Sandra Rothenberg, James Maxwell -Industrial Response to the Banning of FCs

Industrial Response to the~Banning of CFCs: Mappirlg thePaths of Technical Chan~~e1

Sandra RothenbergJJames Maxwell

Massachusetts Institute of Technology ( IT), USA

About the Authors

Sandra Rothenberg is a doctoral student in the organizational st dies programat MIT's Sloan School of Management. Prior to entering the d ctoral pro-gram, she worked as a research associate at the International M tor VehicleProgram aiMIT, studying environment policy and manageme t in the auto-mobile industry. Her research interests presently include firm anagementof environmental performance.

Mailing Address: Massachusetts Institute ofTec~nolofiY' lnternati.onal Motort hicl~ Pro-

gram, Sloan School of Management, 50 Memonal Dnve, Cambndge, MA 0 142-1~47 USA;

Phone: (617) 253.,3847; Fax: (617) 253-2660; E-mail: SLRQTHEN@MIT. U

Dr. Maxwell is a visiting scholar. at the Center for International Studies atMIT and Director of the Center for Environment and Health t Johl[1 Snow,Inc. He received his Ph.D. in the field of Public Policy at MIT. He isl cur-rently supervising studies of siting in the US and foreign count ies, integra-tion of environment and health in policy making, and environ ental man-agement practices in the automobile industry.

Mailing Address: MIT Center for International Studies, 292 Main Street, Ct bridge, MA

02139 USA; Phone: (617) 253-4949; Fax: (617) 253-9330; E-mail:

]IM_MAXWELL@]SLCOM

Descriptorstech?i~ innovation, ozone depletion, environmental.mana~ement, aut?mob~le industry, ~lec-trorncs Industry, CFCs, Montreal Protocol, technologIcal trajectory, envIronmfntal regulatIon

-213

Technology S~~es 1997,4/2, @ 1998 Walter de Gru er

Abstract

Policy makers assume risks when they attemp to m~ndate environmentalchanges in industry practice that go beyond e isting technical capabilities. Inthe case of the international regulation of chI rofluorocarbons (CFCs), theelectronics and automobile industries respond d suqcessfully by making atransition to CFC-free alternatives. Neverthel ss, the nature and timing ofresponses differed significantly between the 0 industries. Large multina-tional electronics firms considered the global estric~ions a challenge, andmoved swiftly to develop alternatives to CFC- ased ~hemicals. Their counter-parts in the automobile industry, however, del yed ib making a commitmentto air-conditioning systems which run on CF substitutes. The differencesbetween the two industries can only be explai ed by a complex mix of tech-nical, organizational, and institutional factors. In light. of these cases, wesuggest that policy makers need to consider re ulatory approaches that aremore tailored to particular industries.

During the past decade, industry has increasi gly refognized that productand process improvements can often simultan ousl}i enhance environmentalper~ormance, while improving C:J~ality a.nd 10 ~ringl~osts. Nevertheless, c?m-parnes often overlook opporturntles for mnov Ion 'tlthout the mandates Im-posed by government, and regulation remains the primary method of induc-ing environmentally related technical changes in industry.

A small body of literature examines the infl encel of environmental regula-tion ?n technological innovation from a socio ogicat and organizational per-

spectIve (Ashford, 1993; Ashford, Ayers, & St ne, 1,985; Ashford & Heaton,

1983; Kemp & Soete, 1992; Rothwell, 1981; chot~ 1992). One commonthread in this literature is that regulations ofte induce environmental inno-vation; however, these regulations often fail to fulfill their potential. Theyforce innovation at high cost, mandate the ad ption of end-of-pipe technol-ogies, or stifle innovation altogether. The nev -end~ng challenge for policymakers is to develop regulations that lead to e viroqmental improvements,but develop them in a way that minimizes the burd~ns on industry. Past re-search on government policies designed to su port irnovation suggests thatif policy makers are to foster innovation, it is ssent1al that they understandthe factors driving the process. As important, hey must understand howthese factors vary across different firms and in ustries (Ashford, 1993; Nelson& Langois, 1983; Utterback, 1974).

In this article, we look at the response of 0 user industries, electronicsand automobiles, to the restriction of, and en ing !?;lobal ban, on CFCs inthe late 1980s. Both industries ultimately resp ndedl successfully to the globalban by developing alternative technologies to FCs~ However, large multina-

I

-

214

,tional firms within the two industries differed in the nature and timi ' g of

their response to the phase-out of CFCs. Large electronic firms consi eredthe global restrictions as a challenge, and moved swiftly to develop al erna-tives to CFC-based chemicals. Many of the large automotive compa 'es,however, delayed in making a commitment to air-conditioning syste s thatran on CFC substitutes, despite the fact that delays could potentially cause alarge and costly retrofit problem for the industry.

We believe that the different responses between the two industries, as wellas among certain firms, cannot simply be explained by existing mode s offirm response to environmental regulation. It was acomp.lex interacti n oftechnological, organizational and institutional factors that influenced thewillingness of firms to innovate and respond "proactively' to technol gy-forcing regulations. The case of CFCs is thus of both theoretical and racricalinterest in identifying factors that shape industry's response to techno ogy-forcing regulation. The information for this study was collected rhro gh in-terviews with representatives from these industries in 1992 and 1993 as wellas extensive research of secondary source material as part of the MassachusettsInstitute of Technology Norwegian Chlorine Study (Technology Businessand Environment Program, 1993). I

Literature Review [The Determinants of Technological Change

Although a significant amount of research has been conducted to understandthe relationships between scientific discovery, innovation, markets, an~ otherinstitutions external to the firm, much remains to be discovered abou thenature of technological progress. There is a growing jbodyof literatur on theprocesses and institutions involved with technological advance (Tus an &Nelson, 1990). Several theorists have proposed that technological cha ge isfar more complex than traditional "market-pull" jor "technology-push" modelssuggest. Instead, they argue, technology moves along an evolutionary path,the direction of which is determined by a number of technical, organi tional,and contextual factors (Dosi, 1982; Nelson & Winter, 1982; Rosenb ,rg,1969; Van den Belt & Rip, 1987). !

Empirical work supports the notion, for example, thatorganizatio~al fac-tors are a critical influence on the nature of technological change. Technologyi~related to the efforts, knowledge and s~lls.ofindividuals in th~Org~niza- tl0n, as well as the structure of the organIzatIon as a whole (Dosl, 1.9 2;

Nelson & Winter, 1982). As described by Dosi (1982), firms operate withinan existing technological paradigm, which is a "pattern of solutions to

; ' elected

technological problems, based on selected principles derived from nat ral

sciences and on selected material technologies" (p.152). This paradig en-

--t-

215

Technology Studies 1997,4/2, @ 1998 Walter de Gruyte

compasses certain cognitive and organizational imits by encouraging thedevelopment of certain skills and behavioural r utines over others. Once atechnological paradigm has been selected, these limits exclude alternativetechnological possibilities and technological pr gress moves down a specificpath or trajectory. Over time, organizational ro tines and capabilities evolveto operate under the current paradigm, and m ement along a given pathbuilds momentum, directing problem solving aFtivit>f and technology devel-opment in a certain direction. i i

Beyond the organization's boundary, changeI s motivated not only by ex-

ogenous needs expressed through market signal, as suggested in the traditio-

nal "market-pull" model, but also by a whole h st of institutional and politi-cal factors (Nelson & Winter, 1977). The influ nce of the operating contexton the path varies. External forces can challeng organizations to focus onnew technological innovation paths or operate nder new technological para-digms. Alternatively social structures, procedure, relationships and norms cansupport existing technological paradigms, enco raging technological changedown the existing path (Tushman & Rosenkop 1992). Tushman and Rosen-kopf (1992) propose that the significance of th se factors in determining thepath of technological change varies with the co plexity of the system at hand;the more complex the system, the more relevan institutional and politicalfactors will be in determining the path of devel pment.

Theoretical and empirical work has attempte to e~tablish when these var-ious factors exert the greatest influence on the p ocess lof technical change. Forthe most part, this work finds that the determi ants 6f technological changevary in importance along with the phases of tec nology development through-out the product life cycle (Anderson & Tushman 1990; Tushman & Anderson,1986; Tushman & Rosenkopf, 1992). Tushman nd Rosenkopf(1992) suggestthat social and organizational factors have the ost effect during periods ofexperimentation, when keen competition exists between old and new tech-nological regimes. Once a dominant design has een ~stablished, technologi-cal change proceeds incrementally. At this stage" social, political, and orga-nizational factors are less influential, and techn logical change is driven moreby the "logic" internal to the technology (Tush an & Rosenkopf, 1992).

It is clearly difficult to separate the influence of myriad of factors on theprocess of technological innovation (Bijker, Hu hes, & Pinch, 1987). AsHughes (1988) points out, technology, organiz tions, and social context canall be part of one "seamless web." Organization I characteristics, such as themanagement structure, depend on the characte istics bf the technologicalartifacts, while technologies are shaped by the 0 ganization in which they areused (Thomas, 1994; Orlikowski, 1992). Simil rly, both technological andorganizational systems can be closely related or ven mirror the institutionalcontext in which they operate (DiMaggio & Po ell, 1991; Hughes, 1987).

-

216

Envir?nmental regulation~ are powerf~l institutional factors motivatiJg tech-ngiogical change. A promInent analytIcal model of this process is off~~ed byA§hford (1993) and.Ashford and Heat?n (1983). The model elucidatfs howvarIOUS types of envIronmental regulatIon lead firms to modify the develop-ment and production of technology, and are technology-forcing to differen~degrees. The firms or industries assume a critical role in the model W~ n they respond to the regulatory stimuli. Ashford and Heaton (1983) hypo esize

t?at. industrys response to regulato.ry ~ti~uli is pri~arily shaped ?y t e loca-tlonof the regulated technology withIn Its overall .lIfe cycle. Specifical y, theyrely on the Abernathy and Utterback (1978) framework which suggests that:(a) innovation is predictable in any given industrial context, (b) the c arac-te~istics of the particular technology are the main determinants of the naturbof technQlogical chang~, and (c) the process of innovation is relatively uni-for~~cross widely different productive segments (Ashford & Heaton, 1983).

A§hfordand Heaton's (1983) model, though useful, has cenain limi ations.First, ~mpirical evidence supports the notion that not all te~hnologies followa predictable life cycle (Henderson, 1994). Second, although Ashford 1993)and Ashford and Heaton (1983) recognize that the technological inn ationprocess is far more complex than the Abernathy and Utterback (1978) modelsuggests, they have not attempted to incorporate other factors into the model.Factors such as organizational and industry culture, market dynamics, regu-latoryhistory, history of technological innovation, and even individu ac- !tions are overlooked, even though they are critical in understanding t ere,.. :sponse to regulation at the industry and firm level.

In ~e next section we will illustrate the varying impact of regulatio ontechnical innovation by looking at the responses of two user industrie , theautomobile and electronics industry, to the phase-out of CFCs called or inthe Montreal Protocol. This case is particularly interesting because the Mon-treal Protocol is the first example of a global accord that calls for the t tal !phase-out of a group of chemicals. Both the automobile and the electr nics Iindustry were faced at the time of the Protocol with major technical c al-lenges, as alternative products and processes did not appear to be imm diate~ly available. Their responses to these constraints differed markedly. i

The Banning of CFCs I

CFCs are chemical compounds derived from simple hydrocarbons wit halo-gen atoms substituted for hydrogen. They have been used in industry n- icreasingly since the 1930s in a growing number of applications. By th late I1970s and early 1980s, CFCs were widely used in a number of sectors The'

--+--

21Y

principal substances were CFCs 11 and 12, wit growing use of HCFC-22.These two together account for 73% of the wo ld oz9ne depleting potential.In 1987, the Montreal Protocol called for a 5000 cut in CFC production bythe year 2000. In the year following the signin of the Montreal Protocol, aseries of events led the United States (US) and orne of the European govern-ments to renew their calls for a total CFC phas -out (Maxwell & Weiner,1993a). Two scientific reports contributed signi lcantly to the hardening ofpublic opinion against CFCs and thegeneratio of sufficient political willrequired to negotiate more stringent controls. nly two weeks after the Pro-tocol's signing, the first results became available from the US-led AntarcticAirborne Ozone Experiment. The results demo strated definitively the linkbetween chlorine and the hole (Kerr, 1987). T e second key report was thatof the NASNWMO Ozone Trends Panel. Publ shed in March 1988, it re-vealed unexpectedly large ozone depletions at iddle/high northern latitudesduring winter (Watson, Prather, & Kurylo, 198 ). Within ten days of thestudy's release, DuPont announced its plans to urtail. production of CFCs,and to speed the transition to alternative chemi als, underlining the techni-cal feasibility of a shift within the decade (DuP nt Corporate News ExternalAffairs, 1988). In August 1988, ICI announce its intention to join DuPontin an orderly phase-out of existing CFCs and i a rapid commercializationof substitutes. In June 1990, the London Revis.ons to the Protocol were sig-ned, establishing a timetable leading to total p ase-o+t ofCFC productionby the year 2000.

All countries which use or produce CFCs in ignificant quantities haveratified or accepted the Protocol, including Chi a, India, and the RussianFederation (United Nations Environment Prog amme, 1992). The Protocolwas amended in 1990 and in November of 1992" The principal limits arethat production and consumption ofCFCs 11, 12, 113, 114, and 115 andhalons are subject to specific limits based on 1 86 levels, which graduallydecrease until no production or use is permitte by 1 January 2000. Con-sumption and production of other fully haloge ated ~FCs are frozen at80% of 1989 levels from January 1993, declini g until they are fully phasedout by January 2000. Several countries have ins ituted stricter provisions thanthe Protocol., particularly in response to new sc entific evidence. In addition,the United States established an excise tax that ncreases rapidly over time onozone depleting substances manufactured or i ported for use in the UnitedStates (United States Environmental Protection Agen~y, 1991). Because the Montreal Protocol and its imple enti g legislation in the sig-

natory countries places bans on the production of C Cs, there are no restric-tions on the quantities individual user compan"es may use. In principle,companies could continue to use fully halogen ted CFCs well past the phase-out date. Despite this, many users are attempti g to reduce or eliminate

218

..

their use of all ozone destroying chemicals as rapidly and substam ally as pos-sible, in some cases going well beyond the regulatory requirement. Much ofthis activity is in anticipation to the rapid reduction of available FCs as thephase;..out is implemented by the producer industries. 'Pressure for change in-creased when several of the world's major producers voluntarily oved theirphase-out schedules up beyond the Protocol and more progressive nationaldeadlines (Maxwell & Weiner, 1993b; Oye & Maxwell, 1994).

Electronics Industry

At the time the Montreal Protocol was under negotiation, the ele~ roniCS in- dustry was one of the most vociferous opponents of mandatory b s. They

co~ntered the environment.aiistswith ~laims ~at there were no al e~native~available and that the costs Involved WIth phasIng out CFCs woul JeopardIzean extremely important and growing industry. For example, the AmericanElectronics Association gave evidence to the US House of Representatives inMarch 1987, saying "The electronics industry has a keen interestI oth in the

continued availabi~ityof this indispensable so~vent ~d in i~ssafe u e: ...while

near-term alternatIves to other CFCs appear IncreasIngly lIkely, th Industryis troubled that no suitable alternatives to CFC-113 appear likely in the nearfuture. ..The issue has profound and troubling implications for the U~; high-tech industry's international competitiveness and for international~trade" (United States House of Representatives, 1987).

In the following year, 1988, one of the world's major electroni manhfac-turers announced its commitment not only to drastic reductions i CFclconsumption, well in excess of requirements, but also to a total pHase-out inthe near future.. This commitment took place prior to a complete ban onCFCswas even negotiated. Many of the other major electronics companiesfollowed shortly afterwards. In the forthcoming sections, we will d!iscuss themajor uses of CFCs by the electronics industry and discuss technicfal pathschosen by actors in this industry. II I I

Overview of Electronics Industry. The principal regulated ozone depletingsubstances used by the electronics industry areGFG-, 113 and, to a lesser ex-tent, methyl chloroform and carbon tetrachloride. GFG-113 is one of theprincipal fully-halogenatedGFGs, with an ozone depletion potential of 0.8(with GFG-ll as abase of 1) and accounts for some 23% of all G Gs pro-duced in 1990 (United States Environmental Protection Agency, 1 91).Prior to the Protocol, the electronics industry made up almosthal of theglobal use of th~ s~bstance (Electronic Engin~er~ng Association, 1 87). I

The electrorncs Industry uses GFG-113 prIncIpally to remove fl fromprinted circuit boards, printed wiring assemblies, and other electro ic assem-

+ I 219

I

blies. The industry also cleans other equipme t such as disk drives, althoughthis makes up only a small proportion of all cl aning operations in the indus-try. Assemblies are cleaned either by CFC-ll spray, immersion in liquid, orimmersion in vapour. In general, the need for leaning increases with the levelof required reliability. Many simple electronic evices have never been cleaned.

A number of options for reducing CFC us have emerged in response tothe Montreal Protocol. These can be summari ed in four general categories:conservation and recovery, hydrocarbons, aqu ous cfeaning, and no cleantechnology: I

1. Conservation and Recovery. CFC solven s wer~ cheap prior to interna-tional controls on production (estimates vary, ut typical prices for CFC-113were around $1.60/Ib, excluding any excise tax s [United States EnvironmentalProtection Agency, 1988]) and made up a ve small part of the overall man-ufacturing costs of electronic equipment. The are also nontoxic and werethought to have no environmental effects. Thi means that in the past theytended to be used wastefully, and that substan ial reductions can be madethrough simple housekeeping methods. The nited States EnvironmentalProtection Agency (1990) reports that many c mpanies have reduced lossesby up to 70-85% through conservation and r covery programs. These reduc-tions allow companies to easily match the redu tions in the production sched-ules laid out in the Montreal Protocol and eve in St me of the most strictnational legislation.

2. Hydrocarbons. Companies have markete several hydrocarbons for semi-aqueous cleaning,2 but the most common are erpenes, chemicals which oc-cur naturally in most essential oils and in the leoresins of plants. They werefirst developed in 1983, for rosin flux removal from printed circuit boards byAT&T and Petroferm. AT&T has shown that they can be used to clean sur-face mounted circuit boards, which is essentia if they are to be long termalternatives to CFCs. Since they are not a dro -in replacement for CFC-113,companies using this method have to make su stantial equipment modifi-cations.

The advantages of terpene solutions stem fr m their workability in closespaces, and at room temperatures, as well as t eir being non-corrosive. Cer-tain potential problems or questions, however, are associated with terpene use..The effects of occupational exposure to these ubstances are currently un-clear. 3 Terpenes also have environmental draw acks. One problem concerns

the waste produced by using them in semi-aq eous cleaning. The secondconcern is that terpenes are volatile organic co pounds, molecules that evap-orate at their use temperature and lead to a p otochemical reaction causingatmospheric oxygen to be converted into trop spheric ozone under favourableclimatic conditions. Mists and vapours there£; re need to be controlled so asnot to contribute to the formation of photoch mical smog. Equipment also

-

220

Sandra R~nb~ J~es Maxwell- Industri~ ~e~n~t? the Banning of CFCs I!"

must be available to control for the odour and flammability of these bom-pounds. I~ I

.3. Aqueous Cleaning. Aqueous clear"ting ~s t~e longest-established ~lterna-tlve to halog~nated solvents for the electronIcs Industry, and many cor Panies have used thIs method to some extent for years. Aqueous cleaning ca be

used with both the conventional rosin fluxes and with water-soluble uxes,on a surface mount as well as through-hole assemblies.

The technique basically involves placing assemblies in an industri wash-ing machine, where they are sprayed with pure water or with a water basedmixture. When the industry began to look for alternatives to ozone epleti~gsubstances, water cleaning was thought to be greatly inferior to CFC 113 asit. has a higher surface tension, which would make it difficult to get aterinto small gaps in the boards, and even harder to get out again. In pr ctice,many companies have found that water can clean small gaps in both urfac~mount and through-hole assemblies, particularly with the help of ne nozzlesand high pressure/high volume sprays.

The advantages of this type of cleaning come from its occupationa safetybenefits and its process flexibility. Design, formulation, and concentr tioncan all be adjusted for cleaning different assemblies with different tyPFs ofcontaminants. This method, however, is not without its disadvantage. Inprocess terms there are problems because the equipment takes up mo e spaceon the factory floor and because this technique requires more careful ngi-neering and control than cleaning with CFC-solvent. In environment terms,the first problem comes from water itself being a scarce resource in anypans of the world, and the high purity water required o&en being ve ex-pensive. The need to dry assemblies makes energy consumption high~r usingthis method than using CFCs. This process also generates contaminatfd wa.$hwater, which has to be treated before discharged to sewers. Therefore, itmakes sense to recycle the waste water for economic and environmental rea-sons. Closed loop and zero discharge systems would make aqueous cl aningan ideal alternative from technical, economic, and environmental per pec-tives; yet these systems are, in general, not fully developed. Zero-disc argesystems are available for some, but not all, water-soluble flux applicati ns andare under development for rosin and organic acid fluxes. '

4. No-Clean Techniques. An option increasingly being chosen by c mpa-;nies committed to early phase-out dates is not to clean the boards at II. Theidea of leaving solder on assemblies is not new. Many manufacturers f "lowperformance" electronic goods have never removed solder residues. 0 thewhole, industry has tended to clean equipment which has to be very liable,or in which corrosive fluxes are used. Managers of plants producing h ghperformance products who are now changing their processes over to n -cleantechnology say that the option was "in the literature," but that there as no

221

Technology Studies 1997,4/2, @ 1998 Walter de Gruyter

incentive to investigate its applicability for their procets (R. Hartbauer, inter-view with author, May 19, 1992). New fluxes coming onto the market arestarting to change this. Typical fluxes contain 15-35% solids, whereas thenew types of flux contain as little as 1 % solids, which leave minimal residuesand therefore will have very little effect on product reliability.

This technique has the obvious advantage of elimin~ting one stage in themanufacturing process. This simplification not only reduces cleaning costs,but also increases efficiency as companies can test the reliability of the assem-blies ("bed of nails testing") immediately after wave soldering, rather thanhaving to wait until after cleaning and drying. This te

t hniqUe, however, may

not be appropriate for certain applications, particular! those with magnetic

components such as disk drives (Technology Business nd Environment Pro-gram, 1993). There is also some concern that, as asseIl1lbly technologies be-come finer and more compact, cleaning will become ecessary (S. Dulmage,interview with author, May 19, 1992).

The principal obstacle to wholesale adoption of no- lean, however, is nottechnical, but organizational. Traditional product testi g methods use clean-liness of the surfaces as a measure of reliability. Boards at have been solderedusing this method, obviously compare badly in this respect with those thathave been cleaned, even if there is no demonstrable difference in reliability.This method of reliability testing is currently being ch~nged. At the sametime as a need for a change in standards arises, compa ies need to changetheir customers' perceptions. Northern Telecom cited ustomer resistance asthe principal obstacle to adopting no-clean technolo (Technology Businessand Environment Program, 1993). This resistance has diminished substan-tially, however, as customers have begun to use equip ent where the boardshave not been cleaned and have seen that performance is not impaired.

Industry Responses. The first company to commit to f total CFC phase-out

was Northern Telecom, which announced in 1988 its ecision to phase-out

all ozone depleting substances by December 1991. Th company committedto all newly acquired sites being CFC free within 15 months of purchase, Toreach this goal, the company has implemented a numbcr of technical options.It considers, however, that no-clean fluxes are the best 4lternative for 80-90%of its through-hole technology boards. It estimates that this program will save$50 million over the next eight years. Northern Teleco was one of the ini-tiators of the Industry Cooperative for Ozone Layer Pr tection (ICOLP), agroup consisting of nine companies and the US Envir nmental ProtectionAgency (EPA). ICOLP is an industry-run information exchange systemthrough which companies can exchange technical info mation and ideas. Italso aims to aid the transfer of technology to assist the phase-out of CFC usein the electronics industry in developing countries.

-

222

Sandra Rothenberg, James Maxwell -Industrial Response to the Banning of FCs

AT&T reduced CFC-113 consumption to 50% of 1989 levels by 1991(1988 consumption was three million lbs.) and committed in 1989 to aphase-out by 1994 (D. Chittock, personal communication, April 18,1991).As with most other electronics firms, these reductions only apply to CFCsused in manufacturing, and do not include those used in refrigeration or airconditioning. AT&T announced that by 1994, it would stop accepting prod-ucts from suppliers who use CFCs. By 1991, the company had stopped ac-cepting foam packaging from outside sources that had been made with CFCs.AT &T has chosen terpenes on the whole, but is increasingly investigatingno-clean. Results from terpenes have been good, and AT&T has shown themto be the cheapest of the alternatives for their needs.

In 1989, IBM published a commitment to eliminate CFCs in its manu-facturingprocess by 1993 (use in 1988 was 113 million lbs). By the end of1991,20 of the 40 plants that had been using CFCs had stopped entirely,and emissions company-wide had reduced by 83% since 1987 (TechnologyBusiness and Environment Program, 1993). For the most part, IBM usesaqueous cleaning as an alternative to CFCs, with saponifiers and air dryers.IBM is currently reviewing the cost implications of changing over to no-cleantechnology, and estimates that the savings in operating costs, through avoid-ing buying solvent, at least outweigh the costs of investing in new capital

equipment.Digital Equipment Corporation began its phase-out program in 1988,

and had reduced consumption and use by more than 90% by 1992. Thecompany has chosen aqueous cleaning for most applications. It has workedwith an equipment supplier to develop a new nozzle, patented technologythat it has made available, free of charge through ICOLP. This is the firsttechnology that offers adequate performance with aqueous solutions cleaningsurface mount boards.

Summary. The response of the electronics industry was marked by two char-acteristics. First, despite protests to the contrary, within a year of the signingof the Montreal Protocol in 1987, in which only a 50% reduction by theyear 2000 was called for, the electronics industry was announcing a rapidphase-out ofCFCs. Lead by Northern Telecom and AT&T, the industry rec-ognized that the Protocol in 1987 was just the first step to an eventual ban.As stated by one industry representative, at that point, "the writing was onthe wall." Second, technological choices of the various companies within theindustry varied, as there were a number of technical options available forcompanies. The factors that lead to this approach will be discussed in com-parison to those in the automobile industry later in the article.

223

Technology Studies 1997, 4/2, @ 1998 Walter de Gruyter

The Automotive Industry

In response to the Montreal Protocol in the late 1980s, automobile manufac-turers began investigating alternatives to CFCs for use in automobile air con-ditioners. Like the electronics industry, the feeling t the time was that nooptions existed and that the new technology just c uld not be developed. Asone industry manager commented, "When we firs started discussing the op-tions for CFC-free air conditioners, I thought 'this .s impossible.'" A numberof chemical options, including blends such as CFC 12/DME, CFC-500, andHCFC-142b/22 were explored. While these mixtures appeared attractive forshort-term reductions in CFC-12, each one presented technical obstacles toimplementation, and the environmental and healt~ performance characteris-tics were less than ideal. !

Two chemicals emerged as the most feasible alternatives.to CFC-12. Onewas HCFC-22, a chemical which has only 5% of the ozone-depletion poten-tial ofCFC-12. The other was HFC-134a, a lesser known chemical with zeroozone depleting potential. Spurred by legislation in Sweden that banned theuse of CFCs by 1995, Swedish manufacturers, Volvo and Saab, spent a sig-nificant amount of time investigating HCFC-22 in automobile air-condi-tioning systems (Keebler, 1989). When it seemed that HCFC-22 wouldemerge as the chosen substitute, some companies took the initiative to pro-pose new alternative technologies using the coolant. Ultimately, however,auto manufacturers were reluctant to undertake the extensive and costly en-gineering changes that were needed for HCFC-22. Because of HCFC-22'sozone depleting potential, many feared future reguJatory controls on thesubstance (Keebler, 1989). i

The automobile industry eventually turned its attention to HFC-134afor use as the alternative coolant. The most attractive feature of HFC-134awas that it had zero owne depleting potential (ODP), compared to ODPof 1.0 for CFC-12 (Mo?treal P~otocoI198~ Refri~eration ~r Conditioningand Heat Pumps Techmcal Options Committee, 1V89). While there wassome fear of regulatory controls on HFC-134a due! to its global warmingpotential, industry officials felt that ozone depletioti was a more tangible andimmediate threat, and there were no other alternatiVes at that time with zeroODP. Data revealing that HFC-134a had global w1lrming potential (approxi-mately one tenth that ofCFC-12) further reducedlthese concerns (Bateman,1992). '

The most immediate concern with HFC-134a ~as the uncertainty sur-rounding its availability. In 1988, however, two large chemical manufacturers,DuPont and ICI, began to accelerate their commercialization schedules tomeet the needs of the auto industry, and HFC-131a became the best alter-native to meet the pressing CFC reduction schedu~.

-

224

~n~dra Rothenbe~~,J!!:~_es Maxwell -I~_u~ri_al Responset?~~~B~nning of ~~.-

Once it had been established that HFC-134a could be supplied iiin adequatequantities, automobile manufacturers and their suppliers began w~rking moreaggressively on redesigning the air-conditioning system to run on ithis substi-tute. Unlike the electronics industry, very little of this technology ihad beenresearched at any great depth prior to the Protocol. Using HFC-13VIa requireda redesign of the entire air-conditioning system. Selection of a .lubticant com-patible with HFC-134a presented the industry with one of the m<j>st prob-lematic challenges in the development of the new air-conditioning system. In~dditi?n, differences in the chemical properties required that systetns operat-mg wIth HFC-134a have new seals, hoses, fans, and more powerfiill com-pressors in order to reach the same level of cooling performance as CFC-12(Yamanda, Sonoda, & Arakawa, 1992).

Unlike other industries affected by the CFC ban, the replaceme~t of CFC-12 in automobile air-conditioning units, presented the automobile industrywit? technical p~oblems that reached much further than just chan$ing thedesIgn of the UnIt. Many of these problems stemmed from the fac~ that thenew units, in order to reach the same level of performance as their, predeces-sors, usually needed larger and heavier components. For most man~facturersthis meant an entire redesign of the front end of the automobile, which isusually extremely tight for space, in order to maintain the same shape. In ad-dition, since reducing vehicle weight is one of the main considerations whenincreasing automobile fuel efficiency, the weight of the new units, whichwere as much as two pounds heavier than the old units, was also of primaryconcern. In an industry where efforts are made to reduce vehicle weight byless than an ounce, introduction of the heavier air-conditioning units meantreducing the weight of the vehicle through material substitutions. Furthercomplicating the issue, each change in vehicle design required a subsequentalteration in production facilities. These changes, although a famili,.r occur-rence in automobile assembly plants, nonetheless presented costly obstaclesto change.

Companies had varied success in designing air-conditioning systems whichwere small and light enough to reduce the impact on total vehicle dbsign. Forthe most part, the new air conditioners were both heavier and largeir. Oneexception in this regard was a system already being looked into by ~issan forefficiency reasons, in which a unique parallel condenser allowed th(:jm to de-sign a unit which was the same size as their old air-conditioning unit, reduc-ing the impact on total vehicle redesign (Technology Business and Environ-ment Program, 1993). Air conditioners based on this and other des~gns start-ed to emerge from the industry and its suppliers, making new air co~ditionersboth smaller and lighter than their CFC-based counterparts. For th~ mostpart, most companies are now utilizing similar technologies in their' newCFC-free air conditioners.

+-~25

Technology Studies 1.997, 4/2, @ 1998 Walter de Gruyter

Companies started to introduce CFC-free air conditioners in 1991, threeyears after the Montreal Protocol was passed. Within tpe industry, somecompanies have been slower than others to phase-out CFC-based air condi-tioners from their products. Early movers in this transition were Volvo, Saab,and Mercedes, with Mercedes producing the first vehicle with the new cool-ant; most other producers started the transition in 1992. The length of thetransition time varied among companies. While some European and Japanesemanufacturers planned to be completely phased out by production year 1993,many producers, such as Toyota, Volkswagen, and Chrysler targeted 1994 astheir phase-out deadline. American producers, GM and Ford, initially settheir phase-out deadline to be 1995. In light of the new US deadline andimpending CFC shortages, several producers, including Chrysler, Toyota,and Ford, have moved up their phase-out schedules by one year. The EPAnow estimates that approximately 90% of all new motor vehicles will haveCFC-free air conditioners by 1994 (L. Nirk, telephone coversation withBen Jordan, June 21, 1993). A complete phase-out was expected by 1995

(Cavenaugh, 1993).

Servicing of Old Vehicles: A Costly Problem. Aside from the expense of de-signing the new air-conditioning systems, the cost of new units is not signifi-cantly higher than the ones built to run on CFC-12 (Wald, 1993). Instead,the primary costs of the phase-out to automobile consumers stem from theservicing of air-conditioning systems that run on CFC-12. For automobilesunder warranty, these costs will likely be borne by automobile companies.The sources of these costs are twofold. First, the price of CFC-12 is rapidlyrising due to its high demand and lowering supply; CFC-12 has already beensold for as much as $20 a pound in areas such as California and Arizona(Keebler, 1993). In the United States, a $3.35/pound excise tax added in1993 contributes to this number (McMurray, 1992). Second, as the supplyof CFCs becomes scarce, the industry and consumers will face significantcosts for retrofitting obsolete air conditioners.

The problem arises from the fact that approximately 80% of the CFCsused as automobile air conditioner coolant is consumed in the service of oldsystems (Montreal Protocol 1991 Refrigeration Air Conditioning and HeatPumps Technical Options Committee, 1991). The United Nations Environ-ment Programme has estimated that, assuming a full global implementationof an alternative coolant in 1995, the demand for CFC-12 will peak in 1994due to the servicing of old vehicles. In 2005, however, a large number of ve-hicles will still require CFC-12 for service (Montreal Protocol 1991 Refrigera-tion Air Conditioning and Heat Pumps Technical Options Committee, 1991).

Recycling ofCFC-12 had been identified as the primary method ofsus-taining the CFC-12 supply for the servicing of existing air-conditioning sys-

c

-226

~~~~thenberg, Jam~~ ~~~=-~~~ustrial Respo~~~~~~~n_~in~ of Cf~s

terns. Although recycling CFCs is not mandated by the Montreal Protocol,several countries have recognized the benefits of sustaining the existing sup-ply ofCFCs and are regulating CFC recycling (Montreal Protocol 1991 Re-frigeration Air Conditioning and Heat Pumps Technical Options Committee,1991). Initial recycling efforts by the automobile industry and suppliers werepushed by this type of regulation. The costs of recycling equipment are sub-stantial. Depending on the level of sophistication, a service station may payfrom $1500 to $5000 to acquire one of these machines. Costs are recoveredby some service shops by charging a service fee for a CFC refill. Due to theskyrocketing costs of CFCs from taxes and a diminishing supply, the manu-facturers of recovery/recycle and recovery-only equipment have contendedthat most of the machines will pay for themselves in a month (Gardner &Baker, 1992). This type of overwhelming financial incentive is pushing rapidadoption of CFC recycling.

Although recycling will helpmeet the demand for CFC-12, it cannot berelied upon to supply all future CFC-12 needs. Spot shortages have alreadyoccurred in the United States as the production of CFC-12 is being phasedout; larger shortages are predicted for future years. Because of the CFC short-age, the Mobile Air Conditioning Society has estimated that as many as 100million vehicles will need to have their air conditioners serviced without theuse ofCFC-12 (Keebler, 1992). Two options are being explored to addressthis probJem. One is the development of a "drop-in" substitute for CFC-12,which would eliminate the obsolescence of the old air-conditioniqg units.This option has met with significant resistance by many in the automotiveindustry because of the ability of these drop-ins to contaminate the supply ofthe other coolants. The fear is that mechanics, not knowing a substitute re-frigerant has been used, will suck the refrigerant into their recycling machinesand contaminate the supply of existing coolants, destroying air conditionersthat run on these coolants (Laugesen & Spencer, 1992).

In the absence of a "drop-in" substitute, automobile owners will have topurchase retrofit technology to enable their air conditioners to run on HFC-134a. Retrofitting is a major concern to the automobile industry becausethe number of vehicles requiring this technology is great and the associatedcosts to them are high. The estimated cost of retrofitting technolog1j, for thosevehicles that have retrofit kits available, varies between $200-1000 for eachvehicle. The costs largely depend on the level of technology of the existingvehicles. Newer vehicles, such as 1992 or 1993 models, will be less expensiveto convert since the air conditioners were designed with technology such asless permeable seals and barrier hoses that could withstand the demands ofR-134a. Older cars will be at the higher end of the price range since morecomponents will need to be changed.

'---, 227

Technology Studies 1997, 4/2, @ 1998 Walter de Gruyter

The EPA has estimated that because of the retrofit problem, pushing thedeadline up from the year 2000 to 1997 for the CFC phase-out, cost theautomobile industry approximately $2.8 billion. These costs were even higherwhen the deadline was pushed up to 1995. The magnitude of the retrofitproblem forced theEPA to ask DuPont to extend its production ofCFC-12to service the millions of air conditioners in existing vehicles. As stated byone consUltant, "the EPA looks like 'the bad guy'. They're begging DuPontto make it [CFC-12]" (Kirschner, 1994).

Summary. In making the transition to systems that can run on the new cool-ant material, the automotive industry moved relatively slowly as compared tothe electronics industry.. Companies started phasing out old technologies in1992, and a complete phase-out was not achieved until 1995. In comparisonto the electronics industry, there were higher costs associated with technologydevelopment. To reduce these costs, many firms undertook a slower pace ofchange. The need for retrofit technologies for automotive air conditioners,however, made these delays a costly option as well. The retrofit problemcould cost the automotive industries, and ultimately, consumers several bil-lion dollars. While most firms have utilized similar technologies to replaceCFC-based air conditioners, they have implemented different schedules forthe phase-in of new air-conditioning units in their automobiles. In the nextsection we will analyze the response of the automobile industry, in compari-son to the electronics industry, and discuss the factors that have contributedto the responses of each industry, as well as firms within these industries.

Discussion

Overall, the Montreal Protocol and its subsequent amendments representa successful example of technology-forcing regulation. Despite initial industryresistance, the transition from CFCs has been achieved with great success inalmost every case. Once faced with inevitable controls on CFCs, the electron-ics industry found ways of reducing consumption, and either developed newalternatives or adapted other cleaning methods they had already been using.In the automobile industry, where the alternatives seemed less aq:ractive, com-panies modified existing technologies and are coming up with even higher-performance CFC-free products.

Both industries eventually shifted to CFC alternatives. In neither industrydid the phase-out of CFCs require radical breakthroughs in technology.Technology-forcing regulations induced both industries to reexamine theirexisting technical options and make improvements. In both industries, thecosts of making a transition to alternatives were not prohibitive. For auto-mobile companies, new technologies initially represented only a small added

-228

cost to the total cost of the vehicle. Many firms in the electronics industryeven saved money by shifting to alternatives.

The two industries differed in the rate and nature of their responses to thetechnology-forcing regulation. Technical differences help to explain some,but not all, of the differences in the responses of the two industries. For theelectronics industry, most of the technologies existed, even if only in the re-search stage. For the automobile industry, the technical challenge was perhapsmore complex and depended on the availability of HFC-134a. In addition,as part of an inherently complex product, the replacement ofCFC-12 in au-tomobile air-conditioning units, presented the industry with technical prob-lems beyond design changes in the unit itself.

Existing models of technical innovation would predict the significance ofthese and other technological characteristics as they exist along the productlife cycle, for understanding the path of technical change (Tushman & Rosen-kopf, 1992; Ashford & Heaton, 1983). For the most part, however, thesemodels do not account for the tight coupling among technological, organiza-tional, and institutional factors that occurred in this case. First, long-standing

,organizational routines within each industry played a role. The electronicsindustry is characterized by innovation, short product cycles, frequent processchanges, and short product lead times. Firms respond quickly to technicaland market changes and their organizations are designed to facilitate continu-ous innovation. Because they were already exploring new technological op-tions, many companies chose the technical solution that they had been usingbefore the environmental concerns arose. Existing technologies created cog-nitive frameworks and behavioural routines that supported the pursuit ofcertain technical options in firms. Digital Equipment Corporation, for exam-ple, has been using an aqueous system for cleaning since 1974, and has nowdeveloped a new aqueous technology to cope with the problems associatedwith cleaning surface mount boards.

Second, there were also influences of corporate environmental strategy.The use of ozone-depleting substances challenged the electronics industry inthe late eighties and managers saw the advantages of moving ahead of regula-tions. Despite uncertainties resulting from varied and shifting national phase-out schedules, many of the electronics firms felt that it was more effective totake voluntary action than be constrained by a future regulation over whichthey might exert little control. Many of the firms in the automobile industryperceived similar regulatory stimuli from a different vantage point. As late as1992, GM, Chrysler, and Ford showed surprise at the Bush administration'sannouncement regarding a new phase-out timetable in the United States(Wald, 1993). Even after the Bush announcement, American automobilecompanies hesitated. They still held the view that allowances would be madefor the servicing of old vehicles. More specifically, they believed that an ex-

-229

Technology Studies 1997,4/2, @ 1998 Walter de Gruyter

tension would be made for CFC production for use in automobile air con-ditioners as an "essential" use of the substance. In 1992, these companies stillrequested more specific guidelines on how to proceed with the phase-out(Technology Business and Environment Program, 1993).

This response of the automobile industry emerged from a long history oftreating environmental regulations as a constraint rather than an opportunityfor innovation; the automobile industry found it difficult to modify theirunderlying approach to environmental regulation. Following Roberts andBluhm's (1981) work on electric utilities, rather than being "positively re-sponsive," companies within the industry seemed to engage in "optimisticself deception." Firms institute unique strategies with respect to environmentalregulations; these strategies are strongly influenced by the organizational his-tory and past experiences of its managers!

Related to overall environmental strategy, implementation of CFC alter-natives in the two industries was also shaped by the level of managementsupport and resources. This factor is especially important for environmentalinitiatives, which often do not yield high rates of return on investment. Someof the faster-moving companies in the automobile industry sought to elimi-nate CFCs as part of their corporate commitment to improving environ-mental performance, and resources were allocated to allow design teams tomeet stringent internal CFC deadlines. In several automobile companies,however, although senior environmental managers decided to move early,their efforts were often hindered by a lack of internal support. Other environ-mental issues were considered far more pressing. This was particularly prob-lematic since CFC dependent products and processes were not regulateddirectly and no strict deadlines existed for the phase-out by user industries.Environmental managers in these automobile companies indicated that theyhad encountered resistance from top management when they sought to de-vote the vast resources that would be needed to phase-out CFCs at a rapidpace. It was only when the potential for a costly retrofit problem became evi-dent that a rapid phase-out schedule was endorsed by these companies. Eventhen, environmental mangers struggled to convince top management to ad-dress this critical issue.5

In the electronics industry, image, cost, the availability of CFC 11 and 12,and concern for the environment, were cited as reasons for a more rapidtransition to alternatives than that mandated by regulation. In the electronicsindustry, senior management in a number of companies committed early toa phase-out date. This commitment gave plant managers an incentive to im-plement changes and to innovate on their own.

In addition to these organizational factors, institutional factors help to ex-plain the differences observed among firms within each the two industries.Regional differences in policy development and the general climate towards

-230

~~~h~nber&,)~~~_~~I-::: Industri~~~:_e to the ~~~cl~c~

environmental issues influenced the implementation schedules of variouscompanies. European automobile companies were encouraged to move swift-ly by the aggressive Swedish and German CFC deadlines that were set earlierthen those established in the Montreal Protocol and its amendments." TheUnited States, on the other hand, waited until 1992 to change its phase-outdate to 1995. As discussed above, this no doubt had a profound impact onthe responses within the industry, as all three American companies based theirinitial schedules on the year 2000 phase-out date established in the 1990London amendments and the 1990 Clean Air Act Amendments.

The major theoretical implication of the cases outlined in this article isthat differences between the two industries cannot be explained by modelsdependent on placing a technology within its overall life cycle. Instead, wefound that industry responses to regulatory stimuli were shaped by complexinteractions among a number of technical, as well as organizational and in-stitutional forces. Managers in the two industries perceived the ozone deple-tion problem differently and reacted accordingly. Senior management in theelectronics industry recognized the salience of the issue, and framed the issueas another technological problem demanding a solution. Environmental man-agers in the automobile industry saw the regulations as a constraint. Theyalso had difficulty obtaining the attention and support of senior managementfor more rapid changes. In addition, institutional factors help to explain thedifferent implementation paths taken by individual firms within the two in-dustries.

Implications for Researchers

1. This experience highlights the challenge for researchers in gaining a deeperunderstanding of the interaction between regulation as a stimulus for changewith the cognitive, organizational, and institutional factors influencing firms'response to technology-forcing regulation. More research is needed to under-stand which organizational and institutional factors are important, how theyinteract, and under what circumstances they are salient. From their work onelectric utilities, Roberts and Bluhm (1981) hypothesize that if a regulatoryagency provokes a crisis situation in a firm, organizations can become "rigid."This rigidity encourages the firm to rely on established routines that act as abarrier to learning, and this needs to be explored more systematically acrossfirms and industries. Other hypotheses about organizational and institutionalfactors also need to be explored.

As discussed earlier, Tushman and Rosenkopf (1992) suggest that organi-zational and social factors have the most influence during periods of experi-mentation. This case builds upon this hypothesis further by suggesting thatsocial factors, such as regulation, are more likely to be responded to progres-

231

Technology Studies 1997,4/2, @ 1998 Walter de Gruyter I, I

sively by industries or firms characterized by an "experimental" culture. Fur-ther work in this area is clearly needed to suggest when various factors maybe more significant than others for the process of technical change.

2. Another issue underlying the different responses to regulatory stimuliin the automobile and electronic industries is the importance of attentionallocation. As seen in this case, while regulations are designed to force organ-izational attention to be devoted to environmental innovation, they do sowith varying degrees of success. Understanding how top managers focus at-tention on the numerous and competing issues facing their organizationsand then subsequently manage attention allocation within the rest of theorganization, is a critical area for future research (Ocasio, 1995). An under-standing of attention allocation within firms can lead to more useful guide-lines for effective environmental management within firms.

Recommendations for Policy Makers

To ensure environmentally responsible innovation, policy makers need amore in-depth understanding of the industry and firms being affected byregulation. Similar to other broad-based policies, such as tax, antitrust, andpatent policies, environmental policies such as chemical bans will influencefirms and technology development within these firms in very different ways(Nelson & Langois, 1983). Several key policy issues emerge from the abovediscussion.

1. This article illustrates that regulations can act as powerful stimuli totechnical innovation, but in this case, firms faced the additional impetus torespond from market-based regulatory policies. The US excise tax on CFCsforced firms to internalize the cost of CFCs to the environment, and gaveenvironmental managers an argument for action that aligned with other in-dustry objectives. Thus, the case suggests that technology-forcing regulationsare likely to work more effectively when combined with financial incentives.

2. Industry's response to the banning of CFCs demonstrates the lack offit that can occur between regulatory policy and industrial product cycles, aproblem of particular salience when regulations apply to several industrieswhich vary in their technical capabilities and planning cycles. When govern-ment officials establish regulations that are inconsistent with product life cy-cles, they run the risk of raising the cost of compliance to industry and stif-ling needed innovations. One approach to this issue is to devise regulatoryapproaches targeted to particular firms and industries. This approach canhelp to reduce costs and enhance flexibility in achieving environmental goals.The EPA's 33/50 program, which negotiated agreements with particularfirms, and the (ICOLP) are recent examples of quasi-regulatory programstargeted to specific firms and industries.

--232

c~ ~

?~dra ~9~~' )a~~~~_~well -Industrial Response to the Banning of gFCs

Issues for the 21 st Century

Our research suggests that technology-forcing regulations can b~ effectivewhen they do not require major technical breakthroughs and when the costis modest. In cases requiring large technical advances and signifij:am capitalcosts, they are likely to meet with stiff resistance, yet the organi'Z'.ational char-acteristics of the industry being regulated are important influences on theprocess of innovation. So too are the industry's or firm's regulatqry historyand standard operating procedures toward environmental regulation. Indus-try's response to the next wave of environmental regulation is likiely to beshaped by the character of industrial innovation, as well as the key organiza-tional characteristics of particular firms and industries.

Endnotes.1. Funding for this research was provided by the MIT International Motor VeJjlicle Program,

Massachusetts Institute of Technology Norwegian Chlorine Study, and the MIT Technolo-gy, Business and Environment Program. The authors would like to thank S*ford Weiner

, for his help in designing the overall study on the phase-out of CFCs. We w~uld also like to

thank Julia Bucknell for her help studying the electronics industry andWendyVitr for herassistance in research on the automotive industry.

2. DuPont, for example, markets a non-terpene semi-aqueous hydrocarbon blend called"AxareI38."

3. The US Halogenated Solvents Industry Alliance (HSIA) has classified them as of "hightoxicity", yet Petroferm, the manufacturers of one of the leading brands, say~ BioAct EC7Rcan be used without harm to workers. United Nations Environment PrograIi1me (1989)and the United States Environmental Protection Agency (1991) report that .here has notyet been sufficient toxicity testing to resolve the issue.

4. Looking at the industry responses to recycling legislation emerging in Europe and more re-cently in the United States, while the automobile industry is attempting to t1ndertake astrategy of "positive responsiveness," they are struggling to do so with the saJ1ne organiza-tional structures and cultures. As a result, one can see similar patterns of reswnse beginningto emerge in the electronics and automobile industry in the area of product ~ake-back andrecycling (MIT Conference, The Greening of Durable Products: What's the Best Route?,March 22-23, 1993; MIT Conference, The Greening of Durable Products: Improving Co-ordination in the Wotld of Recycling, October 3-4, 1994).

5. US industry executive, Interview with Author, 5 May, 1993.

6. Currently, in Europe where concern over global warming is more pronounced, some Ger-man refrigerator manufacturers are indicating that they will be going directl~ to propane-based refrigerants, and skip the "middle step" of HFC-.1 34a (Air Conditionitilg Heating &Refrigeration News, 1993).

233

Technology Studies 1997.4/2. @ 1998 Walter de Gruyter

References

Abernathy, W J., & Utterback, J. M. (1978). Patterns of industrial innovation. TechnologyReview, 80(7), 40-47.

Air Conditioning Heating & Refrigeration News. (1993). So much for ODP: Here comesGWP; Ozone depletion potential, global warming potential of refrigerants. Air Condition-ing, Heating 6- Refrigeration News, 190(7), 16.

Anderson, P., & Tushman, M. (1990). Technological discontinuities and dominant designs: Acyclical model of technological change. Administrative Science Quarterly, 35, 604-633.

Ashford, N. A. (1993). Understanding technological responses of industrial forms to environ-mental problems: Implications for government policy. In K. Fischer & J. Schot (Eds.),Environmental strategies for industry, (pp. 277-307). Washington, DC: Island Press.

Ashford, N. A., Ayers, C., & Stone, R. F. (1985). Using regulation to change the market forinnovation. Harvard Environmental Law Review, 9, 419-466.

Ashford, N. A., & Heaton, G. R. (1983). Regulation and technological innovation in thechemical industry. Law and Contemporary Problems, 46(3), 109-157.

Bateman, D. J. (1992). Current Development Status ofHFC-134a for Automotive Air Con-ditioning (Technical Paper Series No. 900213). Society of Automotive Engineers.

Bijker, W E., Hughes, T. P., & Pinch, T. J. (1987). The Social construction o/technologicalsystems: New directions in the sociology and history o/technology. Cambridge, MA: TheMIT Press.

Cavenaugh, J. (1993). CFC scramble heats up; Chlorofluorocarbon price increases encouragethe adoption of replacement refrigerants; Aliphatic Organics. Chemical Marketing Report~244(12), 5.

DiMaggio, P. J., & Powell, W. W (1991). Introduction. In W W Powell & P. J. DiMaggio(Eds.) , The new institutionalism in organizational analysis, (pp.I-38). Chicago: The Uni-versity of Chicago Press.

Dosi, G. (1982). Technological paradigms and technological trajectories. Research Policy, 11:147-162.

DuPont Corporate News External Affairs. (1988). DuPont Policy Statement: CFCs. DuPont.24 March.

Electronic Engineering Association. (1987). European Industry Position paper on CFC 113.September.

Gardner, D., & Baker, A. (1992). CFC busters. Design News, August 24.76.

Henderson, R. (1994). Of life cycles real and imaginary: The unexpectedly long old age ofoptical lithography (Working Paper, Sloan School of Management No. WP #3661-94).Massachusetts Institute of Technology. February.

Hughes, T. P. (1987). The evolution of large technical systems. In W E. Bijker, T. P. Hughesand T. J. Pinch (Eds.), The Social Construction o/1echnological Systems, (pp. 51-82).Cambridge, MA: The MIT Press.

Hughes, T. P. (1988). The seamless web: Technology, science, et cetera, et cetera. In B. Elliott(Eds.), 1echnology and social process, (pp. 9-19) Edinburgh: Edinburgh University Press.

Keebler, J. (1989). Front-runner R-22 loses in auto refrigerant race. Automotive News,July 24, 24.

Keebler, J. (1992). Environmental issues dominate SAE shpw. Automotive News, 8 March, 1.

-

234

Keebler, J. (1993). Phase our ofR-12 chilling prices of used cars. Automotive,News, 29March, 16.

Kemp, R., & Soete, L. .(1992). The greening of technological progress: An e~olutionaryperspective. Futures, June, 437-457.

Kerr, R. A. (1987). Winds, pollutants drive ozone hole. Science, 238: 156-15t3.

Kirschner, E. (1994). DuPont extends CFC production after EPA warns of s<i:arciry.Chemical Week, 12 January, 18.

Laugesen, W, & Spencer, P. L. (1992). Trouble for your car's air condirioner. Consumers'Research, August, 10-13.

Maxwell, J., & Weiner, S. (1993a). Green consciousness or dollar diplomag:The British re-sponse to ozone depletion. International Environmental Affairs, 5: 19-41.

Maxwell, J., & Weiner, S. (1993b). Learning the right lesson: DuPont's expe~ience withCFCs. Paper presented at The Association of Public Policy Analysis and M~nagement,Washington DC.

McMurray, S. (1992). Air-conditioner firms put chill on plans to phase-out use of chloro-fluorocarbons, Wall Street Journal August, 10-13.

Montreal Protocol 1989 Refrigeration Air Conditioning and Heat Pumps Teqhnical OptionsCommittee. (1989). Technical progress in protecting the ozone layer: Refri~eration, air con-ditioning and heat pumps technical options report. UNEP. June 30.

Montreal Protocol 1991 Refrigeration Air Conditioning and Heat Pumps Technical OptionsCommittee. (1991). Montreal Protocol 1991 assessment report of the refriteration, air con-ditioning and heat pumps technical options committee. UNEP. December,

Nelson, R. R., & Langois, R. N. (1983). Industrial innovation polig: Lesso~ from Americanhistory. Science, 219: 814-818.

Nelson, R. R., & Winter, S. G. (1977). In search of useful theory of innovation. ResearchPoliry 6: 36-76.

Nelson, R. R., & Winter, S. G. (1982). An evolutionary theory of economic change. Cambridge,MA: The Belknap Press of The Harvard Universiry Press.

Ocasio, W. (1995). The enactment of economic adversiry: A reconciliation of theories of fail-ure-induced change and threat-rigidity. In L. L. Cumings & B. M. Staw (Eds.), Research inorganizational behavio1; 17., 287-331.

Orlikowski, W J. (1992). The duality of technology: Rethinking the concept of technology inorganizations. Organization Science, 3(3): 398-427.

Oye, K., & Maxwell, J. (1994). Self interest and environmental management. Journal ofTheoretical Politics, 6(4): 593-624.

Roberts, M. J., & Bluhm, J. S. (1981). The choices of power: Utilities face the environmentalchallenge. Cambridge, MA: Harvard University Press.

Rosenberg, N. (1969). The direction of technological change: Inducement mf;chanisms andfocusing devices. Economic Development and Cultural Change, 18(1): 1-24.

Rothwell, R. (1981). Some indirect impacts of government regulation on ind~strial innova-tion in the United States. Technological Forecasting and Social Change, 19: 57-80.

Schot, J. W, (1992). Constructive technology assessment and technology dyn.mics: The caseof clean technologies. Science, Technology and Human Values, 17(1): 36-56.

235

.

Technology Studies 1997,4/2, @ 1998 Walter de Gruytcr

Technology Business and Environment Program. (1993). Dimensions of managing chlorinein the environment: Report of the MIT/Norwegian Chlorine Policy Study. Center forTechnology, Policy and Industrial Development, MIT. March.

Thomas, R. J. (1994). What machines can't do: Politics and technology in the industrial enter-prise. Berkeley: UC Press.

Tushman, M., & Nelson, R. (1990). Introduction: Technology, organizations and innovation.Administrative Science Quarterly, 35: 1-8.

Tushman, M. L., & Anderson, P. (1986). Technological discontinuities and organizationalenvironments. Administrative Science Quarterly, 31: 439-465.

Tushman, M. L., & Rosenkopf, L. (1992). Organizational determinants of technologicalchange: Towards a sociology of technological evolution. Research in Organizational Behavio1;14: 311-347.

United Nations Environment Programme. (1989). Electronics, degreasing and dry cleaningsolvents, technical options report. June.

United Nations Environment Programme. (1992). Status of Ratification/Accession/Accept-ance/ Approval of the Montreal Protocol. (14th Issue). UN Office of Legal Affairs. June.

United States Environmental Protection Agency. (1988). Assessment of regulatory impacts ofthe phase-out of CFCs and halons. Annual Report.

United States Environmental Protection Agency. (1990). Manual of practices to reduce andeliminate CFC-113 use in the electronics industry. March.

United States Environmental Protection Agency. (1991). Alternatives for CFC-113 andmethyl chloroform in metal cleaning. June.

United States Environmental Protection Agency. (1991). The alternative fluorocarbons en-vironmental acceptability study. November.

United States House of Representatives. (1987). Hearings before the Subcommittee on natu-ral Resources, Agriculture, Research and Environment of the Committee on Science, Spaceand Technology.

Utterback, J. M. (1974). Innovation in industry and the diffusion of technology. Science, 183:620-626.

Van den Belt, H., & Rip, A (1987). The Nelson-Winter-Dosi model and synthetic dyechemistry. In W. E. Bijker, T. P. Hughes, & T. J. Pinch (Eds.), The social construction oftechnological systems, (pp. 135-158). Cambridge, MA: The MIT Press.

Wald, M. L. (1993). Car need freon? Better fill up soon. The New York Times, C1-C2.February 18.

Watson, R.T., Prather, M., & Kurylo, M. J. (1988). Present state of knowledge of the upperatmosphere in 1988: An assessment report. March 15.

Yamanda, A., Sonoda, Y., & Arakawa, Y. (1992). Development of an automotive air con-ditioning system using the HFC-134a refrigerant (Technical Paper Series No. 920216).Society of Automotive Engineers. February.

-

236