Mexico - Labor Market and Productivity Enhancement Project

Document of The World Bank FOR OFFICLAL USE ONLY Report No: 18759 IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MEXICO LABOR MARKET AND PRODUCTIVITY ENHANCEMENT PROJECT (LOAN 3542-ME) December 28, 1998 Mexico Department Human and Social Capital Development Group Latin America and the Caribbean Regional Office This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = New Peso (N$) US$1.00 = 9.9 New Pesos (December, 1998) FISCAL YEAR January 1 - December 31 ABBREVIATIONS & ACRONYMS CIMO Multiple Support Service Program to MSMEs (Programa de Capacitaci6n Industrial de Mano de Obra, re-named in 1992 Programa de Calidad Integral y Modernizaci6n) DGE General Directorate for Employment (Direcci6n General de Empleo) DGIT General Directorate for Information and Telecommunications (Direcci6n General de Informaci6n y Telecomunicaciones) GDP Gross Domestic Product CGPEET General Coordination Unit for Labor Market Policies, Studies and Statistic (Coordinaci6n General de Politicas, Estudios, y Estadisticas del Trabajo) ICR Implementation Completion Report IDB Inter American Development Bank IMF International Monetary Fund INEGI National Institute of Statistics (Instituto Nacional de Estadistica, Geografla e Informdtica) MSMEs Medium, Small and Micro Enterprises NAFIN National Financing Company (Nacional Financiera S.N. C.) NAFTA North American Free Trade Agreement (Tratado de Libre Comercio de America del Norte) PILEOT Project of Local Initiatives for Employment and Temporary Jobs (Proyecto de Iniciativas Locales de Empleo y Ocupaci6n Temporal) PROBECAT Labor Retraining Program (Programa de Becas de Capacitaci6n para Trabajadores) PROSSE Program of Essential Social Services (Programa de Servicios Sociales Esenciales) SEP Public Education Secretariat (Secretaria de Educaci6n Puiblica) SNE National Employment Service (Servicio Nacional de Empleo) SEE State Employment Services (Servicios Estatales de Empleo) STPS Secretariat of Labor and Social Welfare (Secretaria de Trabajo y Previsi6n Social) UPCs Regional Productivity Support Units (Unidades Promotoras de Competitividad) Vice President Shavid Javed Burki Director Olivier Lafourcade Sector Director Xavier Coll Staff Member Anna Sant'Anna  FOR OFFICIAL USE ONLY iii IMPLEMENTATION COMPLETION REPORT MEXICO: LABOR MARKET AND PRODUCTIVITY ENHANCEMENT PROJECT (Loan 3542-ME) Table of Contents PREFACE ............................................... IV EVALUATION SUMMARY ............................................... VI PART 1: PROJECT IMPLEMENTATION ASSESSMENT ................................................I PROJECT BACKGROUND ................................................I PROJECT OBJECTIVES AND DESCRIPTION ..........2.....................................2 Original Objectives........................ ....................- 2 Assessment of Original Project Objectives ........2.....................................2 Original Project Design and Organization ..............................................3 Changes ..............................................4 IMPLEMENTATION RECORD .................4......4 Proroductiiit Enhancement ..4...................4 Labor Mark-et Adjustment .............................................. .6 Information, Policy and Institutional Strengthening ..............................7................ .7 Procuremient of Goods and Services ............................................. 10 ACHIEV EMENT OF ORIGINAL OBJECTIVES ............................................... 10 MAJOR FACTORS AFFECTING THE PROJECT ............................................... 12 Factors not subject to Government control ............................................. 12 Factors suibject to Government control ............................................. 12 Factors subject to the control of the implementation agency ............................................. 12 PROJECT SUSTAINABILITY ............................................... 12 BANK PERFORMANCE ............................................... 12 BORROWER PERFORMANCE ............................................... 13 ASSESSMENT OF OUTCOME ............................................... 13 LESSONS LEARNED ............................................... 14 FUTURE OPERATIONS ............................................... 15 PART 11: STATISTICAL TABLES ............................................... 16 Appendixes: A. Mission's Aide Memoire B. Government's Contribution to the ICR C. Government's Comments on the ICR This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  iv IMPLEMENTATION COMPLETION REPORT MEXICO LABOR MARKET AND PRODUCTIVITY ENHANCEMENT PROJECT (Loan 3542-ME) PREFACE This Implementation Completion Report (ICR) assesses the implementation and results of the Labor Market and Productivity Enhancement Project in Mexico, for which Loan 3542-ME, in the amount US$174 million equivalent, was approved by the Board on December 15, 1992. The Legal Agreements were signed on December 18, 1992 and the loan became effective on March 11, 1993. In 1995, in response to Mexico's economic crisis, both the IDB and the Bank agreed to finance a two-year Program of Essential Social Services (PROSSE) to protect priority programs for the poor. One of the components of PROSSE provided financing for retraining and employment generation, expanding considerably the activities that were being financed through Loan 3542-ME. The present evaluation assesses the implementation and impact of the original loan, but also takes into account the broader context of the expanded program, which was financed in parallel through PROSSE. Loan 3542-ME was closed on June 30, 1998. The final disbursement took place on September 3, 1998. The total amount of the loan disbursed was US$174.0 million and the total project cost amounted to US$313.6 million. The Loan provided 55 percent of total project cost (compared to 49 estimated at appraisal), and the contributions from the Federal Government, the State Governments, and from private enterprises represented respectively 28, 13, and 4 percent, compared to the following shares estimated at appraisal: 32, 5 and 14 percent respectively. The State Governments contributed with US$12.2 million of the Borrower total share of US$99.2 million. The ICR was prepared by Nydia Maraviglia (LCSHD Consultant) after the final supervision mission that took place in February 9 to 13, 1998. The mission consisted of Mari Minowa (LCHSD Task Manager and mission leader), Hong Tan (PSDBE principal economist), Mauricio Mathov (ISGIS, information specialist), and Nydia Maraviglia (LCSHD Consultant). The Aide Memoire for that mission is included as Appendix A. This ICR is based on the findings of the project completion mission, documents in the project file including several evaluation studies based on specially designed surveys, and interviews  v with Government officials, representatives of private enterprises and relevant Bank staff. The last mission Aide-Memoire is included as Appendix A. The ICR was reviewed by Carmen Hamann (Sector Leader, LCHSD/LCC1C) and Anna Sant'Anna (Sector Coordinator, LCC1C). The Borrower prepared its own evaluation of project execution, which is available in the project file as indicated in Appendix B. The Borrower was requested to comment on the Bank ICR and the response has been incorporated into the text.  vi IMPLEMENTATION COMPLETION REPORT MEXICO LABOR MARKET AND PRODUCTIVITY ENHANCEMENT PROJECT (Loan 3542-ME) EVALUATION SUMMARY Introduction In the early 1980's the Government of Mexico developed a number of employment related programs reflecting a strong commitment to build up its human capital in the context of a free market oriented strategy. These programs, consisting of registration of job seekers and vacancies, career counseling, referral, placements, and training, were run mainly through the National Employment Services (SNE) and its network of State Employment Services (SEEs) of the Secretariat of Labor and Social Welfare (STPS). The STPS is, by law, the leading institution in matters related to employment, labor market regulations, in-service training, productivity and working conditions. It was responsible for the National Training and Productivity Program and a for the National Agreement for Increased Quality and Productivity signed by the Government in 1992 with representatives of business and labor. This is the background against which this project was developed. Labor force skills increased as a result of improvements in literacy (now at 90 percent) and in school attendance. Nevertheless, the need for further enhancement in workers' skills keeps increasing as a consequence of the rapid liberalization of the economy and growing competition with labor markets in developed countries, particularly after the North American Free Trade Agreement (NAFTA). Project Objectives and Description The objective of the proposed project was to assist the Government of Mexico in promoting private sector-led growth and in easing the costs of labor mobility and labor market adjustment. This was to be achieved by increasing productivity and competitiveness, by reducing the costs of hiring and job search, by training displaced and unemployed workers (particularly women), by improving labor market information, and by designing, evaluating, and implementing relevant policies and programs in the sector. The project consisted of three main components: (a) Productivity Enhancement (28  vii percent of total costs) aimed at improving productivity and competitiveness of medium, small and micro enterprises (MSMEs), by expanding the coverage and scope of the Multiple Support Service Program to MSMEs (CIMO) while strengthening coordination with other programs, and by promoting the gradual privatization of CIMO; (b) Labor Market Adjustment (61 percent of total costs) was designed to improve the coverage of the SEEs and provide education and training opportunities for displaced workers; and (c) Information, Policy and Institutional Strengthening (11 percent of total costs) aimed at improving the availability, timeliness and dissemination of labor market information and strengthening institutional capacity to design, operate and evaluate labor market policies and programs. A major factor that changed the scope of the overall Labor and Productivity Enhancement Program (of which this project was a part) was the approval, in 1995, of the Program of Essential Social Services (PROSSE). This program was a two-year time slice operation supporting the highest priority social sectors programs. Large items of the project budget for 1995 and 1996, were allocated to PROSSE, thus freeing funds for CIMO expansion. Implementation Experience and Results' Productivity Enhancement. The CIMO program operated on the basis of four main principles applied according to need: (a) identifying opportunities within the enterprises to improve productivity, quality and competitiveness; (b) preparing joint training programs by grouping enterprises by sector and needs; (c) reorganizing work and production to ensure quality; and (d) encouraging enterprise connections and/or integration of productive processes to increase MSMEs product quality and competitiveness. During the project period (1993-1997), 76 percent of CIMO assistance was for micro-enterprises, 17 percent for small firms, and 7 percent for medium enterprises. Program coverage increased from 29 existing Regional Productivity Support Units (UPC) to 62 units and sub-units (a substantial change from the original plan of adding only 5 units) and promoters increased from 89 to 206. By mid-1994 the demand for employment enhancement activities increased further as a result of the economic crisis and the approval of NAFTA. PROSSE financing of other project components allowed a large increase in CIMO activities between 1995 and 1997. Overall, during the whole project period, almost 1.5 million workers attended about 120,000 training and related events. The quality of CIMO services evolved from simply responding to enterprise requests, to a careful diagnosis of needs and problems prior to providing assistance. CIMO has also helped develop, in a number of the enterprises it assisted, a culture of seeking and using technical assistance and training for their workers, and establishing links with other enterprises to maximize production efficiency and improve earnings. CIMO also became a facilitator in developing skills in technical assistance and 'It should be noted that, in addition to inputs from this loan, the implementation described below includes 49,500 training activities and 800,000 fellowships under PROBECAT, that were financed by the PROSSE project during 1995 and 1996.  viii training in the regions where these did not exist. The program is now continuing with financing from an IDB loan. Labor Market Adjustment. In 1992, after ten years of operation, the SEEs still had a limited capacity to effectively improve the operations of labor markets. There was a clear need for improvements in organization and information, and administration of the Labor Retraining Program (Programa de Becas de Capacitaci6n para Trabajadores - PROBECAT). This component was implemented on the basis of strategies that included promoting complementarity in labor related assistance, installing better information systems and making improvements in worker training, retraining and placement. To promote complementarity, the program established Executive Committees of Employment Services, formed by representatives of enterprises, training agencies and other labor market agents. During the project period, 65 such committees were established, greatly exceeding the target of 32. Since 1993, these committees, together with the SEEs, served almost 2.5 million job seekers, advertised about 2 million vacancies and resulted in employment of almost 600,000 workers. These numbers represent, respectively, 132, 99 and 93 percent of the original goals. In addition, SEEs issued labor bulletins, sponsored employment fairs, opened information windows, and operated mobile units for greater outreach. The PROBECAT program and the Project of Local Initiatives for Employment and Temporary Jobs (Proyecto de Iniciativas Locales de Empleo y Ocupaci6n Temporal - PILEOT) programs included four training models. These were: (a) basic training for productive activities (in a school environment); (b) mixed training (in schools and enterprises); (c) re-training and re-conversion; and (d) training for self-employment. These programs had a very rapid growth in recent years, as a result of the injection of considerable funding by the Government and by the Bank, both through this project and through PROSSE. From 1992 to 1997 the annual number of fellowships granted went from 42,000 to over 500,000 - more than a ten-fold increase. In addition, training in PILEOT benefited around 550,000 workers with a mix of basic school education, literacy classes and work related training. However, this mixed approach did not give satisfactory results, due to, among other factors, the low level of income and education of the trainees, which resulted in high desertion rates. Information, Policy and Institutional Strengthening. An information system was developed to handle data on employment, salaries and benefits, education, training, and technology, as well as on collective bargaining agreements and other information relevant to labor market regulations. In 1993, the system was developed using DATAFLEX, but in 1995 data storage showed signs of saturation and sluggishness. In January 1996, a change was made to INFORMIX ON-LINE. The General Directorate for Information and Telecommunications (DGIT) of the STPS, responsible for the system since 1995, introduced INTERNET and INTRANET for external and internal communications respectively. The systems provide, inter alia, information on vacancies, job seekers, training courses, and fellowships. Since the information is available nationwide, it has made possible to identify the most serious regional or local  ix occupational imbalances and take decisions on training that are responsive to specific local needs. Several surveys and studies (listed in Annex 6) were carried out through this project. The main surveys were conducted in coordination with the National Institute of Statistics (INEGI). Study topics included organization of enterprises, management and labor conditions, determinants of productivity and international comparisons, the informal sector and agriculture employment. Of particular relevance to this project were the CIMO, PROBECAT and PILEOT studies. After an evaluation of PROBECAT conducted in 1994, it was decided that the 1995 and 1996 evaluations would concentrate on the program variations comprising training for self-employment and support to local initiatives, rather than continuing with global evaluations. The first evaluation of PILEOT intended to determine the extent to which training does facilitate self-employment and the second evaluation assessed the program contribution to the consolidation of productive projects developed through local initiatives, measuring the impact on income and employability of fellowship recipients. Assessment of Outcome Overall, the project amply met most of the targets set as intermediate objectives, provided a strong background for program changes and, most importantly, gave impetus to labor force modernization while facilitating employment of hundreds of thousands of unemployed workers during a period of economic crisis. Weaknesses in implementation included insufficient monitoring of training and weak enforcement of program standards by state level and local management. The studies constituted commendable efforts in evaluation of the major programs. CIMO, PROBECAT and PILEOT, resulting in many analyses and findings, but there were weaknesses in the administration of research. Intermediate targets met included: (a) improving the coverage and scope of CIMO; (b) strengthening coordination between CIMO and other institutions rendering support to enterprises; (c) strengthening the SEEs and improving the quality and availability of information on local and regional labor markets; (d) providing improved re-training for displaced and unemployed workers; and (e) introducing, on a pilot basis, an education and training program for under-qualified rural migrants. Furthermore, some of the targets were extensively surpassed. During the project period, almost 1.5 million workers attended 120,000 training and ancillary events. The project successfully developed information systems, strengthened STPS manpower and promoted labor related studies and research within the academic community. Other objectives were only partially met or not fulfilled at all. For instance, the CIMO study (from the 1995 survey) indicates that the enterprises assisted by the program showed lower gains in labor productivity and value-added when compared with the control group. Although CIMO enterprises registered increasing profits, these increases were smaller than those in the control group. As another example, the proposed gradual privatization of CIMO in financing and operations was not realized. Private sector  x contributions to CIMO were lower than the targets initially proposed (44 percent of total costs compared to a goal of 85 percent for training costs and 60 percent for technical assistance). The ambitious plan of transforming enterprise organizations (chambers and associations) into privately owned centers of enterprise modernization had not been achieved at project completion. Only 19 UPCs were formally incorporated into enterprise associations, but they were still dependent from STPS for financing and technical support. Major Factors Affecting the Project The economic crisis that affected Mexico in 1995 increased unemployment and placed considerable pressure on the STPS programs. The Government and the labor agency responded appropriately to the crisis by requesting more international assistance, increasing the budget allocations for these programs and making other changes. Except for some delays in budget allocations at the beginning, this project was well funded by the Government, even during periods of economic stress, because of its expected effects in alleviating the unemployment problem. The STPS confronted problems in administering and controlling the quality of some surveys and studies. Although many of the studies provided guidance for future program changes and for improving research design, in some instances, methodological flaws clouded the analysis and conclusions. Stronger and more diverse technical assistance in establishing and implementing the research p gram would have been a worthwhile investment. Also, closer monitoring by state and local level staff regarding selection of program beneficiaries and closer adherence to program standards could have increased the quality of the programs and their impact. Summary of Findings While it was found that the enterprises assisted by CIMO experienced many qualitative improvements in organization, management and incentives to staff, the statistical analysis showed lower gains in labor productivity and value-added among CIMO supported enterprises, when compared with the control group. It should be noted, however, that the surveys on which these results are based were taken in 1995, when many changes in program design were still to be made. PROBECAT fellowships were found to be a good tool for facilitating employment of job seekers with specific characteristics. Young workers receiving training in a school environment, were able to recover training costs after 9 to 17 months of work, but participants between 26 to 55 years of age with recent work experience, were unable to recover the costs of training. However, considering that another objective of PROBECAT is to help equalize labor income, the program may have helped shorten the length of employment search for those who, prior to the training, were at a disadvantage in comparison with the average unemployed worker. This is particularly the case of women. The first PILEOT study on self-employment showed increases in earnings after training (10.5 percent for men and 25 percent for women) although 40 percent spent  xi about 7 months searching for employment before beginning gainful occupations. The self-employed group grew from 27.7 percent to 35.4 percent as a result of the training, but for the overall group of trainees, only 29 percent became self-employed after training. The income of participants in PILEOT improved only slightly, from 1.3 to 1.4 times the minimum salary, after work related training and only 21 percent were found to be working when they were interviewed after training. These low results were in part a consequence of poor selection of trainees (fellowships were sought as a source of income or were given for political reasons). These results have been taken into account in program adjustments. Overall, through participation in PROBECAT (excluding PILEOT), a total of 947,715 trainees applying for jobs, 730,441 (or 77 percent) obtained jobs after training. These results compare favorably with direct job search (without training) which resulted in 586,591 placements among 2,140,299 job seekers (or 27 percent success) or with employment fairs (16 percent success). The study indicated the need for improvements in targeting the unemployed that would benefit most from the program. Two thirds of the trainees were employed after training. Labor earnings increased by about 20 percent among those trainees who had a job before training (25 percent for women and 15.5 percent for men). These increases were larger than those observed among workers earning the minimum wage. Ct cerning the PILEOT program in rural areas, results were generally of low significance in relation to the program objectives. There were no significant differences in labor earnings before and after the program. There were problems in monitoring and in selecting program participants. A few participant groups (21 percent of the total sample) were seriously committed and engaged in income generating activities after training. These groups (which included a total of 72 persons), were interviewed in-depth to determine the characteristics necessary for success. In most cases, the successful groups were long established, having been formed two to nine years prior to the training. In contrast, the other groups participating in the program had been together for a few months or a few days before program inception and showed no great interested in improving production. In general, only two groups indicated improvement in quality of their products or other benefits. Continuation with programs of this type would require a complete reformulation of its design and operation. It should be emphasized that the studies gave, in many instances, only tentative results because of methodological flaws. A more definitive judgment on these programs should await more carefully designed analysis and more recent data. New studies would need to focus on measuring program costs in relation to benefits, effects on productivity, and on the extent to which program participants fared better than they would have without the program.  xii Key Lessons Learned The main lessons learned from this project are summarized below by major programs. With respect to the CIMO program, the following is noted: (a) CIMO would need to focus much more on guiding enterprises toward productivity and client related issues; (b) In order to make enterprises self-sufficient regarding technical assistance and worker training, CIMO would need to advise firms on concrete steps they should take to identify and use relevant and effective training programs for continued staff development and how to access proficient and affordable technical advice; (c) Most of CIMO assistance has been for micro and small enterprises (only 7 percent were medium size firms), yet this group obtained lower benefits than the medium enterprises from training and technical assistance. This outcome indicates that CIMO should continue to cater to small enterprises, but selection should follow strict criteria to determine real needs for the assistance and the economic viability of the firm; (d) CIMO should introduce specific targets, that can be easily monitored by the small firms to measure productive performance of the firm and its workers, as well as changes in sales and value added; (e) Training for small enterprises needs to emphasize production and commercialization of goods and services, with preference to currently taught subjects such as administration, accounting and finance. With regard to the labor market adjustment programs (PROBECAT and PILEOT), he following are the key lessons: (a) Best practices on how to open and administer a business should be added to the training curriculum for workers receiving training for self-employment, and course duration should be increased to 3 months; (b) To enhance the impact of programs encouraging self-employment, prospective trainees should be properly screened with respect to motivation for and financial capacity to open their own business; (c) Women benefit from and make good use of the training programs if they have economic dependents; (d) Young people with more than secondary education are unlikely to seek self-employment or low-level employment after training, and should be directed to other employment assistance programs; (e) Training programs for unemployed workers should be continuously monitored to ensure relevance of contents, selection standards, and desirable employment outcomes; (f) Although capable Mexican institutions and individual consultants participated in this project, they were not well experienced in the novel methodologies (not previously utilized in Mexico) that had been proposed for studying the impacts on the program in the labor market. To minimize this problem in the future, it is advisable that STPS engage in wide competition among qualified institutions prior to appointment of consultants. These institutions should have hands-on experience on difficult evaluation studies. It is also recommended that studies be carefully supervised to ensure an acceptable final product; and (g) New studies need to focus on measuring program costs in relation to benefits, effects on productivity, and the extent to which program participants fared better than they would have without the program.  IMPLEMENTATION COMPLETION REPORT MEXICO LABOR MARKET AND PRODUCTIVITY ENHANCEMENT PROJECT (Loan 3542-ME) PART I: PROJECT IMPLEMENTATION ASSESSMENT PROJECT BACKGROUND 1. Over the last four decades the sector distribution of the labor force in Mexico has substantially changed. The share of primary sector workers fell from about 58 percent in 1950 to 23 percen?t in 1991, while employment in the manufacturing and services increased more than threefold, to 74 percent. These two sectors account currently for about 90 percent of GDP. There have been significant migration flows toward urban areas and high-income states in response to ne>. job opportunities. The skill level of the labor force has increased as a result of improvements in literacy (currently near 90 percent) and school attendance (75 percent of young adults 20-29 years old finished primary school). Nevertheless, the need for further enhancement in workers' education and skills keeps growing as a consequence of the rapid liberalization of the economy and growing competition with labor markets in developed countries, particularly after the North American Free Trade Agreement (NAFTA). 2. Since the early 1980's the Government of Mexico developed a number of employment related programs which reflected a strong commitment to build up its human capital in the context of a free market oriented strategy. These programs are run mainly by the Secretariat of Labor and Social Welfare (STPS), through the National Employment Services (Servicio Nacional de Empleo - SNE) and its network of State Employment Services (Servicios Estatales de Empleo - SEEs). Services offered include registration of job seekers and vacancies, career counseling, referrals, placements, and training. New programs were established in the late 1980s and early 1990: the Education Modernization Program (1989-94); the Industrial Modernization and Export Program (1990-94); the Modernization Development Program for Micro, Small and Medium Sized Industries (1991-94); and the National Training and Productivity Program (1990- 94). The STPS is, by law, the leading institution in matters related to employment, labor market regulations, in-service training, productivity and working conditions. The STPS was made responsible for the National Training and Productivity Program and a National Agreement for Increased Quality and Productivity signed by the Government with representatives of business and labor in 1992.  2 3. At the time of project appraisal, the main labor market issues facing Mexico were: (a) sluggish labor productivity growth compared to other developing countries at similar levels of economic development and to developed countries (these differences could be in part associated with less post-school investment in human capital); (b) unemployment and labor mobility, including higher unemployment than reported due to omission of rural areas, widespread under- employment, and low employment stability; (c) limited capacity of the labor market to absorb new entrants in the labor force and migrant flows from rural areas; and (d) constraints in monitoring and policy-making capacity by labor market agencies, associated with limited information for analysis of labor market issues. 4. The project concept was based on four main strategies. First, to promote the active participation of the private sector in training and labor productivity enhancement activities. Second, to support training for displaced workers to facilitate adaptation to new technologies and changes in production methods. Third, to provide technical assistance to medium and small and micro enterprises aimed at increasing productivity, quality, efficiency and growth. Fourth, to improve data gathering, analysis, and dissemination of information among workers and enterprises, as well as for policy-making. PROJECT OBJECTIVES AND DESCRIPTION Original Objectives 5. The objective of the project was to assist the Government of Mexico in promoting private sector-led growth and in easing the costs of labor mobility and labor market adjustment. This objective was to be achieved by: (a) increasing the productivity and competitiveness of micro, small and medium-sized enterprises; (b) reducing costs incurred by employees and employers in hiring and job search; (c) providing education and training opportunities for displaced and unemployed workers, with special attention to women; and (d) improving the availability, timeliness, and dissemination of labor market information, and, more broadly, the capacity to monitor labor market changes and to design, evaluate, and implement relevant policies and programs in the sector. Assessment of Original Project Objectives 6. The project objectives were consistent with the Bank's country assistance strategy of supporting a shift in the Government's focus toward long-term development issues, such as rationalization of the role of the public sector and increased private sector led growth. The detailed objectives correctly identified the need to promote an increased participation of private industry and service enterprises in worker training and in rising productivity. Indeed, in 1995 only 22 percent of those employed had received any special training and, prior to initiation of this project, labor productivity growth in Mexico was 1.5 percent or one half to one fourth of the grow experienced in developed countries. Furthermore, the objective of improving availability, timeliness and dissemination of labor market information addressed an acute need to facilitate  3 employment search, both for workers and for firms. Finally, the project objective focused on evaluation and policy development was well conceived, as it was expected to provide solid evidence for the future development of labor market programs. Original Project Design and Organization 7. The project consisted of three main components: a) Productivity Enhancement (28 percent of total costs) was intended to support activities aimed at increasing the productivity and competitiveness of medium, small and micro enterprises (MSMEs) by: (a) extending the coverage of the pilot in-service training program (a Multiple Support Service Program to MSMEs - CIMO) initiated under the Manpower Training Project (Loan 2876-ME) in 1987, and expanding the scope of the support services provided to MSMEs; (b) strengthening coordination between CIMO and other institutions rendering support to MSMEs; and (c) promoting the gradual privatization of CIMO by increasing the involvement of enterprise associations in financing, operation and gradual transformation of Regional Productivity Support Units (Unidades Promotoras de Competitividadt-UPCs). To achieve the latter goal, the project intended to increase the share of training costs to be financed by the private sector, from an initial 65 percent, to 85 percent at the end of the period. Additionally, private contribution for technical assistance to enterprises would increase from 30 to 60 percent. b) Labor Market Adjustment (61 percent of total costs) was designed to improve the coverage of the SEEs and provide education and training opportunities for displaced workers, through: (a) strengthening the operation and management of the SEEs, and improving the quality and availability of information on local and regional labor markets; (b) providing improved re- training programs for approximately 300,000 displaced and unemployed workers, meeting stringent selection criteria; and (c) introducing, on a pilot basis, an education and training program for under-qualified rural migrants who were ineligible for the regular re-training program. c) Information, Policy and Institutional Strengthening (11 percent of total costs) aimed at improving the availability, timeliness and dissemination of labor market information and strengthening institutional capacity to design, operate and evaluate labor market policies and programs, by: (a) supporting the development of an information system and the undertaking of in-depth surveys to provide the basis for periodic and systematic labor market data generation and analysis; (b) promoting an extensive program of studies and research on labor market issues, programs and policies; (c) supporting regular publication and dissemination of labor market information; (d) establishing a systematic framework for program evaluation; and (e) helping strengthen STPS' management and technical capacity through human resource development and modernization of the data management system.  4 Changes 8. A major factor that changed the scope of the Labor and Productivity Enhancement Program (of which this project was a part) was the approval of the Program of Essential Social Services (PROSSE) in 1995. This program was designed as a two-year time slice operation supporting the highest priority social sectors programs, including retraining and employment generation. PROSSE was part of the assistance given to Mexico by the Bank, the International Monetary Fund (IMF), the Inter American Development Bank (IDB), the United States Government and other G-10 countries to minimize the negative impacts of the crisis that ensued after the floating of the peso on December 20, 1994. PROSSE added a large amount (approximately US$140.5 million) to the financial resources available to the employment sector. In 1995 and 1996, two main items of the budget for the labor productivity enhancement program (PROBECAT and PILEOT) were absorbed by PROSSE. This change freed funds for the expansion of the CIMO program under Loan 3542-ME. In 1995, when the scope of the project was modified, there were concerns on the part of the Bank that such a large expansion could undermine the quality of the programs. However, the STPS was able to handle quite effectively the transition to a much larger program. IMPLEMENTATION RECORD 9. The current section presents a record of the project activities from inception through loan closing, in thL ollowing components: (a) Productivity Enhancement (CIMO); (b) Labor Market Adjustment; and (c) Information, Policy and Institutional Strengthening. This section describes the implementation of the program by the STPS and includes, in addition to activities under Loan 3542-ME, 49,500 training activities and 800,000 fellowships under PROBECAT that were financed by the PROSSE project during 1995 and 1996. Productivity Enhancement 10. The CIMO program operated on the basis of four main principles, applied according to demand requirements: (a) identification of opportunities to improve productivity, quality and competitiveness within the enterprises; (b) preparation of training programs for all occupational categories in each enterprise, grouping enterprises according to minimum numbers required for efficient utilization of training; (c) reorganization of production to ensure quality; and (d) enhancement of the links between enterprises and providers, marketing agents and/or integration of the productive processes, in order to assist micro, small and medium enterprises to achieve appropriate product-quality and competitive levels of production, by specializing if necessary, and by associating with complementary enterprises, and/or by participating in synchronized and flexible production schemes. CIMO placed a greater emphasis in providing assistance to smaller firms. During the project period, the assistance provided by CIMO was directed 76 percent to micro 17 percent to small, and 7 percent to medium enterprises. 11. The geographic coverage of the program grew from 29 existing training promotion units (UPCs), to 54 units and 8 sub-units, a substantial change from the original plan of adding only 5  5 units, and the number of promoters increased from 89 to 206 (less were needed than estimated in the original target of 250). Several UPCs are located in enterprise organizations (chambers and associations) where they shared the physical space, telephones, fax, messenger services, photocopy machines and utilities. This cooperation has proved useful, especially considering the capacity of these organizations to convene meetings and transmit information to the affiliated enterprises. The project provided additional furniture, computers, other office equipment and a basic library to the UPCs. The original project design was intended to assist enterprise organizations to assume, in time, a developmental role vis-a-vis individual enterprises, but this proved difficult because of technical expertise constraints as well as the relatively high financial inputs that would be needed for sustaining UPC functions. At project closing, 19 UPCs had been fornally incorporated into enterprise organizations, although most of them continue to be dependent from STPS for financing and technical support. 12. During 1993-94 the project financed, through CIMO, about 17,000 events, of which approximately 60 percent were for training and the remaining 40 percent consisted of technical assistance activities. Approximately 225,000 workers belonging to some 80,000 enterprises attended these events, and 25,000 additional enterprises received several other services. The private sector contributed 55.7 percent of the cost of these activities. By mid-1994 it became evident that the original program had been too conservative in relation to market needs. The demand for employment enhancement activities had increased as a result of the 1995 economic crisis and the approval of NAFTA. The Government responded to these new demands by substantially expanding both PROBECAT and CIMO programs, with the assistance of the Bank and IDB. Furthermore, a new program was created to serve the rural areas, PILEOT. Between 1995 and 1997, after approval of PROSSE, the all these programs experienced rapid growth. Overall, almost 1.5 million workers attended about 120,000 events during the project period. Private sector contributions amounted to 44 percent of the total costs. 13. CIMO operated with enough flexibility to adapt the program to specific needs and characteristics of various types of firms. While initially assistance was mainly determined on the basis of needs expressed by enterprises, the program evolved and currently it bases its technical assistance and training on a diagnosis of needs and problems of each firm. Prior to CIMO, micro and small enterprises did not have a culture to seek and use technical assistance, to engage in training activities for their workers, or to establish links with other enterprises to maximize production efficiency and improve earnings. This situation has changed at the firms assisted by the program. In some regions it was difficult to find individuals or firms that could provide technical assistance or could enter into contracts to conduct training. To remedy this situation, CIMO was able to contact firms in other regions (particularly in Mexico City, or other large cities) to procure such services and has also undertaken the task of developing skills in technical assistance and training in the regions where these did not exist. CIMO has also been instrumental in determining appropriate levels of costs for these services, in general being able to pay lower prices than those paid by the medium and large enterprises. It is a common occurrence among firms receiving assistance from CIMO to continue on their own to use technical assistance after the CIMO program is completed, although the extent to which this happens has been difficult to quantify. After project completion, CIMO continues to operate with IDB financing.  6 Labor Market Adjustment 14. In 1992, after ten years of operation, the SEEs still had a limited capacity to effectively assist in the operations of labor markets due to organizational and technical constraints. There was a clear need for improvements in organization and in availability of information on labor markets and in the administration of PROBECAT. There was also a need to seek training and employment solutions for rural migrants to the cities, on a pilot basis. To address these needs, the Labor Market Adjustment component was implemented based on four strategies: (a) promoting complementarity among organizations or institutions participating in the labor markets; (b) generating, using, and disseminating information on labor markets in order to improve worker placement mechanisms and promote alternative types of employment; (c) improving the training, retraining and occupational shifts for workers seeking employment; and (d) increasing the efficiency of the General Directorate for Employment (DGE). The DGE coordinated the implementation of this component at central level, and the SEEs were responsible for operations at state level. The SEEs comprises a network of 99 offices offering employment services in 83 cities and covering approximately 4,187 localities throughout the country. 15. To promote complementarity among agencies working on labor market issues, the program established Executive Committees of Employment Services with representatives from enterprises, training institutions and other agents. During the project period, 65 committees were established, considerably exceeding the initial target of 32. These committees grouped 741 agents representing 161 chambers of commerce and enterprise associations, 177 enterprises, 152 government organizations, 209 training institutions, 19 unions, and 23 other agents. The committees facilitated consultations on appropriate directions for re-training and placement services, identification of training needs and curricula modernization, and guidelines for improving the PROBECAT program to best respond to market needs. These committees, together with the SEEs, have evolved into systems of state employment, which meet periodically to exchange information, modus operandi, evaluation of labor market conditions, and organization of promotion and dissemination events. The project equipped 50 SEE state offices with computers connected in a network, and paid for contracts with information systems technicians to administer the networks. The project also financed individual computer equipment for 51 SEEs. In 22 states, project assistance complemented the equipment provided by the state Governments. SEE staff attended intensive training programs on operation of the systems. Since 1993 when the system was installed, almost 2.5 million job seekers made applications and about 2 million vacancies were advertised. These services contributed to the employment of almost 600,000 workers. These numbers represent 132, 99 and 93 percent of the original goals, respectively. 16. In addition, 69 employment fairs were sponsored with participation of 1,832 enterprises, where almost 100,000 vacancies were advertised, 130,300 job applications were made, and approximately 21,000 applicants found a job. In 1995, the Bank agreed to the acquisition of 48 mobile units for the SEEs, an initiative not originally included in project design. These mobile units considerably expanded the outreach area for placement services and for the selection of applicants for the PROBECAT program. Another initiative of the project was the opening of 33  7 information windows in the SEEs, operated by trained staff. These windows offer advise on self- employment and refer users to others sources of information operated by the Secretariat of Commerce and Industrial Development and by Nacional Financiera SN. C (NAFIN). 17. The project supported training and re-training of the labor force through the PROBECAT and PILEOT programns. Although these programs were financed through PROSSE in 1995 and 1996, the implementation description covers the period from 1992 to project completion in 1997. Four training models were used: (a) basic training for productive activities in a school environment; (b) mixed training in schools and enterprises; (c) re-training and re-conversion; and (d) training for self-employment. These programs grew very rapidly in recent years, as a result of the injection of considerable funding by the Government and by the Bank, both through this project and through PROSSE. While PROBECAT growth was modest in the first to years of project implementation (from 42,000 fellowships per year in 1992, to 46,000 in 1993), during the 1994-1997 period, the number of fellowships per year increased from 200,000 in 1994, to 400,000 in 1995 and to over 500,000 in 1996 and 1997. In addition, training in PILEOT benefited approximately 550,000 workers, who also received training for self-employment. An attempt was made to provide basic education and literacy classes at the same time as work related skills, but it proved less successful. Identification and recruitment of illiterate workers eligible for training was very difficult and dropout rates were high due to the precarious economic situation of the trainees and their low level of education. Lessons learned from these experiences were used in assisting the same target population through other programs. Information, Policy and Institutional Strengthening 18. This component was designed to improve the availability, timeliness and dissemination of labor market information and strengthen the institutional capacity of STPS to design, operate and evaluate labor market policies and programs. As part of the project, two programs were implemented: (a) a program of labor information, studies, and strengthening of the STPS; and (b) another program of support and evaluation of labor policies. 19. Information System. An information system was developed to handle data on employment, salaries and benefits, education, training, and technology, as well as on collective bargaining, agreements with union, and other information relevant to labor market regulations. The project contributed substantially to the advancement of the design of this information system. The system design was finalized in 1993. At that time, the system used DATAFLEX as base. In 1995, data from surveys began to be entered into the system and soon the system showed signs of saturation, as response times to consultations was extremely long. In January 1996 this activity was transferred from the General Directorate for Labor Statistics, to the General Directorate for Information and Telecommunications (DGIT) and the system's base was changed to INFORMIX ON-LINE. DGIT also introduced INTERNET and INTRANET for communications outside and within STPS. During 1997, four hundred WEB pages were generated in the address http:www.stps.gob.mx. The system provides information on vacancies, job seekers, training courses available, and fellowships. These systems are in constant evolution, using the latest information technology as it becomes available. Since the information provided  8 by the system is available nationwide, it makes it possible to identify the regional or local employment imbalances and decide on training needs that respond to specific market conditions. The project also supported the publication of an Information Bulletin, issued every two months, with information on plans and results of training, labor market characteristics and trends, and news on coordinating activities of public and private employment agencies. Between 1994 and July 1997 the SEEs produced 473 issues of the Information Bulletin (or 73 percent of the original goal). In eight SEEs, the state governments absorbed part of the production cost of this publication. 20. Surveys and Studies. A variety of surveys and studies were conducted through this project (see Annex 6). National surveys were undertaken in coordination with the National Institute of Statistics (INEGI). There were, however, extended delays in the surveys of enterprises and households, and several had to be canceled. As of 1998, new collaborative arrangements between STPS and INEGI were established, questionnaires for households were simplified, and the timely presentation of survey results was introduced as a clause in the bidding documents used for contracting surveys. As part of the extensive labor markets study program supported by the project, 29 studies were completed on topics related to labor market characteristics and the determinants of employment, salaries, and productivity. Among the topics covered, the following were analyzed in detail: organization, management and labor conditions of enterprises; socio-demographic aspects of employment, determinants of productivity and international comparisons; the informal sector; and employment in agriculture. As part of the project, the STPS also implemented a program to promote labor-related research among the academic community. At project completion, there were 150 labor market specialists in the network of researchers. Annual awards were given to the best studies and financial assistance was granted for dissertation work dealing with relevant labor topics. The results of these studies and surveys were used in re-directing and improving STPS programs. Of particular relevance to this project were the studies of CIMO, PROBECAT and PILEOT, briefly described below. 21. The CIMO Evaluation. The CIMO study was based on the Enterprise Operations Survey carried out between July and September, 1996, covering a sample of 1,367 enterprises, of which 622 had received assistance from the program (pilot group) and other 745 enterprises did not (control group). The information referred to the periods 1991-93 (Set I) and 1993-95 (Set II) and it was disaggregated according to enterprise size (micro, small and medium) and to whether the firms belonged to the manufacturing or the services sector. The Pilot Group had better results than the Control Group with respect to organizational changes within the firm, value- added from investments in training and technical assistance, and staff performance incentives. However, none of these innovations had significant impacts on productivity of the labor force and value-added of the enterprises. The Control Group showed more favorable results with respect to value added derived from the introduction of quality control systems. This could be an indication of greater interest among the Pilot Group enterprises in improving internal efficiency, while firms in the Control Group might have focused primarily on increasing the quality of their products or services. Micro and small enterprises and manufacturing firms in the Pilot Group showed lower utilization of installed capacity compared to firms in the Control Group. Although these results need to be confirmed through further analysis, they seem to  9 indicate that CIMO might need to modify its targeting criteria and its strategy to improve productivity of MSMEs. 22. The Evaluations of PROBECAT and PILEOT. The PROBECAT program includes both basic (school) training and training for productive activities, or "mixed training." The program was evaluated in 1994. After that evaluation, the STPS decided to redirect its evaluation efforts in 1995 and 1996 to two variations introduced to the program, namely: training for self-employment and support to local initiatives. In both instances, control groups of unemployed persons were compared with persons who received PROBECAT fellowships. In general, fellowships were found to be an effective tool to facilitate the process of obtaining a job. At the same time, specific factors were found to either hinder or help the achievement of the program objectives, and these findings were subsequently used to improve the program. For those receiving training in a school environment, cost recovery was achieved after 9 to 17 months of taking gainful employment. On the other hand, for those persons who were 26 to 55 years of age and had recent work experience, the costs of training were not recovered. However, considering PROBECAT's equity objectives, the program might have helped shorten the job search period for those who were at a disadvantage (particularly women), compared to the average unemployed worker. 23. The evaluation of PILEOT was carried out in 1996 and comprised an assessment of the impact of training in facilitating self-employment, based on a sample of 338 program participants in 5 states; and an analysis of the employment impact of local initiatives, based on a sample of 188 cases in 3 states. Results of the self-employment evaluation indicated positive returns to training. Average earning increased by 10.5 percent and by 25 percent for women. Nevertheless, 40 percent of the program participants spent an average 7 months seeking employment. Among those employed before and after training, self-employment grew from 27.7 percent to 35.4 percent as a result of the training, compared to 29.0 percent self-employed after training among all program participants. With respect to the impact of local initiatives, earnings improved slightly after training (from 1.3 to 1.4 minimum salaries per month), but as many as 21 percent of the sample was not gainfully employed after training at the time of the survey. This less than satisfactory result is attributable to poor selection of participants at the local level, where many enrolled with the only objective of receiving a subsidy during a period of great economic hardship. Program adjustments are planned to incorporate these findings. 24. Institutional Development. This component supported the strengthening the capacity of STPS staff to design, implement, and evaluate labor policies. Implementation began under the General Directorate of Employment (DGE) and was transferred to the General Coordination Unit for Labor Market Policies, Studies and Statistics (CGPEET) in 1996. Similar efforts had been supported between 1987 and 1992 by the earlier Bank financed project (Loan 2876-ME), with very positive results. The general strategy under Loan 3542-ME, consisted in strengthening the decision making process by improving technical capacity at central and local levels through the modernization of equipment and procedures, and improved processes for hiring consultants. The STPS human resources were developed by training trainers and preparing new teaching materials, both at central and local levels, with assistance from universities and other institutions specialized in in-service training. The training carried out during project implementation  10 benefited 851 SEE employees, 170 UPC trainers working at UPCs, and 155 administrative and technical staff. In subsequent years, a total of 728 courses were given to 10,789 employees. Emphasis was given to computer systems, management, accounting, and statistics. Procurement of Goods and Services 25. The equipment procured for the DGE labor market information system included 2 RISC servers, 52 servers and 512 PCs, at a total cost of $3,530,000. The equipment procured for the General Directorate of Labor Statistics consisted of 7 RISC servers, 1 server, 85 PCs and ancillary equipment, at a total of US$805,000. In both cases, more equipment was acquired at lower prices compared to appraisal estimates, due to careful procurement and lower computer costs in recent years. 26. For training centers, the project financed complementary furniture and equipment (including audiovisual) for 184 different occupation specialties in 90 centers. Of these, 89 percent of the institutions are from the government, 10 percent belong to the private sector and I percent to the non-governmental sector (NGOs). ACHIEVEMENT OF ORIGINAL OBJECTIVES 27. The project components dealing with productivity enhancement and labor market adjustment achieved or surpassed most of its intermediate targets of: (a) improving the coverage of the enterprise pilot training program (CIMO) and expanding the scope of technical assistance to MSMEs; (b) strengthening coordination between CIMO and other institutions assisting to MSMEs; (c) strengthening the SEEs and improving the quality and availability of information on local and regional labor markets; (d) providing improved re-training for displaced and unemployed workers; and (e) introducing an education and training program for under-qualified rural migrants on a pilot basis. PROBECAT surpassed several times the target number of fellowships, from 42,000 in 1992, to 46,000 in 1993, to almost 200,000 in 1994, 400,000 in 1995, and over 500,000 in 1996 and 1997. CIMO expanded its geographic coverage from 29 existing training promotion units, to a total of 62 units and sub-units. During the project period, almost 1.5 million workers attended 120,000 training and related events. CIMO operated with enough flexibility to respond well to regional and sector needs, continuously adjusting program features as results from surveys and studies became available. Through the Information, Policy, and Institutional Strengthening component, the project had considerable impact in the development of information systems, strengthening the capacity of STPS staff through specialized training, and in carrying out a successful program of promotion of specialization in labor related studies and research within the academic community. 28. Other objectives were only partially met or not fulfilled at all. For instance, the promotion of the gradual privatization of CIMO through increased participation of enterprise associations in program financing and operations, did not materialize, although substantial cost- sharing was achieved. Private sector contribution to CIMO, although substantial, was lower than the target originally proposed. The private sector was expected to absorb the cost of training  11 workers, increasing from 65 percent at project inception to 85 percent at project completion; and the cost of technical assistance to MSMEs was expected to be born by participating firms at a ratio of 30 percent initially and 60 percent at project completion. On average, the private sector financed 44 percent of the total cost of training and technical assistance provided by CIMO. The ambitious plan to transform enterprise organizations (chambers and associations) into centers of enterprise modernization and a reservoir of technical expertise (thereby transferring the developmental role vis-a-vis individual firms from the public to the private sector), proved difficult to achieve. Difficulties were primarily associated with the technical and financial constraints facing these associations that limited their ability to sustain UPC functions. At project closing, 19 UPCs had been formally incorporated into enterprise associations, but these units continue to be dependent from STPS for financing and technical support. 29. Worker training provided through PROBECAT (excluding PILEOT), benefited a total of 947,715 job seekers, and 730,441 (or 77 percent) of them found gainful employment after training. This compares favorably with direct job search (without training) which resulted in 586,591 placements among 2,140,299 applicants (or 27 percent success); as well as with employment fairs, which resulted in 16 percent success. The evaluation of PROBECAT pointed to a need for improvements in targeting the unemployed that would benefit most from the program. This group comprises women and young male adults of less than 25 years of age, and especially women with economic dependents. The fellowships would be more beneficial for those job seekers with less than high school education, compared to young people with bachelor or higher degrees. This last group proved more difficult to be helped by PROBECAT and should be referred to other assistance programs. 30. Assessing the impact of PILEOT is important, especially considering that it has utilized well over one half of the PROBECAT fellowships since 1995 (other initiatives for utilizing fellowships were proposed in 1995, but did not amount to significant branching of the main program). Two thirds of the PILEOT participants obtained employment after training, and their earnings increased by about 20 percent (25 percent for women and 15.5 percent for men) among those who had a job before training. The increase in earnings was larger for workers earning above the minimum wage, but only 40 percent of them became self-employed after training, possibly due to lack of capital to start a business. Clearly, there is room for improvements in the PILEOT program and some recommendations to this effect are presented in the section on Lessons Learned. 31. The outcome of the pilot program on local initiatives was assessed through a sample survey of activities in 33 municipalities and 39 localities. The objective of this program was to help economically active groups in rural areas to improve productivity and income. The results of this study are not conclusive. More successful groups tended to cluster in agriculture, manual crafts and food production, and were established long before the initiation of the training program (2 to 9 years). In contrast, the less successful groups were established a few months or days prior to program participation. Although the successful groups proved capable of sustained economic production after training, they reported no significant improvements in the quality of their products. In general, empirical evidence indicates that much would need to be changed in the design and operation of this program in order to ensure positive impact on rural productivity and incomes.  12 MAJOR FACTORS AFFECTING THE PROJECT Factors not subject to Government control 32. The economic crisis that affected Mexico in 1995 increased unemployment and placed considerable pressure on the STPS programs. The Government responded to the crisis by requesting more international assistance, increasing the budget allocations to STPS programs and making other changes, such as expanding the geographic coverage of existing programs and reaching into rural areas. Factors subject to Government control 33. Except for some delays in budget allocations at the beginning of the project implementation period, this project was well funded by the Government, even during periods of economic stress, because of its expected impact in alleviating the unemployment problem. Factors subject to the control of the implementation agency 34. The STPS confronted problems in administering and controlling the quality of the surveys and studies part of this project. Although many of the studies provided guidance for improvements in program and research designs, in some instances, methodological flaws clouded the analysis and conclusions. Investment in stronger and more diverse technical assistance to carry out the research program would have been worthwhile. Also, closer monitoring by state level staff of the selection of program beneficiaries (enterprises and trainees) and better adherence to established standards would have increased considerably the quality of the programs and their impact. PROJECT SUSTAINABILITY 35. The capacity demonstrated by the STPS at federal and state levels to carry out this program augurs well for its continuity. Nevertheless, reliance on external financial assistance for employment programs, which require substantial outlays in the federal budget, will continue for the medium term. In the near future, continuity of these programs is assured with assistance from IDB. While increased reliability on the state level agencies will tend to foster sustainability, appropriate evaluation and measurement of program results ought to continue at the federal level in order to ensure that program outcomes justify program costs. BANK PERFORMANCE 36. Bank staff was highly committed and cooperative, and demonstrated technical expertise, both during appraisal and supervision of the project. Although different teams were responsible for project design and appraisal and for subsequent supervision, continuity was ensured. On hindsight, however, it would have probably been better if the Bank team had stronger research expertise available throughout the project implementation period. Such expertise was made  13 available in a rather sporadic manner. More systematic and frequent research supervision would also have led project managers to seek more relevant technical assistance for the main studies. BORROWER PERFORMANCE 37. Overall, the STPS and the State Governments demonstrated an exceptional capability for launching large program expansions, in a short time, without detriment to the good quality of services and training provided. The Government also acted diligently in seeking additional financial assistance (through the PROSSE project) for employment related program expansion, in response to the economic crisis. In spite of the excellent performance of the STPS, some problems should be mentioned so they may be remedied in future investments. Improvements can be made to ensure continuous monitoring of training and other programs, and to make corrections in program design during implementation. Also, as already mentioned, difficulties were encountered in the implementation of the ambitious research and survey program. Although capable Mexican institutions and individual consultants with recognized expertise were used, they were not necessarily experienced in the novel methodologies that had been proposed for the study of program impact on the labor market. Consequently, there were problems in the execution of several studies, ambivalence as to the interpretation of terms of reference agreed with the Bank, difficulties in preparing study protocols and in identifying appropriate consultants. All these problems resulted in tentative conclusions and unclear interpretation of the results of some of the studies. ASSESSMENT OF OUTCOME 38. Overall, the project met most of its objectives and has been rated as highly satisfactory. It provided a strong basis for program improvements and, most importantly, gave impetus to labor force modernization by facilitating employment for hundreds of thousands of unemployed workers during a period of economic crisis. Issues pointing to weak aspects of implementation include the administration of research and insufficient monitoring by state and local level agencies of some of the training and productivity enhancement programs. 39. In spite of their limitations, the studies constituted commendable efforts in the evaluation of major programs: CIMO, PROBECAT and PILEOT. The studies resulted in a wealth of analyses and tentative findings. While it was found that the enterprises assisted by CIMO experienced many qualitative improvements in organization, management, and incentives to staff, the statistical analysis did not show significant differences in labor productivity and value- added when assisted firms were compared to the control group. It should be noted, however, that the surveys on which these results were based were taken in 1995, when many changes in program design were still to be made and there was insufficient elapsed time to gauge the full effects of the program. More definitive conclusions regarding the outcomes should be based on more recent data. New studies will need to focus on measuring prograrn costs in relation to benefits, identifying impact on productivity, and assessing the extent to which program participants fared better than they would have without the program.  14 LESSONS LEARNED 40. The following were the lessons learned from the review and assessment of the project: a) CIMO would need to focus much more on guiding enterprises toward productivity issues. Productivity and value-added were lower for CIMO assisted enterprises than for firms in the control group. In the latter, improvements were caused mainly by introduction of quality controls, an indication that those firms were more focused on increasing the quality of their products or services, and on client satisfaction. b) In order to make enterprises self-sufficient with regard to technical assistance and worker training, CIMO would need to advise firms on concrete steps they should take to identify and use relevant and effective training programs for continued staff development and proficient and affordable technical advice. c) Most of CIMO assistance has been for micro and small enterprises (only 7 percent were medium size firms), yet this group obtained lower benefits than the medium enterprises from training and technical assistance. CIMO should continue to cater to smaller enterprises. However, the selection of these enterprises should follow strict criteria to determine a real need for the assistance and the economic viability of the firm. This would avoid investments in enterprises that would not be able to afford the total costs of the assistance on their own or that would not be able to stand the competition due to pronounced inefficiency or market loss. d) It is important for the CIMO program to introduce specific targets of productive performance that could be easily monitored by the smaller firms, including output per worker, sales and value added. While training should continue to be customized to the needs of each firm, more emphasis should be given to production and commercialization of goods and services, rather than the current emphasis on subjects such as administration, accounting and finance. e) For those workers receiving training for self-employment through PROBECAT, there is a need to include in the courses conceptual frameworks and best practices on how to open and administer a business. To this end, course duration might need to be increased to three months and might need to include also other topics related to productivity, competitiveness and profits. f) To increase the impact of programs encouraging self-employment, prospective PROBECAT beneficiaries should be properly screened regarding their socio-demographic, economic, as well as employment history and their post-fellowship motivation and intentions. Before being accepted for training, the background of the candidates should give reasonable indications that they really wish to become self-employed and are financially capable of opening their own business. For instance, it was found that housewives who never held a job outside the home were most likely not to seek employment after training. Given that participants receive payment from PROBECAT while studying, many use the program as a means to earn some money, without a real interest in post-training employment.  15 g) In general, women benefit from and make good use of the PROBECAT training programs if they have economic dependents. The number of economic dependents tends to be positively related to the women's interest in self-employment or in a paid job after training. h) Young people with more than high school education tend not to seek self-employment or low-earnings employment after training. They should be directed to other employment assistance programs. i) The PROBECAT training programs for unemployed workers, should be continuously monitored with respect to performance of trainers, the extent to which selection criteria is consistently applied, the training offered is relevant, and impacts on trainees with respect to employment and earnings. j) The ambitious research and survey program proved difficult to be carried out. Although the STPS ensured the participation of capable Mexican institutions and individual consultants with recognized expertise, their lack of experience with novel research methodologies proved to be a serious constraint. Better research outcomes could have been achieved if experts were used in the definition and execution of complicated studies, and if wider competition among qualified institutions had been encouraged prior to the selection of consultants. This is particularly important for research requiring various kinds of expertise and hands-on experience with especially difficult evaluation studies. Finally, the administration of the project should have provided more effective follow up of studies to ensure an acceptable final product in all cases. k) New studies would need to focus on measuring program costs in relation to benefits, effects on productivity and the extent to which program participants fared better than they would have without the program. FUTURE OPERATIONS 41. At the request of the Government of Mexico, the Bank is currently preparing a new operation focusing on technical training and certification. The proposed project concept is being developed by a technical team comprising representatives of the STPS, Secretariat of Education (SEP) and CONOCER, an agency comprising government, entrepreneurs and union representatives, responsible for the development of labor competency norms and certification.  16 IMPLEMENTATION COMPLETION REPORT MEXICO LABOR MARKET AND PRODUCTIVITY ENHANCEMENT PROJECT (Loan 3542-ME) PART II: STATISTICAL TABLES Table 1: Summary of Assessments Table 2: Related Bank Credits Table 3: Project Timetable Table 4: Cumulative Estimated and Actual Disbursement Table 5: Key Monitoring Indicators Table 6: Studies included in Project Table 7A: Project Costs Table 7B: Project Financing Table 8: Bank Resources: Staff Inputs Table 9: Bank Resources: Missions Table 10: Status of Legal Covenants  17 Table 1: Summary of Assessments A.t Achtsheg,vemnt* J-ttW A '?st_ Macro policies 4 Sector policies 4 Financial objectives 4 Institutional development 4 Physical objectives 4 Poverty reduction 4 Gender issues 4 Other social objectives 4 Environmental objectives 4 Public sector management 4 Private sector development 4 Other (specify) B. Project SuttblZRi ' I4 fkI < Sustainability 4 Identification 4 Preparation assistance 4 Appraisal 4 Supervision 4 Preparation 4 Implementation 4 Covenant compliance 4 Operation 4 Assessment of Overall Outcome 4  18 Table 2: Related Bank Loans X0gu1,2m...|..W.@,~,