“learning by exporting and high- tech capital deepening in singapore manufacturing industries,...
TRANSCRIPT
“Learning by Exporting and High-tech Capital Deepening in
Singapore Manufacturing Industries, 1974-2006”
โดย ดร. เอกพล จงวิ ล�ยวิรรณ วิ�นพ�ธที่�� 23 กรกฎาคม 255
1 เวิลา 1400. น .
Structural Characteristics Structural Characteristics
Figure 1: Exports and High-tech Capital Formation in Singapore Manufacturing Industries, 1974-2006.
020
0000
040
0000
060
0000
080
0000
0V
alue
(th
ousa
nd)
1970 1980 1990 2000 2010Year
Exports High-tech Capital
Figure 2: The Ratios of Export Intensity (XI) and High-tech Capital Deepening (T) in Singapore Manufacturing Industries, 1974-2006.
.4.5
.6.7
.8.9
1970 1980 1990 2000 2010Year
Export Intensity (XI) High-tech Capital Deepening (T)
Structural Characteristics Structural Characteristics (continued)(continued)
Structural CharacteristicsStructural Characteristics (continued)(continued)
Figure 3: Scatter Plot of the Ratios of Export Intensity (XI) and High-tech Capital Deepening (T) across Singapore Manufacturing Industries.
Food
Textile
Wearing Apparel
Petroleum
Petrochemical Non-metallic
O thers
LeatherWood
Paper
Printing
Pharmaceutical Rubber
Basic Metal
Fabricated Metal Machinery
Electronic
O ptical Instruments
Transportation Equipment
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1
XI
TI
Learning by ExportingLearning by Exporting
Learning-by-exporting hypothesis refers to the mechanism through which exporting activities generate externalities via technological and informational spillovers and lead to productivity growth.
However, the empirical evidence is rather mixed and less clear-cut.
Clerides, et al. (1998), Bernard and Jensen (1999), Aw, et al. (2000) for South Korea, and Greenaway and Kneller (2004) find weak evidence of the learning effect.
This is in contrast with the studies by Aw, et al. (2000) for Taiwan, Baldwin and Gu (2003), and De Loecker (2007).
High-tech Capital AccumulationHigh-tech Capital Accumulation High-tech capital deepening has been a source of debate that it results
in higher productivity performance, thereby enhancing productivity growth and profitability.
As is the learning effect, its impacts on the productivity growth are unclear.
See Morrison and Berndt (1991), Oliner and Siechel (1994), Siegel and Griliches (1992), Siegel (1997), and Oliner and Siechel (2000).
Data Sources and MeasurementsData Sources and Measurements The two-digit SIC manufacturing industries data are retrieved from the
Census of Manufacturing Activities (CMA) from 1974 to 2006, provided by Singapore Economic Development Board (EDB)
Various price deflators are gathered from Monthly Digest of Statistics Singapore (MDSS), provided by Singapore Department of Statistics.
Data Sources and Measurements Data Sources and Measurements (continued)(continued)
TFPG MeasurementsTFPG Measurements
Following Feenstra and Hanson (1997), I assume the translog real value added function.
,where and B is a 4x4 matrix of parameters.
Definition:
With the assumption of linear homogeneity, it is straightforward to show that
Econometric ModelsEconometric Models
Specification I:
Specification II:
Econometric Estimations and DiscussionsEconometric Estimations and Discussions
To control for unobservable industry heterogeneity, I employ the Generalized Least Squares (GLS) estimation based on the Random Effects model.
I employ the Breusch-Pagan test to investigate whether the industry-specific effects exist.
I employ the Hausman’s Specification Test to inspect whether GLS estimates are consistent.
It is not appropriate to impose parameter restrictions at priori.
I obtain unrestricted estimates and test whether they well behave.
Empirical Results: Specification IEmpirical Results: Specification I
Empirical Results: Specification IIEmpirical Results: Specification II
Main Findings and Policy ImplicationsMain Findings and Policy Implications
The learning effect and high-teach capital deepening affect TFPG via non-neutral (factor-biased) technological progress.
A policymaker should also take their income distribution effects into consideration – Not everyone gains from overall productivity growth, Some (Factors) May Lose!Some (Factors) May Lose!
Structural ChangesStructural Changes
Main Findings and Policy ImplicationsMain Findings and Policy Implications
The factor-biased technological progress resulting from the export-led development and the high-tech capital accumulation is period-specific.
It may be interesting to further investigate the factors that bring about the structural changes.
Contributions to TFPGContributions to TFPG
By using the GLS estimates, I evaluate the followings at mean values.
Contributions to TFPG Contributions to TFPG (continued)(continued)
Main Findings and Policy ImplicationsMain Findings and Policy Implications The learning effect can explain TFPG more largely than a high-tech
capital accumulation.
A policymaker should be aware that only some industries gain from the export-led development and high-tech capital deepening.
Some (Industries) May Lose!
THANK YOUTHANK YOU