PRODUCTIVITY LED GROWTH IN OECD COUNTRIES: AN EMPIRICAL ASSESMENT

Abstract

Labour productivity and economic growth relation is investigated in the paper. Labour productivity could be counted an engine for industrial growth and hence for the economy. In this paper, labour productivity and economic growth is investigated between 2004 and 2012 for 18 OECD countries by LSDV panel data and System GMM panel data approaches. Obtained results show that there is a significant relation between labour productivity and economic growth for 18 OECD countries.

Full text

PRODUCTIVITY LED GROWTH IN OECD COUNTRIES: AN EMPIRICAL
ASSESMENT
Ömer Tuğsal DORUK, Lecturer, Ġstanbul Esenyurt University,
Ergül SÖYLEMEZOĞLU, Lecturer, Ġstanbul Esenyurt University
Abstract
Labour productivity and economic growth relation is investigated in the paper. Labour
productivity could be counted an engine for industrial growth and hence for the economy. In
this paper, labour productivity and economic growth is investigated between 2004 and 2012
for 18 OECD countries by LSDV panel data and System GMM panel data approaches.
Obtained results show that there is a significant relation between labour productivity and
economic growth for 18 OECD countries.
Keywords: Labour productivity, economic growth, productivity led growth in OECD
countries
1. Introduction
Labour productivity and economic growth relation is important factor for economy due to
productivity affected on total industry and hence on the economy. The paper consisted 4
subsection; the first section consists general introduction, in the second section; the literature
review and general theoretical assumptions are reviewed, the third subsection in which the
empirical investigation, data and methodology are described and the last section, general
summary and empirical findings are discussed.
2. Productivity Led Growth: Theoretical Underpinnings and Literature
Labour productivity is linked with Solow (1957)’s seminal work and growth model, which
can be evaluated in total factor productivity.
A simple productivity model can be assumed as following;
Y = A f(K,L) (1)
where Y, A, K and L denote GDP, technology, capital and labour, respectively.
Therefore labour productivity is one of the essential variables for GDP in Solow (1957)
model.
Corvers(1997) and Nicolini (2011) pointed out that the labour productivity growth is depends
to human capital and thus human capital is an essential for labour productivity. Even if

Corvers (1997) emphasized that human capital is more important than capital/physical
investment. Nicolini (2011) underlined that human capital is need to be supported by physical
capital. The combination of the pshsical capital and human capital leads to negative or
positive productivity levels according to their levels.
For Jorgensen (2011:159);
“… Productivity growth is the key economic indicator of innovation”
Dixon and Lim (2012) researches, empirically, labour productivity, technological shocks and
policy reforms of productivity relation in Australia. And Dixon and Lim (2012) emphasized
the relations in Australia for productivity.
A well-known fact that called as Verdoorn Law, which is one of the Kaldorian Growth Laws,
is investigating whether the labour productivity growth effects on the output growth.
Verdoorn (1949) found that the growth of labour productivity and the output growth is
positively linked prior to World War II era. Boulier (1984) noticed that the replications of
Verdoorn Law have found same results.
Upender (1996: 7) pointed that labour productivity led growth issue in the literature as
following;
“…Although the importance of labour quality has long been recognised, its impact on
productivity or economic growth has seldom been examined. Some studies have attempted to
identify how labour quality and economic development are related; however, most of these
studies employ proxy variables to represent labour quality”
For Fillippetti and Payrache (2013), the growth performance difference in OECD economies
linked to institutional environment, economic regulation of product and labour markets.
Given the close link between productivity growth and technological progress, cross- industry
differentials are partly related to different patterns of innovative activity and adoption
(Nicoletti et al, 2003: 11-18).
During the Eurozone Crisis, aggregate productivity declined in Greece. European Central
Bank comments for this problem as on below;
“…place great emphasis on the need to increase labour productivity [in
Greece], for in the medium and long term, growth is dependent on this very factor. And

that is why we say that this is a crucial reform . . .” (Polidori, 2010).
For Rada (2007), one of the repeatedly raised central questions in development economics is
concerning the mechanisms through the economic growth and labour productivity
relationship. Thus, higher growth rates lead the declining of unemployment and
underemployment issues.
Rada (2007) underlined that the advanced economies, especially for Japan have managed to
catch up process between rapid growth rates and labour productivity in its economic history.
Labour productivity is more sensible than capital productivity term. And labour productivity
is important due to labour produces commodity outputs.
In terms of labour theory of value; production mean is not a fixed capital, capital is only to
facilitate the labour productivity, in terms of output.
Mohun (2009: 1028) pointed out that;
“… while labour productivity requires a measure of real output, capital productivity is a ratio
of two money magnitudes in different sets of prices. Each of these magnitudes is the product
of a price index and a constant price variable.”
In the literature, labour productivity is used for innovation variable. And capital productivity
is not a sensible as labour productivity for production. And labour productivity was a essential
for development policy of some countries, i.e Japan.
3. Empirical investigation
3.1. Data and Methodology
3.1.1.Maddala and Wu (1996) ADF-Fisher χ2 Panel Unit Root Test
ADF- Fisher χ2 panel unit root, which based on Fisher, is combined with Fisher’s Pλ
testi and some other test thereby suggested by Maddala and Wu (1996).
N denotes unit root test and Pi denotes P value for ith test;
-2

 Pi (2)
The unit root test has 2N degrees of freedom and χ2 distribution which has Ti∞
(Hsiao, 2004:301).

3.1.2.Levin,Lin ve Chu(2002) Panel Unit Root Test
Levin,Lin ve Chu (2002) panel unit root test is extended verson of Dickey Fuller Test
and developed for panel data, which based on the equation as following;
∆Yi,t=αi+ρ Yi,t-1+

 k∆ Yi,t-k+δit+θt+uit (3)
Levin,Lin ve Chu (2002) panel unit root test predicted ρ in pooled OLS (Asteriou &
Hall, 2007: 367).
3.1.3. Bias Corrected Least Squares Dummy Variable (LSDV) Approach
Yi,t = α yi,t-1+ΣiβiXi,t+ui,t+εi,t (4)
ui,t are fixed effects. LSDV approach which based on Kiviet (1999) handles small cross
section problem in panel data.
3.1.4. System GMM Approach
The system GMM approach which is used in this paper based on Blundell and Bond which is
using orthogonal deviation instead of differencing method is used in Arellano Bond method.
The system GMM approach based on Hausman ve Taylor (1978) and described as,
mathematically;
yit = x’itβ + Z’iγ + vit (5)
where β K x 1 and γ gx1 which the variables of xit are time dependent and cross
section dependent. Zi denotes time inconsistent variables (Baltagi, 2005: 142).
The vector form of this equation is;
Yi = Wiη + vi (6)
(Baltagi, 2005)
Arellano and Bover GMM predictor uses the transformed form of the equation on (6) and
found the equation on below;
'
1/
T
C
HT



(7)
Where the line that provides C, C1T = 0 condition becomes a (T-1) x T matrix form of (T-1)
(Tatoğlu, 2012: 86). A residual that transformed according to Arellano-Bover System GMM
as following;

i
ii
i
CV
V HV V




(8)
(Baltagi, 2005).
The instrument matrix for fully transformed system according to Arellano- Bover GMM
Method as below;
'0
'
0'
i
i
i
i
w
Mw
m






(9)
The moment condition is; ;
E(M’i Hvi) = 0 .
And the moment condition for system GMM is;
'
( ) 0
ii
E M HV 
(10)
Where
N
H I H
(11)
and
ˆN
I  
(12)
And we get Arellano Bover predictor as following;
' ' 'M HY M HW M Hv


(13)
The variance covariance matrix as following; ;
'
1
ˆˆ
ˆ
N
ii
i
uu
N




(14)
(Tatoğlu, 2012).

3.2. Data and Findings
The labour productivity variable productivity per unit (henceforth LP) and GDP growth rates
(henceforth gGDP) were used in this paper. The data of the countries were taken from OECD
Stat Database (2013).
Table 1. The countries in the sample
Belgium
Czech Republic
Estonia
France
Greece
Hungary
Ireland
Korea
Luxembourg
Netherlands
Poland
Portugal
Slovak Republic
Slovenia
Spain
Sweden
Turkey
United Kingdom
The country list in the sample is depicted in Table 1. Totally, 18 OECD countries are in the
sample as seen in Table 1.
gGDP and LP for 18 OECD countries are shown in Figure 1. Graphically, the general path of
the variables is seen in Figure 1. As seen in figure 1, gGDP and labour productivity are
closely linked.

Note: For countries: the order is same as in Table 1.
Figure 1. The gGDP and labour productivity path in 2004 and 2012
Table 2. Descriptive Statistics
gGDP
LP
Mean
2.10
1.79
Maximum
10.49
18.04
Minimum
-14.09
-6.92
Std. Dev.
3.89
2.96
Observations
162
162
The descriptive statistics are shown in Table 2. The mean of the gGDP is 2.10 % and of
labour productivity is 1.79 per labour. The maximum value of gGDP is %10.49 and of labour
productivity is 18.04 per labour in the sample. The data were consisted of 162 observations.
-10
010 20
-10
010 20
-10
010 20
-10
010 20
1 2 3 4 5
6 7 8 9 10
11 12 13 14 15
16 17 18
gGDP lp

Table 3. Correlation Analysis
gGDP
LP
0.61
The correlation between labour productivity and economic growth is depicted in Table 3. The
correlation relation between the variables is positive and the effect of labour productivity on
economic growth is %61.
Table 4. Panel Unit Root Test Results
LLC
ADF-Fisher
PP-Fisher
LP
-5.79a
82.36a
86.55a
gGDP
-6.09a
78.98a
70.79a
Notes: a denotes at %1 statistical significance level. And the maximum lags for the unit root
process were selected according to Akaike Information Criterion. The unit root test results
based on no constant and no trend unit root model for all unit root tests.
The panel unit root test results are shown in Table 4. According to the results, all the variables
are I(0), which means that all the variables are stationary in the levels. After we had found the
variables stationary levels, we proceed the system GMM approach and bias corrected Least
Squares Dummy Variable Regression (LSDV) methods.
The system GMM and LSDV results are presented in Table 5. The results according to Panel
LSDV approach; there is significant relation between labour productivity and growth, which
economic growth is positively linked with labour productivity by %88. In system GMM
approach in which one and two lag of labour productivity are used for finding whether there is
a lagged relation between labour productivity and growth in the sample. However, there is no
lagged relation between the variables. However there is a strong and positive link between
labour productivity and gGDP growth as 1.59 at %1 statistical significance level.

Table 5. System GMM and LSDV Results
System GMM
LSDV
C
-0.70 (0.51)
-
LP
1.59 (0.25)a
0.88a (0.08)
ggdpt-1
0.49 (0.15)a
0.43a(0.06)
LPt-1
-0.24 (0.26)
LPt-2
-0.34 (0.23)
F Test
21.20a
m-1
-2.20a
m-2
-1.75
Instruments
14
Hansen Test
results
13.02
Diff-Hansen
for GMM
instruments
10.34
-
Diff-Hansen
for iv
instruments
1.34
-
Notes: a shows significance at %1 statistical significance level. Parentheses show standard
errors. For System GMM; The Wald Test result is significant at %1 statistical significance
level. According to Difference in Hansen Test results there is no any misleading instrument.
M1 and m2 denotes the Arellano Bond autocorrelation test for order 1 and order 2,
respectively. According to m1 and m2 test results, as expected, first order autocorrelation
exist at %5 statistical significance level and there is no second order autocorrelation at %5
significance level. Hansen Test results approved that there is no overidentification problem in
the model. t-2 of gmm instruments were used in the system GMM model. Roodman’s xtabond2
codes were used for estimation.
5. Summary and General Results
Labour productivity might be counted an engine for growth and it shows the innovation in the
economies. Labour productivity is a more sensible factor than capital productivity for
production of commodities and Solow (1957) evaluated labour productivity in the total factor
productivity framework for economies.
Labour productivity and economic growth relationship is investigated in this paper for OECD
countries between 2004 and 2012. Obtained results from system GMM and LSDV panel data
methods/regressions, there is a strong and positive linkage between labour productivity and

economic growth. Furthermore, the linkage is not in lagged form. Our findings approved that
labour productivity is an essential variable or factor for economic growth in OECD countries.
6. References
ARELLANO, M., and BOVER O., Another look at instrumental variables estimation of error-
component models, Journal of Econometrics, Vol. 68, 1995.
ASTERIOU, D. and HALL, S.G., Applied Econometrics: A Modern Approach, New York:
Palgrave Macmillian, 2007.
BALTAGI, B. H., Econometric Analysis of Panel Data, Wiltshire:John Wiley & Sons, 2005.
BLUNDELL, R., and BOND, S., Initial conditions and moment restrictions in dynamic panel-
data models, Journal of Econometrics Vol. 87, 1998.
BOULIER, B.L., What Lies Behind Verdoorn's Law?, Oxford Economic Papers, New Series,
Vol. 36, No. 2, Jun, 1984.
CORVERS, F., The Impact of Human Capital on Labour Productivity in Manufacturing
Sectors of the European Union, Applied Economics, Vol. 29, 1997.
DIXON, R., LIMB, G.C., A Univariate Model of Aggregate Labour Productivity, Applied
Economics, Vol. 44, 2012.
FILIPPETTĠ, A., PEYRACHE, A., Labour Productivity and Technology Gap in European
Regions: A Conditional Frontier Approac, Regional Studies, Jun, 2013.
HAKKALA, K., Corporate Restructuring and Labor Productivity Growth, Oxford University
Press on Behalf of Associazione ICC, Industrial and Corporate Change, Vol. 15, No.4, 2006.
JIANG, Y., An Empirical Study of Openness and Convergence in Labor Productivity in the
Chinese Provinces Econ Change Restruct, Vol. 45, 2012.
JORGENSON, D.W., HO, M.S., SAMUELS J.D., Information Technology and U.S.
Productivity Growth: Evidence from a Prototype Industry Production Account, J Prod Anal.,
Vol. 36, 2011.
KIVIET, J.F., On Bias, Inconsistency, and Efficiency of Various Estimators in Dynamic
Panel Data Models, Journal of Econometrics, Vol. 68, 1995.
LEVIN, A., LIN, C.F. and CHU,C.S.J., Unit Root Tests in Panel Data: Asymptotic and
Finite-Sample Properties,Journal of Econometrics, 2002.
MADDALA, G.S. and WU, S., A Comparative Study of Unit Root Tests with Panel Data and
New Simple Test, Oxford Bulletin of Economics and Statistics, Vol. 61, 1999.
MOHUN, S., Aggregate Capital Productivity in the US Economy, 1964–2001, Cambridge
Journal of Economics Vol. 33, 2009.

NICOLETTI, G., SCARPETTA, S., LANE, P.R., Regulation, Productivity and Growth:
OECD Evidence, Economic Policy, Vol. 18, No. 36, Apr., 2003.
NICOLINI, R., Labour productivity in Spain: 1977–2002, Applied Economics Vol. 43, No. 4,
2011.
POLIDORI, E., Interview with Jean-Claude Trichet. La Repubblica, Conducted on Jun.,
2010.
RADA, C., Stagnation or Transformation of a Dual Economy Through Endogenous
Productivity Growth, Cambridge Journal of Economics Vol 31, May., 2007.
SOLOW, R., Technical Change and the Aggregate Production Function. Review of
Economics and Statistics, Vol. 39, 1957.
TATOĞLU, F.Y., Ġleri Panel Veri Analizi, Beta Yayınları, Ġstanbul, 2012.
TAYLOR, A. J., Labour Productivity and Technological Innovation in the British Coal
Industry, 1850-1914, The Economic History Review, New Series, Vol. 14, No. 1 1961.
UPENDER, M., Elasticity of Labour Productivity in Indian Manufacturing Economic and
Political Weekly, Vol. 31, No. 21, May, 1996.
VERDOORN, P. J., Fattori che regoleno lo sviluppo della produttivita del lavaro, L'Industria;
translated by G. and A. P. Thirlwall, 1949.