IM2012 International Conference on Innovation Methods for Innovation Management and Policy - FOREIGN DIRECT INVESTMENT AND PRODUCTIVITY SPILLOVERS: Firm Level Evidence from Chilean industrial sector. Leopoldo Laborda and Daniel Sotelsek.

Published on: **Mar 4, 2016**

Published in:
Economy & Finance

- 1. FOREIGN DIRECT INVESTMENT AND PRODUCTIVITY SPILLOVERS: Firm Level Evidence from Chilean industrial sector. Leopoldo Laborda Daniel Sotelsek University of Alcalá (Spain) University of Alcalá (Spain) llabordacastillo@gmail.com daniel.sotelsek@uah.es
- 2. The consequences for human welfare involved in questions like these (economic growth) are simply staggering: Once one starts to think about them , it is hard to think about anything else. Robert E. Lucas, Jr. This presentation is organized as follows 1. Introduction, objective, the methodology and hypotesis. 2. The framework proposed to analyze the relation between technical FDI and productivity spillovers in the context of the Chilean industrial sector. 3. Main empirical results obtained. 4 Summary of the main conclusions and policy implications.
- 3. This paper tries to cover: 1. The added efficiency (productivity of the factors) 2. Spillovers from foreign direct investment (FDI) make contribution to productivity growth. Hypothesis: a) Higher competition is associated with larger spillovers from foreign presence in the industry, that is, positive productivity through competition. b) Firms with R&D expenditure gain more productivity spillovers from FDI than those without R&D expenditure c) There is a positivity productivity spillover from FDI d) There are positive FDI spillovers to each component of productivity growth (TEC, TP, and SEC). Introduction In this situation one important questions is: does FDI lead to productivity spillovers?
- 4. Methodology Stochastic frontier approach to estimate FDI productivity spillovers in Chilean manufacturing firms. Apply the method of Data Envelopment Analysis (DEA) and compute the Malmquist index to decompose total factor productivity (TFP) growth into technical efficiency change (TEC), technological progress (TP), and scale efficiency change (SEC). a) Deterministic frontier production functions: the stochastic frontier-inefficiency model. b) Decomposing productivity growth: a generalized Malmquist index [ ]7SECTPTEC 1,1,1,1, 0 ++++ ++= tt i tt i tt i tt iG
- 5. The statistical source used for this analysis is the World Bank’s Enterprise Surveys (ES). Table 1: FP variables
- 6. Empirical Results Chile has been considered as a clasicc emergemt country for 2 reasons_ a) Rate of economic growth during 15 years b) The added value generated by the chilean industry (important differences betwen sectors)
- 7. Empirical Results Chile has been considered as a clasicc emergemt country for 2 reasons_ a) Rate of economic growth during 15 years b) The added value generated by the chilean industry (important differences betwen sectors)
- 8. Empirical Results When observing the whole of the Chilean industry, one can observe very similar behaviors, during the studied periods, in the levels of technical efficiency achieved. In this context, the betterment potential (ease) in the use of their productive inputs is close to 53 %. The estimated efficiency indexes have values between 0 and 1, where the most efficient companies are those closest to 1
- 9. Empirical Results Competition, absorptive capacity and productivity spillovers. In this section hypotheses a) and b) are tested A test of first hypothesis (see table 8) includes a spillover variable, competition variable (HHI) and an interacting variable of spillover and HHI. A test of second hypothesis (see table 9) includes a spillover variable, R&D effort and an interacting variable of R&D effort and spillover in order to evaluate the absorptive capacity of spillovers in the industry. The positive and highly significant coefficients confirm the expected positive and significant output effects of labor and capital The negative and significant coefficient on the spillover variable (spillover) in Models 1, 2, 3 and 4 in Table 9 implies a positive and significant efficiency spillover in the Chilean industrial sectors The coefficient of the research and development dummy (R&D) is positive and significant at the 1% level, suggesting that firms with high R&D effort, on average, have lower efficiencies compared to those with low R&D effort. The negative coefficient of the interacting variable between R&D and spilloverssuggests that firms with high R&D effort gain more spillovers from foreign firms
- 10. Empirical Results Sources of productivity growth and FDI spillovers A positive and significant estimate is found for R&D effort, which indicates that firms with high R&D effort have higher SEC than those with R&D effort. Table 10 shows that the major contribution to productivity growth in the Chilean industrial sectors is from technological progress and for technological efficiency change (with the exception of Biotechnology sector). In contrast, the SEC indices are relatively low, suggesting that this component do not contribute much to productivity growth.
- 11. Conclusions The intra-industry productivity spillovers are examined through the spillover variable, and the roles of competition and R&D in extending spillovers from FDI are evaluated to test a channel of productivity spillovers. Authors like Suyanto et al (2009) suggests that policies for strengthening the absorptive capacity of domestic firms through investing in knowledge and human capital formation may be superior to policies that provide concessions for FDI. In this context more general policies should be pursued, which not only attract FDI but also benefit domestic firms, for example, build modern infrastructure, increasing and strengthening the institutions for accelerating and sustaining economic growth.
- 12. a) Deterministic frontier production functions: the stochastic frontier- inefficiency model. Following Battese and Coelli (1995) the stochastic frontier approach (SFA) is used to estimate a production function and an inefficiency function simultaneously. The Battese–Coelli model can be expressed as follows: ( ) ( ) []1exp;, itititit uvtxfy -×= b In a linear equation, the technical inefficiency effects can be specified as follows: [ ]2ititit wzu += d where itw is an unobservable random variable, which is defined by the truncation of the normal distribution with zero mean and variance, 2 us , such that the point of truncation is ditz- .
- 13. a) Decomposing productivity growth: a generalized Malmquist index. According Orea (2002), if firm i’s technology in time t can be represented by a translog output-oriented distance function ( )txyD itit ,,0 where ity , itx , and t are defined as above, then the logarithm of a generalized output-oriented Malmquist productivity growth index, 1, 0 +tt iG , can be decomposed into TEC, TP, and SEC between time periods t and 1+t : [ ]7SECTPTEC 1,1,1,1, 0 ++++ ++= tt i tt i tt i tt iG where ( ) ( ) [ ]8,,ln1,,lnTEC 1,1,01,1,0 1, txyDtxyD titititi tt i ++++ + -+= ( ) ( ) ( ) [ ]9 ,,ln 1 1,,ln 2 1 TP 1,1,01,1,01, ú û ù ê ë é ¶ ¶ + +¶ +¶ = +++++ t txyD t txyD tititititt i [ ]10ln 11 2 1 SEC ,1, 1 ,1, 1, 1,1, ú û ù ê ë é × ú ú û ù ê ê ë é - + - = + = + + ++ å itn nti N n itn it it nti ti titt i x x e e e e e e From equation [ ]3 , the scale elasticity can be written as [ ]14 2 1 1 å= ++= K k ntnitnknnit tx bbbe The index of scale efficiency change then can be calculated by using equations [ ]10 and [ ]14 .