ReportNo. 15808-MOR Kingdom of Impact Evaluation Report

Public Disclosure Authorized Socioeconomic Influence of Rural Roacds FoL:ulth High\vav\ Project [ oarn 2254-N( )PR

June28, 1996

(iperition El%sEEalLation1 De)artnmen1t Public Disclosure Authorized Public Disclosure Authorized

Document of the World Bank Public Disclosure Authorized Currency Equivalents

Currency Unit = Dirham (Dh) US$1 = 8.63 Dh

Abbreviations and Acronyms douars Hamlet or section of a larger village DRCR Directorate of Road and Road Traffic GDP Gross Domestic Product HDM Highway Design Model MLSS Morocco Living Standards Survey MPW Ministry of Public Works qx 100 Kilograms SUNABEL Sugar Factory VOC Vehicle Operating Costs The WorldBank Washington, D.C. 20433 U.S.A.

Officeof the Director-General Operations Evaluation

June 28, 1996

MEMORANDUlM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT

SUBJECT: Impact Evaluation Report on Morocco Socioeconomic Influence of Rural Roads Fourth Highway Project (Loan 2254-MOR)

Attached is the Impact Evaluation Report (IER) on the Morocco Fourth Highway project (Loan 2254, approved in FY83). The main objective of the impact evaluation was to understand the impact of rural roads, five to ten years after completion of the improvements carried out under the project. The study focused on impacts on: (i) transport infrastructure and services; (ii) agriculture; (iii) social services; and (iv) the environment. The impact study also assessed the economic benefits of the improvements and their sustainability. The study focused on four of the ten rural roads improved under the project; the sample roads were geographically distributed in the North, Center and Center-South of the country to represent a variety of climate, agricultural, and economic conditions. The study was conducted during 1995.

The impact study compared present conditions in the areas of project roads to the situation prior to the improvements, and to conditions in comparison roads located nearby and which were not subject to improvements during the project period. The study's main finding is that the benefits of paving rural roads extend considerably beyond making traffic easier, lowering the cost of operating cars, trucks and other vehicles, and improving the quality of transport services. The extended benefits include triggering major changes in the agricultural economy, including higher outputs; transformation of the agricultural output mix, for example, from low-value cereals to high-value fruit orchards; increased use of modern inputs, especially fertilizers. Improved access to health and education facilities increased enrollment rates in rural education as well as frequency of visits to health care services, and allowed to recruit professional personnel to staff schools and health facilities. The biggest impact was on girls' enrollment in primary education, which more than trebled in the project zones during the study period. Women also benefited as the existence of paved roads sharply increased the affordability of butane for cooking and heating, dramatically reducing women's daily chores for the collection of fuelwood. In preparing the IER, surveys were conducted at the household and village levels, followed by focus groups in each of the villages covered by the study. In January 1996, as part of the preparation of the study's final report, an OED mission, together with personnel from the Ministry of Public Works, attended meetings in the road areas which included local officials from all economic sectors as well as elected, county-level representatives. These meetings were essential to interpret data and draw conclusions.

The study also showed that economic benefits were substantial, resulting in satisfactory economic returns, and that the investments were not premature. The transport, agriculture, and social impacts are likely to be sustainable.

Attachment

1

Contents

Preface ...... 3 Evaluation Summary ...... 5 1. The Project and the Socioeconomic Setting ...... 9 Fourth Roads Project: Rural Roads Component ...... 9 2. Study Methodology ...... 11 Hypotheses for Testing ...... 11 Selection of Project and Control Roads ...... 12 Characteristics of the Roads (Map IBRD 27809) ...... 12 Comparison Between Project and Control Zones and Over Time ...... 14 Other Methodological Aspects ...... 15 3. Impact on Transport Infrastructure and Services ...... 17 Rationale for the Impact ...... 15 Key Features of Morocco's Road Transport System ...... 17 Road Conditions Before the Project...... 18 Main Findings ...... 18 Traffic: National Trends ...... 18 Traffic: Project Roads ...... 19 Transport Costs and Services ...... 19 Motorization ...... 20 Improved Access to Market and Services ...... 21 Household Transport Expenditures...... 22 4. Impact on Agriculture ...... 23 Rationale for the Impact ...... 23 Key Features of the Sector ...... 23 Main Findings ...... 24 Other Related Economic Impacts...... 27 5. Impact on Social Services ...... 29 Rationale for the Impact...... 29 Accessibility to Social Services...... 29 Education ...... 29 Key Features of the Sector...... 29 Main Findings ...... 30

This report was prepared by Mr. Hernan Levy (Task Manager), in collaboration with Mrs. Claudine Voyadzis (Consultant) with input by Mr. Claudio Volonte (Consultant). Mrs. Maryvonne Mauprivez provided administrative assistance. Local consultants-Morocco Trade and Development Services (MTDS)-prepared an interim assessment. The report was issued by the Infrastructure and Energy Division (Mr. Yves Albouy, Chief), of the Operations Evaluation Department (Mr. Francisco Aguirre-Sacasa, Director). 2

Health ...... 33 Key Features of the Sector...... 33 Main Findings ...... 33 Impact on Women ...... 36 Rationale for the Impact ...... 36 Gender Indicators: A National Perspective ...... 36 Main Findings ...... 37 Other Impacts ...... 38 6. Impact on the Environment ...... 41 Expected Impacts ...... 41 Main findings ...... 41 Types of Impacts ...... 41 7. Economic Analysis ...... 45 Economic Analysis at Project Appraisal ...... 45 Findings of Importance to Economic Analysis ...... 45 Project Benefits ...... 46 Traffic ...... 46 Design of Rural Roads...... 49 8. Sustainability of the Impacts ...... 51 The Roads ...... 51 The Socioeconomic Impacts ...... 52 9. Conclusions and Recommendations ...... 55 Conclusions ...... 55 Recommendations ...... 57

Annexes

A. Study Hypotheses ...... 59 B. National Trends in Agriculture, Roads, Health, and Education...... 61 C. Study Methodology ...... 65 D. Survey Results Aggregated by Region, Project and Control Zones, 1985-1995...... 68 E. Calculation of Agricultural Value-Added, Northern Region () ...... 80 F. Economic Analysis of Road Improvements ...... 83 G. Summary of Impacts ...... 86

Bibliography

Maps IBRD Nos. 27809 and 27810 3

Preface

This is an Impact Evaluation Report (IER) of rural roads financed under the Bank's Fourth Highway Project (Loan 2254-MOR, for US$85 million, approved in March 1983 and closed in December 1990) in Morocco. This evaluation aims to understand the impact of rural roads on transport infrastructure and services, as well as on the region's economy and social welfare, five to ten years after completion of the improvements carried out under the project.

Two mrainreasons prompted the Operations Evaluation Department (OED) to undertake this study. First, rural roads account for a significant proportion of Bank clients' rural infrastructure programs, and of Bank lending for roads, thus, better understanding of the impact of these operations would be of wide interest. Second, Morocco has launched a large program to improve rural roads, and the Government expressed strong support for conducting the study. In addition, the study would be of interest to countries that are undergoing fast urbanization and that are seeking to better comprehend the cost-effectiveness of infrastructure investments to improve living conditions in the rural milieu.

The study was started in Morocco in January 1995, with the recruitment of a team of local consultants. A study launch workshop in May 1995, attended by central and local government officials, the consultants, and OED staff, initiated the field work. Subsequent OED missions to Morocco reviewed, together with the Ministry of Publics Works' Road Directorate (MPWRD), the progress in the consultants' work, and discussed with central and provincial and county-level officials the study's findings.

The study selected for analysis four out of the ten roads whose paving was financed by the project. These four roads, located in three distinct -Northern, Central and Central-Southern-were chosen to maximize diversity in agro-climatic conditions, economic characteristics of the zones, and economic function of the roads. For each project road, a control road, located nearby, was selected, which had not been subject to improvement during the study period. Socioeconomic data for the project and the control roads was collected mainly through five types of surveys, of which three were at household and village levels, one at provincial level for public works, and one at local level for agriculture. In addition, focus groups were carried out in each of the villages covered by the study. Once the interim consultants' report was ready, an OED/MPWRD mission conducted meetings in the field in January 1996 that included representatives from all economic sectors and elected county-level representatives; these meetings were essential to interpret data and to refine or correct the consultants' findings.

Copies of the draft IER were sent to the relevant Government officials and agencies concerned for their review and comments. No comments were received.

5

Evaluation Summary

Background

1. Despite sustainedurbanization through the 1980sand 1990s,about 50 percentof Morocco'spopulation remains rural. Rural inhabitantshave benefitedless from the country's economicgrowth over the last decadethan urban dwellers. A result is that over 70 percent of the poor populationlives in rural areas. Consequentwith its objectivesto improvethe welfare of its rural population,the Governmentis launchinga plan of road investmentsfor the period 1996- 2000 that aims to improveor pave 10,000kilometers of rural roads. This report, by examining the results after several years of operationof a rural road componentincluded in a broaderroad project financedby the Bank (the Fourth HighwayProject, approved in March 1983),seeks to understandthe impactsthat emanate from improvingrural roads, and how they filter beyond the physical investmentto the agriculturaleconomy and the social sectors. The intentionis to derive lessonsthat can help Moroccoand other countries better assessthe long-termvalue of investing in rural roads.

Methodology

2. This study assessesthe impactof paving and other improvements(completed between 1987 anid1991 ) to four rural roads located in three differentregions of Morocco:North, Center and South. As a starting point, the studyformulated four sets of hypothesesabout the impactsof the road improvements:(i) direct impact on transport infrastructureand services; (ii) impact on the agriculturaleconomy; (iii) impact on the social sectorssuch as health and education;and (iv) impact on the environment.

3. To test these hypotheses,the study utilized two types of analyses: first, for each of the roads considered,it comparedcurrent conditionswith those before the investmentsand, second, it comparedconditions in the projectroad relativeto a control road which did not benefit from improvementsover the period of the study. Data was obtainedfrom extensivesurveys conductedat the farm, regional and village levels, and focus groups discussionsat these levels helped interpretthe data. A monitoringsystem prepared by consultantsunder the road project, intendedto provide a large set of indicators,was not put in place as public works officials doubted its cost-effectiveness.

4. A study limitationis that since comparisonroads werejudgmentally selected at the end of the project because they had no improvementsover the studyperiod, we cannot definitively attribute changes in the communitiesstudied to road improvements. The communitiesnear improved roads may have been systematicallydifferent from those in areas wherethe roads were not improved. For example, communitiesnear improvedroads may have been targeted for other improvements. In addition,the sample is not generalizable. We believe, however,that viewed as case studies, the four improvedroads and the four unimprovedroads offer useful illustrations of some of the impactsof road improvements.

Impact on TransportInfrastructure and Services

5. All four roads studied were improved from an originally deteriorated gravel or unengineered track condition to an asphalt surface (mostly 4-meter paved width). The 6 most direct impact was elimination of frequent road closures during rainy periods, as the improved roads are open to traffic year-round. The road users benefited in several other ways: the cost of operating vehicles diopped, leading to lower prices for freight and passengers services than in roads that were not improved. Traffic on project roads increased at rates higher than before the improvement, and comprised a bigger proportion of larger, more efficient trucks. TIhesupply of road passenger services increased substantially, especially share-ride taxis offering frequent service, whereas in the past the only service was a rural bus offering as little as one run a day. Ownership of motorized vehicles increased, both of cars and trucks. The access time by the rural population to markets and social services fell drastically. In some cases, the time to access county and village administrative offices, agricultural extension personnel, health centers and rural markets, was cut by at least 50 percent. This improvement was a result of both better roads and new facilities, whose construction was made possible in part by the roads.

Impact on Agriculture

6. The studyfound that in the road projectareas overall levels of agriculturalactivity increasedin volume of production,productivity of the land, and monetaryvalues of the output. The agriculturalproduction mix was transformedas farmerswere able to shift land from low value cerealsto high value fruit orchards,which yield higher profits,thanks to the reduction in perishabilityrisks brought about by the better quality and year-roundoperability of the improvedroads. In two of the three studyregions, land used for vegetablesand fruits increased over 40 percent over the study period. Livestockproduction shifted towards pure breed cows, also a higher yield activity. The use of modern agriculturalinputs, especially fertilizers, increasedas improvedtransport made distributionchannels better. Use of agriculturalextension services by the small farms increasedby more than four times over the studyperiod. The shift to higher value products,combined with improvedyields for traditionalcrops, raised the value added per unit of cultivated land.

7. Improvementsin the agriculturaleconomy led to related economicchanges in workloads,employment on farm and establishmentof new shops; these changes followed different patterns dependingon the region. Off-farmemployment grew overallby more than six times in the project zones (comparedto about three times in the control zones) and happened across all three regions. The studyfound that agriculturalpractices in the control zones,which did not benefit from improvedroads, remain essentiallythe same today as a decadeago.

Impact on Social Services

8. The surveys showedthat while enrollmentin primary educationincreased throughout all areas covered by the study,the gains in the areas servedby the project roads, where enrollment more than doubled between 1985 and 1995,was much higher than in the control roads. In parallel, the quality of education improved,as it became possibleto recruit teachersto staff the schools, and absenteeismof both teachersand studentsdropped. The rural populationalso nearly doubled their use of health care facilities(hospital and primary care), and, similarlyto education,the quality of health serviceswas enhancedas the supplyof medicines improved, health officials launcheda campaignto staff rural health care centers with a doctor, and immunizationand other health preventionprograms became easier to implement. 7

9. Some of the social impactswere especiallylarge for women:girls' enrollment in primary education trebledover the period; expandedor new maternal and child care programswere made available and accessible,and the introductionof butane at affordableprices thanks to the existence of paved roads dramaticallyreduced women's chores of daily collectionof fuelwood for cooking and heating. Rural-urbaninteraction increased several-fold in the two directions: urban dwellersvisiting their rural relatives,and farm householdmembers visitingcities.

Impact on the Environment

10. Changesin transport conditionsand in the agriculturaleconomy had both negativeand positive impactson the environmentalthough, overall, no environmentallysensitive areas were at risk by the road projects, which did not involvenew construction. Negative impactswere those resulting from the increasedtraffic and economicactivity, especially air and noise pollution and road accidents,and the increasedused of fertilizersand other chemicalswhich in all likelihoodcontaminated the water table. Positive impactsresulted in part from the transformationof the agriculturaleconomy, notably curtailmentof extensivegoat and sheep herding-that damages the soil cover-and increasedtree plantations,and from broader use of butane substitutingfor fuelwood,whose demand is larger than the size of Morocco'ssustainable forests.

EconomicAnalysis

11. The improvementsin the agriculturaleconomy and in access to social servicestranslated into increases in the roads' traffic levels at rates substantiallyhigher than on unpavedrural roads. The study quantifiedthe economicbenefits accruingto road users in the form of savings in vehicle operatingcosts comparedto the original,unpaved roads, and the economicgains resulting from people and freight being able to move at any time, withoutthe risk of road closures. Social impacts,although real, could not be isolatedfor attributionto the road investmentand were not quantifiedin the analysis.

12. The economicreturns as well as the timingof the improvementsfor all project roads were found to be satisfactory(economic rates of return ranging between 16 and 30 percent). Even under a worse case scenarioassuming lower traffic growth and reduced benefits from operatingcosts savings, the returns would have remainedacceptable. The benefits from the investmentsare expected to accrueto farmers in the form of expanded use of commercialfreight services substitutingfor non-motorizedcarriage of loads, includingoperation of heavier, more efficient trucks. The rural populationin the roads' areas is expectedto benefit mainly by the availabilityat affordableprices of frequent servicesby share-ridetaxis.

13. Althoughsome of the benefits appearedto be dependenton paved surfacesand the economic returns are satisfactory,this does not ensurethat paving was an optimal economic solution, compared,for example,to rehabilitatingthe roads to a gravel surface maintainedat a good standard.

Sustainabilityof Benefits

14. Historic trends showingsteady traffic growth over longperiods on Morocco'spaved roads, even during droughtyears when agriculturalproduction fell, suggest that the stream of 8

benefits is likely to be sustainable. The main issue is route 603 which currently,seven years after completionof the works and in part becausetraffic reached unexpectedlyhigh levels, is in poor condition. While improvementof this road is includedunder the MW's maintenance program for 1994-1998,inadequate budgetary allocations for road maintenanceraise uncertainty about the level of benefitson this road over the long-term. The transformationand improvement of the agriculturaleconomy is based on sound businessdecisions and is likely to be sustained, with the possible exceptionof the sugar beet planting in the North (a small componentof traffic on route 603) which is uneconomicand may lose market dependingon the outcomeof the privatizationprogram for the state-ownedsugar factories. The gains in agriculture are also dependenton governmenttrade and fiscal policies,and as well as the access of Moroccan produce to European Union markets. The sustainabilityof the social service impacts appear as likely in view of the high value assignedto them by the direct beneficiariesand because of governmentpolicies and the increasedfunding it is allocatingto improve social services.

Recommendations

15. The followingrecommendations are offered:

(a) Establisha practicalrural road monitoringsystem. In the current program to pave 10,000 km of rural roads, an effective monitoringsystem could be set up with the following characteristics:(i) comprisea small number of road, agricultureand social sector indicators; (ii) arrange for each type of indicatorto be collectedby the respectivesectoral, local-level authority, maybe once every two or three years; (iii) integratethe collectionof agricultural and social indicatorsin the annual surveyscarried out by these sectors;(iv) arrange for local authorities, ideallywith assistanceby a university,to compile the data into a single report for each selectedroad, and disseminatethe report to local and central authorities

(b) Increase local communityparticipation in rural roads through:(i) exchangessuch as "village .neetings", at the planning stage, to ensure that road plans are properly and timely taken into accountwithin the planning of all local-levelactivities, and (ii) for road maintenance,by direct participationof local stakeholdersin the fundingallocation for this activity, through, for example,the establishmentof local "road boards".

(c) Considerand require if cost-effective,the introductionof mitigationmeasures such as educationcampaigns and road signs to reduce and or prevent the likely increasein road accidentsand their consequences,resulting from the increasedtraffic levels and speeds when rural roads are paved.

(d) Consideradopting identificationand evaluationmethodologies based on multi-criteria indicatorsfor rural roads that: (i) combinea minimumacceptable economic return with well defined social sector (mainly educationand health) objectivesand with social sectors investmentsintegrated in local governmentbudgets, or, (ii) for very low traffic levels and for a small part of the budget for rural roads (20 percentis a good practice), apply multi-criteria based solely on attaining social objectivesrepresented by well-definedindicators.

(e) Reassessthe optimal pavementwidth for rural roads and comparewith rehabilitationto gravel surface maintainedin good standards. The very latestversion of the Bank'sHighway Design Model (HDM) model allowsto conduct this kind of analysis. 9

1. The Project and the Socioeconomic Setting

Despite sustained urbanizationthrough the 1980s and 1990s,about 50 percent of Morocco's populationremains rural. Rural inhabitantshave eceived less benefitsfrom the country's economicgrowth over the last decade than urbandwellers. A result is that over 70percent of all poor live in rural areas. Thegovernment, through investments in rural roads, combinedwith other infrastructureand socialprograms, is makingefforts to improvethe lot of the rural population. A newplan of road investmentsfor the period 1996-2000aims to improveor pave 10,000kilometers of rural roads. This report, by examiningthe medium-termresults of a rural road componentincluded in a larger roadproject financed by the Bank in the mid-1980s,seeks to understandthe impactsthat emanatefrom improvingrural roads, and how they filter beyond the physical investmentto the agricultutaleconomy and the social sectors. The intention is to derive lessons that can help Moroccoand other countriesbetter assess the value of investing in rural roads.

1.1 Morocco is situatedon the extremeNorthwestern corner of Africa, with a land mass exceeding700,000 square kilometers,and combinedAtlantic-Mediterranean coast lines which total almost3,500 km in length. In 1995,the populationis estimatedto have reached28 million inhabitants,of which and over 50% live in urban areas. Per capita income is estimated at approximatelyUS$1,200.

1.2 Since its independencein 1956,Morocco has, as a response to climaticuncertainty, placed significantpriority on the developmentof irrigated agriculture,and has invested substantialresources in an ambitiousdam constructionprogram and improvingperformance in the industrialcrops sector, especiallysugar, vegetableoil, cereals,dairy, and livestock. In additionto its commitmentto food self-sufficiency,Morocco devotes substantialresources to the mineral industryand is the world leader in phosphateexports (13 million tons annually). Despite these efforts, Morocco'scommercial deficit has continuedto widen and reached29 million Dh in 1994.

1.3 Beginningin the 1980's,Morocco undertook impressive steps to implementa World Bank-IMFcoordinated program of structuraladjustment. The main objectivesof this effort, now largely implemented,included reducing the gap betweenthe demand and supplyof goods and services,decreasing the budgetdeficit, and encouragingexports. The central government committeditself to a greater delegationof activities from the public to the private sector and decentralizationof responsibilityto local government. At the same time as it addressed fundamentalmacro-economic reforms, Morocco pursued broad programsdesigned to improve socioeconomicwelfare in rural areas.

1.4 For administrativepurposes, Morocco is divided into 7 regions and 43 provinces. The present study concernsthree differentregions: the North-WestRegion, the Center Region, and the Tensif Region, located in the central-southernpart of the country.

Fourth HighwayProject: Rural Roads Component

1.5 Concern for improvingrural roads dates from the mid-1970'swhen Morocco's Ministry of Public Works commissioneda studyto identify8,000 km of rural roads in need of constructionand/or upgrading. These efforts were reflected in the preparationof the 1981-1985 DevelopmentPlan. A key objectiveof this plan was to "...improveand expand the provincial 10 road network to help alleviaterural poverty". The Governmentand the World Bank agreed to include in the FourthHighway project, besidesimprovements in the national highwaysystem and constructionof the -Casablancafreeway, a major investmentin secondaryand tertiary rural roads. An importantaim of the Fourth Project to be attainedthrough the rural roads it helped finance was to"...to help reduce incomedisparities between regions".

1.6 The FourthRoads Project containeda specificstudy designedto develop a methodology to measure the impact on agricultureof provincial(rural) road improvements. It was expected that such methodologycould later help improvethe methodsused to identifyrural roads with the greatest potential for agriculturaland socioeconomicimpact. The consultant's report (1989) recommendedthe establishmentof a systemfor monitoringthe impactof rural roads, and included some 40 indicatorscovering different aspects. The systemwas, however,not implemented,as governmentauthorities had strong doubtsthat the cost of setting up and maintainingsuch monitoringsystem would bring commensuratebenefits.

1.7 The governmentrecently established a 5-year program(1996-2000) with a target to improve and pave 10,000km of rural roads. An on-goingWorld Bank highwayproject is expectedto help finance 1,000km under this program. The Directiondes Routes has undertaken to screen, identifyand evaluate such routes, and is interestedin designinga cost-effectivesystem for monitoringrural roads. The present studywill providetimely inputs to that effort. 11

2. Study Methodology

This study assesses the impact ofpaving and other improvements to four rural roads located in three different regions of Morocco: North, Center and Tensift (Center-South). As a starting point, the studyformulatedfour sets of hypotheses about the impacts of the road improvements: (i) direct impact on transport infrastructure and services; (ii) improvements in the agricultural economy; (iii) impacts on the social sectors such as health and education; and (iv) impact on the environment.

To test these hypotheses, this study utilized two types of analyses: first, for each of the roads considered, a comparison of current conditions with those before the investments, and, second, a comparison of conditions in the project road relative to a control road which did not benefit from improvements over the period of the study. Extensive surveys conducted at the farm, regional and village levels allowed the collection of the data requiredfor this analysis, and focus groups discussions at these levels helped interpret the data

Hypotheses for Testing

2.1 A starting point for the design of the study was the development of hypotheses on the socioeconomic impact of rural roads. Such approach is a standard procedure in OED's infrastructure impact studies. The formulation of relevant hypotheses was based on the direct experience of OED staff and consultants on rural road projects, and on the review of literature (e.g., World Bank appraisal reports of rural road projects and relevant OED audit reports as well as studies prepared by the United States Agency for International Development and the Inter- American Development Bank). The range of impacts of transport investments is best described by the quotation below, which served as a framework for the formulation of the study hypotheses.

"Transportplays a key role in getting land into production, in marketing agricultural commodities, and in making forest and mineral wealth accessible. It is a significantfactor in the development of industry, in the expansion of trade, in the conduct of health and education programs, and in the exchange of ideas"'.

2.2 The study team developed four categories of hypotheses:

1. hypotheses dealing with the direct impact on the quality of the roads, and their related effects on traffic level and composition and on the road services offered; 2. hypotheses dealing with improvements in the agricultural economy in zones benefiting from project road construction; 3. hypotheses dealing with improvements of socio-economic indicators (as measured at the farm, village and regional levels); and 4. hypotheses dealing with the impact of project road construction on the environment.

2.3 These hypotheses and the associated socio-economic indicators developed by the study team to test the validity of each of the hypotheses are further detailed in Annex A, as well as in the respective chapters of this report.

' Owens,W., 1964. StrategyforMobility. The BrookingsInstitute: Washington, D.C. 12

Selection of Project and Control Roads

2.4 In order to measure changes in socio-economic indicators, the study team selected a sample comprising four of the ten roads which had received funding under the Fourth Roads Project. The four roads, one in the North (North-West Region, /Chefchaouen areas), one in the Center (Central Region, area), and two in the Tensift Region (Marrakech/Chichaoua areas in Morocco's central-southem area) were chosen to maximize diversity in agro-climatic zone, economic character of the zone, and function of the road. To simplify the presentation in this report, these areas are labeled as North, Center and South.

2.5 During the preparation and appraisal of the Fourth Project, Moroccan officials and consultants had seiected a control road for each of the roads targeted for improvement. In its review of these potential control roads, the study team, OED and Directorate of Road and Road Traffic staff reviewed the appropriateness of retaining the corresponding control roads for the four roads selected for this study. For three of the four control roads, it was determined that the initial selections could no longer serve effectively as control roads. In these cases, the control roads themselves had undergone improvement since the beginning of the project or were so different in character (e.g., transit or local function; type of terrain; agro-climatic conditions) that before-after comparisons would be meaningless. Table 2.1 summarizes the project and control roads ultimately used in this study.

Table 2.1 Roads Selected for the Impact Study No. Area/Province Project Road Initial Control Road Alternative Control Rd.

1 North/Larache RS 603 CT 8305 C1 8203 2 Center/Settat CT 1427 CT 1411 Not required 3 South/Haouz CT 6307 CT 6402 CT 6029 4 South/ Chichaoua CT 6308 CT 6403 CT 6457B

Characteristics of the Roads (Map IBRD 27809)

2.6 North: Route RS 603, located approximately 200 km Northwest of Rabat, cuts across the Larache and Chefchaouen provincial boundary (with the majority of the 95 km road in ), serving as a link road between two rural towns, Ksar-el-Kebir and . Improvement works began in 1987 along kilometers 1-30 and 43-60. The section kilometers 30- 43 was completed in 1991. For most of its length, the road is 4 meters in width, except for the first 13.5 kilometers, where its width is 5.5 meters. The zone of influence is characterized by higher than normal rainfall (over 800 mm), steep slopes along the Rif mountain range, high rates of erosion, with agriculture as the dominant economic activity.

2.7 Instead of maintaining CT 8305 as a control, a road which had experienced improvement along 10 kilometers did not serve a similar link function, and traversed significantly different type terrain and agro-climates, the team selected an alternative control road, CT 8203, which 13

IBRD27810

MOROCCO . kI;

-, -,/ J Iy

'-'-;1 ~ -- A1 U k>sor-eWKobir33- . r \ _ PRCPRO ADR 3OA + ; r \\'5 v~~~~~~~~~~~~~- / MAJ ROADS ;. ~~ / a- .O SEED cnE .() ~ ~ ~ ~ r AWi w t PRav5NCaCAPfTALS rS}-/R J \ 69 ~~~~~~~~~~~~~~~~~~~~~NATICNALCWPrAL

;_ f 27 ~~~~~~Ouazne PROVINCE--- EOUN0ARIE (-w-~ NTERNATIONALKBUNDARES

& X _-* ~~~~~~~~~~~~APPROXaMATEBONURY

2~~~- / I * ' B - ;2-

-U~~~~* , ~ ~~~ ~ ~ ~ ~ ~~~~ -..'aR; ~~~~~~~~~~~~~~~~~~~~~~~~~L¢> I ~

1 05 101520 25 301G o_n roIG6 :i^ -, ., -

O 5 10 ?J 20MiW6 .. \

r- ^ - -,wb-.. >,-;S A. p..Th ji , 1` bd--A _W A. ot;'^L'S Xuolzo

6- lFr- ,-- 1r | r t r r AMLIM 14 remained an unimproved track but did serve a similar link function and traversed similar terrain and agro-climates.

2.8 Center: Road CT 1427 in directly South of , also serves a link function, joining El Bourouj and Oued Zem along a 30 km distance. The 36 km control road CT 1411, which links with , has remained in an unimproved state ("etat de piste") and traverses similar terrain and agro-climate as CT 1427 and, as a result, was maintained for the analysis.

2.9 The zone of influence of CT 1427 is dramatically different from that of RS 603 in Chefchaouen, with poor agricultural resources (land, surface water), and limited rainfall (less than 300 mm yearly). The major agricultural activity of the resident population revolves around extensive sheep grazing. The major economic activity in the region is centered in mining, primarily phosphates. As a result, the economic character of the zone has evolved much more towards commerce and merchandise exchange rather than agricultural forms of activity.

2.10 South: In the Tensift region South of Casablanca, two roads were selected, CT 6307 in the Haouz province, linking Amizmiz to Adassil along a length of 42 kilometers of which 13.4 kilometers were improved during the Fourth Roads Project (to a 4 meter width), and CT 6308, in the , a 13.4 kilometer road (width of 4 meters) which provides access to the population of Eborg. As in Settat Province, the agricultural economy does not benefit from favorable land and water resources. Because there are few alternatives to agricultural economic activity, there is a greater emphasis than in Settat on dryland cereals, particularly barley and dairy production.

2.11 The roads selected in 1984 as control roads for CT's 6307 and 6308 experienced improvement in the intervening year. Instead the team selected CT's 6029, linking Tnine Oudaya to Moulay Brahim along a length of 43 kms. and CT 6457B, an access road linking Souk el Had with Mejjat along a length of 15 kms. Both these roads remained in an unimproved state at the time of this study.

Comparisons Between Project and Control Zones and Over Time

2.12 Several factors call for caution in the comparison of impact results between the control and the project zones:

a) The concept of control. The study team put special emphasis to select control roads which had not been the subject of improvements during the project period, and which were geographically near the project road. The control roads thus selected best simulated the "without project" situation to assess the net impact of the investment.

b) Baseline data. While an intention existed at the start of the project to collect baseline data, and most of such data was defined, in practice no data was collected, neither in the project roads nor in the originaf control zones. Thus, in order to recreate baseline data, retrospective questions were included in surveys. However, the recall ability, 10 years 15

later, may be highly unreliable for some indicators, for example, those requiring a quantitative assessment, such as visits to health services by members of a household. Other indicators, especially those pertaining to benchmark events of an either/or type (e.g., whether a particular type of service was available or not at the time, or whether a facility existed or not), may be recalled with a higher level of precision.

c) A study limitation is that since comparison roads were judgmentally selected at the end of the proje~ctbecause they had no improvements over the study period, we cannot definitively attribute changes in the communities studied to road improvements. The communities near improved roads may have been systematically different from those in areas where the roads were not improved. For example, communities near improved roads may have been targeted for other improvements. In addition, the sample is not generalizable. We believe, however, that viewed as case studies, the four improved roads and the four unimproved roads offer useful illustrations of some of the impacts of road improvements.

2.13 Since, as suggested above, the reliability of 1995 data is higher than that for 1985, this report gives priority in the analysis of impacts to comparisons between project and control in 1995, with references to 1985 generally used in a second order.

2.14 In addition to the comparison of the project with the control zones, information is provided about national trends in the areas where the impacts were assessed (Annex B). This information is limited due to the scarcity of time series covering the whole period under study, as a number of national statistics were either staited or discontinued during this period. In particular, it was not possible to prepare regional trends in the project's areas relevant to the type of impacts to be measured; such trends would have provided an additional basis to clarify the attribution of impacts in relation to the road improvenments.

2.15 A last caveat refers to the weather, and its influence on agricultural and socio-economic conditions. During the 10-year period under study, Morocco experienced periods of drought, when traditional agricultural production, notable cereals, fell sharply. The latest agricultural statistics, which were used in thiz report, were for the year 1993/1994, a year when rainfall was above normal. Thus, comparison of agricultural output with earlier years is somewhat distorted. However, for other aspects of the agricultural economy, as well as for other socio-economic impacts, rainfall conditions are immaterial to the analysis.

Other Methodological Aspects

2.16 The design of the questionnaire, the selection of the survey sample (at village, household and farm levels), the study launch workshop, and data collection and analysis are described in Annex C. Aggregated survey results by region are presented in Annex D.

2.17 In preparing the survey, a special effort was made to generate information that would allow to isolate the impact on gender. Also, the household data was stratified by farm size, as it was anticipated that this variable could be a significant determinant of some impacts.

2.18 One study limitation is that, while it was a stated objective of the project to reduce local income disparities, the causal links between traffic changes and incomes could not be firmed up 16 because the roads sample is smalland there are many factors exogenousto road improvements which must be controlledfor.

I.19 Finally, another importantlimitation is acknowledged: in selectingunimproved roads as controls, the methodologyfocuses on the impact of road improvementbut not on the question of whether and to what extent achievingthis improvementwith Bank's help, rather than without it, makes a difference. Thus, the sutdy answersimportant questions about the nature, magnitude and sustainabilityof impactsfrom one project in the Bank portfolio;but, it cannot answer questionsabcut the effectivenessof the Bank's interventionrelative to other ones. 17

3. Impact on TransportInfrastructure and Services

Alifour roads studied were improved from an originally deteriorated gravel or unengineered track condition to an asphalt surface. The most direct impact was year-round operation of the roads, eliminating frequent road closures during rainy periods. The road users benefited in other ways: as a result of the better road surface, cost of operating vehicles dropped, leading to lower prices for freight and passengers than those on roads that were not improved. Traffic on project roads increased at rates higher than before the imnprovement,and comprised a bigger proportion of larger, more efficient trucks. The supply of road passenger services increased substantially, especially share-ride taxis offering frequent service, whereas in the past the only service was a rural bus offering as little as one service run a day. Ownership of motorized vehicles increased, both of cars and trucks. The access time by the rural population to markets and social services was drastically cut; in some cases, such as access to county and village administrative offices, to agricultural extension personnel, to health centers and to rural markets, the time to reach these services was reduced by at least 50 percent. This improvement was the result of both better roads and new facilites, whose construction was made possible in part by the roads.

Rationale for the Impact

3.1 The project was intended to have a direct impact on the quality of road conditions. This was expected to reduce the operating costs of vehicles, to allow motorized and larger vehicles to operate at all times, and to make road maintenance easier. In turn, such effects could be expected to stimulate the provision of more commercial transport services to the project's areas, and in some cases, the improved roads could become alternative routes for use by long-distance traffic. A complementary effect expected was improved access by the population to markets and social services reducing time and costs, and ensuring access under all weather conditions.

Key Features of Morocco's Road Transport System'

3.2 Morocco's road system comprises about 58,000 km of classified roads, of which about 28,000 are paved. The quality of the network is uneven, notably regarding pavement width: in only 1,500 km do pavement widths exceed seven meters, or the equivalent of two conventional lanes; in 16,000 km, the width is 6 meters or less, and in 20 percent of the network, the paved width is 4 meters (barely more than one conventional 3.5 meter lane) or less. In addition, there are some 30,000 km of unpaved, mostly rural roads.

3.3 Inter-urban bus and truck services are operated mainly by private companies and cooperatives. About one third of road freight is carried in small trucks (less than 8 ton net weight), which operate mostly in the rural areas, although these trucks also do substantial long- distance hauls. Inter-urban road passenger services is mainly by large buses, but collective, large taxis (mostly diesel-powered Mercedes) have become increasingly popular in the last year; nationally, taxis account for some 15 percent of the passengers carried. However, in many rural areas, such taxis are the prevailing mode of transport.

World Bank, i993. Kingdom of AMorocco-TransportSector Strategy Paper. World Bank, Middle East and North Africa Region, 'ashington, D.C. 18

Road Conditions Before the Project'

3.4 Overall, road conditions at the time of project appraisal in 1982 were characterized by low levels of traffic, poor road surface and periods of road closure, as described in Table 3.1.

Table 3.1 Road Conditions in 1982, Before Project Region Road Traf fic Surface Periods of Road Closure (vehlday)

North 603 40 gravel, poor condition about 90 days Center 1427 150 unengineered track4 about 60 days South 6307/08 54 gravel, poor condition during rainy season

Main Findings

3.5 All four roads were improved to an asphalt surface. The improvements were completed in the following dates: route 603 in 1987, route 1427 in 1988 and routes 6307 and 6308 in 1991. Traffic on all roads increased substantially after construction, although there were significant differences in the growth rates among the roads (Table 3.2). The better road surface, and improvements in road alignments, substantially reduced the cost of operating vehicles; at the same time, road closures were practically eliminated.

3.6 In 1995, road conditions were as follows:

Table 3.2. Road Conditions in 1995, After Project Region Road Traffic Traffic growth Surface Periods of Road (vehlday) (°//year, since 1982) Closure North 603 640 23.8% bitumen, 4-5.5m nil Center 1427 275 4.8% bitumen, 4 m nil South 6307/6308 192 10.3% bitumen, 4 m nil

Traffic: National Trends

3.7 Road transport in Morocco measured in vehicle-kilometers over the network has grown at an average 4.4 percent per year since 1982 through today. Growth accelerated to over eight percent since the later years of the 1980s. This growth was mostly due to an increase in the vehicle fleet and to more intensive utilization of vehicles. The road network itself grew little: during the 10-year period 1984-1994, the length of the overall road network increased by only 10 percent (or less than one percent per year on average). The highest expansion in the length of the network, percentage-wise, at 13 percent, occurred in the tertiary network, which comprises most rural roads and which carry low traffic (Annex B).

3Maroc Setec, 1982. Feasibility study: Evaluation Economique d'un 4eme Projet Routier. Maroc Setec, Setec Economie.

4 'A 1'etat de piste'. 19

Traffic: Project Roads

3.8 North: Chefchaouen - Route 603. As noted above, the roads studied in this report experienced different rates of traffic growth. The most striking case is Route 603, where traffic grew at an average annual rate almost five times as big as that of the national network as a whole over the same period. This high growth is explained by the location and role of this route. It connects two major population centers, Ksar-el-Kebir and Chefchaouen; in addition, the improved road became an attractive route to two other major centers in the area of Ksar-el-Kebir, I arache to the North and Souk-el-Arba-du-Rharb to the South. Prior to the improvement of route 603, its poor condition required drivers to use an alternative route (Provincial Road 28), south of route 603. In some cases, this practically doubled the distance: for example, Ksar-el- Kebir to Cliefchaouen is 86 km via route 603, 150 km via PR 28. Thus, after the improvement, Route 603 ceased to be purely a rural road, and became an important road for long-distance travel (see map in next page).

3.9 Centcr: Settat - Route 1427. Traffic grew at about the average for the national network. This .oute is located in an area with a higher road density network than route 603, and this is partly the reason why this road generated a lower traffic increase than the Northern project road. However, since 1982 to 1987, year when the improvements were completed, traffic growth on the unpaved road is likely to have been in the order of two percent per year. This indicates that since completion of the improvements, traffic growth accelerated.

3.10 South: Marrakech - Routes 6307/6308. Traffic increased at a high rate on this road, mainly because of its poor condition prior to the improvement, and because after the improvement the road took a transit function for long-distance traffic which was not originally foreseen.

3.11 Traffic composition. In addition to increases in traffic levels, the composition of traffic changed, as expected, as the improved roads allowed heavier freight vehicles to circulate. The percentage of large trucks on these roads in 1995 ranged between 10 and 13 (compared to between 0 and 2 percent in 1985); for rural roads, this is a high percentage (Table 3.3). It is also high for rural roads when compared to the average percentage of trucks (13%) for 41 national and secondary routes surveyed by the MPW in 1994.

Table 3.3 Composition of Traffic in Project Roads, 1995 (Vehicles per Day) Road Cars&Light Buses Trucks-small Trucks-large Total % % of Large Trucks (less 8 tons) (over 8 tons) Trucks Trucks Rt 603 (North) 479 11 65 85 640 23 13 Rt 1427 (Center) 205 3 32 35 275 24 13 Rts6307/08 145 2 35 20 192 29 10 (South)

Transport Costs and Services

3.12 The improvement in the roads resulted in a significant reduction in the cost of operating vehicles (see Chapter 7). In some cases, such drop in operating costs was reflected in lower transport rates offered by commercial trucking services. For example, data from a focus group 20 for routes 6307/6308 indicates that the rate for a truckload of merchandise between two population centers some 10 km apart (between Amizmis and one of the "douars" 5of the Assif Al Mal county) went down from 300 Dh before the project to less than 150 Dh once the road was improved. 1n some cases, the reduction in the price of transport services paid by the local population was purely due to the better surface condition, which resulted on lower operating costs of the vehicle circulating on them. In other cases, such as the case of Chefchaouen mentioned above, transport prices decreased for many road users because a new road alternative became attractive that was substantially shorter in distance than the altemative existing road.

3.13 Another benefit was an increase in the quality and frequency of commercial transport services. For example, focus groups reported that on routes 6307/6308, prior to the improvement of the road the only passenger service was a daily run of a rural bus. Today, a fleet of some 40 collective taxis serve these roads, with a frequency of several taxis per hour.

3.14 Survey data also revealed that annual transport costs per unit of cultivated area (expressed in current Dh) of farming inputs (fertilizers, herbicides, seeds) decreased drastically in the project area of (no information is available for the other two provinces) compared to the control areas. The annual transport cost of agricultural products to markets decreased in both the project and the control areas between 1985 and 1995, but the gains are substantially greater in the project than in the control areas (Figure 3.1).

Motorization

3.15 Overall, ownership of motor vehicles in the project zones increased. Ownership of cars increased about 3 times, reaching in 1995 one car per every 10 farms, compared to one car per every 30 farms in the control zone. Regionally, the largest increase happened in Settat, where many Moroccans residing in Europe bring cars on their visits home. Ownership of trucks followed a similar pattern, increasing in the project areas by about three fold, to reach in 1995 one truck for every 11 farms, compared to no change in the control zones, where truck ownership remained close to zero.

5 In Morocco's regional administrative system, the "douar" is the lowest level unit in the rural areas. A douar generally comprises anywhere upwards of some 50 agricultural units. 21

Figure 3.3 Transport Costs of Agriculture Inputs and Outputs per Unit of Cultivated Land in Chefchaouen Province 25 Outputs Inputs

20 58% CD

U15

4)~~~-4

0.

85 95 85 95 85 95 85 95 Project Control Project Control

Improved Access to Markretsand Services

3 .16 Surveys and focus groups reported a substantial reduction in the time required to access markets and services resulting from a combination of effects of the improved roads: higher frequency of road transport services, improvements in the quality of transport mean (from wvalkingor animal-drawn cart to motorized vehicle), and establishment of new facilities (markets, school, health centers) triggered by the improved roads. Data for the control zone was not as complete; available information showed that gains were also achieved in these zones, mainly by the establishment of new facilities, but gains were generally smaller than in the project zones. 22

3.17 Figure 3.2 provides an overview of the improvements in access in the project zones. (Changes in transport mode from horse to walk took place when distances shortened following construction of new facilities).

Figure 3.2 Reaching Markets and Services: Modes of Transportation and Time Traveled (hours). (Aggregated for the three provinces in project zones only.)

Modc of Transportation Facility Time to Facility (hours) 1985 1995 Horse Taxi VillageAdm.

Horse Taxi Agr. Extension

Horse Walk CountyAdm.

Horse Walk HealthCenter

Horse Horse Souk

Horse Horse Water Supply 01985

Walk Walk School 0 1995

0 1 2

Household Transport Expenditures

3.18 Household transport expenditures over the I 0-year period to 1995 increased (in current Dh) both in the project and in the control zones, but substantially more in control (188%) than in the project zones, despite the fact that annual household expenditures overall increased more in project (148 percent) than in the control (87 percent) zones. At the same time, as noted later in this report, improvements in the agricultural economy and in social indicators were also better in the project zones. Generally, higher household expenditures and better overall social welfare lead to higher transport expenditures, but this did not happen. A likely reason is that improvement of the roads did cause transport prices, in current terms, to increase less in the project zones than in the control zones, even as the quality and frequency of road transport services improved more in the project zones. 23

4. Impact on Agriculture

The studyfound that overall levels of agricultural activity increased in volume ofproduction, productivity of the land, and monetary values of the output. The agricultural production mix was transformed, as farmers shifted land use from low value cereals to high value fruit orchards, which yield higher profits, thanks to the reduction in perishability risks brought about by the better quality and year-round operability of the improved roads. In two of the three study regions, land usedfor vegetables andfruits increased over 40 percent. Livestock production shifted towards pure breed cows, following the trend toward more profitable investments. The use of modern agriculural inputs, especiallyfertilizers, increased as improved transport made distribution channels better. Use of agricultural extension services by smallfarms increased by more than four times over the study period. The shift to higher value products, combined with improved yields for traditional crops, raised the value added per unit of cultivated land. Improvements in the agricultural economy led to related economic changes in workloads, employment in farms and establishment of new shops; these changes followed different patterns depending on the region. Off-farm employment grew overall by more than six times in the project zones (compared to about three times in the control zones) and happened across all three regions. The study found that agricultural practices in the control zones, which did not benefit from improved roads, remain essentially the same as a decade ago.

Rationale for the Impact

4.1 The road improvements, especially their projected year-round use operability and substantially lower costs of operating freight and passenger vehicles, were expected to benefit the agriculture economy in the projects' area through higher farm-gate prices; lower costs of delivering productive inputs (fertilizers, insecticides, seeds, etc.); cheaper and more reliable availability of trucking services and, better accessibility to agriculture extension services. In turn, these effects were expected to lead to transformation of the agricultural output mix towards higher-value but perishable produce and to increases in the amount of planted area.

Key Features of the Sector6

4.2 Despite a high growth rate of urbanizaton during the 80s of about 4.8 percent per year the rural population in Morocco remains high, close to 50%,. Over 70% of all poor live in rural areas. Currently, agriculture contributes close to one quarter of the GDP, 40 percent of the employment, and 30 percent of the exports. Thus, agriculture is a vital sector, both from the social and the economic viewpoints.

4.3 Most agricultural lands are rainfed. Irrigated areas are only some 12 percent of total cultivable land. It is estimated that the potential for further expanding irrigation is limited, probably no more than by 20 percent of the existing irrigated areas. The potential for extensive agriculture is practically nil, as already lands of marginal value for agriculture are being cultivated. Thus, increasing the efficiency in the use of land is essential for the development of Moroccan agriculture.

6Most of the information in this section is based on World Bank, 1994. Kingdom of Morocco - Agricultural Sector Strategy Paper (Yellow Cover). Middle East and North Africa Region, The World Bank, Washington, D.C. 24

Main Findings

4.4 The roads opened between 1987 and 1991. The latest agricultural data available is from surveys done in 1995. Data prior to road opening was obtained for 1985. Information collected for these two years' reveals significant changes over the 10-year period, in the areas where the roads were improved. The latest completed agricultural year at the time of the survey was 1993- 1994, when there was above-normal rainfall, and thus higher than average agricultural output, especially for grains.

4.5 The overall level of agricultural activity increased (Figure 4.1). Yields for the main crops increased in all three project zones. The most notable increase happened in fruit orchards, which, on average for the three zones, passed from 13.3 qx per hectare in 1985 to 17.4 qx per hectare in 1995, or a 31 percent increase during the period. These increases can be traced to improved seeds that require irrigation, and which the roads, that became passable practically year-round once the works were completed, made possible; the improved roads encouraged farmers to invest in on-farm irrigation, such as wells and pumps. Cereals yields also increased significantly, but such improvements are more difficult to attribute to changed economic conditions in the project zones, since cereal yields were influenced by the above average rainfall in 1993-1994.

Figure 4.1 Changes in Agricultural Productivity (Output per Unit of Cultivated Land) in Project Areas Between 1985 and 1995 (percentage change)

FruitTrees

HighValue 0 Marrakech Vegetables E3 Settat M Chefchaouen Cereals l

20% 40% 60% 80% percentchanges between 1985 and 1995

7Based on agriculturalstatistics and survey data. ' It was not possible to obtain agricultural data at the ioad project level for earlier years, which would have permitted to calculate more normal values (e.g, three year averages) of productior levels. However, the comparisons with the control zones allow to take this factor into account. 25

4.6 The agriculture production mix was substantially transformed (Figure 4.2). This was due mainly to an overall decrease in the land made available for low value agriculture, especially cereals and livestock production, combined with increased land available for the cultivation of high-value irrigated vegetables and fruits. The amount of land used for fruit plantation increased in all three zones: eight percent in Marrakech, and 42 percent in Chefchaouen and in Settat. Land used for high-value vegetables increased in Settat and Marrakech; however, it decreased in Chefchaouen mainly due to a shift to industrial sugar beet production, which also generated higher incomes for the farmers (Box 4. 1).

Figure 4.2 Changes in Agricultural Land Use (Hectares per Type of Use) in Project Areas Between 1985 and 1995 (percentage change)

Fruit Trees

High Value Vegetables O Marrakech . Settat Cer * Chefchaouen

-50% 0% 50% 100% 150% percentchanges between 1985 and 1995

Box 4.1. Introduction of New Crops

Cult.vation of sugar beets in Chefchaouen is an example of the influence of improved roads on the introduction of new crops. Planting of sugar beets was introduced in the region during construction of a dam and putting in place an irrigation system in the Loukos area. However, collecting and distributing the crops was expensive and risky in view of the poor condition of the roads. Near the village of Tatouft, only 4 farmers produced sugar beets in 1985. As route 603 was improved, 90 percent of the farmers located close to the road (about 150 farmers) were producing sugar beet in 1995. Cultivation of sugar beet and other industrial crops, while yielding higher profits for the farmers, are intensive in labor use, and generated substantial on- fzrm employment.

4.7 The transformation of the agricultural economy followed the well known Von Thunen model (Figure 4.3). In this model, the "economic" distance to market is reduced as roads are improved, encouraging farmers to substitute grains by vegetables, which yield higher profits but due to their perishability require reliable and speedy transport. 26

Figure 4.3 The Von T'hunen Model

RENT ($)

Vegetable (expensive to trarisport)

Grain

Land Use Vegetable Grain Undisturbed Forest DISTANCE TO MARKET

4.8 Use of modern inputs increased. This was a key factor in the improvements in the agriculture economy in the project areas. In particular, the use of fertilizers doubled in Chefchaouen and increased by 60 percent in Settat. Overall, Chefchaouen experienced large increases in the use of other inputs such as improved seeds and herbicides, whereas Settat had the largest gains in equipment (tractors, covercrops, trucks).

4.9 Use of agricultural extension services increased This happened especially in the small farms, where yearly contacts with extension centers went up from less than one before the road improvements to more than four after such improvements, while the corresponding increase in the control zones was from 0.5 to 1.5. Such gains in the use of extension services happened across all three project zones considered in this study. The gains in the larger farms, which had more in-house know-how, were substantially smaller than gains in small farms (about 150 percent). In fact, gains were lower in large farms in the project areas than those registered by large farms in the control zones (450 percent). This finding is consistent with that related to the gains made by the small and medium farms, but not the large farms, in the use of modern inputs.

4.10 Transformation in livestock production. Investments decreased in every class of livestock except improved breeds (imported pure breeds or hybrids). Across all regions, farmers reported that they reduced the numbers of all types of local beef/dairy, sheep and goat herds (with the exception of a small increase in goat herding in Marrakech). Local beef production decreased steeply, ranging from a 22% drop in Chefchaouen to a 42% drop in Marrakech. On the other hand, investment in pure breed cows increased by about 150% in the project areas, reflecting the shift to investments that could generate higher profits. By comparison, the number of pure breed cows in the control zones also increased, but by only 50% over the period.

9 As presentedin Chomnitsand Gray, 1994. Roads,Land, Marketsand Deforestation:A SpatialModel of Land Use in Belize. PolicyResearch Department, World Bank, Washington,D.C. 27

4.11 The value added per hectare increased (Figure 4.4). As a result of increases in yields, and transformation of the agricultural and livestock output mix, farmers derived a substantial gain in value added per hectare. The net agricultural value added per unit area of production, increased in each one of three project zones: 46 percent in Chefchaouen, 36 percent in Settat and 261 percent in Marrakech. Details of value added analysis are in Annex E.

Figure 4.4. Changes in Agricultural Value-Added (Dh per Unit of Cultivated Land) in Project Areas Between 1985 and 1995 (percentage change).

Marrakech

Chefchaouen

Settat

0% 50% 100% 150% 200% percentagebetween 1985 and 1995

4.12 The surveys and the time-series data neither corroborated nor negated the hypothesis that the total planted area would increase as a result of the better roads and the concomitant reduction in transport costs. Overall in Morocco, the harvested land increased by close to 25 percent between 1985 and 1994 (Annex B).

4.13 Comparison with the Control Zones. Focus groups discussions and visits by study members familiar with the control zones reported that agricultural practices remain generally similar to those employed a decade earlier. The lack of modernization is widely explained by the lack of transport access to markets where improved inputs can be purchased and where agricultural surplus can be marketed. The differences (crop development, land under cultivation, on-farm water management) between the project zones served by paved roads and the control zones served by tracks are dramatically visible. For example, in both Marrakech and Settat control zones, there has been little or no introduction of improved livestock varieties, high-value vegetables, or tree cropping. Even in the cases where there has been limited introduction of livestock, the dairy productivity (1,200 liters/animal per year) is significantly lower than the rates achieved in the project zone (2,500 liters/animal per year). In the control zones, more than 90 percent of the available agricultural land is taken up with relatively low value cereal cultivation.

Other Related Economic Impacts

4.14 Workload and Employment. The main finding is that the number of days worked outside the farm by members of the farm household increased between 1985 and 1995 in both the project and the control zones. The number of days increased 6 times in the project zones, compared to about 3 times in the control zones. Most of the increase occurred for men, although more women go to work today outside the farm in the project zone (average of three) than in 1985 (average of 0.15). Large farms, that invested heavily in equipment following the road improvements and reduced labor requirements, showed the largest gains in outside employment 28

when compared to the control zones. While the increase of off-farm employment in all zones reflects the increasing integration of the rural with the urban economies, the substantially higher gains in the project zones can be attributed to two effects of the road improvements: i) a direct effect, resulting from a drastic drop in access costs from farms to outside employment, as the improved roads reduced transport costs, and ii) an indirect effect, stemming from the higher modernization in the management of the farms on the projects' zones, which released household labor for outside employment.

4.15 On-farm employment, that is, people hired from outside the farm households to work on the farms, followed a different pattern. It showed a decrease in the project zones (by 7% on the aggregate for all three regions), while remaining practically constant in the control zones. Overall, the findings on household workload, on-farm and off-farm employment appear internally consistent and coherent with other findings related to the modernization of agricultural work in the project zones. Chapter 5 discusses the impact of the changing demand on women.

4.16 Establishment of New Shops. The study showed no clear overall impact when all three regions are considered. However, there was a marked rise in Settat, where the number of shops in the project area increased from one in 1985 to 10 in 1995, while the corresponding increase in the control zone was from 2 to 6. The most striking gain in commercial activity in this zone is the rise in the number of kiosks ('tentes') in the weekly market (souk), that went up from 10 in 1985 to more than 500 now. Another indicator is the increase in the number of basic food stores ('epiceries') in villages, as 20 such shops were established in Larbaa, and five in Lamazia. The increase in shop activity in this region, compared to the other two in the study, can be attributed to the lesser agricultural possibilities in Settat, which led to relatively higher shop activity. The finding on the increase in number of shops and kiosks in the market is corroborated by data provided by focus groups discussions, which reported that tax revenues from commercial activity went up from 1,500 Dh in 1985 to 10,000 Dh in 1995 (equivalent to about 7,000 Dh in constant 1985 prices).

Lox 4 Interview with a New Shop Owner on Road 603, Province of Chefchaouen

A small cafe and grocery store was recently built along the road. The owner, third son of a nearby farmer, had seized the opportunity to build this cafe and derived his income from it. He is 27 years old and still a bachelor. When we visited him, it was about 4 p.m.; the little cafe was full of young men. The grocery store was filled with items such as sugar, flour, coffee, tea and some fruit and vegetables. A television in the cafe was broadcasting the local news to which some clients listened while others played cards or discussed village events. The owner told us that he had built this facility alonig the road once the road was rehabilitated, and that he was better off working in the cafe/grocery store than working in his father's farm. The cafe was a kind of gathering place for these young men who before had nowhere to go as their farm was too isolated. (Reported by study team members, Voyadzis and Benaziz). 29

5. Impact on Social Services

The surveys conducted by the study showed that while enrollment in primary education increased throughout all areas covered by the study, the gains in the areas served by the project roads where enrollment more than doubled between 1985 and 1995, was much higher than in the control roads. In parallel, the quality of education improved, as it became possible to recruit teachers to staff the schools, and absenteeism of both teachers and students dropped. The rural population also nearly doubled their use of health care services (hospital andprimare care), and, similarly to education, the quality of health services was enhanced as the supply of medicine improved, health officials launched a campaign to staff rural health care centers with a doctor, and immunizaton and other health prevention programs became easier to implement. Women benefitted greatlyfrom the project roads: girls' enrollment in primary education trebled over the period; expanded or new maternal ai,d child care programs were made available and accessible, and the introduction of butane at affordable prices, thanks to the existence ofpaved roads, dramatically reduced women's chores of daily collection offuelwoodfor cooking and heating. Rural-urban interaction increased Feveral-fold in the two directions, urban dwellers visiting their rural relatives, andfarm household members visiting cities.

Rationale for the Impact

5.1 The improvement of the roads was expected to facilitate access by the population to existing services, notably education and health, and to encourage the government and local authorities to establish new facilities in those areas. A direct effect was expected to be the increased enrollment in primarv schools and visits to health facilities, with consequent improvements in education and health indicators.

Accessibility to Social Services

5.2 Access to social services is inadequate in rural Morocco. Accessibility is worse in the more isolated areas, due both to scarce social services facilities and to deficient transport infrastructure and services to reach them. Comprehensive, long-term plans for the social welfare of the country were inexistent since social concerns were not a top priority in Morocco". Spending to promote human capital was not considered an investment that would ultimately lead to greater and more sustainable economic growth. It was only in 1993 that the Government issued the Social Development Strategy to extend basic social services to the poor and increase their participation in economically productive activities.

Education

Key Features of the Sector

5.3 In recent years, Morocco has made a great effort to improve education. Its budget for the sector is about one fifth of the central government's budget and over five percent of the GDP. Substantial progress has been made in improving enrollment in secondary and higher education, which has grown at annual rates of 4.3 and 9.1 percent, respectively. However, it has been more

10SAR, No. 15073-MOR. 30

difficult to achieve results in primary education; enrollment rates following an economic recession actually declined between 1983 and 1988, and did not recover its previous level until 1991.

5.4 The current situation is defined by the following indicators. Illiteracy is 58 percent, and it is substantially worse in the rural areas, where it reaches 77 percent (37 percent in the urban areas). In the countryside, schools are in short supply and those that exist are poorly equipped, making it difficult to attract teachers. About 56 percent of the communities have a primary school, and only 3 percent have access to a paved road." School attendance rates in the rural areas are low: 48 percent overall, 64 percent for boys and 32 percent for girls. Important causes for this are on the demand side: parental attitudes (particularly illiteracy of mothers), and the opportunity costs of sending children to school, as many of the rural poor cannot afford to forego children's labor or earnings. On the supply side, causes range from reasonable and accessible school facilities to provision of meals.

Main Findings

5.5 Improvement in Primary School Enrollment Rates. This was the main impact of the road improvements on education. In all three areas combined, primary school enrollment rates in 1995 reached a high level, for rural areas, of 68 percent. This compares to 51 percent in the control zones. Relative to 1985, enrollment rates in the project zones more than doubled from the initial level of 28 percent. While enrollment in the control zones also improved over the period, the rate of improvement was lower.

5.6 There are substantial regional differences in the absolute enrollment levels reached, and in the gains in the period (Table 5.1). The highest enrollment rate was for boys in Settat, reaching 85 percent in 1995. The highest enrollment rate for girls was also attained in Settat, with a high 67 percent in 1995. The highest gain was for girls in Chefchaouen, where enrollment rates increased from 10 percent in 1985 to 38 percent in 1995.

5.7 Girls' gains. As shown in Table 5.1, girls achieved the biggest gains in enrollment in primary school, which reached 54 percent in 1995, more than three times the level in 1985; there is insufficient information about the control zones. More details on the impact on girls' education are given later in this chapter.

Khandker, Lavy and Filmer, 1994. Schooling and Cognitve Achievements of Children in Morocco. Can the Government Improve Outcomes? World Bank Discussion Paper No. 264. The World Bank, Washington, D.C. (page 17). 31

Table 5.1 Boys and Girls (age 7 to 15) Attending School, in Project Zones, Before and After Project (Percentages)

1985 1995 Percentage Change All three regions aggregated Boys 39 81 108% Girls 17 54 220%

Chefchaouen Boys 49 81 65% Girls 10 38 287%

Settat Boys 44 85 93% Girls 26 67 162%

Marrakech Boys 34 80 134% Girls 15 52 250%

5.8 Impact on the Quality of Education. In addition to gains in enrollment, the improved roads had other effects, some directly related to the increased enrollment, some related to improvements in the quality of the schools and of education. Focus groups discussions revealed that:

a) absenteeism of both teachers and students dropped as road closures were eliminated, with beneficial effects on education quality and outcomes; b) the number of primary schools and satellite classes"2 increased (Table 5.2), as existence of a paved road is, according to community officials, a key planning criterion for the construction of primary schools in rural areas (construction of these facilities is decided and financed by the local governments but funding for their operation is provided by the central government, seepara. 5.9); c) recruitment of teachers became easier, as better roads made them willing to settle in villages nearby or at the school (especially in the isolated satellite classes); d) supply of materials improved, both teaching materials and materials for the maintenance and day-to-day operations of the school (such as potable water, often transported by the teacher by hand-cart).

12 Under Morocco's rural, primary education system, a 'mother school' (ecole mere) is first built in a rural area. Then, to improve coverage and accessibility, an initial satellite classroom is built for the first primary grade, located separately but under the administrative responsibility of the mother school. Thereafter, new classrooms are added yearly to the satellite facility as the students move to the next grade level. 32

Table 5.2 Number of Primary Schools for all Villages in the Three Provinces, 1985 and 1995 Project Control 1985 1995 1985 1995

Chefchaoue i 1 5 1 2 Settat 1 7 1 1 Marrakech 1 1.5 0 0

5.9 The findings on the impact of improved, paved roads on education in this report corroborate the findings of a study3 based on a comprehensive survey carried out in 1990-1991 of standard of living throughout Morocco (the Morocco Living Standards Survey, MLSS), notably that: "The presence of a paved road increases school participation for both boys and girls. When there is no paved road in a community, the school attendance rate is 21 percent for rural girls and 58 percent for rural boys. In contrast, the rate increases to 48 percent for rural girls and 76 percent for rural boys if there is a paved road in the community".

5.10 Education Infrastructure. Focus groups and key informant interviews provided information on changes in education infrastructure before-and after-project in the various regions (see also Table 5.2). In Chefchaouen, the number of schools and satellite classrooms increased considerably. Schools increased fourfold in the three villages surveyed and in one of these villages, Tatouft, a coed school was created enabling girls to attend middle school.

5.11 In Settat, before the road improvement there was only one satellite school with five classrooms in Larbaa (which became the chef-lieu of the commune and the biggest market in the area). Since the construction of the road, Larbaa has now three such schools with five classes. In another village where there were no schools before, one school with two classes was created enabling young children to go to school. Most teachers commute from the small town () to the schools. The enrollment of girls still lags behind because of the traditional reluctance to send girls to school, and the lack of a school is another constraint to girls' education in the area.

5.12 Less information was available in Marrakech where the number of schools increased from 4 before 1983 to 10 in 1995. For example, in the province of Chichaoua, three municipalities along project roads did not have a primary school before road improvement; they all have one school now.

13 Khandker, Lavy and Filmer, 1994. Schooling and Cognitve Achievements of Children in Morocco. Can the GovernmentImprove Outcomes? World Bank Discussion Paper No. 264. The World Bank, Washington, D.C. 33

Health

Key Features of the Sector

5.13 The limited health infrastructure and health services in the Moroccan countryside are reflected in poor health indicators. Only 6.5 percent of rural inhabitants have medical services in their douars, and only 27 percent have access to fixed primary care facilities. There are 100 hospitals in Morocco and all are located in urban settings. There are about 260 health centers in urban areas versus 230 in rural areas, and some 150 dispensaries in urban areas versus 1,040 in rural areas.'4 A rural health center serves, on average, 52,000 inhabitants and in some places up to 180,000.

5.14 The gap in health conditions between rural and urban settings is substantial. For example, life expectancy for rural inhabitants is 66, where as that for urban inhabitants is 72. A rural inhabitant is almost 25 km from facilities where a health-related consultation can be sought, compared with less than one third that distance for an urban inhabitant. Probably as a result, fewer than 40 percent of rural inhabitants seek treatment for illness (compared with over 60 percent of urban inhabitants), and lose some two days (about 25 percent) more of potential income-generating activity each year due to illness.

5.15 The fertility rate in rural areas is twice that of urban areas (5.7 and 2.8, respectively), and family planning services are beyond the reach of most women in the countryside. The maternal mortality rate is very high: 362 maternal deaths per 100,000 live births in rural areas against 284 in urban areas; the average in developed countries is less than 10 per 100,000 live births. The infant mortality rate is estimated at 66 deaths per 1,000 live births," which compares to less than 20 in developed countries.

5.16 The public sector is the larger supplier of health care, and the Central Government funding amounts to about one percent of GDP each year.

Main Findings

5.17 Visits to Health Facilities. As shown in Table 5.3, visits to health facilities in all three regions combined increased. The largest gains were for visits to hospitals and health centers, which doubled from 1985 to 1995 once the roads were improved. Improvements also happened in the control zones, but reached lower levels than in the project zones. Nationally, admissions to hospitals increased by about 25 percent during the period (Annex B).

5.18 Health Facilities' Staffing and Supplies. Focus groups revealed improvements in health facilities and services before and after the project:

a) In the Chefchaouen area, the health centers existing at the three main population centers (Tatoufet, Souk El Kolla and Tanakoub) had neither permanent professional personnel nor equipment and supplies. Today, each center is staffed by a medical doctor and two

14 Health centers (centre' de sante) are defined as health establishments with at least one physician. Health centers offer mainly out- patient services. Dispensaries (dispensaires) are defined as establishments offering out-patient health services and staffed by nurses and other para-medical staff (but not physician). In practice, many health centers have not succeeded in recruit a doctor, and there is prtctically no difference with a dispensaire. '5World Bank, 1995. World Development Report 1995. Workers in an Integrating World. World Bank, Washington, D.C. 34

or three nurses. In addition,each of the centers now has an ambulanceto transport patients to the hospitalsin Chefchaouenor Ksar El Kebir. One of the main benefits has been that pregnant womenready for deliveryare now transportedto the hospital by an ambulance,rather than by mule as the past. An additionaleffect has been the establishmentof two pharmaciesalong the improvedroad.

b) In the Settat area, similar impactshave been registered. At the health center in Larbaa, services are now providedby 2 nurses and 3 weekly visits by a doctor,whereas no such personnel existed before the road was improved.

c) In the Marrakecharea, the health servicesalong the road areas have improvedlittle. It appears that the improvedaccess to existingfacilities (at Amizmiz and at Assif Al Mal) provided by the improvedroads and new, or more frequent share-ridetaxis and bus services resulted in an improvementin the population'saccess to health services,and that health authoritieshave not deemedthe constructionof more facilitiesto be a priority.

Table 5.3 Changes in Frequencyand Distancesto Health Services in Project and Control Zones, Before and After Project, for all Three RegionsAggregated Infirmary Health Centers Hospitals Distance Frequency Distance Frequency Distance Frequency (km2) of visits (kn) of visits (kn) of visits (days/year) (days/year) (days/year) Project Zone Before 2.3 4.3 20.7 2.8 60 1 After 2.5 6 8.7 5.6 60 2.4 -_ontrolZone Before 5 3.5 9.6 3.7 40 0.8 After 1.7 4.6 9.8 5 40 1.6 Note:In the 1980s,health centers were conceived as primarycare facilities serviced by a full-timedoctor, whileinfirmaries were supposed to be staffedby nursesonly. In practice,the intendedlevel of staffing wasoften not reached.

5.19 Health Service Strategy and Plans. Information supplied by focus groups and health authoritiesalso revealedthat:

a) as the road access was improvedand, sometimes,combined with the constructionof new facilities, health service authoritieswere able to change the health service strategy for the areas servicedfrom the road, by, for example,moving from occasionalvisits to the areas by medical personnelto permanentavailability of a facility near the users.'6 The existenceof permanenthealth professionals,as opposedto occasionalvisits by doctors

16 Morocco's rural health service strategy comprises three main altematives: (i) a permanent fixed, primarycare facility, which is easily accessible to the local population, called 'strategie fixe"; (ii) or, occasional visits to the areas by a nurse, normally on a motorized bicycle ('velomoteur'), called "strategic itinerante"; (iii) or, occasional visits to the area by a mobile vehicle, called "strategic mobile". 35

and nurses, allowed to implement health programs like pre-natal care and immunization of children; b) implementation of family planning national programs, mainly the distribution of contraceptives and provision of medical advice, are no longer hampered by transport bottlenecks thanks to the roads' improved condition; c) in the areas enjoying improved roads, it has been possible to implement a "medicalization" program whose aim is that every health center is serviced by at least one medical doctor. Local authorities reported that it was , and still is, practically impossible to recruit doctors in areas which do not have good road access; d) due to the better access, it has been made easier to recruit as well as to retain health service personnel; and e) the supply of medicines has increased considerably.

5.20 Comparison of Impacts between the Project and the Control Zones. As noted, there has been improvement in access to health infrastructure and services in the villages around the project zones. However, the fact that the control zones also benefited from such improvements makes it difficult to establish causal links between project investment and reported impacts. Three possible reasons may explain these findings:

a) improvements, percentage-wise, were about the same in the project and the control zones, but starting from and reaching lower levels in the control zones, possibly reflecting that it is easier to achieve gains at the lower end of the spectrum; b) the "recall" factor of the retrospective survey (asking respondents to remember events ten years earlier) has led respondents to recall events/activities in a more favorable light than they actually were, especially as they improved over time. Thus, respondents from the project zones might have had the same health facilities as those from the control zones in 1985, but the memory of hardship experienced earlier had faded over time; and c) the rehabilitation of the roads had no impact on the number of trips to and from health facilities, and the improvements noted are all due to exogenous factors such as the implementation of health programs in the areas. Unfortunately, data at the village, douar or neighborhood levels was not available to verify this. 36

Impact on Women

Box 5.1 Living Conditions of Rural Women

"The rural sector's limited transportation infrastructure limits women's opportunities. Many rural villages in Morocco are isolated from market centers by long distances. The lack of roads makes travel too difficult, dangerous, or time-consuming for women to profitably engage in economic activity involving travel. Consequently, women often lack access to markets and are unable to participate in wage labor or trade. Transportation constraints stemming from the limitations of the road system inhibit women's access to health care. Travel time to the nearest health center is greater than 60 minutes for nearly half the population in rural areas. This particularly affects women who with their reproductive role are more likely to be in need of health services and have less mobility than men."'7

Rationale for the Impact

5.21 The improvement of roads and the resulting better access to services and facilities was expected to benefit women more than men, for a number of reasons. First, women in rural Morocco enjoy substantially less mobility than men. On the education front, the effects of improved access to schools would be expected to be higher for girls in view of their very low enrollment rates compared to those of boys. Similarly, better access to health services could be expected to enable women to get gender-specific health services, such as maternal and infant care. Women would also be expected to derive benefits concerning personal mobility and rural- urban contacts because of their daily time-consuming activities. Female workload is heavy in the absence of easy access to clean water and fuelwood or other heating and cooking sources; thus, improvements in local roads facilitating access to water supply and to sources of energy (both existing sources and those created following the road improvements) could be expected to significantly reduce women's workload.

Gender Indicators: A National Perspective'

5.22 Education Nationally, about 40 percent of all children aged 7-13 are enrolled in primary education. Of those enrolled, 41 percent are girls. In rural areas, the percentage of girls goes down to 32 percent; by 6th grade, the percentage further drops to 25.

5.23 The probability of an average rural girl having access to basic education and surviving through the fifth grade was estimated in 1991 at less than 20 percent, while the corresponding probability for an urban boy was close to 90 percent. These disparities are explained by parental attitudes to schooling and opportunity costs of sending children to school as well as by unavailability of schooling facilities, along with factors such as the lack of provision of meals. Among the poorest expenditure group surveyed in the 1990-91 MILSS,poverty (that is, liability to forgo children's labor or earnings) was cited in about two-thirds of the cases as the reason for not having children attend school.

17WorldBank, 1995. Kingdom of Morocco. Enhancing the Participation of Women in Development. Middle East and North Africa Region, World Bank, Washington, D.C. (paragraph 3.5). "8 lbid, footnote 17. 37

5.24 Health Several indicators illustrate women's lower health conditions in rural settings. While women in urban areas may expect to live 74 years, such number drops to 67 in rural areas. At the same time, the fertility rate in rural areas is twice that in urban areas (5.7 and 2.8, respectively). Family planning services are beyond the reach of most women in the countryside, and cultural factors further discourage their use. Maternal mortality rates are high in rural areas: they reach 362 maternal deaths per 100,000 live births, which compares with 284 in Morocco's urban areas (and less than 10 per 100,000 live births in developed countries). Infant malnutrition is about twice as prevalent in rural as in urban areas.

5.25 Women consistently emphasize the need to have basic health facilities in the village or to have ambulances for easy transport, especially in maternity related emergencies.'9 In rural settings, most deliveries take place at home (72 percent) with the assistance of traditional birth attendants (kabla) in 48 percent of cases and with relatives or friends in 20 percent of cases. Only about thirty percent of the births are attended by a health professional; of these, two thirds of the time by a doctor and one third by a midwife or nurse. Women's lack of information about the health centers further reduces their access to such services.

Main Findings

5.26 Education. There was a dramatic improvement in the project zones in the enrollment of girls in primary school: it went up from 17 percent in 1985 to 54 percent in 1995 (see Table 5.1). This happened consistently across the three zones under study, and all three zones registered enrollment levels higher than the national average for rural areas, especially Settat, where girls' enrollment reached 67 percent. One reason cited for the substantial improvement in girls' enrollment is that several new primary schools and satellite classrooms opened during the period. By comparison, enrollment of girls in primary education in rural areas throughtout Morocco iemained practically constant during the study period (Annex B).

5.27 Despite the large gains in the enrollment of girls in primary education in 1995, in the three zones combined, girls' enrollment levels were still substantially below boys', 54 percent versus 81 percent. It appears that the traditional reluctance of parents to send girls to school- because they contribute more than boys to household chores such as getting water and caring for siblings, and because of perceived dangers of rape when there are long distances and difficult trails to school-is still the primary cause for their lagging behind boys. These findings on girls' education are consistent with those in the study completed by Khandher, Lavy and Filmer (1994).2`

5.28 Health. In the project zones, there was a clear gain in the frequency of visits by women to health services: their visits to a hospital more than doubled (2.4 per year in 1995, compared to 1.10 in 1985) and their visits to the primary care centers also increased (3.1 in 1995 compared to 2.3 in 1985). Men made relatively similar gains percentage-wise, but the frequency of their

19 Ibid, footnote 17.

20 This report found, through regression analysis based on LMS data, that: (a) the presence of a paved road increases the girls' probability of ever attending a school by 40 percent, compared with less than 10 percent for boys; (b) school attendance for rural children increases from 58 percent for boys and 21 percent for girls when there is no paved road in the community, to 76 percent for boys and 48 percent for girls when there is a paved road. Thus, girls' attendance more than doubles, while boys' attendance, which is higher initially, increase by about 15 percent as a result of the existence of a paved road; (c) if a road is blocked during the year (an average of I month per year in the study), there is a 30 percent drop in girls attendance, but only 20 percent in boys attendance. 38 visits to hospitals was much lower than women (0.8 per year in 1995); however, it was about the same for visits to primary care centers.

5.29 Focus groups reported that women gained substantially from health programs in maternal and infant care and family planning, but no statistics specific to the project zones were availab!e.

5.30 Another health impact is the diversification of the diet. Focus groups in Chefchaouen reported that while in the past they rarely ate fish, they do so now at least once a month; they credit this change to the paving of the project road, which has enabled refrigerated trucks to have access to the souks serving their areas. Similarly, it was reported that consumption of fresh vegetables and fruits increased.

5.31 Cooking and Heating. A major gain in women' welfare stemming from the better quality roads was the introduction of butane for cooking and heating. For example, focus groups in Chef-haouen reported that before the improvement of the road, women had to spend an average of 2 hours daily in order to get and carry fuelwood. Butane gas, used extensively in urban areas, did not reach the rural areas due to the high transport and distribution costs. A bottle of butane used to cost 20 Dh; following improvement of the road, the price dropped considerably, to as low as II Dh. which is affordable. A similar phenomenon was reported in the other regions.

Box 5.2 Fuelwood Collection by Women

"Fuelwood coliection for cooking and heating is sometimes 5 km away and most women must go twice a day, collecting up to ten kilos of wood per basket. In some larger households, women have to repeat the voyage four times a day. Quite naturally, the women linked backache, pain in the joints and arthritis to the heavy loads of wood they have on their shoulder." 21

5.32 Work Opportunities. There were area-specific impacts on women which, in some cases, resulted in opposite consequences. For example, focus groups in Settat reported that in the past women were required to work on the fields helping with the wheat crop. Today, as a result of mechanization and the creation of agricultural cooperatives made possible by the improved road, most women can stay home. Since, coincidentally, taxi services have greatly expanded, women now use time for more visits to their relatives and to nearby villages and cities. A rather different case happened in some localities in the Tensift area. The increase in milk cooperatives that resulted from the improved road required more labor to take care of the cows. Thus, municipal programs were instituted encouraging women to work on this activity and a specific incentive was created where women got a share of the revenues stemming from milk sales; the result has been an increase in earning-generating work for women. In view of the low participation of women in Morocco's formal (monetized) economy, this is an important impact.

Other Impacts

5.33 Access to Water Supply. As shown in Table 3.2, in all three areas served by the improved project roads, the average time to access water supply facilities showed a small

21 Ibid, footnote 17. 39

decrease, from an average for all regions of 16 minutes in 1985 to 13 minutes in 1995, while no gains were achieved in the control zones. 22

5.34 Mobility and Rural-Urban Links. The expanded rural taxi23 and bus services over the study period, led to a major increas_,in moHlity and rural-urban links. For example, across all three regions served by the project roads, the number of visits by relatives in the urban areas to their families on the farms increased, on average, from 2 visits per year in 1985, to some 15 per year ten years later; this compares to less than two in the control zones in 1985, and to just 3 in 1995. At the same time, visits by members of the farm households to the nearby cities increased substantially in Settat and in Marrakech, from, on average, 12 in 1985 to over 60 in 1995, while the equivalent numbers in the controls zones changed from 8 to 25.

22 Accessto water, as well as to other services,is reportedby the WorldBank's 1996Participation Sourcebook. To be a high priority by Moroccanwomen: "...we learnedthat whereas menwanted to build bridgesand roads,women' top prioritieswere clinics, ambulances,electricity, running water and collectiveovens"

23 Rural taxis in Moroccoare share-ridetype, that is, they are not limitedto servingone client but can take passengersup to the capacityof the vehicle.

41

6. Impact on the Environment

Changes in transport conditions and in the agricultural economy had both negative and positive impacts on the environment although, overall, no environmentally sensitive areas were at risk by the road projects, which did not involve new construction. Negative impacts were those resultingfrom the increased traffic and economic activity, especially air and noise pollution and road accidents, and the increased use offertilizers and other chemicals which in all likelyhood contaminated the water table. Positive impacts resulted inpartfrom the transformation of the agricultural economy, notably curtailment of extensive goat and sheep that damages the soil cover and increased tree plantations, andfrom broader use of butane substitutingforfuelwood, whose demand is larger than the size of Morocco's sustainable forests.

Expected Impacts

6.1 The environmental impact of paving a rural road was expected to have both negative and positive dimensions. On the negative side, there would be the effects of increased economic activity in the rural economy and resulting higher levels of motorized traffic, with its known cor,sequences of increased use of agricultural chemicals (fertilizers, pesticides) and vehicular pollution. On the positive side, transformation of the agricultural output and the introduction of better agronomic practices could lead to reduced soil erosion and better care of the forests. The road works themselves, invoiving no extension of the existing right of way, could be expected to have only a minor impact on the environment, and limited in time to the period of the execution of the works.24

Main Findings

6.2 The data made available through the study did not permit to quantify the impacts of the rural roads improvements on the environment. However, information from the focus groups interviews, the analysis of the farm level questionnaires and the changes in the regional agricultural economy provide a basis to identify types of impacts, both negative and positive, on the environment. Such impacts are divided here between those resulting directly from the road works and the higher traffic levels they induced or allowed, and those resulting from changes in the agricultural economy and related economic activity.

Types of Impacts

6.3 Impacts Directly Resulting from Road Improvements and Traffic. The main negative impacts are the increase in the levels of:

a) air pollution stemming from the increased fuel fumes resulting from the higher traffic;

b) road noise as a result of the higher vehicular activity; and

c) soil erosion caused by higher traffic levels and cars driving on road shoulders, leading to occasional flooding.

24 The roads works on the four roads analyzedin this reportwould fall underthe Bank's EnvironmentalCategory B. 42

6.4 However, at least in the case of the first two impacts, while the overall levels increased, the levels per unit of traffic decreased as a result of the better road surfaces, which are more fuel efficient and less noisy.

6.5 While there is no statistical evidence, it is likely that the higher traffic levels and especially the higher speeds made possible by thc paved roads, led to an increase in the number of road accidents, and related fatalities, injuries and material damages.

6.6 On the positive side, the asphalt roads substantially reduced, relative to the earth and gravel roads that they substituted, dust pollution; farm workers traveling to the souks or to social services in the villages and children attending school were the main beneficiaries of such reduction in dust.

6.7 Impacts from Changes in the Agricultural Economy and Increased Economic Activity. The main negative impacts were:

a) increased use of agricultural chemicals (mainly fertilizer and pesticides), resulting from the intensification of vegetable gardening, higher-yield cattle, tree cropping and sugar beet production. Measures of such contaminants on the water table, or in locally marketed foodstuffs, are not available for the zones of influence of the project roads;

b) increased liquid and solid waste resulting from livestock slaughtering in souks.

6.8 The main positive impacts, identified by the surveys and the focus groups, were:

a) the curtailment of extensive goat and sheep herding;

b) an increase in the number of tree plantations; and

c) increased reliance on bottled gas for heating and cooking fuel (as access to these supplies is made cheaper and easier by the better roads), substituting for fuelwood.

6.9 In some areas (namely, in the Rif) these impacts combined resulted in decreased run-off from steep slopes, increased preservation of upper watersheds and increased reforestation of the watershed which had been severely eroded due to the uncontrolled scavenging for fuelwood and the decrease in soil cover due to goat herding.

6.10 The shift to high-value agriculture, although it increased the use of polluting chemicals, reduced extensive planting and range practices generally harmful to the environment.

6.11 The positive impacts noted above will benefit the forests. Such benefits are important. It is estimated that 85% of Morocco's need for fuelwood is supplied by forests and non- commercial agricultural vegetation. This level of demand for fuetwood in the country as a whole is estimated at 10 million cubic meters of wood, well above the sustainable forest yield of 3 million cubic meters. Also, some 40 percent of grazing grounds are in forest zones and overgrazing undermines reforestation.25

25 WorldBank. 1994, op. cit. 43

6.12 Overall, it is not possibleto provide a balanceof the various environmental impacts caused by the improvementof the rural roads, since differenttypes of impacts took place whose effects are not directly comparableand quantificationof the individualimpacts is not possible. However, the perceptionfrom the focus groups and from the surveys carried out during the study is that the positive impactsare of similar magnitudeas those impactsthat are harmful to the environment.

45

7. Economic Analysis

The improvements in the agricultural economy and access to social services that the better roads brought about translated into increases in the roads'traffic levels at rates substantially higher than on unpaved rural roads. The study quantified the economic benefits based on savings in vehicle operating costs compared to the original, unpaved roads, and on the economic gains resulting from the avoidance of road closures (due to rains) since the paved roads are open to traffic year-round. Social impacts, although real, could not be isolatedfor attribution to the road investments and were not quantified. The analysis showed that the economic returns as well as the timing of the improvements for all project roads were satisfactory (economic rates of return ranging between 16 and 30 percent), and that even under a worse case scenario the returns would have remained acceptable. The benefits from the investments are expected to accrue to farmers in the form of expanded use of commercialfreight services substituting for non-motorized carriage of loads, including heavier, more efficient trucks. The rural population in the roads' areas is expected to benefit mainly by the availability at affordable prices of frequent services of share-ride taxis. The satisfactory returns do not necessarily mean that paving was the optimal economic decision, comrpared,for example, to rehabilitated gravel roads maintained at good standards.

Economic Analysis at Project Appraisal

7.1 The feasibility study on which the Bank's appraisal report was based assessed benefits of rural roads improvements mainly in terms of increased agricultural volumes stemming from: i) an increase of between 5 and 20 percent in planted areas, and ii) an increase in unit yields of up to 70 percent relative to those existing at the time the project was prepared. The study also quantified savings in vehicle operating costs to the existing traffic.

7.2 Based on this analysis, the ex-ante rates of return for the road projects were as follows:

Table 7.1 Ex-ante Economic Rates of Return, as Presented in the Staff Appraisal Report Route Economic Rate of Return, percent

603 2 1 1427 37 6307/08 39

Findings of Importance to Economic Analysis

7.3 Findings in this study point to the difficulty in forecasting changes in agricultural production levels, value added, and related traffic. The transformation of the agricultural economy found in the three regions went beyond improvements in yields, or increases in planted areas (a common assumption in rural road projects).26 As noted in Chapter 4, such

26 A common model is to assume a market for agricultural products located at a fixed distance for which the ruralarea being analyzed is a marginal supplier and does not influence the market price. Thus, reduced transportcosts stemming from better roads 46

transformation followed the classical Van Thunen model, where improvement in roads will change the composition of agricultural output from low value cereals to high value but highly perishable vegetables.

7.4 At the same time, a transformation of the agricultural economy to higher value vegetables may lead, at least in the short term, to either an increase or a decrease in traffic, depending on precisely how and by how much the agricultural economy was transformed. In the medium term, it is likely that the better agricultural economy will raise disposable income, and will, therefore, lead to an increase in passenger travel.

Project Benefits

7.5. Economic benefits from the road improvements can be divided into two categories: (i) those related to the vehicles using the road had the road not been improved, designated as the road's normal traffic and (ii) those related to the developmental effects of the road improvement. Benefits to normal traffic are measured by the savings in the vehicle operating costs. Benefits stemming from the developmental effect of the road can be measured either by the reduction in transport costs of the generated traffic, or by the value of expanded production. In theory both methods should yield the same results.2" As this retrospective analysis can take advantage that actual traffic data exists for one year before the improvements, and one year after the improvements, this report follows the transport cost savings approach, comparing costs "with" and "without" project. The alternative approach, quantification of production benefits, although it allows to better relate road improvements to the developmental changes taking place, requires making assumptions which, overall, greatly increase the potential for faulty results. Especially difficult and prone to error is the assessment of a road's area of influence from which production is to be quantified, and the estimation of the prices of inputs and outputs, notably as they are subject to large variations due to droughts and domestic and external market conditions.

7.6 The roads' impact on the social sectors, described in earlier chapters, cannot be directly measured for the purposes of economic analysis, and they are normally considered as intangible benefits; nor can the analysis quantify benefit distribution among different beneficiaries. Benefit distribution is discussed later in this chapter. A recommendation regarding consideration of social impacts in the screening and evaluation of rural roads is in Chapter 9.

Traffic

7.7 Changes in traffic levels between 1982 and 1995 are discussed in Chapter 3. The difference between current traffic levels and those in the early 1980s can be attributed to : i) normal growth in traffic, that is, the annual growth rate at which traffic grows on unpaved rural roads; ii) traffic diverted from other roads, due to shorter distance and/or lower vehicle operating costs; and iii) new traffic induced by the road in its direct zone of influence. The relative importance of each factor is directly related to the economic role of the road, both in the area it serves and as a long-distance route. Briefly explained, the roads' economic roles are:

cause farm-gate prices to increase, and allows to extend the distance at which planting for selling in the existing market is profitable. See A. Walters, 1968. The Economics of Road User Charges. World Bank Occasional Page # 5.

27 [ans Adler, 1987. Economic Appraisal of Transport Projects. A Manual with Case Studies. EDI Series in Economic Development. 47

Road 603: It crosses an agricultural region that produces beets for a sugar factory (SUNABEL); is also provides access to an otherwise isolated population.2" At the same time, this road was considered to have a substantial potential as a transit route for long-distance traffic.29 As noted in Chapter 3, this actually happened.

Road 1427: It connects the provinces of Settat and Khourigba and traverses an area devoted to cattle raising and agriculture.30 This road was also considered to have potential as a transit route. Since the traffic growth on this road was similar to the average traffic growth rates in Morocco, the transit potential probably did not happen, or it did to a limited extent only.

Roads 6308/6307. These two roads were more purely rural roads in that they provide access to an isolated population but also are in an area with tourist sites, which attracts travelers going to .

7.8 Quantification of Project Benefits. The savings in vehicle operating costs between the unimproved and the paved roads have been based on indicative figures given in the appraisal report for the road project approved in 1995.3' The quantification of benefits applies such vehicle operating cost savings to the normal traffic, and only half this amount to the induced traffic (a conservative assumption that takes all induced traffic as being developmental, generated traffic). At the same time, since 1982 traffic data did not contain details of non- motorized traffic (NMT), e.g. pedestrians, bicycles, carts. The present analysis omits the benefits that would accrue to the non-motorized resulting from (i) the gains from the better road quality to NMT traffic remaining on the road after its improvement, and (ii) shifting of NMT to motorized vehicles. However, the analysis does quantify separately the benefits derived from avoiding road closures. The analysis also takes into account that in the roads with pavements 4 meters or less, the costs of operating vehicles increases after a traffic threshold, due to vehicles partially riding on the road's earth shoulders.

7.9 There are no records in Morocco of the growth rate of traffic on the rural roads from which to asses how much traffic may have been induced by the investments, as opposed to the "notmal" traffic level that would have been reached in 1995 if historical trends of traffic growth had continued. The assumption adopted here is that traffic growth on rural roads is lower than on the main highways; the hypothesis is that, had the rural roads not been improved, traffic would have grown at a rate of 2.5 percent per year.

7 10 Key data for the economic evaluation is shown in Table 7.2. More detailed information is provided in Annex F.

2t MTDS report, Volume 11,Annex 2, page 2. Report dated November 1, 1995.

29 Maroc-Setec, 1982, op. cit. (footnote 1). The reference to transit traffic is given on page 4.11.

30 World Bank, 1993, op. cit. (footnote 3).

31 Kingdom of Morocco, Secondary, Tertiary and Rural Roads Project. Staff Appraisal Report, World Bank, June 3, 1994. The SAR provides average vehicle operating costs for various categories of rural roads, including paved and unpaved. 48

Table 7.2 Key Data for EconomicEvaluation

Traffic (veh./day) Road Length Const. Cost 1982 Open Year 199S Open Year (km) (mDh/95) Estimate Actual 603 1987 35 28.9 40 450 640 1427 1988 30 17.3 150 202 275 6307 1991 10 5.8 54 161 192 6308 1991 14 8.1 54 161 192

7.11 As shown in Table 7.3 below,the economicrate of return for all the investmentsis satisfactory,and ranges from 16 percentfor road 6308 to 30 percent for road 1427. A sensitivity analysis considering lowertraffic growth rates, lower savings in vehicle operatingcosts (for example, if the pavementsare not well maintained),and higher costs of routine maintenance,as well as a worst case scenario combiningIcwer traffic with lowervehicle operatingcosts savings shows that the economicreturn for all roads would still be acceptable.

7.12 The table also shows that the first year return (FYR, benefitsthe year of road opening relative to investmentcosts) is satisfactoryfor all roads. A satisfactoryFYR indicatesthat the investmentwas not premature.

Table 7.3 EconomicRate of Return for Project Roads Project Roads 603 1427 6307 6308

Best Estimate (ERR) 21% 30% 16% 16% First Year Return 24% 26% 16% 16%

ERR SensitivityAnalysis (A) Total traffic annual growth 1) Higher (5.5%) 22% 30% 17% 17% 2) Lower (3.5%) 21% 29% 16% 16% B) Overestimationof VOC savings 21% 26% 14% 14% (20% lower) C) Higher routine maintenancecost 20% 28% 15% 15% (3%) D) Worse case scenario 20% 25% 13% 14% (A. 1 and B combined)

7.13 The main reasons for the satisfactoryreturns appear to be: (a) the very poor condition of the original roads, which caused a major reductionin vehicle operatingcosts when the roads were improved and paved, (b) the fact that the roads (which is partially wider) were built to a narrow paved width of 4 meters only, with consequentsavings in constructioncosts, and (c) that substantialtraffic was generated,as reflected by the differenceat opening year between the 49 normal traffic (without the improvement) and the estimate of the actual traffic that year based on the 1995 survey.

7.14 While some benefits appear to be dependent on paved roads the satisfactory returns to the rura! roads investment does not ensure that paving was the optimal economic solution. It is possible that, especially the roads with lower traffic such as roads 6307/6308, rehabilitating the roads to a good gravel surface and maintaining them well could have yielded even higher returns.

7.15 Economic Rate of Returns Using HDMModel. MTDS consultants carried out an estimate of the irvestment returns using the Bank's Highway Design Model (HDM) and adopting standard assumptions used by the MPW. The HDM/MPW-based evaluation gave ERRs substantially higher than those presented in Task 7.3, suggesting that some of the assumptions may be overly optimistic. Details of this analysis is in Annex F.

7.16 Distribution of Benefits. The direct beneficiaries of the road improvements are road users: motorized passenger and freight vehicles and non-motorized vehicles comprising bicycles, carts and pedestrians. Since vehicle ownership is still low in Morocco, most motorized vehicles operating on the roads do so for commercial purposes. Farmers benefit from the improved roads in two ways: (i) by expanding use of commercial freight services, substituting for the slower, limited-distance and more expensive non-motorized transport, which on unpaved roads in Morocco's rural areas carry some 50 percent of all goods, and (ii) by being able to use the services of truckers operating larger freight vehicles, which offer lower transport rates. An indirect measure of the transport benefits to farmers is given by the lower increase (than in the control zones) in transport household expenditures, discussed in Chapter 3. The drastic drop in the retail price of butane cited in Chapter 5 is another indicator of these benefits. While these benefits are significant, it is likely that farmers could have derived even greater advantages had Morocco's road freight transport market been less regulated and more competitive. This comment applies especially to the large trucks, over eight gross tons, that are required to do business through the state-owned Office Nationale des Transports (ONT), which operates as a freight-forwarding monopoly.

7.17 Regarding passenger travel, benefits accrued to the rural population mainly in the way of substantially increased share-taxi services which have taken an increasing share of the rural transport market, at the expense of interurban bus services. There are practically no entry controls to the taxi market, and tariffs are competitive. The non-motorized transport has benefited from the substantially better surface quality offered by the paved roads.

Design of Rural Roads

7.18 Two out of the three roads in this study were improved to paved surfaces of 4 meters or less, while the third has a 5.5 meter-wide section. Current thinking in the MPW regarding the large rural program under construction, appears to be not to consider pavements less than 6- meter wide. This would be in conflict with the results in the present study, which show that 4- meter pavements were cost effective for the traffic levels attained on roads where they were tried.

51

8. Sustainabilityof the Impacts

Historic trends showing steady traffic growth over long periods on Morocco's paved roads, even during drought years when agricultural production fell, is a key indicator suggesting the stream of benefits is likely to be sustainable. The immediate issue is route 603 which, seven years after completion of the works and in part because traffic reached an unexpectedly high level, currently is in poor condition. While resurfacing of this road is included under the MPW's maintenance program for 1994-1998, inadequate budgetary allocations for road maintenance raise uncertainty about the level of benefits on this road over the long-term. Agricultural benefits are based on sound economic decisions and are likely to be sustained, with the possible exception of the sugar beet planting in the North (a small component of traffic on route 603) which is uneconomic and may lose market depending on the outcome of the privatization program for the state-owned sugarfactories. The sustainability of the social service impacts appear as likely in view of the high value assigned to them by the direct beneficiaries and because of Government policies aimed to improve social services, and the increasedfunding it is allocating to these activities.

The Roads

8.1 Sustainability of the benefits and impacts of the investment in rural roads will depend on the traffic levels and on the country's ability to maintain the roads to the same standard as they were built. Until now, traffic has grown at high rates both country-wide and on the project's roads. At the same time, traffic is relatively immune to the cyclical droughts that affect Morocco. The biggest threat to keeping or increasing traffic levels is the adequacy of road maintenance.

8.2 The MPW surveys the condition of the main roads every 2-3 years. In the 1994 survey, that is, seven years after completion of Road 603, six years for Road 1427 and four years for Roads 6307 and 6308, the condition of the roads was the following:

Table 8.1 Condition of Project Roads, Based on 1994 MPW Survey Area/Road Surfare Quality (% of road length) Excellent Good Poor Very Poor

North/Rt 603 8 5 23 64

Center/Rt 1427 0 100 0 0

South/Rts 6307/08 100 0 0 0

8.3 In part, the poor condition of route 603 may be explained by the explosive growth of traffic on this route, which is some two to three times the level of traffic on the other roads considered in this study; in addition, this traffic includes some 20-25 percent of trucks, which contribute further to the deterioration of its pavement; also, an asphalt surface seven years old normally is approaching its economic life. 52

8.4 More generally, there are three main issues related to maintenance and the sustainability of the roads benefits: the amount of budgetary resources for maintenance, the criteria for allocation of funding, and the maintenance administration. Inadequate budgetary resources for road maintenance has been a perennial problem in Morocco. The completion report of the fourth project, and, more strongly, the audit prepared by OED, noted this situation. More recently, the appraisal report of the highway project approved in 1995 reported that, as of 1993, there was a country-wide backlog of hignway maintenance works estimated at Dh 1.74 billion. An increase in the overall funding for roads appears unlikely due to a tight budget, and the Government is addressing this issue by shifting, to some extent, more resources for maintenance, relative to new road construction. Thefunding allocation issue is whether the limited funds are correctly allocated to the roads and type of maintenance with the highest pay-offs. The allocation of road funding is done at two levels: first, the DRCR allocated a tentative budget to each province based on four criteria: (a) length of paved roads in bad and very bad condition; (b) traffic on paved roads; (c) total length of roads and (d) length of mountainous roads; second, individual roads maintenance priorities are assessed through a pavement management system. This allocation system is a considerable improvement over past practices, but the EVAL needs improvement to ensure correct counting of the benefits and identification of maintenance priorities. Administration of maintenance, including programming, contracts and supervision of road works is carried out by 40 provincial offices reporting to seven regional offices. While this organization appears to have built-in incentives to optimize public expenditures for roads,32 it could be improved by including road users in the road management process.

8.5 The maintenance of all four roads analyzed in this study fall under the responsibility of the MPW. The MPW's Maintenance Action Plan for the period 1994-1998 includes the resurfacing of route 603. The sustainability of the benefits of this road hinges on the implementation of this plan.

The Socioeconomic Impacts

8.6 Agriculture. The improved project roads, to the extent they are kept well maintained, will be a key factor in ensuring the sustainability of the transformation and modernization of agricultural production that they helped bring about. The agricultural impacts, already in place for some years, are likely to be sustainable as the domestic market for produce is broadening thanks to improved incomes. Agricultural gains are, however, dependent on government trade and fiscal policies and on the opportunities for selling Moroccan fruits and vegetables in the European markets. The development of sugar beet planting in the Chefchaouen area along Route 603 is the agricultural impact most at risk. The sugar processing factories, which are state- owned, lose money, and are being considered for privatization. It is not clear whether, at a sufficiently low purchase price, such factories may become attractive for private sector investment and continued operation.3

8.7 The success achieved by the extension services in helping introduce new agricultural techniques is important for the sustainability of the agricultural benefits particularly as such

32 This analysis relies largely on information provided in the appraisal report of the Secondary, Tertiary and Rura] Roads Project.

33 World Bank, 1994. Royaume du Maroc, Development Agro-Industriel. Constraints et opportunities. World Bank, Washington, D.C. 53 services are likely to be curtailed in view of Government policies to further reduce public expenditures.34

8.8 Social Services. The sustainability of the improvement in access to and use of health services, as well as improvements in women's welfare, appears to be high because beneficiaries realize their direct value. This is the case, notably, of primary health care, whose benefits are tangible and immediate. Sustainability of the supply of services would be enhanced to the extent government decentralization policies are implemented and more budgetary decisions are made by the local governments. The sustainability of education impacts will depend largely on the level of budgetary allocations to primary education and on sustained economic growth; as noted elsewhere in this report, enrollment in primary education fell in the 1980s as a result of economic recession and there is a danger that a similar situation may occur in the future. The World Bank is assisting the social sectors through ongoing and newly-prepared projects, which should help in sustaining the benefits of social programs.35

34Govemment policies on extension services give special emphasis to irrigated areas where, on average, one extension center exists for every 3,000 hectares, whereas the ratio is only one center for every 100,000 hectares in rainfed areas. Of the three project areas, only Chefchaouen is an irrigated one, and is likely to continue to benefit from intensive agricultural extension services. 3' Basic Health and Basic Education Projects. These two projects were expected to be approved by the Bank's Board during fiscal year 1996.

55

9. Conclusionsand Recommendations

Conclusions

9.1 The study, based on a broad database from surveys and focus groups, gathered unique socioeconomic information that allowed to assess the impact of road improvements beyond conventional cost-benefit analysis. Notably, the data permitted to quantify changes in economic and social indicators in the 10-year period covered by the study. Information from control zones where roads were not improved provided a proxy scenario for a "without project situation" and, when combined with feedback from focus groups, helped establish a logical chain of events to assess the relative role of the road improvements in the impacts reported. The main conclusions are summarized below.

9.2 Transport Infrastructure and Services. The improved road conditions led to:

* year-round use of the roads, eliminating frequent road closures during rainy periods; e reduced operating costs of vehicles, and lower rates for freight and passengers, resulting in substantial household transport savings relative to regions that did not benefit from road improvements; * traffic increasing at rates higher than the past trend and comprising a larger proportion of trucks that offer lower transport rates; * a major increase in the supply of rural passenger services, especially share-ride taxis; * a higher degree of ownership of motorized vehicles, cars as well as trucks; and * a substantial reduction in access time by the rural population to markets and social services.

9.3 Agriculture. Changes detected in the agricultural economy in the project areas were:

* the overall level of agricultural activity increased in volume of production, productivity of the land, and monetary values of the output; * the agricultural production mix was transformed, land use shifted from low-value cereals to high-value fruit orchards, and livestock production shifted to pure breed cows; * the use of modern inputs, especially fertilizers, improved as distribution channels became easier and enjoyed lower costs; * as a result of the shift to higher-value products, and the overall increase in yields, the value added per unit of cultivated land increased; * related economic changes in workload, employment on and off-farm, establishment of new shops followed different patterns depending on the region.

9.4 Social sectors. Changes detected in social indicators in the project areas were:

* in education, enrollment of children in primary school more than doubled between 1985 and 1995, and quality of education improved; as new facilities were built it became possible to recruit teachers in the rural areas served by the project, and absenteeism of both from students and teachers dropped; * in health, visits per person to health facilities (hospitals and primary care facilities) nearly doubled over the period and, similarly to education, the quality of health services improved, as new health centers were built (except in the Marrakech area), supply of medicines was facilitated, health authorities launched a program to staff rural health 56

centers with a full-time doctor, and immunization and other health prevention programs became easier to implement; * on women, while the road projects did not originally aim to have specific gender impacts, they did: girls' enrollment in primary education more than trebled; women benefited from matermal and child health care programs; and, equally significant, the introduction of butane at affordable prices (thanks to the existence of paved roads) dramatically reduced women's chores of daily collection of fuelwood for cooking and heating; and * in rural-urban interaction, the improved rural transport services resulted in several-fold increases in social exchanges, both by relatives from the urban areas visiting the farns, and by farm household members visiting nearby cities.

9.5 Regional Distribution of Transport, Agricultural and Social Impacts. Many of the impacts were felt about equally in all three regions considered in the study. In the transport sector, this was the case with the elimination of road closures, increase in the amount of large trucks offering lower rates, and increase in the availability of share-ride taxis, all of which showed a high degree of impact. The Northern region enjoyed the most impact regarding traffic growth, reduction in operating costs and lowering rates of transport. In agriculture, the most consistent gains occurred in the Central region, which scored high in practically all categories of impacts (such as improved farmers' access to markets, reduced prices for inputs and overall economic gains) followed closely by the Southern region. In the social sectors, the Central region also achieved the highest gains, while the other two regions attained somewhat lower gains. A summary of impacts by region is in Annex G.

9.6 Environment. Changes in transport conditions and in the agricultural economy had both negative and positive impacts on the environment, although overall, no environmentally sensitive areas were at risk by the road projects or by their direct and indirect effects:

- negative impacts were those resulting from the increased traffic and economic activity, that is, air and noise vehicular pollution, agricultural fertilizers and other chemicals which in all likelihood contaminated the water table; * positive impacts stemmed mainly from transformation in the agricultural economy: curtailment of extensive goat and sheep herding and increase in tree plantations as well as from increased use of butane for cooking and heating which substituted for fuelwood, whose demand is larger than Morocco's sustainable forests.

9.7 Projects' Economic Benefits. Savings in vehicle operating costs and avoidance of road closures generated by the paving of the roads yielded satisfactory economic rates of return for the investments. The increases in the level of and value added of agricultural outputs, in part triggered by the better road conditions, corroborate this finding.

9.8 Sustainability. The transport, agricultural and social impacts are likely to be sustainable. The only issue is Road 603 in the North which currently, seven years after opening, is in poor condition. The Ministry of Public Works has included this road for repaving under its 1994- 1998 maintenance program but actual implementation, on which the sustainability of the road's benefits hinges, is uncertain due to inadequate funding for maintenance country-wide. 57

9.9 Monitoring of Road Impacts. Despite project objectives to develop and implement an impact monitoring system, this did not happen mainly because government authorities were not persuaded of the cost-effectiveness of the recommended monitoring program.

9.10 Integrating Road Improvements with Other Sectors. The findings of this study indicate that in most cases, the improvement of the roads was a necessary condition for the impact to happen, but other actions were also required, like investments and programs in education and health. While the project focused on roads alone, widespread impacts suggest that, in parallel, local authorities in other sectors did take into account the roads improvement plans in their own plans. It is not clear, however, whether the impact of the roads could have been made greater had more formal mechanisms for coordination of investments been devised from the beginning.

9.11 Comparison with Control Zones. In some cases, improvements in the control zones were equal or better than in the project zones. This indicates that the control zones, although they did not benefit from road improvements, are likely to have benefited from programs in other sectors which did not occur in the project zones.

Recommendations

9.12 The following recommendations are offered:

(a) Establish a practical rural road monitoring system. In the current program to pave 10,000 km of rural roads, an effective monitoring system could be set up with the following characteristics:

(i) select a small number of road, agriculture and social sector indicators;

(ii) arrange for each type of indicator to be collected by the respective sectoral, local-level authority, maybe once every two or three years;

(iii) integrate the collection of agricultural and social sectoral impact indicators in the annual survey carried out by these sectors (such as the agriculture annual survey on cereal production) and coordinate the type of data to be collected;

(iv) arrange for the local authorities (ideally with assistance by a university) to compile the data collected into a single report for each road, and disseminate the report to local and central authorities.

(b) increase local community participation in rural roads, through:

(i) at the planning stage, exchanges such as "village meetings", to ensure that road plans are properly and timely taken into account within the planning of all local- level activities;

(ii) for road maintenance, through direct participation of local stockholders in the 58

funding allocation for this activity, such as, for example, the establishment of local "road boards".36

(c) Consider, and require if cost-effective, the introduction of mitigation measures, such as education campaigns and road signs, to reduce and or prevent the likely increase in road accidents and their consequences, resulting from the increased traffic levels and speeds when rural roads are paved.

(d) Consider adopting for rural roads, identification and evaluation methodologies based on multi-criteria indicators: (i) that combine a minimum acceptable economic return with well defined social sector (mainly education and health) objectives and with social sectors investments integrated in local government budgets, or, (ii) for very low traffic levels and for a small part of the budget for rural roads (20 percent is a good practice), apply multi-criteria based solely on attaining social objectives represented by well defined indicators.

(e? Reassess the cost-effectiveness of 4-meter pavements in rural roads, and further analyze the alternative to rehabilitate rural roads to good gravel standards. The very latest version of the Bank's HDM model allows to conduct such kind of analysis.

36 Road Boards empowered to make decisions on allocation of road maintenance funds are a way to involve road users in the management of roads. See Ian Heggie, 1995. Management and Financing of Roads - An Agendafor Reform. World Bank Technical Paper Number 275. Washington, DC. 59

Annex A Study hypotheses and indicators

Area of Impact Indicator H1ypothesis Impact on transport infrastructureand services 1. increasesmotorized traffic i) traffic counts; ii) use of animal traction for transport; and iii)estimatesof householdtime savings 2. decreases transport costs i) HDM3 data; and ii) estimatesof householdtransportation costs savings 3. improvesaccess to services i) travel time to service Impact on Agriculture 4. improvesfarmers' access to markets and reduces i) transportationtime to souk; prices paid for inputs as well as household ii) amount of produceto markets; expenditures iii)use of agricultureinputs; iv) indexes of prices for agriculturaland non- agriculturalgoods; and v) prices of householdinputs and consumer goods 5. increasesthe amount of land under cultivation, i) land use measures; productivityand production ii) productionstatistics; and iii)yields per hectare 6. attracts investmentsin agriculture i) total land use data; and ii) CNCAactivity 7. increases farmers' incomes i) agriculturalvalue added; and ii) farmer estimates 8. decreases workloadfor men and women i) estimatesof householdtime savings 9. reduces the demand for short-termlabor i) local estimatesof unemployment;and ii) estimates of on farm labor demand and off- farm labor supply 10. increases land and housingprices i) local land and housingprices 11. increases other types of economicactivities i) nights in hotels; ii) numberof hotel rooms; iii)number of employeesin tourist sector; and iv) numberof small businesses(milling, mining, garages,banks, and shops) Impact on Social Services 12. facilitatesthe use of social services i) frequencyof contact with health facilities; and ii) school attendance(gender segregated) 13. leads to increasingsocial and economic i) number of health and educationalfacilities; infrastructure ii) potablewater and electricityavailability; iii)number of small earthen dams; and iv) number of markets Annex A 60

Area of Impact Indicator Hypothesis 14. increases mobility and tightens rural-urbanlinks i) number of visits to next largest town; ii) number of visits of urban relatives; and iii)number of rural and urban off-farm work days Impact on the environment 15. results in adverse impactson the environment i) use of contaminatinginputs; and ii) numberof pollutingactivities 61

Annex B

National Trends in Agriculture, Roads, Health and Education.

Figure B. 1. National trends in the agricultural sector: agricultural land and production of cereal and legume. 1985 = 100 Changes in harvestable and harvested area 130

120

110

100,______901 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 + Harvestable Area -0Harvested Area

1985 = 100 Changes in cereal and legume production 200-

U 150

0

0.

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 + Cereal Production -. Legwne Production Source: Morocco Statistical Yearbooks 62 Annex B

Figure B.2. National trends in the road sector: number of motorized vehicles, vehicle-km and road network classification. Changes in the nunber of motorized vehicles and 1985= 100 vehice-kum 180

140 ° = _

______> 120______

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 . Motorized vehicle fleet G Number of vehicle-km

1985 = 100 Changes in the road network

112 ______

108

o 06

04

102 , ___ 10

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 - Main _ Secondary A Tertiary x Total Source: Comptage Routiere, 1994. Principaux Resultats. (Direction des Routes et de la Circulation Routiere, Ministere des Travaux Publics). 63 Annex B

Figure B.3. National trends in health services: number of physicians working for the public sector and admissions to hospitals.

1985 100

250 ______f ___

200250 _

.0

E 50 0~~~~~~~~~~~~~~~~

x x~~~~~~~~~~

t00 x __ _ _ x

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Physicians -X- Admissions to hospitals

Source: Morocco Statistical Yearbooks Annex B 64

Figure B.4. National trends in the primary education sector: number of students and teachers, and changes in attendance.

1985 = 100 Changes in the number of students and teachers 140 -- m primaryeucation

120 _-______

.~100* z

801- - - 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 0 _ No. of stud:nts No. of teachers Source: Morocco Statistical Yearbooks

Changes in male and femaleattendance: 60% rural schools

40%

20%

0% 1 1982 1994 * % Male D % Female

Source: Statistiques Scolaires, 1993/94. (Direction de la Planification, Division de Statistiques, Ministere de l'Education Nationale). 65

Annex C Study Methodology

Questionnaire Design

1. During the monthsof Marchand April, the team designed five questionnairesfor data collection correspondingto differentlevels of survey:three at the householdand village levels, one at the provincialpublic works representativelevel and one at the local representativeof the ministry of agriculture level. The team trained two field data collectorsresponsible for administeringthe village and householdlevel surveys,field tested the questionnaires(last week of April 1995), and continuallyrefined the questionnaireitems to respondto the levels of analysis required to test study hypotheses.

Primary Data

Household Level Data

2. Originally,sample design called for the collectionof data from 200 small, mediumand large-sizedrural householdsin four projectzones and four control zones, with householdsbeing selected approximatelyequally in each of the four studyareas (fifty per area). Data collection began in Chefchaouenon May 29 and was completedtwo weeks later. Data for Settat was collected the first two weeks of June, and for Marrakech(two projectroads) in the final two weeks of June. The studyteam also conductedfocus group interviewsand collectedvillage level data and sample farm-leveldata (UnqueteExploitation) at the same time as the household-level data. Copiesof all questionnairesare includedin Annex 2.

3. Because of the shortageof large-sizefarms in the Chefchaouenprovince, the study comprises householdlevel data from 188 sample farms. Table I showsthe frequencyof farms sampled by size and region.

Focus Group Data

4. In each of the villages, Rural Sociologistconducted focus group interviewswhich varied between 10 and 20 villagers per interview. The focus group interviewdata providedimportant qualitative insightsfor the analysis of the householdlevel data. See Volume II, Annex A for the data made availablefrom the focus group interviews.

Village Level Data

5. In additionto household-leveldata, village leveldata was collected for 9 projectzone villages and 3 control zone villages. The VillageLevel Data questionnaireis in Annex 2.

Farm Level Data

6. Wheneverhousehold respondents indicated interest in the study,data collectorsrequested whether these villages would be willingto answer a series of detailed questionsconcerning farm budgets. Fourteen villagersagreed to such requests(3 in Chefchaouen,4 in Settat, and 7 in Marrakech)and provideddetailed crop productiondata (sugar beets, chickpeas,olives, barley, 66 Annex C hard and soft wheat) which will be used in assessing the over-all agricultural value added in project zones.

Table 1. Frequency of Sample Farms by Size and Province Province Chefchaouen Settat Marrakech Total

Prjiect Villages 3 2 4 9 Farms: 26 40 90 156 Small 16 22 49 87 Medium 5 12 28 45 Large 5 6 13 24 Total 29 42 94 164

Control Villages I I 1 3 Farms: 12 6 10 28 Small 4 3 6 13 Medium 4 1 3 8 Large 4 10 1 15 Total 13 11 11 35

Study Launch Workshop

7. Before proceeding with the data collection, the team, OEDD, and DRCR agreed to organize a study launch workshop to encourage input from local level officials from the Ministries of Agriculture and Public Works, to finalize project and control roads selection on the basis of that input, to revise questionnaire design as appropriate, and to develop consensus for the local level data collection effort as well as for the recommendations which would ultimately flow from this study. On May 4, 1995, team members met with 11 representatives of the DRCR and provincial level agricultural and public works staff from Larache, Settat and Marrakech. OED staff also attended the one-half day workshop. In addition to reviewing the methodological approach and the selection of project and control roads, participants discussed the local level approvals iequired for the survey, the timing of the data collection, and questionnaire construction. Provincial participants agreed with the approach, indicated their willingness to undertake the local level authorizations for the study, and promised to submit any recommendations for revisions in cluestionnaire design. Five days later, on the basis of the final recommendations from local level staff, the team finalized the questionnaires and the schedule for data collection.

Data Collection and Analysis

8. The household, village level, and provincial level (agriculture and public works) data used in this analysis was collected over a six-week period from the end of May until the beginning of July 1q95.

9. The household and village level questionnaire data was entered into Excel spread sheets; statistical analysis was performed in STATVIEW in order to identify indicators for which changes between 1985 and 1995 in project villages exceeded those changes for control roads. 67 Annex C

The data from the focus group interviews was textually analyzed to determine anecdotal support for the village and household level surveys. In early July 1995, a member of OED traveled to Rabat to review preliminary findings from Chefchaouen, verify World Bank requirements for rural sector impact evaluations and to visit one of the project roads (603 in Chefchaouen). 68

Annex D Survey Results Aggregated by Region, Project and Control Zones, 1985-1995 ProjectZones ControlZones Percentdifference Chefchaouen 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control

IMPACTS ON TRANSPORT INFRASTRUCTURE AND SERVICES

Access to markets and services (time in hours) To souk 0.6 0.4 0.3 0.3 -31% -7% -24% To school 0.4 0.3 0.2 0.2 -32% 14% -46% To water 0.6 0.5 1.6 1.6 -12% 3% -14% To health facilities 0.6 0.3 1.2 0.8 -53% -39% -14% To extension CT 2.7 0.8 3.4 1.5 -71% -57% -15% To Adm. (Chef Lieu) 0.7 0.4 1.8 0.9 -52% -53% 0% To Adm. (Caidat) 2.6 0.7 3.0 1.3 -75% -57% -18% Mean mode of transportation Souk Horse Horse Horse Horse same same same School Walk Walk Walk Walk same same same Water Horse Horse Horse Horse same same same Health facilities Horse Walk Horse No Info change N/A N/A Extension Horse Taxi Horse No Info change N/A N/A Adm (Chef Lieu) Horse Walk Horse No Info change N/A N/A Adm. (Caidat) Horse Taxi Horse No Info change N/A N/A Transport costs (Dh/year) Agriculture Inputs 23.6 14.8 51.3 124.6 -37% 143% -180% Agriculture Outputs: Transport diary products 0 0 0 166.7 N/A N/A N/A Transport sold livestock 0 0 5.8 137.9 N/A 2266% N/A Transport agriculture outputs 54.0 18.8 130.8 172.5 -65% 32% -97% Cost of transport to: (Dh) Souk 0 0.5 0 0 N/A N/A N/A School 0 0 0 0 N/A N/A N/A Water 0 0 0 0.8 N/A N/A N/A Health Facilities 0.3 2 0 0.8 532% N/A N/A Extension 2.4 15.9 0 31.3 576% N/A N/A Adm. (Chief Lieu) 0.4 2.8 0 5.0 634% N/A N/A Adm. (Cadiat) 3.0 10.7 0 23.8 260% N/A N/A

IMPACTS ON SOCIAL SERVICES

Household characteristics No. of persons/household No Info 7.2 No Info 8.1 N/A N/A %ofwomen No Info 41.72 No Info 45.8 N/A N/A N/A % of girls w/in group of people <7 years No Info 40.38 No Info 47.8 N/A N/A N/A % of girls w/in group of people 7-15 years No Info 39.37 No Info 52.4 N/A N/A N/A % of women w/in people >15 years No Info 41.46 No Info 41.8 N/A N/A N/A %ofboysgoingtoschool 49.1 81.3 66.7 75.0 65% 12% 53% Annex D 69

Project Zones Control Zones Percent difference Chefchaouen 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control % of girls going to school 9.9 38.5 No Info 66.7 287% N/A N/A Infirmary Distance (km) 1.0 1.0 0.3 0.3 0% 0% 0% Visits/year 2.6 5.2 2.5 4.6 98% 83% 15% Men to infirmary/year 1.4 2.8 1.0 2.3 104% 133% -29% Women to infirmary/year 1.2 2.4 1.5 2.3 96% 55% 41% Health Centers Distance (km) No info Visits/year No info Men to health centers/year No info Women to health centers/year No info Hospital Distance (km) 21.7 21.7 24.5 24.5 0% 0% 0% Women to hospital/year 0.7 1.7 0.3 1.0 138% 300% -162% Visits to hospital/year 0.9 2.6 0.6 1.4 180% 145% 36% Men to hospital/year 0.2 0.9 0.3 1.0 267% 203% 64% Mobility and rural-urban links Visits/yr. of urban relatives to farm 3.7 17.5 1.5 6.3 381% 322% 59% Visits/year to city by head of household 22.4 134.3 5.1 66.5 500% 1209% -709% Visits/year to city by other household 6.9 38.6 3.4 34 461% 894% -433% Employment in farm (number of people) Total people working 8.7 9.2 6.0 8.4 5% 40% -35% Employees 5.5 4.7 3.8 5.5 -13% 44% -57% Men employees 4.1 3.5 2.8 3.3 -13% 18% -31% Women employees 1.4 1.2 1.0 2.0 -14% 100% -114% Family 3.2 4.2 2.6 3.1 28% 19% 9% Family men 2.2 2.8 1.3 2.2 31% 74% -43% Family women 1.1 1.6 0.9 1.3 41% 45% -3% Employment outside farm by family members (No. of days) All family members 8.9 29.5 10.8 90.0 234% 731% -497% Head of household 3.5 7.6 0 0 120% N/A N/A Other family members 3.1 42.5 10.8 87.5 1280% 708% 572% Other men family members 5.4 47.3 10.8 90.0 779% 731% 48% Other women family members 0.9 1.2 0 0 35% N/A N/A Expenses (Dh/year) Food/year 7673 14140 13217 17983 84% 36% 48% Transportation/year 716 1287 183 1238 80% 575% -495% Clothing/year 382 737 371 696 93% 88% 5% Education/year 150 355 283 839 137% 196% -59% Health/year 915 881 483 1638 -4% 239% -242% Other/year 478 1011 479 1408 112% 194% -82% Total 10313 18410 15017 23801 79% 58% 20%

The high enrollment rate in the control zone in 1995 may be due to proximity to primary school, as respected by the low access time to school. 70 Annex D

Project Zones Control Zones Percent difference Chefchaouen 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control IMPACTS ON AGRICULTURE Extension services Distance (km) to CT 28.5 28.5 31.3 31.3 0% 0% 0% Contacts/year w/CT 2.0 15.9 4.8 26.8 678% 455% 223% Distance (km) to DPA 22.0 22.0 25.0 25.5 0% 2% -2% Contacts/year w/DPA 0.6 7.2 1.1 16.5 1124% 1428% -304% Farming areas (hectares) Areaperfarm 7.13 6.14 10.96 12.79 -14% 17% -31% For legumes 1.43 1.07 1.75 1.63 -25% -7% -18% For trees 57.77 85.69 104.58 124.17 48% 19% 30% For cereals 4.63 5.01 6.88 7.42 8% 8% 0% For forages 0.27 0.04 0 0.17 -85% N/A N/A For "other" 0.63 1.33 2.33 3.5 111% 50% 61% Livestocl (number per farm) Beekeeping hives 4.2 1.2 0 0 -71% N/A N/A Chicken 14.9 17.0 12.5 17.9 14% 43% -29% Cows, crossed breed 0 0 0 1.0 N/A N/A N/A Cows, indigenous breed 4.5 3.4 4.2 3.8 -25% -10% -15% Cows, pure breed 0 0 0 0 N/A N/A N/A Donkeys 1.1 0.6 0.7 0.8 -45% 12% -57% Goats 4.6 1.2 5.1 6.2 -75% 21% -97% Horses 0.9 0.4 0.7 0.5 -62% -25% -37% Mules 1.7 1.6 0.8 1.7 -6% 123% -129% Sheep 31.9 29.8 38.2 30.4 -6% -20% 14% Farm Inputs: quantity Fertilizer Deep qx/ha. 1.1 1.1 1.4 2.3 1% 66% -65% Fertilizer surface qx/ha. 0.5 1.3 0.8 1.5 154% 82% 72% Fuel litter/year 1.2 1.5 0 No Info 34% N/A N/A Herbicides litter/year 0.3 1.53 0.67 1.5 394% 124% 270% Manure T/year 0 0 0 0 N/A N/A N/A Selected seeds qx/ha. 0.8 1.3 1.1 1.3 66% 19% 47% Farm Inputs: cost (Dh/year) Fertilizer Deep 169.1 461.0 235.0 536.7 173% 128% 44% Fertilizer surface 66.2 268.5 122.9 257.9 306% 110% 196% fuel 4.0 21.5 0 7458.3 433% N/A N/A Herbicides 17.3 130.2 45.8 139.2 652% 204% 448% Manure 0 0 0 0 N/A N/A N/A Selected seeds 202.9 466.5 275.8 487.5 130% 77% 53% Veterine expenses 248.1 414.4 904.2 1995.8 67% 121% -54% Farm outputs: yield (qx/ha.) Almonds (kg/tree) No Info Barley 17.5 24.6 19.8 27.0 40% 36% 4% Chick peas 6.5 6.7 8.4 11.7 3% 39% -36% Corn No Info Faba beans No Info Forages No Info Maraich. No Info Milk production (1/year) 1716.0 923.2 1586.3 4899.6 -46% 209% -255% Olive oil (litters/tree) 2.7 3.8 28.6 2.8 39% -90% 129% 71 Annex D

Project Zones Control Zones Percent difference Chefchaouen 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control Wheat 12.1 18.0 15.5 22.6 49% 46% 3% Livestock under production (number) Bovine 4.1 3.4 5.0 4.9 -17% -2% -15% Goat 4.7 1.9 5.1 2.8 -59% -44% -15% Milk cows 0.7 0.6 0.8 1.0 -15% 20% -36% Sheep 31.9 29.8 37.9 30.4 -6% -20% 13% Quantity of farm products sold in markets Almonds (kg/year) 0 0 0 0 N/A N/A N/A Barley (qx/year) 25.0 36.0 45.8 55.3 44% 21% 23% Bovine (No.) 2.2 1.5 1.2 1.3 -28% 14% -42% Chick peas (qx/year) 11.9 6.1 12.9 14.8 -49% 14% -63% Corn (qx/year) 0 0 4.2 4.7 N/A 12% N/A Faba beans (qx/year) 0 0 0 0 N/A N/A N/A Forages (qx/year) 0 0 0 0 N/A N/A N/A Goat (No.) 4.15 0.7 3.8 1.9 -82% -50% -33% Maraich (qx/year) 0 0 0 0 N/A N/A N/A Milk (litter/year) 0 3.85 0 2800.0 N/A N/A N/A Olive oil (litter/year) 103.5 154.0 179.2 235.8 49% 32% 17% Sheep (No.) 27.1 25.5 33.0 26.0 -6% -21% 15% Wheat(qx/year) 26.9 41.7 51.5 67.5 55% 31% 24% Prices of farm products Almonds (Dh/kg) No Info Barley (Dh/ql) 99.4 158.4 121.7 182.2 59% 50% 10% Bovine (Dh/unit) 2752.4 3271.2 3355.6 3162.5 19% -6% 25% Chick peas (Dh/ql) 623.0 877.3 680.0 1058.3 41% 56% -15% Corn (Dh/ql) No Info Faba beans (Dh/ql) No Info Forages (Dh/ql) No Info Goat (Dh/unit) 251.3 253.2 296.7 305.6 1% 3% -2% Maraich. (Dh/ql) No Info Milk (Dh/litter) 2.7 3.2 2.0 3.1 19% 51% -32% Milkcows(Dh/unit) 3362.6 3793.3 3833.3 4966.7 130% 30%,'o -17% Olive oil (Dh/litter) 16.7 25.8 15.3 21.6 54% 42% 13% Sheep (Dh/unit) 405.1 426.1 376.4 405.6 5% 8% -3% Wheat (Dh/ql) 246.5 353.7 255.8 338.3 43% 32% 11% 72

Annex D

Project Zones Control Zones Percent difference Settat 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control

IMPACTS ON TRANSPORT INFRASTRUCTURE AND SERVICES

Access to markets and services (time in hours) To souk 1.1 0.7 1.5 1.2 -35% -19% -16% To school 0.3 0.3 No Info 0.3 -4% N/A N/A To water 0.1 0.1 2.9 0.3 0% -90% 90% To health facilities 1.0 0.5 1.0 0.6 -54% -38% -16% To extension CT 1.6 0.7 No Info No Info -57% N/A N/A To Adm. (Chef Lieu) 0.9 0.6 1.4 1.2 -37% -11% -26% To Adm. (Caidat) 1.8 0.7 1.4 1.4 -65% -1% -63% Mean mode of transportation Souk Horse Taxi Horse Horse change same change School Walk Walk Walk Walk same same same Water Walk Walk Horse Horse same same same Health facilities Walk Horse Horse Horse change same change Extension Horse Taxi Horse Horse change same change Adm (Chef Lieu) Horse Horse Horse Horse same same same Adm. (Caidat) Horse Taxi Horse Horse change same change Tranport costs (Dh/year) Agriculture Inputs 62.4 73.8 40.5 86.0 18% 112% -94% Agriculture Outputs: Transport diary products 0.0 0.0 0.0 0.0 N/A N/A N/A Transport sold livestock 112.7 113.9 83.0 96.0 1% 16% -15% Transport agriculture outputs 187.7 No Info 115.0 188.5 N/A 64% N/A Cost of transport to: (Dh) Souk 0.1 No Info 0.0 0.5 N/A N/A N/A School 0.0 0.0 0.0 0.0 N/A N/A N/A Water 0.0 0.0 0.0 0.0 N/A N/A N/A HlealthFacilities 0.0 No Info 0.0 0.3 N/A N/A N/A Extension 0.0 No Info No Info No Info N/A N/A N/A Adm. (Chief Lieu) 0.0 No Info 0.0 0.0 N/A N/A N/A Adm. (Cadiat) 0.0 No Info 0.8 0.5 N/A -38% N/A IMPACTS ON SOCIAL SERVICES Household characteristics No. persons/household No Info 10.5 No Info 9.2 N/A N/A N/A % of women No Info 49.1 No Info 42.3 N/A N/A N/A % of girls w/in group of people <7 years No Info 49.9 Nc Info 38.4 N/A N/A N/A % of girls w/in group of people 7-15 years No Info 48.2 No Info 40.1 N/A N/A N/A % of women w/in people > 15 years No Info 49.4 No Info 44.2 N/A N/A N/A % of boys going to school 43.8 84.6 No Info 70.4 93% N/A N/A % of girls going to school 25.7 67.4 No Info 55.0 163% N/A N/A Infirmary Distance (km) 7.7 3.2 No Info 3.0 -58% N/A N/A Visits/year 5.7 6.2 No Info 6.0 8% N/A N/A Men to infirmary/yea; 2.8 2.6 No Info 2.9 -7% N/A N/A Women to infirmary/year 3.3 3.5 No Info 3.1 8% N/A N/A 73 Annex D

Project Zones Control Zones Percent difference Settat 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control Health Centers Distance (km) 10.2 8.9 9.6 9.6 -12% 0% -12% Visits/year 2.9 6.8 3.7 4.9 131% 32% 99% Men to health center/year 1.4 2.8 1.7 1.9 107% 12% 95% Women to health center/year 1.6 4.0 2.2 3.0 147% 36% 111% Hospital Distance (km) 57.4 57.4 48.4 48.4 0% 0% 0% Women to hospital/year 0.7 1.8 0.7 1.3 154% 86% 68% Visits to hospital/year 1.0 2.6 0.9 1.7 163% 89% 74% Men to hospital/year 0.3 0.9 0.2 0.4 193% 100% 93% Mobility and rural-urban links Visits/yr. of urban relatives to farm 0.8 26.6 1.3 2.3 3223% 77% 3146% Visits!year to city by head of household 49.0 342.6 96.4 129.8 600% 35% 565% Visits/year to city by other household 14.8 127.1 21.5 32.4 758% 51% 708% Employment in farm (number of people) Total people working 7.5 7.7 6.4 6.7 3% 5% -1% Employees 4.5 4.3 3.4 2.1 -3% -38% 36% Men employees 4.3 4.2 3.1 2.6 -4% -16% 12% Women employees 0.2 0.1 0.3 0.0 -47% -100% 53% Family 2.9 3.4 3.0 4.1 17% 37% -20% Family men 2.0 2.3 1.9 2.8 12% 47% -35% Family women 0.9 1.1 1.1 1.3 25% 18% 7% Employment outside farm by family members (No. of days) Allfamilymembers 23.8 117.1 10.0 11.0 392% 10% 382% Head of household 16.3 18.0 2.0 2.0 11% 0% 11% Others family members 7.5 90.1 8.0 9.0 1096% 13% 1084% Other men family members 7.5 89.6 8.0 9.0 1094% 13% 1082% Other women family members 0.0 9.5 0.0 0.0 31567% N/A N/A Expenses (Dh/year) Food/year 18525.0 33527.0 10972.0 21060.0 81% 92% -11% Transportation/year 2741.3 4032.8 690.0 1374.0 47% 99% -52% Clothing/year 450.8 956.3 395.0 675.0 112% 71% 41% Education/year 630.8 1451.5 162.5 980.0 130% 503% -373% Health/year 1206.3 2613.8 455.0 800.0 117% 76% 41% Other/year 614.7 1970.4 210.0 342.5 221% 63% 157% Total 24168.7 44551.6 12884.5 25231.5 84% 96% -11% IMPACTS ON AGRICULTURE Extension services Distance (km) CT 20.5 20.5 34.5 34.5 0% 0% 0% Contacts/year w/CT 9.9 14.0 0.1 0.4 42% 300% -258% Distance (km) DPA 90.2 90.2 48.2 48.2 0% 0% 0% Contacts/year w/DPA 0.1 1.7 0.0 0.0 1208% N/A N/A Farming areas (hectares) Areaperfarm 16.1 24.3 6.7 6.7 51% 0% 51% For legumes 0.4 0.2 0.6 0.5 -59% -18% -41% For trees 9.0 13.2 0.6 3.0 47% 400% -353% For cereals 15.6 24.3 6.0 6.1 55% 2% 54% For forages 0.1 0.3 0.0 0.0 460% N/A N/A 74 Annex D

Project Zones Contr ol Zones Percent difference Settat 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control For "other" 0.1 0.1 0.0 0.0 0% 0% 0% Livestock (number per farm) Beekeeping hives 0.0 0.0 0.0 0.0 N/A N/A N/A Chicken 9.1 11.0 7.3 8.0 21% 10% 12% Cows, crosse:dbreed 0.6 1.1 0.0 0.2 75% N/A N/A Cows, indigenous breed 3.6 2.1 2.4 2.1 -41% -13% -28% Cows, pure breed 0.2 0.5 0.2 0.3 150% 50% 100% Donkeys 1.0 0.8 0.7 1.2 -18% 71% -90% Goats 3.0 2.4 4.0 1.5 -19% -63% 44% Horses 0.5 0.3 0.4 0.7 -53% 75% -128% Mules 0.7 0.4 0.6 0.1 -46% -83% 38% Sheep 84.1 57.4 32.3 22.6 -32% -30% -2% Farm Inputs: quantity Fertilizer Deep qx/ha. 0.0 0.0 0.2 0.1 N/A -33% N/A Fertilizer surface qx/ha. 0.0 0.0 0.0 0.2 N/A N/A N/A Fuel liter/year 380.5 801.5 21.6 28.0 111% 30% 81% Herbicides liter/year 0.3 0.5 0.6 0.7 41% 17% 25% Manure T/year 1.4 1.4 0.0 0.0 2% N/A N/A Selected seeds qx/ha. 1.0 1.2 1.2 1.5 11% 33% -22% Farm Inputs: cost (Dh/year) Fertilizer Deep 0.0 0.0 28.5 44.0 N/A 54% N/A Fertilizer surface 0.0 0.0 0.0 28.0 N/A N/A N/A Fuel 1535.0 3632.8 82.0 126.0 137% 54% 83% Herbicides 12.6 23.9 16.0 31.0 89% 94% -5% Manure 0.0 0.0 0.0 0.0 N/A N/A N/A Selected seeds 252.6 366.6 285.0 409.0 45% 44% 2% Veterine expenses 671.8 1865.0 323.0 440.0 178% 36% 141% Farm outputs: yield (qx/ha.) 11.2 N/A 4.2 6.0 N/A 44% N/A Almonds (kg/tree) No Info Barley 16.7 22.5 27.5 31.4 35% 14% 20% Chick peas No Info Corn No Info Faba beans No Info Forages No Info Maraich. No Info Milk production (I/year) 2168.1 4120.6 3530.0 2996.0 90% -15% 105% Olive oil (liters/tree) 3.5 3.6 0.0 0.0 3% N/A N/A Wheat 12.0 16.5 23.0 26.9 37% 17% 20% Livestock under production (number) Bovine 4.4 3.4 2.6 2.6 -22% 0% -22% Goat 3.0 2.5 3.0 1.5 -19% -50% 31% Milk cows 0.4 0.2 0.1 0.2 -39% 100% -139% Sheep 83.2 57.1 32.3 22.6 -31% -30% -1% Quantity of farm products sold in markets Almonds (kg/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Barley (qx/year) 16.5 18.5 14.8 21.3 12% 44% -31% Bovine (No.) 1.5 1.2 0.7 0.6 -20% -14% -5% Chick peas (qx/year) 0.0 0.0 1.0 0.3 N/A -75% N/A Corn (qx/year) 0.0 0.0 2.4 4.0 N/A 65% N/A Annex D 75

Project Zones Control Zones Percent difference Settat 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control Faba beans (qx/year) 0.0 2.0 0.0 0.0 N/A N/A N/A Project Control Percent Zones Zones differen ce Set/at 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control Maraich (qx/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Milk (liter/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Olive oil (liter/year) 15.0 21.3 0.0 0.0 42% N/A N/A Sheep (No.) 58.5 38.5 29.2 19.6 -34% -33% -1% Wheat (qx/year) 78.6 334.0 14.3 20.6 325% 44% 281% Prices of farm products Almonds (Dh/kg) No Info Barley (Dh/ql) 76.7 140.0 85.5 137.0 83% 60% 22% Bovine (Dh/unit) 4820.0 6531.8 5150.0 5666.0 36% 10% 25% Chick peas (Dh/ql) No Info Corn (Dh/ql) No Info Faba beans (Dh/ql) No Info Forages (Dh/ql) No Info Goat (Dh/unit) 298.0 360.5 No Info No Info 21% N/A N/A Maraich. (Dh/ql) No Info Milk (Dh/litter) 2.3 3.2 2.0 3.0 44% 50% -6% Milk cows (Dh/unit) 5748.9 8214.2 5220.0 6320.0 43% 21% 22% Olive oil (Dh/litter) 19.1 24.9 No Info No Info 30% N/A N/A Sheep (Dh/unit) 580.6 685.1 455.0 625.0 18% 37% -19% Wheat (Dh/ql) 212.7 291.8 252.5 301.0 37% 19% 18% 76 Annex D

Project Zones Control Zones Percent difference Marrakech 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control

IMPACTS ON TRANSPORT INFRASTRUCTURE AND SERVICES

Access to markets and serjices (time in hours) To souk 1.1 0.4 2.1 2.1 -60% 0.0 -60% To school 0.2 0.2 No Info 0.3 -30% N/A N/A To water 0.1 0.1 0.2 0.2 -40% 0.0 -40% To health facilities 1.0 0.4 1.8 1.8 -60% 0.0 -60% To extension CT 1.0 0.5 No Info No Info -50% N/A N/A To Adm. (Chef Lieu) 1.2 0.4 2.0 2.0 -70% 0.0 -70% To Adm. (Caidat) 2.1 0.7 2.1 2.1 -70% 0.0 -70% Mean mod ol ransportation Souk Horse Taxi Hlorse Horse change same change School Walk Walk Walk Walk same same same Water Walk Walk Walk Walk same same same ilealth facilities Horse Taxi Horse Horse change same change Extension Horse Taxi Horse Horse change same change Adm (Chef Lieu) Horse Taxi Horse Horse change same change Adm. (Caidat) Horse Taxi Horse Horse change same change -raj1sportcosts (Dh/year) Agriculture inputs 118.0 188.5 48.9 59.2 60% 20% 40% Agriculture outputs: Transport diary products 74.5 110.1 0.0 0.0 50% N/A N/A Transport sold livestock 118.0 19.9 24.0 23.5 -80% 0.0 -80% Transport agriculture outputs 218.6 304.4 144.3 202.9 40% 40% 0 i f transport to: (Dh) Souk 11 2.6 0.0 0.0 140% N/A N/A School 0.0 0.0 No Info No Info N/A N/A N/A Water 0.1 0.1 0.0 0.0 -50% N/A N/A Health Facilities 1.8 2.9 0.5 0.8 70% 60% 10% Extension 3.1 3.6 No Info No Info 20% N/A N/A Adm. (Chief Lieu) 0.1 2.8 0.0 0.0 222% N/A N/A Adin. (Cadiat) 0.9 6.0 0.0 0.0 560% N/A N/A IMIPACTS (ON SOCIAL SERVICES o-pstohoidcharacteristics No. persons/household No Info 8.2 No Info 8.9 N/A N/A N/A % of women No Info 47.9 No Info 44.6 N/A N/A N/A of dirls w/,ingroup of people <7 years No Info 47.0 No Info 20.0 N/A N/A N/A %/;oi girls w/in group of people 7-15 years No Info 45.3 No Info No Info N/A N/A N/A ° of womnn w/in people >15 years No Info 50.2 No Info 48.6 N/A N/A N/A %/of boys going to scilool 34.2 79.9 No Info No Info 130% N/A N/A %/of girls going to school 14.8 51.8 No Info No Info 250% N/A N/A t!i madry Distance (krn) 2.5 3.1 10.0 No Info 20% N/A N/A Visits/year 4.2 6.1 4.5 No Info 50% N/A N/A Men to infirmary/year 2.1 3.3 2.5 No Info 60% N/A N/A Women to infirmary/year 2.1 3.1 2.0 No Info 50% N/A N/A 77 Annex D

Project Zones Control Zones Percent difference Marrakech 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control Health Centers Distance (km) 10.5 8 4 No Info 10.0 -20% N/A N/A Visits/year 2.8 4.4 No Info 5.0 60% N/A N/A Men to health center/year 1.2 2.1 No 'tfo 2.7 70% N/A N/A Women to health center/year 1.6 2.4 No Info 2.9 50% N/A N/A Hospital Distance (km) 71.5 71.5 48.6 48.6 0.0 0.0 0.0 Women to hospital/year 0.8 1.4 0.7 1.l 70% 60% 10% Visits to hospital/year 1.2 2.2 1.1 1.8 80% 60% 20% Men to hospital/year 0.4 0.8 0.4 0.7 100% 80% 20% Mobility and rural-urban links Visits/yr. of urban relatives to farm 1.7 9.2 0.5 0.9 430% 80% 350% Visits/year to city by head of household 31.9 118.2 7.3 8.9 270% 20% 250% Visits/year to city by other household 12.4 38.6 2.6 5.4 210% 110% 100% Employment in farm (number of people) Total people working 7.1 7.9 5.1 5.3 10% 0.0 10% Employees 4.8 5.2 2.6 1.2 10% -50% 60% Men employees 4.7 5.0 2.6 1.1 10% -60% 70%/io Women employees 0.2 0.2 0.1 0.1 -20% 0.0 -20% Family 2.3 2.7 2.5 4.1 20% 60% -50% Family men 1.5 1.7 1.6 2.8 20% 80% -60% Family women 0.8 0.9 1.0 1.3 20% 30% -10% Employment outside farm by family members (No. of days) Days work by family outside farm 34.6 49.3 58.0 114.0 0.4 1.0 -0.5 Days work by HH head outside farm 24.7 20.9 2.0 0.0 -0.2 -1.0 0.8 Days work by others outside farm 9.6 28.1 58.0 114.0 1.9 1.0 1.0 Daysworkbyothermenoutsidefarm 10.9 30.1 43.0 114.0 1.8 1.7 0.i Days work by other women outside farm 0.0 0.3 15.0 0.0 N/A -1.0 N/A Expenses (Dh/year) Food/year 7996.6 13167. 6916.0 11908. 0.6 0.7 -0.1 4 0 Transportation/year 841.7 1398.1 175.0 265.0 0.7 0.5 0.1 Clothing/year 305.8 528.2 250.0 335.0 0.7 0.3 0.4 Education/year 253.5 415.0 No Info No Info 0.6 N/A N/A Health/year 516.8 822.7 355.0 405.0 0.6 0.1 0.5 Other/year 155.5 338.4 270.0 395.0 1.2 0.5 0.7 Total 10069. 16669. 7966.0 13308. 0.7 0.7 0.0 8 8 0 IMAPACTSON AGRICULTURE Extension services Distance (km) CT 25.3 17.3 42.0 42.0 -0.3 0.0 0.3 Contacts/year w/CT 1.2 4.6 0.0 0.0 2.9 N/A N/A Distance (km) DPA 77.8 48.6 42.0 42.0 -0.4 0.0 -0.4 Contacts/year w/DPA 0.1 0.3 0.0 0.0 3.9 N/A N/A Farming areas (hectares) Area per farm 8.2 8.2 14.0 9.4 0.0 -0.3 0.3 For legumes 0.0 0.0 0.0 0.0 N/A N/A N,A 78 Annex D

Project Zones Control Zones Percent difference Marrakech 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control Fortrees 137.5 148.1 3.0 5.2 0.1 0.7 -0.7 For cereals 7.6 7.1 13.4 8.8 -0.1 -0.3 0.3 For forages 0.1 0.2 0.0 0.0 0.4 N/A N/A For "other" 0.4 0.4 0.4 0.6 0.0 0.5 -0.5 Livestock (number per farm) Beekeeping hives 0.1 0.1 1.0 0.5 0.0 -0.5 0.5 Chicken 10.0 10.7 7.8 7.5 0.1 0.0 0.1 Cows, crossed breed 0.3 0.8 0.0 0.1 1.4 N/A N/A Cows, indigenous breed 1.7 1.5 2.5 0.6 -0.1 -0.8 0.7 Cows, pure breed 0.0 0.0 0.0 0.0 N/A N/A N/A Donkeys 1.6 1.2 1.4 1.5 -0.2 0.1 -0.3 Goats 2.2 1.6 7.5 3.6 -0.3 -0.5 0.2 Horses 0.1 0.0 0.0 0.0 -0.8 N/A N/A Mules 0.4 0.3 0.3 0.2 -0.2 -0.3 0.2 Sheep 18.4 13.7 30.8 12.0 -0.3 -0.6 0.4 Farm Inputs: quantity Fertilizer Deep qx/ha. 0.3 0.4 0.0 0.2 0.4 N/A N/A Fertilizer surface qx/ha. 0.0 0.1 0.0 46.0 3.0 N/A N/A Fuel liter/year 0.0 0.0 0.0 0.0 N/A N/A N/A Herbicides liter/year 2.0 0.1 0.0 0.0 -0.9 N/A N/A Manure T/year 0.5 0.6 0.4 0.3 0.2 -0.3 0.5 Selected seeds qx/ha. 0.8 0.9 1.1 1.1 0.1 0.0 0.1 Farn Inputs: cos (Dh/year) Fertilizer Deep 41.6 87.2 0.0 40.0 1.1 N/A N/A Fertilizer surface 1.6 15.3 0.0 0.0 8.8 N/A N/A Fuel 0.0 0.0 0.0 0.0 N/A N/A N/A Herbicides 2.0 8.3 0.0 0.0 3.2 N/A N/A Manure 14.0 17.9 10.0 20.0 0.3 1.0 -0.7 Selected seeds 175.6 279.9 263.0 389.0 0.6 0.5 0.1 Veterine expenses 174.9 226.3 110.0 66.0 0.3 -0.4 0.7 Fanm outputs: yie (qx/ha.) Almonds (kg/tree) 6.4 6.0 No Info No Info -0.1 N/A N/A Barley 16.9 21.0 12.1 13.7 0.2 0.1 0.1 Chick peas No Info Corn No Info Faba beans No Info Forages No Info Maraich. No Info Milk production (1/year) 1675.8 2169.1 1175.0 238.0 0.3 -0.8 1.1 Olive oil (litters/tree) 5.1 5.0 No Info No Info 0.0 N/A N/A Wheat 13.7 16.4 10.4 12.9 0.2 0.2 0.0 Livestock under production (number) Bovine 2.1 2.3 2.4 0.7 0.1 -0.7 0.8 Goat 2.1 1.6 7.5 3.6 -0.3 -0.5 0.3 Milk cows 1.4 1.4 1.7 0.5 0.1 -0.7 0.8 Sheep 17.4 13.1 30.8 12.0 -0.2 -0.6 0.4 79 Annex D

Project Zones Control Zones Percent diference Marrakech 1985 1995 1985 1995 in project in control project- 1985-95 1985-95 control Quantity of farmnproducts sod in market Almonds (kg/year) 618.8 628.4 0.0 0.0 0.0 N/A N/A Barley (qx/year) 20.5 33.6 3.0 1.5 0.6 -0.5 1.1 Bovine (No.) 0.7 0.7 1.0 0.3 0.0 -0.7 0.7 Chick peas (qx/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Corn (qx/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Faba beans (qx/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Forages (qx/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Goat (No.) 1.2 0.9 4.3 2.1 -0.3 -0.5 0.2 Maraich (qx/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Milk (litter/year) 0.0 0.0 0.0 0.0 N/A N/A N/A Olive oil (litter/year) 559.9 566.0 3.0 3.0 0.0 0.0 0.0 Sheep (No.) 12.4 8.2 13.3 8.6 -0.3 -0.4 0.0 Wheat (qx/year) 27.8 39.3 32.0 37.5 0.4 0.2 0.2 Prices of farm products Alrnonds (Dh/kg) 41.2 50.3 No Info No Info 0.2 N/A N/A Barley (Dh/ql) 80.5 156.8 73.5 168.0 0.9 1.3 -0.3 Bovine (Dh/unit) 5582.8 6613.0 4750.0 No Info 0.2 N/A N/A Chick peas (Dh/ql) No Info Corn (Dh/ql) No Info Faba beans (Dh/ql) No Info Forages (Dh/ql) No Info Goat (Dh/unit) 344.3 394.4 No Info No Info 0.1 N/A N/A Maraich. (Dh/ql) No Info Milk (Dh/litter) 2.3 2.8 No Info No Info 0.2 N/A N/A Milk cows (Dh/unit) 6337.5 7752.2 5300.0 No Info 0.2 N/A N/A Olive oil (Dh/litter) 17.7 27.0 No Info No Info 0.5 N/A N/A Sheep (Dh/unit) 612.4 652.4 610.0 605.0 0.1 0.0 0.1 Wheat (Dh/ql) 201.5 296.1 183.0 267.0 0.5 0.5 0.0 80

Annex E

Calculation of Agricultural Value-Added Northern Region (Chefchaouen)

1. The study team conductedan analysisof changesin agriculturalvalue added in the Chefchaouenregion, both in the projectand in the control zones,and comparing 1985 and 1995. The key data and results are summarizedbelow (all monetaryvalues in constant 1985 dirhams).

2. The followingtables show the key data used for calculatingthe agriculturalvalue added in the project (Table I and 2) and in the control (Table3 and 4) zones.

Results

3. As shownby the abovetables, the agriculturalvalue added in 1995 was 7583 Dh per hectare in the project zone, comparedto 2894 Dh per hectare in the control zone.

Table 1. Transformationof AgriculturalPractices in the Project Zone of ChefchaouenProvince 1985 1995 Land Use (ha) Cereals 18,698 17,618 High Value Vegetables 5,213 4,248 Fruit Trees 3,528 4,979 Productivity(qxlha) Cereals 12 18 High Value Vegetables N/A N/A Fruit Trees 17 28 Value Added (Dh/ha) Cereals 37,396 60,464 High Value Vegetables 20,852 24,086 Fruit Trees 8,820 12,547 Net Value Addedfor Agriculture(Dhlha) 67,068 97,658 81

Annex E

Table 2. Transformation of Livestock Production Practices in the Prject Zone of Chefchaouen Province 1985 1995 Animal Production (number) Local cows 11,560 8,920 Sheep 75,527 65,105 Goat 5,924 2,654 Productivity Meat (kg/animal) Local cows 102 120 Sheep 25 30 Goat 17 28 Dairy (Litters/animal) Local cows 420 700 Value Added (Dh/animal) Local cows 2,880 3,432 Sheep 640 750 Goat 375 515

Net Value Added for Livestock Production (Dh) 85,865 122,344

Table 3. Agriculture (Crops) Value Added in ControlZone (constant 1985prices) Small Medium Large Total Land Use (ha) Average land holding 3.6 9.6 27.4 Cereal 1.9 5.1 15 Vegetables 0.9 1.7 2.2 Tree Crop 0.8 1.7 1.2 Yields Cereal (tons/ha) 1.3 1.7 2 Cereal (qx) 25 87 300 Vegetables N/A N/A N/A Fruit (qx/ha) 17 17 17 Value Added (Dh) Cereal 5,250 18,270 63,000 86,520 Vegetable N/A N/A N/A 17,609 Fruit Trees N/A N/A N/A 30,240 Net Agriculture (Crops) Value Added (Dh) 73,933 82

Annex E

Table 4. Agriculture (Livestock) Value Added in Control Zone (constant1985 prices) Small Medium Large Total Dairy Production (cows only) Average livestock holdings (number) 2 2 2 Production (litters/animal) 420 420 420 Price per litter (Dh/litter) 2.5 2.5 2.5 Value Added for Dairy (Dh 1995) 2,100 2,100 2,100 6,300 (Dh 1985) (4,410) Meat Production Cows (number) 2 2 2 Weight per animal (kg) 102 102 102 Price per kg (Dh/kg) 28 28 28 Value Added 5,712 5,712 5,712 17,136 Sheeps (number) 10 14 14 Weight per animal (kg) 25 25 25 Price per kg (Dh/kg) 35 35 35 Value Added 8,750 17,500 17,500 33,250

Value Added for Animals (Dh) 43,839 83

Annex F Economic Analysis of Road Improvements

Economic Evaluation Using HDM-III Model

1. The consultantsworking for MTDScarried out an economicevaluation of the three road projects, based on the applicationof the HDM-IIImodel. This model, developedby the World Bank, is regularly used by the MPW/DRCRto assess road investments.

2. To collect traffic data, during the last two weeks of June and all of July 1996,the DRCR, in collaborationwith local level public worksofficials, organized both visual and automatictraffic counts in each of the project zones. Visual countswere conductedover two consecutivedays (one market, one non-market)and included6 categoriesof vehiculartraffic. In addition,the team used data from the National Roads Studiesand Research Center in order to calculate vehicle operatingcosts.

3. For purposes of the HDM analysis,traffic data was compressedinto five categories of vehicles (from the seven categoriesobserved): private passengervehicles; pick-up,jeeps, mini- trucks and tractors; buses; freight trucks, less than 8 tons; freight trucks, more than 8 tons.

4. Road design characteristics,for both control and projectroads, are shown in Table 1.

Table 1. Road design characteristics Region Road Width Shoulder Topography Soil meters meters Type Northern Project 603 4-6 1.5 Mountain S2 Control 8203 4 1.5 Mountain S2 Central Project 1427 4 1.7 Flat Si Control 1411 3 2.0 Flat SI Tensift(South) Project P-6307 4 2 Flat S2 Project P-6308 4 2 Flat S2 Control C-6302 No Info No Info Flat S2

5. Table 2 summarizesthe traffic compositionon projectroads in 1995 by type of vehicle and by road.

Table 2. Observedaverage daily traffic by vehicle type for projectroads (June 1995). (percentage of total count) Road Private Pickups, Buses Doubleaxel Doubleaxel Total vehicles jeeps, trucks trucks tractors (less 8 tons) (more8 tons) 603 54 22 2 10 12 640 1427 54 20 1 12 13 275 6308 51 19 0 15 15 212 6307 24 48 1 22 5 172 84 Annex F 6. Table 3 shows the road and shoulder condition on project roads in 1993.

Table 3. Road and shoulder surface quality in 1993 for project roads. SurfaceQuality (% of road surface) ShoulderQuality (% of road length) Road Excellent Good Poor Very Excellent Good Poor Very Poor Poor 603 8 5 23 64 37 32 25 6 1427 0 100 0 0 0 0 0 100 6307/8 100 0 0 0 99 0 1 0

7. Additional assumptions used in the application of the HDM model include: benefit stream covering a 15-year period since opening of the road; minor adjustments for maintenance condition; traffic in opening year estimated from traffic in 1995, decreased by 4 percent annually. Unit vehicle operating costs, for each type of vehicle, were those currently used by the MPW. Annual maintenance are based on standards used by the MPW.

8. Rate of Returns. The economic rate of returns obtained by application of the HDM model and the data and assumptions noted above are presented in Table 4.

Table 4. Economic rate of returns based on HDM model (monetary values are in constanzt1985 prices) Road Constr&Maint VOCCosts VOCCosts Discounted EconomicRate Costs (paved) (track) BenefitStream of Return 603 9.6 215 329 105 137 1427 12.8 58 92 22 37 6307 5.5 34 53 15 54 6308 3.3 18 30 9 50

Economic Evaluation Without Using HDM model

9. The economic evaluation presented in the main text of the report did not use the HDM model. It used essentially the key parameters and coefficients of the economic analysis presented in the Staff Appraisal Report of the most recent Bank highway project, the Secondary, Tertiary and Rural Roads Project, approved in June 1994. Information in this annex is limited to complement the data and assumptions presented in the main report. Tables 5 and 6 and their footnotes provide this complementary information. Table 5. Key Data for Economic Evaluation Traffic (vehicles/day, annual average) Road Open Length Const. Cost 1982 Open Year Open Year 1995 1995 Year (kmn) (mDhW95)a Normat Estimatec Normald Actual

603 1987 35 28.9 40 45 450 55 640

1427 1988 30 17.3 150 174 202 207 275

6307 1991 10 5.8 54 67 161 74 192

6308 1991 14 8.1 54 67 161 74 192

Notes: L a The 1995Highway SAR showsaverage construction cost, paved rura; road of Dh 826,000/kmfor a 6 meterpavement (route 603). For routes 1427 and 6307/6308,with about4 meterpavement, construction costs estimatedat Dh 578,200(30% lower) b Trafficin the openingyear, assuming2.5% growthper year while an unpavedroad, from the base 1982 level. c Estimatefrom actualin 1995,assuming traffic growthwith paved road is 4.5% per year. d Trafficin 1995 assumingno project,and traffic growthcontinuiirg in unimprovedroad at 2.5% growthper year e VOCBenefits are calculatedassuming unit savingsper vehiclekm betweenpaved road and old road is Dh 2.2 (=5.2 bad road - 3.0 good road) (SAR, 1995).

Table 6. Complementary Data for Economic Evaluation Roads Annual Rate of Traffic Growth VOC Savings Resurfacing Closure Days (Dh/veh-km) frequency Normal Total for old road for open days for closure days 603 2.5% 4.5% 90 2.2 3.6 7 yearsa 1427 2.5% 4.5% 60 2.2 3.6 10 years 6307 2.5% 4.5% 30 2.2 3.6 10 years x 6308 2.5% f 4.5% 30 2.2 3.6 1r0years a Higherresurfacing frequency for Road603 due to its high traffic and deterioratedcondition in 1994after 7 yearsof usage 86

Annex G Summary of Impacts

'fable 8.1. Summary of impacts by regions according to area of impact and hypotheses Area of Impact Region Hypotheses Overall Chefchaouen Settat Marrakech

Transport infrastructure and services increases motorized vehicle higher degree of ownership of cars Low High Medium as well as trucks traffic increasing at rates higher High Low Medium than the past trend (Table 3.2) decreases transportation reduction of operating costs of High Medium Low costs vehicles lower rates of freight transport High No Info No Info passengers transport High Medium Medium larger percentage of trucks in traffic High High High offering lower rates (Table 3.3) improvements in access to year-round use of roads, eliminating services frequent road closures during rainy High High High season major increase in the supply of rural passenger services, especially share- High High High ride taxis substantial reduction in access time High Medium High by the rural population to markets and social services

Agriculture improves farmers' access to Low High Low markets reduces prices paid for Low High Medium inputs increases the amount of land the overall level of agricultural .. . ~~t Medium Medium High under cultivation, activity increased productivity and production attracts investments in increases use of modem inputs, agriculture especially fertilizers (as distribution High High Low channels became easier and lower cost) increases farmers' incomes - agricultural production mix was Low High High transformed, and land use was shifted from low-value cereals to high-value fruit orchards - livestock production was shifted to Low High Low pure breed cows as a result of the shift to higher- value products, and the overall increases in yields, the value added Low Low High per unit of cultivated land increased (para. 4.1 1) Annex G 87

Area of Impact Region Hypotheses Overall Chefthaouen Settat Marrakech decreases workload for men due to related economic changes Low Medium High and women reduces the demand for Inconclusive from available data short-term labor increases land and housing Inconclusive from available data prices increases other types of establishment of new shops Low High Low economic activities Social Sectors facilitates the use of social visits per person to health facilities High High High services nearly doubled between 1985 and 1995 enrollment of children in primary school more than doubled over the period Overall Low High Medium Girls Low High Medium leads to increasing social new education facilities were built Medium High Low- and economic infrastructure new health centers were built High High Low increasing mobility and the improved rural transport tightens rural-urban links services resulted in several-fold increases in social exchanges, both Low High Medium by relatives from the urban areas visiting the farms and by farm household members visiting nearby cities

MAP SECTION

IBRD27809

MOROCCO NORTHERNREGION TANGIERI -

IMPACTEVALUATION STUDY LARACHEPROVINCE -. TUN RURALROADS IMPROVEMENT COMPONENT B

- ~OF THEFOURTH ROAD PROJECT (l.N2254)

IA ~~~~~PROJECTROADS AN-~'

- - ~~~~~~~~~~~~~CONTROLROADS -. ALTERNARTVECONTROL ROADS -MAIN ROADS - ~~~~~~~~~~0SELEC-TEDCTIES IYO, A A H A T~ ~~o 0 PROVINCECAPITALS NATIONALCAMlTALCHFHON RPPROE~~~~~~~~ - - ~~~~PROVINCEBOUNDARIES

-- INTERNATnONALBOUNDARIES -1, 0~~~50T20 I50 2~O22CJ - -- INTERNATIONALBORUNDARIES JAPPROSIMATE)

2 CASABL4NCA ~CENTRALREGION -

* ~~~~ Sc-l-BENE SElAT PROVINCEKNTA

KENITRA L ELA E

t I KACEM~~~~~~~~~~~~~~~~~~~SAIISRRN

KHEMISSET~ ~ SSOLIR

EL N~~~~~~~~~~~~~~~~~~~~~

05 S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~AROFI SS1ZIARHNA1.

1.S-.f-MY El Kd.. 99

I