Ozone Depleting Substances - CiteSeerX

Ex-post estimates of costs to business of EU environmental policies: A case study looking at Ozone Depleting Substances Final Version: June 2006

By Robin Vanner Reviewed by Professor Paul Ekins

This report was commissioned by: European Commission, DG Environment, Unit G.1 Sustainable Development & Economic Analysis, under a framework contract No ENV.G.1/FRA/2004/0081. The contents and views contained in this report are those of the authors, and do not necessarily represent those of the European Commission.

2

Contents 1 2 3 4 5 6 7

Summary results.......................................................................................................... 4 Introduction................................................................................................................. 6 2.1 Policy introduction.............................................................................................. 6 2.2 Introduction to the literature ............................................................................... 6 Historical analysis....................................................................................................... 7 3.1 Pre-implementation............................................................................................. 7 3.2 Post-implementation ........................................................................................... 8 Distributional effects................................................................................................... 8 4.1 Distributional effects within the production chain.............................................. 8 4.2 Geographical distributional effects ..................................................................... 9 Conclusions............................................................................................................... 10 References................................................................................................................. 12 Annex I - Proposed plan of work.............................................................................. 14

3

1 Summary results 1.1 Scope and time frame This case study addresses the international phasing out of Ozone Depleting substances (ODS) under the Montreal Protocol and EU Regulations 3093/94/EC, and 2037/2000/EC. It compares ex ante cost estimates made at the end of the 1980s and in the early 1990s with ex post estimates from the late 1990s and later.

1.2 Differences between ex ante and ex post estimates and their causes This case study provides an early example of how the cost estimates provided by industry in response to proposed (ex-ante estimates) legislation can over estimate the costs which were later generated (ex-post estimates). This report has found that for this case study, the comparison factor (ex-ante/ex-post) can range from at least 1.4 at an overall macro level to 40 for individual case studies, and as high as 125 for administrative costs of compliance. Any such result is complicated by a range of methodological drawbacks, not least the effect of inflation. One way around this problem is to use a like-for-like material based comparison, i.e. ex-ante and ex-post estimates of how many times more expensive will a substitute be relative to the ODS they are designed to replace. Using this approach, the comparison of before implementation of the Montreal Protocol and a 1998 real market cost differential for the substitution CFC 11 and 12 with HFC 134a generates a comparison factor of 2.5. Further down the product chain for consumers, the ex-ante estimates of prices of their products due to the Montreal protocol differed by a factor of about 1.25 (see table 1). Any comparison of ex-ante and ex-post cost needs to be done with reference to a counterfactual reference scenario (i.e. what would have happened if the legislation was not implemented). ODS is a case study where the counterfactual can be assumed to have been the continued use of cheap CFCs at similar low prices. SEI (1999) reports that as the sectors CFC patents had expired, they were an established technology with little available profit. However, extensive research of the literature has not found any suggestion that the industry had any intention or incentive to develop alternatives in the absence of a regulated phase out.

4

Table 1 - Summary of ex-ante and ex-post costs estimates Case study

Ex-ante

Total CFC phaseout in the US

50% @$3.55/kg (1988)

Cost to consumers

20-25% (Reported in SEI 1999)

Social cost of reduction

Costs of substitutes Reporting and record keeping Foam blowing plant- interim use of HCFC-22 Rate of cost reduction

50 % $2.7 billion

10 times CFCs (Industry) $300 million (1988) Up to $2m

Ex-post

Cost ratio Key driver of difference (ante/post) Macro cost estimates The 1988 RIA estimate was for a 50% 100% phase-out over 10 years, the 1992 @$2.20/kg >1.6 estimate was for a 100% phase-out over (1992) 4 years Little or no Proportional increase in consumer cost ~1.25 product prices. Based on low ex-ante (UNEP estimate 1995) In 1988, the EPA estimated that the social cost of a 50% reduction by 2000 100% >1.4 would be $2.7 billion. Complete @$1.9 elimination was estimated to be 30% billion less than this by 1992 Case study cost estimates Comparison of HFC 134a relative to its 3-5 times substitutes (CFC 11 or 12). Ex-post ~2.5 CFCs costs as reported in 1988. Difference due to competition $2.4 Regulators moved the focused from the million 125 use sectors to producers and importers. (1989) $50k 40 Regulator permitting interim use of HCFC-22’S whilst developing hydrocarbons $0.48/kg/yr Median cost of Multi Lateral Fund (MLF) funded projects

Source

Cook 1996, p.7

Hoener 1996, p.50

SEI 1999, p.37 Lee 1996, p.33 Cook 1996, p.5 UNEP 2003, p.13

Notes: ● The social cost of reducing the use of CFCs differs to the other costs in the table as it includes the cost of compliance to all actors in society, including the foregone value for applications which are perceived (exante) not to have a possible substitute available. ● Compliance was not always achieved in the way anticipated during ex-ante estimates. For example, regulators moved their focus from the use sectors as anticipated to producers and importers which proved to be a significantly more cost effective approach. Also, the Montreal Protocol later changed to permit the interim use of HCFCs.

5

2 Introduction The international phasing out of Ozone Depleting substances (ODS) under the Montreal Protocol and EU Regulations 3093/94/EC, and 2037/2000/EC provides an early example of how the cost estimates provided by industry in response to proposed (ex-ante estimates) legislation can over estimate the costs which were later generated (ex-post estimates). This report considers the reasons for the differences and whether they can be said to be the result of exaggeration by the industry or just conservative assumptions in the face of unknown future technological developments. The report also considers how these differences may have differed geographically between the USA and the Europe, and within Europe and the EU. Finally consideration is given to the lessons which can be learnt from this case study and how it might be used in the process of informing future policy.

2.1 Policy introduction Under the auspices of the United Nations the global community agreed to adopt the Vienna Convention to combat the threat of ozone depletion in 1985. The provisions for phasing out the production and use of ODS were laid down in the Montreal protocol in 1987. As a result of subsequent incremental strengthening of the provisions, the Protocol now provides for the worldwide phase-out of all major ODS: CFCs, halons, methyl chloroform (1,1,1 trichloroethane), carbon tetrachloride, partly halogenated CFCs (e.g. hydrochlorofluorocarbons (HCFCs)) and methyl bromide. EU ozone policy dates as far back as 1978, but it was not until 1994 that major steps were taken towards a total phase out of the major ozone depleting substances (CFCs and halons). The existing EU policy relating to ODS consists mainly of two Regulations, one from 1994, Regulation 3093/94/EC, and one from 2000, Regulation 2037/00/EC. Whereas the first regulation, 3093/94/EC, aimed at reducing the production of particularly CFCs and halons within the EU, the latest regulation, 2037/00/EC, aims at controlling both production and consumption of all ODS within the EU. The Regulation 2037/00/EC which was proposed in August 1998, agreed in June 2000, and came into force 1 October 2000, contains new and faster phase-out schedules for HCFCs and methyl bromide, stricter regulations for the handling of ODS including new staff training schemes, mandatory recovery and destruction, and improved monitoring and licensing schemes.

2.2 Introduction to the literature The predictions made by industry when faced with the prospect of having to phase out the use of ODS is a much used example of when ex-ante estimates turned out to be too high (see for example Cook & Lee 1996, SEI 1999, Naess 2001 and ICS 2004). However due to the early nature of this case study, very limited actual quantitative ex-ante estimates were made or are therefore available. The only available economy wide ex-ante estimate were made on behalf of the USA’s EPA by the Rand Corporation (e.g. see Palmer et al 1980) which has often been compared to their later reports, as well as a series of post implementation progress reports written by the United Nations 6

Environment Programme’s (UNEPs) Technical and Economic Assessment Panel (TEAP) (e.g. UNEP 1989, 1994, 1995, 1997 II, 1998a&b, 2002, 2003 and 2005). The UNEP report expresses all costs figures in USA dollars. This report therefore quotes all figures in USA dollars as provided, and generates a costs comparison factor, ex-ante/ex-post cost estimates, to express the difference in estimates. Where possible, expected differentials for substitutes (i.e. a given substitute is expected to be a number of times greater than the ODS it replaces) is provided. This approach has the advantage of eliminating the effects of inflation. Using this approach, the comparison of before implementation of the Montreal Protocol and a 1998 real market cost differential for the substitution CFC 11 and 12 with HFC 134a is 2.5. Detailed estimates of costs are not available an EU level, however geographical differences in costs are dealt with briefly in a qualitative way in SEI 1999, and at a country level by Naess 2001.

3 Historical analysis 3.1 Pre-implementation Early on in the policy formulation process (i.e. in the late 1970’s), the chemical industry who produced the ODS were in opposition to the phase out of ODS. This opposition was founded on the perceived high costs of a substitution away from ODS and the possible impact on jobs, although proper cost estimates were not provided at this stage (ICS 2004). This perception was largely based on the extensive application of CFCs and other ODS throughout the economy, and the capital equipment associated with their use. Later on in the late 1980’s DuPont pointed out that “in the United States alone, there is now more than US $135 billion worth of installed equipment dependent on current CFC products” (reported in SEI 1999). The Association of European Chemical Companies argued in 1987 that regulating CFCs would cause “very large“ costs, leading to "redesign and re-equipping of large sectors of vital industry…, smaller firms going out of business … and an effect on inflation and employment, nationally and internationally“ (reported in SEI 1999). The sector also argued that “Development of alternatives to CFCs having equivalent safety and thermodynamic properties is extremely unlikely” (cited in UNEP, 1991a). In 1980, Du Pont led the creation of the Alliance for Responsible CFC Policy (ARCFCP) with the stated aim of preventing, or at least mitigating, any further regulatory threat to the CFC business (Gabel 1995, p.330). The early ex-ante costs estimated which were produced with the co-operation of industry representatives (see for example Palmer et al 1980), were later shown to have simultaneously under estimated the extent, the feasible rate whilst over estimating the unit cost of phasing out the use of ODS. As shown in Table 1 and reported in Cook (1996), in 1998 it was estimated that “it would costs $2.7 billion to halve U.S. CFC consumption within 10 years at an average cost of $3.55 per kg reduced. Four years later, the RIA [Regulatory Impact Assessment] estimate for a total CFC phase-out by 2000 was $3.8 billion” at a rate of $2.20 per kg. A later 1993 estimate predicted that a total phase out could be feasibly accelerated to 1996 at an overall cost of $6.4 billion (Cook 1996). This evolution in estimates needs to be seen in the context of the sectors’ relatively sudden change in their opposition to the Montreal Protocol during a period of technological development of substitutes leading up to the provision of data.

7

After their initial opposition to the Montreal Protocol, DuPont, who were leading in the race towards substitute development in the 1980s, was in 1986 first to support international controls (SEI 1999). Only one year after predicting economic chaos in case of CFC restrictions, Pennwalt declared its intentions to acquire a leadership position in the substitute market (Cogan, 1988). At this point the producer sector were divided with the remaining opposing companies, reportedly mostly European, having by the end of the 1980’s to accept the fate of CFCs and attempt to catch up and compete with the market leaders (SEI 1999). The cost to EU businesses of this situation is considered later in the section on Distributional effects. Behind this split is a technological divergence between Du Pont and their mostly European competitors. Du Pont invested in HCFC technology; a quick to develop substitute to CFC’s with only 2-10% the environmental impact of the CFC’s. The European sector opted to spend longer developing HFCs which provided a total ozone solution (Gable 1995, p. 337).

3.2 Post-implementation During the implementation of the Protocol the further division emerged between the producers and the diverse groups of users of ODS who maintained their opposition. “Estimates sponsored by user industries remained higher during that period as these sectors remained sceptical of ODS controls” (SEI 1999). The work of the UNEP’s TEAP in the 1990s was focused on exploring specific examples of applications which did not have practicable substitutes, for example ODS were thought to be essential for electronics manufacturing. This focus of the TEAP therefore evolved to consider the cost of enforcing the legislation for such applications. After making no reference to costs in their 1989 report, the TEAP were making numerous references to costs in their reports published in the later 1990’s and early years of this century (for example, there where 79 cost references in UNEP 1998a) as the regulator sought detailed economic analysis to justify any application for exception. The outcome of this process was that by the middle of the 1990s, CFCs in medical metered-dose inhalers were the only significant ODS use accepted as being essential (UNEP 1998), and that virtually all of the global reductions in CFC use (50% by 1995) had come at little or no cost to consumers (UNEP 1995). Refrigerator manufacturers were significant users of CFCs. Early hydrocarbon based ‘Greenfreeze’ substitutes were less energy efficient and had a greater cost. However, by the end of the 1990s, Greenfreeze technology was as energy efficient as the available alternatives and, if there was a cost differential, it was in favour of hydrocarbon Greenfreeze technology (UNEP 1997) and by the late 1990’s this was the dominant technology in Europe (SEI 1999).

4 Distributional effects 4.1 Distributional effects within the production chain This case study not only demonstrates that the ex-ante estimates provided by business were over estimates, it also shows that the businesses who helped generate the estimates 8

would not have been the ones who ultimately paid any additional cost. From the literature it is clear that the mandatory nature of the Montreal Protocol often provided a way for business to pass the costs onto the user sectors, and ultimately to the consumers. In addition to this, those companies who led in the development of CFC substitutes achieved a market rent so long as their competitors were in the process of catching up. These lead companies also opposed legislation, but only until they had managed to develop substitutes, after which they moved to encourage the regulations which they then gained from. The laggards were reported to be the European companies (SEI 1999) who therefore lost out due to their delay in innovation. The ex-post cost estimates showed that the phase out occurred at a much lower cost than the ex-ante estimates had suggested, and in some cases were achieved with costs savings or efficiency benefits in other areas (see for example SEI 1999). This finding highlights a possible additional cost which would not have been captured by either the ex-ante or expost cost estimates. Review of one part of a production process will likely promote unexpected efficiencies as the wider process is captured by the review. It could be argued that if these sectors were not required to phase out ODS, their efforts could have been somehow directed towards more targeted forms of innovation which may have provided more efficiencies than had actually occurred, and in turn provide other social goods in terms of jobs and growth. ODS legislation clearly provided social good in terms of environmental protection at a financial cost which was less than was originally considered socially acceptable. This may however have been achieved at an unacknowledged opportunity cost in terms of foregone efficiencies.

4.2 Geographical distributional effects As already discussed, European producers of ODS developed the total ozone safe HFCs whilst the US owned Du Pont more quickly developed the partially ozone safe substitutes, HCFCs (Gabel 19995, p334). This led to European producers losing out whist HFCs were in development. Naess (2001) explores a number of European country case studies to explore how the legislation was implemented differently. The key relevant outcomes from these country case studies are explored below. Norway’s role as an international frontrunner in terms of ozone policy has been facilitated by the lack of opposing interest groups within Norway. Neither industry, nor consumers have had particular incentives to oppose a phase out of main ODS in Norway (Naess 2001). This is an example therefore where the regulators was not presented with negative ex-ante predictions from their own industry, and were therefore free to follow relatively stringent regulations. Naess (2001) explores the theory that Southern member states often face a larger share of compliance costs, as regulations can often reflect Northern problems and are adapted to Northern conditions. A lack of capacity and capability in these countries is thought to lead to difficulties or additional costs during implementation. “This general trend seems to be valid also in terms of ozone policy where Southern states have scored rather low in terms of both implementation and enforcement” (Naess 2001):

9

•

•

“Spain has been one of the major ODS producers and consumers in the EU…Spain has phased out consumption of CFCs and halons domestically, but still produces CFCs for export to developing countries and for essential uses, which is allowed under the Montreal Protocol, but banned within the EU under the latest regulation 2037/00/EC.” “The most recent data for Italy from 1999 show that Italy is still producing CFCs for essential uses and export to developing countries, and that production has stabilised around 7 000 tons a year of CFCs”.

Naess (2001) then goes on to predict that “many of the problems experienced in Southern Europe may occur also in the CEE countries , such as illegal trade and a implementation deficit, due to lacking administrative and personnel capacity. Naess (2001) then goes on to report that “consumption has stabilised at a low level [by the mid 1990’s] in the CEE countries, international contributions from the Global Environment Facility and the Multilateral Fund of the Montreal Protocol of US$ 111 million has been given to 11 countries in Eastern Europe (Belarus, Bulgaria, the Czech Republic, Hungary, Poland, the Russian Federation, Slovakia, Slovenia and Ukraine) but most CEECs have not fulfilled their obligations as set out in the latest Amendments to the Montreal Protocol, which suggest that potential problems persist due to administrative, legal or institutional failure to implement regulations properly”.

5 Conclusions This report has found that a comparison of ex-ante and ex-post cost estimates can produce a comparison factor from at least 1.4 at an overall macro level to 40 for individual case studies, and as high as 125 for administrative costs of compliance. The relatively modest overall figure of 1.4 is actually hiding a much greater failure to predict the implications of policy implementation. Analysis of the details of this ex-ante cost estimates shows they simultaneously under estimated the extent and the feasible rate of phasing out the use of ODS, whilst over estimated the unit cost. Analysis of the behaviour of the actors involved reveals that these major changes in costs estimates coincided with the leading companies Du Pont successfully developing HCFCs as a substitute to ODS, and therefore relaxing their opposition to the proposed legislation. Du Pont achieved a market rent due to their innovation in anticipation of the legislation. This rent was maintained so long as their mostly European competitors were in the process of developing their alternatives, HFCs. A later negotiation to phase out HCFCs due to their continued partial ozone depleting nature was of course resisted by Du Pont (Gabel 19995). To help inform formulation of future environmental policy it is useful to explore the reasons behind the industries’ failure to correctly predict the costs of implementing legislation. The change in Du Pont’s position is perfectly reasonable in the context of their development of HCFCs. However, this technological development occurred notably ahead of their previous predictions. Furthermore, once competitive pressure was applied, their competitors swiftly followed and developed their own alternatives. This raises the question of were these pessimistic ex-ante estimates the result of exaggeration by the industry or just conservative assumptions in the face of unknown prospects for future technological development?

10

The phasing out of ODS is a very complex and technical matter. Different groups within an organisation would have quite possibly generated very different costs estimates. Those responsible for the prospective development of the CFC substitutes would tend to be conservative in their assumptions about future developments to avoid promising to deliver innovations they were unsure of. Therefore, although explicit exaggeration by the industry cannot be ruled out, this is not an explanation which is required to explain the failure. The implementation of the Montreal Protocol was in many ways the first of its kind in phasing out a widely used substance from our modern economy. However, the chemical industry had previously always managed to develop new compounds to fulfil societal demands to generate profits. It would therefore be difficult to conclude that the failure to accurately predict the costs of developing substitutes was as a result the genuine lack of foreseeability of the outcome. This case study has also showed that the industry did not lose-out due to the Montreal Protocol as the costs were passed on through the product chain to the consumers. It was not of course in their interest to encourage legislation until they either had, or were confident about, developing substitutes. From this analysis it therefore seems sensible to conclude that the failure to successfully predict the true costs of implementing the phase out of ODS was likely to be due to conservative assumptions by some within the industry in the face of uncertain future technological developments. This situation was in the context of there being no real incentive for senior managers to intervene until alternatives were known to be on the way. The age of this case study means that the ex-ante costs estimates can be seen in their full context. Although there are limited ex-ante cost estimates available, this case study has been widely documented and reported on. Both industry and policy makers will have to some degree already learnt from the outcomes of this case study when considering more recent environmental legislation. The results from this report should further help inform such future interactions. Perhaps the most useful finding which has been made possible by the historical nature of the case study is the likely underlying dynamics and interests of the actors involved which led to the failure of the industry to accurately predict implementation costs. This conclusion suggests that policy makers need to consider exante costs estimates in the context of the incentives and interests of the actors who generate them. Formal review procedures of ex-ante cost estimates might go someway to reduce the chances of overestimation. Peer review procedures within organisations and between companies in a given sector might highlight excessive conservatism in assumptions. However, if a sectors’ interest lay in talking-down the prospects of innovation, it is likely that only detailed external reviews of company’s assumptions can critically challenge ex-ante cost estimates.

11

6 References Cogan, D.G., 1988 Stones in a Glass House. Washington: Investor Responsibility Research Center Cook, E., 1996 Overview. In Ozone Protection in the United States - Elements of Success (ed. by Elizabeth Cook). Washington, DC. World Resources Institute, 31-38. ISBN 156973-088-1 Gabel, H., L, 1995 Environmental management as a competitive strategy: the case of CFCs, chapter in Principles of Environmental and Resource Economics: A guide for students and decision makers Ed by H Folmer, The conference o European Rectors 19995 Lee, D., 1996 Trading Pollution. In Ozone Protection in the United States - Elements of Success (ed. by Elizabeth Cook). Washington, DC. World Resources Institute, 31-38. ISBN 1-56973-088-1 ICS (The International Chemical Secretariat) 2004 Cry Wolf – Predicted costs by industry in the face of new regulations, Report 6:04, April 2004 Malakoff, D., Phillips, S., 1996 The spray can ban. In Ozone Protection in the United States - Elements of Success (ed. by Elizabeth Cook). Washington, DC. World Resources Institute, 11-20. ISBN 1-56973-088-1 Moerner, J.A., (1996) Taxing Pollution. In Ozone Protection in the United States Elements of Success (ed. by Elizabeth Cook). Washington, DC. World Resources Institute, 31-38. ISBN 1-56973-088-1 Naess, T., The Effectiveness of the European Union (EU) Ozone Policy, The Fridtjof Nansen Institute, FNI Report 15/2001, ISBN 82-7613-417-3 Palmer, A.R., Mooz, W.E., Quinn, T.H., Wolf, K.A., (1980). Economic Implications of Regulating Chlorofluorocarbons Emissions from Non Aerosol Applications (Report No. R-2524-EPA). Santa Monica, CA: Rand SEI (Stockholm Environment Institute) 1999 Costs and Strategies presented by Industry during the Negotiation of Environmental Regulations Prepared for the Swedish Ministry of the Environment, April 1999 UNEP 1989 Montreal Protocol on Substances that Deplete the Ozone Layer: Economic Panel Report, UNEP, July 1989 UNEP 1991a 1991 Economic Assessment Report. Economic Options Committee, UNEP, 30 November, 1991 12

UNEP 1991b ‘1991 Assessment, Report of the Technology and Economic Assessment Panel’, UNEP, December 1991 UNEP 1994 Scientific Assessment of Ozone Depletion: 1994 (Report No. 37). WMO/UNEP UNEP 1995 Montreal Protocol on Substances that Deplete the Ozone Layer: 1994 Report of the Economics Options Committee, UNEP UNEP 1997 Montreal Protocol on Substances that Deplete the Ozone Layer: Technology and Economic Assessment Panel, Volume II, UNEP, April 1997 Volume II UNEP 1998a Montreal Protocol on Substances that Deplete the Ozone Layer: 1998 Report of the Technology and Economics Assessment Panel, UNEP April 1998 UNEP 1998b Montreal Protocol on Substances that Deplete the Ozone Layer: 1998 Report of the Technology and Economics Assessment Panel, UNEP October 1998 UNEP 2002 Montreal Protocol on Substances that Deplete the Ozone Layer: 2002 Report of the Technology and Economics Assessment Panel, UNEP UNEP 2003 Montreal Protocol on Substances that Deplete the Ozone Layer, May 2003 Report of the Technology and Economics Assessment Panel, Progress report, UNEP UNEP 2005 Montreal Protocol on Substances that Deplete the Ozone Layer, May 2005 Report of the Technology and Economics Assessment Panel, Progress report, UNEP

13

7 Annex I - Proposed plan of work This proposed plan of work was submitted for review on the 8th of July 2005. No response or comment was received. 1. Scope: The analysis will look at the costs of Ozone Depleting Substances (ODS) and their substitutes as developed. Attempts will also be made to consider whether ex-ante estimates correctly anticipated how these costs would pass through the product chains and impact on the wider economy. 2. Timeframe: The time frames will be ex-ante estimates made leading up to the ratification of the Montreal Protocol in 1989, with ex-post estimates made once the substitutes had been fully developed (likely to be from the mid 1990's onwards). Reference will be made to the rent associated with early developed substitutes. 3. Geographical: It is anticipated that the analysis will be done at a Europe wide level and may need to make reference to US as well as EU chemical industry costs. It is likely that high level estimates will be compared with specific case study costs. 4. Approach: There will be a comparison of ex-ante /ex-post estimates. It is intended to compare ex-ante and ex-post estimates of the additional costs of ODS substitutes (i.e. the substitutes being factors of x&y times greater than CFS). This approach should tackle some of the methodological problems associated with different timeframes.

14