Structural Change and Labor Productivity Growth in Indonesia
Lilis Siti Badriah, Armida S. Alisjahbana, Kodrat Wibowo and Ferry Hadiyanto
Faculty of Economics and Business, Padjadjaran University, West Java, Indonesia
Keywords: Structural change; Input reallocation; Labor productivity growth.
Abstract: Structural changes indicate reallocation of inputs from less productive sectors to more productive sectors.
Some literature indicate that the impact of structural change could become a structural bonus for productivity
growth. However, the others indicate that structural changes could become a structural burden. This study
aims to examine the impact of structural changes on labor productivity growth in Indonesia and the
determinants of that growth. The methods used are shift-share analysis and panel data regression. The data
consist of total output and sectorial labors as well as other macro data from 30 provinces of Indonesia during
2003-2014. The combination of both methods show a corresponding result that structural changes have a
weak impact on labor productivity growth in Indonesia. This result implies the need for support of more
relevant government policies by improving the quality of human resources, investment, infrastructure, and
maintaining macroeconomic stability to get more benefits from structural changes.
The data from the Central Bureau of Statistics (2013)
shows that there are changes in Indonesian economic
structure which is marked by the shifting of
dominance role of economic sector in GDP
formation, from agriculture sector to non-agriculture
sector, both from the output side and the labor side.
The interesting thing is the condition of the industry-
al sector. Despite the fact that the share of output
dominates in the formation of GDP, throughout 2004-
2013, the share shows a consistently decreasing trend
from 28.37% (2004) to 25.54% (2013). It shows that
the development in the industrial sector is stagnant
and even tends to decline. These conditions have an
impact on the performance of the Indonesian
economy, which can be seen, among others, from the
aggregate productivity indicators created. According
to data of Asian Productivity Organization (2014), the
growth of Indonesian TFP during 1970-2012 was
fluctuated with a downward trend. The average per
year was 0.9%. These conditions indicate that the
productivity which accompanies structural changes is
not sustainable. Whereas some literature reviews
indicate that manufacturing is “the engine of
economic growth” (McMillan, 2014; Kaldor in
1960s in UNIDO, 2011; and Ocampo, 2005).
The structural changes indicate the reallocation of
production inputs, from less productive sectors to
more productive sectors. The process of reallocating
these inputs can have a significant positive effect on
productivity growth that encourages the economic
growth overall (the Structural Bonus Hypothesis).
Timer and Szirmai (2000) argue that the shifting of
resources from the early industry to the middle and
the late industries illustrates the process of techno-
logical improvement and encourages bonus for
aggregate productivity growth in the manufacturing
sector. This is in line with McMillan (2014) that
any shift in resources from low productivity activities
to high productivity activities can result in structural
change bonus known as “growth-enhancing structural
change”. However, structural changes can also have
a weak impact and even a negative impact on the
growth of productivity (The Structural Bur-den
Hypothesis). Baumol’s Hypothesis of unbalanced
growth states that the difference between industries in
the opportunity to increase labor productivity (at a
certain level of demand) shifts the share of labor from
a “progressive” industry to a “stagnant” industry. In
the long-run, this condition tends to decrease the
prospect of per capita income growth (Baumol,
In 2003, Penender’s empirical study of industrial
structure and aggregate growth in OECD countries
Badriah, L., Alisjahbana, A., Wibowo, K. and Hadiyanto, F.
Structural Change and Labor Productivity Growth in Indonesia.
In Proceedings of the 2nd International Conference on Economic Education and Entrepreneurship (ICEEE 2017), pages 397-402
ISBN: 978-989-758-308-7
Copyright © 2017 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
during 1990-1998, with shift- share analysis and
panel data regression model, showed determinants
affecting per capita output and its growth, i.e. changes
in economic structure, demographic, business cycle
conditions, labor market rigidity, physical capital
investment, and human capital development. The
results showed that on average, structural change had
a weak impact on the growth of aggregate labor
productivity (robust structural burden).
Carree (2003) did comment on Fagerberg (2000)
research on technological progress, structural change,
and productivity growth using ISIC 5 industry data
from 20 OECD countries, during 1972-1992, divided
into 4 sub periods. The results of Caree's study
showed that changes in industrial employment share
did not have a significant impact on productivity
growth, industrial employment share at the beginning
of the period had a significant positive effect on
productivity growth in highly technologically
progressive industries, and initial productivity levels
have a significant negative effect on productivity
growth. It means that the inter-industries technology
convergence has been occurred.
Concerning productivity growth, Paus (2004)
studied the growth of productivity in Latin America
by observing factors affecting productivity growth,
i.e. technological change (short and long run),
domestic technological capabilities and conductive
social and economic environment influenced by
macroeconomic and political stability, access to
techno-logical know-how, requisite physical
infrastructure, and human capital development.
Another factor that can affect productivity growth is
wages. According to the efficiency-wage theory, high
wages can make workers more productive (Mankiw,
Based on the phenomenon of structural changes
occurring in Indonesia, which are associated with
some relevant empirical study results, this research is
addressed to answer how does the structural change
influence the growth of labor productivity in
Indonesia and what are the determinant factors of that
labor productivity growth?
Previous empirical studies tend to focus on the
relationship between structural change and
productivity growth in the manufacturing sector of
some countries, focusing on industry data, or just
looking at interrelationships from shifting between
economic sec-tors. This study, therefore, observed the
relationship of structural change and productivity
growth occur-ring in one country (Indonesia) in more
depth, with research objects using sectoral data from
each province to capture detailed characteristics of
behavioral change in the Seconomy of various
provinces. Hence, it is expected to provide
recommendation for development policy makers in
Indonesia to get more benefits of the structural
2.1 Data
This study used panel data of 30 provinces during
2003-2014. The data obtained from the Central
Bureau of Statistics, including real GDP at constant
2000 prices; the number of sectoral and national
workers; the proportion of economic sector
contribution to GRDP; the proportion of workers; and
the productivity of workers in various economic
sectors, investment, average length of school,
infrastructure, inflation rate, and wages.
2.2 Research Model
2.2.1 Shift-Share Analysis Model
This model can show structural bonus or structural
burden conditions in relation between structural
change and productivity growth (Peneder, 2003).
This research adopted shift-share decomposition
model used by Peneder (2003), in which the factors
affecting labor productivity growth were decomposed
into static-shift effect, dynamic-shift effect, and
within-shift effect, by the following formula:
Where LP = labor productivity; By = base year of
study; Fy = final year of study; T = sigma, whole
sector i; Si = share of sector i workforce in total em-
ployment; i = 9 economic sectors: (1) agriculture; (2)
mining and quarrying; (3) industry; (4) Electricity,
Gas and Water Supply; (5) Construction; (6) Trade,
Hotel, Restaurant; 7) Transportation and
Communication; (8) Financial, Real Estate, and
Business Service; (9) Services.
The first part of equation (2) is the static-shift
effect. If the static-shift effect is positive value, it
indicates a structural bonus. The second part of that
equation is the dynamic-shift effect. If the dynamic-
ICEEE 2017 - 2nd International Conference on Economic Education and Entrepreneurship
shift effect is negative value, it indicates a structural
burden. The third part of that equation is the within-
shift effect that shows growth of labor aggregate
productivity assuming no structural shifts during the
initial year.
2.2.2 Econometric Analysis Model
To analyze the determinant factors of productivity
growth, this study adopted the model of Carree (2003)
with the addition of some relevant control variables
from Peneder (2003), Paus (2004), and other
researchers. The study period was divided into 4 sub-
periods, each consisting of 3 years (M = 3): 2003-
2005, 2006-2008, 2009-2011, and 2012-2014. The
division into four sub-periods was intended to capture
intra-period variations and to increase the sensitivity
of changes in business cycle (Carree, 2003). Thus, the
model of this study is as follows:
Where Ln (Yi,t)/(Yi,t-M) = growth in labor
productivity; Yi,t-M = initial labor productivity; Xi,t
- Xi,t-M = changes in labor share of the industrial
sector; Xi,t-M = initial labor share of the industrial
sector; INVTt-1 = total investment in previous
period; ΔINVT = total investment change; HC =
development of human capital, with indicator of the
average number of years in education; INFST =
Infrastructure, with indicator of length of provincial
road; INFLS = Inflation rate; W = Wage rate; M =
number of years range in a sub-period (4); i = i (rank
of) Province.
3.1 Result of Shift-Share Analysis
The result of Shift-Share analysis of aggregate
economic sector can be seen in Table 1.
Based on Table 1, the growth of labor productivity
in Indonesia is influenced by static-shift effect,
dynamic- shift effect, and within-shift effect. In
aggregate, average labor productivity growth reached
0.8386. In line with previous research results
(Peneder, 2003, Fagerberg, 2000; Timmer and Szir-
mai, 2000; McMillan, 2014), the within- shift
effect still dominates the contribution of labor
productivity growth. This means, the reallocation of
labor between sectors has only a weak net impact on
overall productivity growth.
The total static-shift effect is positive at 0.3665.
This means that sectors with high productivity levels
are able to attract more labor resources, increasing the
shares of the sectors in total employment. While on
the other hand, dynamic-shift effect is negative at -
0.1002. This means that the economic sectors with
high labor productivity growth are unable to manage
its shares of labor in total employment. It caused a
decline in the shares of labor. This condition indicates
that the share of employment shifts from a
progressive economic sector to an economic sector
has lower labor productivity growth.
Table 1: Decomposition of aggregate productivity growth in Indonesia, during 2003-2014 period.
Economic Sector
Static shift
Dynamic shift
Within shift
mining and quarrying
Electricity, Gas and Water Supply
Trade, Hotel, Restaurant
Transportation and Communication
Financial, Real Estate, and Business
Source: Results of data processing (2017)
Structural Change and Labor Productivity Growth in Indonesia
Based on the decomposition of the economic
sectors, the value of static-shift effect is positive
while the dynamic-shift effect is negative. It can be
said that the structural changes in Indonesia shows
the tendency of structural burden, where labor
shifts from high productivity sector to low
productivity sector, although initially the sector
productivity level is high. Therefore, it can be said
that the structural bonus is relatively weak. This is
consistent with the conclusion of Peneder (2003).
3.2 Result of Econometric Analysis
The results of estimation model can be seen in
Table 2.
Based on the data in Table 2, the independent
variables in the equation model (1) only consist of
the main variables affecting the labor productivity
growth, whereas in the equation model (2), the
model (1) is expanded by 6 control variables.
Based on Table 2, in both equation models, initial
productivity variable and structural change
variables, i.e. changes in labor share of the
industrial sector and initial labor share of the
industrial sector, consistently show significant
negative values. The variable of initial labor
productivity with negative value significantly
indicates the existence of technological
convergence in the inter-provincial industrial
sector, affecting the growth of labor productivity of
the industrial sector in Indonesia. The results of
this study are in line with the research of Fagerberg
(2000) and Carree (2003).
Table 2: Estimation results of labor productivity growth (fixed effect model).
Labor productivity growth of industrial sector
Source: Data Processing, 2017
** = significant at α = 1%, * = significant at α = 10%.
The variable of industrial labor share change is
significant negative value in both models. This
means, the reallocation of labor moves towards the
industrial sector with lower productivity levels. The
variable of initial labor share is significant negative
value. This means, the reallocation of labor moves
towards the industrial sector with lower productivity
levels at the beginning of the period. Both variables
of structural change negatively affect the labor
productivity growth of industrial sector in Indonesia.
This indicates the occurrence of structural burden.
The results of these estimation are in line with the re-
search of Timmer and Szirmai (2000), Fagerberg
(2000), Peneder (2003), and Caree (2003) that
structural changes have a weak impact on improving
labor productivity growth. The estimation results are
also in line with the results of shift-share
decomposition analysis, which tends to prove the
ICEEE 2017 - 2nd International Conference on Economic Education and Entrepreneurship
structural burden hypothesis proposed by Baumol
The variable of total investment in the previous
period and the variable of total investment change are
positive values significantly. These means that in-
vestment/capital deepening in short and long run term
have a significant positive impact on labor
productivity growth in the industrial sector. This
results are in line with Peneder (2003).
The variable of average number of years in
education, as a proxy of human capital, is positive
significantly. That is, the workers with higher
education will increase the growth of labor
productivity in the industrial sector. The results of this
study are in line with Fagerberg (2000), Jorgenson
and Stiroh (2001), Peneder (2003), and Paus (2004).
The variable of infrastructure, with proxy of
provincial road length, is negative significantly at α =
10%. This can happen because, based on existing da-
ta, road conditions in each province are different and
still inadequate. This condition causes complicated
congestion and distribution process, thus less sup-
porting economic activity, among others, causing the
increase of transportation/distribution cost which can
reduce net result of output value. In addition, based
on data from The Global Competitiveness Report
2016-2017 (The World Economic Forum, 2016),
Indonesia’s infrastructure performance is still
relatively low. Of the 138 countries studied,
Indonesia is ranked 80th for the overall infrastructure
aspect, while in terms of road quality, Indonesia is
ranked 75th.
The variable of inflation rate. As a proxy of
macroeconomic stability, is positive significantly.
This means, an increase in inflation rate will increase
the labor productivity growth. This result is different
from some existing studies that inflation has a
negative effect on economic performance (Jaret and
Selody, 1982; Clark, 1982; Hondroyiannis and Pa-
papetrou, 1997; Bitros and Heat, 2001; Tsionas,
2003a; Christopoulos and Tsionas, 2005). Neverthe-
less, Mankiw (2007) states that in terms of supply
side, inflation reflects an increase in aggregate
demand. It will encourage the company to increase its
production capacity to make a profit, as shown in the
dynamics of the aggregate supply curve, where there
is a positive relationship between the price and the
quantity of output. Indeed, to some extent, with-in the
relatively low level of inflation (less than 10%), it is
needed in order to encourage the supply side
The variable of wage is insignificant positive
value. The wage rate indicator used in this study is
provincial minimum wage. The insignificant impact
of wage because the provincial minimum wage is
made as a reference for employers to determine
wages for their workers, so it is likely that many
employers who pay less than the provincial minimum
wage. The International Labor Organization (2015)
states that while it is the right of workers to receive
remuneration equal to the minimum wage, high levels
of vulnerability and informality in the labor market
and limited labor inspection capacity causing one-
third of workers receive wages less than the
provincial mini-mum wage. According to Mankiw
(2007), wage measurement should be based on total
compensation covering wages in cash and intangible
compensation (fringe benefit). In situations of
intangible compensation, such as pension funds and
health insurance being a major part of compensation,
wages in cash are generally not in line with
The results of this study indicate that structural
changes in Indonesia that lead to the increasing role
of industrial sector in the formation of national out-
put is not necessarily able to increase labor
productivity in the respected sector. The structural
changes that occur tend to be structural burden, which
unfavourable for the growth of labor productivity of
industrial sector.
The results of this study imply that structural
changes occurring along with the economic growth
should be supported by various other elements. They
are relevant government policies as efforts to improve
the quality of human capital and to provide better
infrastructure through an adequate development
budget allocation, maintaining the stability of the
macro-economy through appropriate controlling of
inflation rate Hence, it may create conducive
conditions to encourage of capital accumulation
through various investment activities.
Asian Productivity Organization. 2014. Productivity Data
book, Asian Productivity Organization. Tokyo: Keio
University Press Inc.
Central Bureau of Statistics. 2013. Statistik Indonesia.
Jakarta: BPS.
Baumol, William J. 1967. Macroeconomics of Unbalanced
Growth: The Anatomy of Urban Crisis. The American
Economic Review, (Vol. 57, No. 3), (Jun, 1967): 415-
Bitros, C. C. and Panas, E. E. 2001. Is there an inflation
productivity trade-off? Some evidence from the
manufacturing sector in Greece. Applied Economics,
33, 19619.
Structural Change and Labor Productivity Growth in Indonesia
Carree, M.A. 2003. Technological Progress, Structural
Change, and Productivity Growth: A Comment.
Structural Change and Economic Dynamics 14 (2003):
Christopolous, D. K. and Tsionas, E. G. 2005. Productivity
growth and inflation in Europe: Evidence from panel
cointegration tests. Empirical Economics (30):13750.
Clark, K. P.1982. Inflation and the productivity decline.
American Economic Review, Papers and Proceedings
(72): 149-54
Fagerberg, Jan. 2000. Technological progress, structural
change, and productivity growth: a comparative stud.
Structural Change and Economic Dynamics 11 (2000):
Hondroyiannis, G. & Papapetrou, E. 1997. Seasonality
cointe2wgration and the inflation, productivity and
wage growth relationship in Greece. Social Science
Journal (34): 23547.
ILO. 2015. Indonesia: Upah dan produktivitas untuk
pembangunan berkelanjutan. International Labour
Jarrett, J. P. & Selody, J. G. 1982. The productivity inflation
nexus in Canada 19631979. Review of Economics and
Statistics (64): 3617.
Mankiw, N. Gregory. 2007. Macroeconomics. 6
New York: Worth Publisher.
McMillan, Margaret, Dani Rodrik, Inigo Verduzco-Gallo.
2014. Globalization, Structural Change, and
Productivity Growth, With an Update of Africa. World
Development (63):11-32.
Ocampo, J.A. 2005. The Quest for Dynamic Efficiency:
Structural Dynamics and Economic Growth in
Developing Countries. In Beyond Reforms: Structural
Dynamics and Macroeconomic Vulnerability, ed.
Ocampo, J.A., Santiago: Economic Commission for
Latin America and the Caribbean.
Paus, Eva A. 2004. Productivity growth in Latin Amerika:
The Limit of Neoliberal Reform, World Development
(Vol. 32, No. 3): 427-445.
Peneder, Michael. 2003. Industrial Structure and Aggregate
Growth, Structural Change and Economic Dynamics
14 (2003): 427-448.
Stiroh, Kevin J. 2001. What Drives Productivity Growth?
FRBNY Economic Policy Review, March 2001.
The World Economic Forum. 2016. The Global
Competitiveness Report 2016-2017.
Timmer, Marcel P. & Adam Szirmai, 2000. Productivity
growth in Asian manufacturing the structural bonus
hypothesis examined. Structural Change and economic
dynamic 11 (2000): 371-392.
Tsionas, E. G. 2003a. Inflation and productivity in Europe:
an empirical investigation, Empirical, 30, 3962.
UNIDO. 2013. Industrial Development Report 2013.
Sustaining Employment Growth: The Role of
Manufacturing and Structural Change, Overview.
United Nations Industrial Development Organization.
ICEEE 2017 - 2nd International Conference on Economic Education and Entrepreneurship