437 HE G E ING O TECHNOLOGICAL PROGRESS An evolutionary perspective Rem5 Kemp and Luc Soete This article provides insight into technolo gy-econo my-ecology linkages which may help to define and accomplish environmentally sustainable development. An evolutionary perspective is adop ted in which economic growth and technologi cal change are viewed as a complex, non-lin ear, path-dependent process, drive n by short-term benefits instead of longer- term optimal ity. The article discusses the externality issues of technolo- gical change and the need for insti tutio nal adaptation, and talks about the relat ionship between economic growth and particular trajectories of technological change. It is stated that some of the present technological trajectories have reached their environme ntal limits and need to be replaced by environment-friendlier trajectories. However, such transi- tions are hindered by technical, economic and institutional barriers since the new trajectories have not yet benefited from ‘dynamic scale and learnin g effects’ and because the ‘selection environment’ is adapted to the old regime. The determinants of the decision processes to generate and adopt cleaner technol ogies are identif ied and analysed, and some policy issues of stimulating environment-friendlier technologies are dis- cussed. Technological progress has changed and shaped the world. Whether it has always been beneficial is the subject of ongoing debate. Although technolo- gical progress in the industrialized countries has led to great prosperity, at least in a materialistic sense, the burdens on the lives of some individuals were often high. In a world that is more and more affected by technological Rene Kemp is a researche r at the Maastricht Economic Research Institute on Innovation and Technology (MERIT), University of Limburg, P. 0. B ox 616, 6200 MD Maastricht, the Nether- lands. Luc Soete is the director of MERIT and professor of international economics in the Faculty of Economics and Business Administration of the University of Limburg. This article is a further development of ‘Inside the Green box: on the economics of technological change and the environment’, published in C. Freeman and L. Soete (editors), New Explorations in the Economics of Technological Change (Pinter, London/New York, 1990). FUTURES June 1992 0016-328 7/05437- 21 @ 1992 Butterworth-Heinemann Ltd
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
respe c ts: they dep end inc rea singly on sc ientific knowled ge and eq uipment ,
and further, in orde r to op erate , they assume an orga niza tiona l fab ric which
itself is com plex. One must inde ed spea k of system s here and not only o f
tec hnologies: one is immed iately involved in a netw ork of soc iote c hnica lrelations involving fac tors of supp ly, ma intena nce, insuranc e etc , without
which use of the tec hnologica l produc t wo uld be imp ossible. And the more
complex the soc iotec hnica l system, the more vulnerab le the soc ial orga niza -
tion is to ac c ident or the ob struction of just one p art of the system. At the
sam e time, the knowledge need ed to understand the tec hnic al op eration of
the system ha s be come so spe c ialized as to be esote ric to the ma jority of
pe op le. Spe c ialists in increasingly na rrow fields ha ve be en cut off from ea c h
other by the ir respec tive skills; with even grea ter rea son, the multitude of
non-tec hnicians has be en cut off from sc ientists and eng ineers.
The ve ry sc ale and complexity of the sc ientific and tec hnological
ente rprise mea n that its po tential c onseq uenc es a re unprec ed ented. As
ment ioned above , as the tec hnological risks of the past (pit explosions,
railwa y a c c ide nts, da m bursts) c an usually be averted now , toda y’s ‘m ajor
tec hnological risks’ threa ten large r a rea s and for a longer time. In the event
of a disaster, these a reas c an no long er be ea sily isolate d and hence
evac uated . Moreover, toxic emissions and rad ioa c tive co ntam ination ma y
have effec ts that are not dete c tab le for very ma ny yea rs or whic h last for
several generations.5
This b rings us quite naturally to the increasing importanc e of dynamic
‘externality’ issues. Here too , the de ba te surround ing the introd uc tion of
new tec hnologies has addressed ma ny issues w hich a t present dominate the
de ba te over environmenta l c osts and long-term da ma ge . Within a dynam ic ,
evo lutionary perspec tive, long-term externa lities are, in Nelson a nd Winter’s
wo rd s, no longer
susceptible to definitive once and for all categorization and are more intimately
related to particular historical and institutional contexts. To a large extent, the
problems involved are aspects of economic change. The processes of change are
continually tossing up new ‘externalities’ that must be dealt with in some manner or
other. In a regime in which technical advance is occurring and organizational
structure is evolving in response to changing patterns of demand and supply, new
non-market interactions that are not contained adequately by prevailing laws and
policies are almost certain to appear, and old ones may disappear. Long-lasting
chemical insecticides were not a problem eighty years ago. Horse manure polluted
the cities but automotive emissions did not. The canonical ‘externality’ problem ofevolutionary theory is the generation by new technologies of benefits and costs that
old institutional structures ignore.6
The huge environmenta l prob lems and threa ts that the wo rld is fac ing, suc h
as the de struc tion of the ozone laye r and the glob al warming, imp ly that
present institutiona l reg imes are inad eq ua te.’ Up until no w the use or
ab use of the environment ha s be en reg ulated , if at all, by em ission norms,
prod uc t sta ndards and ba ns, and -in some c ases-cha rge s and subsidies. It
is found that these measures have be en largely insuffic ient and have
prima rily led to the de velop ment and use of ‘c leaning’ tec hnology such as
‘e nd-of-pipe’ tec hnology and other treatment tec hnology instea d of ‘c lea n’
tec hnology (cleaner p rod uc tion proc esses) through which environmenta l
In our view, economic growth, similarly to technological development,
is primarily driven by short-term economic benefits and characterized as a
non-linear, evolutionary process, typically path-dependent with many bifur-
cations and possibilities of ‘locked-in’ development, in which past historyand historic events are important. I2 The direction and rate of technological
progress and economic growth are in our view not considered as being
autonomous and ‘inevitable’ but as an endogenous process related to the
structure of economic incentives, the accumulated knowledge and expertise
in scientific and business organizations and the socioinstitutional surround-
ing (regulation, values and social norms, lifestyles etc). Using the termino-
logy now commonly used in the economics of technological change
literature, economic growth is likely to be characterized by clusters of
economically interrelated technological trajectories, which might give
stimulus to growth in the whole economy.
Such clusters of technological trajectories have been identified withnew technological systems13 and new technoeconomic paradigms.14 The
network of technological trajectories related to cheap oil-based energy,
combined with mass utilization of the automobile as cheap individualized
transport system has often been identified with the post-war period of rapid
growth. In a similar fashion, other networks (eg electricity} have been
identified with respect to previous periods of rapid growth.15
As each system of ‘network’ technologies and infrastructure grows and
develops further, more and more negative externalities occur. Congestion,
nuisance of all kinds etc will gradually increase, so that the growth trajectory
will eventually reach its limits. Canals in the 18th-19th centuries are a good
example, as is horse transport in inner cities at the end of the 19th century.
From such a perspective, we would argue that present environmental
problems signal in a similar way to earlier congestion problems the limits of
the particular growth trajectory. A brief historical analogy might clarify the
point.
At the end of the 19th century the city of London was facing enormous
congestion and environmental problems related to the use of horses as a
means of transport. It is estimated that a horse produces no less than
15 kilos of manure per day. Most street corners in the city of London were
stationed by so-called crossing sweepers, whose task was not to keep the
roads clean, but to clear the way for pedestrians. At the end of the 19th
century there were around 6000 crossing sweepers in London. Alternative
means of transport had been available for years, but were not used because
of restrictive regulations: the red flag amendment, for instance, set a speed
limit for steam engines at 8 mph. The small-scale production restricted the
realization of dynamic learning and scale effects, and the lack of infrastruc-
For suc h c hanges, insight into the interrela ted ness of paths of ec ono mic
growth a nd tec hnological change , and into the po ssibilities of red irec ting it,
be co me imp ortant. For this we have ad op ted an evolutiona ry pe rspe c tive in
which ec onomic growth and tec hnolog ic al progress are view ed as a co m-plex, non-linea r and typica lly pa th-dep end ent p roc ess. In the ab senc e of
ap prop riate ince ntives for environmenta l protec tion, due to a lac k of
prope rly de fined p rop erty rights and inad eq uate go vernment regulation,
tec hnologica l progress and ec onom ic growth are driven in a direc tion that
c aused environmenta l de grad ation and the de terioration of ‘e nvironmenta l
capital ’ .
What is need ed is an ec ologica l mod ernization or cha nge in tec hnoec o-
nomic pa rad igm. The c hange of technological and ec onomic pa tterns
how ever is ob structed by the do minance of prevailing tec hnologica l traject-
ories that have be nefited from dynam ic sca le and lea rning effects, ac c om-
mod ated by the ad ap ted supp ly and de ma nd side . The a djustment andreplac ement of the past trajec tories requ ires, therefore, not only strong
ince ntives in the sense of rela tively rad ica l go vernment me asures, but a lso
the remova l of institutiona l ba rriers, new princ iples of environme nta l policy,
a reorientation of ed uca tion and science po lic y, a nd the integration of
ec olog ica l aspe c ts in othe r areas of po lic y. The short-term c onseq uenc es in
the sense of lower grow th or eve n loss of we lfare, in a ma terialistic sense,
c ould b e d ram atic. Not only should large pa rts of the industrial ac tivities
either be c losed or drastica lly c hang ed , but a lso pe op le’s lifestyles, espe-
cially in the ric h c ountries, wo uld have to ad ap t to the acute environmenta l
c risis. Whether the new grow th tra jec tories p rovide us with more we lfare
and pe rsona l freed om is less relevant from the pe rspec tive that our we lfa re
and freed om of c hoice are alrea dy de c rea sing a s a result of the environ-
mental degradation.
Coming bac k to the polic y implica tions, we sugg est the follow ing. First,
in line with w ha t wa s stated a bove , a stricter environmenta l po lic y ha s to be
implemented, in most ca ses be ing nothing more than simp ly to app ly the
polluter-pays principle.
Sec ond, we propo se a c hange in po lic y mix tow ards the use of
ec ono mic instruments suc h as cha rges (for em issions, wa ste and produc ts),
trade ab le po llution pe rmits and dep osit-fund systems (for instance for
wa ste). The b ene fits of suc h instruments a re ma ny: (1) c harges and trad e-
ab le po llution pe rmits are more effic ient be c ause every po lluter is given the
choice b etw een c omplianc e and pa ying the po lluter’s b ill, (2) there is a
financ ial ince ntive to d iminish all po llution-not merely to the level of
em issions sta nd ards, (3) suc h a system dep end s less on the ava ilab ility of
po llution c ontrol tec hnology-therefore it c an be introd uce d more quickly,
and onc e such a system is imp leme nted it generally p rovid es stronger
ince ntives for the de velop ment and ad op tion of c leane r tec hnolog ies than
current policies of tec hnology-b ased em ission sta nd ards, and (4) it stimu-
late s muc h more proc ess-integ ra ted tec hnology (inc luding rec ycling tec h-
nology) instea d of ‘ end -of-pipe ’ tec hnology. Whether suc h a system c an be
implemented, at low enough adm inistrative c osts, a nd whe ther such a
system is po litica lly and soc ially ac c ep tab le, has to be stud ied on a
case-by-case basis.54
Third, environmenta l polic y should be more oriented tow ards the
prevention of environmental problems instead of taking care of some of the
worst effects of the environmental problems. Economic instruments such as
charges and tradeable pollution permits serve this goal. Another way would
be to price a// products, pollution and industrial waste according to theirenvironmental damage. Such a system would favour preventive technologies
and discourage treatment technology that often leads to the transfer of
environmental problems (mostly in the form of new or other hazardous
waste problems). Pollution-prevention-pays programmes as in the USA
might also be useful .55 A lot of firms are not familiar with profitable
environmental measures because of a low priority of environmental manage-
ment and a lack of organization (thus being examples of what Leibenstein
calls X-inefficiency).
Fourth, a reorientation of education and science programmes is
needed, especially on the technical, economic, legal and health aspects of
environmental problems and policy. Both firms and the government are inneed of people with adequate knowledge of environmental aspects (either
technical, economic or legal). As stated above, firms often lack knowledge
about technical opportunities to reduce pollution and waste. At the govern-
ment level, policy making is hindered by lack of knowledge about the
health consequences of pollution and the costs and benefits of regulatory
actions. Also, the enforcement of environmental policy is seriously hindered
by lack of expertise and personnel.
Fifth, the integration of ecological aspects in other areas of policy is
required. In many policy areas ecological aspects are insufficiently inte-
grated, which hinders a transition to more sustainable economic and
technological paths. Examples abound: agricultural policies are still primarily
oriented at higher productivity and increasing production, despite the
understanding that agriculture is responsible for serious environmental
problems (water and soil pollution through the use of pesticides, herbicides
and fertilizers, acid rain etc). There is also a clear lack of the environmental
dimension in most technology policy, which seems much too focused on
other technologies, especially high-prestige projects such as space and
computer programmes, of which the costs are high and the benefits
uncertain. Public policy with respect to transport is still more oriented at
improving and extending the road infrastructure for cars and trucks instead
of stimulating the use of public transport services such as trains and buses.
And in the case of industrial policies there is not only a lack of integration
of ecological aspects but the policies are in fact aimed at protecting the
national polluting sectors against regulation, especially when regulation is
less strict in other countries. Finally, the constitutional structure, in the form
of tax policy and the regulatory framework of norms and standards,
sometimes unnecessarily hinders technological change towards environ-
mental protection. Recycling and treatment of waste material, for instance,
in some cases is obstructed by emission standards and product norms that
are developed for other, less urgent reasons. Public procurement policy can
also be better used for environment purposes, especially with respect to
building materials.56
Sixth, despite its evident importance as a principle to guide policy, the
‘polluter-pays principle’ is less useful in the international context. Becauseof the lack of international legislative and enforcement power, countries
that are the victims of the pollution of other c ountries ma y have to pa y for
environm enta l measures in othe r c ountries in orde r to prevent being
po lluted by their neighb ours. This ma y be come an important issue b ec ause
‘po llution know s no borde rs’ . We d o not want to be nega tive ab out this.Snce less de veloped co untries c an de ma nd to be co mp ensated for not
using CFCs in their refrige rato rs or for not emitting CO2 or other trac e gases
that c ontribute to the greenho use effec t, they ha ve some kind of an ‘asset’
for which they c an let rich c ountries pay for. Also in the case of interna-
tional environmenta l prob lems, industries in one country ma y de c ide to pay
for environmenta l measures in other countries that are more cost-effect-
ive-just as electric ity p rod uc ers in the Netherland s are now planning to pay
for forest p lanta tion in Brazil to com pensate for CO2 em issions of a new
pow er plant. All this of c ourse ma kes pe rfec t sense from an ec ono mic po int
of view.
Finally, the c onc erns a nd issues related to the environmenta l impa c t ofgrowth and tec hnologica l ad vanc e have sudd enly re-emerged in a co ntext
very d ifferent from that of the mid-1970s when the issues were first brought
to the forefront in the Club o f Rome report. To b eg in with, the evidenc e on
the environmenta l da ma ge in terms o f a ir, wa ter and soil po llution is by
now far more ove rwhe lming. Sec ond , the complexity and time-lag s of the
interac tions be twe en po llution and the ec olog ica l system and the surround -
ing ec onomic and tec hnological environments are still no t we ll, b ut never-
theless be tter unde rstoo d. Third, the pub lic percept ion of the environ-
menta l prob lems is fa r mo re acute . The hym n to ma terial p rogress with the
environment to be ad ap ted to the need s and requireme nts of such progress
appe ars no longer to be sung w ith the sam e conviction. Fourth, at a time
whe n na tional go vernments a re wa king up to the imp ortanc e of the issues
involved , it is the globa l d imension which is most ac ute. Particularly with
respec t to environme nta l issues, the na tiona l sta te, in Daniel Bell’s word s,
ap pea rs indee d ‘ too big for the sma ll prob lems of life and too sma ll for the
big p rob lems of life’ .57
Notes and references
1 .
2.
3.
4.
5.
6.
7.
a .
In this article we deal with the environmental problems of past industrialization and
application of technology and not with the present and future environmental problems
which are related to population growth and poverty. Further, the focus is more on the
problems of pollution than on the problem of overexploitation of natural resources.A similar perspective is in R. U. Ayres, ‘Industrial metabolism’, in J. H. Ausubel and H. E.
Sladowich (editors), Technology and Environment (Washington, DC, National Academy
Press, 1989), pages 23-49.
A typical example of this is the announced extra spending on environmental care by the
provincial authorities of Noord-Brabant in the Netherlands out of an unplanned amount of
tax income derived from increased car ownership.
H. Brooks, ‘Technology assessment as a process’, UNESCO, international Social Sciences
journal, 25(3), 1973.
P. Lagadec, La Civdisation du Risque (Paris, Le Seuil, 1981); and P. Lagadec, Le Risque
between 76% and 81%. See V. J. Hartje and R. L. Lurie, Adopting Rules for Pollution
Control Innovations: End-of-Pipe versus integrated Process Technology (Berlin, Wissen-
schaftszentrum Berlin, International Institute for Environment and Society (IIUG), 1984),
page 368.
9. We do not want to ban the use of traditional instruments of environmental policy such asstandards and charges-although we would welcome a shift in the policy-mix in favour of
economic instruments such as charges and tradeable emissions permits (see section 5).
10. For a discussion of this issue, see R. E. Litan, ‘The safety and innovation effects of US
liability law: the evidence’, American Economic Review, 87(2), 1991, pages 59-64.
11. Although our approach is more economic in character, it corresponds to the idea of
co-evolutionary development of R. B. Norgaard, ‘Coevolutionary development potential’,
Land Economics, 60(2), 1984, pages 160-173.
12. The idea of technological change as an historic, path-dependent process, with possibilities
of ‘locked-in’ development is worked out theoretically in Paul David’s model of localized
learning and Brian Arthur’s model of increasing returns with adoption. In the words of P. A.
David, Technical Choice, Innovation and Economic Growth: Essays in the Nineteenth
Century (New York, Cambridge University Press, 1975), page 4: ‘Because technological
“learning” depends upon the accumulation of actual production experience, short-sighted
choices about what to produce, and especially about how to produce it using presently
known methods, also in effect govern what subsequently comes to be learned. Choices of
technique become the link through which prevailing economic conditions may influence
the future dimensions of technological knowledge. This is not the only link imaginable. But
it may be far more important historically than the rational, forward-looking responses of
optimizing inventors and innovators which economists have been inclined to depict as
responsible for the appearance of market- or demand-induced changes in the state of
technology.’ More recently, it is Frank Hahn who writes: ‘The path of history is the
outcome of individual decisions and in turn helps to fix the latter. This is really the main
message: the information available to agents at any time is determined by the particular
path followed. The economy could have followed a different path and generated quite
different information. There is something essentially historical in a proper definition of
equilibrium and of course in the dynamics itself’. F. Hahn, ‘Information dynamics and
equilibrium’, paper to Conference of Scottish Economists, as quoted by C. Freeman, in G.
Dosi, C. Freeman, R. R. Nelson, G. Silverberg and L. L. G. Soete (editors), Technical
Change and Economic Theory (London and New York, Pinter Publishers, 1988), pages 4-5.
13. C. Freeman, J. Clark and L. L. C. Soete, Unemployment and Technical Innovation: a Study
of Long Waves in Economic Development (London, Pinter, 1982).
14. C. Freeman and C. Perez, ‘Structural crises of adjustment, business cycles and investment
behaviour, in Dosi et a/, op tit, reference 12.
15. See Freeman et a/, op tit, reference 13.
16. E. E. Montrol l and W. W. Badger, Introduction to Quantitative Aspects of Social Phenomena
(New York, Gordon and Breach, 1974), page 224.
17. A. Griibler, Rise and Fall of Infrastructures, Dynamics of Evolution and Technological
Change in Transport, PhD Dissertation, Technische Universitat, Vienna, 1988.
18. For a discussion of the factors affecting the diffusion of ISDN, see eg P. A. David and W. E.
Steinmuller, ‘The ISDN bandwagon is coming, but who will be there to climb aboard?
Quandaries in the economics of data communication networks’, Economics of innovations
and New Technology, I, 1990, pages 43-62.
19. J. H. Ausubel, ‘Regularities in technological development: an environmental view’, in J. H.
Ausubel and H. E. Sladowich (editors), Technology and Environment (Washington, DC,
National Academy Press, 1989), pages 70-91.
20. Ibid.
21. Ibid, page 72.
22. C. Freeman, ‘Prometheus unbound’, futures, 76(5), October 1984, pages 494-507.
23. Nationaal Milieubeleidsplan (NMP) (National Environmental Policy Plan), Tweede Kamer,
vergaderjaar 1988-1989, 21137, no l-2, The Netherlands, 1989.
24. R. R. Nelson and S. G. Winter, ‘In search of useful theory of innovation’, Research Policy, 6,
1977, pages 36-76.
25. For an overview of the literature, see G. Dosi, C. Freeman, R. R. Nelson, C. Silverberg and
L. L. G. Soete (editors), Technical Change and Economic Theory (London and New York,
Pinter Publishers, 1988), and for a wide collection of important contributions, see G. Dosi,
‘The nature of the innovation process’, in ibid.
26. Ibid, page 223.27. Nelson and Winter, op tit, references 6 and 24.
also p ub lishe d in N. Rosenbe rg, Inside the Black Box (Cambridge, Cambridge University
Press, 1982), p ages 104-119.
29. G. Dosi, ‘Tec hnolog ica l pa radigm s and tec hnolog ica l trajec tories: a sugg ested interpretation
of the determinants and direc tions of tec hnica l c hange , Research Policy, 11, 1982, pages146-162. For a discusssion of a ll these co nc ep ts, see G. Dosi, ‘ Sourc es, p roc ed ures an d
microecono mic effects of innovation’, /ournat of Economic Literature, 26(3), 1988, pages
1120-1171; and R. P. M. Kemp , ‘A n ec ono mic an alysis of c leane r tec hnolog y: theo ry a nd
evidence’, pa pe r presented at the ‘G reening of Industry’ Conferenc e in Noordw ijk, the
Nethe rla nd s, 17-19 Nove mb er 1991.
30. Nelson a nd Winter, op tit, referenc es 24 a nd 27.
31. Nelson an d Winter, op tit, referenc e 27, pa ge s 262-263. The distinc tion b et we en the
generation of an innovation and its selection doe s not imply that these tw o phenom ena are
not linked . On the c ontrary, the de c ision to de velop an innovation, whic h involves R&D
co sts, d ep end s on the c harac teristics of the selec tion environm ent, and the selec tion
environment is shap ed by tec hnological cha nge.
32. Fou r kinds o f net wo rks c a n be distingu ished : (i) the firm a s a net wo rk of a c tivities (R&D,
ma nufac turing, ma nag eme nt, ma rketing); (ii) the ec ono my a s a netw ork of interac ting
supp liers and users, and involving a ce rtain infrastructure; (iii) the tec hnolog ica l system (thewh ole of ma chines, tec hniques, co mp onents, tec hnica l stand ards); (iv) the soc ioinstitutiona l
network (educ ation and training, regulation, values, p ow er ba lance, attitudes to techno logy
and cha nge etc). It is within and through these networks that learning and ad ap tation takes
pla c e that may be referred to as netw ork externalities (see sec tion 2).
33. Lea rning cu rves, or expe rienc e c urves, are a we ll know n phe nom enon in industry. Despite
their widesprea d prac tical acc epta nce, theoretica l research on the learning curve has bee n
limited , and there ha ve be en relatively few pu blished em piric al studies (M. B. Liebe rman,
‘The learning c urve and pricing in the c hem ica l proc essing industries’, Rand journal of
Economics, 7X2), 1984, pa ge 213). A not ab le exc ep tion is the Project PIPPA (Post-
Innova tion Performa nce and Policy Ana lysis) that studied the incide nc e and origin o f po st-
innovation in a large numb er of prod ucts and their man ufac turing proc esses (in L.
Ce orgh iou, J. S. Me tc a lfe, M. Gib b ons, T Ray an d J. Eva ns, Post-innovation Performance:
Technological Development and Competition (London, Macmillan, 1986)).
34. K. J. Arrow , ‘The ec ono mic implica tions of learning b y do ing’, Review of Economic Studies,
29,1962, pages 155-173.
35. Rosenberg (1982), op tit, reference 28.
36. B. A. Lund va ll, ‘Innova tion as an interactive proc ess: from user-produc er iteration to the
na tional system of innovation’ , in Dosi et a/, op tit, reference 25.
37. W. I. Ab ernathy and 1. M. Utterbac k, ‘A dyn am ic mo de l of proc ess and prod uct innovation’ ,
Omega, 3(6), 1975, p ag es 639-656, hav e de velop ed a dyna mic mo de l of proc ess and prod uct
innovation. In their mod el, in the ea rly stag e of de velop ment, whe n the produc t market is
ill de fined, the rate of prod uc t innovat ions is relatively high in the ea rly stag e of
de velop men t whe reas in the later stag es of the prod uct lifec yc le, w hen the prod uct is more
ad ap ted to user-need s, and whe n price compe tition bec omes more important, process
innovations, aime d at reduc ing co sts, a re more freq uent.
38. The impo rtance of improveme nts of an innova tion in favo ur of a wid e diffusion is
some thing we ll know n to historians of tec hnolog y. N. Rosenb erg, Perspectives on Techno-
logy (Ca mb rid ge , Ca mb ridg e University Press, 1976), pa ge 195, write ‘. mo st invent ions
are relatively c rude a nd ineffic ient at the date whe n the y are first reco gnized as co nstituting
a ne w innovation. They are, of nec essity, ba dly ad ap ted to ma ny of the ultimate uses to
whic h they will eventua lly be put; therefore, they may offer only very small ad vantag es, or
pe rhaps none at all, ove r previously existing tec hniques. Diffusion unde r these circum-
stanc es will nec essarily be slow.’ Rosenb erg also points to the impo rtance of sec ond ary
innovat ions (for examp le in ma c hine and tool de velop men t), infrastruc tural fac ilities a nd the
av ailab ility of tec hnica l skills required (wh ich correspo nds to our co nce pt of selec tion
environment).
39. Although ‘c leaner’ or ‘environment-saving’ technology wo uld be a better term, w e use the
more com mon term ‘c lean tec hnolog y’ in the following (for a discu ssion of different
c onc ep ts a nd terms, see R. P. M. Kem p, A. A. Olstho orn, F. H. Oo sterhuis a nd H.
Verbruggen, Instrumenten voor de stimulering van milieutechnologie (Polic y instrume nts to
stimulate clea ner tec hnolog y), (Leidsc hend am , the Nethe rlan ds, VROM , 1991-a shorter
version of the repo rt is p ub lishe d b y EZ, Den Haa g); R. P. M. Kem p, A. A. Olstho orn, F. H.
Oosterhuis, H. Verbrugge n, ‘Policy instruments to stimulate c leane r tec hnology’ , pa pe rpresented at the EAERE-Confe renc e in Stoc kholm, II-14 June 1991. Clea n tec hno log y is
used here as a ge neral te rm for all tec hniques, proc esses and prod ucts that av oid or
diminish environmental damage and/or help save raw materials, natural resources and
energy. As such it is a bit of a misnomer: first, no technology of course is clean in a strict
sense, and second, as stated before, a distinction should be made between ‘clean ’ and
‘cleaning’ technology.
40. Dosi, op tit, reference 25, page 222.41. D. C. Mowery and N. Rosenberg, ‘The influence of market demand upon innovation’,
Research Policy, 6, 1979, pages 102-153.
42. R. C. Levin, ‘A new look at the patent system’, American Economic Review, American
Economic Association Proceedings, 1986, pages 199-202.
43. Dosi, op tit, reference 25.
44. P. Ijlst, C. T. M. Stokman and E. T. Visser, tnformatieoverdracht en informatiebehoefte in
de milieuproduktiesector in Nederland (Information transfers and information needs in the
environmental industry in the Netherlands), (Zoetermeer, EIM, 1988).
45. R. R. Nelson, Understanding Technical Change as an Evolutionary Process (Amsterdam,
North Holland, 1987).
46. Levin, op tit, reference 42.
47. Ibid.
48. For a discussion of this, see ibid.
49. Although lower prices increase demand, potential buyers may also decide to postpone theirpurchase when further price falls are expected (see Rosenberg, op tit, reference 28).
50. A. Frank and H. J. J. Swarte, Milieutechnologieen: toepassing in kleine en middelgrote
ondernemingen (The use of clean technologies in small and medium size firms), (Rotterdam,
The Netherlands, Erasmus Studiecentrum voor Milieukunde, 1986); and H. E. Williams,
J. Medhurst and K. Drew, ‘Corporate strategies for a sustainable future’, paper presented at
the ‘Greening of Industry’ Conference in Noordwijk, the Netherlands, 17-19 November
1991.
51. Frank and Swarte, op tit, reference 50.
52. Abernathy and Utterback, op tit, reference 37.
53. We do not plan to go deeper into the topic of policy instruments to stimulate clean
technology. For a discussion on this topic see J. Cramer, J. W. Schot, F. van den Akker and
G. Maas Ceesteranus, ‘Stimulating cleaner technologies through economic instruments:
possibilities and constraints’, tndustry and Environment Review, 73, April-May-June 1990,
pages 46-53; and Kemp et al, op tit, reference 39.
54. High administrative costs for charges can make such a system less efficient than a system of
standards, which contradicts the traditional theoretical argument by economists that a
system of charges is more efficient. In addition, whereas the costs of pollution control for a
certain sector will be lower in the case of charges compared to the costs of pollution
control under a regime of standards, the sum of the total costs of pollution control and the
payments for the remainder of the emissions will generally be higher than the total costs of
pollution control under direct regulation, unless the polluting sector is financially compen-
sated.
55. D. Huisingh, L. Martin, H. Hilger and N. Seldman, Proven Profits from Pollution Prevention:
Case studies in resource conservation and waste reduction (Washington, DC, Institute for
Local Selfreliance, 1986).
56. Of course, a cleaner environment is not the only government objective. The point here is
that technological transitions generally proceed slowly since they require all kinds of
institutional changes that do not occur rapidly.
57. J.-J. Salomon and A. Lebeau, L’ecrivain Public et I’Ordinateur: Mirages du developpement