Parallel Session 5: FDI and development

Transcription

1 ASIA-PACIFIC RESEARCH AND TRAINING NETWORK ON TRADE ARTNeT CONFERENCE ARTNeT Trade Economists Conference Trade in the Asian century - delivering on the promise of economic prosperity rd September 2014 Parallel Session 5: FDI and development

2 The Worldwide Shift of FDI to Services - How does it Impact Asia? New Evidence from Seventeen Asian Economies. Nadia Doytch CUNY Brooklyn College and Asian Institute of Management (Non-Resident Research Fellow)

3 0 0.5 Trillion of USD Total Sectoral FDI in Trillion USD Aggregate FDI increased 9 times (from 203 Bln. USD to 1995 Bln. USD) Mining FDI increased 10 times (from 14.3 Bln. USD to Bln. USD) Manufacturing FDI increased 5 times (from 50.4 Bln. USD to Bln. USD) Total Services FDI increased 15 times (from 74 Bln. USD to 1,146 Bln. USD) Financial Services FDI increased 14 times (from 39.4 Bln. USD to Bln. USD) Glabal Sectoral FDI Aggregate FDI Manufacturing FDI Financial FDI Mining FDI Total Services FDI

4 0 Bln. of USD Total Sectoral FDI in Billion USD, South and East Asia and the Pacific, Total FDI Manufacturing FDI Financial Services FDI Business services FDI Mining FD Aggregate Services FDI Trade Services FDI FDI flows by industry, Totals for South and East Asia and the Pacific Countries: Australia, Bangladesh, Brunei, Cambodia, China, Hong Kong, India, Indonesia, Japan, Rep. of Korea, Lao, Malaysia, Pakistan, Philippines, Singapore, Thailand, Vietnam

5 0 50 Bln. of USD Total Sectoral FDI in Billion USD, South and East Asia and the Pacific, Mining FDI Aggregate Services FDI Trade Services FDI Manufacturing FDI Financial Services FDI Business Services FDI FDI flows by industry, Totals for South and East Asia and the Pacific

6 Diverging Manufacturing and Services FDI Shares of Total FDI in Asia, Manufacturing FDI share in Total total services FDI share of total Manufacturing and Services FDI shares of Total FDI, 17 South and East Asian Economies

7 Why Study the FDI-Growth Relation? FDI has been looked at as a stable development engine, because of its hypothesized positive spillovers on economic growth. FDI is a stable component of capital flows to emerging economies. FDI became a large component of capital flows to these economies.

8 Productivity Spillovers Cost A3 A1 A2 AC0 AC1 Q A voluntary or involuntary transfer of MNEs intangible assets to domestically owned firms lower AC curve (positive spillover). Domestic firms are imperfectly competitive. MNE competes in quantity and captures market. Firm spreads fixed costs over smaller market moves up AC curve (negative spillover) MNE drains resources away (e.g. labor) Costs go up (negative spillover)

9 This the first systematic analysis of sectoral spillover of FDI disaggregated down to the industry level for South and East Asian economies. Study Objectives and Contribution Disaggregate total FDI and examine the growth effects of nine different types of FDI flows, in addition to aggregate FDI- mining, manufacturing, construction, aggregate service, and within services: financial, trade, tourism, business services, and transport & communications. Address how the different sector- and industry-level flows affect growth of manufacturing and service sectors in addition to the growth of aggregate economy. (We run 30 different models). We explore different institutional quality control variables, such as government stability and anti-corruption. Employ a dynamic panel GMM estimator that controls for endogeneity to a new data set for: Australia, Bangladesh, Brunei, China, Hong Kong, India, Indonesia, Japan, Rep. of Korea, Malaysia, Pakistan, Philippines, Singapore, Thailand, Vietnam) for (unbalanced) that was not previously available.

10 Main Findings The manufacturing sector experiences positive productivity spillovers from its own sector (manufacturing) FDI and negative spillovers from Business Services FDI. The services sector experiences positive productivity spillovers from particular services FDI inflows: financial services, trade, and transport & communications. It has negative spillovers from extractive industries FDI. The overall effect on aggregate economic growth from FDI is positive. Particularly important are the services flows of financials, trade, and transport & communications, as well as extractives industries FDI.

11 Two Approaches to Study the Growth Effect of FDI Macro approach Cross-country growth effects of total FDI (Borstein, degregorio, Lee, 1998, Blomstrom, Lipsey and Zejan, 1994, Alfaro et al., 2004) FDI inflows benefit growth. But does not explain intersectoral differences. Micro approach Plant-level productivity of firms in manufacturing industry (Haskel, Pereira and Slaughter, 2002, Blalock and Gertler, 2003, Aitken and Harrison, 1999, Gorg and Stroble, 2001, Lipsey (2003, 2004) Results ambiguous. Cross-country analysis not possible. Sector-level Previously Doytch and Uctum (2011) found some negative spillovers of non-financial services FDI on manufacturing growth based on geographical region and level of development. Due to lack industry-level service FDI data, further investigation was not conducted in this study.

12 Aggregate FDI shares of GDP, Australia Hong Kong SAR, China Korea, Rep. Singapore Total FDI, High Income Economies Brunei Darussalam Japan Macao SAR, China The leaders in increasing FDI shares of GDP between 1999 and 2012 are: - High Income countries: Hong Kong (+0.85) and Australia (+1.84); - Middle Income Countries: Indonesia (+2.15) - Low Income Countries: India (+3.06), Cambodia (+0.68), and Bangladesh (+0.52) China Malaysia Thailand Indonesia Philippines Bangladesh India Pakistan Cambodia Lao PDR Vietnam Total FDI share of GDP, Middle Income Economies Total FDI share of GDP, Low Income Economies

13 Manufacturing, Australia Hong Kong SAR, China Korea, Rep. Singapore Manufacturing FDI, High Income Economies Brunei Darussalam Japan Macao SAR, China The leaders in increasing FDI shares of GDP between 1999 and 2011 are: - High Income countries: Brunei (+ 2.29) and Australia (+0.99); - Middle Income Countries: Thailand (+0.20) - Low Income Countries: India (+3.96), Pakistan (+1.15), and Bangladesh (+1.28) China Malaysia Thailand Manufacturing FDI share of GDP, Middle Income Economies Indonesia Philippines Bangladesh India Pakistan Manufacturing FDI share of GDP, Low Income Economies Cambodia Lao PDR Vietnam

14 Business Services, Australia Hong Kong SAR, China Korea, Rep. Singapore Business Services FDI, High Income Economies Brunei Darussalam Japan Macao SAR, China The leaders of increasing FDI shares of GDP between 1999 and 2012 are: - High Income countries: Hong Kong (+3.71), and Singapore (+3.61); - Middle Income Countries: Thailand (+1.65) and Philippines (+15.39) - Low Income Countries: Vietnam (+17.48) China Malaysia Thailand Business FDI, Middle Income Economies Indonesia Philippines Bangladesh India Pakistan Business Services FDI, Low Income Economies Cambodia Lao PDR Vietnam

15 Trade Services, Australia Hong Kong SAR, China Korea, Rep. Singapore Trade Service FDI, High Income Economies Brunei Darussalam Japan Macao SAR, China The leaders of increasing FDI shares of GDP between 1999 and 2012 are: - High Income countries: Hong Kong (+3.19), Brunei (49.26). - Middle Income Countries: Malaysia (+1.83) - Low Income Countries: Bangladesh (+27.70), India (+1.81) Vietnam (+4.51) China Malaysia Thailand Indonesia Philippines Bangladesh Lao PDR Vietnam India Pakistan Trade Services FDI, Middle Income Economies Trade Services FDI, Low Income Economies

16 A Traditional Growth Model y it conv log( yi, t 1) 1W it 2Xit i t it (1) y i, t 1 = lagged real per capita income of country i, y it = real per capita GDP country i, W i = a vector of traditional determinants of Solow growth model, X i = recent determinants (FDI, institutional factors) i, t = country-specific and a time-specific effect.

17 Empirical Model y k it k j 0 1 log( yi, t 1) 2 fit 3xit 4 D t i it, i. i. d(0, ), i.i.d.(0, ) i ~ i it, and E[ ] 0 ~ k y it = real income per capita (constant 2005 prices). k yi t 1, = lagged level of income per capita j f it = FDI net inflows share of GDP k= GDP, manufacturing value added, and services value added. i it j= total FDI, mining, manufacturing, construction, aggregate services, financials, trade, tourism, business services, transport & communications. i= 1,..,17 and t= 1,..,27 ( the cross-sectional and time dimensions of the data) x it = gross fixed capital formation share of GDP, gross secondary school enrolment ratio, natural resources rents share of GDP, intuitional quality (government stability and anti-corruption run interchangeably). t D = 15 -dummy variable

18 Empirical Methodology and Data Advantage of the dynamic panel GMM estimators Arellano-Bond (AB) and Blundell-Bond (BB) Control for joint endogeneity while preserving the dynamic nature of the regression BB: Instruments = lagged level observations for differenced variables and lagged differenced observations for level variables. Sources: FDI net inflows: central banks web sites, ASEAN, OECD and UNCTAD. Domestic Investment share of GDP, secondary school enrollment ratios, natural resources rents: WDI. Government stability and Anti-corruption: ICRG

19 Necessary Conditions for BB Dynamic Panel Estimation (i) No second order autocorrelation in the error term E E E k [ y i, t s 1 ( it i, t )] 0 for s 2 and t=3,.t [ x i, t s 1 ( it i, t )] 0 for s 2 and t=3,.t j [ f i, t s 1 ( it i, t )] 0 for s 2 and t=3,.t; (ii) Country-specific effects not correlated with firstdifferenced dependent or independent variables: k k E[( y i, t 1 yi, t 2)( i it )] 0 E[( x i, t 1 xi, t 2)( i it )] 0 j j E[( f i, t 1 fi, t 2)( i it )] 0

20 Independent variables GDP per Capita Growth, Institutional Control- Government Stability (1) Total (2) Extractive Industries (3) Manufacturing (4) Total Services (5) Financial (6) Construction (7) Trade (8) Tourism (9) Business (10) Transport Log of Lagged Real Services Value added per capita Fixed Capital share of GDP Natural Resources Rents share of GDP.998*** (219.24) (6.19) *** (-2.98) 1.001*** (258.04) (8.61) (-1.42).993*** (322.15) (6.06) (-1.46).985*** (249.47) (6.37) 4.06e-06 (0.03).996*** (266.04) (7.52) (-1.13).995*** (262.46) (7.76) (-1.01).992*** (248.26) (8.21) (-1.00).994*** (450.74) (6.49) (-1.28).998*** (251.47) (7.12) *** (-2.84).996*** (259.35) (8.03) (-1.26) Gross Sec. School Enrollment Ratio () (-0.81) (-1.54) (0.40).0005** (2.36) (-0.82) -6.61e-06 (-0.04) (0.30) (-0.39) (-1.12) (-0.78) Government Stability Index (-0.10).001 (0.82).001 (0.68).0007 (0.45).001 (0.79).0003 (0.20).002 (0.96).001 (0.87).003** (2.10).0006 (0.48).068** (2.22).021** (2.22).148 (0.93).170*** (4.04).443* (1.75).160 (0.13).455* (1.86) (-0.08).025 (0.49).235* (1.65) Observations Number of countries

21 GDP per Capita Growth, Institutional Control- Government Stability Independent variables (1) Total (2) Extractive Industries (3) Manufacturing (4) Total Services (5) Financial (6) Construction (7) Trade (8) Tourism (9) Business (10) Transport Log of Lagged Real Manuf Value Added per capita Fixed Capital share of GDP Natural Resources Rents share of GDP.988*** (196.36).003*** (6.27).0003 (1.10) *** (211.10).004*** (7.21) (-1.22).992*** (282.90).003*** (3.76).0002 (0.57) *** (188.85).003*** (5.00) (-1.21) *** (241.72).004*** (6.35).0001 (0.32).995*** (218.36).004*** (7.10).0004 (0.91) 1.005*** (177.94).004*** (5.75) (-0.35).996*** (257.19).005*** (10.87) -.001** (0.40).999*** (197.75).003*** (3.57) (-1.52) 1 (117.04).004*** (5.04).002 (1.20) Gross Sec. School Enrollment Ratio () (0.33) *** (-3.35) (-0.22) (-1.61) (-1.68) (-1.25) *** (-3.31) (-0.42) *** (-2.84) ** (-2.03) Government Stability Index.0007 (0.22).002 (0.75) (-1.46).002 (0.63) (-0.62) (-0.89) (-0.12) -.016*** (-4.89).003 (1.32) -.007** (-2.85).033 (0.31).018 (0.73).681*** (2.99) -.170* (-1.86).341 (0.46) (-0.92) (-1.00) (0.39) -.439*** (-5.69) (-0.49) Observations Number of countries

22 GDP per Capita Growth, Institutional Control- Government Stability Independent variables (1) Total (2) Extractive Industries (3) Manufacturin g (4) Total Services (5) Financial (6) Constructio n (7) Trade (8) Tourism (9) Business (10) Transport Log of Lagged Real GDP per capita.993*** (315.97).999*** (268.58).994*** (342.17).987*** (217.97).997*** (331.35).997*** (289.98).993*** (265.72).998*** (213.11).997*** (274.39).998*** (323.12) Fixed Capital share of GDP (7.00) (9.02) (5.98) (5.46) (7.51) (6.48) (7.74) (5.67) (7.09) (17.72) Natural Resources Rents share of GDP (-1.05).0003** (2.21).0001 (0.89).0001** (0.97) (0.34).0001 (0.93).0001 (0.81).0002 (0.19) (0.24) (-0.20) Gross Sec. School Enrollment Ratio () (0.06) (-1.09) (0.26).0004* (1.70) (-1.23) (-0.30) (0.33) (-1.19) (-0.86) (-1.38) Government Stability Index.003** (2.03).001 (0.69).0008 (0.53).0004 (0.23).0008 (0.45).0007 (0.38).001 (0.82).001 (1.22).002 (1.42) (-0.27).055 (1.52) -.071*** (-3.76).209 (1.16).167*** (2.61).401* (1.62).232 (0.21).530* (1.87) (-0.29).027 (0.43).372** (2.12) Observations

23 The overall effect on aggregate economic growth from FDI is positive. Particularly import are the services flows of financials, trade, and transport & communications, as well as extractives industries FDI. Empirical Results &Conclusion The own sector spillovers are positive both for manufacturing and services FDI. The services sector is positively impacted by particular services FDI inflows: financial services, trade, and transport & communications. The intersectional spillovers, if any, are negative: mining FDI on the services sector and business services FDI on manufacturing sector.