Abstract
The literature on the impact of research output (RO) on economic growth (EG) has been rapidly expanding. However, the single growth processes of technological laggard countries and the mediating roles of human capital (HC) and structural change have been overlooked. Based on cointegration analyses and Granger causality tests over 40 years (1980–2019) for Portugal, five results are worth highlighting: (1) in the short run, RO is critical to promote EG; (2) the long run relation between RO and EG is more complex, being positive and significant in the case of global and research fields that resemble capital goods (Life, Physical, Engineering & Technology, and Social Sciences), and negative in the case of research fields that resemble final goods (Clinical & Pre-Clinical Health, and Arts & Humanities); (3) existence of important short run mismatches between HC and scientific production, with the former mitigating the positive impact of the latter on EG; (4) in the long run, such mismatches are only apparent for ‘general’ HC (years of schooling of the population 25 + years), with the positive association between RO and EG being enhanced by increases in ‘specialized’ HC (number of R&D researchers); (5) structural change processes favouring industry amplify the positive (long-run) association and (short-run) impact of RO on EG. Such results robustly suggest that even in technologically laggard contexts, scientific production is critical for economic growth, especially when aligned with changes in sectoral composition that favour industry.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
The influence of research output (RO) on economic growth (EG) is a key issue of science policy that serves to define research priorities and understand the R&D investment efficacy (Abramo et al., 2022; Inglesi-Lotz & Pouris, 2013; Lee et al., 2011; Pourghaz et al., 2023). In this context, assessing the extent to which RO effectively influences EG, and finding out the fields of science that are likely to impact more on countries’ EG, emerge as an essential undertaking of empirical analysis (Pinto & Teixeira, 2020; Jaffe et al., 2020; Barret et al. 2021).
Studies that analysed the impact of RO on EG can be grouped into two main categories. Those that consider a relatively large sample of countries over time (e.g., Dkhili & Oweis, 2018; Demir 2019; Pinto & Teixeira, 2020; Jack et al., 2021; Sepehrdoost et al. 2021; Azmeh, 2022; Pourghaz et al., 2023) and single country, longitudinal, studies (e.g., Inglesi-Lotz et al., 2014; Lee et al., 2011; Ntuli et al., 2015; Yasgül & Güris, 2016; Zaman et al., 2018). In general, both groups of studies suggest that global RO tend to be associated with countries’ economic prosperity. Notwithstanding, studies that are based on single country analysis have addressed mainly very well-positioned countries in terms of science from Europe (e.g., Germany, Netherlands, Switzerland, UK), America (e.g., Canada, USA), Asia (e.g., Japan, Singapore, South Korea, Taiwan) and Oceania (e.g., Australia) (Inglesi-Lotz et al., 2014; Lee et al., 2011; Ntuli et al., 2015), as well as emerging economies such as Brazil, Russia, India, China, and South Africa (Inglesi-Lotz et al., 2015; Lee et al., 2011; Pourghaz et al., 2023). Moreover, these single country-based empirical studies have analysed RO in global terms without discriminating the diversity of research areas (Inglesi-Lotz et al., 2014; Pourghaz et al., 2023) or have focused on a specific field (Jin, 2010; Jin & Jin, 2013; Yasgül & Güris, 2016; Zaman et al., 2018). To the best of our knowledge, no study has to date explored the impact of RO by different fields of research using a single country analysis, most notably, focusing on countries far from the science frontier such as Portugal. As Khan (2022, p. 157) exposes, laggard type of countries “… faces underrepresentation in the economic growth literature.”
The Portuguese case can be scientifically interesting because, on the one hand, it provides evidence of some laggardness in terms of technology and innovation performance (Teixeira & Fortuna, 2004, 2010; European Union, 2023). On the other hand, it has experienced notable dynamics in terms of RO (Heitor et al., 2014), although its performance in terms of human capital and structural change has been debatable (OECD, 2023; Teixeira et al., 2014). Despite the apparent dynamics in terms of scientific output, the studies that have analysed Portuguese EG overlooked such growth factor, failing to scrutinize the extent to which RO has contributed to EG.
This study aims to contribute to fill this gap by conducting an empirical analysis of the long-term relationship (and potential impact) of RO, globally and by fields of science, on Portugal’s EG. We further consider the mediating role of human capital and structural change in this relationship as we expect that the impact of RO on EG may be influenced by the level of human capital (Silva & Teixeira, 2011) and the productive specialization profile of the economy (Teixeira & Queirós, 2016). It is crucial to understand which types of research are aligned with the countries’ absorptive capacity and pace of structural change from a scientific and policymaker’s point of view. Such impacts have not yet been empirically tested.
Methodologically, in the line with the existing studies in this area (e.g., Jin, 2009, 2010; Lee et al., 2011), we have employed Johansen cointegration and Granger causality tests, involving time series of real GDP per capita, RO (global and by fields of research), human capital, and structural change, from 1980 to 2019.
This paper is organized as follows. “A literature review on the impact of RO on EG considering the mediating role of human capital and structural change” section summarizes the relevant literature. “Methodology” section defines and discusses the methodology. The empirical results are detailed in “Empirical results” section. Finally, the Conclusions put forward the study’s main contributions, limitations, and policy implications.
A literature review on the impact of RO on EG considering the mediating role of human capital and structural change
The impact of RO on EG: main hypotheses
Economic growth (EG) can be referred to as an increase in the capacity of an economy to produce goods and services, compared from one period to another (Raisová & Ćurþová, 2014). A rich corpus of theoretical approaches has analyzed the determinants of economic growth (Doré & Teixeira, 2023), which are often grouped into two main categories (Di Maio, 2013): the mainstream perspective, essentially associated to the supply side neoclassical (Solow, 1956; Swan, 1956) and endogenous (Lucas, 1988; Romer, 1986) growth theories, and a heterodox perspective, which include, among others, the structuralist (Pasinetti, 1981), the evolutionary/ neo Schumpeterian (Freeman & Louçã, 2001; Freeman et al., 1982; Hodgson, 1996; Nelson, 1998; Pérez, 1983), and the Keynesian demand-led (Harris, 1974; Kaldor, 1957) approaches.
The mainstream’s, supply side, EG factors include growth of production factors, technical progress, and innovation, being the latter often associated to externalities in human and technical capital formation (Ireland, 1994; Begg et al., 1999). Albeit there is a tendency to explain EG by reference to supply-side rather than demand-side forces (Smith, 2012), aggregate long run EG can be explained, according to alternative, heterodox, theoretical perspectives, most notably the Keynesian demand-led theory of growth, by changes in income and output associated with the adjustment of productive capacity to aggregate demand (Barbosa-Filho, 2000). Accordingly, EG is a complex process, entailing structural change of the economic system, in which key factors explaining EG, particularly, income distribution, technological progress, and political-economic institutions, contribute to EG through their effect on the growth in demand (Smith, 2018).
In both mainstream and heterodox perspectives, (labor) productivity growth is an important source of economic growth, with technological change being seen as an endogenous driver of productivity. Future technological needs critically depend on knowledge capital, which is built, to a large extent, upon scientific research output (RO) (Sarpong et al., 2023).
Scientific RO, most notably, codified knowledge associated to scientific publications (Kumar et al., 2016; Solarin & Yen, 2016), is one of the ways that new knowledge is produced (Inglesi-Lotz & Pouris, 2013; Ntuli et al., 2015; Yasgul & Güriş, 2016). Such knowledge is likely to engender positive externalities on the productive capacity of economies (Inglesi-Lotz & Pouris, 2013; Inglesi-Lotz & Pouris, 2013), generating innovations (Meo et al., 2013), improving the quality of human capital (Ntuli et al., 2015; Schofer et al., 2000) and, ultimately, leading to EG (Pegkas et al., 2019). Additionally, the amount of RO reflects, to a large extent, the capabilities of a country’s labour force and the potential of the economy for attracting foreign and domestic investments (Kumar et al., 2016).
Mainstream economic theories, namely the neoclassical and endogenous growth theories, formally explain the relationship between knowledge and EG (Solarin & Yen, 2016). In the neoclassical theory, knowledge associated to technology is exogenous, falling like ‘manna from heaven’ (Solow, 1956). In the new (endogenous) growth models, technical change as the outcome of deliberate efforts by profit-maximizing firms (Romer, 1986), is the result of the resources devoted to R&D activities, and of the degree of appropriability of innovative rents (Castellacci, 2004). Accordingly, increasing R&D, which in large part is constituted by scientific production (basic R&D), contributes to innovation and EG. Heterodox, historically oriented, approaches highlight the fact that EG is a complex process of evolution and transformation, shaped by the complex endogenous interactions between technology, economy, institutions, and social factors, characterized by reverse causality relations rather than a simple transition along a steady state growth path (Khan, 2022).
Although the importance of knowledge for EG has been long recognized at the theoretical level, empirically the literature on EG only more recently started paying attention to the impact of RO on EG (e.g., Kumar et al., 2016; Solarin & Yen, 2016; Yasgul & Güriş, 2016; Dkhili & Oweis, 2018; Pinto & Teixeira, 2020; Pourghaz et al., 2023).
While recognizing that economic dynamics of nations and the availability of R&D funds (which fosters RO), feed each other, highlighting the existence of a bidirectional relationship between RO and EG (Uyar et al., 2022), several studies have identified a positive relationship with causality running from RO to EG (e.g., Inglesi-Lotz et al., 2014; Jack et al., 2021; Lee et al., 2011; Ntuli et al., 2015; Solarin & Yen, 2016). Investment in R&D activities fosters the production of RO that is an open source for innovation (Inglesi-Lotz et al., 2015; Ntuli et al., 2015; Solarin & Yen, 2016) that can lead to higher EG through the increase in productivity and labour (Ntuli et al., 2015; Solarin & Yen, 2016).
The current debate on EG and RO has mostly focused on which area of scientific knowledge has best promoted EG (Antonelli & Fassio, 2016; Azmeh, 2022; Jaffe et al., 2020; Pinto & Teixeira, 2020). Different fields of knowledge are likely to impact differently on the countries’ EG (Jin & Jin, 2013; Yasgul & Güriş, 2016). Following the framework of Antonelli and Fassio (2016), we identify two types of scientific knowledge, ‘capital good’ and ‘final good’. Capital good works as “intangible capital and intermediary inputs”, i.e., as a necessary input in the production of other goods (Antonelli & Fassio, 2016, p. 559). Promoting this type of knowledge can foster technological change and might lead to a rise in EG, as it is characterized by high levels of appropriation and a wider scope of application, being usually associated to hard (e.g., life and physical sciences, engineering, and technology) and social (e.g., economics and business) sciences (Antonelli & Fassio, 2016; Pinto & Teixeira, 2020). Hard sciences related knowledge tends to contribute most to EG because it leads directly to the introduction of technological innovations in a varied array of industries; and social science-related knowledge fosters organizational innovations and improvements in business practices, being fundamental to EG (Antonelli & Fassio, 2016; Pinto & Teixeira, 2020).
The second type of knowledge that can be treated as a ‘final good’, with low levels of appropriation and limited capacity for application, is often associated to the humanities and/ or to the medical sciences. Its impact on EG is small when compared to the previous type of knowledge as it tends to contribute mostly to increasing the utility of final consumers instead of directly increasing EG (Antonelli & Fassio, 2016).
Complementarily, Jaffe et al. (2020) showed that higher productivity in the basic sciences, namely physics and chemistry, induces stronger impact on EG when comparing with the relatively lower productivity, and lower growth impact, of the applied sciences, most notably, medicine and pharmacy.
We hypothesize that:
H1
RO positively impacts on EG.
H1a
The impact of RO on EG is elevated in the areas of science where knowledge is like capital goods than in those resembling final goods.
The mediating role of human capital and structural change: further hypotheses
An economy characterized by high levels of human capital (education/ training) tends to be more productive (Bodman & Le, 2013; Wößmann, 2003), leveraging EG (Jin & Jin, 2013). The theoretical models of human capital (Becker, 1962; Mincer, 1958; Schultz, 1961) establish that investment in knowledge and human capital can directly lead to increases in productivity and, consequently, boost EG. By acting as a productivity booster for research activities, human capital indirectly enhances RO (Pinto & Teixeira, 2020; Silva & Teixeira, 2011). Thus, highly human capital endowed countries are possibly more efficient in performing R&D activities, which lead to increased levels of RO (Pinto & Teixeira, 2020; Romer, 1990; Teixeira & Fortuna, 2010).
According to Inglesi-Lotz and Pouris (2013), the channel that can explain the impact of improved human capital on EG through RO is the following: better human capital leads to improvements in the production of research, which generates more and/ or a better knowledge basis and consequently promotes EG.
Based on the above, we hypothesize that:
H2
Human capital enhances the impact of RO on EG.
Structural change, defined as a change in the economy’s productive structure (Quatraro, 2010), is as an important determining factor of EG (Silva & Teixeira, 2011). The effect can be direct (Quatraro, 2009, 2010) or indirect through the production of RO (Pinto & Teixeira, 2020). The indirect impact occurs essentially when there is a match between changes in productive structure and the evolution of RO by scientific areas (Quatraro, 2010; Silva & Teixeira, 2011). By match we mean that the research activities developed in a country are aligned with the current needs of the industries, reflecting science and industry close linkages that can revitalize ideas and knowledge, which bolster higher economic performance.
Accordingly, we hypothesize that:
H3
Structural change favouring industry increases the impact of RO on EG.
The impact of RO on EG: overview of the empirical results
Global research output
The impact of RO on EG has mainly addressed in global terms (e.g., Inglesi-Lotz et al., 2015; Ntuli et al., 2015; Pourghaz et al., 2023; Solarin & Yen, 2016) or considering a limited number of specific fields of research, such as Chemical Engineering (Hart & Sommerfeld, 1998), Economics (Jin, 2009, 2010), Biotechnology (Yasgül & Güris, 2016), Engineering (Jack et al., 2021), Economics and Business (Jin & Jin, 2013), and the Sciences and Social Sciences (Zaman et al., 2018). One exception to this pattern is the recent study by Azmeh (2022) who analyses, for 15 MENA (Middle East and North Africa) countries, the impact of RO on EG, globally and by 27 research fields.
The literature that analyse the global impact of RO on EG may be organized into two main types of studies.Footnote 1 Those that analyse relatively large samples of countries without discriminating the countries (e.g., Azmeh, 2022; De Moya-Anegón & Herrero-Solana, 1999; Dkhili & Oweis, 2018; Jack et al., 2021; Khan, 2022; Onyancha, 2020; Pourghaz et al., 2023; Solarin & Yen, 2016) and studies that have analysed countries individually either considering sets of several countries (e.g., Hatemi-J et al., 2016; Inglesi-Lotz et al., 2015; Kumar et al., 2016; Lee et al., 2011; Ntuli et al., 2015) or single country studies (e.g., Inglesi-Lotz et al., 2014; Odhiambo & Ntenga, 2016).
The studies that consider relatively large samples of countries without discriminating them have achieved mixed results. The earlier study by De Moya-Anegón and Herrero-Solana (1999), focusing on Latin America countries, and resorting to correlation techniques, found a positive association between RO and EG. Involving a larger number of countries (169 and 39 countries, respectively) and more sophisticated econometric techniques (System GMM and Simultaneous longitudinal models, respectively) Solarin and Yen (2016) and Pourghaz et al. (2023) concluded that RO significantly foster EG. A similar result was achieved by Onyancha (2020) who investigated 48 countries in Sub-Saharan Africa, between 1991 and 2011, and further found evidence of reverse causality (that is, EG also positively and significantly impacted EG). In contrast, Dkhili and Oweis (2018), examining 43 countries in Sub-Saharan Africa in the period 1996–2015, and Khan (2022), scrutinizing a sample of 82 low- and middle-income countries for the period 2005–2019, failed to encounter a statistically significant relation between RO and EG. Considering the period of analysis 2000–2017, Azmeh (2022) found that RO, when measured in terms of the number of publications, was detrimental to the EG of 15 MENA countries whereas the number of citations, a proxy for RO quality, emerged positively and significantly relate to EG.
Regarding the studies that focus on RO in global terms and analyse the countries individually, it is possible to find some examples of one-way causality from RO to EG (Hatemi-J et al., 2016; Inglesi-Lotz et al., 2015; Lee et al., 2011; Ntuli et al., 2015). Lee et al. (2011), who employed time series methodologies, analysed 25 countries individually between 1981 and 2007. Their results evidenced that the causality ran from RO, measured by the number of publications, to EG, measured by nominal GDP, in the case of Austria, Australia, Germany, the Netherlands and India, despite these countries’ different competitive advantages in scientific research. Similar results were found by Inglesi-Lotz et al. (2015) for India, which is often considered a new emerging R&D destination for international projects in different fields of research and stands among the fastest-growing of the emerging BRICS countries. In their analysis, Inglesi-Lotz et al. (2015) employed the bootstrap panel Granger causality approach for the period between 1981 and 2011. These authors considered the real GDP, instead of the nominal GDP used in Lee et al. (2011), and RO was measured by the share of the country’s number of publications to the rest of the world. In contrast to Lee et al. (2011), who found a causality effect from RO to EG for Austria, Australia, Germany, and the Netherlands, Ntuli et al. (2015) found no causality in that direction for those countries, and Hatemi-J et al. (2016) for the German case. Such disparate results can be related to differences in the measurement of EG (nominal versus real GDP). Using the same method and period as Inglesi-Lotz et al. (2015), Ntuli et al. (2015) found positive causality from RO (total number of articles published) to EG for Finland, Hungary, Mexico, and the USA. The latter country was also analysed by Inglesi-Lotz et al. (2014), who again encountered a positive causality running from RO (share of the country’s number of publications to the rest of the world) to EG for a similar period, also using a Granger causality relationship indicated by the bootstrap rolling window causality tests. In contrast, the reverse direction of causality, i.e., from EG to RO was found by Lee et al. (2011) and Kumar et al. (2016). The latter authors used research publications per worker to proxy the RO and real GDP per worker for EG between 1981 and 2012, evaluating the Chinese and US case based on cointegration analyses. Hatemi-J et al. (2016) found no causality for almost all the G7 countries including the United States, Canada, France, Germany, Japan, and Italy, in the period from 1981 to 2012. The United Kingdom is another case in which some discrepancies were found among the existing studies: Hatemi-J et al. (2016) discovered causality from RO to EG, whereas Ntuli et al. (2015) identified the reverse direction, and Lee et al. (2011) found no causality.
Exploring the case of South Africa, Inglesi-Lotz and Pouris (2013) and Odhiambo and Ntenga (2016) found causality from RO to EG. In contrast, Inglesi-Lotz et al. (2015), using the same proxy as Inglesi-Lotz and Pouris (2013) but with a wider period (1981 to 2011), found no causality for South Africa.
In summary, and based on the extant empirical evidence, it is not possible to establish a clear relationship between EG and the scientific performance of each country, and the direction of causality between RO and EG.Footnote 2
Research output by fields
In the small set of studies that analyse RO by fields of knowledge—Biotechnology, Chemical Engineering, Economics, Engineering, Sciences and Social Sciences—and countries individually, evidence was found that in the field of Chemical Engineering, publications have a strong positive correlation with EG in 5 countries: USA, Canada, Great Britain, Australia, and India (Hart & Sommerfeld, 1998). Additionally, Jin (2009), focusing on the field of Economics for 5 East Asian countries (South Korea, Taiwan, Hong Kong, Japan and Singapore) between 1969 and 2004, uncovered that the causality ran from RO (publication per million people in Economics) to EG (nominal GDP) only in the case of South Korea and Taiwan. According to Jin (2009), such an outcome is consistent with the countries’ investment in purchasing overseas publications to be competitive with foreign universities. Exploring the case of Japan, for the period from 1970 to 2004, Jin (2010) found, similarly to his previous study (Jin, 2009), evidence of reverse causality, that is, causality running from EG to RO.
Regarding the relationship between Turkish EG and its RO in Biotechnology over the period from 1981 to 2013, Yasgül and Güris (2016) found, based on bootstrapped Granger causality analysis, that causality ran from RO to EG. This means that the RO in the field of Biotechnology was one of the factors that led to EG for the period in analysis. According to the authors, this field of research involves new technologies, requires interdisciplinary research and potential dissemination to the traditional sectors, thus generating EG.
Exploring the relationship between research productivity (number of publications in the physical and social sciences, research and development (R&D), expenditures and researchers involved in R&D activities) and EG (real GDP) in the period from 1980 to 2011 using cointegration and Granger causality, Zaman et al. (2018) found causality going from the number of publications to EG in Turkey, Russia, South Korea, Canada, the UK, and China.
Considering a larger number of countries (56), Jack et al. (2021) found that the number of publications in Engineering is positively and significantly related to EG; however, whereas the Asian countries get benefits from specialization in Engineering research, for the Latin American countries no effects are found. Also analysing a relatively large number of countries, Jin and Jin (2013), Jaffe et al. (2013) and Antonelli and Fassio (2016) have found some heterogeneity between groups of scientific fields. Specifically, Jin and Jin (2013) concluded that RO in Basic Science and Engineering evidenced a larger effect in EG than RO in Economics and Business, although both positively contribute to countries’ EG. To some extent like these results, Jaffe et al. (2013) evidenced that countries with higher relative productivity in Basic Sciences growth more compared to countries with a higher relative productivity in Applied Sciences. The results obtained by Antonelli and Fassio (2016) suggest that fields that resemble to capital goods (hard sciences and social sciences) contribute more to total factor productivity growth than fields that resemble to final goods, such as Medical related Sciences or Arts and Humanities. The results obtained by Inglesi-Lotz and Pouris (2013), Laverde‑Rojas and Correa (2019), and Azmeh (2022) seem to confirm the stronger growth effect of the fields of science that resemble to final goods, with the latter two studies evidencing some diversity of results of the impact of RO on EG according to countries’ level of development/ income. Indeed, Laverde‑Rojas and Correa (2019) demonstrate that for low-income countries both the global RO and Biochemistry, Genetics and Molecular Biology, Engineering, and Mathematics RO failed to be statistically and significantly related to EG, whereas in high-income countries RO, global and by scientific fields, significantly promotes EG. Focusing on MENA countries, a technology and innovation laggard region (Morrar, 2019), Azmeh (2022) demonstrated that for 22 out 27 scientific fields the impact of RO on EG is either not significant (14 research fields) or negative (8 research fields); the few scientific fields that appears to foster MENA countries’ EG include those that Antonelli and Fassio (2016) named fields that resemble to capital goods, most notably Business, Management and Accounting, Computer Science, Energy, Mathematics, and Physics and Astronomy.
Empirical evidence for Portugal
Several studies have addressed Portugal’s long-term growth in isolation at the aggregate/national level (e.g., Santos et al., 2018) or regional level (Manso et al., 2015). However, most of such studies has neglected to address determinants specifically related to RO. In fact, these studies have focused mainly on determinants related to macroeconomic conditions (Santos et al., 2018), international trade (Andraz & Rodrigues, 2010), and labour and demographic conditions (Morgado, 2014).Footnote 3
Although there are no studies on Portuguese EG that have explored the impact of RO/scientific knowledge, Teixeira and Fortuna (2004), using vector autoregressive and cointegration analyses, assessed the impact of R&D, which includes RO, on Portugal’s long-run growth between 1960 and 2001. The authors concluded that R&D intensity, which reflected internal innovation efforts, was crucial to the EG process in Portugal during that period. Moreover, Teixeira and Fortuna (2004, 2010) showed that human capital bolstered the impact of the internal stock of knowledge/ innovation capability and importing advanced technology from abroad on EG. In summary, human capital acted as a mediating factor in the relation between the internal stock of knowledge/innovation capability/importing advanced technology from abroad and EG.
Some studies have identified structural change as a critical factor for Portuguese EG (e.g., Lains, 2008; Rocha, 1997). Concerning the period from 1960 to 1970, structural change (i.e., increases in industry and decreases in agricultural employment/ product shares) was related to the acceleration of Portuguese EG (Rocha, 1997). Analysing the period between 1960 and 2004, Lains (2008) evidenced that in Portugal the change in the structure of employment occurred due to a reduction in the labour force share employed in traditional sectors (e.g., agriculture, forestry, fishing; textiles, leather, footwear, and clothing; food, drink, and tobacco) and an increase of that share in modern sectors (e.g., non-market services, other services).
Methodology
Main hypotheses to be tested, relevant variables and their proxies
The main purpose of this study is to evaluate the impact of research output (RO), global and by field of research, on Portuguese economic growth (EG), mediated by human capital and structural change, from 1980 to 2019. As detailed in “A literature review on the impact of RO on EG considering the mediating role of human capital and structural change” section, we seek to test four main hypotheses:
H1
RO positively impacts on EG.
H1a
The impact of RO on EG is elevated in the areas of science where knowledge is like capital goods than in those resembling final goods.
H2
Human capital enhances the impact of RO on EG.
H3
Structural change favouring industry increases the impact of RO on EG.
Accordingly, we need to select the indicators (proxies) and the corresponding data sources for the relevant variables, most notably: Economic Growth (EG), Research Output (RO), Human Capital (HC), and Structural Change (SC).
The variable proxy for EG is often expressed both in levels (GDP per capita) (e.g., Jin, 2010; Jaffe et al., 2013; Jack et al., 2021) or in growth rates (annual growth rate of GDP per capita) (e.g., Azmeh, 2022; Jin & Jin, 2013). Given that the level of GDP per capita is more adequate to capture differences in welfare in the long run (Hall & Jones, 1999), in our baseline estimations we opted for levels instead of growth rates.Footnote 4 Several studies related to RO and EG (e.g., Jin, 2010; Odhiambo & Ntenga, 2016) and empirical studies based on the neoclassic theory (see Barro & Sala-i-Martin, 1997) have used the GDP (per capita) in levels.
Real GDP per capita (constant 2016 prices, in €) is collected by the Portuguese National Statistical Institute (INE) and it is available in the platform PORDATA. Figure 1 depicts the time series of the Portuguese real GDP per capita, which evidence an upward trend between 1980 to 2019. At constant prices, in 2019, the Portuguese GDP per capita was 19,731 euros, more than double that of 1980, which was 9463 euros, and on average each Portuguese had 15,179 euros. Since 2003, worsened by the great recession following the 2008 financial crisis, Portugal experienced a substantial decrease in growth and a decrease (from 2008 to 2013) of its real GDP per capita, in part explained by the strong restrictive policies (monetary, income and fiscal), associated to the Economic Adjustment Programme (2011–2014) negotiated between Portuguese authorities and the International Monetary Fund, European Commission, and the European Central Bank (the ‘Troika’). After 2015, Portugal observed a recovery in terms of standard of living.
Evolution of the Portuguese real GDP per capita (constant 2016 prices, in €) Source: INE, in PORDATA
The measurement of RO is not a simple task (Inglesi-Lotz & Pouris, 2013). Two main proxies have been proposed: patents and research publications (Inglesi-Lotz et al., 2018). There are several drawbacks that advise against the use of patents as a proxy for RO. First, even in highly advanced countries, patents are a very small part of the outcome of research activity (Inglesi-Lotz & Pouris, 2013; Lee et al., 2011). Secondly, in countries far from the technological frontier, innovative and research activities seldom involve patents given the productive specialization of such countries (mainly based on low or medium–low technology-based industries) and the embryonic stage of intellectual property rights institutions (Yasgul & Guris, 2016).
There are alternative metrics to measure research publications: number of articles published (e.g., Jin, 2009; Lee et al., 2011; Solarin & Yen, 2016), the ratio of the total number of scientific publications in a given country to the total number of scientific publications in the world (Inglesi-Lotz et al., 2014, 2015), and the number of citations (Correa et al., 2021; Uyar et al., 2022) or High Quality Science Index (HQSI) (Allik et al., 2020), which attempt to reflect the relative quality of the publications (King, 2004).
Given that we intend to explore the impact of RO on EG, our focus is on how the quantity of publications is likely to impact on a country’s EG.Footnote 5 Moreover, as we are analysing only the Portuguese case, we have opted to consider the number of Portuguese scientific publications without relating it to the total number of scientific publications in the world but rather to the total population. Similar options were taken by studies in this area (e.g., Jin, 2009; Lee et al., 2011; Solarin & Yen, 2016).
We gathered the RO data from the InCites dataset from Web of Science (WoS) and Scopus. Although most of the extant literature focusing on RO opted for one bibliographic database [one exception is Jack et al. (2021)], given that these databases have distinct degrees of coverage by research fields (Suárez et al., 2022; Visser et al., 2021), we decided to consider both databases, which works as a sensitivity analysis/robustness check of the results.
We selected the articles by research area, grouped in the two main categories proposed by Antonelli and Fassio (2016): research fields that resemble capital good—Physical Sciences (PHYSIC), Engineering & Technology (ENG&TEC), Life Sciences (LIFE), Social Sciences (SOCIAL)-, and research fields that resemble to final good Clinical, Pre-clinical & Health Sciences (C&HEALTH), and Arts & Humanities (ARTS&H)-, using, in the case of WoS Subject areas, the GIPP scheme,Footnote 6 and in case of Scopus a rearrangement of the main Subject areas.Footnote 7
Figure 2 depicts the evolution of RO, in global terms and by research field, measured by the number of publications per 1000 inhabitants, gathered from WoS and Scopus bibliographic databases. Regardless the data source, the trend of publications is clearly upward in both global terms and by field of research. The average number of publications per thousand inhabitants is slightly higher in Scopus for Global RO, Physical Sciences, Clinical and Pre-Clinical & Health, and Social Sciences, whereas for Life Sciences, Engineering & Technology, and Arts & Humanities the number is higher in WoS. In 2019, according to the WoS (Scopus), Portugal registered 1.9 (2.6) publications per thousand inhabitants. The research fields registering higher number of publications in that year were Physical Sciences, and Life Sciences, and the lowest, Social Sciences, and Arts and Humanities.
Research Output (RO)—number of publications per 1000 inhabitants—globally and by fields of research, Portugal, 1980–2019 Source: Authors’ elaboration based on data from InCites dataset (Web of Science) and Scopus
Regarding the proxy for human capital stock, there are different alternatives in the literature, namely literacy rates, school enrolment ratios, and average years of schooling, to mention the most used (Abraham & Mallatt, 2022; Teixeira, 2005). The literacy rates omit significant elements of human capital, such as “numeracy, logical and analytical reasoning, and scientific and technological knowledge” taking only into consideration the elementary level (Le et al., 2005, p. 18). The school enrolment ratios just take into consideration the number of students that are registered at a specific level of education, thus reflecting the future and not the present human capital stock (Benos & Zotou, 2014; Le et al., 2005). The average years of schooling serves to quantify the accumulated investment in education and the total amount of formal education attained, being therefore considered a reasonable proxy for human capital stock (Benos & Zotou, 2014). Specifically, in line with other relevant studies (e.g., Moral-Benito, 2012; Teixeira & Queirós, 2016), we used in our baseline estimations the average years of schooling of the adult population (individuals aged 25 years or more), which tends to reflect the general stock of human capital of a country (Kessler & Lülfesmann, 2002).Footnote 8 The data comes from the United Nations Development Programme (2019), encompassing the period from 1990 to 2019, combined with data from de La Fuente and Domenech (2002) comprising the period from 1980 to 1990.
The lack of human capital has been frequently identified as a key obstacle to the Portuguese economic growth (Guichard & Larre, 2006). In 2019–2020, Portugal continues to be the most laggard country of the European Union in terms of the average number of years of formal education of the population aged 25 and plus with 9.3 years of schooling, very far from the 14 years of schooling of Germany. Still, the evolution of Portuguese human capital in the last forty years has been remarkable. As observed in Fig. 3, in 1980 an adult Portuguese citizen possessed, on average, less than 6 years of formal education, reaching in 2019, as referred, 9.3 years.
Structural change is defined as the evolution over a period of the weight of a given sector (e.g., primary, secondary, tertiary), in terms of employment, production or value-added, (Teixeira & Queirós, 2016). In this study, we considered the weight of industry in total production (Fig. 4). The data combines information from INE, in Banco de Portugal (1980 to 1995) and INE, in PORDATA (1996 to 2019).
Structural change (weight of the industrial product in total product), Portugal, 1980–2019 Source: 1980–1995—INE, in Banco de Portugal, Séries Longas do BdP; 1996–2019—INE, in PORDATA
Over the period of 40 years, between 1980 and 2019, Portugal experienced considerable change in its economic structure, in which the weight of the industrial product in total product decreased from 28% to 17.5%. However, the evolution was not linear. From 1988 to 2009, the share of industry fell considerably, reaching its lowest Fig. (12.6%) in 2009, but afterwards it was observed some re-industrialization, with the share of industry increasing to 17.5% in 2019 (Fig. 4).
The econometric specification and the estimation technique
Existing research on the effects of RO on EG of individual countries use both single country samples (Inglesi-Lotz & Pouris, 2013; Inglesi-Lotz et al., 2014; Jin, 2010; Odhiambo & Ntenga, 2016; Yasgül & Güris, 2016) and multi-country samples (Azmeh, 2022; Jack et al., 2021; Jin, 2009; Khan, 2022; Kumar et al., 2016; Lee et al., 2011; Pourghaz et al., 2023; Zaman et al., 2018), involving the analysis of annual time series. Such research has applied different methods of analysis: vector autoregressive (VAR) (Inglesi-Lotz et al., 2014; Lee et al., 2011; Zaman et al., 2018); cointegration (Johansen tests) and Granger causality (Yasgül & Güris, 2016), autoregressive distributed lag (ARDL) (Inglesi-Lotz & Pouris, 2013; Kumar et al., 2016; Odhiambo & Ntenga, 2016), and System GMM (Azmeh, 2022; Dkhili & Oweis, 2018; Jack et al., 2021; Solarin & Yen, 2016). In the case of single country analysis, the preferred econometric techniques have been cointegration Johansen tests, complemented by Granger causality tests.
The starting point in the time series analysis (for selecting the appropriate technique for analysing time series) is identifying whether the relevant variables are stationary or non-stationary, that is, if their value tend to revert to their long-run average value or not. If all the variables of interest are stationary, ordinary least square (OLS) or VAR models can provide unbiased estimates; however, if they are non-stationary or of mixed type (some are stationary and others are non-stationary), Johansen cointegration test or Autoregressive distributed lags (ARDL) models are, respectively, preferable (Shrestha & Bhatta, 2018). In the case cointegration exists, this implies that there is a long-term, or equilibrium, relationship between the relevant variables (Johansen & Juselius, 1990), and expectedly a causality among these variables in at least one direction (Engle & Granger, 1987).
The following equation captures the reduced form of the relationship between the variables under analysis:
where t represents time, \(u\) the random perturbation term, and Y, RO, HC, and SC the proxies for economic growth (EG), research output (RO), human capital (HC), and structural change (SC), respectively.
The econometric specification in (1) is estimated for global RO and by scientific area, namely: Physical Sciences (PHYSIC), Engineering & Technology (ENG&TEC), Life Sciences (LIFE), Social Sciences (SOCIAL), Clinical, Pre-clinical & Health Sciences (C&HEALTH), and Arts & Humanities (ARTS&H).
Empirical results
Unit root tests
For assessing whether the relevant variables are stationary or non-stationary, extant literature has usually employed unit root tests, most notably the augmented Dickey-Fuller (Dickey & Fuller, 1981), the Phillips-Perron (Phillips & Perron, 1988), and the KPSS tests (Kwiatkowski et al., 1992).
The visual inspection of the variables in levels and first differences suggests that the variables in levels are non-stationary (i.e., have a trend), whereas in the first differences they are stationary.Footnote 9 This evidence is corroborated by the formal unit root tests. The ADF tests whether a variable follows a unit-root process, being the null hypothesis that the variable contains a unit root (i.e., is non-stationary) against the alternative that the variable was generated by a stationary process (Dickey & Fuller, 1979). As shown in Table 1, we cannot reject the null hypothesis that the variables in levels have a unit root, but we reject the null hypothesis that the variables in their first differences have a unit root. Similar results are obtained when we use the Phillips–Perron unit-root test. Both tests confirm that all the variables in levels are non-stationary, i.e., the null hypothesis that there is one-unit root cannot be rejected considering the variables in levels, whereas in their first differences the null hypothesis is rejected, that is, they are stationary in the first differences. In summary, according to the ADF and Phillips-Perron tests, all the variables are integrated of order one, I (1). Thus, the series can be cointegrated (Dickey et al., 1991), in other words, there can be one or more stationary linear combinations of the series, pointing to a stable long-run relationship between them.
Johansen cointegration test and long-run relationships
This study intends to assess if there is a statistically significant long-run relationship between RO (global and by scientific field—PHYSIC, ENG&TEC, LIFE, SOCIAL, C&HEALTH, and ARTS&H), human capital (HC), structural change (SC), and EG. The Stata® 18 software has been used to perform the estimations.
The unit root tests (“Unit root tests” section) show that all the relevant variables are integrated of the same order, I (1). Thus, we can proceed with testing whether there are long run cointegrated relationships between the relevant variables using the Johansen cointegration test (Johansen & Juselius, 1990).
To uncover the number of cointegration vectors, we apply the trace test which tests the null hypothesis of r cointegrating vectors against the alternative hypothesis of n cointegrating vectors. The trace test generally rejects the null hypothesis that EG, RO, and the interaction variables between RO and human capital/ structural change have no cointegrating relationship (that is, the null hypothesis of the number of linearly independent cointegrating relationships (r) is 0) at the 5% level (Table 2). The trace test rejects the null hypothesis of no cointegration (r = 0) in all the models and fails to reject the null hypothesis of at most four (PHYSIC, SOCIAL, and ARTS&H) or five (Global RO, LIFE, ENG&TEC, C&HEALTH) cointegrating equations (r = 4 or 5). Thus, we accept the null hypothesis that there are 4 or 5 cointegrating equations in the multivariate models.
As we seek to establish the long-run influence of the other variables on EG and we do not have an underlying solid theoretical reasoning for imposing further restrictions on the parameters of the long-run relationships, we have opted for the Johansen normalization procedure which restricts the coefficient on EG to a unit, in line of the procedure adopted by Teixeira and Fortuna (2004, 2010). Table 3 presents the estimated long-run relationships (Vector Error Correction Model—VECM) between EG (EG—GDP per capita) and research output (RO) plus human capital (HC), structural change (SC), where RO interacted with HC and SC.
The overall Jarque–Bera statistics do not reject (at 5% significance) the null hypothesis that the disturbances in the VECM are normally distributed. Furthermore, according to the Lagrange multiplier test, we cannot reject the null hypothesis that there is no (second order) autocorrelation in the residuals. Finally, the eigenvalue stability condition shows that the estimated cointegration equations are stationary as required (the estimated roots are not close to 1). In short, the VECM related to the cointegration relation are well specified.
The results of our baseline estimations (Table 3) show that, in the long run, the relationship between global RO (number of publications per 1000 inhabitants) and EG (real GDP per capita) is positive and significant at the 1% level. Specifically, an increase of 1% in global RO is associated with a 0.541% (WoS)—0.606% (Scopus) increase in EG. Such results do not change when we consider the GDP per capita growth, as proxy for EG (Table A6, SED), or researchers involved in R&D activities, as proxy for HC (Table A12, SED). However, when we measure RO in terms of normalized citations (Table A9, SED), that is, when we focus on the ‘quality’/ scientific influence of the RO instead of quantity, we fail to encounter a statistically significant long-run relationship between RO and EG.
These results are aligned with the theoretical expectations that scientific knowledge and EG evolve jointly (De Moya-Anegón & Herrero-Solana, 1999; Solarin & Yen, 2016). Given that Portugal has been internationally recognized for the evolution in its scientific production in the last few decades (OECD, 2019; Simões, 2022), our results suggest that globally and over the long run such policy efforts seem to have payoff.
We further found that, in the long run, the relationship between RO for all the research fields classified as hard sciences and EG is positive and significant at the 1% level, using WoS and Scopus (except for Life Sciences) data; this result is robust for the proxy of EG (GDP per capita growth or normalized citations) and HC (except for Physical Sciences). Taking as the baseline models those estimated resorting to WoS data, the long run elasticity between RO and EG e particularly high in ENG&TEC and LIFE.
In contrast, the results regarding the relationship between the soft sciences RO (i.e., research fields that resemble to final goods) and EG suggest a significant (at the 1% level) and negative long run relationship between RO and EG, for both C&HEALTH and ARTS&H using WoS data, and for ARTS&H using Scopus data (C&HEALTH’s coefficient is statistically non-significant in this case). This result, however, is conditional on the proxy used for measuring HC in the case of C&HEALTH, and on the proxy used for measuring EG and RO in the case of ARTS&H. In concrete, when HC is measured by the number of researchers involved in R&D activities, that is, when we consider the specialized HC, the health-related RO comes significantly and positively related to EG in the long run.Footnote 10 Moreover, in the case of ARTS&H, when EG is measured by the growth of the GDP per capitaFootnote 11 or RO is measured by the normalized citations (‘quality’ of the RO),Footnote 12 the long run elasticity between RO and EG is statistically significant and positive.
By research fields, the baseline results based on WoS data are partially in line with those of Antonelli and Fassio (2016), who explored the contribution of several fields of research to EG ranging from the hard to soft sciences and concluded that hard sciences contribute more to total factor productivity (TFP). However, they did not find a negative relationship in any of the areas of RO and EG.
Albeit the mainstream approach, or what Schofer et al. (2000) calls the ‘instrumental perspective’, conveys that RO tends to improve socioeconomic efficiency and generate new products through the production of new knowledge, innovation, and technical applications, more institutional-related approaches challenge the picture of a simple and straightforward positive relationship between RO and EG (Shenhav & Kamens, 1991). On the one side, we may fail to encounter a statistically significant relation between RO and EG because in the case of innovation laggard countries links between science and industry (‘technical development’) are too weak, or RO produced is useless/mismanaged, and thus the incentives to produce new knowledge are too limited (Arrow, 1962). On the other side, RO may have significant and negative impact on EG (Schofer et al., 2000; Shenhav & Kamens, 1991), particularly in areas that resemble more to final goods, namely C&HEALTH and ARTS&H, as RO expansion “legitimate a broad progressive agenda of social amelioration (e.g., by identifying environmental and health problems, and social welfare and human rights issues) that can result in regulation and direct constraints on productive economic activity…” (Schofer et al., 2000, p. 866).
Except for the model related to global output, when significant (i.e., in the models of ENG&TEC, SOCIAL, C&HEALTH, and ARTS&H), human capital emerges in the baseline estimations with WoS data as positively associated with EG in the long run. For Scopus data (Table 3) or WoS data but measuring HC by the number of researchers involved in R&D activities (Table A12, SED), that is, ‘specialized HC, the long run relation between HC and EG is significant and positive both for global and by research area. This suggests that an economy characterized by high levels of human capital (education/training), in particular ‘specialized human capital’, tends to be more productive and innovative, leveraging EG (Bodman & Le, 2013; Wößmann, 2003). We further found that, except for the soft sciences (C&HEALTH, and ARTS&H), the long-run relationship between ‘general’ human capital as a mediator of RO and EG is statistically non-significant or significant and negative (Table 3). However, when we use the indicator of ‘specialized HC’, that is, number of researchers involved in R&D activities (Table A12, SED), the long run effect of RO (global and by research area) on EG comes (apart from SOCIAL) enhanced. This evidences that for Portugal, between 1980 and 2019, increases in the years of schooling mitigated the positive association between RO and the real GDP per capita, whereas increases in the number of researchers involved in R&D activities amplifies the positive long run association.
Regarding the association between structural change and EG, it was found to be positive and significant for LIFE, ENG&TEC and SOCIAL. Since the 1950s and 1960s, many countries increased their living standards by reallocating resources from agriculture to higher-productivity sectors, namely the industrial and services sectors (Gabardo et al., 2017). These shifts led to a positive structural change that boosted productivity and, consequently, sustained EG paths (Martins, 2019).
Moreover, we found that, apart from ARTS&H, there is a positive and significant long-term correlation between structural change, RO and EG. In other words, in the long run, structural change tends to amplify the positive association between RO and EG.
Granger (non-)causality test
In the previous section, we confirmed the existence of at least one cointegration relationship between economic growth (EG), research output (RO), human capital (HC), structural change (SC), and that RO interacted with HC and SC. Thus, we proceeded to test for Granger (non-)causality.
Taking our core time series—economic growth (EG) and research output (RO)-, RO is said to Granger-cause EG if the latter can be better projected using the memoirs of both RO and EG. Thus, we can test for the non-existence of Granger causality by estimating the vector autoregressive models (VAR).
The results show that, in the short run, there is a positive causality running from global RO to EG, as the null hypothesis of Granger causality from global RO to EG is rejected at the 1% level of significance (see Table 4). We thus conclude that between 1980 and 2019, increases in global RO strongly fostered improvements in real GDP per capita (economic growth). Therefore, H1 (RO positively impacts on EG) is validated. Such results unambiguously show the critical relevance of RO production for Portuguese EG in the last forty years. The findings are in line with some earlier empirical studies that have identified causality running from RO to EG for several countries and periods: Australia, Austria, Germany, India, and The Netherlands from 1981 to 2007 (Lee et al., 2011); the USA, from 1981 to 2011 (Inglesi-Lotz et al., 2014; Ntuli et al., 2015); and Finland, Hungary, and Mexico (Ntuli et al., 2015).
The evidence for Portugal in the last forty years fits theoretical principles according to which a higher level of knowledge, resulting from research activity, promotes EG through the development of innovation, which leads to improvements in productive capacity and labour quality (Hatemi-J et al., 2016; Ntuli et al., 2015). It also agrees with disparate evidence and accounts from other studies focusing on Portugal’s EG. For instance, an increase in R&D intensity was extremely important to the EG process in Portugal during the period from 1960 to 2001 (Teixeira & Fortuna, 2004).
Analysing the relationship between RO by fields and EG, the results of Granger causality suggest that the RO of both ‘hard sciences’ and ‘soft sciences’ positively impacts on EG. Moreover, the impact is strong for both ‘hard sciences’ and ‘soft sciences’, with the magnitude of the impact being particularly stronger in Life Sciences and Clinical & Pre-Clinical Health.Footnote 13 As such, our data do not corroborate H1a (The impact of RO on EG is elevated in the areas of science where knowledge is like capital goods than in those resembling final goods).
These results are thus not completely aligned with the findings of Antonelli and Fassio (2016). These authors, exploring the period between 1998 and 2008 in 13 countries, concluded that knowledge associated to the hard and soft sciences leads to different effects on EG. Specifically, they found that the hard sciences, which produce knowledge with a high scope of application and appropriation can, to a larger extent than soft sciences (which produce knowledge with a smaller scope of application and appropriation), lead to increases in final consumer utility, promote technological change, and generate EG.
The interaction between human capital and RO (global and by areas) impacts significantly (Granger causes) but negatively on EG. This suggests that high levels of human capital mitigate the positive impact of RO (global and by areas) on EG. Consequently, H2 (Human capital enhances the impact of RO on EG) is rejected by our data. Such results are robust (see Tables A7, A10, A13 in SED) and evidence important mismatches between formal education and scientific production in Portugal in the short run. The increase in terms of the average number of years of formal schooling (and in the number of R&D researchers) does not seem to be aligned with Portuguese RO in terms of their joint short run impact on EG.
Structural change has unambiguously and positively (Granger) caused EG in Portugal in the last forty years. Moreover, and mostly important, high levels of structural change towards industry production has significantly leveraged the impact of global RO and PHYSIC, SOCIAL, and C&HEALTH (but not LIFE, ENG&TEC, and ARTS&H) RO on EG. Furthermore, RO (Granger) causes EG. In this vein, H3 (Structural change favouring industry increases the impact of RO on EG) is partially validated. These results are in line with findings by Pena-Vinces et al. (2019), who analysed South American economies between 2003 and 2013; they concluded that scientific capacity and manufacturing development had a larger combined effect on international competitiveness than their individual marginal effects.
Conclusion
It is unquestionable that all areas of research are important and can provide benefits to society (Sutherland et al., 2011) that go beyond the effects they may have in terms of EG (Antonelli & Fassio, 2016). However, assessing whether (global and by areas) RO impacts on EG and is aligned with human capital and structural change is fundamental for both the scientific and policy spheres.
Despite its late awakening Heitor & Horta (2013), Portugal has experienced significative dynamics in terms of RO over the last forty years (MCTES 2017; Simões, 2022). Notwithstanding, it remains a laggard country, characterized by considerable backwardness regarding technology and innovation performance, including weak links between science and industry (Teixeira, 2007; Teixeira & Monteiro, 2018). Furthermore, the impact of human capital development and structural change on the country’s productivity dynamics and growth processes has been debatable (Pereira & Lains, 2012; OECD, 2014).
Based on time series analyses for Portugal from 1980 to 2019 and employing cointegration and Granger causality analyses, this study assesses the role of RO in the EG performance of the country and examines whether structural change processes favouring industry and human capital stock have amplified or mitigated the direct impact of RO (global and by areas) on EG.
The study contributes to the literature on three main levels: theoretical, methodological, and empirical. At the theoretical level, the study explores and adapts, from a novel perspective, the contribution by Antonelli and Fassio (2016), considering RO in two groups of knowledge, hard sciences (LIFE, PHYSIC, ENG&TEC, and SOCIAL) and soft sciences (C&HEALTH, and ARTS&H). The former is characterized by knowledge as a capital good, whereas the latter is characterized by knowledge as a final good. At the methodological level, the study contributes to the scanty literature and employs time series analyses that consider the direct and indirect (interaction) effects of (global and by areas) RO on EG via human capital and structural change. At the empirical level, this study offers new and challenging evidence of the long- and short-run effects of RO on EG in a technologically laggard country.
The results of this study unambiguously underline the important role of RO in fostering EG in Portugal. Specifically, global RO and hard and soft sciences ROs significantly boosts EG in the short run (and in the long run in the former two cases). Additionally, structural change favouring industry emerged as a mediating factor that significantly amplifies the positive impact of RO on EG.
In the short run, high levels of (‘general’ and ‘specialized’) human capital mitigate the impact of RO on EG evidencing the existence of important human capital-RO mismatches. In the long run ‘specialized’ (i.e., the number of R&D researchers), but not ‘general’ (i.e., years of formal education of population aged 25 +), human capital amplifies the positive and statistically significant association between RO and EG.
These results have important policy implications. First, they suggest that to achieve higher EG it is essential for Portugal to invest in science, regardless of the scientific domain. Second, the long run significant and negative association between scientific areas that resemble final goods, namely Clinical & Pre-Clinical Health and Arts & Humanities, suggests that in these domains science addresses goals that are beyond strict economic growth, focusing instead on sustainable, social progress which may require new economic regulations that can be conflicting with economic growth (Luciani, 2020; Zhong et al., 2021). Third, policy incentives directed at fostering a strong industrial basis are likely to bolster EG effects derived from investments in science/RO. Thus, public policies should target specific instruments and programmes that promote the relationship between science and industry. Fourth, it is essential to overcome the short and long run mismatches between human capital and RO, which can be achieved by improving the dialogue between education, science, and industry, seeking to design education and training offers closer to industry needs, encouraging the effective integration of PhD holders in companies and the mobility of researchers between industry and academia (OECD, 2019). Finally, it is urgent to improve existing and/or implement new efficient mechanisms of transferring knowledge developed in universities and research labs to industry and the marketplace (Gibson & Naquin, 2011). In this latter dimension, the noticeable expansion of technological infrastructures (TTOs, Business Incubators, Science Parks) observed in the last twenty years in Portugal (Ratinho & Henriques, 2010; Cartaxo & Godinho, 2017) needs to be followed by an effective improvement in their efficiency levels in terms of technology transfer if significant growth effects are to be acquired (Teixeira & Monteiro, 2018).
Although this study conveys some novel contributions, it has several limitations that may comprise interesting paths for further research. First, besides assessing the impact of the quantity of RO (number of publications) and ‘quality’ (number of citations) of that RO on EG it is critical to go beyond the ‘instrumentalist’ perspective of the scientific production by analysing the role of institutions and the overreaching impact of science on social welfare, namely by scrutinizing the societal transformations that may generate long-term noneconomic benefits or even costs (Schofer et al., 2000; Shenhav & Kamens, 1991) associated with demographic and climate transitions. Second, it is important to consider the heterogeneity of human capital by including not only the aggregate number of schooling years, but also the stock of human capital in secondary and tertiary education and/or courses that could contribute to further elaborate on the human capital mismatches found. Finally, to the extent that RO may exhibit nonlinear and asymmetric behavior, one potential, challenging and interesting avenue for further research would involve resorting to nonlinear error correction models and the estimation of nonlinear, nonstationary relationships through parametric, nonparametric and semiparametric cointegration models that have been recently proposed (see Mnasri et al., 2023).
Notes
Table A2 in the Supplementary Electronic Document (SED) summarises these contributions.
See Table A3 in the Supplementary Electronic Document (SED).
See the synthesis of the literature on Portuguese growth determinants available in Table A1, in the Supplementary Electronic Document (SED).
Nevertheless, we undertook some robustness check/ sensitivity analyses by estimating all the relevant models using the growth of GDP per capita instead of the levels—see Tables A5–A7 in the Supplementary Electronic Document. For the long-term relations between EG and RO, with exception of the case of Arts & Humanities, results obtained are similar to those that used EG measured by the level of the GDP per capita.
We undertook several robustness check/ sensitivity analyses and estimated all the relevant models using the normalized citation impact, often considered a measure of the ‘quality’ of RO (see Azmeh 2022). For both long-term and short-term relations between EG and RO, results are similar to the baseline estimations (i.e., those that consider the number/quantity of articles published). The unique exception is the long-term relation between global RO and EG, which is statistically non-significant when we measure RO in ‘quality’ (number of citations per article) instead of ‘quantity’ (number of articles per 1000 inhabitants)—see Tables A8–A10 in the Supplementary Electronic Document (SED).
See http://help.prod-incites.com/inCites2Live/filterValuesGroup/researchAreaSchema/gippDetail/gippCategories.html, last accessed August 2023.
Scopus four main Subject areas are Life sciences, Physical Sciences, Social Sciences, and Health Science. To obtain comparable research fields to that of WoS, in Physical Sciences we excluded Engineering, considering the latter as an autonomous category, and in Social Sciences we excluded Arts and Humanities, considering the latter also an autonomous category.
As a robustness check, we also used the number of researchers involved in R&D activities as a proxy for more specialized type of Human Capital (HC). Although the results for the long-run core relation RO → EG did not differ significantly, the estimates associated to the HC variable and/ or its interaction with RO change dramatically in the case of research fields resembling capital goods, becoming positive and statistically significant—see Table A12 in the Supplementary Electronic Document (SED).
See all the relevant graphics in Table A4 in the Supplementary Electronic Document (SED).
See Table A12 in SED.
See Table A6 in SED.
See Table A9 in SED.
Robustness estimations (see Tables A7, A10, A13 in SED) confirm these results.
References
Abraham, K. G., & Mallatt, J. (2022). Measuring human capital. Journal of Economic Perspectives, 36(3), 103–130.
Abramo, G., D’Angelo, C. A., & Di Costa, F. (2022). Revealing the scientific comparative advantage of nations: Common and distinctive features. Journal of Informetrics, 16(1), 101244.
Allik, J., Lauk, K., & Realo, A. (2020). Factors predicting the scientific wealth of nations. Cross-Cultural Research, 54(4), 364–397.
Andraz, J. M., & Rodrigues, P. M. M. (2010). What causes economic growth in Portugal: Exports or inward FDI? Journal of Economic Studies, 37(3), 267–287.
Antonelli, C., & Fassio, C. (2016). Academic knowledge and economic growth: Are scientific fields all alike? Socio-Economic Review, 3, 537–565.
Arrow, K. (1962). Economic welfare and the allocation of resources for invention. In R. R. Nelson (Ed.), The rate and direction of inventive activity: Economic and social factors (pp. 609–626). Princeton University Press.
Azmeh, C. (2022). Quantity and quality of research output and economic growth: Empirical investigation for all research areas in the MENA countries. Scientometrics, 127(11), 6147–6163.
Barbosa-Filho, N. H. (2000). A note on the theory of demand-led growth. Contributions to Political Economy, 19(1), 19–32.
Barrett, P., Hansen, N.-J., Natal, J.-M., & Noureldin, D. (2021). Why Basic Science Matters for Economic Growth? Public investment in basic research will pay for itself. IMF Blog. Retrived August 18, 2023, from https://www.imf.org/en/Blogs/Articles/2021/10/06/blog-ch3-weo-why-basic-science-matters-for-economic-growth.
Barro, R. J., & Sala-i-Martin, X. (1997). Technological diffusion, convergence, and growth. Journal of Economic Growth, 2, 1–26.
Becker, G. (1962). Investment in human capital: a theoretical analysis. The Journal of Political Economy, 70(5, part 2), 9–49.
Begg, I., Dalum, B., Guerrieri, P., Pianta, M. (1999). The impact of specialization in Europe, in The Economic Challenge for Europe, (Eds) J. Fagerberg, P. Guerrieri and B. Verspagen. Aldershot: Edward Elgar.
Benos, N., & Zotou, S. (2014). Education and economic growth: A meta-regression analysis. World Development, 64, 669–689.
Bodman, P., & Le, T. (2013). Assessing the roles that absorptive capacity and economic distance play in the foreign direct investment-productivity growth nexus. Applied Economics, 45(8), 1027–1039.
Cartaxo, R. M., & Godinho, M. M. (2017). How institutional nature and available resources determine the performance of technology transfer offices. Industry and Innovation, 24(7), 713–734.
Castellacci, F. (2004). A Neo-Schumpeterian approach to why growth rates differ. Revue Économique, 55(6), 1145–1169
de la Fuente, A. & Doménech, R. (2002). Educational attainment in the OECD, 1960–95 (May 2002). SSRN: https://ssrn.com/abstract=317689.
De Moya-Anegón, F., & Herrero-Solana, V. (1999). Science in America Latina: A comparison of bibliometric and scientific-technical indicators. Scientometrics, 46, 299–320.
Demir, O. (2019). Does high tech exports matter for economic growth a panel approach for upper middle-income economies? AJIT-e. Bilişim Teknolojileri Online Dergisi, 9(31), 43–54.
Di Maio, M. (2013). Are mainstream and heterodox economists different? An empirical analysis. The American Journal of Economics and Sociology, 72(5), 1315–1348.
Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74, 427–431.
Dickey, D. A., & Fuller, W. A. (1981). Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica: Journal of the Econometric Society, 49, 1057–1072.
Dickey, D. A., Jansen, D. W., & Thornton, D. L. (1991). A primer on cointegration with an application to money and income. Federal Reserve Bank of St. Louis Review, 73, 58–78.
Dkhili, H., & Oweis, K. A. (2018). Does research output matter for economic growth in Sub Saharan African Countries? Quantity and Quality Analysis, 7(3), 221–242. https://doi.org/10.17583/rimcis.2018.3108
Doré, N. I. & Teixeira, A.A.C. (2023). The role of human capital, structural change, and institutional quality on Brazil’s economic growth over the last two hundred years (1822–2019). Structural Change and Economic Dynamics, 66, 1–12
Engle, R., & Granger, C. W. (1987). Co-integration and error correction: Representation, estimation, and testing. Econometrica, 55, 251–276.
European Union. (2023). European innovation scoreboard 2023. Publications Office of the European Union.
Freeman, C., Clark, J., & Soete, I. (1982). Politics unemployment and technical innovation. Frances Pinter.
Freeman, C., & Louçã, F. (2001). As time goes by: From industrial revolution to the information revolution. Oxford University Press.
Gabardo, F. A., Pereima, J. B., & Einloft, P. (2017). The incorporation of structural change into growth theory: A historical appraisal. Economia, 18(3), 392–410.
Granger, C. W. (1988). Some recent development in a concept of causality. Journal of Econometrics, 39(1–2), 199–211.
Guichard, S., & Larre, B. (2006). Enhancing Portugal’s Human Capital, OECD Economics Department Working Papers, No. 505. OECD Publishing, Paris.
Hall, R. E., & Jones, C. I. (1999). Why do some countries produce so much more output per worker than others? The Quarterly Journal of Economics, 114(1), 83–116.
Harris, D. (1974). The price policy of firms, the level of employment and distribution of income in the short run. Australian Economic Papers, 13, 144–151.
Hart, P. W., & Sommerfeld, J. T. (1998). Relationship between growth in gross domestic product (GDP) and growth in the chemical engineering literature in five different countries. Scientometrics, 42(3), 299–311.
Hatemi-J, A., Ajmi, A. N., El Montasser, G., Inglesi-Lotz, R., & Gupta, R. (2016). Research output and economic growth in G7 countries: New evidence from asymmetric panel causality testing. Applied Economics, 48(24), 2301–2308.
He, Q., & Xu, B. (2019). Determinants of economic growth: A varying-coefficient path identification approach. Journal of Business Research, 101, 811–818.
Heitor, M., & Horta, H. (2013). Science and technology in Portugal: From late awakening to the challenge of knowledge-integrated communities, In Higher Education in Portugal 1974–2009: A Nation, a (Eds.), Generation, G. Neave and A. Amaral. Chapter 8, Springer Dordrecht.
Heitor, M., Horta, H., & Mendonça, J. (2014). Developing human capital and research capacity: Science policies promoting brain gain. Technological Forecasting and Social Change, 82, 6–22.
Hodgson, G. (1996). An evolutionary theory of long-term economic growth. International Studies Quarterly, 40(3), 391–410.
Inglesi-Lotz, R., Balcilar, M., & Gupta, R. (2014). Time-varying causality between research output and economic growth in US. Scientometrics, 100, 203–216.
Inglesi-Lotz, R., Chang, T., & Gupta, R. (2015). Causality between research output and economic growth in BRICS. Quality and Quantity, 49, 167–176.
Inglesi-Lotz, R., Hakimi, A., & Pouris, A. (2018). Patents vs publications and R&D: Three sides of the same coin? Panel smooth transition regression (PSTR) for OECD and BRICS countries. Applied Economics, 50(45), 4912–4923.
Inglesi-Lotz, R., & Pouris, A. (2013). The influence of scientific research output of academics on economic growth in South Africa: An autoregressive distributed lag (ARDL) application. Scientometrics, 95, 129–139.
Ireland, P. N. (1994). Supply-side economics and endogenous growth. Journal of Monetary Economics, 33(3), 559–571.
Jack, P., Lachman, J., & López, A. (2021). Scientific knowledge production and economic catching-up: An empirical analysis. Scientometrics, 126, 4565–4587. https://doi.org/10.1007/s11192-021-03973-4
Jaffe, K., Caicedo, M., Manzanares, M., Gil, M., Rios, A., Florez, A., Montoreano, C., & Davila, V. (2013). Productivity in physical and chemical science predicts the future economic growth of developing countries better than other popular indices. PLoS ONE, 8(6), 1–10.
Jaffe, K., ter Horst, E., Gunn, L. H., Zambrano, J. D., & Molina, G. (2020). A network analysis of research productivity by country, discipline, and wealth. PLoS ONE, 15(5), 1–15.
Jin, J. C. (2009). Economic research and economic growth: Evidence from East Asian economies. Journal of Asian Economics, 20(2), 150–155.
Jin, J. C. (2010). Research publications, economic growth and causality: Japan’s experience. Pacific Economic Review, 15(5), 666–673.
Jin, J. C., & Jin, L. (2013). Research publications and economic growth: Evidence from cross-country regressions. Applied Economics, 45(8), 983–990.
Johansen, S., & Juselius, K. (1990). Maximum likelihood estimation and inference on cointegration—with applications to the demand for money. Oxford Bulletin of Economics and Statistics, 52(2), 169–210.
Kaldor, N. (1957). A model of economic growth. Economic Journal, 67, 591–624.
Kessler, A. S., & Lülfesmann, C. (2002). The theory of human capital revisited: On the interaction of general and specific investments, CESifo Working Paper, No. 776, Center for Economic Studies and ifo Institute (CESifo), Munich.
Khan, M. S. (2022). Absorptive capacities and economic growth in low-and middle-income economies. Structural Change and Economic Dynamics, 62, 156–188.
King, D. A. (2004). The scientific impact of nations. What different countries get for their research spending. Nature, 430, 311–316.
Kumar, R. R., Stauvermann, P. J., & Patel, A. (2016). Exploring the link between research and economic growth: An empirical study of China and USA. Quality and Quantity, 50(3), 1073–1091.
Kwiatkowski, D., Phillips, P. C., Schmidt, P., & Shin, Y. (1992). Testing the null hypothesis of stationarity against the alternative of a unit root. Journal of Econometrics, 54(1–3), 159–178.
Lains, P. (2008). The Portuguese economy in the Irish mirror, 1960–2004. Open Economies Review, 19(5), 667–683.
Laverde-Rojas, H., & Correa, J. C. (2019). Can scientific productivity impact the economic complexity of countries? Scientometrics, 120(1), 267–282.
Le, T., Gibson, J., & Oxley, L. (2005). Measures of human capital: A review of the literature. New Zealand Treasury Working Paper 05/10, pp.1–55.
Lee, L. C., Lin, P. H., Chuang, Y. W., & Lee, Y. Y. (2011). Research output and economic productivity: A Granger causality test. Scientometrics, 89(2), 465–478.
Lucas, R. E. (1988). On the mechanics of economic development. Journal of Monetary Economics, 22(1), 3–42.
Luciani, G. (2020). The impacts of the energy transition on growth and income distribution. In M. Hafner & S. Tagliapietra (Eds.), The geopolitics of the global energy transition. (Vol. 73). Cham: Springer.
Manso, J. R. P., de Matos, A. J. F., & Carvalho, C. C. (2015). Determinants of regional growth in Portugal: An empirical analysis. Economics & Sociology, 8(4), 11.
Martins, P. M. (2019). Structural change: Pace, patterns and determinants. Review of Development Economics, 23(1), 1–32.
MCTES (2017). Science, technology and tertiary education in Portugal – Perspectives for 2030, Background report to the OECD joint-review of Science, Technology and Tertiary Education in Portugal, draft document, Ministry of Science, Technology and Higher Education.
Meo, S.A., Al Masri, A.A., Usmani, A.M., Memon, A.N., & Zaidi, S.Z. (2013). Impact of GDP, spending on R&D, number of universities and scientific journals on research publications among Asian countries. Plos One, 8(6), 1–8
Mincer, J. (1958). Investment in human capital and personal income distribution. Journal of Political Economy, 66(4), 281–302.
Mnasri, A., Mrabet, Z., & Alsamara, M. (2023). A new quadratic asymmetric error correction model: Does size matter? Empirical Economics, 65, 33–64.
Moral-Benito, E. (2012). Determinants of economic growth: A Bayesian panel data approach. The Review of Economics and Statistics, 94(2), 566–579.
Morgado, S. M. A. (2014). Does health promote economic growth? Portuguese case study: From dictatorship to full democracy. The European Journal of Health Economics, 15(6), 591–598.
Morrar, R. (2019). Innovation in the MENA Region, in IEMed. Mediterranean Yearbook, 2019, 262–269.
Nelson, R. R. (1998). The agenda for growth theory: A different point of view. Cambridge Journal of Economics, 22(4), 497–520.
Nelson, R., & Winter, S. (1982). An evolutionary theory of economic change. Belknap Harvard.
Ntuli, H., Inglesi-Lotz, R., Chang, T., & Pouris, A. (2015). Does research output cause economic growth or vice versa? Evidence from 34 OECD countries. Journal of the Association for Information Science and Technology, 66(8), 1709–1716.
Odhiambo, N. M., & Ntenga, L. (2016). Research publications and economic growth in South Africa: An empirical investigation. International Journal of Social Economics, 43(7), 662–675.
OECD (2014). OECD Economic Surveys: Portugal, Paris: OECD Publishing.
OECD. (2019). OECD review of higher education: Portugal. OECD Publishing. https://doi.org/10.1787/9789264308138-en
OECD. (2023). OECD economic surveys: Portugal 2023. OECD Publishing. https://doi.org/10.1787/2b8ee40a-en
Onyancha, O. B. (2020). A meta-analysis study of the relationship between research and economic development in selected countries in sub-Saharan Africa. Scientometrics, 123(2), 655–675.
Pasinetti, L. L. (1981). Structural change and economic growth: A theoretical essay on the dynamics of the wealth of nations. Cambridge University Press.
Pegkas, P., Staikouras, C., & Tsamadias, C. (2019). Does research and development expenditure impact innovation? Evidence from the European Union countries. Journal of Policy Modeling, 41(5), 1005–1025.
Pereira, A. S., & Lains, P. (2012). From an agrarian society to a knowledge economy? The rising importance of education to the Portuguese economy, 1950-2009. In G. Neave & A. Amaral (Eds.), Higher education in Portugal, 1974-2009. A nation, a generation (pp. 109–134). Springer.
Pérez, C. (1983). Structural change and assimilation of new technologies in the economic and social systems. Futures, 15, 357–375.
Phillips, P. C., & Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75(2), 335–346.
Pinto, T., & Teixeira, A. A. C. (2020). The impact of research output on economic growth by fields of science: A dynamic panel data analysis, 1980–2016. Scientometrics, 123, 945–978.
Pourghaz, A., Samani, E. B., & Shokri, B. (2023). Analysis of the impact of research output on economic growth with using a multivariate random effects model. Scientometrics, 128, 2259–2282.
Quatraro, F. (2009). Innovation, structural change and productivity growth: Evidence from Italian regions, 1980–2003. Cambridge Journal of Economics, 33(5), 1001–1022.
Quatraro, F. (2010). The economics of structural change in knowledge. Routledge.
Ratinho, T., & Henriques, E. (2010). The role of science parks and business incubators in converging countries: Evidence from Portugal. Technovation, 4, 278–290.
Raisová, M., & Durcová, J. (2014). Economic growth-supply and demand perspective. Procedia Economics and Finance, 15, 184–191
Rocha, E. (1997). Portugal, anos 60: Crescimento económico acelerado e papel das relações com as colónias. Análise Social, 13(51), 593–617.
Romer, P. M. (1986). Increasing returns and long-run growth. Journal of Political Economy, 94(5), 1002–1037.
Romer, P. M. (1990). Endogenous technological change. Journal of Political Economy, 98(5), S71–S102.
Sala-i-Martin, X., Doppelhofer, G., & Miller, R. I. (2004). Determinants of long-term growth: A Bayesian averaging of classical estimates (BACE) approach. American Economic Review, 94(4), 813–835.
Santos, J., Domingos, T., Sousa, T., & Aubyn, M. S. (2018). Useful exergy is key in obtaining plausible aggregate production functions and recognizing the role of energy in economic growth: Portugal 1960–2009. Ecological Economics, 148, 103–120.
Sarpong, D., Boakye, D., Ofosu, G., & Botchie, D. (2023). The three pointers of research and development (R&D) for growth-boosting sustainable innovation system. Technovation, 122, 102581.
Saviotti, P. P., & Pyka, A. (2004). Economic development by the creation of new sectors. Journal of Evolutionary Economics, 14(1), 1–35.
Schofer, E., Ramirez, F. O., & Meyer, J. W. (2000). The effects of science on national economic development, 1970–1990. American Sociological Review, 65(6), 866–887.
Schultz, T. W. (1961). Investment in human capital. The American Economic Review, 51(1), 1–17.
Schumpeter, J. A. (1912). The theory of economic development (10th ed.). Transaction Publishers.
Sepehrdoust, H., Tartar, M. & Davarikish, R. (2021). Does scientific productivity stimulate intensified technology exports in developing economies. Journal of the Knowledge Economy, 12, 2111–2135.
Shenhav, Y. A., & Kamens, D. H. (1991). The ‘costs’ of institutional isomorphism: Science in non-western countries. Social Studies of Science, 21, 527–545.
Shrestha, M. B., & Bhatta, G. R. (2018). Selecting appropriate methodological framework for time series data analysis. The Journal of Finance and Data Science, 4(2), 71–89.
Silva, E. G., & Teixeira, A. A. C. (2011). Does structure influence growth? A panel data econometric assessment of ‘relatively less developed’ countries, 1979–2003. Industrial and Corporate Change, 20(2), 457–510.
Silverberg, G., & Verspagen, B. (1994). Learning, innovation and economic growth. A long run model of industrial dynamics. Industrial and Corporate Change, 3, 199–224.
Simões, V. C. (2022). In L. Melchor & K. Krieger (Eds.), Science for policy in Portugal. Publications Office of the European Union.
Smith, M. (2012). Demand-led growth theory: A historical approach. Review of Political Economy, 24(4), 543–573.
Smith, M. (2018). Demand-led growth theory in a classical framework: Its superiority, its limitations, and its explanatory power, Centro Sraffa Working Papers n. 29. http://www.centrosraffa.org/public/923ecb9d-2371-48a0-8425-13913be3c4c3.pdf
Solarin, S. A., & Yen, Y. Y. (2016). A global analysis of the impact of research output on economic growth. Scientometrics, 108(2), 855–874.
Solow, R. M. (1956). A contribution to the theory of economic growth. The Quarterly Journal of Economics, 70(1), 65–94.
Suárez, D., Peralta, M. J., Piedra, Y., & Navarro, M. A. (2022). Thematic coverage of CRIS in WoS, Scopus and Dimensions (2000–2020). Procedia Computer Science, 211, 170–186.
Sutherland, W. J., Goulson, D., Potts, S. G., & Dicks, L. V. (2011). Quantifying the impact and relevance of scientific research. PLoS ONE, 6(11), e27537.
Swan, T. W. (1956). Economic growth and capital accumulation. Economic Record, 32(2), 334–361.
Teixeira, A. A. C. (2005). Measuring aggregate human capital in Portugal: 1960–2001. Portuguese Journal of Social Science, 4(2), 101–120.
Teixeira, A.A.C. (2007). Inovação e relações universidade-empresa no caso português. Boletim Mensal de Actividade Económica—Junho de 2007, pp. 15–20. https://www.gee.gov.pt/pt/documentos/estudos-e-seminarios/working-papers/930-d000870/file.
Teixeira, A. A. C., & Fortuna, N. (2004). Human capital, innovation capability and economic growth in Portugal, 1960–2001. Portuguese Economic Journal, 3(3), 205–225.
Teixeira, A. A. C., & Fortuna, N. (2010). Human capital, R&D, trade, and long-run productivity testing the technological absorption hypothesis for the Portuguese economy, 1960–2001. Research Policy, 39(3), 335–350.
Teixeira, A.A.C. & Monteiro, A. (2018). The efficiency of Portuguese Technology Transfer Offices and the importance of university characteristics, GEE Papers, Number 93, February 2018.
Teixeira, A. A. C., Silva, E., & Mamede, R. (2014). Structural change, competitiveness and industrial policy. Painful lessons from the European periphery. Routledge.
Teixeira, A. A. C., & Queirós, A. S. S. (2016). Economic growth, human capital and structural change: A dynamic panel data analysis. Research Policy, 45(8), 1636–1648.
United Nations Development Programme (2019), Human Development Report 2019. Beyond income, beyond averages, beyond today: Inequalities in human development in the 21st century, New York: United Nations Development Programme.
Uyar, A., Kuzey, C., & Karamahmutoglu, M. K. (2022). Macroeconomic factors, R&D expenditure and research productivity in economics and finance. Managerial Finance, 48(5), 733–759.
Visser, M., van Eck, N. J., & Waltman, L. (2021). Large-scale comparison of bibliographic data sources: Scopus, Web of Science, Dimensions, Crossref, and Microsoft Academic. Quantitative Science Studies, 2(1), 20–41.
Wößmann, L. (2003). Specifying human capital. Journal of Economic Surveys, 17(3), 239–270.
Yasgül, Y. S., & Güris, B. (2016). Causality between research output in the field of biotechnology and economic growth in Turkey. Quality and Quantity, 50, 1715–1726.
Zaman, K., Khan, H. U. R., Ahmad, M., & Aamir, A. (2018). Research productivity and economic growth: A policy lesson learnt from across the globe. Iran Economic Review, 22(3), 627–641.
Zhong, M., Wang, P., Ji, M., Zeng, X. H., & Wei, H. X. (2021). Promote or inhibit: Economic goal pressure and residents’ health. Frontiers in Public Health, 26(9), 725957.
Acknowledgements
We are indebted to two referees for their comments and suggestions helped, which helped to improve the paper.
Funding
Open access funding provided by FCT|FCCN (b-on). Aurora Teixeira’s research has been financed by Portuguese public funds through FCT—Fundação para a Ciência e a Tecnologia, I.P., in the framework of the project with reference UIDB/04105/2020.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
We have no conflicts of interest to disclose.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Pinto, T., Teixeira, A.A.C. Research output and economic growth in technological laggard contexts: a longitudinal analysis (1980–2019) by type of research. Scientometrics 129, 1197–1230 (2024). https://doi.org/10.1007/s11192-023-04893-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11192-023-04893-1