Firm-Level Determinants of Export Performance: Evidence from the Philippines 45 th Annual Meeting Philippine Economic Society 14 November 2007 Ma. Teresa S. Dueñas-Caparas
Research Background Export activity as a driver of economic growth Research work dealt with the contribution of export to growth, understanding the factors affecting trade performance (exchange rate and productivity), and the design of industrial and trade policies Macroeconomic setting
Research Background Emergence of firm-level studies All firms face the same macroeconomic condition YET they respond and perform differently in their export behavior If so, then, there must be firm-specific characteristics that significantly affect its performance in the world market.
Research Thrust Issue Identify firm-level characteristics on Philippine manufacturing firms Value-added No firm-level studies have been conducted for the Philippines
Trade Paradigms Trade pattern is largely explained by comparative advantage (Ricardo, Hecksher-Ohlin) Assumptions were relaxed and extended to increasing returns to scale and oligopolistic structure. Neo-technology trade theorist emphasized the role of innovation in creating markets and conferring cost advantages
Trade Paradigms (cont d) Technology and knowledge are considered part of the production function and technological differences of nations determine export activity (Posner 1961, Vernon 1966). Shortcomings; Model predicts technologically-advanced nations will export new products No dynamics for technological catch-up Learning mode not incorporated Insufficient treatment of technology
Capability Approach (Schumpeterian) Salient features; Innovation is an important determinant of competitiveness Firms carry out innovative activities Importance of monopolistic and oligopolistic market structures (large firms) Continuous search and learning process that can have varying results Advocates building technological capability
Export Performance and Technological Development Sector Annual Growth, % (1985-98) Patents Exports Computing and data processing 109.0 33.8 Electricity and electric power 2.8 24.8 Electronics and components classes 97.2 62.4 Optics/ photography 30.4 (7.4) Communications and networking 95.0 14.8 Electronics, Physics 42.9 25.2 Pharmaceuticals 34.2 38.1 Surgery-body care-cosmetics 62.6 8.3 Source: Montobbio and Rampa, 2005.
Critical Observations Sectors which offered technological opportunities show the largest improvements in world export shares (causality not tested) With technological advancement, structural changes occur mainly dominated by medium-high technology products Evolution of trade theory is accompanied by different approaches in understanding trade performance.
Trade Performance (Firm-level) Significant role of the entrepreneur Knowledge and learning process Mastering technology is costly and firm-initiated (source of comparative advantage) Requires capital accumulation complemented by technological capability measures Important task: identify determinants of export performance, design appropriate policies
Firm-level Studies (Developed and Developing Countries) Author Research Issue Result Montobbio and Rampa, 2005 (LA and Asian firms) Bleaney and Wakelin, 1999 (British firms) Analyze the impact of structural change on the sectoral distribution of export activities Factor/s affecting trade performance FDI, technological specialization, skills and R&D affect export activities; Structural change affect growth trajectories Technologically innovative activities measure by R&D expenditures Sterlacchini, 2001 (Italian firms) - do- Firm size Lefebvre and Lefebvre, 2001 (Canadian SMEs) -do- Import activities, R&D expenditures, distribution access, knowledge intensity and firm size
Firm-level Studies (Developed and Developing Countries) Author Research Issue Result Alvarez, 2002 [Chile] Factor/s affecting trade performance Productivity, firm size, human capital, foreign technical licenses Sarpong and Wolf, 2004 [Ghana] - do- Firm age and firm size Van Dijk, 2002 [Indonesia] -do- Sectoral variation; Relative size, foreign affiliation, firm age (common factor); Skilled labor differs across sector, R&D for mature industries Wignaraja, 2002 [Mauritius] -do- Technological capability index Technological index and foreign ownership
Philippine Setting
Philippine Trade Performance Notable performance of the Philippine products in the world market Dramatic drop in export share of traditional exports, rise of non-traditional export goods Fundamental weakness in the export structure, i.e., concentrated on three products-import dependent and weak backward/forward linkages Resulted in productivity gap where export growth is not translated to output growth
Philippine Trade Performance Country Growth in Crisis, % (1997-98) Growth Post-crisis, % (1999-2000) GDP Export GDP Export Korea (0.9) 1.0 9.8 15.5 Singapore 4.8 (6.2) 7.9 12.4 Thailand (6.1) (1.5) 4.2 13.5 Indonesia (4.2) 0.8 2.8 15.0 Philippines 2.4 19.8 3.6 13.8 Source: Asian Development Bank.
Philippine Trade Studies Author Research Issue Result Pante and Medalla 1990 Medalla, Tecson, Bautista, Power and Associates 1996 Yap 1999 Lall 2000 Evaluated trade and industrial programs Do policy reforms matter? Efficiency issues in the industry Link between trade patterns and competitiveness in the manufacturing sector Export performance and competitiveness Reforms were selective in revitalizing the country; lowered protection but did not alter the biases against export and agriculture Export-oriented industries were efficient but structural weaknesses remain No empirical link established; stressed the dichotomy between domestic manufacturing and export sector Competitive base is narrow, low technological activities, weak technical support for domestic firms
Research Gap Need to have a deeper understanding on trade performance and role of technology variables in the Philippine setting Provide empirical evidence to the rising literature on firm-level export performance in developing countries
Conceptual Framework (Determinants of Export Performance) Firm size Firm age Human Capital Research and development expenditure to compete globally, you have to be big ; capable of bearing large investment and high risks associated with exporting Ambiguous effect Older firms may have accumulated stock knowledge but younger firms are more flexible Strongly related with technological capabilities where skills and education are preferred features for the foreign market Associated with technology and knowledge acquisition Training Foreign Ownership Capital intensity Enhances learning and accumulates additional skills which can improve productivity Access to superior production, technology, management know-how, and network Past innovations and knowledge are embodied in the capital good
Conceptual Framework (Sectoral Variation-Pavitt Taxonomy) Firm Category Determinants of Technological Trajectory Measured Characteristics Core Sector Source of Technology Source of Process Technology Relative Size of innovating firm Supplierdominated Agriculture, housing, private services, traditional manufacturing Suppliers, research extension services, bog users Suppliers Small Scale-intensive Bulk materials like steel and glass, assembly like automobile Product engineering, suppliers, R&D In-house; suppliers Large Specialized supplier Machinery and equipment Design and development users In-house; consumers Small Science-based Electronics, electrical machineries, chemicals R&D, public science, production engineering In-house; suppliers Large Source: Pavitt 1984.
Empirical Model
Empirical Model Source of Data: ADB Investment Climate Survey conducted in year 2002 covering Philippine manufacturing firms, (1) food and food processing (2) textiles (3) garments and (4) electronics and electrical machinery Selected firms on random basis using stratified simple random design Total of 716 firms were surveyed across key urban areas, frequency weight is applied to make it representative across the population size
Empirical Model Variable Definition Measure EXPORTP2 Export propensity Export/sales RSIZE Relative firm size No. of employees in firm j/total no. of employees in sector I where firm j belongs RSIZE2 (relative firm size) 2 SKILLED Skilled manpower Share of skilled workers to total workers in firm j RNDSALES00 R&D expenditure Share of R&D expenditure to total sales in year 2000 TRAINING Skill training Dummy variable where 1= firm offers training, 0=otherwise MNC Foreign affiliation Dummy variable where 1=firm has foreign equity participation more than 505, 0=otherwise CAPINTENSITY Capital intensity Capital stock/labor cost AGE Firm age Year 2003-operating year of the firm AGE2 (Firm age) 2
Empirical Model Sectoral estimations were conducted; sectors (2)&(3) were lumped as one. Dependent variable 0 < EXPORTP2 < 1 Use of ordinary least squares estimation is inappropriate because it fails to recognize the limits of the variable definition while TOBIT estimation fails to recognize endogenous variables bound by zero or positive by definition (censored dataset). ADB dataset is estimated using generalized linear model (GLZ) with Logit as the link function.
Descriptive Statistics, CPBI 1999 Sector No. of Establishments Total Employment Value of Output (P) Food (% share) 1,427 (19.2) 212,707 (19.5) P443,424,756 (25.2) Clothing (% share) 1,303 (17.5) 199,376 (18.3) P85,007,408 (4.8) Electronics (% share) 878 (11.8) 283,889 (26.0) P513,845,489 (29.2) Others (% share) 3,842 (51.6) 393,856 (36.1) P716,407,618 (40.7) Philippines 7,450 1,089,837 P1,758,685,271
Descriptive Statistics, ADB Survey 2002 Activity Food Clothing Electronics Total Non-exporter; EXPORTP2=0 Partial exporter; 0<EXPORTP2<1 Full exporters; EXPORTP2=1 Missing information 204 156 31 391 32 48 24 104 5 88 58 151 52 54 35 141 Total 241 292 113 646 Cleaned firms 189 238 78 505
Key Observations-Food Sector Average years of operation is 23 years Food processing firms Six of the 9 firms are exporters 85% of the survey sample are non-exporters Only 9 firms have foreign affiliation
Key Observations-Clothing Sector Only 30% of the sample are full exporters Clothing firms Firms w/o foreign affiliation are non-exporters Only 52 firms have foreign affiliation Oldest firm is 79 years old; Average age is 15 years
Key Observations-Electronics Sector Majority of the surveyed firms are exporters Electronic firms Oldest firm is operating for 56 years Youngest firm is only 2 years in operation Exporting firms have foreign affiliation
Mean and (Standard Deviation) Variable Food Clothing Electronics RSIZE 0.4 (0.01) 0.3 (0.0) 0.9 (0.2) SKILLED 57.4 (0.3) 79.1 (0.2) 82.4 (0.2) RNDSALES00 0.7 (0.03) 0.7 (0.05) 0.7 (0.04) TRAINING=1; EXPORTP2>0 MNC=1; EXPORTP2>0 8% 9% 48% 3% 17% 58% CAPINTENSITY 4.5 (10.9) 2.6 (8.3) 6.2 (22.6) AGE 23 (20) 17 (13) 15 (10)
Estimation Results (z-values in parenthesis) Variable Food Processing Clothing Electronics Constant -4.7 (-6.2) -1.8 (-4.3) -8.3 (-2.5) RSIZE 18.1 (0.2) 295.3 (2.2)*** 372.8 (0.5) RSIZE2-147.52 (-0.2) -9885.1 (2.4)*** -3734.9 (-0.4) SKILLED 3.4 (1.6) -0.3 (-0.6) 1.3 (1.5) RNDSALES00-54.8 (-1.2) -10.1 (-3.7)*** 275.5 (3.45)*** TRAINING 1.1 (-0.8) 0.5 (0.5) 2.77 (4.9)*** MNC 1.2 (3.5)*** 5.8 (3.3)*** 2.94 (4.2)*** CAPINTENSITY -0.0 (-0.6) -0.0 (-0.04) 0.2 (3.1)*** AGE -0.6 (-1.0) 0.1 (1.8)*** 0.4 (1.5) AGE2 0.0 (1.4) -0.0 (-1.5) -0.01 (-1.7)*** NO. OF FIRMS 189 238 78
Analysis and Explanation: Food Sector Only MNC affiliation has significant effect on export performance Food processing firms Only few firms are exporters; Inconclusive results Few firms have foreign affiliation, spillovers not likely evident Marginal effect: for every 1% increase in foreign equity, Export performance will Increase by 6%
Analysis and Explanation: Clothing Sector MNC and AGE have positive influence implying foreign interest And firm maturity are important; Scale economies and vintage effect Clothing firms Export performance in relation to firm size has an optimal point, beyond which benefits are not positive R&D variable significant But opposite sign; Subcontracting is the Source of technology Size variables are significant, Inverted U-shape implying non-linear effects of size (Wagner 2001)
Analysis and Explanation: Electronic Sector R&D, Training, MNC and capital intensity are positive and significant factors affecting export performance Electronic firms Non-linear effect of AGE is significant; cumulative effects of learning and experience does not matter beyond the threshold level Empirically confirms the predictions of capability literature for high technology firms in developing countries Partly confirms capital-labor complementarity where higher capital intensity requires skilled employees
Recap Determinants of export performance on the Philippine manufacturing firms, 2002 Factors were identified guided by the capability literature Sectoral variation is stressed ADB Firm-level survey
Major Findings Influence of foreign affiliation is similar across the three sectors positive and significant suggesting foreign interest is a good source of knowledge and technology R&D activity, improvement of human capital and capital intensity are important factors for science-based firms. Firm size has an inverted U-shape across all sectors, but significant only for the clothing firms. Firm age has a non-linear influence on electronics while vintage effect is suggested for the clothing sector.
Why are some firms better exporter than others? Foreign interest Skill development Capital accumulation Export performance Age Size
Key Points Research highlights the importance of firm-level characteristics in determining export performance; and Underscores the relevance of foreign direct investment in the firm s pursuit of increasing and/or sustaining global competitiveness.
Firm-Level Determinants of Export Performance: Evidence from the Philippines 45 th Annual Meeting Philippine Economic Society 14 November 2007 Ma. Teresa S. Dueñas-Caparas