The effects of intellectual property rights protection in the technology transfer context on economic growth: the case of developing countries (2024)

10According to the North–South theoretical literature, the nature of a stronger patent system in developing countries’ influence on technology transfer is ambiguous and dependscrucially on the channel of international production transfer from North to South, and on the channels through which technology can be transferred from North to South (foreign direct investment–FDI –, licensing, and trade).

11Helpman (1993) argues that tighter IPRs reduce technology flows from developed countries to developing ones if imitation is the channel of international production transfer. Stronger IPRs increase imitation cost, they restrict southern imitation, and shift production back to the North. In the long run, innovation in the North may fall because fewer resources are available for innovation there. Lai (1998), on the other hand, argues thatstronger IPRs protection increases the rate of technology transfer if multinationalization is the channel of production transfer, but has the opposite effect if production is transferred through imitation. When FDI is the channel of production transfer, strengthening IPRs protection is seen as an incentive given by the South to encourage northern technology flows. Stronger IPRs protection in the South would improve the incentives for innovation in the North, and encouragesnorthern FDI. This makes more northern labor available for research.

12According to other studies where licensing is assumed, the means by which technology can spread internationally, the implications of stronger IPRs protection in the South on technology transfer from the North to the South, are not clear cut. Yang and Maskus (2001) argue that stronger IPRs in the South incite firms in the North to license advanced technologies. They increase the licensor’s share of rents and reduce the costs of licensing contracts. Consequently, additional resources would be available for R&D. Hence, stronger IPRs in the South are associated with increased innovation in the North and technology transfer. Similarly, Yang and Maskus (2009) show that stronger IPRsin developing countries would enhance technology transfer through licensing and reduce the South firm’s marginal production cost, thereby increasing its competitiveness in the international markets. They argue that stronger IPRs protection may improve the ability of firms in developing countries to enter export markets. In contrast to these studies, Yang etal. (2016) show that a stronger patent system’s ability to attract better technology depends on the strength of two factors: its effect on technology transfer cost and its effect on knowledge spillover. A stronger patent system decreases (increases) the quality of the technology to be transferred in the reforming country if its effect on knowledge spillover is stronger (weaker) than its effect on the technology transfer cost.

13Like the other channels of technology transfer, ambiguous conclusions on the impact of IPRs protection in the South on international trade are drawn from the theoretical literature. According to Maskus and Yang (2013), strengthening IPRs protection in developing countries encourages inward technology transfer via trade. They argue thatbecause strengthened patent rights can expand technology transfer, such reforms could in turn improve a country’s export performance by enhancing the productivity of local enterprises. However, Smith (2001) argues that stronger IPRs protection positively affects technology transfer via trade, under certain circ*mstances. More specifically, the impact of stronger IPRs protection on technology transfer occurring through imports depends on two contradictory effects: a ‘market expansion’ effect and a ‘market power’ effect. Strengthening IPRs protection can positively affect trade if the ‘market expansion’ effect dominates the ‘market power’ effect. According to the first effect, trade is positively correlated with the effectiveness of protection in the importing country. Strengthening IPRs protection can reduce counterfeits, resulting in an increase in the request for innovative businesses. According to the second effect, strengthening IPRs protection may have a negative effect on trade flows. Greater protection allows the patent holder firm to reduce the quantities exported through its monopoly power in the host country market, and increases its price. The total effect depends on the imitation capacity of the importing country (Smith, 1999).

14The effects of IPRs protection on economic growth have been analyzed by a number of North-South models in theoretical literature, but mixed results have been found. Much depends on the assumptions about the activities of imitation and innovation in both regions.

15From a North-South model, Grossman and Helpman (1991, Chapter11) analyzed the consequences of imperfect IPRs protection on technical progress and economic growth by assuming that innovation takes place only in the North and imitation only takes place in the South. According to these authors, imperfect IPRs protection encourages imitation activity in the South. Thanks to the increase in the number of imitation goods, the southern stock of knowledge increases. Conversely, a strong IPRs protection could reduce the rate of imitation. Due to the increased difficulty to imitate, the duration of monopoly profits of the northern innovator is longer, and this monopoly position lasts. The productivity of southern imitators decreases, and the country’s stock of knowledge will not be stimulated. In contrast, the imitation activity generates two contradictory effects on the North. The positive effect is that technical progress is stimulated to the extent that the northern firm innovates to survive under the pressure of low-cost southern imitators. The negative effect is due to the disappearance of the rent of the innovator from the moment the variety of a product is imitated.

16However, assuming a co-existence of imitation and innovation in developing countries, Hwang etal. (2016) show that the economic effects of IPRs protection vary according to the level of economic development. Tighter IPRs protection reduces the varieties of the intermediate goods and raises the production cost of the final good. These effects rise and then decline with income. Therefore, the relationship between the less-developed countries’ IPRs protection and economic development is U-shaped. This implies that a country’s willingness to strengthen its IPRs protection first decreases and then increases with its income.

17To sum up, theoretical predictions about the effects of IPRs protection on technology transfer and economic growth are ambiguous. A clear picture can appear from an empirical study. In this paper, an econometric approach is chosen to answer this question by focusing on only one channel of technology transfer: imports of capital goods.

21For the first equation, we assume that technology transfer through capital goods imports is a determinant of TFP growth rate. In the technology transfer literature (Connolly, 2003; Coe etal., 1997; Grossman, Helpman, 1991), thanks to openness on international trade, imports of a country may relate to varieties of capital goods that are not available on the domestic market. Similarly, following the liberalization of capital goods, domestic producers in a country can use domestic and foreign inputs for the production of the final good. Moreover, trade in capital goods gives an advantage to developing countries in terms of new knowledge made in advanced countries.

22Indeed, the import of capital goods is not the only way by which international trade may affect TFP growth. Grossman and Helpman (1991) have identified the channels through which trade can have an effect on growth. Alongside its role in technology transfer, trade contributes to improving the production efficiency of a country, insofar as it allows it to benefit from the positive effects of competition and economies of scale. Openness promotes efficient allocation of resources and leads to an optimization of production processes. To capture the latter effects, we have opted for the inclusion of a measure of trade openness as a determinant of TFP growth rate. The volume of exports and imports as a share of total GDP; often referred to as the “trade openness ratio” will be used as a measure of the openness of a country.

23Furthermore, developing countries import capital goods from developed countries since they are more advanced in terms of technology. These imports are expected to promote technological catch-up across countries and lead to faster rates of economic growth in the importing countries, which would imply some form of convergence. However, this process of convergence is conditional in the sense that the growth rate depends, for a given initial technology gap, on national policies and other variables that influence the growth rate of the importing country. To test the hypothesis of convergence, the majority of empirical studies introduce a measure of the initial situation in their growth regressions. Some study the link between GDP per capita growth rate and the logarithm of initial per capita GDP (Romer, 1993; Mankiw etal., 1992; Barro, 1991). However, Edwards (1998) explained the growth rate of TFP based on a log of initial per capita GDP. In the various cases, a negative relationship between the growth rate and the initial situation allows the convergence for the considered sample of countries to be highlighted.

24Moreover, the catching-up process of the developed countries’ technological level is achieved as the developing country adopts foreign technology and is dependent on a high absorption capacity. Most economists (Ben Habib, Spiegel, 1994; Grossman, Helpman, 1991; Nelson, Phelps, 1966) see in human capital the main factor that determines not only the ability of a country to innovate, but also to assimilate foreign technology and catch up the technological level of developed countries. Specifically, they assume that there is a link between the level of education of a country and its capacity to absorb foreign technology. To test these arguments and the hypothesis of convergence, we include human capital and initial technological level as determinants of TFP growth rate.

25Lastly, we also include IPRs protection as an explanatory variable in the first equation in order to examine if it has a direct effect on TFP growth rate. Relying on existing studies, IPRs protection motivates innovation and encourages research, it thereby increases economic growth (Falvey, Foster, 2006; Park, Ginarte, 1997; Gould, Gruben, 1996).

26In the literature, the indicators of IPRs protection are two in number: the Rapp and Rozek indicator (1990), and that of Park and Ginarte (1997). In this work, we will use this last indicator, which is most used in the recent literature for empirical research. It covers a relatively larger number of countries and a longer period of time. This indicator is constructed by Park and Ginarte (1997) and it is updated by Park (2008). The indicator on IPRs protection ranges from zero (no patent system) to five (strongest level of protection). The value of the index is obtained by aggregating the following five components: extent of coverage, duration of protection, absence of restrictions on rights, membership of international treaties, and statutory enforcement provisions. [2]

27Thus, equation1 examines the determinants of the growth rate of TFP. The explanatory variables are the initial technological level, human capital, the import of capital goods, the degree of openness (openness rate) and IPRs protection. Table1 summarizes all these variables.

28The first equation is written as follows:

29

30where i refers to a country and t to the time period. ε is an error term.

31The expected signs of the coefficients: β₁<0, β₂> 0, β₃> 0 and β₄> 0. The sign of the estimated coefficient of In(IPR); β₅ is to be determined.

32In equation2, it is assumed that technology transfer through the importation of capital goods is the endogenous variable. We include IPRs protection as an explanatory variable.

33Some empirical studies, such as those of Ivus (2010), Awokuse and Yin (2010), Smith (1999), explained trade in goods by IPRs protection regimes in the importing countries. For example, Ivus (2010) explained the northern countries’ exports (24 OECD countries) to developing countries (55developing countries) through the southern IPRs policy. The author distinguished between exports of sensitive industries by patent protection, and those of insensitive industries by patent protection. Awokuse and Yin (2010) explained the imports of Chinese technology intensive sectors (chemistry, electronics, instruments and machines) by the IPRs policy. Smith (1999) explained the bilateral exports of the United States economy by the regimes of IPRs protection, the tariff policy of the importing country, as well as other explanatory variables (per capita income, population, the distance between partner countries).

Table 1

Variable Definition Source Dependent Variable TFP TFP ??: growth rate of TFP the average annual growth rate of TFP in a developing country calculated over periods 1970-74,1975-79, 1980-1984,1985-89, 1990-1994, 1995-1999, 2000- 2004 and 2005-2009 – To calculate the average annual growth rate of TFP, we relied on the work of Bernanke and Gürkaynak (2001). – Database Alan Heston, Robert Summers and Bettina Aten, Penn World Table Version 7.1, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, July 2012 Explanatory Variables In(y0) : initial technological level the logarithm of initial per capita GDP of a developing country calculated for the initial year of each period, i.e. for 1970, 1975, 1980, 1985, 1990, 1995, 2000 and 2005 Database Alan Heston, Robert Summers and Bettina Aten, Penn World Table Version 7.1, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, July 2012 In(HK) : human capital Three different measures of human capital are adopted: In(atys), In(ayss) and In(ayts) Basic data of Barro, R. and Lee, J. W., A New Data Set of Educational Attainment in the World, 1950-2010. Website: http://www.barrolee.com/ For more detail about data set of educational attainment, see Barro and Lee (2013) In(atys) the logarithm of the average years of total schooling (for the population aged over 15 years) calculated for the initial year of each period, i.e. for 1970, 1975.1980, 1985, 1990, 1995, 2000 and 2005 In(ayss) the logarithm of the average years of secondary schooling (for the population aged over 15 years) calculated for the initial year of each period, i.e. for 1970, 1975.1980, 1985, 1990, 1995, 2000 and 2005 In(ayts) the logarithm of the average years of tertiary schooling (for the population aged over 15 years) calculated for the initial year of each period, i.e. for 1970, 1975, 1980, 1985, 1990, 1995, 2000 and 2005

In(Mtech) : import of high technology goods the logarithm of the average of the share of imports of capital goods in GDP. The average is calculated for each of the periods 1970-1974, 1975-1979, 1980-1984, 1985-1989, 1990-1994, 1995-1999, 2000-2004 and 2005-2009 We use the share of imports Machinery and transport equipment (class 7, SITC, Standard Classification for International Trade) in GDP. The data source is different numbers of UNCTAD Handbook of Statistics, the United Nations Conference on Trade and Development (UN) UNCTAD Handbook of Statistics. Imports are expressed in US dollars (current).

34As presented above, a strengthening of IPRs protection has two contradictory effects on trade flows. They are a ‘market expansion’effect and a ‘market power’ effect (Smith, 2001). The results of our empirical analysis will suggest which effect dominates the other for the sample of countries.

35Similarly, in Equation 2, it is assumed that technology transfer via the import of capital goods depends on the degree of openness of the country, in addition to IPRs protection. A growing openness to international trade promotes the number of foreign varieties of capital goods in an economy which promotes technology transfer. Among the openness measures defined in the literature are indicators related to trade policies. The index of freedom to trade internationally is used in this work as a determinant of technology transfer through the import of capital goods. This index of freedom to trade internationally varies on a scale from one to ten. It is an aggregating measure of restraints that affect international exchange: (i)taxes on international trade,(ii)regulatory trade barriers, (iii)black market exchange rates, and (iv)controls of the movement of capital and people. [3]

36On the other hand, there are some threshold levels of development below which a country may not be able to adopt new technologies (Nelson, Phelps, 1966). So we predict that the higher the technology level of the host country, the higher the import of high-tech goods will be. In our analysis, the technology level of a country is measured by the ratio of per capita income relative to that of the United States, and is included as an explanatory variable in the technology transfer equation. We test the impact of the technology level of a country on the volume of inflows of foreign technology.

37The choice of the United States is largely owing to its position in the global economy. In many studies, it is taken to be the leading country, i.e. the economic dominant country. Moreover, this country was the largest global exporter of capital and equipment goods for many years (Thorbecke, 2012).

38Finally, it is assumed that a country’s membership of the WTO boosts its imports of foreign technologies embodied in capital goods since, by signing the TRIPS Agreement, the country agrees to strengthen IPRs protection.

39Table2 describes the evolution of the growth rate of the IPRs indicator of Park and Ginarte (calculated on average) over the 1960-2005 period for all developed countries and the sample of developing countries. The sample of developed countries includes the following countries: Australia, Austria, Belgium, Canada, Denmark, Spain, USA, Finland, France, Greece, Iceland, Ireland, Italy, Japan, Luxembourg, Malta, Norway, New Zealand, the Netherlands, Portugal, the United Kingdom, Sweden and Switzerland.

40The first conclusion that emerges from this table is that during the period 1975-1990, the growth rate of the IPRs indicator (calculated average) in our sample of developing countries is lower than those achieved in developed countries. Conversely, the situation was reversed during the period 1990-2005. The growth rates of IPRs indicator became relatively higher in developing countries.

41This phenomenon can be explained by the fact that most developed countries started the establishment of a system of IPRs protection relatively earlier. However, developing countries have accelerated the growth of IPRs protection and have implemented reforms in this area after 1990 in order to gain access to the WTO.

42Table2 shows that developing countries who were signatories to the TRIPS Agreement in 1995 registered the highest growth rate of IPRs protection during 1990-1995. However, since developing countries signed the agreement in 1996, the highest growth rate of IPRs protection was recorded during the period 1995-2000. Similarly, Jordan signed the TRIPS Agreement in 2000 and greatly strengthened IPRs protection between 1995 and 2000. In addition, Nepal registered a not-null growth rate of IPRs protection only during the period 2000-2005. The country signed the agreement in 2004.

43It appears that strengthening the protection of IPRs characterized the period in which the TRIPS Agreement was signed by countries in the sample.

44We then create a dummy variable referred to as TRIPS regarding the period during which the TRIPS Agreement was signed by a sample of countries. The value of 1 is assigned to the period in which the TRIPS Agreement was signed by the country. The value zero is allocated to other periods. We include this dummy variable as our final determinant of technology transfer.

45To summarize, equation2 examines the determinants of the import of high-tech goods. The explanatory variables are IPRs protection, the degree of openness of the country, its technological level, and the TRIPS variable. Table3 summarizes these variables.

Table 3

Variable Definition Source Dependent Variable In(Mtech) : import of high technology goods This is the same variable defined in equation 1 and table 1 Explanatory Variables In(IPR) : IPRs protection This is the same variable defined in equation 1 and table 1. In(ftrade) : Freedom to Trade Internationally The logarithm of the indicator “Freedom to Trade Internationally”. It is set for the initial year of the following periods: 1970-74, 1975-79, 1980- 84, 1985-1989, 1990-1994 and 1995-1999. For the 2000-2004 and 2005-2009 periods, the average of this indicator is calculated Data Base James Gwartney, Robert Lawson, and Joshua Hall, Fraser Institute, year: 2012 URL: http://www. freetheworld.com/datasets efw.html In y y USA 0 0 : technological level The logarithm of the ratio of GDP per capita of a country relative to that of the US economy. It is set for the initial year of each period, i.e. for 1970, 1975, 1980, 1985, 1990, 1995, 2000 and 2005. Database Alan Heston, Robert Summers and Bettina Aten, Penn World Table Version 7.1, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, July 2012 URL: http://pwt.econ. upenn.edu/.

TRIPS : TRIPS Agreement A dummy variable regards the period during which the TRIPS Agreement was signed by a country. The value of 1 is assigned to the period in which the TRIPS Agreement was signed by the country. The value zero is allocated to other periods. World Trade Organization - Agreement on Trade-Related Aspects of Intellectual Property Rights-Contracting Parties/Signatories

46The second equation can be written as follows:

47

48i denotes a country. t denotes the time. ε’ is an error term.

49The expected sign of δ₁ coefficient is to be determined. The expected signs of δ₂, δ₃ and δ₄ coefficients are positive.

50The expected sign of δ₁ coefficient is to be determined because, as was discussed above, the impact of strengthening IPRs protection on trade flows depends on two contradictory effects; ‘market expansion’ or ‘market power’ effects. The first effect is positive, while the second effect is negative. The total effect is then to be determined. If the sign of δ₁ is positive and significant, the ‘market expansion’ effect outweighs the ‘market power’ effect. IPRs protection is likely to encourage imports of capital goods.

51Therefore, the simultaneous equations model is written as follows:

52

53The variables , In(y₀), In(HK), In(Mtech), (M+X/Y), In(IPR), In(ftrade), TRIPS and In(y₀/y_0USA) have the same definitions as in equations n°1 and n°2, (Tables1 and 3).

54The correlation between the dependent and independent variables in equations1 and 2 is presented in Table4. The correlation coefficients between explanatory variables of each equation are low. However, the three measures of human capital adopted in our study are highly correlated with each other.

55The import of capital goods, which is an explanatory variable in the first equation, is itself the dependent (endogenous) variable associated with the second equation in the full system.

56An important step before estimating the model is to focus on the endogeneity of this variable (MTech). The last row in Table5 displays the p value of Durbin-Wu-Hausman tests. The evidence points to the exogeneity of the MTech variable.

57Because the dependent variables are correlated and IPRs protection is a common predictor for both equations, there may be a ‘contemporaneous’ correlation among the errors across the two equations. Table5 shows that the pvalue of Breusch-Pagan tests of independence rejects the hypothesis that the correlation of the residuals across equations is zero.

58Since both are over-identified equations, the model will be over-identified. The Seemingly Unrelated Regressions (SUR) method is applied. Note that the software STATA 11.0 is used. The SUR method allows us to test multi-equation models, while taking into account the fact that the equations are not independent. If we estimate the model equation-by-equation using standard ordinary least squares (OLS), we will obtain consistent estimates, although generally not as efficiently as the SUR method. The latter allows us to obtain estimates for each equation, adjusted for the non-independence of the equations. Yet the individual equations cannot be considered separately because the independence assumption, as stated by economics, is not fulfilled. It is well known from the literature that economic growth factors are correlated.

59The empirical analysis uses a panel data set consisting of eight separate 5-year periods, 1970-74, 1975-79; 1980-1984, 1985-89; 1990-94, 1995-99; 2000-2004 and 2005-2009, for 48developing countries. For these countries of our sample, the available data contain some missing values over the whole period 1970-2009.

60We have estimated the model by iterative, seemingly unrelated regression (ITSUR). According to this method, the estimation is run iteratively until convergence is achieved. We have also estimated the model by OLS. As indicated above, the OLS estimates are obtained while ignoring any correlation between the error terms of different equations. We do not show them here. Our results showed that the SUR estimator performed better in investigating the impacts of IPRs protection than the OLS estimator.

61Table5 shows the results of estimation of the simultaneous equations model. As noted in Table1, three different measures of human capital are adopted: the average years of total schooling (In(atys)), the average years of secondary schooling (In(ayss)), and the average years of tertiary schooling (In(ayts)).

62The first three columns of Table5 present estimates for the base modelspecified above. In columns (4) to (6) of this table, we exclude the variable In(ftrade); ‘Freedom to Trade Internationally’; from the second equation of the system. The tests of independence of the residuals of the two equations obtained from estimating the model by ITSUR are also shown in Table5.