Evangelista 2000

This article was downloaded by: [Universite De Paris 1]
On: 18 November 2012, At: 22:42

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK
Economics of Innovation and New Technology

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/gein20
Sectoral Patterns Of Technological Change In

Services
a
Rinaldo Evangelista
a
Institute for Studies on Scientific Research and Documentation, National Research
Council of Italy, Via C. De Lollis 12, Rome, 00185, Italy
Version of record first published: 28 Jul 2006.
To cite this article: Rinaldo Evangelista (2000): Sectoral Patterns Of Technological Change In Services, Economics of
Innovation and New Technology, 9:3, 183-222
To link to this article: http://dx.doi.org/10.1080/10438590000000008
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss, actions,
claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or
indirectly in connection with or arising out of the use of this material.
Econ Innov. New Tcchn.. 2000, Vol. 9, pp. 183-221 8 2000 OPA (Ovuwas Publishers Associalion) N.V.
Rcprina available directly fmm the publisher Published by license under
Pholaopying pennilled by liCenSC only the Hamood Academic Publishes imprint.
pan of the Gordon and Breach Publishing Gmup.
Rinted in Malaysia
SECTORAL PATTERNS OF
TECHNOLOGICAL CHANGE IN SERVICES
Downloaded by [Universite De Paris 1] at 22:42 18 November 2012
Institute for Studies on Scientific Research and Documentation, National Research

Council of Italy, Wa C. De Lollis 12, 00185 Rome, Italy
(Received March 15, 1999; In final fonn July 30, 1999)
This article provides a comprehensivepicture of the characteristics of innovation in services,

using the results of the 1993-95 Italian innovation survey in services. Technological change
does play a role in services: around one third of service finns have introduced a technological
innovation in the period 1993-95. Process innovation, innovative investment and the acquisi-
tion and internal development of software represent the most important channels through
which service firms innovate. R&D activities represent an important innovation source only
for a small number of science and technology-based service industries. The sectoral analysis
highlights the variety of innovative patterns, which should discourage any simple generaliza-
tion about innovation in services. Accordingly, a sectoral taxonomy is proposed in which
service industries are clustered according to the overall innovative performance of firms, the
nature of the innovation activities carried out, the different knowledge bases underlying the
innovation processes, and the different patterns of interaction through which service f m s
innovate.
JEL classification: Technological change (03); Innovation and Invention: Processes and
Incentives (031); Industry Studies: Services (General) (080)
Innovation in services has become a topic of increasing interest among

economists and scholars of technological change. There are several rea-
sons for the emergence of such a concern: first, services are nowadays a
* Correspondence author: Tel. +39-064l879260; Fax: +39-06-4463836, E-mail: evan-
geli @www.isrds.rmcnr.it
t The author is grateful to Aldo Del Santo and Giulio Perani of Istat for providing the data
used in this article and preparing the methodological appendix. I am also indebted to Maria
Savona for her help in the elaboration of data, preparation of the statistical material and
research assistance and to three anonymous referees for their very useful criticisms and sug-
gestions.
184 RINALDO EVANGELISTA
major component of economic systems accounting for almost two thirds of

jobs in most of the OECD countries (OECD, 1998). Second, it is widely
felt that technological change and the diffusion of information and com-
munication technologies (ICT) are key factors behind the growth of serv-
ices in the last decades (OECD, 1997; Motohashi, 1997): Third, services
are at the centre of the analysis of the growing knowledge and information
content of modem economies and societies (OECD, 1996b, 1996c, 1997).
The shift from tangible to intangible investment, the paradigmatic change
linked to the emergence of ICT, and the new emphasis given to human
resources, are .all assumed to have a particular relevance in services. This
is seen to be due to some "peculiar" characteristics of services identified
by the literature, such as their intangible nature, the high information con-
tent of their activities and the specific location of many services in modem
economies which make them natural interfaces and carriers of new tech-
nologies, knowledge diffusion and organizational change throughout the
economic system (Miles, 1993,1996; Hauknes, 1996).
We are however a long way from having a satisfactory picture of the
extent, role and nature of innovative activities in the service sector, In the
last two decades much of the theoretical and empirical literature on tech-
nology and innovation has focused on the manufacturing sector which has
been traditionally seen as the major producer and user of technology. With
respect to this sector we have learned a great deal about how innovation
activities take place, what factors induce firms to innovate, and how tech-
nological patterns vary across industries and f m s . On the contrary, the lit-
erature on innovation in services is still in a fluid state: concepts and
statistical methodologies have not yet found a proper systematization, and
data are scarce and often not comparable across countries and different
service branches. Most of the stylised "characteristics" pointed out by the
literature have mostly been drawn from specific sectoral studies which do
not give justice to the highly heterogeneous technological and economic
characteristics of services. Moreover, the traditional view which portrayed
services as "sheltered" sectors, characterised by low productivity and poor
technological performance, has been superseded by one emphasising the
high technological performance of sectors such as ICT services, telecom-
munications, or high value added business services such as those involved
in technical consultancy and transfer of know-how. So far there have been
* For an economic analysis of the paradigmatic change linked to the diffusion of ICT see
Antonelli, 1999.
CHANGE IN SERVICES 185
very few attempts to bring some order to this field and most of these have
clashed against the lack of comprehensive and reliable data.*
The purpose of this article is to establish a more balanced and empiri-
cally grounded picture of the nature of innovation activities in the service
sector. We will use data provided by the innovation survey carried out in
Italy in 1997 by the National Statistical Office (ISTAT) in collaboration
with the Institute for Studies on Scientific Research and Documentation of
the National Research Council. This survey represents one the first
large-scale statistical attempt to collect systematic information on innova-

tion activities in the service sector on the basis of the guidelines indicated
in the revised version of the OECD "Oslo Manual" (OECD-EUROSTAT,
1997).~In particular, the richness of the Italian data-base resides in the size
and representativeness of the sample consisting of 6,005 firms drawn from
a universe of 19,300 market service companies with more than 20 employ-
ees. The results of the survey will be used in this article to:
i) outline a broad picture of firms' innovation strategies and perforrn-
ances in the service sector;
ii) highlight major similarities and differences with technological
change in manufacturing;
iii) identify the main patterns of innovation in services through a secto-
ral taxonomy.
The empirical analysis will be carried out by examining type of innova-
tion introduced (service/process), type of innovation inputs (R&D, design,
software, training, investment in machinery and equipment, marketing),
objectives of innovation (efficiency, quality enhancing), and the sources of
information used for innovating.
The remaining part of this article consists of four sections. Section 2
addresses some conceptual and methodological issues in the measurement
of innovation in the service sector and clarifies the methodological
approach adopted by the OECD Oslo Manual and by the Italian innovation
survey. Section 3 looks at the main characteristics of the innovation proc-
ess in the service sector as a whole comparing the data, when appropriate,
* It is enough to say that the pioneering contribution of Soete and Miozzo (1989) which
has attempted to draw a first technological categorization of different services branches has
remained since then unpublished and has not find any proper empirical validation; see also
Hauknes (1996) for a review of the theoretical and empirical literature on innovation in serv-
ices.
t The methodology of this s w e y is described in the appendix.
with the results of the 1995 innovation survey on manufacturing (ISTAT,

1995).* Section 4 explores the variety of innovation patterns in services
and proposes a sectoral taxonomy. Section 5 summarises the empirical
findings of the article.
I1 MEASURING INNOVATION ACTIVITIES ]IN SERVICES:

CONCEPTUAL ISSUES AND METHODOLOGY
The necessity for systematic data collection on innovation activities in the

service sector is nowadays widely recognised. As far as R&D statistics are
concerned, some steps have been made in terms of actual coverage and
collection of data. As a result more accurate figures on the contribution
provided by the service sector in most OECD countries in the process of
generation of new technological knowledge are now becoming available.
These figures show that service industries nowadays perform up to one
third of total business R&D (Young, 1996; OECD, 1996a, 1998)~
However, innovation activities are a much broader and complex phe-
nomenon than R&D (Kline and Rosenberg, 1986). In recent years innova-
tion surveys have increasingly been recognised as a useful tool to provide
information on innovation activities compared to traditional technological
indicators such as R&D, patents and innovation counts. Concepts, defini-
tions and methodologies used for the collection of data on innovation in
the manufacturing sector were first set out in 1989 by the first edition of
the OECD "Oslo Manual" (OECD, 1989). The underlying model of inno-
vation adopted by the manual was the "chain-linked" model. The latter
envisages that R&D is only one of the driving forces of the innovation
process, and that innovation is a multifaceted phenomenon which takes
place within the "national systems of innovation" which include f m s ,
government laboratories, regulatory agencies, universities, funding organi-
sations, the government (Kline and Rosenberg, 1986). Since the publica-
tion of the Oslo Manual in 1989, innovation surveys have been carried out
in many OECD and non-OECD countries, providing systematic empirical
* Section 3 summarizes the results presented in an early work by Siriiii and Evangelista
(1998).
t It is still not clear, however, how much of this increase is due to the changes in the cate-
gorisation of the same firms from manufacturing to services and to the coverage of R&D sta-
tistics, and the extent to which it is due to a real increase of the R&D innovative efforts of
service fums.
CHANGE IN SERVICES
evidence on the varied nature of innovation activities in industry, and more

in particular on the relative importance of product and process innovation,
on the type of innovation inputs used, on the objectives pursued with inno-
vation, on the obstacles faced by f m s in introducing innovations, on the
sources of information used for innovating, and the economic impact of
technological change.
Both the 1989 edition of the Oslo Manual and the first round of innovation
surveys originated from it were tailored to measure innovation activities in
the manufacturing sector. The necessity of extending the sectoral coverage

of innovation surveys in order to include the service sector has since then
been felt more and more strongly by national statistical offices, international
organizations such as EUROSTAT and the OECD, and scholars of techno-
logical change. However, the question of whether, and to what extent, the
methodological and conceptual framework developed with reference to
manufacturing activities could be used for analysing and measuring innova-
tion activities in the service sector emerged very soon as a critical one
(Evangelists and Sirilli, 1995). This is because of some peculiar features of
service activities identified in the literature which are supposed to bear
important implications for the conceptualization and measurement of inno-
vation. Without having the ambition to be exhaustive, those more commonly
recalled in the literature can be summarized in the following:
For most services sectors the production and consumption phases
occur at the same time and cannot be distinguished both on conceptual
grounds and in practice. It is argued therefore that such co-terminality
between pmduction and consumption makes the distinction between
product and process innovations (much used in manufacturing) less
clear-cut, or even meaningless, in services (Gallouj and Weinstein,
1997). Miles (1993, 1996) has introduced a new category of innova-
tions, i.e. delivery innovation, in order to take into account the "deliv-
ery nature" of many service activities.
In a large part of the service sector, economic activities consists of the
customisation of specific technologies, pieces of equipment, organiza-
tional models and strategies, to answer to a wide range of users' needs.
It is often argued that the processes of customization should be
regarded as "innovative" as long as they consist of "problem-solving"
activities, requiring creative work, high qualifications, specific compe-
tencies and the combination of different kinds of know-how.
The intangible nature and the information-based characteristics of

services give to the generation and use of ICT a central role in firms'
innovation activities and their performances. On the one hand, the use
of ICT has represented a major driving force which has led most serv-
ice sectors to close their "productivity gap" vis-i-vis the manufacturing
industry (OECD, 1996~);on the other, ICT form a technological plat-
form upon which new services can be designed and traded @arras,
1986).* In synthesis, the centrality of ICT in most service industries
implies that data on the generation and diffusion of information tech-

nologies should be collected on a systematic basis, also through inno-
vation surveys.+
To the intangible nature and high information content of service activi-
ties is also likely to be associated a limited appropriability of the
results of innovative activities. This is because the former characteris-
tics give to some services the attxibutes of public goods (which are
non-rival and non-excludable). Appropriability problems can also orig-
inate form the close user-producer interactions characterizing the pro-
vision of most knowledge-based services (i.e. technical consultancy).
Ownership rights might be difficult to be established in the case of
co-produced innovations, and "appropriation conflicts" could represent
a disincentive for undertaking innovation activities.
The predominant role played by human factors in the organisation and
delivery of most services is associated with substantial investment in
human resources.* Despite the fact that training activities are not usu-
ally considered as innovative inputs, they should be explicitly consid-
ered as one of the main channels for upgrading the technological
capabilities and competencies of firms in the service sector.
The much emphasised importance of organisational factors in the
service sector suggests that we need to enlarge the concept of innova-
* The role of ICT as "enabling technologies" is at the basis of the "reverse product cycle"
model proposed by Barras (1986) for representing the dynamics of the innovation process in
services.
t Specific questions on the importance played by ICT have been included in the Gennan
questionnaire. The results of the German innovation survey in services have shown that com-
puters, EDP, other hardware and software are regarded by service finns as the most important
types of technology (Licht et al., 1997).
$ In this respect a recent Italian survey carried out by ISTAT has shown that firms' training
expenditure (in relation to total wage costs) are higher in sectors such as Insurance, Banking,
Telecommunicationsand Software than in manufacturing industries such as Chemicals and
Pharmaceuticals,Motor-vehicles and Machiney @el Santo and Forlani, 1995).
tion in order to include organisational changes which can either be

linked to, or be independent from, the introduction of technological
innovations.*
Non-technological types of knowledge, know-how and capabilities (i.e.
those which do not have an ultimate scientific-engineering base) might
also be important in explaining firms' performances, and they may rep-
resent an important part of f m s ' competitive strategies. Knowledge
about markets, consumer habits and tastes as well as institutions, might
be a crucial strategic asset in services. Gallouj and Weinstein (1997)

suggest to adopt a rather broad approach to innovation in services, that
is one encompassing "...not only technologies in the narrow sense of
the term and the competencies relating to those technologies, but also
the 'technologies' specific to services (legal, financial, commercial,
etc.) and the competencies corresponding to them" (Gallouj and Wein-
stein, 1997, p. 554).?$
Despite providing a large number of suggestive insights, the ever grow-

ing literature on services is still far from being conclusive regarding the
distinctive features of innovation in this sector. Furthermore, any attempt
of identifying a list of peculiarities of service activities clashes against two
stylized facts which are also stressed in the literature: i) the very heteroge-
neous nature of service activities which makes difficult any simple gener-
alization about the nature of innovation in services ii) the on-going process
of convergence between manufacturing and service sectors in terms of
type of inputs used, organizational models, modalities of serving the mar-
* This is the approach chosen in the German innovation survey in the service sector (Licht
et al., 1997).
t In particular, Gallouj and Weinstein (1997) suggest to analyse the innovation process in
services taking into account different set of characteristics defining each service (or even a
good): i) the "tangible" technical characteristics of the service (i.e. IT, logistical technologies
and other embodied technologies used in the delivery andlor consumption of the service) ;ii)
the "intangible" technical characteristics of services (legal or financial expertise, economic
and financial modelling, operational research methods embodied in the service) iii) the set of
competencies which need to be directly mobilized for the provision and use of the service iv)
the "final" characteristics of the services as they are perceived by the consumer. The need of
adopting a broad, not strictly technological, view on innovation in services is also suggested
by Hauknes (1996) and Sundbo (1997).
$ Miles (1995) identifies additional specificities of the products, processes, organisations
and markets of services, namely low levels of capital equipment, non continuous nature of
production processes, key role played by the process of delivery of the services, limited role
played by economies of scale, high regulated regimes of markets and products. The features
identified by Miles contain, however, a high degree of generalisation which contrasts with the
heterogeneous nature of the service sector.
190 IUNALDO EVANGELISTA
kets, all features which make more and more meaningless the traditional
manufacturing/service divide (Quinn, 1992; Hauknes, 1996). In the light
of this, it is not surprising that in the revised version of the Oslo Manual
(OECD-EUROSTAT, 1997) it was decided to adopt a unified framework
for analyzing and measuring innovation activities in the manufacturing
and service sectors.
The revised version of the Oslo manual is still focused on technological
innovation. An effort has nonetheless been made in order to accommodate
some of the peculiarities of innovation in services stressed in the literature.

In particular, technological product and process innovation have been con-
ceptualized and defined in a very broad sense. The discriminating criteria
for identifying a technological innovation is in fact based on "its degree of
novelty" andlor on "its objective performances". Furthermore, a sufficient
condition for identifying a technological innovation is that it should
involve investments in (or require the use of) new knowledge. The adop-
tion and diffusion of new technologies is also considered as an integral
part of the innovative process, a sufficient condition being that the product
or process innovations introduced are "new for the firm" (OECD-EURO-
STAT, 1997, p. 18-19).* What are ruled out as technological innovations
are "...changes in products which provide largely subjective improved
customer satisfaction based on personal taste and aesthetic judgement,
andlor derived from following fashions, and or brought about largely by
marketing" (OECD-EUROSTAT, 1997, p. 18). Pure organizational
changes are also not covered by the Oslo Manual. The reason resides on
* The 1997 version of the Oslo Manual provides the following definitions of technological
product and prccess innovations:
"A technologically new product is a product whose technological characteristics or intended
uses differ significantly from those of previously produced products. Such innovations can
involve radically new technologies, can be based on combining existing technologies in new
uses, or can be derived from the use of new knowledge."
"A technologically improved product is an existing product whose performance has been sig-
nificantly enhanced or upgraded. A simple product may be improved (in terms of better per-
formances or lower cost) through use of higher-performance components or materials, or a
complex product which consists of a number of integrated technical sub systems may be
improved by partial changes to one of the sub-system".
"Technological process innovation is the adoption of technologically new or significantly
improved production methods, including methods of product delivery. These methods may
involve changes in equipment, or production organization, or a combination of these changes,
and may be derived from the use of new knowledge. The methods my be intended to produce
or deliver technologically new or improved products, which cannot be produced or delivered
using conventional production methods, or essentially to increase the production or delivery
efficiency of existing products" (OECD-EUROSTAT, 1997, p. 48-49).
the fact that "...organization is a crucial dimension of innovation, but its

measurement appears to be very difficult both conceptually and in prac-
tice", and that "...organizational change is highly firm-specific, making it
still more difficult to summarize in aggregate, sector or economy-wide sta-
tistics" (OECD-EUROSTAT, 1997, p. 43). Processes of customization as
well are not regarded as technological innovation unless the one-off items
display significantly different attributes to products or services that the
firm has previously made or delivered (OECD-EUROSTAT, 1997, p. 56).
As far as innovation activities are concerned the 1997 version of the

Oslo Manual also adopts a broad perspective, investment in knowledge
becoming the qualifying criteria for drawing the boundaries of innovation
activities.* The manual suggest to collect data on firms' expenditures related
to activities as diverse as R&D, design, the acquisition of patents, licences
and know-how, investment in new machinery and equipment, marketing,
training activities and the acquisition and development of software.
The Italian innovation survey in the service sector follows the general
approach and the definitions contained in the 1997 version of the OECD
"Oslo ~ a n u a l " . It~ focuses therefore on technological innovation (in the
broad sense defined above), while activities of customisation and organi-
sational changes have not been explicitly taken into account. The question-
naire used in the Italian innovation survey has however been designed in
order to shed some light on some of the most debated issues in the litera-
ture. In particular, specific questions have been included on the usefulness
of maintaining the distinction between product and process innovation, on
the relative importance of technology v i s - h i s other factors in explaining
the economic performance of firms, and on the impact of innovation on
employment. Finally, an attempt has been made to keep the overall struc-
ture of the questionnaire used in the service sector as close as possible to
the one used in the previous survey camed out in the manufacturing sector.
* According to the 1997 version of the Oslo Manual, '"Technological product and process
innovation activities are all those scientific, technological, organizational,financial and com-
mercial steps, including investment in new knowledge, which actually, or are intended to,
lead to the implementation of technologically new or improved products or processes. Some
may be innovative in their own right. others are not novel but are necessary for implementa-
tion" (OECD-EUROSTAT, 1997, p. 58).
t The definitions of product and process technological innovations adopted in the Italian
questionnaire are based on the OECD Oslo Manual and are reported in the methodological
appendix. There are only few marginal differences between the definitions used in the Italian
survey on services and those contained in the "Oslo Manual". This is because when the sur-
vey questionnaire was prepared the 1997 version of the Manual was not yet available.
192 RINALW EVANGELISTA
This in order to leave some room to compare the two data-set (see the
methodological appendix for more details).
111AN OVERVIEW OF INNOVATION ACTIVITIES IN SERVICES
The presence sf innovation activities in services

One of the first aims of innovation surveys is to establish how widespread

innovation is 'within the industrial structure. This information allows us
starting to shed some light on the actual importance of innovation in serv-
ices, and the extent to which services are backward vis-8-vis the manufac-
turing sector. The diffusion of innovation activities in services can be
estimated by looking at Table I which shows for the service sector as a
whole, and for the main firm size classes, the number of firms covered by
the survey and the percentage of these which introduced at least one tech-
nological innovation in the period 1993-95.*
Table I shows that a little less than one third of service firms (3 1.1%)
introduced technological innovations over the period 1993-1995. This fig-
ure is quite similar to that of manufacturing: 33.3% of Italian manufactur-
ing firms introduced product or process innovations in the period 1990-
1992 (Evangeiista et al., 1996). Table I also shows a positive relationship
between the share of innovating firms in services and their size: the per-
centage increases from 25.9% for firms with between 20 and 49 employ-
ees to 79.0% for firms with more than 1000 employees. Again this pattern
is quite similar to that found for the manufacturing sector. It may be
explained, at least in part, by the scale of production and service activities:
large firms are more likely to report that they have introduced innovations
as they typically offer a broader range of products and services and lines of
business.
Table I1 shows significant differences in the percentage of innovating
f m s across service sectors. In R&D, Banking, Insurance and Other finan-
cial services the percentage is 60% or over; in Retail trade, Hotels and
Restaurants, Cleaning and Security innovative firms are between one tenth
* The figures on services presented in this article are those obtained after applying the
grossing-up factors to the original data-set. This has been done in order to get estimated fig-
ures for the total universe of 19,300 Italian services finns (see the methodological appendix
for more details).
and one fifth of the total. A surprisingly low percentage of innovating

firms was found in the Post and Telecommunication sector. This is due to
the presence in this sector of a couple of very large innovating firms, sur-
rounded by a large group of smaller and less innovative firms.
The evidence presented in Tables I and I1 confirms that industrial sector
and fm size are important factors for determining the presence of innova-
tion activities in services (as in manufacturing industry) and in particular
that only a minority of small firms are involved in any innovative activity.
These data are also consistent with the results of the innovation surveys in
services canied out in other countries such as Australia and The Nether-
lands which have shown that the number of innovating firms varies
between one third and one fourth of the total, and that the percentage of
innovating f m s varies widely across industries and increases with firm
size (Gault and Pattinson, 1995; Brouwer and Kleinknecht, 1995).*
TABLE I Innovating and non-innovating firms in services and manufacturing by firm size
-- -
Servicefirms (1993-95) Manufacturingfirms (1990-92)

Classes
of employees Total firms (*) % Innovatingfim Total firms % Innovating
on total firms firms on t o t a l j i m
20-49 13934 25.9 15109 25.9
50-99 2844 38.6 4142 40.8
100-199 1269 43.4 2012 48.0
200499 822 50.9 1041 58.5
500-999 238 65.1 292 74.0
1000 and over 195 79.0 191 84.3
Total 19302 31.0 22787 33.1
TABLE I1 Innovating and non-innovating service firms by industry
Toralfimrr % Innovatingfim
Sentice sectors
on totalfirms
Trade and repair of motorvehicles 1319 28.0
Wholesale trade (excl. rnotorvehicles) 4469 35.0
Retail trade 2522 20.1
* An innovation survey carried out in Germany, covering the period 1993-95, has shown a
rather high percentage of innovating finns in the service sector (more than 60%).However,
this high figure is affected by the very broad definition of innovation adopted in the German
questionnaire, which also included organizational changes (Licht et al., 1997).
-- -
To,alfinns * % Innovatingfirms
Service sectors
on totalfinns
Hotel and Restaurants 2186 19.6
Land transportation 1510 29.8
Shipping and Sea transportation 117 33.3
Travel and Transport services 1201 29.0
Post and Telecommunication 55 10.9

Banking 917 61.8
Insurance 157 63.1
Other financial services 163 61.3
Computing and Software 972 54.3
Legal, Accounting
Engineering
Technical consuttancy 105 41.9
Advertising
Security
Cleaning 1220 10.5
Other business services

Waste disposal
Total 19024 31.3
*. The two totals are diffe'erent due to the exclusion of four service sectors with a low number
of answeringfirms
Product and process innovation in services

The product/process innovation mix is deemed to be an important dimen-
sion of tirms' innovation strategies and an aspect investigated by innova-
tion surveys. While product innovations are usually associated with the
creation of new markets or to the quality enhancement of existing prod-
ucts, process innovations are typically introduced for reducing costs,
rationalising or increasing the flexibility and performance of production
processes (Sirnonetti et al, 1995). However, in services the applicability of
the distinction between product and process innovation is more problem-

atic, due to the co-terminality between production and consumption which
characterises most service activities (Miles, 1993, 1996; Gallouj and
Weinstein, 1997). There is, however, little quantitative and systematic evi-
dence on the actual difficulty faced by service f m s in separating product
and process innovation. In order to shed some light on this issue f m s
were asked to specify the type of innovation introduced (distinguishing
between product and process/delivery innovation), and to indicate whether
they had introduced innovations for which the distinction between product
and process innovation was not applicable.
Table 111 shows that only one fourth (25.6%) of innovating iirms in the
service sector were not able (in any case) to distinguish between product
and process-delivery innovations.* For the remaining innovating f m s
(i.e. those able to apply the product/process distinction) the introduction of
new or improved ways of producing and delivering services represents by
far the most common type of innovation, with 72.6% of the firms indicat-
ing that they have introduced process innovation. Process innovation is
introduced by almost all firms in service sectors as diverse as R&D
(94.1%), Insurance (91%), Banking (83.7%) and Technical consultancy
(85.7%). Product (service) innovations have been introduced by 53.4% of
innovating f m s in services.
In the manufacturing sector, the most diffused typology is also process
innovation, which has been introduced by 83% of the firms. A first impor-
tant difference between the service and the manufacturing sector is that,
while the most innovative industries in the manufacturing sector show a
clear orientation towards product innovation, an opposite pattern character-
ises the service sector: the introduction of new processes or delivery systems
represents a fundamental component of firms' strategies in services. This
holds up in the case of the most innovative service sectors such as R&D,
Engineering and Technical consultancy (Evangelists and Savona, 1998).+
A second important difference between the service and the manufactur-
ing sector is that while in the manufacturing sector most firms introduce
* 8% of the service innovating firms have introduced a product andlor a process innova-
tion and at the same time have indicated to find difficult to apply such a distinction for some
of the innovations introduced.
t A prevalence of process innovations in the service sector has also been found in Ger-
many (Licht et al., 1997). while product innovations have been found to be the most frequent
type of innovation introduced by service firms in The Netherlands (Brouwer and
Kleinknecht, 1995).
both type of innovations in the service sector only one fourth of the firms
has innovated introducing new services, production processes or new
delivery systems. In other words, surprisingly enough, compared to the
manufacturing sector, in most service industries there is a much more neat
demarcation between product and process oriented innovation strategies.
TABLE Ill Innovating service firms by type of innovation introduced
%offirms %offim
%o f f i m
% offirms introducing declaring
intmducing introducing
Service sectors pmcess and the distinction
product pmcess
product is not
innovation innovation
innovation applicable
R&D
Post and Telecommunication
Advertising
Security
Cleaning
Shipping and Sea transportation
Banking
Engineering
Computing and Software
Land transportation
Legal, Accounting
Hotel and Restaurants
Retail trade
Wholesale trade (excl. motorv.)
Trade and repair of motorvehicles
Technical consultancy
Waste disposal
Insurance
Travel and Transport services
Other business services
Other financial services
Total services
Total manufacturing
The nature of innovation activities in sewices

The varied nature of innovative activities and their sectoral specificity
have been underlined by a vast literature, although almost exclusively with
reference to the manufacturing industry. Besides activities generating new
technological knowledge such as R&D, special attention has also been
attached to processes of technology adoption and diffusion (OECD, 1992;
Evangelista, 1999). As already pointed out the "Oslo Manual" adopts this
more comprehensive perspective on innovation activities, and identifies
for the service sector various innovative activities: R&D, design, the
acquisition of know-how, the acquisition and internal development of new
software, training and marketing activities necessary to introduce innova-
tions and the purchasing of technologically new equipment and machinery.
Data provided by innovation surveys allow us to assess the relative impor-
tance of the various types of innovative sources by looking at the expendi-
tures sustained by f m s to carry out such activities.*
Figure 1 shows the breakdown of expenditure incurred by Italian service
f m s in 1995 to introduce innovations. The picture which emerges from
Figure 1 is rather clear-cut. Innovative activities in the service sector con-
sist, first and foremost, of the purchase and use of "embodied" technolo-
gies (innovative machinery and plants). These account for 46% of total
expenditure on innovation. R&D activities absorb 23,7% of total innova-
tion costs, and software activities represent 14.1% of total innovation
expenditure. The other components of innovation expenditure (design,
acquisition of know-how, training and marketing) play a relatively minor
role, constituting together 16.1% of total innovation expenditure.+.
Figure 2 shows the distribution of innovation expenditure for the Italian
manufacturing firms. Although the lists of innovation items used in the man-
ufacturing and the service questionnaires are not exactly the same, a clear
similarity in the overall structure of innovation expenditure can be identi-
fied. Investment and R&D represent in both sectors more than two thirds of
total innovation expenditures (Evangelista et al., 1996; Evangelista, 1999).
* A more detailed description of the relative importance of the different innovative activi-
ties carried out in the service sector is provided in Sirilli and Evangelista (1998). and Evange-
lists and Savona (1998). The defmitions of the different innovative activities used in the
questionnaire are reported in the methodological appendix.
t Also in Germany, investment has emerged as the most important innovation expenditure
category in the service sector (Licht et al., 1997).
RINALDO EVANGELISTA
Innovative
investment
46.0%
Marketing Tmining
1.9% 2.6%
FIGURE 1 Breakdown of innovation costs in the service sector (1995)
innovative
investment /
FIGURE 2 Breakdown of innovation costs in the manufacturing sector (1992)

Table IV allows us to look at how the relative importance of the different

sources of innovation varies across service industries. Excluding the
anomalous cases of R&D and Engineering service industries, the service
sectors for which R&D plays an important role are Post and Telecommuni-
cations (25.4%) and Computer and Software (18.3%). Industries which
devote substantial shares of their innovation expenditure to software are
Advertising (62.5%), Other financial services (39.2%), Banking (36.5%),
Trade and Repair of motor-vehicles (39.1%) and Insurance (29.0%).
Industries innovating mainly through the acquisition of new machinery

and equipment are .Land transportation (93.4%), Shipping and Sea trans-
portation (89.2%), Waste disposal (89.0%) and Other business services
(73.0%). Overall, investment represents a large chunk of innovation
expenditure across most industries. It should also be noted that the rela-
tively large share of innovation costs devoted to R&D in services is heav-
ily affected by the R&D sector, which concentrates more than 50% of the
total R&D expenditures of the firms surveyed. If the R&D service f m s
are excluded, the R&D component shrinks from 23.7% to 13.7% of total
innovation costs, while the investment component increases its share from
46.0% to 52.6%. In other words, if R&D services are excluded, the techno-
logical dependence of the services sector from the acquisition of ICT
related new equipment becomes more evident.
An indicator of innovation intensity, the total innovation expenditure per
employee by sector, is also reported in the last column of Table IV. The
data show that market service firms spent in 1995 on average 2,315 Ecu
per employee, while in 1992 the equivalent figure for the manufacturing
sector was more than three times higher (7,831 Ecu). Three service sectors
- Engineering, Technical consultancy and Computing and software -
spend on innovation an amount comparable with, or even above, the man-
ufacturing average (between 5,000 and 10,000 Ecu per employee). Among
the sectors which spend less on innovation are Security and Cleaning,
Retail trade and Hotels and Restaurants. Once again, if the R&D sector is
excluded, the technological gap between the services and the manufactur-
ing sector widens substantially.
TABLE IV Breakdown of innovation costs and innovative intensity by industry N
Service sectors Distribution of i ~ o v a t i o ncosts (5% values) Innovation costs

per employee
R&D Design Know-how Software iVaining Marketing Investment Totals (Ecu, 1995)
R&D 83.9 6.7 0.7 1.O 0.9 0.2 6.5 100 49632
Engineering 70.6 5.5 3.0 5.8 2.5 0.3 12.3 100 10003
Technical consultancy 53.8 1.8 0.2 6.1 4.6 7.0 26.5 100 7398
Computing and Software 18.3 32.9 9.1 12.8 4.5 1.8 20.6 100 5201
Other financial services 0.6 10.7 0.3 39.2 2.2 0.3 46.7 100 4740 5
Advertising 5.4 8.2 1.9 62.5 1.7 7.0 13.3 100 3577 k
0
Wholesale trade (excl. motorvehicles) 4.5 3.7 2.1 18.4 3.3 2.3 65.8 100 3026
Insurance
Waste disposal
9.1
0.0
18.5
5.8
6.4
0.0
29.0
1.5
5.2
2.2
1.1
1.4
30.7
89.0
100
100
2630
2576
Z
Legal, Accounting 10.1 21.6 5.6 18.6 13.7 2.0 28.3 100 2340 E9
Land transportation 1.5 1.1 0.5 2.1 0. 5 0.5 93.8 100 2244
Other business services 0.0 5.2 4.1 15.7 0.7 1.2 73.0 100 1823
Post and Telecommunication 25.4 0.6 1.5 5.4 0.0 4.2 63.0 100 1808
Banking 3.3 10.9 8.3 36.5 5.8 1.6 33.6 100 1658
Trade and repair of motorvehicles 2.1 11.0 0.4 39.1 6.1 0.4 40.9 100 1571
Travel and Transport services 4.3 7.1 4.0 19.6 2.6 1.2 61.2 100 1211
Shipping and Sea transportation 0.8 3.5 0.5 5.0 0.7 0.5 89.2 100 1111
Service sectors Distribution of innovation costs (% values) Innovation costs

per employee
R&D Design Know-how Software Training Marketing Investment Totals (Ecu, 1995)
Hotel and Restaurants 2.1 4.8 1.5 25.3 3.2 11.5 51.6 100 575
0
z
Retail trade 1.0 4.8 4.3 22.0 2.0 4.8 61.1 100 494 >
Z
3.5 1.6 2.7 21.2 9.5 1.1 60.4 100 328
Cleaning
8
Security 0.3 0.0 1.2 19.9 3.9 1.0 73.8 100 24 1 2
Total services
Total services (without R&D serv.)
Manufacturing sector
*. Items not included in the questionnaire used for the manufacturing sector
Sources of information, objectives and hampering factors to innovation
Qualitative aspects of firms' technological activities investigated by inno-

vation surveys include the importance attached by service and manufactur-
ing firms to different sources of technological information (Table V),
different objectives pursued through innovation activities (Table 6), and
different obstacles (Table VII) to the introduction of innovation. Tables 5,
6 and 7 show the percentages of firms which have indicated the factors as
being "very important" (i.e. scores 4 or 5 on a 1-5 scale). The last columns
in the Tables show the ranking of the same factors (where applicable) for
manufacturing firms, in the period 1990-92. This allows us to make a
comparison of the importance of different information sources, objectives
and obstacles to innovation in services and manufacturing.
Sources of information
Table V shows that production and delivery departments are the most
important information sources in services: 36.7% of innovating firms indi-
cated such information sources as very important or crucial. The other
internal sources, i.e. R&D, marketing and others, play less important roles.
Among the external sources, clients and customers, and suppliers of equip-
ment, materials and components, are mentioned as very important infor-
mation sources by more then 30% of the firms. Research institutes (both
public and private ones) and patents and licences are perceived as very
important sources of information by a small minority of service firms.
Looking at the importance of the same factors in manufacturing, it
emerges that internal sources are very important in both service and manu-
facturing. The same holds true for clients and customers and suppliers of
equipment, material and components. Private and public research insti-
tutes play a role as providers of technological information for very few
firms also in the manufacturing sector. On the whole, these data suggest
the relevance in both service and manufacturing sectors of technological
information generated internally through "learning by doing" processes (in
the production departments), and externally through both up-stream and
down-stream user-producer interactions.
Objectives of innovation activities

Table VI shows the importance attached by service firms to different
objectives pursued with the introduction of innovation. Improving service
CHANGE IN SERVICES
quality is indicated as being a "very important" objective by a large per-

centage of firms (76.1%). Other relevant objectives are increasing the mar-
ket share (59.5%),extending the service range (48.1%).and lowering
production costs (47.4%).Overall, with the exclusion of the objective of
developing environmentally friendly products, innovation activities in
services appear to pursue a variety of objectives at the same time. Table VI
shows that manufacturing f m s pursue the same kind of innovation strate-
gies. Extending the servicelproduct range is, however, a more important
innovation strategy in services than in the manufacturing sector.
Obstacles to innovation
Finally, Table VII allows us to investigate the importance of different fac-
tors which hamper or hinder the introduction of innovations in firms. Eco-
nomic factors are considered to be the most important hampering factors
to innovation. In particular, for 22.9%of service firms the lack of appro-
priate sources of finance is regarded as a very important obstacle to the
introduction of innovation, while the cost of innovation and excessively
long pay-off periods for innovation are regarded as "very important"
obstacles by 21.9%and 13.9% of firms respectively. It is interesting to
note that the constraints due to legislation, norms, regulation and standards
emerge as the fourth most important liampering factors. Conversely, fac-
tors linked to the lack of information, technological opportunities and
appropriability conditions, which are very much emphasised in the litera-
ture, are regarded as "very important" by only a minority of service firms.
The importance of economic factors as well as institutional constraints due
to legislation, norms and standards are confirmed also among manufactur-
ing firms. The risk of being imitated by competitors is indeed a more
important hampering factor in manufacturing than in services. Appropria-
bility conditions seem, therefore, more important determinants of techno-
logical change in manufacturing than in services.
RINALDO EVANGELISTA
TABLE V Sources of information for innovation
S e r v i c e f i m for which the

source is very important Ranking in the
Source % on total manufac'uring
Number innovating sector
offirms ,inns
Internal sources:
Marketing
Other internal sources
External sources:
Clients or customers
Suppliers of equipment, materials and
comp.
Consultancy firms
Competitors
Conferences, seminars, specializedjour-
nals, etc.
Fairs and exhibitions
Other external sources
Patents, licences etc.
Universities and higher educational insti-
tutes
Public research institutes (excluding uni-
versities)
Private research institutes
*. The questionnaire used for the manufacturing sector does not distinguish between the
different "internal sources".
TABLE VI Objectives of innovation

S e r v i c e f i m for which the
objective is very important Rnnkinp" in the
Sources , rota, mndacruring
sector
wtin~
Improving service quality 4368 76.1 1

Increase market share 3232 59.5 2
Extend the service range 2602 48.1 6
Lower production costs 2567 47.4 3
Servicejnns for which the

objective is very imponanr Rmking in the
Sources % on tota, manufac'uriw
Number Of innovating sector .
firms
firms
Align the fum's technology to other firms' 252 1 46.7 -
Improving working conditions 2373 42.6 4
Improve production flexibility 2035 39.1 5
Maintain market share 2003 38.4 -
Enter in new markets 1947 37.5 7

Modify the service range 1467 28.1 9
Develop environment friendly products 557 11.0 8
Other objectives 140 6.8 -
(-): Objectives not included in the questionnaire used for the manufacturing sector
TABLE VII Factors hampering innovation
- -
Service firms for which the
obstacle is very imponant* in
Obstacles
Number of
finm
,?zvz!i manufacturing
sector
fim
Lack of appropriate sources of finance 2876 22.9 2
Innovation costs too high 2774 21.9 1
Pay-off period of innovation too long 1708 13.9 3
Constraints due to legislation, regulations 1540 12.6 4
and standards
Lack of skilled personnel 1429 11.6 6
Lack of customer's response 1314 10.6 10
Innovation potential (R&D, design, etc.) 1194 9.9 11
insufficient
Resistance to change within the firm 1180 9.7 12
Innovation costs hard to control 1036 8.6 8
Perceived risk too high 882 7.3 7
Lack of information on technologies 816 6.7 13
Lack of information on markets 718 5.9 9
Lack of appropriate external technical serv- 675 5.6 10
ices
Lack of technological opportunities 653 5.4 14
Risk to be imitated by competitors 267 2.2 5
* Both innovating and non-innovating firms have been taken into account
IV SECTORAL PAT'FERNS OF INNOVATION XN SERVICES
Almost paradoxically, one "peculiarity" of the service sector commonly

cited in the literature is its heterogeneity. The service sector encompasses
industries as diverse as Retail, Hotels, Telecommunication and R&D serv-
ices. A high degree of heterogeneity is consequently also expected with
reference to firms' innovation behaviours and performances. The sectoral
break-down of the data on innovation costs presented in Table IV has
already hinted at the presence of a wide variety of technological patterns in

services. The richness of the data provided by the Italian innovation survey
allows us to explore more in depth the heterogeneous nature of innovation
in services looking jointly at:
the type of innovative output (product or process innovation);
the overall innovative performance of industries ;
the type of technological source used by firms to innovate;
the patterns of interactions in innovation, that is the channel through
which firms acquire the relevant technological knowledge both within
the firms and from the wider institutional setting in which they operate.
the objectives and strategies pursued with innovation.
Both for descriptive and interpretative purposes the large set of variables
contained in the data-set needs however to be synthesised and reduced to a
manageable number of indicators able to grasp the key dimensions of
innovation activities in services. In order to do so, a factor analysis will be
performed on selected variables provided by the survey. Service sectors
will be then grouped (through a cluster analysis) according to their "simi-
larity" and "proximity" along the new factors extracted.
The variables used in the factor analysis

The list of variables used in the factor analysis, and their interpretative
meaning, is contained in Table VIII. Firm level data have been aggregated
at the level of 22 service sectors.
The process/product orientation of innovation activities is measured by
the predominance in each industry of firms introducing product or process
innovation (PROD-PROC).
As a proxy for innovation performance we use total innovation expendi-
ture per employee (1N.EXP.).
TABLE VIIl List of variables used in the factoral analysis
Lube1 Description Used as a proxy for

PRODK'ROC N. of firms introducing service innovation minus N. of firms introducing process product/process orientation (index rang-
innovation, all diveded by the sum of firms introducing product andlor process ing from -1 to 1)
innovation
IN.EXP. Total innovation costs per employee Total innovative intensity
R&D- DESIGN% R&D and Design expenditures as % of total innovation expenditures Importance of R&D and Design activi-
ties
OTH. DISEMBQ Innovation expenditures on Software. Acquisition of know-how, Training and Importance of other disembodied com-
INVEST%
Marketing as % of total innovation expenditures
Innovative investment as % of total innovation expenditures
ponents of innovation activities
Importance of embodied innovation
e
activities
k%2
S&T INTER % of firms indicating R&Ddepartment, Universities, Public research institutes Importance of S&T based interactions 2
and Conferences as very important information sources cn
m
USER-PROD INTER % of firms indicating Production departments, Suppliers and Clients as very Importance of learning by doing and
important information sources user-producer interactions !5i2
MARKET Jh'ETER % of finns indicating Marketing departments, Competitors and Consultancyfim Importance of market oriented interac-
as very important information sources tions
NEWMKT % of finns indicating Increasing market shares and Enter in new markets as very Importance of market oriented strate-
important objectives gies
DIVERSIF % of Fums indicating Modify and Exrend the service range as very important Importance of diversification strategies
objectives
COST/EFFIC % of firms indicating Lowering production costs, Improving productionflexibility Importance of cost- reducing and
and Improving production quality as very important objectives quality- enhancing strategies
*: R&D-DESIGN%+%OTH.DISEMB%+INVEST% = 100%;All variables have be computed as sectoral averages
The importance of the different technological sources used by firms

across industries will be measured by the share of total innoyation costs
devoted respectively to R&D and design (R&D-DESIGN%), to invest-
ment (INVEST%) and to a set of other disembodied cost items (acquisi-
tion and development of software, acquisition of know-how, training and
marketing - OTH. DISEMB%).
Data on the importance attached by f m s to different information
sources and objectives pursued with innovation (see tables V and VI) have
been used to build two set of indicators measuring the relevance of differ-
ent patterns of interactions and type of innovation strategies.*
S&T INTER measures the average importance attached by firms to uni-
versities, public research institutes and conferences as sources of techno-
logical information. Such flows and interactions are located "up-stream"
in the knowledge creating process and can be labelled as Science and
Technology based.
USER-PROD INTER measures the average importance attached by
firms to a different set of sources such as suppliers, customers, production
departments. This indicator thus captures the importance of more tacit
forms of knowledge and more "down-stream" links and interactions spe-
cifically oriented to answer specific needs of the final users or consumers
of services.
MARKET INTER measures the importance of marketing departments,
consultancy firms and private research institutes (likely to be "marketing"
oriented) and competitors as sources of technological innovation. This
indicator is therefore used to capture the importance of patterns of market
oriented interactions and knowledge flows.
NEW MARKET measures the importance of strategies aimed at enter-
ing in new markets or increasing market share. This indicator captures
therefore rather aggressive and market-oriented innovative behaviours.
DIVERSIF measures the importance of strategies aimed at modifying
and extending the service range, a strategy which is likely to be less inno-
vative than the previous one.
* The percentage of firms indicating the different information sources and objectives listed
in Tables 5 and 6 as "important" or ''crucial" (scores 4 and 5) have been grouped together and
averaged. Tfie different information sources and objectives have been grouped together look-
ing at the linear correlations between the average scores (see Evangelista and Savona, 1998).
CHANGE IN SERVICES
COST-EFFIC measures the importance attached by f m s to strategies

aimed at reducing costs, increasing the flexibility and enhancing the effi-
ciency of production process and delivery systems.*
The results of the factor analysis

The factor analysis was quite effective in synthesising the different dimen-
sions of innovative activities proxied by the variables listed in Table 8. As
shown in Table IX,out of the 11 original variables incorporated in our

analysis three factors have been extracted, explaining together 70% of the
total ~ a r i a n c eA. ~"rotated factor matrix" was also produced, reporting the
correlation coefficients between the original variables and the three factors
extracted (after rotation) (see Table I X ) . ~ This technique enhances the
interpretability of the factors extracted.
The first factor relates to the overall innovative intensity of service
industries (IN.EXP), the importance of internal innovative efforts rep-
resented by R&D and design activities (R&D-DESIGN%) and S&T
based patterns of interaction (S&T INTER). It relates negatively to the
importance of innovative efforts consisting of the absorption of tech-
nologies through the purchasing of new equipment (INVEST%).
The second factor relates to the importance of other disembodied
sources of technology (both internal and external) such as those con-
sisting of the acquisition of know-how and software as well as training
and marketing activities related to innovation (OTH. DISEMB%). This
factor also reflects the importance of knowledge flows of a more tacit
nature, coming from production/delivery departments, interactions
* Using the same data-base in another paper Evangelista and Savona (1998) have shown
that strategies aiming at "reducing costs" and "improving production flexibility" are comple-
mentary to strategies aiming at "improving the quality of services". An interpretation of this
finding is that in many service sectors improving the efficiency of the production process
leads to a substantial improvement of the quality of the service delivered.
t Factors have been extracted using the "Rincipal component analysis". The latter method
"extracts" from a given set of variables some new variables which are linear combinationsof
the original ones and are therefore able to summarize the interlinked phenomena measured by
the original variables. The statistical effectiveness of the factor analysis can be assessed by
the extent to which the new factors are able to reduce the original total variance, that is to
reduce the multidimensional nature of the phenomena investigated. On more interpretative
grounds, factor analysis is particularly effective when the factors extracted are few and they
can be meaningfully interpreted by looking at the way they are linked to the original varia-
bles.
$ The varirnax method has been used for rotation. This method attempts to minimizes the
number of variables highly correlated to each factor.
with clients (USER-PROD INTER) as well as from internal marketing

departments and consultancy firms (MARKET INTER). This factor is
negatively related to the importance of investment as a source of tech-
nological change (INVEST%) and innovation strategies finalised to
reducing costs and enhancing the efficiency of production processes
and delivery systems (COST-EFFIC).
The third factor summarises the importance of market-oriented inno-
vative strategies based on the introduction of new services
(PRODE'ROC), and of innovative efforts aiming at entering, creating

or diversifying into new markets (NEWMKT and DIVERSIF).
TABLE IX Results of the factor analysis (22 service sectors)
% of variance
Factor Eigenvalue Cumulative %
explained
1 3.95 36.0 36.0
2 2.14 19.5 55.4
3 1.71 15.6 71.0
Extraction method: Principal component analysis.
Rotated Factor Matrix
FACTOR 1 FACTOR 2 FACTOR 3
S&T based tech. other internal Marketlproduct
sources sources + userlproducer oriented innovation
and innovation interactions vs. strategies
intensity embodied tech-based
sources
IN.EXP 0.92 0.07 -0.07
R&D-DESIGN% 0.95 0.11 0.01
S&T INTER 0.90 0.11 0.18
OTH. DISEMB% -0.45 0.77 -0.10
USER-PROD INTER 0.45 0.57 4.14
MARKET INTER 0.17 0.76 0.26
INVEST% -0.‘60 -0.63 0.06
COST-EFFIC -0.32 -0.47 0.41
NEW MKT 0.11 0.06 0.92
DIVERSIF 0.14 0.17 0.86
PRODPROC -0.12 -0.21 0.46
Rotation method: Varimar
CHANGE IN SERVICES
A sectoral taxonomy of innovation in services
In order to extract homogeneous groups of sectors a cluster analysis was

performed using the first two factors identified above.* The statistical pro-
cedure was stopped when 9 clusters had been formed. This level was cho-
sen taking into account the statistical significance and technological
interpretability of the cluster formed.+ Figure 3 shows the different clus-
ters according to their approximate position with respect to the two main
factors: the vertical axis in the figure represents factor I and factor 2 is
represented by the. horizontal axis. For each sectoral cluster, average val-
ues of the main innovation ,variables used in the factor analysis are also
shown, so to provide a more detailed description of the technological pro-
files of the different sectoral groups.
The different groups of sectors identified by the cluster analysis can be
described as follows:
Technology users
In the bottom-left comer of Figure 3, clusters 1,2 and 3 are low innovation
intensity sectors, which rely upon technologies developed elsewhere. The
innovative pattern characterizing these industries resembles the supplier
dominated trajectory identified by Pavitt (1984) in his taxonomy. Invest-
ment represents by far the most important technological source in these
industries, accounting for a very large share of total innovation costs.
R&D, the design of new services and even the acquisition of know-how
play marginal roles. In all three groups of sectors technological interac-
tions reflect traditional user-producer links, which are likely to be confined
* Cluster analysis is a technique which identifies relatively homogeneous groups of obser-
vations by taking into account any set of quantitative and qualitative characteristics selected
by the analyst. Such a technique consists of grouping (stage by stage) the original observa-
tions in more aggregated groups in order to minimize the internal variance (within each
group) and maximize the inter-group variance. In this study the complete linkage method has
been used. There are several clustering methods, depending on how the "distances" between
cases and groups of observations are measured. In this study the complete linkage method
has been used. The use of other clustering methods (i.e. single linkage and average linkage
methods) has given very similar results.
t The clustering process is usually stopped when too heterogeneous sectors, or groups of
sectors. are combined. Although there are statistical parameters which help to decide the
number of clusters needed to best represent the data, an element of subjective judgement is
always involved in such a choice. An element of subjectivejudgement is also involved in the
choice of the variables to be used in the cluster analysis. Here it was decided to use only the
two main principal components since the inclusion of the third one has led to a much more
complex and less interpretable picture of the main innovation patterns in services.
to the acquisition of technologies from the manufacturing sector. Interac-

tions and knowledge flows with universities and other research institutes
are almost absent.
Waste, Land and Sea transportation form the most innovative and tech-
nology user cluster in this category of service industries. More than 90%
of total innovation costs in these sectors consists of the acquisition of new
machinery, equipment and the setting up of several kind of technological
infrastructures. This group of sectors could fall into the "scale intensive"
and "physical-network" category identified by Soete and Miozzo (1989).

Sectors in cluster 2 (Security, Cleaning and Other business) and 3 (Legal
services, Other financial, Travel and Retail) are the least innovative sectors
in services.
S& T-based
R&D, Engineering and Computing are typical science and technol-
ogy-based service sectors, being major propagators and diffusers of tech-
nological knowledge both within the service sector and in the
manufacturing industry. They have most of the characteristics of the Sci-
ence based firms and sectors identified first by Pavitt (1984) in his taxon-
omy referring to the manufacturing sector, and then by the sectoral
taxonomy of services proposed by Soete and Miozzo (1989). These are
very innovative industries which devote a large part of their innovative
efforts towards the generation and development of new technological
knowledge, as emerges by the large share of their innovation costs devoted
to R&D and design (90% in the case of the R&D services and 60% in the
case of Engineering and Computing services). The innovative pattern of
these sectors is also characterized by the presence of close interactions
with universities and research institutes and they also draw relevant tech-
nological information from conferences. It might be argued that these
industries are located very upstream in the knowledge-generating chain.
Interactive and IT based

Clusters 4 and 5 have been labelled as "interactive and IT based" since the
distinctive features of the innovation process in these sectors are the close
interaction with the users or customers, and large investments in software.
Innovative activities in these sectors do not rely upon R&D and design,
nor on the acquisition of technologies embodied in new machinery and
equipment. A large share of innovation costs in these sectors is devoted to
the acquisition of know-how from consultancy f m s , the purchasing and

development of new software, and other innovation sources such as train-
ing and marketing. The innovation pattern characterising these industries
thus reflects the central role of customization, as well as the importance of
"learning by doing" and "learning by interacting" processes. Advertising
is by far the most innovative sector among this category of service indus-
tries. Firms in this sector devote a large share of their innovation expendi-
tures to the development and acquisition of software and also give a great
importance to technical information drawn from consultancy firms and

competitors. Banks, Insurance and Hotels show a lower innovative inten-
sity and attach great importance to the acquisition of software and to inter-
actions with clients.
Technical consultancy
Technical consultancy seems to have both the characteristics of S&T-based
and ~nkractiveand IT based sectors. Firms in this industry devote large
financial resources to innovation, and a large share of them are used to
carry out R&D and design activities. The innovation process in this sector
is also characterized by the presence of close interactions with both final
customers, consultancy firms as well as with private research institutes.
What this sector does is to provide custom-specific answers to a variety of
technical needs and requirements of clients, exploiting the technologies
available in the market and in the broader science and technology system.
Post and Telecommunications sector is somewhere at the cross road of
the four main patterns identified in Figure 3. Firms in this sector perform
both R&D and design activities and also innovate through investment.
Firms draw technological information both from universities and research
institutes (S&T INTER) and from consultancy firms and competitors
(MARKET INTER). Such a mixed innovative pattern might also be
explained by the very different innovative characteristics of the two sec-
tors composing such a cluster which cannot be disentangled with the data
at our disposal.
V CONCLUSIONS
This article has provided original empirical evidence on some basic char-
acteristics of innovation activities in services and has also identified a few
214 R I N A L D O EVANGELISTA
RLDm,es SITBASED I
TECHNICAL
E23 CONSULTANCY
a
relevance or
R a D and
Lkalpn
Post and '
~etswrnn~unicatims
TECHNOLOGY
El ,%@-I
USERS
4 **I..
s.Tr.7,. "Sam - 9
a- n. I-". .,c
11111 1.
Advart!zing
Banks
I. ".
.*Y
.I.~
." ~
Relavanco Trndrlrep~nm o l w
or ertomal Other financial :
tschnologlcrl ..- . --
. -
r lrrrvir-.
~ , a v-. Hotels
sources Cleaning Rela11 ; INTERACTIVE
Ilnvostmentl a h e r bus. & I T BASED
I
User-producer inlarcelionr l n t e m a l l ~ c lInnovation
t soulcer F-2
Depondenee horn and
(rnanufscturlnp)aupplbrs Lmportanco ot sarvlco
pro~ldelDIcliontoIntor~*tions
F I G U R E 3 Sectoral patterns o f technological change in services
distinctive sectoral patterns of technological change in the heterogeneous

universe of service activities. The most relevant results of this article can
be summarised as follows.
Technological innovation is quite a diffused phenomenon in market
services: just less than a third of Italian service films have introduced at
least one technological innovation during the period 1993-95. Process
innovation, innovative investment and the acquisition and internal devel-
opment of software represent the most important channels through which
service firms innovate. R&D activities represent an important innovation
source only for a small number of science and technology-based service
industries. The other components of innovation expenditure (design,
acquisition of know-how, training and marketing) play a minor role. On
the whole, the empirical evidence presented in section 3 has shown that
service and manufacturing sectors show more similarities than differences
with respect to the basic dimensions of innovation processes taken into
account in this article.
The sectoral analysis has highlighted the variety of innovative patterns
which tend to discourage any simple generalization about innovation in
services. A sectoral taxonomy has been proposed in which service indus-
tries have been clustered according to the overall innovative performance
of firms, the nature of the activities carried out, the different knowledge
bases underlying the innovation processes and different interactive pat-
terns through which service firms innovate. Four main types of service
industries have been identified. T W of ~ them, namely the Technology users
and S&T based patterns resemble, respectively, the supplier dominated
and science-based trajectories identified by Pavitt (1984) in his taxonomy,
as well as two of the four technological patterns of innovation identified
by the author of this article, using the results of the first Italian innovation
survey in the manufacturing sector (Evangelista, 1996, 1999).

The technological trajectory which has been labelled as Interactive and
IT based (including Advertising, Banks, Insurance, Hotel, Trade and repair
of motor-vehicles) shows, on the contrary, a rather distinctive innovation
pattern. Innovation processes in these indusmes are neither focused on
internal technological efforts such as R&D and design, nor on the acquisi-
tion of new machinery and equipment. The acquisition and internal devel-
opment of software, and the acquisition of know-how, training and
marketing activities, absorb more than half of total innovation costs in
these sectors. Furthermore the close and continuous interaction with cli-
ents and final users represents an important source of innovation in these
industries.
A more comprehensive picture of the variegated nature of innovation in
services will be possible when features such as organisational change, cus-
tomisation, and non-technological innovation, will be amenable to appro-
priate statistical investigation. To reach that point, a substantial deal of
conceptual work is still needed to re-draw the boundaries of technology
and innovation activities. Although there is an increasing uneasiness with
respect to the still dominating narrow technologically-basedview on inno-
vation, it is still unclear in which directions, to what extent and how such
boundaries should be re-drawn. It is a strong belief of the author of this
article that such a conceptual effort is needed and would be much benefi-
cial to our general understanding of the innovation processes not only in
services but also in the manufacturing sector.
References
Antonelli, C. (1999). The microdynamics of technological change. London: Routledge.
Archibugi. D., P.Cohendet, A. Kristensen and K. A. Schiiffer (1995), Evaluation of the Com-
munity Innovation Survey, IKE Group, Aalborg, SprintIEims Report.
Barras, R. (1986), "Towards a theory of innovation in services", Research Policy, 15, pp.
161-173.
RINALDO EVANGELISTA
Brouwer. E. and A. Kleinknecht (1995). "An innovation survey in services: The experience
with the CIS questionnaire in The Netherlands", ST1 Review, 16, pp. 141-148.
Del Santo, A. and L. Forlani (1995). "I risultati dell'indagine ISTAT sulla fonnazione con-
tinua delle imprese", in Avveduto, S. and M. Pianta, Innovazione e risorse m n e
neN 'economiadella conoscenza, Cnr-Isrds, Roma.
Evangelista, R. (1996), "Embodied and disembodied innovative activities: Evidence from the
Italian innovation survey", in OECD, 1996d, pp. 139-162.
Evangelista, R. (1999), Knowledge and investment: The sources of innovation in industry,
Cheltenham: Edward Elgar.
Evangelista, R. and G. Sirilli (1995). "Measuring innovation in services", Research Evalua-
tion. 5(3), pp. 207-215.
Evangelista, R., G. Perani. F. Rapiti and D. Archibugi (1996). "Nature and impact of innova-
tion in manufacturing industry: Some evidence from the Italian innovation survey",
Research Policy, 26, pp. 521-536.
Evangelista, R. and M. Savona (1998). "Patterns of innovation in services. The results of the
Italian innovation survey", paper presented to the "VIII Annual RESER Conference",
Berlin, 8-10.0ctober.
Evangelista, R., T. Sandven, G. Sirilli and K. Smith (1998). "Measuring innovation in Euro-
pean industry", International Journal of the Economics and Business 5(3), pp. 311-333.
Gallouj, F. and 0. Weinstein (1997), "Innovation in services", Research Policy, 26, pp. 537-
556.
Hauknes, J. (1996). "Innovation in the service economy", STEP Report 7/96, Oslo.
ISTAT (1995), "Indagine sull'innovazione tecnologica", Anni 1990-92, Notiziario, Serie 4,
Foglio 41, Giugno, Rome.
Licht, G., C. Hipp, M. Kukuk and G. Munt (1997). "Innovationen im Dienstleistungssektor",
Schriftenreihe des ZEW, 24.
Kline, S. J. and N. Rosenberg (1986). "An overview on innovation", in Landau, R. and N.
Rosenberg, 1986.
Landau. R. and N. Rosenberg (eds.) (1986). The positive sum strategy, Washington DC:
National Academy Press.
Miles, I. (1993), "Services in the new industrial economy", Futures, 25(6), pp. 653-672.
Miles, I. (1995), "Services innovation, statistical and conceptual issues", Working Group on
Innovation and Technology Policy, OECD (Doc. DSTyEASISTP/NESnI(95)12), Paris,
28 March.
Miles, I. (1996). "Infrastructure and the delivery of new services", in OECD. 1996~.
Motohashi, K. (1997), "ICT diffusion and its impact in OECD countries", STIReview, 20, pp.
13-36.
OECD (1989), "Oecd proposed guidelines for collecting and interpreting technological inno-
vation data - Oslo Manual", OECD/GD(92)26, Paris.
OECD (1992). Technology and the economy: The key mlatiomhips, OECD, Paris.
OECD (1996a). Technology, productivity and job creation. Vol. 2 Analytical report, OECD,
Paris.
OECD (1996b). The knowledge-based economy, Science Technology Industry, OECD, Paris.
OECD (1996~).Employment and growth in the knowledge-based economy, OECD, Paris.
OECD (1996d), Imvation, patents and technological strategies, OECD, Paris.
OECD (1997). Information technology outlook, OECD, Paris.
OECD (1998), Science, technology and industry outlook, OECD, Paris.
OECD-EUROSTAT (1997). Proposed guidelines for collecting and interpreting technologi-
cal innovation data - Oslo Manual, OECD. Paris.
Pavitt, K. (1984), "Sectoral patterns of technological change: Toward a taxonomy and a the-
ory", Research Policy, 13, pp. 343-373.
Perani, G. and A. Del Santo (1998), "Statistical surveying of scientific research and techno-
logical innovation in services: Methodological problems from Istat surveys", paper pre-
sented to the "13th Voorburg Group Meeting on Service Statistics", Rome, 21-24
September.
Quinn, J.B. (1992). Intelligent enterprise: A knowledge and service based paradigm for
industry, New York: Free Press.
Simonetti, R., D. Archibugi and R. Evangelista (1995), "Product and process innovation:
How are they defined? How are they quantified?", Scientornetrics, 1, pp. 77-89.
Sirilli, G. and R. Evangelista (1998), "Technological innovation in services and manufactur-
ing: results from Italian survey", Reseamh Policy, 27, pp. 881-899.
Soete, L. and M. Miozzo (1989), 'Trade and development in services: A technology perspec-
tive", Working paper n. 89-031, MERIT, Maastricht.
Sundbo, J. (1997). "Management of innovation in services", The Service Industries Journal,
3. pp. 432455.
Vivarelli, M., R. Evangelista and M. Pianta (1996). "Innovation and employment in the Ital-
ian rnanufacturing.industry",Research Policy, 25, pp. 1013-1026.
Young, A. (1996), "Measuring R&D in services", ST1 Working Papers, OECD/GD(96)132.
RINALDO EVANGELISTA
THOD DO LOGICAL APPENDIX

The data set forth in this article are the results of a pilot survey on service
firms carried out in Italy by ISTAT (National Statistical Institute) in col-
laboration with the Institute for Scientific Research and Documentation of
the National Research Council (ISRDS-CNR).
The statistical population covered by the survey is composed of the total
number of service firms with more than 200 employees and of a sample of
firms employing between 20 and 200 employees. The sample is made of

6,005 firms, of which 1,245 with more than 200 employees and 4,760
sampled through a stratified random sample based on economic activity,
firm size and geographical location, from a universe of 19,300 firms.
The following service sectors have been covered:
- Trade, maintenance and repair of motor vehicles
- Wholesale trade (excluding motor vehicles)
- Retail trade
- Hotels and Restaurants
- Land transportation
- Shipping and Sea transportation
- Travel and transport services
- Postal and Telecommunication services
- Financial services
- Insurance and pension funds
- Auxiliary activities for financial intermediation
- Computer and informatics-related services
- Research and development services
- Legal, Accounting, Fiscal services
- Architectural, Engineering, Technical services
- Advertising
- Security and Cleaning services
- Other business services
- Waste disposal
The data have been collected through a mail questionnaire prepared on
the basis of the standards and definitions of the OECD "Oslo Manual".
There are however some little definitional differences due to the fact that
at the time of printing of the questionnaire the 1997 "Oslo Manual" had
not yet been published. The questionnaire was sent out in November 1996;
a remainder was sent to non responding firms on February 1997.
The questionnaire is made of 11 questions on the following issues: type
of innovations introduced (service, process/delivery), innovation expendi-
ture, sources of information for innovation, objectives, impact of innova-
tion on sales and on employment, obstacles to innovation, future
programmes for innovation.
The following definitions have been used in the questionnaire.
Technology can be defined, in a broad sense, as the complex set of

knowledge, capabilities, routines, competencies, equipment and technical
solutions necessary to produce a product or deliver a service.
Service and process technological innovations comprise technologically
new or significantly improved products and services. An innovation has
been implemented if it has been introduced on the market (service innova-
tion) or has been used within a production or delivery process (process
innovation). Usually, the introduction of service and process technological
innovations involves a series of scientific, technological, organisational,
financial and commercial activities.
A service is considered to be technologically innovative when its charac-
teristics and modalities of use are either completely new or have been sig-
nificantly improved from a qualitative point of view, or in terms of their
performance and technologies used. The introduction of a technologically
innovative service can require the use of radically new technologies, a
combination of pre-existing technologies or new knowledge. A service
consisting of a set of a number of integrated technical sub-systems can be
technologically improved by changing one or more of the sub-systems.
A process innovation consists of the adoption of a production or a deliv-
ery method which is new from a technological point of view. Such adop
tion may involve changes in equipment, organisation of production or a
combination of both. The introduction of process innovation may be
intended to produce or deliver innovated services which cannot be pro-
duced or delivered using pre-existing production methods or to improve
the production or delivery efficiency of existing services.
Firms have been asked to indicate the kind of innovation activities per-
formed, and the amount of financial resources devoted to them. The fol-
lowing definitions of innovation activities have been provided to
respondents:
Research and Development (R&D) comprises creative activities carried

out in order to acquire new knowledge or to us e new knowledge for new
applications.
Such activities can be carried out within the fm (intramural) or R&D
services can be purchased from other firms and institutions (extramural).
Design comprises activities aimed at defining procedures, specifications
and operational features (including tests and trial production) necessary
for the introduction of new services, processes and delivery systems.
Acquisition of know-how includes the acquisition of technologies from

external sources in the form of patents, non-patented innovations, licences,
know-how, trade marks, drawings and other consultancy services. Soft-
ware consists of the acquisition, development and updating of software
(excluding software incorporated in the hardware).
Training includes training activities connected to the introduction of new
services and processes only.
Marketing consists of the expenditures for launching new services
(expenditure for the building up of distribution networks and advertising
should be excluded).
Innovative investment includes the expenditures for the acquisition of
machinery and equipment linked to the introduction of new services, pmc-
esses and delivery systems. It includes the acquisition of the software
embedded in the new equipment and excludes the expenditures for land
and buildings.
The information collected through the questionnaire refers to innova-
tions introduced into the firm in the period 1993-1995 and innovation
expenditures sustained in 1995.
The data presented in this article and referring to the manufacturing
industry come from a previous innovation survey carried out in Italy by
ISTAT and ISRDS-CNR in the period 1993-94 and referring to innovation
activities undertaken during the period 199Q-92. This survey was part of
the Community Innovation Survey (CIS), co-ordinated by EUROSTAT.
The survey has covered all Italian manufacturing firms with more than 20
employees. The sample is made of 22,787 manufacturing h s (which
returned the questionnaire) out of a universe of more than 33,000 firms.
The data have been collected through a mail questionnaire prepared by
EUROSTAT on the basis of the standards and definitions of the OECD
"Oslo Manual". The overall structure of the questionnaire and the defini-
'
tion used in the ISTAT innovation survey in the manufacturing sector is
CHANGE IN SERVICES 22 1
very similar to the one used in the following survey in the service sector.
This similarity guaranties a good deal of comparability of the data pro-
vided by the two surveys. A more detailed description of the methodology
used in the CIS is contained in Archibugi et al., 1995, Evangelista et al.,
1998 and Perani and Del Santo, 1998.

Evangelista 2000

Uploaded by

Copyright:

Evangelista 2000

Uploaded by

Document Information

Original Description:

Copyright

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Evangelista 2000

Uploaded by

Copyright:

This article was downloaded by: [Universite De Paris 1]

On: 18 November 2012, At: 22:42

Economics of Innovation and New Technology

Sectoral Patterns Of Technological Change In

To link to this article: http://dx.doi.org/10.1080/10438590000000008

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

Institute for Studies on Scientific Research and Documentation, National Research

(Received March 15, 1999; In final fonn July 30, 1999)

This article provides a comprehensivepicture of the characteristics of innovation in services,

Innovation in services has become a topic of increasing interest among

major component of economic systems accounting for almost two thirds of

large-scale statistical attempt to collect systematic information on innova-

with the results of the 1995 innovation survey on manufacturing (ISTAT,

I1 MEASURING INNOVATION ACTIVITIES ]IN SERVICES:

The necessity for systematic data collection on innovation activities in the

evidence on the varied nature of innovation activities in industry, and more

the manufacturing sector. The necessity of extending the sectoral coverage

The intangible nature and the information-based characteristics of

implies that data on the generation and diffusion of information tech-

tion in order to include organisational changes which can either be

be a crucial strategic asset in services. Gallouj and Weinstein (1997)

Despite providing a large number of suggestive insights, the ever grow-

some of the peculiarities of innovation in services stressed in the literature.

the fact that "...organization is a crucial dimension of innovation, but its

As far as innovation activities are concerned the 1997 version of the

111AN OVERVIEW OF INNOVATION ACTIVITIES IN SERVICES

The presence sf innovation activities in services

One of the first aims of innovation surveys is to establish how widespread

and one fifth of the total. A surprisingly low percentage of innovating

Servicefirms (1993-95) Manufacturingfirms (1990-92)

TABLE I1 Innovating and non-innovating service firms by industry

Trade and repair of motorvehicles 1319 28.0

Wholesale trade (excl. rnotorvehicles) 4469 35.0

Retail trade 2522 20.1

Hotel and Restaurants 2186 19.6

Land transportation 1510 29.8

Shipping and Sea transportation 117 33.3

Travel and Transport services 1201 29.0

Post and Telecommunication 55 10.9

Banking 917 61.8

Insurance 157 63.1

Other financial services 163 61.3

Computing and Software 972 54.3

Other business services

Product and process innovation in services

the distinction between product and process innovation is more problem-

TABLE Ill Innovating service firms by type of innovation introduced

The nature of innovation activities in sewices

FIGURE 1 Breakdown of innovation costs in the service sector (1995)

FIGURE 2 Breakdown of innovation costs in the manufacturing sector (1992)

Table IV allows us to look at how the relative importance of the different

Industries innovating mainly through the acquisition of new machinery

Service sectors Distribution of i ~ o v a t i o ncosts (5% values) Innovation costs

Service sectors Distribution of innovation costs (% values) Innovation costs

Sources of information, objectives and hampering factors to innovation

Qualitative aspects of firms' technological activities investigated by inno-

Objectives of innovation activities

quality is indicated as being a "very important" objective by a large per-

innovation strategy in services than in the manufacturing sector.

TABLE V Sources of information for innovation

S e r v i c e f i m for which the

TABLE VI Objectives of innovation