J. Service Scie nce & Management, 2009, 3: 209-214
doi:10.4236/jssm.2009.23025 Published Online September 2009 (www.SciRP.org/journal/jssm)
The Relationship between Producer Service and Local
Manufacturing Industry: Empirical Evidence from
Shanghai
Shouming CHEN, Jie LI, Meijuan JIA
School of Economics and Management, Tongji University, Shanghai, China.
Received March 11th, 2009; revised May 13th, 2009; accepted July 17th, 2009.
ABSTRACT
Previous studies suggested that the development of producer service resulted from manufacturing outsourcing. On the
other hand, competitive producer service can promote the progress of manufacturing industry. Theoretically, correla-
tion may exist between them. According to co-integration theory, this paper empirically tests the time series data col-
lected from both producer service and manufacturing industry in Shanghai, indicating that co-integration does exist
between producer service and local manufacturing industry. Empirical results of this paper partially support the con-
clusions of prior studies addressed above.
Keywords: producer service, manufa cturing industry, co-integration testing, empirical analysis, shanghai
1. Introduction
Producer service is one of the modern service industries
that have been the forces promoting economic develop-
ment in developed countries. Progress in producer ser-
vice industry not only benefits itself but also signifi-
cantly facilitates advanced manufacturing industry. A
generally accepted view is that advanced manufacturing
industry could not ex ist in China without the presence of
advanced producer service. Subsequently, it is much
more difficult for manufacturing industry to maintain
comparable and competitive advantages. Therefore, it
seems reasonable to suppose that core cities such as
Shanghai should promote modern services mainly com-
posed of producer service to serve manufacturing indus-
try better in adjacent provinces and other parts of China.
Is this assumption rational? Additionally, whether pro-
ducer service in Shanghai really serves manufacturing
industry in adjacent provinces and other parts of China
need further and deeper investigation. This paper con-
ducts empirical analysis and tries to briefly answer these
questions.
2. Theories and Hypothesis
Previous literatures define producer service from two
perspectives. One definition focuses on objected ac-
cepted services. Greenfield mentions that producer ser-
vice mainly serves enterprises, non-profit organizations,
and governments, rather than offering products and ser-
vices to end users [1]. Harrington and Lombard state that
producer server offer intermediate services to en terprises.
[2] Stull and Madden suggest that producer service as-
sists enterprises or organizations to offer other products
or services, rather than providing services to private sec-
tors or family sectors [3]. Th e other definition focuses on
services that producer service can offer. This definition
is an extension of original concept. According to
Browning and Singleman, producer service is mainly
composed of knowledge-intensive industries, for in-
stance, finance, insurance, legal indu stry, and so forth [4].
Howells and Green state that producer service functions
as intermediate agency promoting operational efficiency
and bringing value to production and other services [5].
Daniels concludes several existing classifications of
producer service and suggests that producer service
could be classified to three categories. The first category
is information processing service involved of banking,
insurance, marketing, accounting, asset management,
advertising, and information editing skill. The second is
services related to business, such as sales, traffic man-
agement, installation and maintenance of basic equip-

SHOUMING CHEN, JIE LI, MEIJUAN JIA 210
ments, and maintenance and servicing for communica-
tion equipment. The third category refers to personal
support services, for example, welfare, food supply, in-
dividual travel, and residence [6].
Basically, producer service can be characterized by
following traits. The first trait is intermediate input, in-
dicating producer service does not serve to end users but
promotes intermediate consumption in order to create
more value. The second is relatively stronger industrial
correlation in agriculture, manufacturing, and other in-
dustries. Third, most services producer service provides
are technology-intensive or knowledge-intensive and
contain more stock of human resource. Forth, producer
service trends external, suggestin g that internal p rograms
continually depart from original enterprises and become
professional services providers that are independent to
original enterprises. Thus, kinds of producer service in-
crease quickly.
Many researches conducted before explain the reason
why producer service progresses so fast. Using in-
put-output methodology, Tschetter investigates postwar
producer service and tries to find why it progresses so
quickly. Among many reasons, that ways of enterprises
doing business changed is an important explanation.
Unbundling of production has a more significant impact
[7]. Coffey and Bailly analyze extemalization of pro-
ducer service, suggesting that six factor including inter-
nal technological limits and risk aversion lead manufac-
turing and services providers buy producer services from
outside [8]. According to transaction cost theory, some
scholars of New Institutional Economics analyze pro-
ducer services outsourcing [9,10], and subsequently state
that enterprise prefers outsourcing due to cost reduction
resulted from outsourcing and that external transaction
costs are less than internal costs due to impacts of
economy of scale. Meanwhile, outsourcing can effec-
tively reduce risk. Some researchers analyze producer
services from strategic perspective, suggesting that risk
and uncer tainty cou ld be r edu ced by o u tsour cing and that
through outsourcing enterprises can concentrate avail-
able resources on the most competitive phases in value
chain and then gain advantages [11]. Correlation be-
tween producer service and manufacturing industry is
another focus to which researchers pay much attention.
Though no consensus on this topic has reached, most
researchers agree that producer service promote manu-
facturing industry. Additionally, some researchers state
that essences for the development of manufacturing in-
dustry are external to itself [12] and that further
achievements of Chinese manufacturing industry could
not reached without advanced producer service.
Therefore, we can conclude: producer service is re-
sulted from outsourcing of manufacturing enterprises.
Subsequently, scale of manufacturing industry deter-
mines the scale of producer service to a certain extent.
Thus, we get the hypothesis of this analysis.
Hypothesis: Scale of producer service in Shanghai
positively relates to scale of local manufacturing industry
in Shanghai.
3. Method and Data
3.1 Data Collection
It is difficult to collect appropriate data from producer
service to conduct empirical analysis for a long time. In
early studies, researchers just simply combine the output
of sub-sectors in service industry and then use the com-
bination as the output of producer service; however,
there is no agreement on which subsector is appropriate
to be classified into producer service. A normal used
classification is composed of finance, insurance, business
services, and real property industry. Impertinency of this
classification can be found easily. First, it does not in-
clude services of other services industries offered to en-
terprises. Second, most services offered by these indus-
tries addressed above are consumed by end users, for
example, houses of real property industry and personal
financial services.
We estimate producer services output by subtracting
consumer and government services output from total
service sector output, using a technique first developed
by Grubel and Walker [13]. Tao and Wong used the
technique to estimate producer service in Hong Kong
[14]. Steps of using this method will address below. All
sorts of services industries can be classified into three
categories. The first offering service to enterprise be-
longs to producer service. The second offers service to
government and the third to end users. The amount of
service offered to end users can be estimated by the to tal
amount end users pay annually. We suppose that the
amount that end users outside Shanghai pay for service
offered by local providers is equal to the amount that end
users in Shanghai pay for service offered by ecdemic
providers. We combine service offered to government
into producer service because the amount of services
offered to governments is relative fewer. In other words,
we estimate the annual its value added or GDP of pro-
ducer service by deducting payout that Shanghai end
users pay for service they receive from the annual GDP
that all services industries create.
Data collected and used in this study are listed in Table 1.

SHOUMING CHEN, JIE LI, MEIJUAN JIA 211
Table 1. Value added of producer service and gross output value of manufacturing industry (1980-2004)
(Calculated at current prices)
Year Nominal value added of producer service Nominal gross output value of manufacturing in-
dustry
1980 60.40 560.75
1981 64.26 575.97
1982 68.44 590.39
1983 76.46 614.96
1984 90.56 678.62
1985 112.22 791.31
1986 124.17 871.06
1987 146.20 970.00
1988 173.67 1157.80
1989 184.45 1341.96
1990 221.03 1442.35
1991 283.50 1703.96
1992 369.85 2091.66
1993 520.98 2809.20
1994 710.71 3543.66
1995 905.87 4425.66
1996 1130.39 4835.10
1997 1412.99 6094.83
1998 1631.75 6151.95
1999 1821.70 6409.07
2000 2084.26 7467.26
2001 2261.86 8243.64
2002 2446.84 9088.32
2003 2676.27 11799.35
2004 3117.12 14759.49
Source: Author collected and rearranged from Shanghai Statistical Yearbook, 1980-2005.
Data of manufacturing industry in Table 1 are based
on the gross output value and those of producer service
are based on value added (GDP) due to the availability of
data collected from statistical yearbook. All the data are
calculated at current prices. Data of manufacturing in-
dustry changed from 1995 relatively significantly be-
cause statistical method changed in that year. Value
added tax is excluded from 1995. Thus we found two
types of data in 1996’s Shanghai Statistical Yearbook,
one including value added tax and the other excluding
value added tax. We compared these data and made
some modifications to the data after 1996. A national
economic survey was conducted in 2004, leading to sig-
nificant change in the data of services industry. In the
same way, we modified the data of producer service be-
fore 2004. Table 1 lists the data after modification.
3.2 Data Rearrangement
Based on data in Table 1, Figure 1 illustrates how value
added of producer service and gross output value of
manufacturing industry changed annually. We can see
clearly that trend of producer service resembles that of
manufacturing industry. In other words, both of the data
are non-stationary time series, respectively.
To control for the impact of price index on nominal
value, each value added of producer service of these 25
years was transformed to the value at 1980 constant
prices, so was the gross output value of manufacturing
industry. These transformation s were co mpleted by using
Consumer Price Index (CPI). Table 2 shows the results
of these transformations.
Figure 1. Producer service and manufacturing industry in
Shanghai (1980-2004)

SHOUMING CHEN, JIE LI, MEIJUAN JIA 212
Table 2. Value added of producer service and gross output value of manufacturing industry (1980-2004)
(Calculated at 1980 constant prices)
Year Actual value added of producer service
(PSt) Actual gross output value of manufacturing
industry (Mt)
1980 60.40 560.75
1981 63.42 568.52
1982 67.30 580.61
1983 75.06 603.67
1984 86.98 651.79
1985 93.58 659.84
1986 97.39 683.17
1987 106.10 703.96
1988 104.95 699.65
1989 96.14 699.44
1990 108.38 707.28
1991 125.80 756.14
1992 149.20 843.77
1993 174.86 942.87
1994 192.54 960.00
1995 206.74 1010.04
1996 236.24 1010.50
1997 287.27 1239.14
1998 331.75 1250.75
1999 364.89 1283.74
2000 407.33 1459.32
2001 442.04 1611.05
2002 476.44 1769.66
2003 519.89 2292.11
2004 592.71 2806.46
PSt refers to value added of producer service and Mt to
gross output value of manufacturing industry. We found
that each logarithm of the two time series is linear. Sub-
sequently, we used unit root test to figure out whether
the time series is stationary and the order of integration
of non-stationary series. Among commonly used unit
root test such as DF test, ADF test, and Philips
non-parameter test (PP test), ADF test based on residual
analysis and introduced by Engle and Granger was
adapted in this study. We used Eviews4.0 to conduct this
test and found that ln(PSt) and ln(Mt) are non-stationary.
However, ln(PSt) and ln(Mt) became stationary after first
order differencing, which means that ln(PSt) and ln(Mt)
are series of integration of Order 1, respectively.
3.3 Co-integration Testing
For two series of integration of order 1, we used the
Engle-Granger two-step approach to test the co-integra-
tion relationship .
First, we used ln(PSt) as the dependent variable and
established the regression equation below,
01
ln()ln( )
t
PSM u
after estima ti on we got Equation (2),
ˆ ln()5.67 1.58ln()
t
PSM u
tt

  (2)
t (-10.5) (20.06)
R2=0.95, Adjusted R2=0.94, D.W. =0. 35
Figures in parentheses under Equation (2) are corre-
sponding t values. Coefficient in Equation (2) shows that
when gross output value of manufacturing industry in-
creases 1%, value added of producer service increases
1.58%.
Second, we used unit root test to evaluate residuals of
Equation (2) an d got,
ˆln()1.58ln()5.67
tt t
uPS M

 (3)
Table 3. Results of unit root test to
t
u
ˆ
ADF test statistic -3.310353 1%
Critical value* -3.9635
5%
Critical value -3.0818
10%
Critical value -2.6829
*MacKinnon critical values for rejection of hypothesis of a
unit root.
tt

   (1)

SHOUMING CHEN, JIE LI, MEIJUAN JIA 213
We also used unit root test to test. According to AIC
and SC principle, the best lag order is 9. Test results are
showed in Table 3.
Test results in Table 3 suggest that rejects H0 at
the 5% significance, showing that no unit root exists.
Thus,
t
u
ˆ
t
u
ˆ is a stationary series, which means )0(
~
ˆ
I
u
t
.
Co-integration exists between ln(PSt) and ln(Mt), and
Co-integrated vector is (1, -1.58).
3.4 Result and Analysis
The empirical result shows a long-term relationship be-
tween producer service and local manufacturing industry
in Shanghai. Namely, it is possible that Shanghai’s pro-
ducer service interrelates with the development of local
manufacturing industry. Such empirical result is incon-
sistent with Shanghai’s future industry direction that
Shanghai as an important metropolis in China should
focus on the development of modern producer service,
not on the development of manufacturing industry. The
potential premise behind producer service oriented strat-
egy of Shanghai relies on th e footloose trait of advanced
producer service. In the other word, the advanced pro-
ducer service is not limit to the condition of factor of
production and to the adjacency o f market an d raw mate-
rial. However, the empirical result from Wernerheim and
Sharpe who study the footlooseness of producer service
in metropolitan Canada does not support the premise
[15]. They found that the advanced producer service,
such as Information, Communication and Technology
(ICT) sector, is no t as footloos e as expected [15]. Further,
the result of this paper is consistent with the result from
Wernerheim and Sharpe and shows that the producer
service is insignificantly footloose and that the producer
service has a long-term relationship with local manufac-
turing industry. Local manufacturing industry in Shang-
hai promotes the development of producer service in
Shanghai.
4. Conclusions
Based on data collected from Shanghai, this empirical
study with time series analysis shows that co-integration
do exist between producer service and local manufactur-
ing industry, partially supporting the hypothesis that
producer service interrelates with manufacturing industr y.
The conclusion is also meaningful for other cities or
provinces in china to formulate the strategy of producer
service development. Meanwhile, we recognize the limi-
tations in the application of our conclusion. First of all,
the conclusion of this study based on data collected from
Shanghai only represent the situation in a specific past
period not the future situation. For example, the rela-
tionship between producer service and manufacturing
industry may vary after the elimination of regional col-
laboration barrier. Second, we did not consider a rela-
tionship between the producer service in shanghai and
the manufacturing industry in Yangtze delta. If the rela-
tionship is positiv e, the development of producer service
in shanghai could also be explained by the growing scale
of manufacturing industry in Yangtze delta. This is our
future research issue.
The domain that producer service in Shanghai ser-
viced may cover a much larger area, from Yangtze delta
to the whole country, even the whole world. However,
we could not have a blind faith in the footloose trait of
advanced producer service. In the regional development,
the strategy that coordinates both developments of local
producer service and local manufacturing industry
should be persi st e d c onst antly.
REFERENCES
[1] H. L. Greenfield, “Manpower and the growth of producer
services,” Columbia University Press, New York, 1966.
[2] J. W. Harrington and J. R. Lombard, “Producer-service
firms in a declining manufacturing region,” Environment
and Planning A, No. 21, pp. 65-79, 1989.
[3] W. Stull and J. Madden, “Post-industrial Philadelphia:
Structural changes in the metropolitan economy Phila-
delphia,” University of Pennsylvania, 1990.
[4] H. C. Browning and J. Singelmann, “The emergence of a
service society,” National Technical Information Services,
Springfield, 1975.
[5] J. Howells and A. Green, “Location, technology and in-
dustrial organization in UK services,” Progress in Plan-
ning, Vol. 26, No. 2, pp. 83-184, 1986.
[6] P. W. Daniels, “Service industries in the world econ-
omy,” Blackwell Publishers, 1993.
[7] J. Tschetter, “Producer services industries: Why are they
growing so rapidly?” Monthly Labor Review, Vol. 110,
No. 12, pp. 31-40, 1987.
[8] C W. J. offey and A. S. Bailly, “Producer services and
flexible production: An exploratory analysis,” Growth
and Change, Vol. 22, No. 4, pp. 95-117, 1991.
[9] P. W. Daniels, “The geography of services,” Progress in
Human Geography 9, pp. 443-451, 1985.
[10] W. R. Goe, “Producer services, trade and the social divi-
sion of labor [J],” Regional Studies, No. 24, pp. 327-342,
1990.
[11] W. J. Coffey and J. J. McRae, “Services industries in

SHOUMING CHEN, JIE LI, MEIJUAN JIA 214
regional development,” Institute for Research on Public
Policy, Montreal, 1990.
[12] C. Q. Guo, “Evolution of manufacturing model and ten-
dency of manufacturing development, doctorate forum of
Anhui manufacturing development,” Hefei University of
Technology Publishing House (in Chinese), 2004.
[13] H. G. Grubel and M. A. Walker, “The Canadian service
industries [M],” The Fraser Institute, Vancouver, 1989.
[14] Z. G. Tao and Y. C. R. Wong, “Hong Kong: From an
industrialized city to a centre of manufacturing-related
service [J],” Urban Studies, Vol. 39, No. 12, pp. 2345–
2358, 2002.
[15] C. M. Wernerheim and C. A. Sharpe, “‘High order’ pro-
ducer services in metropolitan Canada: How footloose
arethey?” Regional Studies, No. 37, pp. 469–490, 2003.