Transcript Slides Ito & Tanaka
1
Does Material and Service Offshoring Improve Domestic Productivity?
Evidence from Japanese manufacturing industries
Keiko Ito (Senshu University) Kiyoyasu Tanaka (Institute of Developing Economies) WIOD Conference on Industry‐Level Analyses of Globalization and its Consequences Technische Universitaet Wien, Vienna May 26-28, 2010
2
Motivation
Progress in fragmentation of production processes and the international division of labor in East Asia
Increase in offshoring to Asia contributed to skill upgrading in Japan (Ahn et al. 2008, etc.)
Statistical evidence so far that offshoring enhances productivity is still weak
Particularly, on the effects of service offshoring, empirical evidence is scarce
3
Definition of offshoring
Olsen (2006) “Offshoring” refers to : “Relocation of jobs and processes to any foreign country without distinguishing whether the provider is external or affiliates with the firm” “International outsourcing”: relocation of business activities to unaffiliated foreign firms + “International insourcing” : relocation of business activities to affiliated foreign firms
4
The purpose of this study
Utilizing the comprehensive I-O Tables, relatively detailed industry-level data, and trade statistics, we measure the size of offshoring of material inputs and services inputs, and examine the trend and characteristics of offshoring for Japanese manufacturing Using industry-level data, we investigate the impact of offshoring on domestic productivity for the case of Japanese manufacturing Test for the presence of productivity-enhancing effects for both material offshoring and service offshoring Examine how the effects differ for offshoring to different regions and of different activities
5
Potential impacts of offshoring on productivity
By relocating inefficient tasks to low-cost countries, the unit cost of the firm’s product falls (=cost savings) By offshoring less productive stages of production process, firms can shift corporate resource to high productivity activities (product development, process innovation, etc.) (=restructuring) The use of new varieties of imported material or service inputs may increase productivity (=variety effect)
Technological innovations (ICT, transportation), erosion of trade barriers
6
Previous studies (Industry-level studies)
Egger & Egger (2006): EU mfg. 1992-97 Offshoring has a +tive effect on low-skilled worker labor productivity in the long run Amiti & Wei (2006, 2009): US mfg. 1992-2000 (TFP & LP) Service offshoring (++) Material offshoring (+) Daveri & Jona-Lasinio (2008): Italian mfg. 1995-2003 Material offshoring (+) Service offshoring (0) Lin & Ma (2008): Korean mfg. 1985-2001 Marial offshoring (+) Service offshoring (-)
7
Previous studies (Micro-level studies)
Görg & Hanley (2005): Irish electronics firms (1990-95) Material offshoring (+) only for plants with low exp intensities Service offshoring (0) Görg, Hanley & Strobl (2008): Irish mfg Service offshoring (++) only for exporters -- service offshoring requires knowledge to search for partners Hijzen, Inui & Todo (2006): Japanese mfg (1994-2000)
Offshoring (+) 4x greater effect than domestic outsourcing Fari ñas & Mart ín-Marcos (2008): Spanish mfg (1990-2002) Offshoring (+)
8
Unclear effects of service offshoring
McKinsey Global Institute (2005, 2008) US service offshoring to India
+ on US economy Germany & France
-tive Daveri & Jona-Lasinio (2008) - transitional adjustment costs? (
flexible labor mkt, low job creation) low re-employment rates, less - most productive services are offshored to escape existing inefficiencies at home? (
insufficient liberalization)
Lin & Ma (2008): language barriers MGI (2005) : wage levels of destination countries, competitiveness of the home country, ownership structure of affiliated offshore providers
9
Measurement of offshoring
Offshoring =(imported intermediate inputs from all industries) / (total non-energy intermediate inputs)
z i
N j
1
m ij Y i z i
j N
1
m ij IM j IM j Y i
Material offshoring: z=MO (j=mfg. industries) Service offshoring: z=SO (j=offshorable services) - telecommunications, insurance, finance, business services, information services Data: 1990, 1995, 2000 I-O tables (benchmark) Trade statistics (MOF, Japan), METI extended I-O Tables, BOP statistics, JIP database 2009
Table1. Material & service offshoring in Jpn mfg
1990 (%) Imported inputs as a percentage share of total inputs
Materials
of which: from North America from EU
5.95
2.16
1.21
from Asia
US: 12-17 %
from China from ASEAN4 1.60
0.29
0.48
1995 (%)
6.29
2.08
1.14
2.24
0.60
0.61
Services
of which: from North America from EU from Asia from China from ASEAN4
0.21
0.09
0.06
0.04
0.01
0.01
0.20
0.09
0.04
0.05
0.01
0.01
10 2000 (%) 0.11
0.05
0.05
0.01
0.01
2004 (%) Change 1990-2004 (% points) (%)
7.69
0.23
8.89
0.19
2.95
-0.02
49.57
2.29
1.32
1.96
1.56
-0.21
0.35
3.23
0.98
0.96
4.40
1.73
1.16
2.80
1.44
0.68
-9.63
28.81
175.56
505.10
140.19
-11.50
0.08
0.06
0.04
0.01
0.01
-0.003
-0.008
-0.008
0.003
-0.002
-3.72
-11.72
-18.78
47.09
-20.16
Estimation model
Production function
Q i
,
t
T
SO i
,
t
,
MO i
,
t
S i
,
t
,
M i
,
t
,
L i
,
t
,
K i
,
t
LnQ i
,
t
0 1
SO i
,
t
2
MO i
,
t
1
LnS i
,
t
2
LnM i
,
t
3
LnL i
,
t
4
LnK i
,
t
i
t
e i
,
t
11
Regression Model
LnQ i
,
t
0 1
SO i
,
t
i
i
t
2
MO i
,
t t
i
,
t
1
LnS i
,
t
2
LnM i
,
t
3
LnL i
,
t
4
LnK i
,
t
Include industry dummies & year dummies Include one-year lags of offshoring variables Endogeneity --- Use Value added per worker as a dependent variable Data for S, M, L, K --- JIP Database 2009, 1988-2004
Table 3. OLS estimates of the effect of offshoring on TFP (1)
Dependent variable: Δln (Real Output) Δ Service Offshoring Δ Service Offshoring, t-1 Δ Material Offshoring Δ Material Offshoring, t-1 Δ ln(Material Input) Δ ln(Service Input) Δ ln(Number of Workers) (1) -0.58
(1.627) 0.408* (0.198) 0.542*** (0.048) 0.07
(0.036) 0.084
(0.057) (2) 0.387
(1.071) 0.35
(0.296) 0.518*** (0.045) 0.068* (0.029) 0.072
(0.055) (3) -0.588
(1.511) 0.374
(0.706) 0.445* (0.183) 0.258
(0.230) 0.544*** (0.050) 0.071* (0.035) 0.074
(0.065) Δ ln(Number of Nonproduction Workers) Δ ln(Number of Production Workers) Δ ln(Capital Stock) Year Dummies Industry Dummies Observations 0.262* (0.105) yes no 800 0.641
0.279* (0.105) yes no 800 0.621
0.267* (0.109) yes no 750 0.635
(4) -0.588
(1.497) 0.31
(0.753) 0.380* (0.152) 0.207
(0.191) 0.542*** (0.050) 0.075* (0.035) 0.159
(0.130) -0.079
(0.131) 0.267* (0.109) yes no 750 0.637
Table 3. OLS estimates of the effect of offshoring on TFP (2)
Dependent variable: Δln (Real Output) Δ Service Offshoring Δ Service Offshoring, t-1 Δ Material Offshoring Δ Material Offshoring, t-1 Δ ln(Material Input) Δ ln(Service Input) Δ ln(Number of Workers) (5) -0.613
(1.716) 0.151
(0.132) 0.479*** (0.049) 0.062
(0.032) 0.104
(0.057) (6) 0.268
(1.066) 0.139
(0.193) 0.467*** (0.046) 0.061* (0.029) 0.109
(0.055) (7) -0.635
(1.603) 0.136
(0.674) 0.19
(0.130) 0.101
(0.171) 0.484*** (0.052) 0.062
(0.033) 0.09
(0.068) Δ ln(Number of Nonproduction Workers) Δ ln(Number of Production Workers) Δ ln(Capital Stock) Year Dummies Industry Dummies 13 Observations R-squared 0.124* (0.055) yes yes 800 0.700
0.142** (0.050) yes yes 800 0.688
0.152** (0.056) yes yes 750 0.693
(8) -0.658
(1.604) 0.095
(0.691) 0.175
(0.129) 0.092
(0.163) 0.486*** (0.052) 0.063
(0.033) 0.103
(0.136) -0.012
(0.133) 0.149* (0.057) yes yes 750 0.693
Robustness checks
Labor productivity (VA per worker) specification Material offshoring
+tive & significant (Table 4) TFP growth specification Material offshoring
+tive & significant Additional controls (Table)
14
Material offshoring: +tively associated with domestic productivity Service offshoring: insignificant
Table 5. OLS estimates of the effect of material offshoring to Asia on TFP
Dependent variable: Δln (Real Output) Δ Service Offshoring (1) -0.638
(1.649) Δ Service Offshoring, t-1 Δ Material Offshoring to Asia Δ Material Offshoring to Asia, t-1 Δ ln(Material Input) Δ ln(Service Input) Δ ln(Number of Workers) Δ ln(Capital Stock) Year Dummies Industry Dummies Observations 1.037** (0.303) 0.536*** (0.045) 0.071* (0.035) 0.091
(0.061) 0.258* (0.104) yes no 800 0.641
(2) 0.304
(1.033) 1.242* (0.547) 0.509*** (0.044) 0.074** (0.027) 0.091
(0.051) 0.268* (0.102) yes no 800 0.621
(3) -0.481
(1.528) 0.034
(0.674) 1.015*** (0.220) 0.963** (0.356) 0.530*** (0.046) 0.077* (0.032) 0.094
(0.067) 0.260* (0.105) yes no 750 0.635
(4) -0.489
(1.523) 0.031
(0.691) 0.966** (0.288) 0.915** (0.310) 0.530*** (0.046) 0.078* (0.032) 0.261* (0.105) yes no 750 0.637
(5) -0.637
(1.720) 0.466* (0.225) 0.480*** (0.047) 0.063
(0.032) 0.105
(0.060) 0.127* (0.055) yes yes 800 0.7
(6) 0.232
(1.048) 0.728
(0.425) 0.465*** (0.045) 0.064* (0.028) 0.116* (0.053) 0.141** (0.049) yes yes 800 0.688
(7) -0.565
(1.613) -0.027
(0.652) 0.622* (0.234) 0.690* (0.337) 0.484*** (0.049) 0.066* (0.031) 0.098
(0.069) 0.156** (0.056) yes yes 750 0.693
(8) -0.575
(1.615) -0.037
(0.652) 0.606* (0.287) 0.675* (0.320) 0.484*** (0.048) 0.067* (0.031) 0.154** (0.055) yes yes 750 0.693
Table 6. OLS estimates of the effect of information services offshoring on TFP
Dependent variable: Δln (Real Output) (1) Δ Information Services Offshoring Δ Information Services 10.75
(21.480) Offshoring, t-1 Δ Material Offshoring 0.409* (0.197) Δ Material Offshoring, t-1 Δ ln(Material Input) Δ ln(Service Input) Δ ln(Number of Workers) Δ ln(Capital Stock) Year Dummies Industry Dummies Observations R-squared 0.542*** (0.049) 0.071
(0.036) 0.084
(0.057) 0.261* (0.105) yes no 800 0.641
(2) 82.62** (24.36) 0.347
(0.285) 0.517*** (0.044) 0.068* (0.030) 0.07
(0.057) 0.272* (0.107) yes no 800 0.621
(3) 48.84
(27.71) 116.4** (33.40) 0.377* (0.164) 0.204
(0.213) 0.542*** (0.050) 0.074* (0.036) 0.071
(0.068) 0.258* (0.112) yes no 750 0.635
(4) 43.99
(32.42) 108.5** (37.80) 0.343* (0.144) 0.178
(0.184) 0.541*** (0.050) 0.076* (0.036) 0.259* (0.111) yes no 750 0.637
(5) -0.341
(24.46) 0.152
(0.132) 0.480*** (0.049) 0.063
(0.032) 0.105
(0.057) 0.124* (0.055) yes yes 800 0.700
(6) 84.85** (25.02) 0.148
(0.186) 0.465*** (0.045) 0.061* (0.030) 0.105
(0.057) 0.130* (0.052) yes yes 800 0.688
(7) 65.72
(43.04) 146.5** (49.01) 0.13
(0.124) 0.056
(0.160) 0.482*** (0.052) 0.066
(0.034) 0.082
(0.072) 0.135* (0.062) yes yes 750 0.693
(8) 65.40
(47.45) 145.9* (55.44) 0.129
(0.126) 0.055
(0.155) 0.482*** (0.052) 0.066
(0.034) 0.135* (0.061) yes yes 750 0.693
Summary of the results
Material offshoring: +tively associated with domestic productivity Service offshoring overall: insignificant
17
Material offshoring to Asia : +tive & significant Material offshoring to NA & EUR: insignificant
Characteristics of destination country do matter
Offshoring to low-cost region brings greater productivity gains
Information service offshoring: +tive & significant Business service offshoring: insignificant
Effect of service offshoring on productivity
Business services involve a substantial amount of judgment & communication b/w clients & professionals
These tasks need to be standardized & reorganized in order for foreign firms to provide them from a distance
Peculiarities of Japanese language & business culture may be a barrier
18
19
Conclusion
Material offshoring has a positive effect on productivity while service offshoring does not. (in line with previous results of Italy & Korea, but at odds with results of U.S.) Although the level of service offshoring is still low, increase in info service offshoring would improve productivity.
However, it has been pointed out that the corporate organization of Jpn firms is not well suited to standardized IT systems. (decision making processes require substantial interdivisional comm. ) (Motohashi 2008) Firms which restructure their IT management method would be able to benefit from offshoring of services. Material offshoring to Asia has a positive effect on productivity. The large differences b/w factor prices at home & Asian countries, coupled with geographic proximity, appears to have yielded a productivity-enhancing effect.
20
Future research
Peculiarities of the Japanese business practice or peculiarities of the Japanese langugage & culture? business services
examine effects of “international” outsourcing and “domestic” outsourcing of
Effects of offshoring on various types of workers (low-skilled or high-skilled workers)
Productivity effect of service offshoring by region
21