Do domestic innovations promote trade openness? Empirical evidence from emerging economies

Abstract

Purpose

Innovation activities have gained much importance due to their pivotal role in achieving economic growth – directly by increasing productivity and – indirectly by increasing the degree of trade openness. This study aims to focus on the indirect channel, a rarely explored area of research, especially in the context of emerging economies.

Methodology

To achieve the aim of the study, four proxies of innovation (resident patent applications, nonresident patent applications, scientific and technical journal articles, and research and development expenditures) are used to establish a robust relationship between innovation activities and trade openness in BRICS economies. Panel data from 2000 to 2020 is obtained from World Development Indicators and Penn World Tables. Econometric techniques of panel data such as fixed effect and generalized least squares are employed to extract results from the specified models.

Findings

The findings of the study revealed that three proxies of innovation (i.e., resident patent applications, nonresident patent applications, scientific and technical journal articles) have a significant positive role in improving trade openness in the BRICS economies. However, the fourth proxy of innovation i.e., research and development expenditures had a negative impact on the degree of trade openness. Besides, innovation activities such as inflation rate and foreign direct investment have also influenced the degree of trade openness positively and significantly. Conversely, GDP per capita had a negative relationship with trade openness. Moreover, domestic investments showed a positive influence on the degree of trade openness while employment had a negative and insignificant influence on the degree of trade openness. Finally, the causality analysis revealed a one-way relationship running from innovations to trade openness.

Implications

In view of the results obtained, the policymakers of the BRICS economies might focus on encouraging innovation activities to enhance the degree of trade openness. Increased trade openness will consequently contribute to economic growth enormously and thus the attainment of sustainable development goals (SDG-8). Policymakers are also suggested to encourage FDI inflows and further ensure a moderate inflation rate to improve the degree of trade openness and hence accelerate economic growth.

Originality

This study focused on examining the nexus between innovation activities and trade openness in emerging economies, which is indeed an interesting but rarely explored area of research. The findings of the study might help the policymakers of the BRICS economies in formulating policies regarding trade openness and innovation activities.

1. Introduction

Trade openness is one of the major determinants of economic growth [[1], [2], [3]]. Several economies, particularly, those located in the East Asian region have accelerated and sustained economic growth because of their liberalized trade policies [4]. Openness to international trade is a pre-requisite for achieving higher economic growth, specifically in developing and emerging economies as it ensures greater access to extended international markets, advanced technologies and machineries, important raw materials, and further encourages healthy competition. Nevertheless, some studies revealed a negative relationship between trade openness and economic growth [5,6]. In this instance, Rodriguez and Rodrik [7] claimed that earlier studies on the trade-growth positive relationship were only based on measurement and methodological grounds. Conversely, several studies have provided more concrete evidence about the positive relationship between trade openness and economic growth, for example, [4,8,9]. Almost all the empirical studies have endorsed a positive relationship between trade openness and economic growth [10]. Some recent studies indicated the adverse effects of trade openness on environmental quality. For example, Udeagha and Ngepah [11] demonstrated the asymmetric influence of trade openness on economic growth in the South African context by utilizing the data for the period 1960–2016. Moreover, Udeagha and Ngepah [12] endorsed the positive role of trade openness in explaining the patterns of industries in provinces of the South African economy. It might be argued that despite certain methodological and measurement problems, there exists a correlation between trade openness and economic growth.

The prevailing consensus in the research is that the degree of trade openness has a significant role in achieving higher economic growth. However, the main question that remains to be under investigation is: what exactly determines the degree of trade openness in economies? The stated question has not yet been answered comprehensively by past empirical research. Recent literature has paid much attention to exploring the determinants of trade openness. For instance, Tahir et al. [9] Showed that education, investment, and GDP per capita are the driving forces behind increased trade openness. Similarly, Osei et al. [13] provided significant evidence about the positive influence of GDP per capita on the degree of trade openness in the context of low-income economies. Besides, Ngouhouo et al. [14] focused on 36 SSA economies for the period 1996–2017 and reported that institutional factors such as government effectiveness, rule of law and regulatory quality are positively linked with the increased degree of trade openness. Also, Ngepah and Udeagha [15] demonstrated a positive impact of trade agreements on international trade using the well-known gravity model of international trade. All these studies have only focused on the traditional determinants of trade openness. It is a fact that innovation activities in the domestic economy could also be linked with the degree of trade openness. Innovations are important for economic growth – directly through enhancement in productivity and – indirectly through raising the degree of trade openness. However, the innovation-based determinants of trade openness are largely ignored by prior literature. If innovation-based determinants of trade openness appeared to be more important for achieving a higher degree of trade openness, then policymakers would have a straightforward policy option to pay more attention to encouraging innovations in the domestic economy. Increased innovations would enhance economic growth through the channel of increased trade openness. As such, innovations are considered important channels by which economic openness flourishes in economic growth [16]. Higher economic growth is indeed desirable because it affects the living standard of the masses enormously and helps in attaining sustainable development goals [[17], [18], [19], [20], [21]].

Amid this backdrop in the literature, this research paper focuses on the relationship between domestic innovation activities and the degree of trade openness. This is indeed an interesting but relatively the least investigated research area as far as empirical literature is concerned. Although, theoretically, the relationship between innovation and economic growth is well documented in literature. However, the role of innovation activities on the degree of trade openness is yet to be examined comprehensively. There is very limited research on the relationship between innovation and trade openness [22]. Past research endorsed that differences in technological activity internationally are the prime factor behind differences in countries’ imports, exports, and income levels. It means that innovation activities in the domestic economy are linked with the degree of trade openness. Amendola et al. [23] used patents and investment as proxies for innovation in their model and demonstrated that technological variables shape the export shares of countries. Similarly, Ang et al. [24] showed that innovation stocks and competitiveness are the main drivers behind the export performance of Taiwan, Korea, Japan, and Singapore over the years. Thus, the motivation behind the current study is taken from the limited research on the relationship between innovation activities and trade openness. In terms of the sample section, the present paper focuses on the member economies of BRICS. All member economies are continuously going through structural and technological changes. It is recently noted that the BRICS members have shown an increasing trend in technological activities [25]. The study of Franco and de Oliveira [26] indicated a drastic increase in innovation activities in the BRICS region over the years. China and Russia have diversified their technological base and further entered dynamic frontier areas which is remarkable [25]. The increased innovative and technological improvements shown by the BRICS members could have positive trade-enhancing repercussions. However, the importance of these increased technological and innovative activities on the degree of trade openness is yet unknown. This is primarily because of the lack of available limited literature. Although there are several studies available on the relationship between green innovations and environmental performance in the context of BRICS economies. Where environmental sustainability has been the top priority of BRICS economies. However, domestic innovations and trade openness remains the least investigated area of research in BRICS economies. Therefore, this study is going to fill this research gap by focusing on the relationship between increased domestic innovation activities and trade openness in the context of BRICS economies.

This paper contributes to the literature in three ways. Firstly, the paper sheds light on the relationship between innovations and trade openness which is rarely investigated by prior empirical literature. Although there are studies available on the relationship between innovations and economic growth [27,28]. However, there are not many studies available on the explicit relationship between innovations and trade openness [22]. Hence, the current study deviates from traditional literature and focuses on establishing a relationship between innovations and trade openness. The lack of literature on the relationship between innovation activities and trade openness is one of the prime motivations behind the present study. Secondly, rather than focusing on the single measurement of innovations, this paper uses four proxies for domestic innovations to establish an explicit relationship between innovations and trade openness. This is because all aspects of innovations may not be equally important for increasing the degree of trade openness. Therefore, by employing four proxies of innovations, the current study will provide comprehensive evidence about the relationship between innovations and trade openness. Thirdly, this paper makes a contextual contribution by focusing on the BRICS economies which is ignored by previous literature. The BRICS economies are emerging economies paying significant attention to innovation activities in recent times. Further, the BRICS economies are relatively more open to foreign trade and constantly going through technological and structural changes. Therefore, investigating the relationship between innovations and trade openness in the context of BRICS economies is indeed a vital contribution as far as the empirical literature is concerned. Consequently, the outcome of the current study will certainly draw the attention of researchers, policy practitioners, and the wider public.

The present study is important on several grounds as it contributes enormously to the context of BRICS economies by focusing on a new and interesting area in the literature. Recently, it has been noted that innovation activities are constantly increasing among the BRICS members and hence it is important to explore whether these activities have any implications as far as international trade is concerned. Prior literature has only paid attention to exploring the influence of innovation activities on economic growth. On the other hand, the influence of innovation activities on the degree of trade openness is yet to be explored comprehensively. The prime motivation behind the current study is the lack of literature on the relationship between innovation activities and trade openness, specifically in the context of BRICS economies. If innovation activities result in enhancing the degree of trade openness, the policymakers of BRICS economies would focus on encouraging these activities. Increased trade openness no doubt will put these economies on the right track of economic growth. Although trade openness can be good for speeding up the growth process, it may also degrade the quality of the environment [29]. The present study will bring some new and comprehensive insights about the relationship between innovations and trade openness which is rarely investigated in the context of BRICS economies. The outcome of the study will therefore help policymakers a great deal while formulating suitable policies both related to innovations and international trade.

This paper is divided into several connected sections. Relevant literature on the determinants of trade openness is articulated in section 2. The third section is primarily devoted to model specification and estimating methodology. Statistics on innovation and trade openness for the entire sample as well as for the individual economies of BRICS are reported and discussed in section 4. Results of regression analysis are discussed in section 5 while section 6 is devoted to sensitivity analysis. The penultimate section includes causality results while the final section presents concluding remarks and further suggests policy implications based on results.

2. Literature review

2.1. Theoretical literature

The relationship between innovation activities in the domestic economies and the degree of trade openness has recently received considerable attention from both policymakers and researchers in recent years. It is primarily because of the fact that trade openness matters a lot for achieving higher economic growth as evident from prior literature. Theoretical literature has highlighted the importance of innovation activities as far as international competitiveness is concerned. For instance, Vernon [30] has emphasized the timing of innovation, the effects of scale economies, and the roles of ignorance in influencing trade as compared to comparative cost theory. Similarly, Posner [31] documented that technical change and development may influence trade. Fagerberg [32] documented that international competitiveness is directly dependent on technological competitiveness and the ability to compete on delivery across countries. These theoretical studies have indicated that the patterns of trade among countries can be shaped more by technological change as compared to the comparative advantage theory. In other words, technological change is an important driver of international trade among countries. In the modern globalized world, trade between countries, especially intra-industry trade, is primarily dependent on innovation instead of comparative or absolute advantage theories. Trade among the developed economies is intra-industry and hence it is determined by the rising innovations in developed economies.

2.2. Empirical literature

The empirical literature has recently paid significant attention to exploring the relationship between innovations in the domestic economy and the degree of trade openness. For instance, Vetsikas and Stamboulis [22] investigated the relationship between innovation activities (R&D expenditures and patent applications) and trade openness for 10 European economies by utilizing the ARDL framework. Their findings indicated a positive and significant influence of innovations on the degree of trade openness. Kacani [33] suggested that emerging economies enhance their innovative capabilities for improving their trade openness and integration into global value chains. Amendola et al. [23] carried out a comprehensive study on the relationship between innovation and trade openness. They used patents and investments as proxies for innovation in their model and demonstrated that technological variables shape export shares. Similarly, Ang et al. [24] showed that innovation stocks and competitiveness are the main drivers behind the export performance of Taiwan, Korea, Japan, and Singapore. The dominant message from these studies is that the causality is running from innovations toward trade openness.

On the other hand, some studies have focused on investigating the influence of international trade on innovations. For example, documented that competition generated by trade openness enhances innovations in the domestic economy [34]. Moreover, suggested that innovations are important channels by which economic openness flourishes in economic growth [16]. Innovations such as green finance and financial technologies are also important for the promotion of environmental sustainability in BRICS economies [29]. Moreover, Udeagha and Ngepah [29] pointed out that BRCIS economies should focus on encouraging green innovation and renewable energy R&D to create a sustainable environment and get rid of increased carbon emissions.

Another strand of literature has focused on the macroeconomic determinants of trade openness. For instance, in the context of transitional economies, Tsaurai [35] showed that education and the interaction of education with FDI along with mining sector growth are the main contributors to trade openness. Similarly, by focusing on SAARC economies and employing panel data techniques of fixed effects, Tahir et al. [9] have indicated that education, investment, and income level matter in achieving higher trade openness. Focusing on two emerging economies of Pakistan and China, Latif et al. [36] highlighted the importance of FDI for improving the degree of trade openness. Further, their findings show that exchange rate and labor force are detrimental to the degree of trade openness. GDP per capita could be one of the main factors behind trade openness as it is the reflection of the purchasing power of the population. In this regard, the recent study of Puntoon et al. [37] showed a positive relationship between GDP per capita and trade openness by focusing on 85 economies for the period 1990–2017. However, Osei et al. [13] endorsed that the positive influence of income on trade openness is only valid for low-income economies while for the lower-middle income, this relationship is negative.

Moreover, some studies have highlighted the role of institutional factors while investigating the determinants of trade openness. For instance, Ngouhouo et al. [14] have focused on 36 SSA economies for the period 1996–2017 and reported that institutional factors such as government effectiveness, rule of law and regulatory quality are positively linked with increased degree of trade openness. It implies that economies must focus on improving their institutional quality to significantly increase their degree of trade openness as trade openness is the engine of achieving higher economic growth as evident from prior literature. Guttmann and Richards [38] demonstrated that the most important determinants of trade openness are population and distance between the trading partners. Finally, some studies have highlighted the importance of trade agreements in promoting trade openness. For instance, using data from 53 African economies for the period 1995–2014, Ngepah and Udeagha [39] demonstrated that regional trade agreements enhance international trade. Similarly, Ngepah and Udeagha [15] also displayed a positive impact of trade agreements on international trade using the well-known gravity model of international trade.

2.3. Summary of the literature

It could be concluded from the brief review that the empirical literature on the relationship between innovation and trade openness is not very extensive. Only a few empirical studies have focused on investigating the determinants of trade openness despite the available strong theoretical support in favor of a positive relationship between innovations and trade openness. Similarly, the role of domestic innovation activities in improving the degree of trade openness is yet to be assessed comprehensively. The available have used only a few proxies for innovations such as R&D expenditures, patent applications, and investment [22,23]. It is also a fact that prior literature has ignored the BRICS economies while investigating the determinants of trade openness. Therefore, the current study tries to fill this knowledge gap by focusing on the role of innovative activities in enhancing the degree of trade openness by focusing on the member economies of BRICS. Unlike the previous studies, the current study uses four proxies for measuring innovation to establish a robust relationship between innovation and trade openness. We expect that the current study will be useful for the policymakers of BRICS and future researchers interested in innovation-trade relationships.

3. Methods and estimations

3.1. Model specification

Specifying a model is an important step in applying research studies as it helps in achieving the objectives of the study. The main objective mentioned earlier is to establish a link between innovation activities and trade openness for the BRICS member economies. The following baseline model is specified for measuring the influence of innovations on the degree of trade openness.

$${\text{TRADE}}_{\text{it}}={\mathrm{\gamma }}_0+{\gamma }_1{\text{INNOVATIONS}}_{\text{it}-1}+{\gamma }_2{\mathrm{X}}_{\text{it}}\mathrm{B}+{\mathrm{\eta }}_{\text{it}}$$ (1)

Where ${\mathrm{\eta }}_{\text{it}}={\mathrm{\mu }}_{\mathrm{i}}+{\mathrm{\delta }}_{\text{it}}$.

Trade openness is used as a dependent variable in model 1, which is measured by (Exports + Imports/GDP) *100. Innovation is the independent variable that is measured through four proxies – namely – resident patent applications, nonresident patent applications, scientific and technical journal articles, and R&D expenditures. In the specification model, innovations represent four different proxies, therefore, four different models were developed in this study to measure the four different specifications of the innovations. Previous literature has also used these indicators for measuring domestic innovations [22,40,41]. Our modeling framework is based on the new trade theory where innovations and economies of scale play a dominant role in shaping the degree of trade openness. The term $\left({\mathrm{X}}_{\text{it}}\mathrm{B}\right)$ includes control variables that could impact the degree of trade openness. However, the trade openness of countries could also be linked with several other important factors. Previous empirical literature has identified several factors that are important for the degree of trade openness [9]. Foreign direct investment (FDI), domestic investment, inflation rate, employment level, and GDP per capita are considered the main driving forces behind increasing the degree of trade openness of countries [9,13,37]. The mentioned factors are therefore used as control variables in the estimated models. FDI is measured by taking the net FDI inflows as a percent of GDP while inflation is approximated by taking the growth of the consumer price index. Employment level is measured by taking the number of people engaged while GDP per capita in real terms is used for capturing the size of the market.

3.2. Sample and sources of data

All five economies of BRICS are selected for analysis to provide detailed empirical evidence about the impact of innovation activities on the degree of trade openness. The list is provided in the appendix section. Data on all variables is sourced from the World Development Indicators provided by the World Bank. Detailed information is provided in the appendix section about the construction of variables.

3.3. Estimation tools

The data gathered for analysis is basically panel in nature as it has two dimensions. The first dimension is a cross-section as there are five countries while the second dimension is the time dimension as there are multiple years (2000–2020). The fixed (FE) and random effects (RE) are powerful tools for handling the model of panel data [[42], [43], [44], [45]]. Both estimating tools have certain advantages as well as disadvantages [[46], [47], [48]]. The FE modeling is efficient and suitable when the chances of serial correlation are greater between the error term and independent variables of the model [49]. However, the FE modeling is not able to assess the influence of time-invariant factors. The countries included in the sample are quite heterogeneous and hence there are apparent significant variations among them. These variations need to be controlled to obtain reliable estimates which are desirable. The RE modeling tool is unlikely to produce reliable results when there is a chance of serial correlation between the error term and independent variables. However, it can effectively assess the influence of time-invariant factors. Therefore, RE modeling is rarely used in applied studies due to its inability to control the most likely serial correlation problem between the error term and independent variables. The decision between choosing the FE and RE could be made using the well-known Hausman test (1978). For the Hausman test, the number of cross-sections must be more than the number of parameters. However, in our case, the number of cross-sections is less than the number of parameters. Therefore, we have used the FE estimator for the estimation purpose. It is quite logical to think about the possibility of serial correlation among the chosen variables and the error term of the model. In such a scenario, the use of the FE is more appealing. However, Semykina and Wooldridge [50] mentioned that it is always useful to estimate the panel data models using the FE procedure owing to the likely serial correlation between the error term and independent variables.

In addition, the current study also utilized the Generalized Least Square (GLS) method as well for the estimation purpose. The purpose behind employing the GLS method is that it is used by previous literature as the robustness test for the results based on the FE estimator [[51], [52], [53]]. Results are provided in section 4 based on the FE estimation as well as on GLS-based estimation. Moreover, the study conducted the Pesaran CD test of cross-sectional independence, and the results shown in the appendix have confirmed the cross-sectional independence which is desirable. Previous literature has also utilized the CD test for examining cross-sectional independence [54].

4. Statistics on innovations in BRICS

Detailed information on innovation activities and trade openness for the members of BRICS is shown in Table 1. Data is averaged for the starting and end years of the study for all five members. The percentage changes are calculated by using the formula (current-Previous/Previous) *100. All four indicators used for measuring innovation activities (resident patent applications, nonresident patent applications, scientific and technical journal articles, and R&D expenditures) have shown an increasing trend during the study period. Patent applications by residents of the BRICS members have increased exponentially by 2254.807% between 2000 and 2020 which is indeed a remarkable improvement. Similarly, patent applications by non-residents have also increased by more than 290 % between 2000 and 2020. The rise in nonresident patent applications is an indication that the member economies of BRICS have attracted the best brains from around the world. Research publications by the BRICS members have increased by approximately 488 % during the last couple of decades in BRICS members which is remarkable. On the other hand, Research and Development expenditures expressed in percentage of GDP have increased by approximately 39 %. However, it is also a fact that the R&D expenditures are slightly about 1 % of the GDP which is too low. The wise policy decision would be to allocate more funds towards R&D expenditures as they are the key to long-run economic growth. Interestingly, the trade openness degree (trade as % of GDP) has slightly declined by about 5 % which is surprising.

Table 1.

Innovation and trade statistics in BRICS.

Variables	Definition	2000	2020	Change
RPAT	Residents patent applications	11869.67	279507.8	2254.807 %
NPAT	Nonresidents patent applications	11400.89	44480.2	290.146 %
PUBLICATIONS	Scientific and technical journal articles	26077.41	153391.3	488.215 %
R & D	R&D expenditures	0.86283	1.202674	39.387 %
TRADE	Trade as a % of GDP	42.65477	40.47477	−5.110 %

Note:Authors calculation using data from World Development Indicators.

4.1. Individual statistics on innovations in BRICS

In addition to statistics on innovation for the whole sample, in this section, we presented statistics for individual economies to provide more in-depth insights as demonstrated in Table 2. Patent applications by residents have exponentially raised from 25,346 in 2000–1,344,817 in 2020 showing a net increase of more than 5205 %. The Indian economy also showed a massive increase in resident patent applications as they increased from 2206 in 2000–23,141 in 2020 indicating a net rise of 949 %. The economy of Brazil also showed a rise of 66 % in resident patent applications while the increase in resident applications for the economy of Russia is negligible. Interestingly, in South Africa, the patent application by residents has decreased from 895 in 2000 to 542 in 2020 showing a net decline of more than 39 % which is surprising.

Table 2.

Innovation and trade statistics country-wise.

Country	Variables	2000	2020	Change
Brazil	RPAT	3179	5280	66.08997 %
	NPAT	14,104	19,058	35.12479 %
	PUBLICATIONS	12783.45	58651.41	358.8074 %
	R & D	1.04751	1.164407	11.15951 %
	TRADE	22.63976	32.89045	45.27738 %
Russia	RPAT	23,377	23,759	1.634085 %
	NPAT	8960	11,225	25.27902 %
	PUBLICATIONS	32224.01	74697.73	131.8077 %
	R & D	1.04983	1.09803	4.59122 %
	TRADE	68.09391	45.96195	−32.5021 %
India	RPAT	2206	23,141	949.0027 %
	NPAT	6332	33,630	431.1118 %
	PUBLICATIONS	21770.72	126871.5	482.7621 %
	R & D	0.75699	0.661876	−12.5648 %
	TRADE	26.90092	37.80535	40.53553 %
China	RPAT	25,346	1,344,817	5205.835 %
	NPAT	26,560	152,342	473.5768 %
	PUBLICATIONS	53064.35	493906.2	830.7684 %
	R & D	0.89316	2.40093	168.813 %
	TRADE	39.41101	34.58925	−12.2346 %
South Africa	RPAT	895	542	−39.4413 %
	NPAT	2400	6146	156.0833 %
	PUBLICATIONS	3897.55	12829.42	229.1663 %
	R & D	0.659	0.68813	4.420334 %
	TRADE	46.22072	51.12685	10.61457 %

In terms of patent applications by nonresidents, the Chinese economy again did well among the BRICS members. The patent application by non-residents has increased from 26,560 in 2000–152,342 in 2020 showing a net increase of 473 %. The economy of India has also shown a massive increase of 431 % in nonresident patent applications during the study period. It is interesting to note that South Africa has done well as the nonresident's patents have increased from 2400 in 2000–6146 in 2020 indicating a net increase of approximately 156 %. The decline in resident patent applications and an increase in nonresident applications during the study period is an indication that the South African economy has attracted the best brains from the rest of the world. Finally, Brazil and Russia have also had significant improvements in nonresident patent applications between 2000 and 2020.

Research publications have also increased in all BRICS members. The highest increase of 830 % in research publications was seen in China followed by India (482 %). For the Brazilian economy, an increase of 358 % was witnessed in research publications between 2000 and 2020. The South African economy has also shown an impressive increase of 229 % in research publications. Finally, the performance in terms of research publication in the case of the Russian economy is lowest among the BRICS members.

The R&D expenditures have increased in all BRICS member economies except the Indian economy which is astonishing. The Chinese economy has increased its R&D expenditures by 168 % between 2000 and 2020. The current statistics show that the Chinese economy is spending more than 2 % of its GDP on R&D expenditures which is highest among the BRICS member economies. In the case of the Indian economy, the R&D expenditures have decreased from 0.756 in 2000 to 0.661 in 2020 showing a net decline of more than −12 %. This apparent decline in R&D expenditures could be one of the main reasons behind the growth differences between India and China. The Brazilian economy has increased R&D spending by more than 11 % between 2000 and 2020 which is satisfactory. In the case of Russia and South Africa, the R&D expenditures have increased marginally during the study period.

Statistics on trade openness show that all members of BRICS except Russia and China have improved their trade openness. The trade openness index has declined from 68.093 in 2000 to 45.961 in 2020 showing a decline of −32.502 % which is indeed undesirable. Similarly, the Chinese economy has also shown a decline in trade openness during the study period. Brazil has improved its trade openness index from 22.639 in 2000 to 32.890 showing a rise of 45.277 % which is quite impressive. Similarly, the Indian economy has also done well in terms of improving its trade openness index. The statistics show the overall trade openness of the Indian economy has increased by more than 40 % which is quite satisfactory. Lastly, the South African economy has increased its trade openness index by more than 10 %.

5. Results and discussion

5.1. Descriptive statistics

According to descriptive statistics presented in Table 3, on average the trade openness of BRICS member economies is 44.261 while the maximum and minimum values are 68.093 and 22.105 respectively. The maximum value of trade openness is recorded for the Russian economy for the year 2000 while the minimum value is recorded for the economy of Brazil.

Table 3.

Descriptive statistics.

	TRADE	RPAT	NPAT	PUB	RAD	GDPC	HC	INF	FDI	INV	EMP
Mean	44.261	114344.0	32516.97	89817.79	1.067	5892.643	2.559	5.981	2.291	25.290	282.502
Median	46.518	8841.000	16236.00	41666.17	1.027	6209.366	2.473	5.184	2.073	21.321	86.353
Maximum	68.093	1,393,815.	157093.0	528263.3	2.400	10358.26	3.434	21.477	5.368	44.518	799.306
Minimum	22.105	542.000	2400.000	3897.550	0.615	757.668	1.782	−0.731	0.205	13.716	13.373
Std. Dev.	11.984	305641.2	38750.81	123693.1	0.407	2926.157	0.474	4.035	1.243	9.470	296.305
Observations	105	105	105	105	105	105	105	105	105	105	105

5.2. Discussion on regressions results

Regression results are displayed in Table 4. The specified model is estimated step by step by incorporating the four proxies used for capturing the innovation activities in the domestic economies. In model 1, the patent application by residents is used as a proxy for innovation while in model 2, patent applications by nonresidents are utilized. In the third and fourth models, the number of publications in scientific journals and R&D expenditures is employed as a proxy for innovations, respectively.

Table 4.

Regression results.

Variables	Model-1	Model-2	Model-3	Model-4
	Coefficients	Coefficients	Coefficients	Coefficients
GDPC	−1.221*** (0.239)	−0.625*** (0.097)	−0.710*** (0.067)	−0.373*** (0.124)
INV	0.009 (0.158)	−0.134 (0.173)	0.083 (0.141)	0.200 (0.168)
FDI	0.066*** (0.018)	0.039** (0.016)	0.076*** (0.017)	0.040* (0.021)
HC	−1.068* (0.572)	−0.332 (0.641)	−0.425 (0.675)	0.659 (0.650)
EMP	−0.432 (0.626)	−0.490 (0.752)	−1.056 (0.674)	−1.360** (0.584)
INF	0.029*** (0.004)	0.027*** (0.005)	0.027*** (0.004)	0.035*** (0.005)
RPAT	0.224*** 0.073)	–	–	–
NPAT	–	0.261*** (0.058)	–	–
PUBLICATION	–	–	0.310*** (0.045)	–
R&D	–	–	–	−0.243** (0.101)
CONSTANT	14.837 (3.039)	9.393 (3.772)	11.487 (3.173)	12.024 (3.336)
Diagnostics	R2: 0.922 R2 (Adj): 0.890 F-Test: 28.20***	R2: 0.925 R2 (Adj): 0.893 F-Test: 29.12***	R2: 0.932 R2 (Adj): 0.904 F-Test: 32.62***	R2: 0.916 R2 (Adj):0.880 F-Test: 25.78***

Note: The asterisk (***), (**) and (*) represent 1, 5, and 10% level of significance. Values in parentheses stand for standard errors.

The FE results displayed an expected positive influence of the patent application by the residents on trade openness. The coefficient of patent application by residents is positive and significant statistically. It implies that the degree of trade openness will be impacted positively by the increase in patent applications by residents. The results of the second column show that the patent applications by non-residents have also enhanced the degree of trade openness. It implies that higher patent applications by non-residents are also needed to positively influence the degree of trade openness. The third column shows that journal publications which are also used as a proxy of innovation have also contributed to the degree of trade openness. The results imply that scientific research in the form of journal publications will increase the degree of trade openness of BRICS economies. On the other hand, the R&D expenditures have negatively impacted the degree of trade openness. However, the descriptive statistics reported earlier have shown that the current level of R&D expenditures is still below 2% of the GDP which is alarming. Although the R&D expenditures have increased significantly between 2000 and 2020. One of the possible reasons could be that R&D expenditures may impact trade openness through the channel of patents. Similarly, it is also possible that all R&D expenditures could not be translated into patents. The overall positive influence of innovation activities in the domestic economy and the degree of trade openness is quite encouraging. It implies that the BRICS economies must pay significant attention to encouraging innovations in their economies to enhance the degree of trade openness. Increased trade openness no doubt will put these economies on the right track of higher economic growth. Higher economic growth would also help the huge population of BRICS economies to get rid of the poverty problem which is the end of the objective of all economic activities. The overall results of the current study about the relationship between innovation and trade openness are in line with the recent research study by Vetsikas and Stamboulis [22]. The study demonstrated a positive connection between innovation and trade openness by focusing on 10 European economies for the period 1983–2018. In the light of the obtained results, the BRICS economies are suggested to encourage domestic innovation activities as they are necessary for enhancing the degree of trade openness which is the engine of economic growth as evident from the prior literature [1,2,4].

The results further displayed a positive and statistically significant influence of FDI inflows on the degree of trade openness. It could be said that FDI inflows are important for enhancing the degree of trade openness. FDI inflows are the main ingredient of globalization, and they are responsible for economic transformation, job creation technological development as pointed out by Cantah et al. [55]. FDI inflows mostly coupled with advanced technologies due to which the productive capabilities enhance in the domestic economy. In other words, the overall production increased due to which the share of exports rose enormously. The BRICS economies are therefore suggested to take some serious steps by providing business business-friendly environment coupled with improved law and order situation to encourage the inflows of FDI from advanced economies. Besides FDI inflows, the inflation rate also appeared to be an important factor behind increased trade openness as it has a positive and significant coefficient in all estimated models. However, higher inflation is harmful to an economy as it shatters the confidence of investors [4]. Therefore, moderate inflation in the form of single digits could be ensured to enhance the degree of trade openness.

GDP per capita, which is used as a proxy for market size, appeared to be casting an adverse influence on the degree of trade openness. In all estimated models, the coefficient of GDP per capita is negative and statistically significant. It implies that the degree of trade openness will be impacted adversely by increased GDP per capita. Prior literature has also shown a negative influence of GDP per capita on the degree of trade openness [13]. The possible reasons could be that higher economic size may force the people to get involved in domestic trade instead of international trade. Economies having larger sizes in terms of per capita may increase domestic trade within the economies instead of international trade. This view is consistent with Frankel and Romer [3] views that greater size in terms of population may encourage more domestic trade.

Domestic investment has not had the desirable significant positive influence on the degree of trade openness. The coefficient of investment is positive but insignificant in all estimated models. However, it does not mean that domestic investment is irrelevant to improving the degree of trade openness. Domestic investment is the key factor of production and hence it directly enhances production and consequently the volume of exports. The possible reasons could be that domestic investment in all BRICS economies is not enough to significantly influence the degree of trade openness. Similarly, the insignificant influence of domestic investment could also imply that instead of traditional investment, investment in innovation activities is required to enhance the degree of trade openness. Similarly, human capital and employment have negatively influenced the degree of trade openness. However, neither is significant in the majority of the specifications estimated.

6. Sensitivity analysis

In this section, we have focused on testing the robustness of the findings reported in the earlier section. For this purpose, the GLS estimator is utilized, and results are reported in Table 5. The GLS-based results have provided solid support to earlier results. According to the results, the three proxies of innovation namely, a patent application by residents, a patent application by nonresidents and journal publications have positively and significantly impacted the degree of trade openness which is consistent with earlier results. Similarly, the R&D expenditures have again cast a negative influence on the degree of trade openness which is also consistent with the fixed effects results. Similarly, FDI and inflation rate have maintained their positive and significant influence on the degree of trade openness which is in line with previous results. Moreover, the negative but significant influence of GDP per capita on the degree of trade openness also remained robust in the GLS-based estimation. Finally, the insignificant influence of employment didn't change in the GLS-based estimation while the insignificant positive impact of trade openness turned significant in some of the specifications.

Table 5.

Sensitivity analysis.

Variables	Model-5 GLS	Model-6 GLS	Model-7 GLS	Model-8 GLS
	Coefficients	Coefficients	Coefficients	Coefficients
GDPC	−.418*** (0.144)	−0.492*** (0.075)	−0.469*** (0.070)	−0.119 (0.090)
INV	0.271** (0.129)	−0.114 (0.169)	0.146 (0.134)	0.252** (0.113)
FDI	0.025** (0.010)	0.025*** (0.013)	0.065*** (0.011)	0.019 (0.012)
HC	−.165 (0.388)	−0.578 (0.556)	−1.092*** (0.573)	0.424 (0.480)
EMP	0.594 (0.391)	0.913*** (0.512)	0.695 (0.526)	−1.120 (0.445)
INF	0.013*** (0.003)	0.015*** (0.003)	0.016*** (0.002)	0.021*** (0.003)
RPAT	0.094* (0.051)	–	–	–
NPAT	–	0.405*** (0.046)	–	–
PUBLICATION	–	–	0.293*** (0.047)	–
R&D	–	–	–	−0.171** (0.081)
CONSTANT	2.416 (1.518)	0.273 (2.058)	1.624 (2.11)	2.852 (1.780)
Diagnostics	R2: 0.883 R2 (Adj): 0.869 F-Test: 63.85***	R2: 0.870 R2 (Adj): 0.854 F-Test: 56.68***	R2: 0.863 R2 (Adj): 0.847 F-Test: 53.56***	R2: 0.879 R2 (Adj): 0.864 F-Test: 61.48***

Note: The asterisk (***), (**) and (*) represent 1, 5, and 10% level of significance. Values in parentheses stand for standard errors.

7. Causality testing

In this section, we adopted the Dumitrescu and Hurlin [56] approach for identifying the direction of the relationship between innovation and trade openness. Results are presented in the following Table 6. The results clearly indicated that the three proxies of innovation such as publication, residents patent application and R&D expenditures are unilaterally causing trade openness. On the other hand, non-resident patent applications and trade openness are bidirectionally related to each other. It could be concluded that the innovations in the domestic economy are impacting the degree of trade openness.

Table 6.

Causality testing.

Null Hypothesis	Zbar-Stat	Probability
From NPAT to TRADE	3.43470***	0.0006
From TRADE to NPAT	3.05510***	0.0022
From PUBLICATION to TRADE	1.95131*	0.0510
From TRADE to PUBLICATION	1.44009	0.1498
From R&D to TRADE	3.57758***	0.0003
From TRADE to R&D	0.74497	0.4563
From RPAT to TRADE	1.93643*	0.0528
From TRADE to RPAT	−1.18635	0.2355

Note: The asterisk (***) and (*) represent 1 and 10% level of significance.

8. Conclusion and way forward

The present study has focused on the BRICS member economies and has used four proxies for innovation to assess their impact on the degree of trade openness. The study utilized panel data spanning from 2000 to 2020 and employed suitable econometric techniques including the FE and GLS estimators. The results confirmed the importance of innovation activities for improving the degree of trade openness in the BRICS member economies. Three proxies of innovation such as (resident patent applications, nonresident patent applications, and scientific and technical journal articles) have positively and significantly impacted the degree of trade openness. However, the R&D expenditures which are used as a fourth proxy of innovation have negatively influenced the degree of trade openness which is surprising. However, it is a fact that the R&D expenditures are too low in BRICS economies as the majority of its members are spending even below 1% of the GDP on the R&D sector. The causality analysis confirmed the one-way relationship running from innovations to trade openness. The results further confirmed the positive relationship between inflation, FDI, and trade openness while GDP per capita has adversely influenced the degree of trade openness.

8.1. Policy implications

Based on the comprehensive and robust empirical findings, the study presents the following policy implications.

• 1)In light of the results, three proxies of innovations are found to be enhancing trade openness. Therefore, targeted efforts are required to encourage innovative activities in the domestic economy. Further, spending on R&D activities must be enhanced as the current level of R&D expenditure is not satisfactory. Overall encouraging innovations by the BRICS member economies will enhance the degree of trade openness and hence the speed of economic growth will flourish. Consequently, this will help in achieving the SDG-8 in BRICS economies and thereby strengthen the contributions towards the global agenda of achieving sustainable development. The policy makers of BRICS economies are therefore required to focus on innovation activities with more urgency to achieve sustainable development through the channel of trade openness.
• 2)The results provided sound evidence about the affirmative role of FDI inflows as far as the degree of trade openness is concerned. Therefore, FDI inflows from advanced economies need to be encouraged by the BRICS nations as these inflows are necessary for enhancing trade openness.
• 3)Similarly, the inflation rate, which is the driver of the increased degree of trade openness based on results, needs to be kept in a moderate range preferably in single digits. Moderate inflation will positively influence trade openness, which is desirable. However, higher inflation needs to be controlled using both fiscal and monetary policies as it shatters the confidence of all stakeholders in the economy.
• 4)The results confirmed the insignificant role of domestic investment in improving the degree of trade openness. However, it is a fact that investment in the economy directly increases production and hence the export sector flourishes enormously. For this purpose, the rate of interest needs to be reduced to a reasonable level as low rates of interest flourish investment.

8.2. Limitations and future research directions

This study has some unavoidable limitations. For instance, the first limitation of the study is the use of primitive methodologies, specifically due to the short dimension of the panel. Future studies are suggested to use advanced methodologies such as GMM and panel cointegration analysis. Advanced methods are expected to provide more robust results which could be used in policy formulation with more confidence. In addition to advanced methodologies, future studies are also suggested to incorporate some more control variables such as exchange rate and rate of interest and see how the degree of trade openness responds to innovations. Secondly, due to the unique features of the BRICS economies, the results could not be generalized. Future studies could test the specified models of this study for other regions/countries having similar characteristics. This might include a comprehensive comparative study focusing on both developed and developing economies. Lastly, the current study has investigated the influence of domestic innovations on trade openness. It is possible that innovations may influence the components of trade such as imports and exports differently. Therefore, future studies are suggested to examine the influence of innovation activities on both imports and exports separately. In addition, the influence of innovations on export firms’ performance is also a potential area of research. We leave these investigations to future researchers.

Data availability statement

The data used in this study is available on the World Development Indicators and Penn World Tables.

CRediT authorship contribution statement

Mohd Naved Khan: Writing – review & editing, Writing – original draft, Resources, Formal analysis, Conceptualization. Ahmad Ali Jan: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology. Mohammad Asif: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Funding acquisition, Data curation, Conceptualization. Fong-Woon Lai: Writing – review & editing, Validation, Supervision. Muhammad Kashif Shad: Writing – review & editing, Software, Formal analysis. Saima Shadab: Writing – review & editing, Formal analysis, Data curation.

Declaration of competing interest

The authors declare no conflict of interest. The authors would like to confirm that this paper is an outcome of the authors' own work and has not been published elsewhere neither is considered for publication elsewhere.

Appendix. Section

Table 1 A.

Results of pesaran CD testing.

Pesaran CD Test	Statistic	D.F	Prob.
Model-1	0.377	10	0.705
Model-2	1.772	–	0.076
Model-3	1.824	–	0.068
Model-4	1.192	–	0.233

Table 2 A.

Redundant testing.

Models	Effects Test	Value	d.f.	Prob.
Model-1	Cross-Section/Period F	6.676	(24,73)	0.000
	Cross-Section/Period Chi-square	121.963	24	0.000
Model-2	Cross-Section/Period F	8.644	(24,73)	0.000
	Cross-Section/Period Chi-square	141.333	24	0.000
Model-3	Cross-Section/Period F	10.986	(24,73)	0.000
	Cross-Section/Period Chi-square	160.513	24	0.000
Model-4	Cross-Section/Period F	8.090	(24,73)	0.000
	Cross-Section/Period Chi-square	136.233	24	0.000

Table 3 A.

Variables description.

Variables	Definition	Source
TRADE	Trade (% of GDP)	World Development Indicators
Lnrpat	Patent applications, residents	World Development Indicators
Lnnpat	Patent applications, nonresidents	World Development Indicators
Lnpub	Scientific and technical journal articles	World Development Indicators
Lnrad	Research and development expenditure (% of GDP)	World Development Indicators
Lny	GDP per capita (constant 2015 US$)	World Development Indicators
Lninv	Gross fixed capital formation (% of GDP)	World Development Indicators
Lnfdi	Foreign direct investment, net inflows (% of GDP)	World Development Indicators
lnhc	Human capital index, based on years of schooling and returns to education	Penn World Tables (10)
Lnemp	Number of persons engaged (in millions)	Penn World Tables (10)
Inflation	Inflation, consumer prices (annual %)	World Development Indicators

Table 4 A.

Countries.

Brazil	Russian
India	China
South Africa