Intellectual Property Protection Strategies of Foreign Companies in Korea

This paper aims at analyzing the intellectual property (IP) protection strategies of foreign companies in Korea to suggest policy implications. To this end, the Korean Innovation Survey (KIS) data conducted by Science and Technology Policy Institute (STEPI) is utilized to analyze various IP protection methods − secrecy, lead time and complex design, as well as patent. Specifically, the following three questions were in focus: What are the methods used by foreign companies to protect their IP? How effective are the methods? Can they make more diversified IP protection strategies, compared with domestic companies? Major empirical findings are as follows. Compared with domestic companies, foreign companies are more likely to use secrecy and lead time (rather than patent) and evaluate their effectiveness higher. Also, they tend to utilize more profiles of IP protection strategies than domestic companies. However, the results are true only in the service sector, where competitiveness is relatively low, but not in the manufacturing sector. This paper emphasizes the importance of spillovers from FDI or foreign R&D to Korean companies, especially in the service sector, and suggests policy implications for maximizing spillovers - the facilitation of academy-industry knowledge transfer channels, the improvement of IP protection environments and so on.


I. Introduction
Multinational companies (MNCs) try to enhance their competitiveness, obtaining core technology/capacity through overseas investment or R&D globalization. Especially, since the 1990s, they have been actively utilizing local experts and R&D networks to accelerate business abroad, shifting from simply building production bases or adapting their technology overseas in the 1960s (Gassmann & Zedtwitz 1999). These activities have played a major channel of technology transfer to hosting countries (Mohnen 2001;Veugelers and Cassiman 2004). Namely, technology or know-how generated from foreign companies has been transferred to local companies as a form of spillovers through contract or manpower mobility. Literature also emphasizes benefits or spillovers of the FDI in terms of increases in productivity (Lichtenberg and van Pottelsberghe de la Potterie 1998) and of knowledge transfers (Branstetter 2000).
However, there exist few studies that deal with this issue in terms of the 'appropriability strategy' of a company. Foreign companies also use this for protecting their innovation outcome or intellectual property (IP), which may function as a negative factor of "preventing knowledge spillovers" to hosting countries (Faria and Sofka 2008). The appropriability of economic returns from R&D is important for innovating firms, and foreign companies are not exceptions in this respect. According to Nadiri (1993), 50% of companies' innovations cannot be appropriated by themselves, leaking out as knowledge spillovers, so it is natural for them to have strong incentives to protect their own. Veugelers and Cassiman (2004) also underline that foreign companies are less active in technology transfers than domestic companies.
Korea is one of the countries that host large amounts of FDI. According to the statistics of the Ministry of Knowledge and Economy (MKE 2011), the amount recorded the highest, less than $16,000 million, in 2009, although it decreased to around $12,000 million just recently. A noticeable change is the % of FDI in the service sector, which recorded up to 60~70% of the total in 2008 from 35.2% in 2004. According to the Kim and Kim's study (2010), during 6 對外經濟硏究 제15권 제3호 2011년 가을호 1990~2008, the backward linkage effect of FDI has increased in both manufacturing and service sectors, and the latter increased more than the former.
However, there exists no study that deals with the FDI inflow to Korea in terms of the appropriability strategy of foreign companies. Since 1997, the Korean government has been just in a hurry to make various policies and environments favorable to FDI. Now, it is time to examine the appropriability or IP protection strategies of foreign companies, including the service sector whose economic importance is noticeably increasing recently, and to prepare measures for maximizing the spillovers.
To this context, this paper aims at analyzing the IP protection strategies of foreign companies to suggest policy implications. Specifically, this tries to answer the following three questions: which methods foreign companies tend to use for protecting their IP; how effective is each method; and whether or not they have more profiles of protection strategies, compared with domestic companies. Answers will include differences in their strategies between the manufacturing and service sectors. This study has several academic contributions.
First, it is the first empirical study in the IP protection strategies of foreign companies in Korea -a case of latecomer country. Second, it deals with several IP protection methods, secrecy, complex design and lead time beyond patent, and analyzes their strategies from various aspects. Third, it tries to compare their strategies in both manufacturing and service sectors. This paper is structured as follows. The next section (II) reviews the existing literature in corporate appropriability or IP protection strategies. The following section (III) explains the data and the methodology to be used in empirical analyses. Then the subsequent section (IV) presents empirical results and finally the last section (V) makes summary and conclusion.

II. Theoretical Approach
Intellectual property includes knowledge, skill and other intangible assets that can be converted into usable resources to generate competitive advantages (Harabi 1995). The Basic Law on Intellectual Property, which took an effect in Korea on July 20, 2011, defines it as 'knowledge, information and skill, ideology or emotional expression, mark of business or object, variety or genetic resource and other intangible things through which asset value could be realized(Ministry of Justice 2011).' These imply that it is crucial companies to invest their assets in innovations to strategically appropriate economic returns.
IP protection strategies can be divided by two − formal and informal ones (Harabi 1995). Formal protection strategies, e.g. patent, copyright, trademark and industrial design, are on the legal base, granting a right holder with exclusive rights through application. Formal protection is strong, but its effectiveness is doubted sometimes due to the 'inventing around' by competitors or to the high costs of application. On the other hand, informal protection strategies, e.g. secrecy, complex design and lead time, are on the organizational base. Secrecy or complex design may block new knowledge from being utilized by competitors, which continues until additional expertise is combined, and lead time also mitigates the leakage of knowledge spillovers, allowing original companies first-mover advantages. Informal protection is relatively weak to formal ones, but effective in terms of the protection of tacit knowledge, the low cost of protection and no risk of inventing around like patent (Faria and Sofka 2010).
Previous studies have focused on patents, but current studies try to include informal IP protection strategies. Levin et al. (1987) conducted the Yale Survey on US manufacturing companies to analyze the effectiveness of various appropriability strategies for the first time. According to them, patent is the least ineffective means against imitation; rather, innovative companies prefer lead time, secrecy or complex designs. However, they underline the fact that patent 8 對外經濟硏究 제15권 제3호 2011년 가을호 is prominently used in pharmaceuticals and chemistry where it has traditionally played an important role. Regarding the limitation in patent, Harabi (1995) mentions the disclosure of too much information and the competitor's capacity of inventing around. Using another survey on this issue − the Carnegie Melon Survey on US manufacturing companies, Cohen et al. (2000) argue that patent is just one of the important protection mechanisms, in spite of its increasing importance among large companies. According to them, patent is significant only in terms of preventing litigation or strengthening a bargaining power over the third party. Through the empirical analysis in Dutch companies, Brouwer and Kleinknecht (1999) figure out small companies are less likely to file or apply patent than large companies, due to high costs of application and maintenance.
They also prove that there exist strong sector differences in the company's propensity of patent, explaining the reason in terms of the speed and cost of imitation by competitors. Arundel (2001), using the Community Innovation Survey (CIS) data of 1993, analyzes the relative importance of secrecy over patent to find that companies evaluate secrecy higher than patent for appropriation but that their R&D has no significant impact on the relative value of the two.
Unlike manufacturing sector, only a few studies exist regarding the IP protection strategies in the service sector. As Kuusisto and Paallysaho (2008) point out, service innovation is different from traditional technological innovation, so it needs a different approach. 1) According to them, manufacturing innovation is based on tangible products, while service innovation, on intangible elements and human capitals or interactions. In other word, companies are dependent on employees' knowledge or skill rather than on physical capital, the latter of which has no significant role at least in the business service.
Accordingly, in the service sector, patent or IPRs account for a small portion in corporate IP protection strategies, and companies often introduce various informal protection methods. 2) In their comparative studies in Finland and UK, 1) They argue that IPRs have emerged basically for "catering the needs of physical products." Intellectual Property Protection Strategies of Foreign Companies in Korea 9 Kuusisto and Paallysaho argue that service or knowledge are so diversified in the service sector that they cannot be protected by IPRs effectively and that they need various informal protection strategies. 3) Amara et al. (2008), based on the 2003 Statistics Canada Innovation Survey on Service, estimate eight IP protection methods using the multivariate probit model to find the complementary, substitute and independent relationships among them. Most of all, they underline that informal protection methods support or complement formal ones.
As of now, the literature regarding the IP protection strategies of foreign companies − survey or comprehensive analysis − is rare. There is an analysis on European countries conducted by Faria and Sofka (2008), which is based on the CIS data. Using the CIS III data (1998~2000) of 1,800 manufacturing companies in Portugal and Germany, they examine the location choice of MNCs and the way that they protect knowledge. Here, they make indices for the 'breadth' and 'depth,' of IP protection methods, following Laursen and Salter (2006). According to them, foreign affiliates use more diverse or clearer appropriability strategies beyond patent than domestic affiliates, combining both formal and informal strategies. They also prove that their appropriability strategies are 'less restrictive' in the country with 'more opportunities for knowledge sourcing,' 4) (here, it is Germany) explaining the reason such that they can exchange knowledge with domestic companies by reciprocity.
For foreign investing companies, as being small-sized or affiliates, in general, patent may be an onerous strategy in terms of the high cost or the disclosure of information from mother companies. In this respect, they are more likely to adopt informal protection methods of evaluation higher than patent. Moreover, 2) However, they still emphasize that both formal and informal protection methods are not mutually exclusive, but they support each other in different ways. 3) They explain sector differences in the IP protection strategies for three industries: software sector is most active in using IPRs; advertising sector is not familiar to IP issues, but loyalty building among personnel is important here; and business/ management consultancy takes secrecy and fast-moving innovation cycle rather than IPRs. 4) As to the 'breath' of IP protection strategies, the coefficient is significant only in Portugal but not in Germany.
they may apply more diverse protection methods, informal as well as formal, than domestic companies expecting complementary or synergy effects. Moreover, knowledge and innovation outcome in the service sector are rather tacit and tangible, so the above phenomena will be significant compared with the manufacturing sector. In the subsequent sections, this paper examines how foreign companies make the IP protection strategies in Korea, through the statistical and the empirical analyses.

Data
This Foreign companies are defined as those whose ratio of direct investment accounts for over 10% of the total assets, as specified in the Surveys: 130 firms (9.3%) are in the manufacturing sector and 40 firms (4.6%), in the service sector.
IP protection strategies are based on the companies' answers to the question,  Second, for estimating the effectiveness of each IP protection method, the ordered probit mode is used. The dependent variable is measured as 0 (no use), 1 (less effective), 2 (moderate), 3 (effective), 4 (very effective) and 5 (highly effective), representing ordered categories − how effective each IP protection method was. To the notion of these multiple thresholds, this study applies ordered probit mode to estimate the effectiveness.
Third, for estimating the breadth of IP protection strategies, the negative binominal model is used. The dependent variable is measured by summing up the number of protection methods that were used by companies, ranging 0 (no use) to 4 (using all four methods − patent, secrecy, complex design and lead time). As this variable is a count data including 0 and positive numbers, the OLS estimates may be inefficient, inconsistent and biased. The poisson model may be appropriate as an estimation method in this case, but the negative binominal model is used instead in order to solve the over-dispersion problem of the former (Hilbe 2007).
Models for the empirical analysis are specified by firm size, R&D intensity, R&D cooperation, outwardness, foreign ownership, and industry dummies:

      
Firm size (Lsize) is measured as a log value of employees. Large companies, which are competitive or rich in resources for patent-filing, may file or apply patents more than medium-small enterprises (Brouwer and Kleinknecht 1999).
However, empirical result is rather obscure. Arundel (2001)  Foreign ownership (D_FOR) 6) is measured as 1 if a company is foreign, whose ratio of direct investment accounts for over 10% in total assets (as defined in III.1), and 0 otherwise. Affiliates tend to rely on their parents and to use secrecy rather than patent (Amara et al. 2008). This may be noticeable in the case of foreign investing companies, using secrecy, complex design or lead time, to protect their innovations. R&D cooperation (D_COOP) is measured as 1 if a company makes R&D cooperation with external partners and 0 otherwise.
Literature points out the fact that patent functions as clarifying the ownership among cooperators and that R&D cooperation may increase the likelihood of a company's patent-filing (Brouwer and Kleinknecht 1999;Arundel 2001).
Outwardness (Outwardness) is measured as a ratio of exports in sales. Exporting or outward companies may file more patents to protect their IP overseas, along with various IP protection strategies. Industry (D_IN) is measured as 1 if a company belongs to a specific industry among nine industries in manufacturing sector and five industries in service sector, and 0 otherwise. 7) Brouwer and Kleinknecht (1999) underline that patent is an effective method for IP protection in sectors of higher technological opportunity or those of which the products are easily developed but price for imitation is cheap. Arundel (2001) argues that companies are likely to depend on secrecy rather than on patent when disclosure 6) Regarding control variables, explanations are made mainly in terms of patent rather than informal protection methods − secrecy, complex design and lead time, due to the lack of literature regarding the latter.

Intellectual Property Protection Strategies of Foreign Companies in Korea 15
is disadvantageous to patenting. On the other hand, there also exist some empirical studies that find the opposite or the insignificant results regarding the sector characteristics (Amara et al. 2008;Park 2006). 7) IV. Empirical Results

Propensity and Effectiveness 8)
[ 8) It should be notice that 'effectiveness' is not the objective index but the evaluation that was made by companies on how effective each as an IP protection method. method (model 2), estimated using the ordered probit model, the coefficient of D_FOR is not statistically significant. This also implies that no difference exists even in the evaluations on the effectiveness between the two groups.
The results in the service sector are a bit different, however (Table 5) In sum, no significant difference exists in IP protection strategies between foreign and domestic companies in the manufacturing sector, but it does exist in the service sector: In the latter, foreign companies use secrecy and lead time more and evaluate secrecy more effective. The result that informal methods are preferred to patent as IP protection strategies is consistent with the Arundel's study (2001) and with other survey results (Levin et al. 1987;Cohen et al. 2000), but it is applicable only to the service sector in Korea. This results from the innate characteristics of knowledge in service sector, being more tacit and intangible, such that their IP can be protected through informal methods effectively. This explains the low competitiveness of this sector. Foreign investing companies are likely to make stronger and informal IP protection strategies in hosting countries for fear of technology leakages, which is significant when the competitiveness to which they belong is lower. Namely, this result reveals that foreign companies protect their innovations through informal methods in Korean service sector where its competitiveness is low. Note: 1) D_IN1 (food), D_IN2 (textiles and garments), D_IN3 (wood and wood products), D_IN4 (chemical and chemical products), D_IN5 (rubber and non-metals), D_IN6 (machinery), D_IN7 (electronics), _IN8 (transportation).
2) * represents the Wald test's p-value.  2) * represents the Wald test's p-value. Regarding the other variables, the coefficient of Lsize is shown to be positive and statistically significant. This is consistent with the Amara et al. 's result (2008) such that large companies tend to actively make IP protection strategies.
The coefficient of RD is statistically significant only for patent in the manufacturing sector. This implies that the level of codification of knowledge is high in this sector, so R&D-intensive companies protect their innovations through patent. However, this is not applicable for the service sector where the codification is not normal or less important. The coefficient of D_COOP is statistically significant as seen in Arundel (2001), which underscores that more cooperative companies file or apply patents, clarifying the ownership with their partners. They protect their innovations through informal methods, as well.
Unlike expectation, however, Outwardness is shown to be not significant,

Breath
Next, [ Table 6] presents empirical results on the breadth of IP protection 9) For example, in the case of pharmaceuticals, it takes much time a patent to be created and filed, with a large amount of investments, but this cross-section data cannot cover the whole period. Instead, it reflects the state that pharmaceutical companies keep the relevant information to secret to a certain period, due to high development costs.  companies are more likely to use and evaluate informal IP protection methods higher than patents, and this is true only in the service sector. Second, foreign companies have more profiles of IP protection strategies than domestic companies in service sector. Based on the findings above, foreign companies make their IP protection strategies more actively in service sector rather than in manufacturing sector.
This paper suggests some policy implications. First, considering the fact that foreign companies prefer informal IP protection methods to patent, contact or R&D cooperation needs to be utilized as a channel to their knowledge. For this, incentives or opportunities should be encouraged to lead them participate in this cooperation, particularly, with academic institutions -universities or public research institutes (PRIs). PRIs can be safer partners for foreign companies, rather than competitors, in terms of technology leakages. Moreover, R&D cooperation with PRIs is rather basic, so research outcomes from it can take the form of paper or patent. Through cooperation between PRIs and foreign companies, tacit knowledge of the latter could be patented or partly shared.
Moreover, manpower mobility during this cooperation can be an efficient way of learning tacit knowledge, particularly, in the service sector where foreign companies keep their IP to secret. In these respects, industry-university (IU) cooperation should be encouraged, and a policy concern needs to be made on solving the ownership problem that is inherent between them: for example, the stylized guidelines to IU cooperation could be helpful. Second, IP-friendly environments should be established, such that foreign companies can actively register to commercialize their patents or other forms of IPRs. The rapid increase in patent-filings in the U.S., since the 1980, has benefited from the government's pro-IP policy. As of 2010, the ranking of Korea in IPR protection is 32nd, positioning at the middle-low level, according to the IMD (2011). Foreign investing companies in Korea will get to depend on IPR systems more, increasing filing patents or other forms of IPRs, if its protection environments Intellectual Property Protection Strategies of Foreign Companies in Korea 23 become improved. Here, the Basic Law on Intellectual Property, which took an effect on July 20, 2011, will be helpful: this Law declares the establishment of IP respecting culture as one of the five major policy objectives. Third, in order to enhance the competitiveness in service sector, more opportunities should be made for knowledge exchanges or network-building between residents and foreigners, e.g. international conferences, seminars or consulting. For the Korean service sector, where competitiveness needs to be upgraded, the transfer or spillover of advanced knowledge from foreign companies is crucial. This is particularly the case for knowledge-based industries where knowledge is generated in combination of new or tacit knowledge. Therefore, it is important to improve the competitiveness, by enlarging opportunities or network-building for knowledge exchanges between residents and foreign companies in the short-run.
The contribution of this paper is that it deals with various IP protection methods to analyze the strategies of foreign companies in Korea. As a latecomer economy, Korean service industry is not yet developed, where knowledge spillovers from foreign companies are limited due to their strict IP protection strategies. This is the first try to examine the IP protection strategies of foreign companies in a latecomer economy, expanding the empirical analysis in manufacturing to service sector. One limitation of this study, however, is that it covers a particular period of 2002~04 for manufacturing sector and 2003~05 for service sector. Thus, it would be interesting to try a pooled or panel data set to verify if there would be any maturation.