DOI: 10.28934/ea.22.55.2.pp49-65 First Online: December 6, 2022 ORIGINAL SCIENTIFIC PAPER Does the Increase in the Number of Registered Patents Affect Economic Growth? - Evidence from Romania and Bulgaria Ivana Domazet16F* | Darko Marjanović1 | Deniz Ahmetagić2 | Marija Antonijević1 1 Institute of Economic Sciences, Belgrade, Serbia 2 Faculty of Economics in Subotica, Serbia ABSTRACT The goal of this paper is to determine whether the increase in the number of registered patents per million inhabitants, as a measure of market verification of the results of R&D activities, affects economic growth and the increase in the country's innovation index. The empirical research covered two countries - Romania and Bulgaria. Given that the main task of the research was to accurately measure the investigated phenomena and discover the connection between them, the analysis was based on a quantitative research design. The analysis used secondary data from the international databases of the World Bank and World Intellectual Property Organization, covering the period from 2008 to 2018. The results of the empirical research showed that no correlations were found, which means that in the cases of Romania and Bulgaria, there is no dependence between the increase in the number of registered patents per million inhabitants and the growth of the innovation index and GDP per capita. Key words: innovation, patents, innovation index, GDP per capita, development JEL Classification: O3 INTRODUCTION Economic growth is a key element for improving living standards, reducing poverty and achieving common progress (World Bank, 2022). Countries that generate new technologies and encourage their adoption, as well as those that create innovations, grow faster than those countries that do not promote these activities (Domazet et al., 2021). The same authors state that patenting in certain industries is recognized as an important tool for protecting intellectual property and creating a sustainable competitive advantage. According to OECD (2004), patents have an important role in achieving innovation and economic performance. The changes that occurred in the last two decades regarding the patent policy of the OECD member countries encouraged the creation of patents to initiate investments in innovation and improve wedge. The OECD publication also points out that patents are intended to boost innovation in the private sector by enabling their inventors to profit from the invention. Moser (2013) argues that policies aimed at spreading ideas and modifying patent rights to stimulation could be an adequate way to encourage innovation. The important role of patents in stimulating innovation and economic growth was emphasized in 2012 by the then-president of the European Patent * Corresponding author, e-mail: ivana.domazet@ien.bg.ac.rs 50 Economic Analysis (2022, Vol. 55, No. 2, 49-65) Organization (EPO), Benoît Battistelli, during his speech at the Bulgarian Inventors of the Year event (EPO, 2022). It is important to understand that for patents to have a positive impact on economic growth, it is necessary to apply them in practice (Atun et al., 2007). Since the goal of every country is to achieve economic progress, numerous authors have examined the nature of the relationship between patents and economic growth. In this regard, some researchers have determined the existence of a positive relationship (Wurster, 2021; Maradana et al., 2017; Pelinescu, 2017; Nae & Grigore, 2014; Hasan & Tucci, 2010; Akçomak & Ter Weel, 2009; Sinha, 2008; Blind & Jungmittag, 2008; Crosby, 2000), while others identified the existence of a negative relationship between the mentioned variables (Silaghi & Medeşfălean, 2014; Domazet et al., 2022). On the other hand, Myszczyszyn (2020) and Blind et al. (2021) argue that there is no relationship between patents and economic growth in the long run. Bearing in mind that Bulgaria and Romania are countries that acceded to the EU at the same time (2007) and that they are characterized by similar issues concerning patents – long-term process, high costs, as well a lack of incentives (Silaghi & Medeşfălean, 2014; World Bank, 2010), it is important to determine the nature of the relationship between patents and (a) economic growth and (b) innovation. In accordance with all the above, this paper aims to examine the relationship between the number of registered patents and economic growth and innovation in Bulgaria and Romania. The paper is structured as follows: the results of relevant studies concerning the relationship between patents, economic growth and innovation are presented in the Literature review section. The methodology and results section describes the sample and applied statistical tests used to examine the nature of the relationship between patents, economic growth and innovation, and then presents the research results and their explanation. The last part of the paper deals with the conclusions of the conducted study. LITERATURE REVIEW Analyzing the nonlinear effect of R&D, patents and exports of high-tech products on economic growth in 35 OECD countries for the period 1992-2006, Ersin et al. (2022) have determined that there are significant marginal effects of economic growth rates, which are followed by threshold effects dominated by the participation of research and development in GDP. According to the research results of Pelinescu et al. (2019) there is a positive and significant impact of growth in R&D expenditure on GDP growth per capita. However, the authors point out that each country should adjust patent rights according to the country's development, as well as the development of the respective industry. In addition, they determined that stronger patent protection is not desirable in underdeveloped countries. Depending on the development of the economy, the contribution of patents to economic growth can differ (Eliasson et al., 2004). Pradhan et al. (2020) point to the role of intellectual property as one of the key drivers of economic growth. Bearing in mind that developed countries are characterized by a significant fund of technological knowledge and human capital (Alnuaimi et al., 2012), as well as appropriate institutional frameworks related to the protection of patents, that is, intellectual property rights (Candelin- Palmqvist et al., 2012), it can be concluded that developed countries may have better abilities to maximize utility from patents. Based on the analysis of data from 99 countries in the period 1996-2018, Rubilar-Torrealba et al. (2022) claim that the more developed a country is, the more patents it tends to have. The increase in the number of patents can contribute to the development of innovations that can have a positive impact on economic growth (Marjanović et al., 2019; Caseiro & Simões, 2019; Pradhan et al., 2019). The number of patents is considered to be an adequate indicator for evaluating the success of innovative activities (Dang & Motohashi, 2015). Patents can be used as a proxy for innovation Crosby (2000) or as an indicator of innovation (Blind et al., 2021). Patent data are more closely related to innovation than research Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 51 and development data. In addition, patent data are more widely available since they cover a longer period and can be used in time series analysis Crosby (2000). The research results, which in 1998-2016 included 43 countries (26 developed economies and 17 developing countries), showed that the number of new patents in the field of information and communication technologies (ICT patents) has a positive one-way impact on economic growth. It is important to note that this type of patent has a positive long-term impact. Also, the authors of the study identified a significant positive impact of patents on economic growth in developed countries. In contrast to developing countries, the impact of patents on economic growth is negative (Nguyen & Doytch, 2022). Analyzing 4 European countries and 12 sectors, Blind & Jungmittag (2008) found that patents contribute to economic growth. Likewise, the study points to the conclusion that patents are more important for growth in R&D-intensive industries. A positive impact of the number of patents on economic growth in the long term was identified in the study of Josheski & Koteski (2011), which covered the G7 countries in the period 1963–1993, while a negative impact was found in the short term. Sinha (2008) examined the relationship between the number of patents and economic growth in Japan and South Korea over the period 1963-2005. In the case of Japan, a two-way causal relationship was identified between the number of patents and real economic growth. In contrast to South Korea, there is a one-way relationship between real GDP growth to the growth in the number of patents. The results of the regression analysis showed that a 1% increase in key technological patents leads to an average increase in GDP per capita by 0.108% (Wurster, 2021). Maradana et al. (2017) determined the existence of a significant connection between patents and economic growth per capita in the long term based on data for 19 European countries in the period 1989–2014. In Romania, Portugal, Belgium, Italy, Germany, Finland, Greece, the Netherlands and the UK, it was found that there is unidirectional causality from the number of resident patents to GDP growth. Additionally, the results point to the conclusion that there is unidirectional causality from economic growth per capita to the number of patents of residents in Hungary, Norway, the Czech Republic, Denmark and Ireland. In addition, a two-way causality was established between the number of resident patents and economic growth per capita. In the case of the number of non-resident patents in Romania, Belgium, the Czech Republic, France, the Netherlands, Spain and Sweden, there is unidirectional causality from the number of non-resident patents to economic growth per capita. Unidirectional causality from economic growth to the number of non-resident patents was identified in Norway, Greece, Finland and Germany. Bidirectional causality was found in Portugal, the UK, Denmark, Ireland and Hungary. The results of the research conducted by Crosby (2000) based on the data on the number of patent applications in Australia in the period 1901-1997 showed that the increase in patents leads to the economic growth of the country. Crosby (2000) established the existence of a positive relationship between patents and economic growth in the short term; however, Schmookler (2013) claims that this relationship should be negative in the short term, while it should be positive in the long term. Akçomak & Ter Weel (2009) claim that there is a positive impact of the number of patent applications per population on the growth of GDP per capita based on the analysis of data from European countries in the period 1990 - 2002. Hasan & Tucci (2010), based on data for 58 states in the period 1980 - 2003, revealed a positive influence of the ratio patents / R&D expenses on the growth of GDP per capita. The existence of a positive relationship between patents and economic growth was established in Romania (Nae & Grigore, 2014). Similar conclusions were obtained in a study conducted by (Pelinescu, 2017) based on data from UNESCO and Eurostat databases for the period 2000-2015. A negative correlation between the number of patent applications and the GDP growth rate was identified in Malaysia, China and Indonesia based on the data from the period 2000-2009 (Saini & Jain, 2011). A study conducted by Silaghi & Medeşfălean (2014) based on the data for the period 1990-2010 in Romania shows that patents have a statistically significant negative impact on economic growth. In addition, it was determined that an increase in the number of 52 Economic Analysis (2022, Vol. 55, No. 2, 49-65) patents by 1% leads to a decrease in GDP by 0.089%. The authors claim that technologies in Romania are most often imitated from abroad. It is also worth mentioning that companies that carry out R&D activities patent their products where they will sell them, which is usually outside the country. The biggest issues are considered to be the time required to obtain a patent, high costs and an underdeveloped market for trading the patented product. Iwaisako & Futagami (2013) explain the negative impact as a result of a high level of patent protection that significantly reduces the demand for capital, which consequently has a negative impact on output. Myszczyszyn (2020) claims that in the long run, there is no relationship between the number of patents and economic growth in Germany based on data from 1872–1913. Similar are the conclusions of Blind et al. (2021), who analyzed data related to the period 1981 - 2014 for 15 EU countries and found that there is no significant impact of patents on economic growth in the long term. DATA ANALYSIS AND FINDINGS Patents represent the product of innovative knowledge and facilitate the spread of technology, thereby stimulating economic growth. The main goal of the work was to determine whether the increase in the number of registered patents per million inhabitants affects economic growth, measured by GDP per capita, and the growth of innovation, measured by the innovation index of a country. The analysis was based on a quantitative research design (comparative approach). The empirical research covered two countries (Romania and Bulgaria) in the period from 2008 to 2018. The analysis used secondary data from internationally recognized databases on the phenomena that were the subject of research in this paper: 1. World bank and 2. World Intellectual Property Organization. The secondary data used in the paper were based on several advantages that these data contain, and for that reason, were taken as relevant. The main reasons for choosing these databases are the ability to access identical data, their immediate availability, and the mutual comparability of data for Romania and Bulgaria. Keeping in mind the aspiration to avoid the possible existence of different measuring instruments for the same phenomena in the national statistics of the countries included in the analysis, the data of international organizations were chosen (Table 1) rather than the data of national statistics. Table 1. Number of patents, innovation index and GDP per capita Country Romania Variable / Year Total patents Total population Number of patents* GII** GDP *** 2008 - 20,537,875 - 2.44 6,730 2009 1,150 20,367,487 56.46 2.92 6,410 2010 1,501 20,246,871 74.13 3.22 6,190 2011 1,599 20,147,528 79.36 36.83 6,350 2012 1,244 20,058,035 62.02 37.80 6,510 2013 1,241 19,983,693 62.10 40.33 6,760 2014 1,252 19,908,979 62.88 38.08 7,020 2015 1,235 19,815,481 62.32 38.20 7,330 2016 1,255 19,702,332 63.70 37.90 7,720 2017 1,452 19,587,491 74.13 39.16 8,320 2018 1,501 19,473,936 77.08 37.59 8,700 Country Bulgaria 2008 - 7,492,561 - 2.12 5,140 2009 397 7,444,443 53.32 2.85 4,990 2010 391 7,395,599 52.87 3.26 5,050 2011 395 7,348,328 53.75 38.42 5,300 Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 53 2012 371 7,305,888 50.78 40.70 5,350 2013 500 7,265,115 68.82 41.33 5,400 2014 467 7,223,938 64.65 40.74 5,530 2015 512 7,177,991 71.33 42.16 5,790 2016 427 7,127,822 59.91 41.42 6,050 2017 425 7,075,947 60.06 42.84 6,310 2018 459 7,024,216 65.34 42.65 6,550 Notes: *number of patents per million inhabitants; **GII = global innovation index; ***GDP per capita Source: WIPO (2019), Country profile – Romania, Bulgaria, available at: https://www.wipo.int/ipstats/en/statistics/country_profile/; World Bank (2019), Total population by country - Romania, Bulgaria, available at: https://data.worldbank.org/indicator/SP.POP.TOTL Since the secondary data available in international databases had the character of time series, appropriate econometric models for time series were used in the analysis. Statistical testing of correlations between the variables shown in Table 1 was performed through simple linear regression for each pair of independent and dependent variables individually. This was performed because the small sample size (n < 30) and the nature of the formulated hypothesis (only one independent and two dependent variables) did not allow the application of a more complex regression technique such as multiple regression or multivariate multiple regression, which require the existence of at least two independent variables and significantly a larger number of observations (Hair et al., 2014). Given that no data were available on the total number of registered patents for Romania and Bulgaria for 2008, it was not possible to calculate the number of patents per capita. If you look at the data on the innovation index from Table 1, you can see that the data for the first three years (2008, 2009 and 2010) were presented using a different methodology compared to the other observation periods. For the results of the statistical analysis to be precise and clear, the research used data for the period from 2011 to 2018. Statistical testing of relationships between variables was performed through simple linear regression for each pair of independent and dependent variables individually for Romania and Bulgaria. In the first case, in the example of Romania, the task was to determine whether all six assumptions (Ass.1 - Ass.6) were fulfilled for both observed variables (the dependent variable is the innovation index; the independent variable is the number of patents per million inhabitants). The results of the simple linear regression are presented in Table 2. Table 2. Verification of fulfillment of assumptions - Case I (Romania) Variable / assumption Number of patents per million inhabitants (n = 8) Innovation index (n = 8) The nature of the variable Metric Metric Distribution diagram Value of indicators of Durbin-Watson statistics d = 2.558 d = 1.539 54 Economic Analysis (2022, Vol. 55, No. 2, 49-65) Variable / assumption Number of patents per million inhabitants (n = 8) Innovation index (n = 8) Histogram P-P normality diagram Source: Authors' research Ass. 1. The task was to determine whether the variables have a continuous nature. The analysis showed that both observed variables have a metric measurement and are therefore treated as metric variables measured on a ratio scale. The assumption is fulfilled. Ass. 2. and Ass. 3. In the conducted analysis, it was not established that there is a linear relationship between the dependent and independent variables. Also, the absence of atypical points was not determined (distribution diagram, table 2). The assumptions are not met. Ass. 4a. Based on the conducted Durbin Watson statistic and the obtained results shown in table 2, it was determined that there is no independence of observations when it comes to the number of patents per million inhabitants (d=2.558). The assumption is not met. Ass. 4b. Based on the conducted Durbin Watson statistic and the obtained results shown in table 2, it was determined that observations are independent when it comes to the country's innovation index (d=1.539). The assumption is fulfilled. Ass. 5. and Ass. 6. Based on the performed analysis and obtained results shown in table 2 (histograms and P-P diagrams of normality), it was determined that there is no absence of heteroskedasticity and normal distribution of residual errors. Assumptions are not made. In the second case, in the example of Romania, the task was to determine whether all six assumptions (Ass.1 - Ass.6) were fulfilled for both observed variables (dependent variable = GDP per capita; independent variable = number of patents per million inhabitants). The results of the simple linear regression are presented in Table 3. Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 55 Table 3. Verification of fulfillment of assumptions - Case II (Romania) Variable / assumption Number of patents per million inhabitants (n = 8) GDP per capita (n = 8) The nature of the variable Metric Metric Distribution diagram Value of indicators of Durbin-Watson statistics d = 0.258 d = 1.140 Histogram P-P normality diagram Source: Authors' research Ass. 1. The task was to determine whether the variables have a continuous nature. The analysis showed that both observed variables have a metric measurement and are therefore treated as metric variables measured on a ratio scale. The assumption is fulfilled. Ass. 2. and Ass. 3. In the conducted analysis, it was not established that there is a linear relationship between the dependent and independent variables. Also, the absence of atypical points was not determined (distribution diagram, table 3). Assumptions are not met. Ass. 4. Based on the conducted Durbin Watson statistic, it was determined that (a) there is no independence of observations when it comes to the number of patents per million inhabitants (d=0.971), (b) there is no independence of observations when it comes to GDP per capita (d=1.052). The assumption is not met. Ass. 5. and Ass. 6. Based on the performed analysis and the obtained results shown in table 3. (histograms and P-P diagrams of normality), it was determined that in the case of the independent variable (number of patents per million inhabitants) there is an absence of heteroscedasticity and normal distribution of residual errors. In contrast, in the case of the dependent variable (GDP per capita), this was not the case. The assumptions are partially fulfilled. Given that the obtained results showed that these assumptions were not met or only partially met, the next task was to transform the data based on the logarithm (table 4). 56 Economic Analysis (2022, Vol. 55, No. 2, 49-65) Table 4. Results of linear regression Model Summaryb Variables The number of patents per million inhabitants and the country's innovation index R R Square Adjusted R Square SE of the Estimate 0.417a 0.174 0.036 0.01177 a. Predictors: (Constant), patent_transf b. Dependent Variable: indeks_i_transf Variables The number of patents per million inhabitants and GDP per capita R R Square Adjusted R Square SE of the Estimate 0.329a 0.108 -0.041 0.05052 a. Predictors: (Constant), patent_transf b. Dependent Variable: BDP_i_transf ANOVAb Variables The number of patents per million inhabitants and the country's innovation index Sum of Squares df Mean Square F p Regression 0.000 1 0.000 1.260 0.305a Residual 0.001 6 0.000 Total 0.001 7 a. Predictors: (Constant), patent_transf b. Dependent Variable: indeks_i_transf Variables The number of patents per million inhabitants and GDP per capita Sum of Squares df Mean Square F p Regression 0.002 1 0.002 0.726 0.427a Residual 0.015 6 0.003 Total 0.017 7 a. Predictors: (Constant), patent_transf b. Dependent Variable: BDP_i_transf Coefficientsa Variables The number of patents per million inhabitants and the country's innovation index Unstandardized Coefficients Standardized Coefficients t p B Std. Error Beta (Constant) 1.777 0.174 -0.417 10.225 0.000 patent_transf -0.107 0.095 -1.122 0.305 a. Dependent Variable: indeks_i_transf Variables The number of patents per million inhabitants and GDP per capita Unstandardized Coefficients Standardized Coefficients t p B Std. Error Beta (Constant) 3.228 0.746 0.329 4.326 0.005 patent_transf 0.347 0.408 0.852 0.427 a. Dependent Variable: BDP_i_transf Source: Authors' research In the simple linear regression model for the variables number of patents per million inhabitants and the country's innovation index, a correlation coefficient of R = 0.417 was determined, which, according to Cohen's criteria, can be considered a medium. Based on the Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 57 obtained results, R2 = 0.174 (coefficient of determination) and Adj.R2 = 0.036 (corrected coefficient of determination), it is concluded that a total of 17.4% and 3.6% of changes in the dependent variable, the country's innovation index, can be explained by changes in the independent variable, the number of patents per million inhabitants. Based on the results of the ANOVA test (F (1,6) = 1.260 and p = 0.305), it can be concluded that the regression model at the p < 0.050 level was not statistically significant. That result provides additional information about the relationship between the independent and dependent variables included in the regression model and shows that the change in the number of patents per million inhabitants does not provide a statistically significant explanation for changes in the country's innovation index. The obtained results showed that the value of the ordinary regression coefficient is B = 1.777 (SE B = 0.174), while the value of the standardized regression coefficient is β = -0.417. Given that the coefficients of correlation and determination had a small value, with the absence of statistical significance of the regression model, it can be concluded that there is no statistically significant relationship between the number of patents per million inhabitants and the country's innovation index in the case of Romania. In the simple linear regression model for the variables number of patents per million inhabitants and GDP per capita, a correlation coefficient of R = 0.329 was determined, which according to Cohen's criteria, can be considered as large (significant). Based on the obtained results R2 = 0.108 (coefficient of determination) and Adj.R2 = -0.041 (corrected coefficient of determination), the conclusion is that a total of 10.8% of changes in the dependent variable GDP per capita can be explained by changes in the independent variable number of patents per million inhabitants. However, this result should be taken with a grain of salt, while the entire explanatory power of the regression model should be assessed as insignificant. Based on the results of the ANOVA test (F (1,6) = 0.726, p = 0.427), it can be concluded that the regression model at the p < 0.050 level was not statistically significant. That result provides additional information about the relationship between the independent and dependent variables included in the regression model and shows that the change in the number of patents per million inhabitants does not provide a statistically significant explanation for changes in the GDP per capita. The obtained results showed that the value of the ordinary regression coefficient is B = 3.228 (SE B = 0.746), while the value of the standardized regression coefficient is β = 0.329. Given that the coefficients of correlation and determination had a small value, with the absence of statistical significance of the regression model, it can be concluded that there is no statistically significant relationship between the number of patents per million inhabitants and the GDP per capita in the case of Romania. In the first case, using the example of Bulgaria, the task was to determine whether all six assumptions (Ass.1 - Ass.6) were fulfilled for both observed variables (dependent variable = innovation index; independent variable = number of patents per million inhabitants). The results of the simple linear regression are presented in Table 5. 58 Economic Analysis (2022, Vol. 55, No. 2, 49-65) Table 5. Verification of fulfillment of assumptions - Case I (Bulgaria) Variable / assumption Number of patents per million inhabitants (n = 8) Innovation index (n = 8) The nature of the variable Metric Metric Distribution diagram Value of indicators of Durbin-Watson statistics d = 1.220 d = 2.029 Histogram P-P normality diagram Source: Authors' research Ass. 1. The task was to determine whether the variables have a continuous nature. The analysis showed that both observed variables have a metric measurement and are therefore treated as metric variables measured on a ratio scale. The assumption is fulfilled. Ass. 2. and Ass. 3. In the conducted analysis, it was not established that there is a linear relationship between the dependent and independent variables. Also, the absence of atypical points was not determined (distribution diagram, table 5). The assumptions are not met. Ass. 4a. Based on the conducted Durbin Watson statistic and the obtained results shown in table 5, it was determined that there is no independence of observations when it comes to the number of patents per million inhabitants (d=1.220). The assumption is not met. Ass. 4b. Based on the conducted Durbin Watson statistic and the obtained results shown in table 5, it was determined that observations are independent when it comes to the country's innovation index (d=2.029). The assumption is fulfilled. Ass. 5. and Ass. 6. The results shown in table 5 (histograms and P-P diagrams of normality) were intended to show the absence of heteroskedasticity and the normal distribution of residual errors in the dependent and independent variables. The assumptions are partially fulfilled. In another case, using the example of Bulgaria, the task was to determine whether all six assumptions (Ass.1 - Ass.6) were fulfilled for both observed variables (dependent variable = GDP per capita; independent variable = number of patents per million inhabitants). The results of the simple linear regression are presented in Table 6. Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 59 Table 6. Verification of fulfillment of assumptions - Case II (Bulgaria) Variable / assumption Number of patents per million inhabitants (n = 8) GDP per capita (n = 8) The nature of the variable Metric Metric Distribution diagram Value of indicators of Durbin-Watson statistics d = 0.322 d = 1.711 Histogram P-P normality diagram Source: Authors' research Ass. 1. The task was to determine whether the variables have a continuous nature. The analysis showed that both observed variables have a metric measurement and are therefore treated as metric variables measured on a ratio scale. The assumption is fulfilled. Ass. 2. and Ass. 3. In the conducted analysis, it was not established that there is a linear relationship between the dependent and independent variables. Also, the absence of atypical points was not determined (distribution diagram, table 6). Assumptions are not met. Ass. 4a. Based on the conducted Durbin Watson statistic and the obtained results shown in table 6, it was determined that there is no independence of observations when it comes to the number of patents per million inhabitants (d=0.322). The assumption is not met. Ass. 4a. Based on the conducted Durbin Watson statistic and the obtained results shown in table 6, it was determined that observations are independent when it comes to GDP per capita (d=1.711). The assumption is fulfilled. Ass. 5. and Ass. 6. The results shown in table 6 (histograms and P-P diagrams of normality) were intended to show the absence of heteroscedasticity and the normal distribution of residual errors in the dependent and independent variables. The assumptions are partially fulfilled. Given that the obtained results showed that these assumptions were not met or only partially met, the next task was to transform the data based on the logarithm (table 7). 60 Economic Analysis (2022, Vol. 55, No. 2, 49-65) Table 7. Results of linear regression Model Summaryb Variables The number of patents per million inhabitants and the country's innovation index R R Square Adjusted R Square SE of the Estimate 0.560a 0.313 0.199 0.01349 a. Predictors: (Constant), patent_transf b. Dependent Variable: indeks_i_transf Variables The number of patents per million inhabitants and GDP per capita R R Square Adjusted R Square SE of the Estimate 0.319a 0.102 -0.048 0.03583 a. Predictors: (Constant), patent_transf b. Dependent Variable: BDP_i_transf ANOVAb Variables The number of patents per million inhabitants and the country's innovation index Sum of Squares df Mean Square F p Regression 0.000 1 0.000 2.739 0.149a Residual 0.001 6 0.000 Total 0.002 7 a. Predictors: (Constant), patent_transf b. Dependent Variable: indeks_i_transf Variables The number of patents per million inhabitants and GDP per capita Sum of Squares df Mean Square F p Regression 0.001 1 0.001 0.680 0.441a Residual 0.008 6 0.001 Total 0.009 7 a. Predictors: (Constant), patent_transf b. Dependent Variable: BDP_i_transf Coefficientsa Variables The number of patents per million inhabitants and the country's innovation index Unstandardized Coefficients Standardized Coefficients t p B Std. Error Beta (Constant) 1.320 0.179 0.560 7.395 0.000 patent_transf 0.165 0.100 1.655 0.149 a. Dependent Variable: indeks_i_transf Variables The number of patents per million inhabitants and GDP per capita Unstandardized Coefficients Standardized Coefficients t p B Std. Error Beta (Constant) 3.370 0.474 0.319 7.106 0.000 patent_transf 0.219 0.265 0.825 0.441 a. Dependent Variable: BDP_i_transf Source: Authors' research In the simple linear regression model for the variables number of patents per million inhabitants and the country's innovation index, a correlation coefficient of R = 0.560 was Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 61 determined, which according to Cohen's criteria, can be considered large (significant). Based on the obtained results R2 = 0.313 (coefficient of determination) and Adj.R2 = 0.199 (corrected coefficient of determination), it is concluded that a total of 31.3% and 19.9% of changes in the dependent variable, the country's innovation index, can be explained by changes in the independent variable, the number of patents per million inhabitants. Based on the results of the ANOVA test (F (1,6) = 2.739 and p = 0.149), it can be concluded that the regression model at the p < 0.050 level was not statistically significant. That result provides additional information about the relationship between the independent and dependent variables included in the regression model and shows that the change in the number of patents per million inhabitants does not provide a statistically significant explanation for the changes in the country's innovation index. The obtained results showed that the value of the ordinary regression coefficient is B = 1.320 (SE B = 0.179), while the value of the standardized regression coefficient is β = 0.560. Given that the coefficients of correlation and determination had a small value, with the absence of statistical significance of the regression model, it can be concluded that there is no statistically significant relationship between the number of patents per million inhabitants and the country's innovation index in the case of Bulgaria. In the simple linear regression model for the variables number of patents per million inhabitants and GDP per capita, a correlation coefficient of R = 0.319 was determined. According to Cohen's criteria, it can be considered a medium. Based on the obtained results R2 = 0.102 (coefficient of determination) and Adj.R2 = -0.048 (corrected coefficient of determination), the conclusion is that a total of 10.2% of changes in the dependent variable GDP per capita can be explained by changes in the independent variable number of patents per million inhabitants. However, this result should be taken with a grain of salt, while the entire explanatory power of the regression model should be assessed as insignificant. Based on the results of the ANOVA test (F (1,6) = 0.680, p = 0.441), it can be concluded that the regression model at the p < 0.050 level was not statistically significant. That result provides additional information about the relationship between the independent and dependent variables included in the regression model and shows that the change in the number of patents per million inhabitants does not provide a statistically significant explanation for changes in the GDP per capita. The obtained results showed that the value of the ordinary regression coefficient is B = 3.370 (SE B = 0.474), while the value of the standardized regression coefficient is β = 0.319. Given that the coefficients of correlation and determination had a small value, with the absence of statistical significance of the regression model, it can be concluded that there is no statistically significant relationship between the number of patents per million inhabitants and the GDP per capita in the case of Bulgaria. CONCLUSION The national innovation system of Romania includes numerous institutions, organizations and agencies that have a highly developed organizational structure. However, the developed organizational structure in the case of the national innovation system of Romania does not imply its efficiency because the system is excessively fragmented and needs to provide equal access to all actors in terms of financial support. Financial investments in the field of research, development and innovation in Romania in the analyzed period (2008-2018) are very low, the results of the policy in this area are modest, the demand for innovation is low while the supply of human resources is inadequate. In 2014, a new Law on Innovations was adopted with the task of interpreting private property rights and improving the development of patents, which was supposed to have a direct impact on the economy. Nevertheless, there was still a low interest of companies in activities in the field of research, development and innovation, and the majority of companies justified their attitude with large allocations, the riskiness of the venture and the uncertainty of the results. A small number of innovations derive from the structure of the Romanian economy itself, in which low and medium technologies prevail. 62 Economic Analysis (2022, Vol. 55, No. 2, 49-65) On the other hand, during the analyzed period, Bulgaria's innovation system, although institutionally developed, suffered from two pressing challenges - inefficiency and the absence of significant financial resources that would ensure its development. In addition, the main problems faced by Bulgaria in the field of research, development and innovation are their small application in the business sphere, weak demand for innovations, and poor cooperation between the academic community and the economy. However, despite all the developments that have taken place in the previous 10 years or so, Bulgaria's innovation system is still facing major challenges, among which are the basic ones: how to ensure the continuity of a satisfactory intensity of investment in research and development, how to improve cooperation with the business sector; and how to create a climate and framework for the introduction of public- private partnership in the research and development system. Of all the forms of intellectual property and technological innovation, patents are most often associated with them since they protect the essence of technological innovation. To create innovations to introduce new and improved products or services to the market, or to introduce new and improved production processes, companies must invest in research and development. Therefore, the conducted research aimed to determine whether the increase in the number of patents per million inhabitants affects economic growth and the increase in the innovation index of Romania and Bulgaria. Based on the obtained coefficients of correlation and determination and the absence of statistical significance of the regression model, the obtained results of the empirical research showed that in the case of Romania and Bulgaria, the following applies: (a) there is no statistically significant correlation between the number of patents per million inhabitants and the innovation index (b) there is no statistically significant correlation between the number of patents per million inhabitants and GDP per capita. If the national innovation systems of Romania and Bulgaria were to be compared, it could be concluded that there is an evident lack of cooperation between the business sector and the academic community. In addition, innovation in the economy is at a very low level. This is reflected in the entire system, which is inefficient and underfunded. All this supports the fact that the results obtained through this research are not a big surprise. In the analyzed countries, there are still many open dilemmas regarding cooperation between science and business. Thereupon, there is little demand for innovation, but also an open question of the success of applying for patents in the business sector. Based on all of the above, it is highly debatable whether Romania and Bulgaria would have more significant results if: • there was an increase in the number of patents per capita, • they invested more funds in the area of research and development, and • they hired a significantly larger number of better-quality researchers. The results of the conducted research, to a certain extent, can be of importance to decision- makers in Romania and Bulgaria through certain recommendations, which refer to the efficiency of national innovation systems and the level of demand for innovation within each national economy. Therefore, it is very important that policymakers in the field of innovation adequately present the impact that innovation has on the country's economy. An increase in the number of patents per million inhabitants does not guarantee that in these countries, there will be an increase in the innovativeness of the economy and an increase in well-being. For this to happen, it will be necessary for patents to have their application and for all actors in that process to understand that only the application of patents can enable significantly better results in business. The conducted research also has several limitations. The first drawback is the relatively small sample, considering that the analysis was performed only for two countries - Romania and Bulgaria. For subsequent research, it is proposed to increase the sample to four or more countries, with the recommendation that they are countries of the European Union or countries Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 63 of one region. The small number of indicators analyzed is another limitation of this research. Given that the focus of this paper was on determining the impact of the number of patents on innovation and economic growth of the country, the recommendation for future research is to expand the list of indicators to be analyzed (e.g., production growth, employment growth, competitiveness). In this research, the analysis was performed for the period 2008-2018, which represents the third limitation of this work. In some future research, it is recommended to use data from as long a time series as possible (e.g. 30 years). ACKNOWLEDGEMENTS This paper is financed by the Ministry of Education, Science and Technological Development of the Republic of Serbia. REFERENCES Akçomak, S., & Ter Weel, B. (2009). Social Capital, Innovation and Growth: Evidence from Europe. European Economic Review, 53(5), 544-567. Alnuaimi, T., Singh, J., & George, G. (2012). Not with my own: Long-term effects of cross- country collaboration on subsidiary innovation in emerging economies versus advanced economies. Journal of Economic Geography, 12(5), 943-968. Atun, R.A., Harvey, I., & Wild, J. (2007). Innovation, patents and economic growth. International Journal of Innovation Management, 11(02), 279-297. Blind, K., & Jungmittag, A. (2008). The impact of patents and standards on macroeconomic growth: a panel approach covering four countries and 12 sectors. Journal of Productivity Analysis, 29(1), 51-60. Blind, K., Ramel, F., & Rochell, C. (2021). The influence of standards and patents on long-term economic growth. The Journal of Technology Transfer, 1-21. Candelin-Palmqvist, H., Sandberg, B., & Mylly, U.M. (2012). Intellectual property rights in innovation management research: A review. Technovation, 32(9-10), 502-512. Caseiro, C., & Simões, M.C.N. (2019). Business Sector Innovation and Economic Growth a Comparative Analysis Between EU Countries. Economic Analysis, 52(1), 1-22 Crosby, M. (2000). Patents, innovation and growth. Economic Record, 76(234), 255-262. Dang, J., & Motohashi, K. (2015). Patent statistics: A good indicator for innovation in China? Patent subsidy program impacts on patent quality. China Economic Review, 35, 137-155. Domazet, I., Marjanović, D., & Ahmetagić, D. (2022). The Impact of High-Tech Products Exports on Economic Growth: The Case of Serbia, Bulgaria, Romania and Hungary. Ekonomika preduzeća, 70(3-4), 191-205. Domazet, I., Marjanović, D., Ahmetagić, D., & Bugarčić, M. (2021). The Impact of Innovation Indicators on Increasing Exports of High Technology Products. Ekonomika preduzeća, 69(1-2), 31-40. Eliasson, G., Johansson, D., & Taymaz, E. (2004). Simulating the new economy. Structural Change and Economic Dynamics, 15(3), 289-314. Ersin, Ö., Ustabaş, A., & Acar, T. (2022). The Nonlinear Effects Of High Technology Exports, R&D And Patents On Economic Growth: A Panel Threshold Approach To 35 OECD Countries. Romanian Journal of Economic Forecasting, 25(1), 26-44. European Patent Office (EPO). (2022). Bulgaria marks 10 years of membership in the European Patent Organisation. https://www.epo.org/news-events/news/2012/20121022.html Hair, J.F., Black, W.C., Babin, B.J., & Anderson, R.E. (2014). Multivariate Data Analysis. Harlow: Pearson Education. Hasan, I., & Tucci, C.L. (2010). The innovation–economic growth nexus: Global evidence. Research policy, 39(10), 1264-1276. http://ebooks.ien.bg.ac.rs/1613/ http://ebooks.ien.bg.ac.rs/1613/ https://www.epo.org/news-events/news/2012/20121022.html 64 Economic Analysis (2022, Vol. 55, No. 2, 49-65) Iwaisako, T., & Futagami, K. (2013). Patent protection, capital accumulation, and economic growth. Economic Theory, 52(2), 631-668. Josheski, D., & Koteski, C. (2011). The causal relationship between patent growth and growth of GDP with quarterly data in the G7 countries: cointegration, ARDL and error correction models. https://mpra.ub.uni-muenchen.d /33153/1/MPRA_paper_33153.pdf Maradana, R.P., Pradhan, R.P., Dash, S., Gaurav, K., Jayakumar, M., & Chatterjee, D. (2017). Does innovation promote economic growth? Evidence from European countries. Journal of Innovation and Entrepreneurship, 6(1), 1-23. Marjanović, D., Ahmetagić, D., & Beraha, I. (2019). Comparative Analysis of High Technology Exports and Selected Innovation Indicators for Serbia and CEE Countries. Economic Analysis, 52(2), 93-103. Moser, P. (2013). Patents and innovation: evidence from economic history. Journal of Economic Perspectives, 27(1), 23-44. Myszczyszyn, J. (2020). The Long-run Relationships between Number of Patents and Economic Growth. European Research Studies, 23(3), 548-563. Nae, G.G., & Grigore, C. (2014). Economic growth at regional level and innovation: Is there any link? Global Journal of Business, Economics and Management, 4(1), 16-21. Nguyen, C.P., & Doytch, N. (2022). The impact of ICT patents on economic growth: An international evidence. Telecommunications Policy, 46(5). OECD. (2004). Patents and innovation: Trends and policy challenges. OECD Publishing. Pelinescu, E., Pauna, C., Saman, C., & Diaconescu, T. (2019). Human capital, innovation and economic growth in the EU countries. Romanian Journal of Economic Forecasting, 22(4), 160- 173. Pelinescu, E. (2017). The Impact of Human Capital on Innovation and Economic Growth in Romania. Internal Auditing & Risk Management, 12(2), 44-50. Pradhan, R.P., Arvin, M.B., Nair, M., & Bennett, S.E. (2020). Sustainable economic growth in the European Union: The role of ICT, venture capital, and innovation. Review of Financial Economics, 38(1), 34-62. Pradhan, R.P., Arvin, M.B., Nair, M., Bennett, S.E., & Hall, J.H. (2019). The information revolution, innovation diffusion and economic growth: an examination of causal links in European countries. Quality & Quantity, 53(3), 1529-1563. Rubilar-Torrealba, R., Chahuán-Jiménez, K., & de la Fuente-Mella, H. (2022). Analysis of the Growth in the Number of Patents Granted and Its Effect over the Level of Growth of the Countries: An Econometric Estimation of the Mixed Model Approach. Sustainability, 14(4). Saini, A.K., & Jain, S. (2011). The Impact of Patent Applications Field on Sustainable Development of Selected Asian Countries. International Journal of Information Technology, 3(2), 358-364. Schmookler, J. (2013). Invention and economic growth. Harvard University Press. Silaghi, M.P., & Medeşfălean, R. (2014). Some insights about determinants of economic growth in Romania. An empirical exercise. Theoretical & Applied Economics, 21(6), 23-36. Sinha, D. (2008). Patents, innovations and economic growth in Japan and South Korea: evidence from individual country and panel data. Applied Econometrics and International Development, 8(1), 1-23. Word Bank. (2010). Bulgaria's Hidden Growth Potential lies in Research, Innovation and Technology. https://www.worldbank.org/en/news/press-release/2010/06/21/bulgarias- hidden-growth-potential-lies-research-innovation-technology World Bank. (2022). Charting a Path to Future Prosperity with the Long Term Growth Model. https://www.worldbank.org/en/news/feature/2022/02/15/charting-a-path-to-future- prosperity-with-the-long-term-growth-model/ https://mpra.ub.uni-muenchen.d/ http://ebooks.ien.bg.ac.rs/1402/ http://ebooks.ien.bg.ac.rs/1402/ https://www.worldbank.org/en/news/press-release/2010/06/21/bulgarias-hidden-growth-potential-lies-research-innovation-technology https://www.worldbank.org/en/news/press-release/2010/06/21/bulgarias-hidden-growth-potential-lies-research-innovation-technology https://www.worldbank.org/en/news/feature/2022/02/15/charting-a-path-to-future-prosperity-with-the-long-term-growth-model/ https://www.worldbank.org/en/news/feature/2022/02/15/charting-a-path-to-future-prosperity-with-the-long-term-growth-model/ Ivana Domazet, Darko Marjanović, Deniz Ahmetagić, Marija Antonijević 65 Wurster, M. (2021). Intellectual property – How key technology patents stimulate economic growth. https://globaleurope.eu/globalization/intellectual-property-how-key-technology- patents-stimulate-economic-growth/ Article history: Received: 26.9.2022. Revised: 28.11.2022. Accepted: 5.12.2022. https://globaleurope.eu/globalization/intellectual-property-how-key-technology-patents-stimulate-economic-growth/ https://globaleurope.eu/globalization/intellectual-property-how-key-technology-patents-stimulate-economic-growth/