Innovation and productivity of SMEs in vietnam firm level panel data evidence

UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS INNOVATION AND PRODUCTIVITY OF VIETNAMESE SMALL AND MEDIUM ENTERPRISES FIRM LEVEL PANEL DATA EVIDENCE BY HO THI MAI ANH MASTER OF ARTS IN DEVELOPMENT ECONOMICS HO CHI MINH CITY, December 2013 UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS INNOVATION AND PRODUCTIVITY OF VIETNAMESE SMALL AND MEDIUM ENTERPRISES FIRM LEVEL PANEL DATA EVIDENCE A thesis submitted in partial fulfilment of the requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS By HO THI MAI ANH Academic Supervisor: Dr PHAM DINH LONG HO CHI MINH CITY, December 2013 DECLARATION This is to certify that this thesis entitled “Innovation and Productivity of Vietnamese Small and Medium Enterprises: firm level panel data evidence” which is submitted in fulfillment of the requirements for the degree of Master of Art in Development Economic to the Vietnam – The Netherlands Programme The thesis constitutes only my original work and due supervision and acknowledgement have been made in the text to all material used CERTIFICATION ii | P a g e ACKNOWLEDGEMENT Above all, I would like to thank and gratefully express the special appreciation to my supervisor - Dr Pham Dinh Long for all of his guidance, useful recommendations and valuable comments I could not be able to complete this thesis without his help and support I would like to acknowledge the great works from the Vietnam – The Netherlands Programme team, especially all of lecturers who have put their enthusiastic into the lectures for students Personally, I would like to thank Dr Truong Dang Thuy and Dr Pham Khanh Nam, who have greatly supported me in the courses and in thesis writing process Special thanks go to my family, friends and colleagues for their motivation and encouragement during my study in this program iii | P a g e ABBREVIATIONS SME Small and Medium Enterprise CDM Crepon Duguet Model CIEM Central Institute of Economic Management FE Fixed effect model GDP Gross Domestic Products GSO General Statistics Office of Vietnam HCMC Ho Chi Minh City ISIC International Standard Industrial Classification ILSSA Institute of Labor Science and Social Affairs MFP Multi Factor Productivity OECD Organization for Economic Co-operation and Development R&D Research & Development RE Random effect model SFA Stochastic frontier analysis TFP Total Factor Productivity WTO World Trade Organization iv | P a g e ABSTRACT Innovation is considered an essential factor for motivating the productivity of nations and firms Innovation and productivity are connected by multidimensional relationships and investigated in many countries However, there is very limited research in this field for Viet Nam This paper examines the relationship between innovation and productivity of Small and Medium Enterprises (SMEs) by using Viet Nam SMEs survey balanced panel data in 2007 and 2009 Cobb-Douglas production function and the fixed effect model are employed throughout the thesis The author has found that the presence of innovation has positive effects on a manufacturing firm’s productivity In addition, this study also looks at the impact of firm size, firm location and manufacturing sector on the relationship between innovation and SMEs’ productivity Key words: innovation, productivity, SMEs Viet Nam v|P a g e TABLE OF CONTENTS LIST OF TABLES VIII LIST OF FIGURES IX CHAPTER 1.1 1.2 1.3 1.4 1.5 INTRODUCTION PROBLEM STATEMENT RESEARCH OBJECTIVES RESEARCH QUESTIONS SCOPE OF THE STUDY STRUCTURE OF THE STUDY CHAPTER LITERATURE REVIEW 2.1 PRODUCTIVITY: CONCEPTS AND MEASUREMENTS 2.2 INNOVATION: CONCEPTS AND MEASUREMENTS 2.3 RELATIONSHIP OF INNOVATION AND PRODUCTIVITY 2.3.3 EMPIRICAL REVIEW OF INNOVATION AND PRODUCTIVITY RELATIONSHIP 10 2.3.4 DETERMINANTS OF THE INNOVATION IMPACT 13 2.3.5 INNOVATION AND FIRM PRODUCTIVITY IN VIET NAM 16 2.3.6 CHAPTER REMARK 16 CHAPTER 3.1 3.2 3.2.1 3.1.2 3.3 3.4 3.5 3.6 3.6.1 3.6.2 3.6.3 3.7 RESEARCH METHODOLOGY AND DATA 17 AN OVERVIEW OF SMES IN VIET NAM 17 CONCEPTUAL FRAMEWORK AND MODEL SPECIFICATION 23 CONCEPTUAL FRAMEWORK 23 MODEL SPECIFICATION 23 RESEARCH HYPOTHESES 26 DEFINITIONS OF VARIABLES AND CONCEPTS 26 DATA COLLECTION 29 METHODOLOGY 29 RANDOM EFFECT REGRESSION MODEL (RE) 30 FIXED EFFECT REGRESSION MODEL 30 SELECTION BETWEEN RE AND FE MODEL BY HAUSMAN TEST 30 MEASUREMENTS OF VARIABLES 31 CHAPTER DATA ANALYSIS 34 EMPIRICAL RESULTS 34 CHAPTER REMARK 41 4.1 vi | P a g e CHAPTER 5.1 5.2 CONCLUSIONS 42 CONCLUSION AND POLICY IMPLICATION 42 RESEARCH LIMITATION AND FUTURE STUDY 44 REFERENCES 46 APPENDIX A:DESCRIPTION OF DATASET 49 APPENDIX B: REGRESSION RESULTS 50 APPENDIX C: HAUSMAN TEST RESULTS 54 APPENDIX D: INDUSTRY CLASSIFICATION 57 APPENDIX E - EMPIRICAL STUDIES ON RELATIONSHIP BETWEEN INNOVATION AND PRODUCTIVITY 58 vii | P a g e LIST OF TABLES Table 1: Summary of main definitions of SME in selected economies 17 Table 2: Number of Enterprises by Sector 2006 – 2011 .19 Table 3: Labor Productivity by Firm size and Location .20 Table 4: Labor Productivity by Sector 20 Table 5: Innovation Rates in Manufacturing SMEs 21 Table 6: Diversification and Innovation Rates (%) .22 Table 7: Diversification and Innovation Rates, by Sector (%) 22 Table 8: Definitions & Measurement of Variables and Expected Sign of Coefficients .27 Table 9: Descriptive Statistic of Variables 34 Table 10: Regression results 36 Table 11: Hausman Test Results 37 Table 12: Regression results with regards to employee size .37 Table 13: Regression results with regards to employee size - groups 38 Table 14: Regression results with regards to firm location 38 Table 15: Regression results with regards to firm location - groups 39 Table 16: Regression results with regards to manufacturing sector 40 Table 17: Regression results with regards to manufacturing sector- groups 40 viii | P a g e APPENDIX A:DESCRIPTION OF DATASET Table A-1: Declaration of panel data xtset newid year panel variable: time variable: delta: newid (strongly balanced) year, to 9, but with gaps unit Source: Author’s calculation Table A-2: Summary statistics of Variables summarize OAperL KAperL MAperL Variable Obs Mean OAperL KAperL MAperL lnOAperL lnKAperL 4220 4220 4220 4220 4220 141696.4 223338.2 97703.32 11.35029 11.55843 lnMAperL lnL employee innovation 4186 4220 4222 4222 10.7425 1.870319 14.33775 4599716 lnOAperL lnKAperL Std Dev lnMAperL lnL employee innovation Min Max 252314.8 406992.8 212574.5 9480911 1.257623 2866.887 254.8344 7.960982 5.540614 6956979 1.06e+07 6472794 15.75526 16.17743 1.257466 1.141638 27.63651 4984542 3.123566 0 15.68312 6.214608 500 Source: Author’s calculation 49 | P a g e APPENDIX B: REGRESSION RESULTS Table B-1: Regression result Model (1) - RE xtreg lnOAperL lnKAperL lnMAperL lnL innovation, fe Fixed-effects (within) regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8234 between = 0.8164 overall = 0.8192 corr(u_i, Xb) F(4,2072) Prob > F = 0.2043 = = lnOAperL Coef lnKAperL lnMAperL lnL innovation _cons 0403773 6119739 -.0935365 0289898 4.480318 0074558 006756 0148831 0136353 1142666 sigma_u sigma_e rho 36445967 26717584 65044999 (fraction of variance due to u_i) F test that all u_i=0: Std Err t 5.42 90.58 -6.28 2.13 39.21 F(2109, 2072) = P>|t| 0.000 0.000 0.000 0.034 0.000 2415.82 0.0000 [95% Conf Interval] 0257556 5987246 -.1227239 0022495 4.256228 2.23 054999 6252231 -.0643491 0557301 4.704407 Prob > F = 0.0000 Source: Author’s calculation Table B-2: Regression result Model (1) - FE xtreg lnOAperL lnKAperL lnMAperL lnL innovation Random-effects GLS regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8131 between = 0.8829 overall = 0.8699 Random effects u_i ~ Gaussian corr(u_i, X) = (assumed) lnOAperL Coef lnKAperL lnMAperL lnL innovation _cons 0753398 6411617 071272 0186494 3.459456 sigma_u sigma_e rho 21000373 26717584 381883 Std Err .0047197 0046414 0054832 0107798 0597612 Wald chi2(4) Prob > chi2 z 15.96 138.14 13.00 1.73 57.89 P>|z| 0.000 0.000 0.000 0.084 0.000 = = 24908.72 0.0000 [95% Conf Interval] 0660894 6320647 0605251 -.0024786 3.342326 0845902 6502587 0820189 0397773 3.576586 (fraction of variance due to u_i) Source: Author’s calculation 50 | P a g e Table B-3: Regression result Model (2) - RE xtreg lnOAperL lnKAperL lnMAperL lnL innovation InLsize Random-effects GLS regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8128 between = 0.8832 overall = 0.8700 Random effects u_i ~ Gaussian corr(u_i, X) = (assumed) lnOAperL Coef lnKAperL lnMAperL lnL innovation InLsize _cons 075251 6411797 0754226 0025202 022743 3.453783 sigma_u sigma_e rho 20876594 26716442 37911621 Wald chi2(5) Prob > chi2 Std Err .004719 0046407 0064351 0172401 019028 0598597 z 15.95 138.16 11.72 0.15 1.20 57.70 P>|z| = = 24936.81 0.0000 [95% Conf Interval] 0.000 0.000 0.000 0.884 0.232 0.000 066002 6320841 06281 -.0312698 -.0145511 3.33646 0845001 6502753 0880351 0363102 0600372 3.571106 (fraction of variance due to u_i) Source: Author’s calculation Table B-4: Regression result Model (2) - FE xtreg lnOAperL lnKAperL lnMAperL lnL innovation InLsize, fe Fixed-effects (within) regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8235 between = 0.8167 overall = 0.8194 corr(u_i, Xb) F(5,2071) Prob > F = 0.2050 = = lnOAperL Coef lnKAperL lnMAperL lnL innovation InLsize _cons 0402848 6120385 -.0973427 0474205 -.0268129 4.486592 007456 006756 0152904 0217827 0247135 114408 sigma_u sigma_e rho 36428572 26716442 65025231 (fraction of variance due to u_i) F test that all u_i=0: Std Err t 5.40 90.59 -6.37 2.18 -1.08 39.22 F(2109, 2071) = P>|t| 0.000 0.000 0.000 0.030 0.278 0.000 2.22 1933.06 0.0000 [95% Conf Interval] 0256628 5987893 -.1273287 0047023 -.0752788 4.262225 0549068 6252877 -.0673566 0901387 0216529 4.710959 Prob > F = 0.0000 Source: Author’s calculation 51 | P a g e Table B-5: Regression result Model (3) - RE xtreg lnOAperL lnKAperL lnMAperL lnL innovation InLocation Random-effects GLS regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8123 between = 0.8841 overall = 0.8706 Random effects u_i ~ Gaussian corr(u_i, X) = (assumed) Wald chi2(5) Prob > chi2 lnOAperL Coef lnKAperL lnMAperL lnL innovation InLocation _cons 0737981 6406835 0688563 -.0016768 058822 3.486165 0047343 0046376 0055204 0123062 0172885 0602105 sigma_u sigma_e rho 20797586 26682879 37792371 (fraction of variance due to u_i) Std Err z 15.59 138.15 12.47 -0.14 3.40 57.90 P>|z| 0.000 0.000 0.000 0.892 0.001 0.000 = = 25017.87 0.0000 [95% Conf Interval] 064519 631594 0580364 -.0257965 0249372 3.368154 0830773 649773 0796762 0224429 0927068 3.604175 Source: Author’s calculation Table B-6: Regression result Model (3) - FE xtreg lnOAperL lnKAperL lnMAperL lnL innovation InLocation, fe Fixed-effects (within) regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8240 between = 0.8100 overall = 0.8143 corr(u_i, Xb) F(5,2071) Prob > F = 0.1993 lnOAperL Coef lnKAperL lnMAperL lnL innovation InLocation _cons 0400784 612038 -.0942923 0556297 -.0706182 4.484393 0074471 0067473 0148668 0172175 0279287 1141296 sigma_u sigma_e rho 37039794 26682879 65834811 (fraction of variance due to u_i) F test that all u_i=0: Std Err t 5.38 90.71 -6.34 3.23 -2.53 39.29 F(2109, 2071) = P>|t| = = 0.000 0.000 0.000 0.001 0.012 0.000 2.22 1938.96 0.0000 [95% Conf Interval] 0254739 5988058 -.1234477 0218644 -.1253894 4.260573 0546829 6252701 -.0651369 0893951 -.015847 4.708214 Prob > F = 0.0000 Source: Author’s calculation 52 | P a g e Table B-7: Regression result Model (4) - RE xtreg lnOAperL lnKAperL lnMAperL lnL innovation In_High Random-effects GLS regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8132 between = 0.8829 overall = 0.8699 Random effects u_i ~ Gaussian corr(u_i, X) = (assumed) Wald chi2(5) Prob > chi2 lnOAperL Coef lnKAperL lnMAperL lnL innovation In_High _cons 0753286 6411869 0712497 0268699 -.0100902 3.45932 0047202 0046422 0054841 0193808 0197762 0597695 sigma_u sigma_e rho 21008212 26718228 38204784 (fraction of variance due to u_i) Std Err z 15.96 138.12 12.99 1.39 -0.51 57.88 P>|z| = = 24903.14 0.0000 [95% Conf Interval] 0.000 0.000 0.000 0.166 0.610 0.000 0660772 6320884 060501 -.0111158 -.0488508 3.342174 08458 6502854 0819984 0648556 0286704 3.576466 Source: Author’s calculation Table B-8: Regression result Model (4) - FE xtreg lnOAperL lnKAperL lnMAperL lnL innovation In_High, fe Fixed-effects (within) regression Group variable: firmid Number of obs Number of groups = = 4186 2110 R-sq: Obs per group: = avg = max = 2.0 within = 0.8235 between = 0.8161 overall = 0.8190 corr(u_i, Xb) F(5,2071) Prob > F = 0.2048 = = lnOAperL Coef lnKAperL lnMAperL lnL innovation In_High _cons 0400982 6118667 -.0939256 0507251 -.0261341 4.485119 0074618 0067571 0148891 0266599 0275453 1143814 sigma_u sigma_e rho 36486488 26718228 65094416 (fraction of variance due to u_i) F test that all u_i=0: Std Err t 5.37 90.55 -6.31 1.90 -0.95 39.21 F(2109, 2071) = P>|t| 0.000 0.000 0.000 0.057 0.343 0.000 2.23 1932.74 0.0000 [95% Conf Interval] 0254648 5986153 -.1231249 -.001558 -.0801534 4.260804 0547317 6251181 -.0647264 1030082 0278853 4.709433 Prob > F = 0.0000 Source: Author’s calculation 53 | P a g e APPENDIX C: HAUSMAN TEST RESULTS Table C-1: Hausman Test result Model (1) hausman fixed random Coefficients (b) (B) fixed random lnKAperL lnMAperL lnL innovation 0403773 6119739 -.0935365 0289898 (b-B) Difference sqrt(diag(V_b-V_B)) S.E -.0349625 -.0291878 -.1648085 0103404 0057718 0049093 0138362 0083497 0753398 6411617 071272 0186494 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(4) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 154.75 Prob>chi2 = 0.0000 Source: Author’s calculation Table C-2: Hausman Test result Model (2) hausman fixed random Coefficients (b) (B) fixed random lnKAperL lnMAperL lnL innovation InLsize 0402848 6120385 -.0973427 0474205 -.0268129 075251 6411797 0754226 0025202 022743 (b-B) Difference sqrt(diag(V_b-V_B)) S.E -.0349663 -.0291412 -.1727652 0449004 -.0495559 0057726 0049099 0138703 013314 01577 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(5) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 170.40 Prob>chi2 = 0.0000 Source: Author’s calculation 54 | P a g e Table C-3: Hausman Test result Model (3) hausman fixed random Coefficients (b) (B) fixed random lnKAperL lnMAperL lnL innovation InLocation 0400784 612038 -.0942923 0556297 -.0706182 0737981 6406835 0688563 -.0016768 058822 (b-B) Difference -.0337197 -.0286455 -.1631486 0573065 -.1294402 sqrt(diag(V_b-V_B)) S.E .0057485 0049009 0138038 0120415 0219344 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(5) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 175.83 Prob>chi2 = 0.0000 Source: Author’s calculation Table C-4: Hausman Test result Model (4) hausman fixed random Coefficients (b) (B) fixed random lnKAperL lnMAperL lnL innovation In_High 0400982 6118667 -.0939256 0507251 -.0261341 0753286 6411869 0712497 0268699 -.0100902 (b-B) Difference sqrt(diag(V_b-V_B)) S.E -.0352303 -.0293202 -.1651754 0238552 -.0160439 0057791 0049101 0138423 0183067 0191741 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(5) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 155.48 Prob>chi2 = 0.0000 Source: Author’s calculation 55 | P a g e Table C-5: Het Test result regress lnOAperL lnKAperL Source SS lnMAperL lnL innovation df MS Model Residual 3233.44605 483.288806 4181 808.361512 115591678 Total 3716.73486 4185 888108687 lnOAperL Coef lnKAperL lnMAperL lnL innovation _cons 0764296 6475528 0815 012375 3.435645 Std Err .0045721 0045694 0049266 011145 0579639 t 16.72 141.71 16.54 1.11 59.27 Number of obs F( 4, 4181) Prob > F R-squared Adj R-squared Root MSE P>|t| 0.000 0.000 0.000 0.267 0.000 = 4186 = 6993.25 = 0.0000 = 0.8700 = 0.8698 = 33999 [95% Conf Interval] 0674659 6385943 0718411 -.0094752 3.322005 0853932 6565113 0911588 0342252 3.549285 Source: Author’s calculation hettest lnKAperL lnMAperL lnL innovation Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho: Constant variance Variables: lnKAperL lnMAperL lnL innovation chi2(4) Prob > chi2 = = 5955.95 0.0000 Source: Author’s calculation Table C-5: Correlation test result cor lnOAperL (obs=4186) lnKAperL lnMAperL lnL innovation InLsize InLocation In_High lnOAperL lnKAperL lnMAperL lnOAperL lnKAperL lnMAperL lnL innovation InLsize InLocation In_High 1.0000 0.4505 0.9216 0.2724 0.1584 0.0246 0.2111 0.1418 1.0000 0.3839 0.1661 0.1173 0.0426 0.1981 0.1046 1.0000 0.1791 0.1261 0.0321 0.1557 0.1174 lnL innova~n 1.0000 0.3140 -0.2122 0.3021 0.2611 1.0000 0.6597 0.4933 0.8317 InLsize InLoca~n 1.0000 0.1896 0.5433 1.0000 0.4768 In_High 1.0000 56 | P a g e APPENDIX D: INDUSTRY CLASSIFICATION OECD Classification - Manufacturing industries classified according to their global technological intensity - ISIC Revision (Note: in the study, there are only two subgroups: low-tech and high-tech sector The medium high-tech and medium low-tech are combined to form an aggregated high-tech sector) OECD classification High-technology Aerospace Computers, office machinery Electronics-communications Pharmaceuticals Scientific instruments Medium-high-technology Motor vehicles Electrical machinery Chemicals Other transport equipment 10.Non-electrical machinery Medium-low-technology 11 Rubber and plastic products 12 Shipbuilding 13 Other manufacturing 14 Non-ferrous metals 15 Non-metallic mineral products 16 Fabricated metal products 17 Petroleum refining 18 Ferrous metals Low-technology 19 Paper printing 20 Textile and clothing 21 Food, beverages, and tobacco 22 Wood and furniture ISIC Revision 3845 3825 3832 3522 385 3843 3832 3522 3842+3844+3849 3825 355+356 3841 39 372 36 381 351+354 371 34 32 31 33 Source: http://epp.eurostat.ec.europa.eu/cache/ITY_SDDS/Annexes/hrst_st_esms_an9.pdf 57 | P a g e APPENDIX E - EMPIRICAL STUDIES ON RELATIONSHIP BETWEEN INNOVATION AND PRODUCTIVITY # Authors & Title Data Variables & Concepts Model & Methodology Results Macro level Le Van (2008) "New Technology, Human Capital, Total Factor Productivity and Growth Process for Developing Countries!" Färe, Grosskopf et al (1994) “Productivity growth, technical progress, and efficiency change in industrialized countries." TFP, growth, technology Solow, Ramsey Model change, human capital Convex-Concave Production Function Developing countries Human Capital Growth Model and developed countries th in the 20 century Cobb-Douglas production function OLS regression Pooled time-series data of 71 non-oil exporting countries (Barro & Lee) Latin America, Asia Growth in developing countries based mainly on physical capital accumulation Growth in developed countries motivated essentially by human capital and technology change 17 OECD countries Technical change, Malmquist index of total factor U.S productivity growth is slightly over the period 1979- productivity growth productivity growth higher than average, all of which is 1988 due to technical change Japan's productivity growth is the highest in the sample, with almost half due to efficiency change Micro level Masso and Vahter (2008) Estonia "Technological innovation and productivity in latetransition Estonia: econometric evidence from innovation surveys." Community Innovation Survey, CIS (19982000) with 3161 firms & CIS (2002-2004) with 1747 firms in Technological innovation, CDM Model – developed by Crepon productivity (1998) This model explained the productivity of firms in correlation with innovation output, knowledge and investment in research development (Figure 1) Larger firms engage in innovative activities than small firms Firm size has an insignificant impact on product and a positive impact on the process innovation only process innovation has a positive significant effect on TFP 58 | P a g e effect of innovation on productivity – not only on the productivity in the last year of the innovation survey, but also one and two years after the survey Tobit Model Fritsch and Meschede (2001) German R&D, innovation, firm size 1800 enterprises "Innovation and productivity in developing countries: A study of Argentine manufacturing firms’ behavior (1992– 2001)." Miguel Benavente (2006) "The role of research and innovation in promoting productivity in Chile." Klepper (1996a, b) R&D expenditure rises less than proportional with size R&D expenditure = asizeb "Product innovation, process innovation, and size." Chudnovsky, López et al (2006) Cohen and approach small enterprises that perform R&D tend to be more innovative than  ln (R&D expenditure) = ln a + b ln large enterprises size small enterprises spend a much standard linear regression methods higher proportion of their R&D budget on new products than on new processes Argentinean manufacturing firms Innovative inputs Fixed effect estimators on innovation Labor productivity is, on average, expenditure intensity and productivity 14.1% higher in innovators than in Innovative outputs non-innovators Innovation surveys Multinomial logit on innovation Firm’s productivity (sales outcomes Large firms have a higher Panel data for 1992- per employee) probability of engaging in 2001 Four groups of innovation outcomes: innovation activities and becoming non-innovators, product innovators, innovators process innovators and combined innovators Chile – less developed country R&D investment CDM Model – ALS methodology Innovation intensity Tobit & Probit 1995-1998 Market share Productivity Labor skills have positive and significant effect on the estimation of productivity Neither research expenditure nor innovation has a significant impact on innovation sales and productivity 59 | P a g e in the short-run Productivity is measured as value added per worker Roper and Love (2002) innovation and export performance: evidence from uk and german manufacturing plants UK, German Innovation Plan-level data Export performance Product Survey Development RD Spill-over effects 1700 UK manufacturing plants and 1300 German plants Probit Model Innovation is positively related to export probability in both countries UK: the scale of plants’ innovation is related positively to export propensity Be effective in their ability to exploit spill-overs from the innovation Germany: negative relationship between the scale of innovation activity and export performance non-innovators are to absorb regional and supply-chain spill-over effects Innovative firms are generally found to discourage exporting Huergo and Jaumandreu (2004) Firms’ age, process innovation and productivity growth Productivity growth Process innovation Firm age firms enter the market experiencing high productivity growth and that above-average growth rates tend to last for many years, but also that productivity growth of surviving firms converges Process innovations at some point then lead to extra productivity growth, which also tends to persist somewhat attenuated for a number of years 60 | P a g e Dhawan (2001) Using a large panel of Firm size publicly traded US Productivity firms, parameters of the production technology for large and small firms are estimated for the 1970–1989 period Cobb Douglas production function Ghosal and Nair-Reichert (2009) North America and Europe Cobb Douglas production function Investments in modernization, innovation and gains in productivity: Evidence from firms in the global paper industry 19 firms in pulp and paper industry Firm size and productivity differential: theory and evidence from a panel of US firms Time series data: 19952004 Firm performance: profitability and productivity Control variables: mergers and acquisitions, R&D expenditure, capital intensity Fixed and Random effect model Performance = f (modernization investment, innovation activity, control variables) smaller firms exhibit a higher profit rate, lower survival probability and difficulty in accessing the capital market small firms are significantly more productive but also more risky than their large counterparts Small firms facing market uncertainties, capital constraints and other challenges undertake actions that make them more efficient than large firms but is achieved at the cost of increasing their riskiness Productivity are different across the firms R&D expenditures don’t significant contribution have OLS regression Sources: Compustat, Thompson Financial Cassiman, Golovko et al (2010) Innovation, exports and productivity panel of Spanish Innovation manufacturing firms Export Spanish Productivity ESEE survey Kolmogorov–Smirnov equality-ofdistributions test, We compare the productivity distributions of different samples of firms Innovating firms show higher productivity levels and grow faster than non-innovators SME firms running from 1990 until 1998 Lööf and Heshmati (2006) Sweden 1996-1998 product innovation –not process innovation – affects productivity and induces small non-exporting firms to enter the export market Dependent Variables: Generalized 61 | P a g e On the Relationship 3000 service and between Innovation and manufacturing firms Performance: A Sensitive 1300 in innovation Analysis sample R&D intensity Innovation sales worker Labor productivity Tobit per Semi-parametric Least Squares (SLS) Independent Variables: Size, human capital, innovation sales per worker Mohnen, Mairesse et al Seven (2006) Countries Innovation: A Comparison Across Seven European Countries European Innovative intensity R&D intensity CIS1 micro-aggregated Size data in 1992 Generalized Tobit 8000 firm and 5700 innovation sample Viet Nam Lang, Lin et al (2012) Viet Nam Innovation Mediate effect of technology innovation capabilities investment manufacturing firms Technological capabilities capability and firm performance in Viet Nam survey questionnaire was designed innovation collected data from enterprises in the high-tech manufacturing industry & SMEs in Viet Nam Competitive performance The samples were randomly selected form “The Role of Technology, the study identified that Viet Nam enterprise are required to invest to enhance seven dimensions of TICs learning, R&D, resource allocation, manufacturing, market, organizational, and strategic planning capabilities Investment capability can lead directly competitive performance through TICs Investment and Ownership Structure in Productivity Performance of the Manufacturing Sector in Viet Nam – BSPSCIEM project” Investment capability had positive effects on their TICs were positively correlated to their firm competitive 62 | P a g e performance Nguyen, Quang Pham et al (2008) “Innovation and exports in Viet Nam's SME sector." SMEs Viet Nam Survey Export, Innovation – 2005 Innovation is measured by new products, new production process or improvement of existing products (Le 2010) SMEs Viet Nam Survey Technical efficiency performance of Vietnamese manufacturing small and medium enterprises Cross-sectional data in 2002, 2005, 2007 of manufacturing SMEs Technical performance Logit / Probit model Export = β0 + β1X + θ2 Innovation + ε Three measures of innovation are significant determinant of exporting and instrumental variable (Z) Innovation = γZ + ε efficiency Stochastic frontier analysis (SFA) – Vietnamese non-state manufacturing using Cobb-Douglas production SMEs operated at high level of function and Translog production technical efficiency function Influencing factors of technical efficiency are: firm age, size, location, subcontracting, long-term cooperation, government assistance have negative impact on technical efficiency New product innovation has positive impact on technical efficiency in 2002, negative impact in 2005 and no significant impact in 2007 63 | P a g e ... technology innovation and productivity of a firm? (ii) What are the roles of firm size, firm location, and manufacturing sector on the impact of innovation and firm productivity? 1.4 SCOPE OF THE... RELATIONSHIP OF INNOVATION AND PRODUCTIVITY 2.3.3 EMPIRICAL REVIEW OF INNOVATION AND PRODUCTIVITY RELATIONSHIP 10 2.3.4 DETERMINANTS OF THE INNOVATION IMPACT 13 2.3.5 INNOVATION AND. .. process innovation, innovation activity and expenditure, impacts of innovation, other sources and findings of innovation of European enterprises Most of the firms in Europe and other countries