure to be 3% of the GDP and the EU average of private R&D expenditure to 2% of GDP. They intend to reach these objectives by increasing the intensity in R&D funding, which “will boost exports and competitiveness” (Sandu & Ciocanel, 2014). Government support is a strong source for stimulating private R&D, as the results have showed, and therefore this should seriously be considered if countries want their firms to increase R&D investment.
The European Commission is promoting these policies more than ever. Meanwhile this thesis is being finalized, they have published their country-specific recommendations within the European Semester, which is the framework for the coordination of economic policies across the European Union (European Commission, n.d.-c). For the first time, they advise policy makers to focus on R&D investments (Wallace, 2019) as “strengthening research and innovation activities are key for Europe’s growth” (European Commission, 2019).
The main conclusion is, therefore, that R&D intensity should be encouraged from the bottom, improving the education system and providing easier access to resources both for students and universities. Also, it would be necessary to increase the cooperation between education and businesses. This would provide companies with the high-skilled people they need to perform R&D and it would offer easier access to employment for students who are willing to perform these activities.
[27]
case of EU member states, also with the European policies. This would avoid duplication in policies and would ease the public funding process.
8. Annexes
1. Comparison of GOVERD, HERD and BERD between countries (2001 and 2017).
[29]
2. GDPppp vs BERDppp for OECD countries (2017). Source: OECD MSTI
Prepared by the author based on data from the OECD (OECD, n.d.-a).
3. Table of correlations (Stata output)
BERD HERD GOVERD GDPgrowth GDPpc HTECHEX EDUINDEX
9. Bibliography
Arrow, K. (1962). Economic Welfare and the Allocation of Resources for Invention. In
The Rate and Direction of Inventive Activity: Economic and Social Factors (pp. 609–
626). National Bureau of Economic Research, Inc.
Bebczuk, R. N. (2002). R&D Expenditures and the Role of Government Around the World.
Estudios de Economíca, 29(1), 109–121.
Bobillo, A. M., Rodríguez-Sanz, J. A., & Tejerina-Gaite, F. (2018). Corporate governance drivers of firm innovation capacity. Review of International Economics, 26(3), 721– 741. https://doi.org/10.1111/roie.12321
Carboni, O. A. (2017). The effect of public support on investment and R&D: An empirical evaluation on European manufacturing firms. Technological Forecasting and Social
Change, 117, 282–295. https://doi.org/10.1016/j.techfore.2016.11.017
Carter Hill, R., Griffiths, W. E., & Lim, G. C. (2012). Principles of Econometrics (Internatio). Wiley.
Cincera, M., Cozza, C., Tubke, A., & Voigt, P. (2012). Doing R&D or Not (in a Crisis), That Is the Question ... European Planning Studies, 20(9), 1525–1547. https://doi.org/10.1080/09654313.2012.709064
Coccia, M. (2011). Prometheus : Critical Studies in Innovation The interaction between public and private R & D expenditure and national productivity, 29(May 2012), 37– 41.
Cuervo-Cazurra, A., & Annique Un, C. (2010). Why some firms never invest in formal
R&D. Strategic Management Journal, 31(7), 759–779.
https://doi.org/10.1002/smj.836
Cunha, F., Heckman, J. J., & Schennach, S. M. (2010). Estimating the Technology of Cognitive and Noncognitive Skill Formation. Econometrica, 78(3), 883–931. https://doi.org/10.3982/ECTA6551
Damijan, J. P., Knell, M., Majcen, B., & Rojec, M. (2003). The role of FDI, R&D accumulation and trade in transferring technology to transition countries: evidence from firm panel data for eight transition countries. Economic Systems, 27(2), 189– 204. https://doi.org/https://doi.org/10.1016/S0939-3625(03)00039-6
Dohoo, I. R., Ducrot, C., Fourichon, C., Donald, A., & Hurnik, D. (1997). An overview of techniques for dealing with large numbers of independent variables in
epidemiologic studies. Preventive Veterinary Medicine.
https://doi.org/10.1016/S0167-5877(96)01074-4
Dugal, S. S., & Morbey, G. K. (1995). Revisiting Corporate R&D Spending During a Recession. Research Technology Management, 38(4), 23–27.
Eesti Energia. (2012). Annual Report 2010-2011.
Erdal, L., & Göçer, İ. (2015). The Effects of Foreign Direct Investment on R&D and Innovations: Panel Data Analysis for Developing Asian Countries. Procedia - Social
[32]
European Commission. (n.d.-a). Europe 2020 strategy. Retrieved February 8, 2019, from https://ec.europa.eu/info/business-economy-euro/economic-and-fiscal-policy-
coordination/eu-economic-governance-monitoring-prevention-correction/european-semester/framework/europe-2020-strategy_en
European Commission. (n.d.-b). Horizon 2020. Retrieved June 16, 2019, from https://ec.europa.eu/programmes/horizon2020/en/what-horizon-2020
European Commission. (n.d.-c). The European Semester. Retrieved June 11, 2019, from https://ec.europa.eu/info/business-economy-euro/economic-and-fiscal-policy-
coordination/eu-economic-governance-monitoring-prevention-correction/european-semester_en
European Commission. (2010). Europe 2020: Commission proposes new economic
strategy in Europe. Press Release Database. Retrieved from
http://europa.eu/rapid/press-release_IP-10-225_en.htm
European Commission. (2011). Effect of the economic crisis on R&D investment. In
Innovation Union Competitiveness report (pp. 60–72). Retrieved from
https://ec.europa.eu/research/innovation-union/pdf/competitiveness-report/2011/chapters/part_i_chapter_2.pdf
European Commission. (2013). Innovation Union Competitiveness report 2013. Eur
25650 En. https://doi.org/10.2777/44320
European Commission. (2019). European Semester: country-specific recommendations. Brussels. Retrieved from https://ec.europa.eu/info/sites/info/files/2019-
european-semester-country-specific-recommendations-commission-recommendations-communication_en.pdf
Eurostat. (n.d.). Statistics Explained: Glossary. Retrieved from
https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Main_Page Eurostat. (2018). Smarter, greener, more inclusive? INDICATORS TO SUPPORT THE
EUROPE 2020 STRATEGY. 2018 Edition. Smarter, greener, more inclusive? Indicators to Support the Europe 2020 Strategy. https://doi.org/10.2785/101636
Falk, M. (2006). What drives business Research and Development (R&D) intensity across Organisation for Economic Co-operation and Development (OECD) countries?
Applied Economics, 38(5), 533–547. https://doi.org/10.1080/00036840500391187 Frascati Manual 2015. (2015). OECD. https://doi.org/10.1787/9789264239012-en
Furman, J. L., & Hayes, R. (2004). Catching up or standing still? National innovative productivity among ‘follower’ countries, 1978–1999. Research Policy, (33), 1329–
1354. Retrieved from
https://ac-els-cdn-com.proxy- ub.rug.nl/S0048733304001179/1-s2.0-S0048733304001179-
main.pdf?_tid=34dd1d09-888a-422c-bbd8-04514b4f9bd3&acdnat=1551435246_b97cadbfb1ae85e25ed1056d5e897339 Griffith, R., Redding, S., & Reenen, J. Van. (2004). MAPPING THE TWO FACES OF R&D:
PRODUCTIVITY GROWTH IN A PANEL OF OECD INDUSTRIES.
[33]
225–243. https://doi.org/10.1080/10438590290004555
Haddad, M., & Harrison, A. (1993). Are there positive spillovers from direct foreign investment?: Evidence from panel data for Morocco. Journal of Development
Economics, 42(1), 51–74. https://doi.org/https://doi.org/10.1016/0304-3878(93)90072-U
Hansen, H., & Rand, J. (2006). On the Causal Links Between FDI and Growth in
Developing Countries. The World Economy, 29(1), 21–41.
https://doi.org/10.1111/j.1467-9701.2006.00756.x
IWULSKA, A. (2012). Country Benchmarks: R&D Policy (The United States). In Golden
Growth: restoring the lustre of the European Economic Model. World Bank.
Kanwar, S., & Evenson, R. (2003). Does intellectual property protection spur
technological change? Oxford Economic Papers, 55(2), 235–264.
https://doi.org/10.1093/oep/55.2.235
Lach, S. (2000). DO R & D SUBSIDIES STIMULATE OR DISPLACE PRIVATE R & D? EVIDENCE FROM ISRAEL. NBER WORKING PAPER SERIES.
Lederman, D., & Maloney, W. F. (2003). R & D and Development. Retrieved from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=402480
Leiponen, A. (2005). Skills and innovation. International Journal of Industrial
Organization, 23(5), 303–323.
https://doi.org/https://doi.org/10.1016/j.ijindorg.2005.03.005
Lin, C., Lin, P., & Song, F. (2010). Property rights protection and corporate R&D: Evidence
from China. Journal of Development Economics, 93(1), 49–62.
https://doi.org/https://doi.org/10.1016/j.jdeveco.2009.04.006
OECD. (n.d.-a). Main Science and Technology Indicators (MSTI). Retrieved from https://stats.oecd.org/Index.aspx?DataSetCode=MSTI_PUB
OECD. (n.d.-b). OECD. Retrieved from https://www.oecd.org/about/history/
OECD. (n.d.-c). OECD Economic Outlook: Statistics and Projections. Retrieved from https://www.oecd-ilibrary.org/economics/data/oecd-economic-outlook-statistics-and-projections_eo-data-en
OECD. (2001). Glossary of Statistical Terms. Retrieved February 21, 2019, from https://stats.oecd.org/glossary/detail.asp?ID=399
OECD. (2012). Innovation in the Crisis and beyond. In OECD Science, Technology and
Industry Outlook 2012 (pp. 21–57). OECD.
OECD. (2018). OECD Economic Surveys: Spain. OECD Economic and Studies.
Parameswaran, M. (2010). International trade and R&D investment: evidence from manufacturing firms in India. International Journal of Technology and Globalisation,
5(1/2), 43. https://doi.org/10.1504/ijtg.2010.033665
Piva, M., & Vivarelli, M. (2009). The role of skills as a major driver of corporate R&D.
[34]
Prof, A., & Ustabaş, A. (2016). The Effects of R & D and High Technology Exports on Economic Growth : A Comparative Cointegration Analysis for Turkey and South Korea, 347–348.
Raghupathi, V., & Raghupathi, W. (2019). Exploring science-and-technology-led innovation : a cross-country study, 0.
Ruttas-Küttim, R., & Stamenov, B. (2016). RIO COUNTRY REPORT 2015: Estonia.
Sandu, S., & Ciocanel, B. (2014). Impact of R&D and Innovation on High-tech Export.
Procedia Economics and Finance, 15(14), 80–90.
https://doi.org/10.1016/s2212-5671(14)00450-x
Scherer, F. M., & Huh, K. (1992). Top managers education and R&D investment. Research
Policy, 21(6), 507–511. https://doi.org/https://doi.org/10.1016/0048-7333(92)90007-Q
Schumpeter, J. A. (1942). Capitalism, socialism and democracy. Capitalism, Socialism
and Democracy (2003rd ed.). Taylor & Francis e-Library.
Sokolov-Mladenović, S., Cvetanović, S., & Mladenović, I. (2016). R&D expenditure and economic growth: EU28 evidence for the period 2002–2012. Economic
Research-Ekonomska Istraživanja, 29(1), 1005–1020. https://doi.org/10.1080/1331677X.2016.1211948
Souitaris, V. (2002). Firm-specific competencies determining technological innovation:
A survey in Greece. R and D Management, 32(1), 61–77.
https://doi.org/10.1111/1467-9310.00239
The World Bank. (n.d.). World Bank Open Data. Retrieved from
https://data.worldbank.org/
Turen, U., & Gökmen, Y. (2013). The Determinants of High Technology Exports Volume:
A Panel Data Analysis of EU-15 Countries. International Journal of Management, Economics and Social Sciences 2304 – 1366 (Vol. 2).
United Nations. (n.d.). Education Index. Retrieved from
http://hdr.undp.org/en/content/education-index
van Pottelsberghe de la Potterie, B. (2008). Europe’s R&D: Missing the Wrong Targets?
Intereconomics, 43(4), 220–225. https://doi.org/10.1007/s10272-008-0254-y
Varsakelis, N. C. (2001). The impact of patent protection, economy openness and national culture on R&D investment: A cross-country empirical investigation.
Research Policy, 30(7), 1059–1068. https://doi.org/10.1016/S0048-7333(00)00130-X
Wallace, N. (2019). EU Commission calls on national governments to focus on research
and innovation investment. Science Business. Retrieved from
https://sciencebusiness.net/framework-programmes/news/eu-commission-calls-national-governments-focus-research-and-innovation
Wallsten, S. J. (2000). The Effects of Government-Industry R&D Programs on Private R&D: The Case of the Small Business Innovation Research Program. The RAND
[35]
Yanikkaya, H. (2003). Trade openness and economic growth: A cross-country empirical
investigation. Journal of Development Economics, 72(1), 57–89.
