The Role of R D on Firms Export Performance

The Role of R&D on Firms’ Export Performance A case of Thailand

Definitions n Thai firms - Firms where 100% of its capital is held by Thai people. n Non-Thai firms - Firms report any share of foreign capital. n Exporting firms (EF) - Firms report the share of exports of total sales is greater than zero (not sold 100% on the domestic market). n n Large-share exporting firms (LSE)- % export in total sale is greater than or equal to 50% Small-share exporting firms (SSE)- % export in total sale is larger than 0% but less than 50% n Non-exporting firms (NE) - Firms that not export at all (sold 100% on the domestic market) 2

Statement of the problem n Shifting in world demand from labor-intensive to technology-intensive products (Lall, 2000) pressures manufactured firms in Thailand improve technological capabilities. n One strategy implemented to develop technological capabilities is performing research and development (R&D). n R&D expenditures in the private sector in Thailand have been increasing. n Is R&D improve export performance of firms in Thailand? 3

Statement of the problem R&D Expenditures Classified by Sector 1996 1997 1999 million baht 2001 Total R&D expenditures 5, 528 4, 811 5, 021 11, 064 Government sector 2, 708 2, 666 1, 833 5, 019 High education sector 1, 219 1, 631 2, 570 1, 950 Business enterprise sector 1, 243 466 587 4, 009 Private non-profit sector 357 46 30 85 Source: Office of the National Research Council of Thailand 4

Conceptual Discussion 1) R&D and export performance n However, the effects of R&D on export performance of firms in developing countries remain ambiguous. n One the one hand, some previous studies (Posner, 1961; Hirsch, 1965; Vernon, 1966) proposed that R&D determines export performance of developed countries only. n One the other hand, there are studies argued that imported technology is often costly, risky and difficult to codified (Kumar and Siddharthan, 1994; Lall, 2000; Rodriquez and Rodriquez, 2005). n There is still no conclusion. 5

Conceptual Discussion 2)The nature of R&D in developing countries n The role and organisation of R&D in Technology. Followers (Forbes and Wield, 2000) R&D as a complement to shop-floor innovations. n R&D as the formal learning unit of the firms n R&D as a measure to build an independent product design capability n R&D as a source of intangible spin-off benefits for the firms n 6

Conceptual Discussion 2)The nature of R&D in developing countries (cont. ) n Industrial Structure and the role of R&D (Kumar and Siddharthan, 1994; Dijk, 2002) In developing countries, R&D determines the export performance of firms in low-technology and medium-technology industries. n However, R&D does not affect export performance of firms in the high-technology industry. n 7

Conceptual Discussion 3. The Multiplant Economy of Scale and R&D Behavior of MNEs (Markusen, 1984) n Economies of multiplant operation. Explains MNEs behavior in host countries. n MNEs may be composed of many plants. One plant (or more) carries out R&D and use its output as “joint inputs” of all plant productions. n Thus, the R&D behavior of MNEs are different from the R&D behavior of local firms and should be separately analyzed. n 8

Remarks for the study of Thailand n The role of R&D in technology follower firms may differ from technology leader firms (Forbes and Wield, 2000). n MNE firms may behave differently in R&D activity from Thai locally owned firms. (Markusen, 1984) n Effects of R&D on export performance vary across industries. (Kumar and Siddharthan, 1994) 9

Objective of the study 1) To estimate the effect of R&D intensity on export performance of firms in Thailand. 2) To investigate the differences in R&D behavior between exporting firms and nonexporting firms. 10

Method of the study n To answer the 1 st objective - using econometrics model to estimate effects of R&D intensity on firm’s export performance (estimated whole sample). - examine the effects of R&D on export performance of Thai firms and non-Thai firms. - examine the effects of R&D on export performance across groups of industries. n To answer the 2 nd objective - using ANOVA to investigate R&D behaviors among R&D firms. 11

Objectives Organization of the study Instruments Examine overall effect of R&D intensity on export Estimating whole sample Estimating Thai-firms and non-Thai firms separately Check R&D behaviors of MNEs in Thailand different indifferent Check sector variations Test censored tobit model against Cragg’s two stages specification and using LR-test to find appropriated model Estimating whole sample classified by industries Estimating Thai-firms classified by industries Using method of estimation suggested by LR-test Interpretations the econometric results Investigating R&D-related factors that differentiate firm’s export Comparative studies on R&D related factors and export (compare R&D firms in selected sample) Using ANOVA Interpretations the ANOVA results 12

st objective The 1 Estimate the effect of R&D intensity on export performance of manufacturing firms in Thailand 13

Theoretical Model From is the optimal level of the firm’s total output. is the optimal level of output sold in the domestic market is the output sold in the foreign market or the level of the firm’s export product. is the cost variable that is common to production for both markets is the specific cost of each market (k= d, f). is a vector of the specific factors of firm j, such as the productivity, 14 size, and ownership.

Theoretical Model Focusing on optimal level of firm’s export n n n From Crepon, Duguet, and Mairesse (1998), R&D can improve productivity, then reduce the exporting cost, and increase export level. R&D can also ease firms to develop new products. Applying equation (4. 13) in practice, firm’s export performance is influenced by the exporting cost and firm’s characteristics 15

Theoretical Model n n Alternatively, we use knowledge from previous literatures as proxy variables. R&D intensity- reduces costs of exported goods and extend product ranges. Skilled labors – can be either complementary factors for R&D activities or raising production costs. Firm’s size – Large firms tend to have cost advantages. Empirical evidences also found inverted U-shaped relationship between firm’s size and export Ownership – Foreign and Joint venture firms have superior knowledge, export experience, and can easier access advanced technology. Hence, costs of export market penetration are lower than local firms. 16

The Model 17

Data descriptions 18

Whole sample estimated results 19

Estimated results by ownership 20

Notifications from estimated results n Effects of R&D intensity on export performance of non-Thai firms and Thai firms are different n Export performance of non-Thai firms is not determined by R&D intensity n The results indicate that R&D behaviors of non. Thai firms can not well explained by this model. (see Markusen, 1984) n Hence, it is reasonable to consider effects of R&D intensity on export performance of Thai firms rather than whole firms. 21

Estimated results of Thai firms n LR-test shows that tobit model is more appropriate n Export performance of Thai firms is determined by R&D intensity, firm’s size, and size 2 n The expected signs of all variables are correct with the assumption. n R&D intensity positively correlates with export performance. If R&D intensity increases (decreases) one million baht, the conditional expected of export intensity increases (decreases) 1. 8746 unit. n Hence, R&D intensity can improve export performance of Thai firms 22

Estimated results of Thai firms n A marginal effect of skilled labors has positive sign and is equal to 0. 0159. If share of skilled labors in total employees increases (decreases) one unit, the conditional expected of export intensity increases (decreases) 0. 0159 unit. n However, this variable is insignificant. Skilled labors do not affect export performance of manufacturing firms in Thailand. 23

Estimated results of Thai firms n As for firm’s size. If numbers of employees increase (decrease) by one hundred persons, the conditional expected of export intensity increases (decreases) by 0. 0039. n Size 2 has negative sign n Firm’s size and export performance reveal the inverted U-shaped relationship 24

The effect of R&D intensity with sector variation Sector characteristics Supplier dominated Scale intensive Specialized suppliers Science based Industry Technology accumulation Agriculture, construction, Import technology / Services, Traditional Little R&D manufacturing Source of technology Specialized suppliers Petroleum and coal, iron Design, Operation , and steel, motor learning by doing Internal vehicles, transportation (Process innovation) equipments Machinery building, instruments Design & development Clients Chemical products, Electronics products, Pharmaceutical products Academic research, Search for new technology In-house R&D, external collaboration 25

Data descriptions 26

Estimated results by industries 27

Estimated results classified by industrial structure n R&D intensity determines export performance of firms in supplier dominated industries and scale intensive industries only. n Our results are similarly with the results of Kumar and Siddharthan (1994) for India and Dijk (2002) for Indonesia. 28

nd objective The 2 To investigate the differences in R&D behavior between exporting firms and nonexporting firms 29

Analytical framework Based on Forbes and Wield (2000) and other previous literatures, R&D behaviors can be represented by Criteria Measurement A presence of in-house R&D % distribution of in-house R&D activities (in all R&D activities) Numbers of scientists and The number of engineers and scientists in engineers total employees The numbers of own design product Types of R&D Objectives of R&D Level of external collaboration in R&D 1)%Products designed by firms according to customer requirements in total sales 2)%Products designed by firms and sold under own brand % distribution of R&D expenditures on Basic, applied, experimental R&D outputs (new or improved product/ process) 1 -5 Likert scale (evaluated the intensity of 30 collaboration)

Analytical Results All Firms with R&D NE SSE LSE %Export =0 (0<%Export <50%) Export≥ 50% The in-house R&D 0. 90 (0. 26) 0. 93 (0. 21) 0. 88 (0. 28) 0. 90 (0. 26) Numbers of scientists and engineers 0. 07 (0. 10) 0. 06 (0. 09) 0. 10 (0. 12) 0. 06 (0. 07) Share of products designed by firms according to customer requirements in total sale* 0. 15 (0. 29) 0. 10 (0. 21) 0. 22 (0. 35) Share of products designed and sell under own brand in total sale* 0. 35 (0. 43) 0. 26 (0. 41) 0. 51 (0. 45) 0. 25 (0. 39) 164 51 63 N 50 31

Analytical Results All Firms NE SSE LSE %Export =0 (0<%Export< 50%) Export≥ 50% Basic Research 0. 64 (0. 31) 0. 59 (0. 31) 0. 63 (0. 31) 0. 70 (0. 32) Applied Research 0. 28 (0. 30) 0. 31 (0. 29) 0. 29 (0. 31) 0. 25 (0. 29) Experimental Development 0. 08 (0. 11) 0. 10 (0. 12) 0. 08 (0. 12) 0. 05 (0. 09) 164 51 63 50 N 32

Analytical Results NE SSE LSE All Firms %Export =0 (0<%Export <50%) Export≥ 50% Improve production process 0. 39 (0. 49) 0. 50 (0. 51) 0. 29 (0. 46) 0. 40 (0. 49) Develop new production process 0. 32 (0. 47) 0. 29 (0. 46) 0. 33 (0. 48) 0. 34 (0. 48) Improve quality of existing products 0. 62 (0. 49) 0. 75 (0. 44) 0. 56 (0. 50) 0. 58 (0. 50) Develop new products 0. 67 (0. 61) 0. 61 (0. 49) 0. 76 (0. 76) 0. 62 (0. 49) Granted patents 0. 09 (0. 28) 0. 10 (0. 30) 0. 05 (0. 21) 0. 12 (0. 33) 164 51 63 N 50 33

External collaboration in R&D 34

Analytical Results n Average % of in-house R&D of NE, SSE, LSE are n n not statistically different. On average, SSE have more numbers of scientists and engineer than other sub-samples, but the distinction is not statistically different. Design capabilities of LSE, SSE, and NFE are statistically different. LSE typically design and produce products according to customer requirements. SSE design and produce products, and then sell under their own brands. 35

Analytical Results n The results roughly present that exporting firms (both SSE and LSE) have design capabilities more than non-exporting firms. n On average, non-exporting firms mostly perform experimental R&D, and also successfully improve efficiency of production process as well as improve quality of existing products. n One possible explanation is that non-exporters carry out experimental R&D and conduct R&D with the purpose of succeed in the domestic markets, not for export markets. 36

Analytical Results n On average, exporting firms (SSE and LSE) cooperate with the external agents more than non -exporting firms. n Non-exporting firms intensely collaborate with public agents n However, the degrees of external collaborations in R&D do not statistically different among three sub-samples. 37

Summary of the Finding n R&D intensity positively determines export performance of Thai firms, but not influence export performance of non-Thai firms. n The results support the explanation on R&D behaviors of MNEs by Markusen (1984). n Regarding Thai firms, R&D intensity affect export performance of firms in supplier dominated industries and scale intensive industries in positive way. n For specialize supplier and science based industries, R&D intensity has no effects on export performance. 38

Summary of the Finding So far we can finish both objectives, n 1 st objective - R&D intensity positively influences an export performance of manufactured firms in Thailand, but its impacts vary by industrial structures and ownership. n 2 nd objective - Thai R&D firms that export report higher designed capabilities than other firms. Hence, apart from R&D intensity, it can be roughly said that designed capabilities are important. 39

Limitations and Suggestions for further studies Restrictions of the data. n It is difficult to measure R&D output. This study use R&D input instead. n The data is cross-section data, which do not allow us to examine the effects of R&D more than one reference year. According to the fact that R&D require some period of time to yield benefits, studying on one year may too short. n In addition, many previous studies propose that, alternatively, export also causes R&D. Using crosssection data restrict to effectively analyze in dual ways. 40

Limitations and Suggestions for further studies Limitations of quantitative analysis. n It does not give much information on firm’s R&D behaviors and export decision. In-depth study should be continued henceforth. n Furthermore, non-formal R&D also important for developing countries, further studies should draw attention on non-formal R&D activities as well. 41

End of presentation 42