L_Leonetti_FINAL_Report_Assignment_II_final.pdf

Acknowledgements

December 2012 This publication was produced for review by the United States Agency for International Development. It was prepared by Tetra Tech ARD.

KOSOVO NEW

OPPORTUNITIES FOR

AGRICULTURE PROGRAM

TECHNICAL ASSISTANCE FOR THE ANALYSIS OF

PROFITABILITY OF SELECTED FARM ACTIVITIES

FINAL REPORT: TECHNICAL ASSISTANCE

ASSIGNMENT II

This report was prepared by the New Opportunities for Agriculture project team of Tetra Tech ARD based on a Final Report prepared by:

Luciano Leonetti Rome, Italy

Prepared for the United States Agency for International Development, USAID Contract Number EDH-I-00-05-00006-00, TO AID-167-TO-11-0000, Kosovo New Opportunities for Agriculture Program, under the Rural and Agricultural Incomes with a Sustainable Environment (RAISE Plus) Indefinite Quantity Contract (IQC).

Implemented by:

Tetra Tech ARD 159 Bank St., Suite 300 Burlington, VT 05401

KOSOVO NEW

OPPORTUNITIES FOR

AGRICULTURE

PROGRAM

TECHNICAL ASSISTANCE FOR THE

ANALYSIS OF PROFITABILITY OF SELECTED

FARM ACTIVITIES

FINAL REPORT: TECHNICAL ASSISTANCE

ASSIGNMENT II

DECEMBER 2012

DISCLAIMER

The authors’ views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government.

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED i

FARM ACTIVITIES

FINAL REPORT ASSIGNMENT II

TABLE OF CONTENTS

TABLE OF CONTENTS ............................................................................................................ I ACRONYMS AND ABBREVIATIONS .................................................................................... III EXECUTIVE SUMMARY ....................................................................................................... IV

BACKGROUND.................................................................................................................... IV MAIN FINDINGS OF THE MISSIONS ....................................................................................... IV OUTCOMES OF THE PROCESS .............................................................................................. V OUTCOMES OF THE ANALYSES ........................................................................................... VI CRITICAL POINTS OF THE ANALYSES .................................................................................. VII RECOMMENDATIONS ......................................................................................................... VII

1.0 INTRODUCTION

2.0 OBJECTIVES OF THE MISSION

3.0 ORGANIZATION OF ACTIVITIES

4.0 OUTCOMES OF THE MISSION

4.1 DEVELOPMENT OF MODELS, ORGANIZATION OF DATA COLLECTION AND DATA

COLLECTION AND PROCESSING

4.2 OUTCOMES OF THE MODELS

5.0 RECOMMENDATIONS AND CONCLUSIONS

5.1 MAINTENANCE OF EXISTING MODELS AND UPDATING OF FIGURES

5.1.1 Maintenance of the Models

5.2 FURTHER DEVELOPMENT OF EXISTING MODELS

5.3 PREPARING MODELS AND ANALYSING PROFITABILITY OF MORE CROPS AND

AGRIFOOD PRODUCTIONS

5.4 CONCLUSIONS AND RECOMMENDATIONS

5.4.1 Building an Integrated Information Platform

5.4.2 Maintaining the Models and Update the Information

5.4.3 Access to Information from External Subjects

5.4.4 Conclusions on the Outcomes of the Models

ANNEX 1: TERMS OF REFERENCE

ANNEX 2: ASSESSMENT OF MODELS AND OF THE DATA COLLECTION,

PROCESSING AND VALIDATION PROCESS

A2.1 OVERALL ASSESSMENT OF THE MODELS AND OF DATA

A2.2 STRUCTURE OF THE MODELS

A2.3 METHODOLOGY OF DATA COLLECTION AND VALIDATION

A2.3.1 The Applied Methodology A2.3.2 Evaluation of the Methodology after its Application to Seven Products

Profitability Analyses

A2.4 ASSESSMENT OF DATA QUALITY

A2.5 SUSTAINABILITY OF THE SYSTEM OF DATA COLLECTION AND DATA PROCESSING

A2.5.1 Consistency between Maintenance and Updating Needs of the Models, Availability of Secondary Data Sources and Resources for Primary Data Collection and Processing

A2.5.2 Technical Sustainability

A2.6 LIMITS AND CRITICAL POINTS OF THE MODELS AND OF THE ANALYSES

A2.6.1 Striking a Balance between Completeness and Complexity of the Models

A2.6.2 Descriptive Capacity of the Models A2.6.3 The Range of Variants and the Flexibility of the Models A2.6.4 Degree of Confidence of the Results and Development of Cost

Classifications A2.6.4 Resources and Conditions to Improve Quantity and Quality of Data and Confidence of Results

A2.7 OPTIONS FOR FURTHER DATA COLLECTION

ii KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED

FINAL REPORT ASSIGNMENT II

ANNEX 3: OUTCOMES OF THE ANALYSES

A3.1 WINTER WHEAT

A3.1.1 Cost Configurations Included in the Analysis A3.1.2 Analysis of Profitability A3.1.3 Estimated Degree of Confidence of the Model A3.1.4 Critical Points A3.1.5 Further Needs of Analysis and Data Validation

A3.2 MAIZE

A3.2.1 Cost Configurations Included in the Analysis A3.2.2 Analysis of Profitability A3.2.3 Estimated Degree of Confidence of the Model A3.2.4 Critical Points A3.2.5 Further Needs of Analysis and Data Validation

A3.3 ALFALFA

A3.3.1 Cost Configurations Included in the Analysis A3.3.2 Analysis of Profitability A3.3.3 Estimated Degree of Confidence of the Model A3.3.4 Critical Points A3.3.5 Further Needs of Analysis and Data Validation

A3.4 GRAPES FOR WINE

A3.4.1 Cost Configurations Included in the Analysis A3.4.2 Analysis of Profitability A3.4.3 Estimated Degree of Confidence of the Model A3.4.4 Critical Points A3.4.5 Further Needs of Analysis and Data Validation

A3.5 APPLES

A3.5.1 Cost Configurations Included in the Analysis A3.5.2 Analysis of Profitability A3.5.3 Estimated Degree of Confidence of the Model A3.5.4 Critical Points A3.5.5 Further Needs of Analysis and Data Validation

A3.6 DAIRY CATTLE PRODUCTION

A3.6.1 Cost Configurations Included in the Analysis A3.6.2 Commercial Milk Production A3.6.3 Semi-Commercial Milk Production

A3.7 SHEEP PRODUCTION

A3.6.1 Cost Configurations Included in the Analysis A3.6.2 Analysis of Profitability A3.6.3 Estimated Degree of Confidence of the Model A3.6.4 Critical Points A3.6.5 Further Needs of Analysis and Data Validation

ANNEX 4: PROPOSED PRIORITIES FOR ADDITIONAL PRODUCTION COST AND

PROFITABILITY ANALYSES

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED iii

FINAL REPORT ASSIGNMENT II

ACRONYMS AND

ABBREVIATIONS

EAU Economic Analysis Unit (later termed the Economic Analysis Department- EAD)*

EC European Commission

EU European Union

EUROSTAT European Union Trade Statistics Database

FADN Farm Accountancy Data Network

HPK Horticultural Promotion in Kosovo project

IEP Information Exchange Protocol

KARDP Kosovo Agriculture and Rural Development Project

MAFRD Ministry of Agricultural, Forestry and Rural Development

MIS Market Information System

NGO Nongovernmental Organization

NOA New Opportunities for Agriculture Program

PPP Plant Protection Products

TOR Terms of Reference

UP University of Prishtina

USAID United States Agency for International Development

* For the purposes of this report and to avoid confusion, the author consistently uses the term

Economic Analysis Unit (EAU) even though this “unit” was renamed a “department” towards the end of 2012.

iv KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED

FINAL REPORT ASSIGNMENT II

EXECUTIVE SUMMARY

BACKGROUND

Within the framework of the gradual expansion of Ministry of Agricultural, Forestry and Rural

Development (MAFRD) activities and services, the Economic Analysis Unit (EAU, later termed the

Economic Analysis Department [EAD]) was established in January 2012, with the mission to perform agriculture related economic researches. A capacity building program for the Unit was designed in

January 2012. As part of this program, short term technical assistance was provided by USAID- New

Opportunities for Agriculture (NOA) Project to build the capacity of EAU in preparing production cost and profitability analyses. For this purpose, two missions were performed in May, 2012 and

September 2012 by an international consultant.

Seven productions were selected to develop different production cost and profitability models and to build EAU team members capacity to shape and adapt the models, to collect process and validate the data and to disseminate the results of the analyses. The selected products included annual crops

(maize and winter wheat), fruit crops (apples and grapes), and forage crops (alfalfa) and animal productions (dairy production and milk-oriented sheep breeding). These activities were selected by

MAFRD because of their importance in Kosovo agriculture and because they are included (except alfalfa) in MAFRD support schemes to farmers.

MAIN FINDINGS OF THE MISSIONS

The main findings of the missions can be summarized as follows:

1. The data already available at the beginning of the technical assistance inputs and used to calculate production costs and profitability were generally better and more abundant than expected, and access to information was easy; however when going beyond basic information there is a steep gap in terms of skills, resources and cooperation between MAFRD services that could hamper the production of more complex information; for some data the range of different estimates was particularly wide and improving the stability of these data proved particularly difficult. In general, the data that appeared weak or unstable by the end of the first mission were only marginally improved by the end of the assignment (second mission), i.e., after the process of verification and validation, it is clear that much more extensive data collection is required in order to obtain better figures . However, no one of these data is really critical for the overall outcome of the relevant model, except in the case of yields of marketable apples

2. Potential users have a limited knowledge of what information is available and have a partially unjustified prejudice regarding the quality of such information. There is much work to be done in training potential users to make best use of available information and to increase awareness about what information is produced where. To improve this situation, internal informative seminars should be periodically organized to streamline the circulation of information inside MAFRD.

3. There is a need to establish cooperation protocols between research institutions and MAFRD.

During the missions some trials were made to establish a more regular cooperation between EAU and research organizations, such as the University of Pristina or the Research Institutes of

MAFRD, focusing on specific tasks (e.g., analyzing the average family work inputs for certain

1 For example, it proved very difficult to reach a consensus on the number of working days necessary for 1 ha of cereal crops, as fractions of working days are required for the each activity. In other cases, such as average yields of fodder crops in normal weather conditions, the range of estimates was very wide, due to the fact that in most cases there is no precise measurement of the harvest.

2 The most common figures do not consider losses, for which there is a large variation of estimates

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED v

FINAL REPORT ASSIGNMENT II

crops). There was always an open attitude towards cooperation, but the difficulties found in agreeing on concrete steps to start informally such cooperation call for the preparation of more formal cooperation protocols, detailing objectives and responsibilities.

4. There is a relative abundance of good quality investment plans, but few data on the actual situation. Most of the previously developed models and available data are relevant to investment plans and to results of trials carried out with external support. As a result, the profitability shown in the available models is higher than in the majority of actual exploitations. There are few data on actual use of inputs and profits.

5. Insufficient efforts are done to circulate available information and to promote the use of standardized parameters for cost analysis and investment planning. Very often the same analyses are independently performed by different subjects using different parameters and information sources, without consulting information already available.

OUTCOMES OF THE PROCESS

The overall result of the technical assistance input was gradually achieved in the two missions performed in May and September in Kosovo, with follow-up activities performed as work from the consultant’s home base. All activities and relevant outcomes were achieved in the framework of teamwork with permanent EADD team members.

During the first part of the input, performed between May (in Kosovo) and June (work from home base), the following results were achieved:

1. The methodology for data collection and validation was defined and agreed and most of the data collection work was performed.

2. Six draft models were developed. The Model for cattle milk production was produced in two variants (commercial and semi-commercial cattle milk production). The first variant was developed during the first mission.

3. The data validation process was initiated.

During the second part of the consultancy, performed between September (in Kosovo) and November

(work from home), the following results were achieved:

1. The draft models were gradually refined and improved

2. The data validation and refining process was continued; by the end of the second mission most of the data were considered sufficiently stable.

The overall results achieved during the two missions combined can be summarized as follows:

1. The seven models foreseen in the ToR were developed; each model includes some variants, showing different production systems (as in the case of commercial and semi-commercial cattle milk production) and/or various cost configurations, showing the impact on a production system of changes made in some parameters, such as irrigation, type of product, or size of the activity.

All models were revised and discussed with EAU team members and with sector experts. By the end of the mission EAD team members acquired the capacity to revise, maintain, adapt and develop the models and a limited capacity to develop new models.

2. A sustainable data collection and processing system was established for the seven models. Tools and procedures for data collection, processing and validation were developed focusing on cost/effectiveness and sustainability, i.e., were developed to obtain a reliable result compatibly with the presently available resources, without relying on further additional inputs. The methodology adopted is using panels of experts and a small sample of farms for each product to collect the data required for the model and to reach a consensus on such data. This methodology has some limits in the accuracy of results, but is consistent with the available resources and skills vi KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED

FINAL REPORT ASSIGNMENT II

for data collection and processing and with the needs of expected users. Tools and procedures were extensively tested with EAU team members. By the end of the missions the EAU team was able to replicate the process and to apply it to other products.

As part of the tools developed for data collection, the “reference tables” were introduced as a self-standing product. These tables provide up-to-date prices of most agriculture inputs, farm machinery, utilities used in agriculture and some ratios that can be used for “rule of thumb” calculations. It is expected that these tables will provide a very useful tool to facilitate and harmonize the work of specialists working with cost and investment analyses.

3. The process of linking EAU with a network of information sources was initiated. Being a newly established Unit, EAU team members had to establish a network of direct connections with other

Ministry Departments, external partners (development agencies and projects, research institutes etc.) and other information providers and sources and to acquire a direct knowledge of the potential and limits of the available information sources. These links should be gradually consolidated into more formal procedures. As part of this effort, it was necessary to create a more regular and streamlined information flow between EAU and other Ministry departments. Also, EAU had to build up its credibility as a reliable source of information.

By the end of the second mission, the EAU team had made some progress in establishing regular flows of information with technical departments of the Ministry and with the Statistical Unit (now part of the same Department of EAASD), but had still a limited outreach in terms of relations with external subjects and limited communication tools and skills.

OUTCOMES OF THE ANALYSES

The outcomes of the production cost and profitability analyses as a result of the joint work of the

EAU team and of the consultant between May and September, 2012 can be summarized as detailed below:

Cereal and fodder crops are profitable. The not irrigated production regime is by far the most common one, but cost and revenues of both irrigated and not irrigated production regimes were calculated and are in all cases positive. The market of fodder crops analyzed (maize for silage and alfalfa) is limited: in most cases the fodder crop is cultivated for the farm’s own consumption in breeding activity; in these cases profitability is figurative, rather than actual, having positive impact on the overall profitability of the farm.

Apple growing is profitable, but less than generally estimated, as revenues are often overestimated. The recovery of investments is reached after seven to eight years, depending on estimate of marketable yields (i.e., after accounting for losses).

Grape growing for wine production is not really profitable. In recent years a combination of high cost of inputs (especially Plant Protection Products) and low prices for grapes generates a contribution margin than 1,000 Euro/ha, which is not sufficient to cover for to cover family work and amortizations.

Cattle milk production shows several critical points, due to low milk yields as compared with feed intakes. This can be due to low quality of feed and/or to limited genetic potential of cows. When all inputs are computed at market price, milk production is generating a loss. When part of animal feed is produced on the farm and replacement heifers are also raised on the farm, profitability is much better. There are also important economies of scale and the difference of profitability between farms with 10 or 20 cows is high.

Sheep milk production is profitable, mostly due to low production costs. Most of the sheep breeders are also processing milk into cheese using traditional premises and methods. In spite of profitability, weak demand for sheep milk cheese is negatively affecting the prospects of the

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED vii

FINAL REPORT ASSIGNMENT II

sector; in fact, some breeders and even larger specialized farms are reconverting their activity toward meat-oriented productions.

CRITICAL POINTS OF THE ANALYSES

The production cost and profitability figures and analyses were drafted on the basis of estimates provided by panels of experts; the overall draft of the models were finalized after submitting them to panel members for feedback; during the validation phase, a small sample of farmers was also consulted to identify possible major discrepancies between the estimates provided by experts and by farmers. When major discrepancies emerged, more specific research was carried out with a larger number of farmers. The farmers are chosen among those considered by local MAFRD officers as representative of the average exploitation.

This methodology is different from that used for similar activities in other countries, which is based on larger stratified samples of farms. The chosen methodology has the advantage of being faster and requiring less resources, at the cost of providing a range of data, rather than a figure obtained following a rigorous statistical analysis; however, since most of the farms in Kosovo do not keep a complete accounting system, the data possibly provided by farms would have problems with reliability, consistency and bias.

The models are conceived as a work in progress rather than as an ultimate tool and they are expected to evolve and grow in complexity in parallel with the growth of the capacity of the EAU team as information provider and with the sophistication of the users’ needs. As a consequence, the structure of the models is rather basic and it should be appropriate to add many more functions, such as a scenario-building capability and a wider range of pre-set presentations of results.

Another critical aspect is related to some of the data included in the models, such as the estimates of family work inputs and the comparison between cost of hired services (e.g., plowing a field) and costs of the same service performed by the farmer using his/her own machinery. This was the first time in which an organic effort to estimate these values was performed in Kosovo and the quality of estimates is still in need of improvement.

RECOMMENDATIONS

The recommendations emerging from the mission findings and outcomes can be summarized as follows:

1. Models and data should be circulated as a draft as soon as possible to MAFRD technical officers and to external partners, to get feedback and suggestions for improvement. At the present stage it is more important to raise awareness and interest on production cost and profitability analysis, rather than trying to come out with a final product. In addition, the reference tables should be also published in the website as stand-alone information.

2. The deepening of the existing models (further improvement in quality of data, introduction of additional functions and presentations) should proceed in parallel with the widening of the range of products covered. EAU team members should build up in parallel their capacity in model-making and in data collection and processing.

3. An integrated MAFRD information platform should be gradually built up, replacing the present system made by separate and autonomous sets of information, often informally collected and processed by separate departments or by development projects supporting such departments.

Also, existing information is not fully exploited. In general, MAFRD officers find it difficult to know what data are available and where.

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED 1

FINAL REPORT ASSIGNMENT II

1.0 INTRODUCTION

Due to the importance of agriculture for Kosovo development, in particular, in terms of employment

(about 26 percent of active population ) and principal or subsidiary contribution to family budgets

(about 90 percent of population own some land and about 50 percent of rural population has some livestock ), increasing attention and resources are being addressed to agriculture and rural development, also as a tool to contrast the increasing international trade deficit recorded in food products.

The Ministry of Agriculture, Forestry and Rural Development (MAFRD) of Kosovo, having the responsibility to regulate and support the agrifood sector and rural development is presently in the process of restructuring and expanding its services, pursuing the Agriculture and Rural Development

Plan 2007 - 2013 (as revised in 2009 and 2010) and the Action Plan for the Economic Vision of

Kosovo 2011 – 2014. As part of these activities, the Economic Analysis Unit (EAU) was established in January 2012, with the mission to perform agriculture related economic research. The EAU was established as a staff unit, reporting directly to the Secretary General.

In the second half of 2012 EAU was put under a newly established Department, having responsibility over the EAU and over the Agriculture Statistics Division, which was previously part of a different

Department.

Due to its key role in the production of information, the Department is the beneficiary of several initiatives implemented by development projects and agencies and is directly or indirectly involved in most of the ongoing Ministry activities aimed at improving the range and quality of services provided by MAFRD to farmers and other support measures for the rural sector.

In the second half of 2012, support has been provided by EC-Taiex and EC-Twinning programs to the

Statistical Division (for improvement of FADN and MIS) and by USAID-NOA Project to the EAU.

The support provided to EAU aimed to build the capacity of the unit in the analysis of profitability of enterprises, including production cost analysis, gross margin analysis and prevalent cost analysis. The capacity building process was initiated in May 2012 and included the provision of technical assistance inputs for total of 60 man/days between May and October 2012. Two missions were performed by an international consultant in May and September 2012. The report on the second field mission and on the outcomes of the overall technical assistance input are provided in this document.

3 There are different calculations of employment in agriculture The World Bank reports for 2005-2006 a share of 26% (Kosovo Poverty

Assessment 2007). For the same years, the European Commission Kosovo Report 2006 indicates a share of 49%, including part-time workers, corresponding to 26% of active population in terms of full employment. Other sources, such as USAID (Kosovo Economic

Performance Assessment 2008, CIA Factbook) indicate a smaller share, in the range of 23% to 25%

4 The World Bank Country Economic Memorandum, Chapter 6, reporting data from WB Kosovo Poverty Assessment 2007.

2 KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED

FINAL REPORT ASSIGNMENT II

2.0 OBJECTIVES OF THE

MISSION

The overall objective of the short term technical assistance was to build the capacity of the EAU and jointly develop appropriate and sustainable models for production cost and gross margin calculation and analysis of profitability of seven products, produced at commercial and semi-commercial level, namely: i) winter wheat; ii) maize (grain production and silage production); iii) alfalfa; iv) grapes for wine and table grapes; v) apples; vi) milk production and; vii) sheep breeding.

The analyses are expected to be mainly used for the following purposes: i) policy-making, including design and adaptation of economic support programs; ii) assessment of policy implementation measures; iii) tools for extension services. However, they will be made available for the needs of all

MAFRD offices and of external partners and stakeholders.

The second mission had the following expected results:

1. Complete the cost calculation models for apple and milk production; for this purpose, discuss and agree with EAU team and panel members on the structure of the model, the detail of costs to be included and the organization of inter-linked excel tables.

2. Refine the models (both the completed ones for grain crops and grapes and those to be completed), working on the most complex cost components, such as:

a. Endowment of equipment of the “typical” (i.e., representative) farm (to calculate more accurately investments and depreciation) and;

b. Average number of work/days used for each crop.

3. Validate the seven cost configurations whose analysis started in May

4. Perform the first step of a cost analysis (prior to validation) of four additional crops and animal productions, namely: i) goat breeding and relevant milk and meat production; ii) open field vegetables (primarily tomatoes and peppers, but other crops can be added, as production cost analyses are quite similar); iii) vegetables production in protected environment (greenhouse production).

5. Keep up-to-date the input costs tables (so called “reference tables”), integrating it according to the needs arising for the work performed during the second mission.

6. Assist in defining and producing contents, modalities and procedures for communication and exchange of information between EAU and development partners. In particular, cooperate with

EAU in building up its website (or section of website within MAFRD website).

7. Consolidate the sustainability of achievements, building experience and self-confidence of EAU team, gradually transferring to them the responsibility of data collection and processing, model updating, and presentation of results. The capacity to develop new models will be also partially built up.

As part of the capacity building process, the development of procedures and the definition of responsibilities for data collection and processing and for dissemination and access to information

(Information Exchange Protocols – IEP) were also foreseen.

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED 3

FINAL REPORT ASSIGNMENT II

3.0 ORGANIZATION OF

ACTIVITIES

The second consulting mission was performed in the period from September 2–28, 2012. The main activities carried out to achieve the mission’s expected results and the objective of this technical assistance were the following:

1. Complete the two milk production (dairy cattle and sheep) and the apple growing models.

2. Refine and check all the models.

3. Carry out the validation of data included in the models, assessing quality of results and refining the quality of models and data. This activity included extensive review of models with Ministry technical specialists and external experts and a couple of field visits, devoted to milk production.

This activity absorbed more resources than expected and eventually became the focus of the whole mission.

4. Define priorities for development and improvement of existing models and development of new models and perform preliminary activities for new models. In the draft work plan this activity was wider, including the full development of three more models. Based on comments on the initial draft report,, it was concluded that the feasibility of this option was to be discussed in the briefing, at the beginning of the mission. It was eventually decided to devote a much reduced time to this activity.

5. Define and introduce in the models cost configuration variants, based on findings and needs emerging from the data validation phase.

6. Analyze the most feasible and sustainable options for regular updating and improvement of the models; perform trials to test the options.

7. Interact with other EAU team members to build up an autonomous capacity to develop and maintain the models and to collect and validate the data feeding the models.

8. Develop a pattern for the main Information Exchange Protocols (IEP).

All activities were carried out through EAU teamwork, with the consultant acting, for the duration of the mission, as an EAD team member.

The responsibility for data collection and validation and for adaptation of the models was gradually transferred to the EAD team members.

By the end of the second mission, the EAU team had achieved an almost completely autonomous capacity to collect the data necessary to feed models, a full capacity to revise the models for possible errors and editing, i.e., a full understanding of the models’ logic, a complete knowledge of the formulas and linkages that make the models and an understanding of the technical and economic rationale behind each formula and presentation. The EAD had also achieved a certain capacity in adapting and developing the models themselves and in developing new models.

The planned and actual mission schedules are provided in Diagram 3.1 below. As can be seen in the diagram, the organization of actual activities differed from initial planning.

4 KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED

FINAL REPORT ASSIGNMENT II

Diagram 3.1: Planned and Actual Activities

Activities 1 st week 2 nd week 3 rd week 4 th week

Complete three old models

1 Refining old seven models

2 Validate cost configuration old seven models

3 Develop three new production cost models

4 Data collection for additional cost configurations *

5 Update; analysis of data sources for existing models

6 EAU capacity building in modeling and data process

7 Development of IEP

Milestones B B C

W D B

B D B

Legend:

Planned activities Actual activities

B: Briefing with team and EAU; BC: Presentation of basic cost tables; DB; Debriefing with team and EAU;

Note: * Three different cost configurations of previously existing models were developed

The changes introduced and the relevant reasons can be summarized as follows:

1. More time than foreseen was devoted to finalizing the milk production models. The model for larger scale (commercial) milk production from dairy cattle was expanded to include more presentations as compared with the one developed by the end of the first mission, while the models for smaller (semi-commercial) milk production from dairy cattle and sheep milk production were substantially reviewed; in particular, the last model was modified to include the production of cheese at farm level and to include self-consumption patterns.

2. Less time was devoted to refining the existing models. The other models (for winter wheat, maize, alfalfa, grapes and apple growing) during the validation phase proved sufficiently complete, requiring only a few changes, so that most of the effort was devoted to validating the data and developing new cost configurations, rather than to further development of the models.

3. Validation of cost configurations proved to be the most complex and time-consuming activity, taking more time than expected. Data were revised with Ministry sector specialists and external experts from USAID-NOA project and IADK. In some cases, data turned out to be less stable than expected (i.e., the range of the best estimate given by different experts was wider than expected), while in other cases data were stable and the structure of models was considered

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED 5

FINAL REPORT ASSIGNMENT II

reliable by experts, but total profitability was lower than generally assumed (e.g., milk production from dairy cattle is generally considered profitable, but once all expenses are accounted for, the models show a loss or a very small margin). In all these cases an iterative process of data and model refining was applied, thus requiring repeated consultations with sector experts and specialists.

4. Developing new models. It was anticipated that we would develop new models for vegetables production in open field and protected environments. This activity was not foreseen in the initial

ToR, which required only the development of seven models (which were mostly developed during the first mission). The time devoted to this activity was substantially reduced, preferring to focus on developing an autonomous capacity of EAU to adapt, modify and develop new models. As a result, only the preliminary activities (preparation of check lists and identification of the most similar model to be adapted) were completed, but such activity was completely performed by the

EAD team.

5. Identifying and testing data sources and procedures to optimize the process of updating the models and the reference tables. The time devoted to these activities corresponded to the plans.

This group of activities included the in-depth screening of different data sources. The conclusions of this activity are detailed in Section 5.0 below. In particular, the following possible sources were screened:

Data from FADN were analyzed. The exercise included the following: i) identification of the data which are regularly sourced directly from farmers and those estimated or calculated based on data of previous years, ii) evaluation of the significance of the sample used for FADN and of the strata included in the sample, to assess which data subsets can be used to analyze a specific cluster, such as farms located in a certain municipality or farms specialized in a certain production and, iii) assessment of the possibility of processing automatically the data (this was possible only after exporting data from MS Excel to formats suitable for database software).

FADN dataset is important because, in spite of its limits (small sample, reduced range of primary data, problems with data validation, focus on farm rather than on crops), it is the most comprehensive primary dataset yearly collected in the whole country and the only source for some information, such as the availability of farm machinery at farm level. FADN will also be gradually expanded and improved , so that it will become an increasingly important source of information for cost production and profitability analyses. The importance of FADN explains the relatively large amount of time devoted to analyzing and discussing with the Ministry staff, and the technical assistance to FADN regarding existing data, the formulas used to generate part of the data included in FADN database and the method used to update the inventory of farm machinery owned by farms.

The possibility to use the baseline survey financed by the Kosovo Agriculture and Rural

Development Project (KARDP) was evaluated. The survey will be carried out for three consecutive years starting from 2013. It can be used to complement data from FADN and to collect additional data that cannot be obtained from other sources; the possibility to adapt the questionnaire taking into consideration EAU needs was agreed.

The data collected by the Payment Department were considered. The department has a large and reliable base of data relevant to the applicants and beneficiaries of the State support schemes. Some of this information can be used in production cost and profitability analyses, 5 A Pilot FADN was established with a test phase involving 50 farms in 2004. This was expanded to 159 farms in 2005 and to 350 in

2010. The request by EU is to further increase the size of the sample, reaching 1,000 units; also the range of primary data collected every years and that one of data calculated or extrapolated will be revised

6 Technical assistance is provided by EC- TAIEX.

7 The Department is in charge to evaluate the applications submitted by farmers to the State Support Schemes, to disburse the amounts which successful applicants are eligible for and to control that the ex-ante and ex-post conditions attached to the schemes are satisfied.

6 KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED

FINAL REPORT ASSIGNMENT II

such as milk yields (as the milk support scheme is related to production) and average size and structure of cereal farms (as the support is given by hectare; having this figure, the calculation of amortization on owned equipment is more accurate). These data can be used and processed in full respect of privacy concerns and can also provide more detailed information at the commune level.

About three working days were devoted to evaluate the possible use of FADN data set.

Less time (about one working day) was devoted to facilitating coordination between EAD and the

KARDP baseline survey and to discussing the use of data from the Payment Department. In the case of the baseline survey, it was easy to reach an agreement on the involvement of EAD in the preparation of the questionnaire, while in the case of the Payment Department the nature and possible use of available data was rather clear since the beginning.

Three more working days were devoted to updating data in previously developed models and reference tables.

6. Building EAU capacity in reviewing, updating and adapting existing models and developing new models. A major share of available resources was devoted to revising in detail the models with the

EAU team, discussing the rationale of the models, the kind of parameters used, the formulas and links embedded in the model, the structure of presentations in function of different users and the different classifications of costs. Critical data and information (i.e., those with highest impact on final results) were identified for each production, while building some practical experience in consolidating data from different sources and evaluating relevant reliability.

7. Development of Information Exchange Protocols (IEP). This task was delayed and eventually not performed, as priority was given to finalizing the models, validating the figures in the models and identifying the sources of data that will minimize future work in data collection and processing.

However, IEP are basically procedures for collecting and communicating information, which will become consolidated with practice, so that, at the present stage, only an example of IEP could be developed, which is done in Annex 5 below.

Milestones. The workshops foreseen in the Terms of Reference were performed during the first mission. The possibility to carry out a debriefing extended to the technical departments of the

Ministry and with external subjects was discussed with Ministry decision makers, but the conclusion was that it was preferable that an internal presentation be made by the EAU team to the Ministry

Board of Directors, after the completion of the mission. The briefing and the debriefing were regularly carried out.

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED 7

FINAL REPORT ASSIGNMENT II

4.0 OUTCOMES OF THE

MISSION

4.1 DEVELOPMENT OF MODELS, ORGANIZATION OF DATA

COLLECTION AND DATA COLLECTION AND PROCESSING

By the end of the mission an overall capacity of the EAD to manage, maintain and develop cost production and profitability analyses was built, with some limits in the areas of data validation, development of models and dissemination of information. The skills developed by EAD team members include the production of cost and profitability models, the collection and processing of data needed in the models, and the process of validation of models and information. In particular, the following specific results were accomplished:

1. Eight production cost and profitability analysis models were developed, relevant to the products foreseen by the ToR. The models included: i) winter wheat, ii) maize in grains and for silage, iii) alfalfa production, iv) grape growing for wine production, v) apple growing, vi) intensive

(commercial) cattle milk production, vii) semi-extensive (semi-commercial) cattle milk production and; viii) sheep milk production. The contents and outcomes of each model are detailed and discussed in Annex 3.

In most cases different cost configurations were developed for each model (e.g., irrigated and non-irrigated maize production or breeding activities of different size); each model includes an investment costs analysis (including also amortizations), a current cost analysis and an overall analysis of profitability. The outcomes of the models are provided in different pre-set presentations, tailored to different types of users (more detailed for technicians, more synthetic for decision-makers).

Whenever necessary, as in the case of cattle milk production, the outcomes of the models are also presented in different cost classifications, to emphasize specific aspects of the business (e.g., contribution margin, difference between cash flows and overall profitability at market prices, etc.).

All the models are developed in MS Excel, can be exported and modified in open source software

(e.g., Open Office) and are portable, i.e., can be transferred from a computer to another without particular problems.

2. Data sources and data collection and validation procedures were established and EAD team acquired an autonomous capacity to collect and process data. This result included the creation of

“reference tables,” i.e., tables which provide an up-to-date catalogue of the cost of the most common agriculture inputs, farm machinery and services for agriculture , the preparation of check lists for conducting the panel meetings and interviews with farmers , the use of available data from FADN; MIS (Market Information System) and other Ministry sources and the acquisition of

8 The reference tables indicate also the evolution of prices in different months of the year, the fiscal amortization period of fixed assets

(i.e., duration of the amortization period in accordance to Kosovo fiscal laws) and the cost of the most common utilities and general purpose inputs (water, electricity, fuel, agricultural work etc.)

9 Interview guidelines and check lists are alternatives to questionnaires. As compared with these last, check lists are result-oriented, rather than process-oriented, resulting much more flexible, but requiring much more skilled interviewers, who are usually the specialists who will later process the data. A consequence of the use of check lists instead of questionnaires is that the number of interviews is usually smaller (as there are few skilled interviewers).

8 KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED

FINAL REPORT ASSIGNMENT II

practical experience in conducting interviews with farmers, reviewing models and data with sector experts and consolidating the information obtained from different sources.

3. Most of the data in the models have been validated. All the data that could be validated with available sources (secondary data, reports, interviews with local experts and selected farmers) within the given time were screened and evaluated. Some data (such as use of family work and farms’ availability of farm machinery) are not available and will require specific and more in-depth research. Among the most uncertain data are those pertaining to marketable yields (in particular for apples, maize for silage and alfalfa), i.e., yields that are actually sold or consumed, after deducting losses.

The estimate of the consultant is that outcomes of the profitability analyses (profit or loss) should be considered basically correct, but should be correctly read, too: in fact many models show a loss when all factors are calculated at market prices, i.e., calculating all inputs as purchased on the market. When self-produced inputs are calculated at production cost (e.g., part of animal feed used in cattle breeding), or family work is calculated as cost/opportunity (i.e., what alternative incomes are lost when working time is devoted to agriculture production) the final results are much better. When appropriate, different presentations are pre-set to show different cost configurations and classifications.

4. The sustainability of the whole process has been achieved. The methodology for data collection, processing and validation and the models has been extensively reviewed with the EAU team members and increasingly difficult and independent tasks have been assigned to each member. By the end of the second mission, each EAU team member was able to collect and process the data, to lead the process of data validation and to review most of the models. A certain capacity to develop new models was also achieved. Also, each EAU member was also becoming specialized in a group of products (e.g., animal production), in up-dating the reference tables and in finalizing the presentations for the potential users.

The above should be considered a remarkable achievement of the EAU team members, as during the same period the team members had also to carry out other tasks

, and to develop the skills required for these other activities.

Some efforts were devoted to increase awareness within the MAFRD about the main issues relevant to information in agriculture, with the objective to improve the use of already available information and to arrive at a shared understanding among producers and users of information on the range and quality of data and information that is available via which sources and means. In fact, the overall process of data collection and validation and model refinement was a good occasion to demonstrate the need and advantages of establishing regular and more formalized exchanges of information among the different Ministry departments; the research of secondary data collection was also an opportunity to evaluate the quantity and quality of the already available information, which turned out to be more abundant and of better quality than initially expected and to understand the different uses for such information.

At present, many units and departments are autonomously collecting and processing the information they need, with frequent duplications and overlapping. The efforts of EAU to gather and interlink the available sources of information showed the advantages – and the difficulty – to change this approach.

The experience of creating the production cost models showed the importance of substantially improving the circulation of available information in the Ministry and also the need for training not only the information providers, but also the users of information, including Ministry decision-makers.

10 During the period from May to September, the AU has also provided support to other Departments in preparing sector analyses, in introducing farm accounting, participating to the establishment of Ministry extension services.

KOSOVO NOA: TECHNICAL ASSISTANCE FOR THE ANALYSIS OF PROFITABILITY OF SELECTED 9

FINAL REPORT ASSIGNMENT II

It is too early to evaluate the impact of such efforts; however, improving the flow of information inside the Ministry and between the Ministry and its external partners is a long term objective that should be constantly pursued.

4.2 OUTCOMES OF THE MODELS

According to the data collected and to the modeling in their present form, the outcome of the production cost and profitability analyses can be summarized as follows:

1. Cereal crops are profitable under present price system. This statement is valid for winter wheat, maize in grains and maize for silage. Data on labor input have been confirmed in the validation phase, but are probably underestimated, as they are lower than in most production cost analyses performed in other countries of the region and in EU member states.