Précis W O R L D B A N K O P E R A T I O N S E V A L U A T I O N D E P A R T M E N T S U M M E R 2 0 0 2 N U M B E R 2 2 7 High-Efficiency Lighting in Mexico R oughly 2.6 million high-efficiency light bulbs were sold below-cost to residential consumers of electricity in two Mexican cities as part of a pilot demonstration project for high-efficiency lighting. If the project's main objective was to show how many high-efficiency light bulbs could be successfully distrib- uted, it was well crafted. The ILUMEX project surpassed its physical targets and successfully demonstrated the technical feasibility of reducing both greenhouse gas emissions and local environmental contamination through the widespread installation of fluorescent lighting. The payback in learning about technical designs was high, but overall it failed to deliver lessons about the economic and finan- cial design of sustainable models for the marketing of high-efficiency light bulbs, especially among poor consumers. The project designers did not ask if the model was replicable or financially sustainable. Cutting Electricity Demand energy-saving light bulbs was to be subsi- The project was intended to demonstrate dized because households and other light the technical and financial feasibility of consumers of energy rarely take advantage reducing both greenhouse gas emissions of energy-efficient technologies, even when and local environmental contamination it can save them money. Additional study through the widespread installation of objectives were to build institutional high-efficiency lighting. The main project capacity for technological change and component was the promotional sale of energy conservation, to provide a replica- about 1.7 million high-efficiency light ble model for managing energy demand, bulbs among residential users in Guadala- and to strengthen institutional capacity at jara (Jalisco State) and Monterrey (Nuevo CFE (Mexico's electric utility) for manag- Leon State), at prices 60 percent below ing energy demand. cost. Significant energy savings and envi- The ILUMEX project was co-financed ronmental benefits were envisaged because by the World Bank's Global Environment high-efficiency lights consume 75 percent Facility (GEF, which provided a grant of less energy than equivalent standard about $US10 million) and the Kingdom of incandescent bulbs, and last 15 times Norway (with a grant of about $3 mil- longer. The initial residential purchase of lion). 2 World Bank Operations Evaluation Department Project Outcomes services, commerce, and industry; (3) there were no retail By the grant's closing date, the project slightly surpassed price reductions; and (4) by project's end, more retailers the original target of 1.7 million fluorescent light bulbs stocked and displayed fluorescent light bulbs. sold--and CFE continued selling them through January 31, Based solely on energy savings, the project had an 1999, for a total of 2.6 million sold. Project costs remained acceptable economic rate of return. The rate was signifi- remarkably close to original estimates and sales prices went cantly lower than projected at appraisal because of the up, so the average subsidy was only 49 percent--not 60 overestimate of capacity savings, but if local and global percent--of the bulb's cost. The energy savings and environmental benefits were factored in, it could increase reduced CO2 emissions were in line with initial estimates, substantially. The distribution analysis (without environ- but savings in power capacity (because reduced power con- mental benefits) shows that all gains went to participant sumption delays the need to invest in new capacity) were households. Those gains exceeded the appraisal estimate only one-third the initial estimates. because the marginal electricity tariffs to high-consumption Moreover, the project lost its original focus on poor con- households were two to three times higher than the utility's sumers because sales were slow among low-consumption marginal costs. CFE lost profits, mostly through electricity households. (Overall, 9.6 percent of sales were to low-con- sales forgone, but CFE's financial losses were reduced by sumption households, 31.3 percent to intermediate users, partial coverage of project costs by grants from the GEF and 59.2 percent to households consuming a lot of electric- and Norway. ity.) Concerns about sales also led to increasing to 10 the Internationally, the project compares well with similar original limit of 6 light bulbs per household, and expanding initiatives in terms of number of light bulbs distributed, marketing efforts to the cities' entire metropolitan areas. unit cost, and level of subsidy. But the project design and The program did not fulfill its expectations of promoting implementation never addressed the issue of sustainability the sale of high-efficiency light bulbs to other consumers. of sales in the long term, in the absence of subsidies. The project did not systematically collect information to Whether fluorescent light bulbs will continue to be used estimate its "free-driver effect" (the increase in fluorescent after the project ends remains unknown. light bulb sales that could be attributed to the demonstra- The project ultimately managed to deliver more than 2.6 tion effect), but the staff of the project implementation unit million fluorescent light bulbs at more reasonable unit suggest (partly on the basis of anecdotal evidence) that (1) prices than in similar projects elsewhere--a considerable during the project's life, retail sales to households actually achievement for a project this size. The project audit con- declined because of competition from the project (and the cludes that the project succeeded in demonstrating the tech- expectation that subsidized sales would be extended to nical feasibility of reducing both greenhouse gas emissions other cities); (2) some people transferred their home experi- and local environmental contamination through the wide- ence to the workplace, thereby increasing demand among spread installation of high-efficiency lighting, but it failed The Project Compared with Other Similar Initiatives Country/ Number of FLs Cost per FL (US$) project origin Project design delivered For the buyer For the project Total Mexico (GEF) Subsidy to consumers 1,712,361 7.1 6.8 13.1 Poland (GEF) Subsidy to manufacturers 1,600,000 9.0 3.2 12.2 Jamaica (GEF) Subsidy to consumers 85,000 5.8 6.8 12.6 Thailand (GEF) Bulk purchases 1,500,000 9.0 2.9 11.9 Brazil Give-away 89,000 -- 8.3 8.3 Denmark Small subsidy plus quality control programs 1,000,000 11.0 4.4 14.4 U.K. 1 Subsidy to consumers 3,000,000 12.4 1.8 14.2 U.K. 2 Give-away 814,000 -- 11.3 11.3 Peru Publicity campaign, no subsidies 400,000 20.0 16.8 36.8 U.S. (several) Subsidy to consumers Rebates between Usually none 20% and 50% of Paid out of a sur- retail prices charge on elec- tricity bills Source: E. Martinot and N. Borg (1998) "Energy-Efficient Lightning Programs. Experience and Lessons from Eight Countries." Précis 227 3 to demonstrate the endeavor's financial feasibility and con- ect was sustainable or replicable and what lessons could be tributed only modestly to building the institutional capacity learned from it. Among various conflicting positions for technological change, energy conservation, and argued at the Bank were the following. demand-side management (the promotion of energy-saving Any project with a large subsidy component is unsus- devices) at CFE, the electric utility. tainable and difficult to replicate. Few will contest that Therefore the project's audit ratings were marginally sat- some activities--such as elementary school and primary isfactory for outcome, modest for impact on institutional health care--are regularly subsidized, and subsidies would development, and uncertain for sustainability. be one way to pay for positive environmental externalities (such as CO2 abatements). The problem with ILUMEX Lessons Learned was not the presence of a subsidy but that nobody ever Despite its shortcomings, this pilot demonstration project asked, "Why is this level of subsidy necessary?" or "How pioneered a new and increasingly important field for the will fluorescent light bulbs be marketed following the end borrower, the Bank, the GEF, and the development com- of the subsidies?" When the actual subsidy went from 60 munity at large. It also provided several important lessons: percent to 49 percent, for example, the project did not dis- It is important to build into the design of demonstra- cuss whether the appropriate level of subsidy would be 60 tion projects ways to consider whether the project is replic- percent, 49 percent, or 24 percent (as in the follow-up pro- able and sustainable. For any pilot demonstration project, gram). Among the many studies required by the Bank, not and more so for a GEF grant, central questions to be a single one asked for an analysis of alternative ways to addressed include whether the project is replicable, whether finance the future distribution of light bulbs (for example, a its outcomes are sustainable, in the long term, in the surcharge on the price of electricity, selling CO2 abate- absence of subsidies, and what lessons can be learned from ments, an environmental investment fund, or term sales at it. Project design should explicitly spell out the strategy, full price). activities, resources, and monitoring committed to answer- Another position was that sustainability was outside the ing these questions. Sticking to and meeting a physical tar- project's reach because it depended on Mexico's willingness get is important, but not sufficient. By definition a pilot to raise electricity rates. There is no basis for this argument. project is a small-scale endeavor that is not expected to Between 1995 and 1997, the relevant marginal price of elec- make a large difference on the ground until it is replicated tricity per kilowatt hour (the amount paid by households) and the lessons learned from it can be applied. was two to three times the utility's marginal cost, and grow- Well-planned implementation is central to project per- ing. With kilowatt hours priced so high, households will formance. The project's implementation scheme--based on gain so much in energy savings that the project is highly decentralized project implementation units and the estab- beneficial, and it should be easy in the future to make lishment of independent trust funds with BANOBRAS (the households pay full price for the fluorescent light bulbs. National Bank of Public Works and Services)--was very Financial gains to households should not be construed as successful. It gave Mexico's utility oversight of the project project economic gains because most of these gains are while ensuring a smooth flow of funds to the project imple- mere transfer payments from CFE and the donors. And it mentation units. may be true that households will pay full price for fluores- Bulk procurement can leverage funding. As similar proj- cent light bulbs in the future. However, the project design ects have already shown, bulk procurement of fluorescent was based on the assumption that households are reluctant light bulbs was an effective way to get technical improve- to do so, and the project results provide no clues to chal- ments and significant price rebates from the light bulb lenge that assumption. manufacturers. The pilot project is justified because it is the first project in a new area, and there is much to learn. This is precisely Assessing If a Pilot Demonstration Project Is where the shortcomings of the ILUMEX project are evi- Sustainable and Replicable dent. Its payback in learning about technical designs (for The ILUMEX project was the first of its kind in the Bank example, about procurement, lamp specifications, opportu- and Global Environment Facility portfolios. Assessing its nities for greenhouse gas verification, and activities imple- outcome elicited a lively discussion among Bank staff. No mented jointly) was high. But by and large it failed to one disputes the project's achievement in terms of light deliver lessons about the long-term financial sustainability bulbs distributed, but views differ about whether the proj- of high-efficiency lighting projects. 4 World Bank Operations Evaluation Department Recent OED Précis 209 Participation in Development Assistance 226 The Next Ascent: An Evaluation of the Aga Khan Rural Support 208 India: Improving the Development Effectiveness of Assistance Program 207 Adapting Transport Institutions to Romania's Transition Needs 225 Assisting Russia's Transition: An Unprecedented Challenge 206 Reforming India's Energy Sector (1978­99) 224 Grant Programs: Improving Their Governance 205 Progress in Java's Villages 223 Supporting Health Reform in Eastern Europe 204 ARDE 2000: From Strategy to Results 222 Bolivia Water Management: A Tale of Three Cities 203 World Bank Forest Strategy: Striking the Right Balance 221 Bridging Troubled Waters: A World Bank Strategy 202 Poverty Reduction in the 1990s: The World Bank Strategy 220 Cultural Properties in Policy and Practice 201 The Drive to Partnership: Aid Coordination and the World Bank 219 ARDE 2001: Making Choices 200 Evaluating Gender and Development at the World Bank 218 IDA's Partnership for Poverty Reduction 199 Ghana: Building a Stronger Transportation System 217 Community Forestry in Nepal 198 Agricultural Extension: The Kenya Experience 216 Promoting Environmentally Sustainable Development 197 Toward a Comprehensive Development Strategy 215 Rural Water Projects: Lessons Learned 196 Evaluation and the Development Challenge 214 Uganda: Policy, Participation, People 195 Poverty Assessments: Maximizing Impacts 213 Developing African Capacity for Monitoring and Evaluation 194 Involuntary Resettlement: The Large Dam Experience 212 Chile's Model for Educating Poor Children 193 Partnership for Education in Jordan 211 Strengthening Tunisian Municipalities to Foster Local Urban 192 Reforming Bolivia's Power Sector Development 191 Bangladesh: Progress Through Partnership 210 Connecting with the Information Revolution 190 Transport in Indonesia Director-General, Operations Evaluation: Robert Picciotto Director, Operations Evaluation Department: Gregory K. Ingram Manager, Sector and Thematic Evaluation: Alain Barbu Task Manager: Andres Liebenthal, Lead Evaluation Officer, OEDST This Précis is based on the evaluation work of Andres Liebenthal Précis are available to Bank Executive Directors and staff from the Internal Documents Unit and from regional information service centers, and to the public from the World Bank InfoShop. Précis are also available at no charge by contacting the OED Help Desk: eline@worldbank.org or calling 1-202/458-4497. Précis Manager, Partnerships and Knowledge: Osvaldo Feinstein · Editor-in-Chief: Elizabeth Campbell-Pagé · Series Editor: Caroline McEuen · Dissemination: Juicy Qureishi-Huq DISCLAIMER: OED Précis are produced by the World Bank Operations Evaluation Department, Partnerships and Knowledge Group (OEDPK), Outreach and Dissemination Unit. The views in this paper are those of the Operations Evaluation staff and editors and should not be attributed to the World Bank, its affiliated organizations, or its Executive Directors. Précis aussi disponible en français Précis en español tambien disponible @ http://www.worldbank.org/oed Précis 227 High-Efficiency Lighting in Mexico ISSN 1564-6297