PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

NOV 3 C 2003

 

120103

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p. 15

7* ————_—— __—._.

ECONOMIC EVALUATION OF BEAN—RESEARCH
INVESTMENT IN MEXICO

By

Horacio Gonzalez-Ramirez

A DISSERTATION
Submitted to
Michigan State University
in partial ﬁrlﬁllment of the requirements
for the degree of
DOCTOR OF PI-HLOSOPHY

Department of Agricultural Economics

2003

ABSTRACT
ECONOMIC EVALUATION OF BEAN-RESEARCH INVESTMENT IN MEXICO
By

Horacio Gonzalez-Ramirez

In Mexico, dry beans are the second most important crop after maize, both in
terms of production and consumption. Approximately 85% of the country’s bean crop is
grown under rainfed conditions. During the 1990s, the total harvested bean area
decreased by 2.0%, average yield decreased by 2.5%, and production declined by 4.5%.
As a result, there has been an increasing trend in requiring bean imports to meet domestic
demand, especially aﬁer Mexico joined NAFTA in 1994. During the 19905 the Mexican
government, through the Secretariat of Agriculture (SAGAR), started two programs——
PROCAMPO and Alliance for the Countryside—to support farmers and promote the
adoption of improved varieties through the Kilo per Kilo program.

In 1982, the Bean/Cowpea Collaborative Research Support Program (CRSP)
signed an agreement with Mexico's National Research Institute for Forestry, Agriculture
and Livestock (INIFAP) to collaborate in developing improved bean varieties for the
semiarid highlands of Mexico’s North-Central region. During 1990-2000, INIFAP
released several improved bean varieties that were distributed via the Kilo per Kilo
program and adopted by farmers in the semiarid region.

The economic impact assessment of bean research is essential to provide

decision-makers with information to improve the allocation of research resources.

The objectives of the study are to describe Mexico's bean subsector,
analyze the factors associated with adoption of the improved bean varieties released by
INIFAP in the 19903, identify factors that contributed to explaining the participation of
farmers in the government’s seed distribution program (Kilo per Kilo), and estimate the
net social gains generated by public investment in agricultural research and extension to
develop and distribute improved bean varieties in northern Mexico.

The study area includes the states of Chihuahua, Durango, and Zacatecas, which
account for 62% of the Mexico’s rainfed bean production area. The results reported in
this study include a rapid appraisal assessment of the bean subsector, an evaluation of
government support policies affecting the bean subsector, a statistical and econometric
analysis of improved bean seed adoption and farmer participation in the Kilo per Kilo
program (based on survey data), and an estimation of the economic returns to public
investment in bean research and extension (using the economic surplus method).

The adoption analysis indicates that the improved bean varieties Pinto Villa and
Pinto Mestizo have been widely adopted in Chihuahua and Durango, that these varieties
have yields that are 20.6% higher that traditional pinto bean varieties and that the yield
diﬁ‘erence is statistically signiﬁcant. The economic analysis indicates that if a closed
economy model is assumed, the ﬁnancial and economic NPVs are positive and the [RRs
are 17.5 and 21.4 %, respectively. If an open economy model is assumed, the ﬁnancial
and economic NPVs are positive and the 11le are 21.3 and 20.7 %, respectively. The
results from both models are consistent and suggest that public investment in bean
research and extension is proﬁtable (opportunity cost of capital=10%). Thus, the

government should continue investing in bean research in northern Mexico.

DEDICATION

This dissertation is dedicated to:
My parents
Alvaro Gonzalez-Franco and Estela Ramirez-Echevarria
My grandmother
Flor de Maria Echevarria P. de Ramirez
My dear aunt
Maria del Consuelo Ramirez-Echevarria
Patricia and our three sons
Leon David, Horacio Alejandro, and Edgar Orlando.

My brother, sister, and friends.

iv

ACKNOWLEDGEMENTS

I express my sincere gratitude to Dr. Richard Bemsten, my major professor for his
attentive, supportive, and critically constructive comments and suggestions during the
dissertation process. I also thank the other members of my guidance committee: Drs. Eric
Crawford, Stephen Harsh, James Kelly, James Oehmke, John Staatz, and Irvin Widders,
for their timely and valuable comments.

This research would have been virtually impossible without the logistic and
ﬁnancial support from the Instituto Nacional de Investigaciones Forestales, Agricolas y
Pecuarias (INIFAP) and their directors—especially Dr. Jesus Moncada de la F., Dr.
Ramon Martinez R, Dr. Sebastian Acosta N., Dr. Rodrigo Aveldaﬁo 8., Dr. Homero
Salinas G., and MC. Arnulfo Pajarito R.—and the Consejo Nacional de Ciencia y
Tecnologia (CONACYT) en Mexico for awarding me a graduate training fellowship. In
addition, I greatly appreciate the assistance and suggestions provided by my colleagues at
INIFAP, especially Jorge Acosta G., Francisco J. Ibarra P., Evenor I. Cuellar R., Adan
Castillo R., Mario Valdes 6., Jesus Lopez H., Rigoberto Rosales S., Pablo Fernandez H.,
Hector Perez T., Efrain Acosta D., Benito Cazares E., and Roman Zandate H.

I also thank thee Bean/Cowpea Collaborative Research Support Program (B/C
CRSP) for its ﬁnancial support of my ﬁeld work in Mexico (through USAID grant #
DAN-G-SS-86-00008-OO), the support of the Department of Agricultural Economics at
Michigan State University (MSU) for providing funding for me to ﬁnish my doctor of

philosophy program, and the support of the Comunidad Latina for my family.

Special thanks for support and friendship of students of the Department of
Agricultural Economics at MSU. I am especially grateﬁil to my study group during most
part of the program: Yoko Kijima, Patricia Masego Makepe, Asfaw Negassa Muleta, and
Edward Pepukayi Mazhangara. Also for the friendship and advise of David Mather,
Youssouf Camara, Denise Mainville, and Juan Estrada-Valle.

In Durango, Mexico, special thanks to the staff of the Secretaria de Agricultura,
Ganaderia, Desarrollo Rural, Pesca y Alimentacion (SAGARPA) of the Federal
Government, the Secretaria de Agricultura, Ganaderia y Desarrollo Rural (SAGDR) of
the Government of the State of Durango, and of the Instituto Nacional de Estadistica,
Geograﬁa e Informatica (INEGI), especially to Felipe Aviles L.

Most important, I haven’t ﬁnished the dissertation without God’s help.

vi

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. xi
LIST OF FIGURES ............................................................................... xiv
CHAPTER 1. INTRODUCTION .................................................................................... 1
1.1 General Background .................................................................................................. 1
1.2 Problem Statement ..................................................................................................... 3
1.3 Justiﬁcation of the Study ........................................................................................... 4
1.4 Research Objectives .................................................................................................. 5
1.5 Dissertation Organization .......................................................................................... 7
CHAPTER 2. LITERATURE REVIEW AND METHODS ........................................... 10
2.] Literature Review .................................................................................................... 10
2.1.1 Bean Subsector Overview ............................................................................... 10
2.1.2 Technology Adoption Studies ......................................................................... 12
2.1.3 Economic Impact Analysis ............................................................................. 14
2.2 Methods of Analysis ................................................................................................ 17
2.2.1 The Area of Study .......................................................................................... 17
2.2.2 Rapid Appraisal .............................................................................................. 20
2.2.3 Farmers’ Survey ............................................................................................. 21
CHAPTER 3. AN OVERVIEW OF THE MEXICAN BEAN SUBSECTOR ................. 26
3.1 Characteristics of Supply ......................................................................................... 26
3.1.1 Area, Production, and Yields .......................................................................... 27
3.1.2 Area, Production, and Yields by Season ......................................................... 29
3.1.3 Production by Water Condition ...................................................................... 31
3.1.4 Impact of Production Seasonality on Bean Marketing ..................................... 32
3.1.5 Bean Imports and Exports .............................................................................. 33
3.2 Characteristics of Demand ....................................................................................... 37
3.2.1 Changes in Preferences ................................................................................... 37
3.2.2 Apparent Consumption ................................................................................... 38
3.2.3 Bean Consumption Per Capita ........................................................................ 41
3.2.4 Consumption by Market Class ........................................................................ 44
3 .3 Characteristics of Prices ........................................................................................... 48
3.3.1 Real Bean Prices ............................................................................................ 48
3.3.2 Wholesale Price Seasonality ........................................................................... 50
3.3.3 International Prices ......................................................................................... 52
3.3.4 Marketing Margins ......................................................................................... 53
3.4 Bean Marketing System ........................................................................................... 54
3.4.1 Bean Marketing System ................................................................................. 55
3.4.2 Bean Marketing Channels ............................................................................... 56
3.4.3 Bean Processing Industry ............................................................................... 59

vii

3.5 Summary of the Chapter .......................................................................................... 62

CHAPTER 4. GOVERNMENT POLICIES AND BEAN SUPPORT PROGMM ........ 66
4.1 Government Agricultural Policies ............................................................................ 66
4.1.1 Support Programs ........................................................................................... 66
4.1.2 Price Policy .................................................................................................... 68
4.1.3 Trade Agreements .......................................................................................... 70
4.1.4 Farm Credit .................................................................................................... 73
4.1.5 Crop Insurance ............................................................................................... 74
4.2 Research, Technical Assistance, and Seed Production .............................................. 75
4.2.] Research and Technology Transfer ................................................................. 75
4.2.2 The National Research Institute for Forestry, Livestock, and Agriculture ....... 76
4.2.3 Technical Assistance Program ........................................................................ 78
4.2.4 The National Service for Seed Inspection and Certiﬁcation ............................ 79
4.2.5 The National Producer of Seeds ...................................................................... 80
4.3 Seed Distribution in the North of Mexico ................................................................ 83
4.3.1 National Program for Seed Distribution .......................................................... 83
4.3.2 Seed Distribution in Chihuahua ...................................................................... 85
4.3.3 Seed Distribution in Durango ......................................................................... 89
4.3.4 Seed Distribution in Zacatecas ........................................................................ 94
4.4 Summary of the Chapter ........................................................................................ 100
CHAPTER 5. IMPROVED VARIETIES ADOPTION ANALYSIS ............................ 104
5.1 Characteristics of Farmers and Farms in Northern Mexico ..................................... 104
5.1.1 Characteristics of the Farmers ...................................................................... 104
5.1.2 Characteristics of the Farms ......................................................................... 106
5.1.3 Sources of Technical Information ................................................................. 107
5.2 Crop Practices, Production Costs, and Capital ........................................................ 108
5.2.1 Crop Practices and Chemical Inputs ............................................................. 108
5.2.2 Production Costs, Credit, and Equipment ..................................................... 111
5.2.3 Family Labor and Hired Labor ..................................................................... 112
5.3 Program Participation, Storage and Prices of Beans ............................................... 114
5.3.1 Participation in Government Programs ......................................................... 114
5.3.2 Storage of Beans and Farm Location ............................................................ 115
5.3.3 Selling Prices of Beans by Market Class ....................................................... 116
5.4 Area Planted to Improved Varieties and Yields ...................................................... 120
5.4.1 Area Planted to Improved Varieties .............................................................. 120
5.4.2 Yields of Traditional and Improved Varieties ............................................... 123
5.4.3 Adopters’ Perception of Improved Varieties Characteristics ......................... 124
5.5 Econometric Analysis of Program Participation ..................................................... 125
5.5.1 Adoption of [VS and Program Participation .................................................. 126
5.5.2 The Model for the Econometric Analysis ...................................................... 127
5.5.3 Description of the Variables ......................................................................... 129
5.5.4 Results of the Analysis ................................................................................. 133
5.6 Summary of the Chapter ........................................................................................ 135

viii

CHAPTER 6. BEAN RESEARCH ECONOMIC ANALYSIS .................................... 139

6.1 Data for the Analysis ............................................................................................. 140
6.1.1 Area, Production and Yields ......................................................................... 140
6.1.2 Value of Production and Prices ..................................................................... 142
6.1.3 INIFAP and B/C CRSP Research Costs ........................................................ 143
6.1.4 Extension and Adoption Costs ...................................................................... 144
6.1.5 Consumer Price Index and Exchange Rate .................................................... 145

6.2 Basis of the Economic Analysis ............................................................................. 146
6.2.1 Supply and Demand Elasticities ................................................................... 146
6.2.2 Supply and Demand Functional Forms ......................................................... 147
6.2.3 Research-Induced Supply Shift ..................................................................... 148
6.2.4 The Logistic Adoption Curve ....................................................................... 148
6.2.5 Economic Surplus Formulas for a Closed Economy ..................................... 149
6.2.6 Net Present Value and Rate of Return ........................................................... 150

6.3 Financial Analysis ................................................................................................. 151
6.3.] Financial Cost Stream .................................................................................. 151
6.3.2 Financial Beneﬁt Stream .............................................................................. 152
6.3.3 Financial Rate of Return ............................................................................... 153

6.4 Economic Analysis ................................................................................................ 155
6.4.1 Economic Cost Stream ................................................................................. 156
6.4.2 Economic Beneﬁt Stream ............................................................................. 156
6.4.3 Economic Rate of Return ............................................................................. 157

6.5 Sensitivity Analysis ............................................................................................... 159
6.5.1 Changes in Adoption and Yields .................................................................. 159
6.5.2 Changes in Prices and Costs ......................................................................... 160
6.5.3 Changes in Elasticities .................................................................................. 162

6.6 The Small Open Economy Model .......................................................................... 165
6.6.1 Financial Analysis ........................................................................................ 165
6.6.2 Economic Analysis ....................................................................................... 167
6.6.3 Closed versus Small Open Economy ............................................................ 168

6.7 Summary of the Chapter ........................................................................................ 173

CHAPTER 7. SUMMARY AND CONCLUSIONS .................................................... 180

7.1 Summary of Findings ............................................................................................ 180
7.1.1 Challenges Facing the Bean Subsector ......................................................... 180
7.1.2 Government Intervention to Increase Productivity ........................................ 182
7.1.3 Farmers’ Adoption of Improved Bean Varieties ........................................... 184
7.1.4 Econometric Analysis of Program Participation ............................................ 185
7.1.5 Returns to Bean-Research Investment .......................................................... 186
7.1.6 Beneﬁts for Consumers and Producers ......................................................... 192

7.2 Policy Implications ................................................................................................ 193
7.2.1 Implications for Government Decision-Makers ............................................ 193
7.2.2 Implications for Bean Researchers ................................................................ 194
7.2.3 Implications for Extension Services .............................................................. 195

7.3 Limitations of the Study ........................................................................................ 196

7.4 Future Research Needs .......................................................................................... 197

ix

APPENDICES ............................................................................................................ 1 99

Appendix A: Farmer Questionnaires in Spanish and English ....................................... 200
Appendix B: Table 1. Bean area, Production and Yield in Mexico, 1990-2000. ........... 214
Appendix C: Table 2. Apparent Consumption of Beans in Mexico, 1980-2000 ............ 215
Appendix D: Table 3. Real Producer Prices for Beans in Mexico, 1990-2000. ............. 216
Appendix E: ARTTP Regulations ................................................................................ 217
Appendix F: Strategy of INIFAP ................................................................................. 218
Appendix G: Requirements for Eligibility of PEAT ..................................................... 219
Appendix H: Requirements to Apply for a Breeder’s Patent ........................................ 220
REFERENCES ............................................................................................................ 221

LIST OF TABLES

Table 2.1 Characteristics of the Improved Bean Varieties Developed by INIFAP for the

Rainfed Conditions of Northern Mexico .................................................................. 18
Table 2.2 Number of Face-to—Face Interviews in Durango, Mexico 2001 ....................... 23
Table 3.1 Area, Yield and Production of Beans by Season in Mexico, 1990-2000 .......... 30
Table 3.2 Imports, Exports, and Net Imports of Dry Beans in Mexico, 1991-2000 ......... 36
Table 3.3 Bean Consumption Per Capita in Mexico, 1990-2000. ................................... 44

Table 3.4 Preferences for Bean Commercial Classes by Region in Mexico, 1993-1994. 45
Table 3.5 Annual Apparent Consumption by State and Market Class, Mexico 1993-1999.

............................................................................................................................... 46
Table 3.6 Bean Wholesale Price Seasonality in Mexico, 1990-1999. ............................. 52
Table 4.1 Direct Cash Payments of PROCAMPO in Mexico. ........................................ 67
Table 4.2 Phase-out for Bean Import Tariffs and Quotas under the NAFTA. ................. 71
Table 4.3 Imports from US. and Canada as a Percentage of the NAFTA Import Quota. 71
Table 4.4 Area Planted with BANRURAL Credit 1990-2000, Mexico. ......................... 73
Table 4.5 Bean Area Protected with Crop Insurance 1990-2000, Mexico. ...................... 74
Table 4.6 Statistics of the Kilo per Kilo Program for Beans in Chihuahua 1996-2000,

Mexico .................................................................................................................... 86
Table 4.7 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program 1997,
Mexico .................................................................................................................... 86
Table 4.8 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program 1998,
Mexico .................................................................................................................... 87
Table 4.9 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program 1999,
Mexico .................................................................................................................... 88
Table 4.10 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program
2000, Mexico. ......................................................................................................... 89
Table 4.11 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program
1996-2000, Mexico. ................................................................................................ 89
Table 4.12 Statistics of the Kilo per Kilo Program for Beans in Durango 1996-2000,
Mexico .................................................................................................................... 90
Table 4.13 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1996,
Mexico .................................................................................................................... 91
Table 4.14 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1997,
Mexico .................................................................................................................... 91
Table 4.15 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1998,
Mexico .................................................................................................................... 92
Table 4.16 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1999,
Mexico .................................................................................................................... 92
Table 4.17 Bean Seed Distribution in Durango through the Kilo per Kilo Program 2000,
Mexico .................................................................................................................... 93
Table 4.18 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1996-
2000, Mexico. ......................................................................................................... 94

xi

Table 4.19 Statistics of the Kilo per Kilo Program for Beans in Zacatecas 1996-2000,

Mexico .................................................................................................................... 95
Table 4.20 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 1996,
Mexico .................................................................................................................... 96
Table 4.21 Bean Seed distribution in Zacatecas through the Kilo per Kilo Program 1997,
Mexico .................................................................................................................... 96
Table 4.22 Bean seed distribution in Zacatecas through the kilo per kilo program 1998,
Mexico .................................................................................................................... 97
Table 4.23 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 1999,
Mexico .................................................................................................................... 98
Table 4.24 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 2000,
Mexico .................................................................................................................... 98
Table 4.25 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 1996-
2000, Mexico. ......................................................................................................... 99
Table 5.1 Characteristics of Farmers in Durango, Chihuahua, and Zacatecas 2001,
Mexico .................................................................................................................. 105
Table 5.2 Characteristics of Farms in Durango, Chihuahua, and Zacatecas 2001, Mexico.
............................................................................................................................. 106

Table 5.3 Percentage of Farmers by Source of Information, Durango 2001, Mexico. 107
Table 5.4 Soil Preparation and Planting Date in Durango, Chihuahua, and Zacatecas 2001

Mexico .................................................................................................................. 108
Table 5.5 Cultural Practices and Chemical Inputs in Durango, Chihuahua, and Zacatecas
2001, Mexico. ....................................................................................................... 110
Table 5.6 Production Costs, Credit, and Equipment in Durango, Chihuahua, and
Zacatecas 2001, Mexico, ....................................................................................... 111
Table 5.7 Family Labor, Hired labor, and Off-farm Employment in Durango, Chihuahua,
and Zacatecas 2001, Mexico. ................................................................................ 113
Table 5.8 Program Participation in Durango, Chihuahua, and Zacatecas 2001, Mexico.
............................................................................................................................. 114
Table 5.9 Storage and Farm Location in Durango, Chihuahua, and Zacatecas 2001,
Mexico .................................................................................................................. 116
Table 5.10 Selling Prices of Beans by Market Class in Chihuahua Mexico, 2000. ....... 117
Table 5.11 Selling Prices of Beans by Market Class in Durango Mexico, 2000 ............ 118
Table 5.12 Selling Prices of Beans by Market Class in Zacatecas Mexico, 2000. ......... 119
Table 5.13 Percentage of Area Planted by Variety in Chihuahua Mexico, 1997-2001. . 121
Table 5.14 Percentage of Area Planted by Variety in Durango Mexico, 1997-2001. 121
Table 5.15 Percentage of Area Planted by Variety in Zacatecas Mexico, 1997-2001 122
Table 5.16 Descriptive Statistics for Bean Yields in Northern Mexico, 1997-2001. ..... 123
Table 5.17 Percentage of Adopters Considering Improved Better than Traditional
Varieties, Mexico 2001. ........................................................................................ 125
Table 5.18 Program Participation and Adoption of Improved Varieties in Durango
Mexico, 2001. ....................................................................................................... 127
Table 5.19 Identiﬁcation of Variables in the Probit Model for Program Participation in
Durango, Mexico, 2001 ......................................................................................... 130
Table 5.20 Estimates of the Probit Model for Participation in the Kilo per Kilo Program
in Durango, Mexico 2001 ...................................................................................... 133

xii

Table 5.21 Skewness and Kurtosis Test for Normality for Residuals. .......................... 134

Table 6.1 Area, Production, and Yields for Dry Beans under Rainfed Conditions in
Chihuahua and Durango Mexico, 1990-2000. ....................................................... 140

Table 6.2 Value of Production, Grower Real Domestic Price (1994:100) for Beans in
Chihuahua and Durango Mexico, and FOB price for US. Pinto Beans, 1990-2000.

............................................................................................................................. 142
Table 6.3 Research Cost of Bean Research in Chihuahua, Durango, and Zacatecas
Mexico 1983-2000. ............................................................................................... 143
Table 6.4 Budget of the Kilo per Kilo Program in Northern Mexico, 1996-2000 .......... 144
Table 6.5 Consumer Price Index and Exchange Rate in Mexico, 1983-2000. ............... 145
Table 6.6 Total Research and Extension Real Costs in Northern Mexico, 1983-2000... 152
Table 6.7 Social and Net Gains from Research in the North of Mexico, 1983-2000. 153
Table 6.8 Social and Net Gains from Research in Northern Mexico, 1983-2000. ......... 157
Table 6.9 Sensitivity Analysis of Decreasing Changes in Adoption Rate and Incremental
Yield. .................................................................................................................... 159
Table 6.10 Sensitivity Analysis for Changes in Prices and in the Ofﬁcial Exchange Rate.
............................................................................................................................. 161
Table 6.11 Sensitivity Analysis for Changes in Research, Extension, and Adoption Costs.
............................................................................................................................. 162
Table 6.12 Sensitivity Analysis for Different Estimates of Supply and Demand Elasticity.
............................................................................................................................. 162
Table 6.13 Sensitivity of Beneﬁt Distribution to Different Estimates of Supply Elasticity.
............................................................................................................................. 163
Table 6.14 Sensitivity of Beneﬁt Distribution to Different Estimates of Demand Elasticity
............................................................................................................................. 164

xiii

LIST OF FIGURES

Figure 2.1 The Area of Study in Northern Mexico. ........................................................ 19
Figure 3.1 Total Planted and Harvested Area for Beans in Mexico, 1990-2000 .............. 27
Figure 3.2 Total Bean Production and Average Yield in Mexico, 1990-2000. ................ 28
Figure 3.3 Harvested Area, production, and Yield under Rainfed and Irrigated Conditions
in Mexico, 1990-2000. ............................................................................................ 32
Figure 3 .4 Seasonality of Bean Production through the Months of the Year in Mexico,
1990-2000 ............................................................................................................... 33
Figure 3.5 Total Imports of Dry Beans into Mexico, 1991-2000. ................................... 34
Figure 3.6 Total Exports of Dry Beans from Mexico, 1991-2000 ................................... 35
Figure 3.7 Apparent Consumption of Beans in Mexico, 1980-2000. .............................. 39
Figure 3.8 Net Imports as a Percentage of Apparent Consumption in Mexico, 1991-2000.
............................................................................................................................... 40
Figure 3.9 Bean Consumption Per Capita in Mexico, 1980-2000. .................................. 42
Figure 3.10 Price per Kilo of Beans, as a Percentage of Mexico’s Daily Minimum Wage,
Mexico 1991-1999. ................................................................................................. 43
Figure 3.11 Real Producer Prices for Beans in Mexico, 1990-2000 ................................ 49
Figure 3.12 Bean Wholesale Price Seasonality in Mexico, 1990-1999. .......................... 51
Figure 3.13 Dry Beans: U.S. Production and Grower Price, 1990-2000. ........................ 53
Figure 3.14 Marketing Margins for Domestic Beans in Mexico, 2000. .......................... 54
Figure 3.15 Bean Marketing Channels in Mexico, 1999 ................................................. 57
Figure 3.16 Production of the Bean Processing Industry in Mexico, 1994-2000 ............. 60

Figure 3.17 Real Cost of Processing Beans by the Industry in Mexico, 1994-2000. ....... 61
Figure 6.1 Logistic Curve for Estimating the Rate of Adoption in Mexico, 1991-2000. 154

xiv

CHAPTER 1. INTRODUCTION
1.1 General Background

During the decade of the 19903, the agricultural sector’s share of Mexico GDP
declined by 49 percent. In 1990, the agriculture, livestock, and forestry sector accounted
for an average of 7.2% of total GDP (MXS 738,898 million; approximately US$ 262,933
million). In contrast, in 2000 the sector1 accounted for an average of only 3.7% of total
GDP (MXS 5,491,373 million; approximately US$ 575,615 million).

Several factors have had a signiﬁcant impact on the agricultural sector, in general,
and the bean subsector in particular, including rapid population growth and urbanization,
slow growth in rural GDP, and market liberalization policies.

During the decade, Mexico’s population increased by 19 percent. In 1990,
Mexico’s total population was 83.2 million, of which 60.3 million was urban and 22.9
million was rural. In contrast, in 2000 total population was 98.9 million, of which 73.6
million was urban population and 25.3 million was rural population. While the urban
population increased rapidly (22%), the rural population remained relatively constant.

In addition, while Mexico’s nominal GDP per capita increased from USS 3,142 to
US$ 5,758 (83%), GDP per capita in the agricultural sector averaged approximately one-
fourth of the average for the whole economy, highlighting the importance of promoting
agricultural development (INEGI 2002, FAO 2002a, IPED 2002).

During the 19903 Mexico implemented policies to liberalize its economy. Since
joining the GATT in 1986 the Mexican government has promoted an open trade policy in

order to expand trade with its most important trading partners. Trade liberalization has

 

1 The most important economic activity was crop production (62.9% of sectoral GDP), followed by
livestock (32.0%) and forestry (5.0%).

taken place gradually, primarily by replacing import licenses on many agricultural
products with tariffs and by decreasing tariffs over time according to arranged schedules.
In addition, most quantitative export restrictions for agricultural products have been
eliminated. Finally, Mexico has signiﬁcantly simpliﬁed its sanitary and phytosanitary
regulations for importing agricultural products, fertilizers, and chemical agricultural
inputs.

In 1994, when Mexico entered the North American Free Trade Agreement
(NAFTA), it made a commitment to eliminate all the tariff and non-tariff barriers among
Canada, Mexico, and the US. Gradually, import licenses are being replaced by a system
of equivalent tariffs and quotas. While tariffs were eliminated immediately for some
products, for others tariffs were scheduled to be phased-out in periods of ﬁve, ten, and
ﬁfteen years. Quotas, which are tariff-exempt, will increase gradually over time,
generally at a rate of 3% per year, and will eventually be eliminated along with tariffs.
Although all agricultural products were included in the NAFTA agreement, long phase-
out periods were negotiated to protect import-sensitive agricultural products. Speciﬁcally,
for both maize and beans, the most important agricultural products in Mexico, a phase-
out period of ﬁfteen years (1994-2008) was established to facilitate the process of
transition. While NAFTA is Mexico’s most important trade agreement, Mexico has also

signed bilateral trade agreements with other countries that promote trade liberalization.

1.2 Problem Statement

After corn, common bean (Phaseolus vulgaris L.) is the second most important
basic grain crop in Mexico, in terms of area planted. In addition, beans are the second
most important component in the diet of the majority of the Mexican population.

During the 19908, the harvested bean area averaged 1.9 million hectares, with an
average yield of 632 kg/ha, and an average production of 1.2 million Mt. However,
during the last one-half of the decade, the bean subsector experienced negative growth.
During 1996-2000, the total harvested area decreased by 2.0% and average yield
decreased 2.5%, compared to 1990-1995. As a result, production declined 4.5% in 1996-
2000, compared to 1990-1995.

In the 19905, the annual value of dry bean production (in real terms, base-
year=l999) averaged MXS 8,050 million, equivalent to approximately USS 847 million.
This value represented only 6.1% of the total value of agricultural production in Mexico.
Considering that dry beans averaged 10.6% of total planted area during the same period,
it is possible to conclude that beans are a relative low value crop. However, it is
important to take into account that most of this crop is currently grown in very dry
regions of the country where farmers have few cropping options that are more proﬁtable
than beans (Tellez 1994; FIRA 2001).

On June 1"t 1982, the Bean/Cowpea-Collaborative Research Support Program2
(CRSP) signed an agreement with Mexico's National Research Institute for Forestry,
Agriculture and Livestock (INIFAP) to collaborate in developing improved bean varieties

for the Mexico’s semiarid highlands of the North-Central region. During the past ten

 

2 The Bean/Cowpea Collaborative Research Support Program is a research and training program that
supports international research partnerships to increase the availability of beans and cowpeas.

years (1990-1999), INIFAP has released several improved bean varieties that have been
adopted by farmers in the semiarid region (e. g., Pinto Villa and Pinto Mestizo).

However, to date the economic impact of these varieties has not been evaluated.
Thus, a study to assess the level of farmer adoption of improved bean varieties and to
estimate the ﬁnancial and economic rate of return to investments in bean research in
Mexico is required to determine the extent to which these research investments have
beneﬁted farmers and consumers and to provide information that will assist breeders in
setting ﬁrture bean research priorities. Thus, this study focused on assessing the economic
impact (i. e., costs, beneﬁts, and returns) of improved bean varieties in northern Mexico,

and on identifying the factors that have contributed to realizing the observed impact.

1.3 Justification of the Study

Numerous studied have demonstrated the critical role that increasing agricultural
production plays in the process of economic development and the key contribution of
research in promoting growth in agricultural production. However, now facing tighter
budgets, research administrators are increasingly being asked to provide evidence than
the costs of public-sector funded research are justiﬁed by the beneﬁts (Alston et a1. 1998,
1999)

The government of Mexico needs to justify its investments—as do other
governments and donors—because the economic value of public investment may not be
obvious. It is particularly diﬁicult to observe the impact of bean research because the
beneﬁts are diffused over many years and millions of dispersed producers and

consumers. Without an economic analysis, it is difﬁcult to assess the social value of new

technologies and to make informed judgments about the trade-offs in allocating scare
scientiﬁc resources (Masters 1996).

Thus, the subject matter of this dissertation—measuring the economic impact of
bean research targeted at developing improved bean varieties—is justiﬁed by the
importance of dry beans in the agricultural Mexico’s agricultural economy and the need
to provide decision-makers with information that they require to improve the allocation
of research resources (Norton and Davis 1981a). In addition, this study's contribution to
the literature is to support or reﬁrte "conventional wisdom" that investments in
agricultural research and development generate large dividends to society, more than
enough to support continued ﬁrnding in the ﬁrture.

Furthermore, this study generates insights that meet the information needs of the
main stakeholders of bean research investments: 1) Government decision—makers, who
desire information on the payoﬁ’ of agricultural research, since it competes with
alternative uses for public ﬁrnds; 2) Research administrators, who desire information on
the expected payoffs from ﬁrnds allocated to alternative research investments, and 3) The
general public (consumers and producers included), who has become increasingly

concerned about the productivity of their tax payments and government investments.

1.4 Research Objectives

The general objective of the study is to describe Mexico's bean subsector, analyze
the factors associated with adoption of the improved bean varieties released by INIFAP
in the 19905, identify factors that contributed to explaining the participation of farmers in

the govemment’s seed distribution program “Kilo por Kilo” in northern Mexico, and

estimate the net social gains of public investment in agricultural research to develop
improved bean varieties.

The speciﬁc objectives of this study are to:

1) Describe the historical and current status of the Mexican bean subsector
including trends in: a) bean supply (cropped area, production and yields, imports and
exports), b) demand for beans (changes in consumer preferences, aggregate consumption,
consumption per capita, and consumption by market class), c) bean prices (real and
wholesale bean prices, price seasonality and international prices), and (1) marketing (bean
marketing, marketing channels and the domestic bean industry). The purpose of the
subsector analysis is to identify the major constraints facing the Mexican bean subsector
and demonstrate how research can contribute to relaxing some of these constraints.

2) Analyze the role that government programs—including "Alianza para el
Campo", "PROCAMPO", and "Kilo por Kilo”—have played in supporting research,
agricultural extension, technology adoption, and growth in bean production in Mexico.
The purpose of the analysis is to identify the government policies that have constrained or
facilitated bean research at INIFAP and farmer adoption of improved varieties, including
the seed distribution program of the Secretariat of Agriculture.

3) Measure the rate of adoption of improved bean varieties grown under rainfed
conditions in the serrriarid highlands of northern of Mexico, including the states of
Chihuahua, Durango, and Zacatecas. In addition to estimating the aggregate level
adoption of improved varieties, data collected through a formal survey of bean farmers

are analyzed to identify the factors explaining farmer adoption and program participation.

4) Carry out an ex post impact assessment of the investments in bean research and
extension in Mexico, using the economic surplus method. The purpose of the ex post
impact analysis is to estimate the rate of return of the investments made by the
Bean/Cowpea CRSP and by the Mexican government through INIF AP to generate
improved varieties and by SAGAR to promote the adoption of improved bean varieties
under rainfed conditions in the states of Chihuahua, Durango, and Zacatecas.

5) Propose recommendations to guide future investments in bean research and
extension that can help to accelerate the rate of technology development and diffusion of

new improved bean varieties, and to increase productivity in Mexico’s bean subsector.

1.5 Dissertation Organization

The dissertation is divided into seven chapters. Chapter 2 is divided into two
sections. Section one presents a review of previous studies and concepts related to the
bean subsector, technology adoption, and economic impact analysis. Section two
describes the methods used to select the area of study, to carry out the rapid appraisal,
and to implement the farmer’s survey.

Chapter 3 describes the most important characteristics of Mexico’s bean subsector
and the most relevant constraints and opportunities that the bean subsector has faced
during the past decade. The chapter is divided into ﬁve sections. Section one describes
the characteristics of bean supply by analyzing the trends in total bean production, bean
production by season and by water condition, production seasonality, and the evolution of
bean imports and exports. Section two analyzes trends in the characteristics of bean

demand by examining changes in preferences, apparent consumption, consumption per

capita, and bean consumption by market class. Section three analyzes trends in bean
prices, including the evolution of real prices to producers, wholesale price seasonality,
international prices, and marketing margins. Section four describes the characteristics of
the bean marketing system, marketing channels, and production and costs of the domestic
bean industry. Section ﬁve summarizes the chapter.

Chapter 4 examines governmental policies and programs that have inﬂuenced
bean research, seed distribution, and farmer adoption of improved varieties of beans in
Mexico. The chapter is divided into four sections. Section one describes the
govemment’s support programs, price policy, trade agreements, farm credit, and crop
insurance related to bean production and marketing. Section two describes the most
important programs and organizations that promote bean research, technical assistance,
and the multiplication of improved seed. Section three analyzes the role that the Kilo per
Kilo program has played in the distribution of improved bean seed in the rainfed areas of
northern Mexico in the states of Chihuahua, Durango, and Zacatecas. Section four
summarizes the chapter.

Chapter 5 presents an estimation of the rate of adoption of improved bean
varieties and examines the factors that contribute to explaining farmer adoption and their
participation in the seed distribution program, based on the statistical analysis of data
collected from the surveys of bean farmers. The chapter is divided into six sections.
Sections one and two present data that characterize the surveyed farms and farmers in
northern Mexico. Section three analyzes factors that are associated with farmer
participation in government programs, farmers’ storage practices, and prices they

received for their bean crop. Section four reports the results of the analysis of area

planted, yields, and characteristics of the improved bean varieties. Section ﬁve analyses
the effect of the seed distribution program on farmers’ adoption of improved bean
varieties and reports the results of an econometric analysis of the factors explaining the
participation of farmers in the program. Section six summarizes the chapter.

Chapter 6 estimates the returns to investment in agricultural research for the
development of improved bean varieties in the rainfed areas of northern Mexico. The
chapter is divided into seven sections. Section one describes the data used for the
analysis. Section two explains the basic assumptions of the economic surplus method and
the formulas used to estimate the returns to research investment in a closed economy.
Sections three and four report the results for the ﬁnancial and economic analysis. Section
ﬁve uses sensitivity analysis to explore the effects that changes in key factors have on the
rates of return estimated and on the distribution of the beneﬁts between consumers and
producers. Section six applies the small open economy model for the economic analysis.
Section seven summarizes the chapter.

Finally Chapter 7 summarizes the most important ﬁndings of this dissertation and
presents the most important conclusions and policy implications for government
decision-makers, bean researchers, and extension services in Mexico. The chapter ends

with a discussion of the limitations of the study and ﬁrture research needs.

CHAPTER 2. LITERATURE REVIEW AND METHODS

This chapter is divided into two main sections. Section one reviews the literature
related to the topics analyzed in this dissertation, including studies that have been carried
out recently by several authors in different countries. This review identiﬁes insights from
previous studies, which guides the analysis reported in this study. Section two explains
how the study area (Chihuahua, Durango, and Zacatecas) was selected and the methods
that were used for collecting secondary and primary data, determining the sample size,

selecting the sample of farmers, and designing the questionnaire.

2.1 Literature Review

2.1.1 Bean Subsector Overview

A subsector has been deﬁned as “the vertical set of economic activities in the
production and distribution of a closely related set of commodities” (Shaffer 1973), or
alternatively as “an interdependent array of organizations, resources, laws and institutions
involved in producing, processing and distributing an agricultural commodity” (Marion et
al. 1986). The concept of subsector studies and the methodology was ﬁrrther developed
and utilized by several authors in the 19905 as a tool for analyzing microenterpise
programs (Haggblade and Gamser 1991), agribusiness subsector assessments (Miles
1994), and agricultural research program planning (Boughton et al. 1995).

According to Staatz (1997), the subsector approach examines how production and
distribution activities for a commodity are organized within the economy and asks how
the productivity of those activities can be increased, either through improved

technologies or better institutions and policies, to better coordinate various stages of

IO

production and distribution. The studies mentioned above provided the theoretical
framework to carry out the bean subsector overview presented in this dissertation.

In recent years, bean subsector studies have been conducted in Honduras and
Guatemala. The Honduran bean subsector study focused on the subsector’s historical
development, current status, and policy issues—including trends in production, adoption
of improved varieties, and implications for policy makers, researchers and extension
(Martel-Lagos and Bernsten 1994; Martel-Lagos 1995). Similarly, the Guatemalan study
presented an overview of the domestic bean subsector, focusing on constraints and
opportunities for vertical coordination and expansion of the bean processing industry
(Estrada-Valle 2001).

In the case of Mexico, recent studies from research centers (Rinderman et al.
1998), government agencies (Tellez 1994), and universities (Calva 1992, 1997) have
analyzed the evolution of the agricultural sector and its interaction with the rest of the
economy. Speciﬁc studies, carried out mainly by the Secretariat of Agriculture in Mexico
(SAGAR 1998, 2000), closely related branches (ASERCA 1997, 1998), and ﬁnancial
organizations (F IRA 1994, 2001) focused on various aspects of the bean subsector in
Mexico. These empirical studies provided important background information that is used
incorporated in this study’s overview of the Mexican bean subsector.

A formal subsector analysis requires a considerable amount of time and ﬁnancial
resources. Given limited resources, this bean subsector overview utilized rapid appraisal
or reconnaissance techniques to obtain the most relevant information needed to
understand the key aspects of the bean subsector in Mexico. As stated by Holtzman

(1986) and Chambers (1981), a rapid appraisal can be used to generate a broad and

11

preliminary overview of the organization, operation and performance of a food system or
its components, which is designed to identify system constraints and opportunities. In
addition, it can be used to identify system dynamics, linkages and overall problems,
which can then be examined more intensely during follow-up programs of research. The
rapid appraisal techniques used for this dissertation included informal interviews with
key informants, a review of literature, and the collection of information and secondary
data. Insights gained from the rapid appraisal guided the design of the survey of bean
farmers in Chihuahua, Durango, and Zacatecas.

2.1.2 Technology Adoption Studies

The adoption process was deﬁned by Rogers (1962) as “the mental process an
individual passes from ﬁrst hearing about an innovation to ﬁnal adoption”. However, a
distinction should be made between individual (farm-level) adoption and aggregate
adoption. Final adoption at the level of the individual farmer is deﬁned as the degree of
use of a new technology in the long-run equilibrium, when the farmer has ﬁrll
information about the new technology and its potential. In the context of aggregate
adoption behavior, the diffusion process has been deﬁned as “the process of spread of a
new technology within a region”. Aggregate adoption is measured by the aggregate level
of use of a speciﬁc technology within a given geographical area or a given population
(Feder et al. 1985).

There is a great diversity of studies on adoption of technology. Some focus more
on models of the adoption behavior of individual farms, while others focus more on the
models of aggregate adoption. Most of the theoretical studies of the adoption behavior of

individual farmers use static analysis that relates the degree of adoption to factors

12

affecting it. Most of the aggregate adoption models are dynamic and derive analytically
the behavior of the diffusion process over time (Feder et al. 1985). The adoption of
technology has been extensively analyzed in the survey on adoption of agricultural
innovations in developing countries (F eder et al. 1985) and in the study on technology
adoption decisions in small farms (Rauniyar and Goode 1992). These studies indicate that
constraints to the rapid adoption of innovations include factors such as the lack of credit,
limited access to information, aversion to risk, inadequate farm size, inadequate
incentives associated with farm tenure arrangements, insufﬁcient human capital, absence
of equipment to relieve labor shortages, problems of supply of complementary inputs,

and inappropriate transportation infrastructure.

The econometric approach for analyzing technology adoption was used by
Griliches (1957, 1980) and Dixon (1980) for exploring the economics of technological
change in the case of hybrid corn in the US. by using logistic growth functions. Logistic
curves also were used by Jarvis (1981) to predict the diffusion of new technologies in
Uruguay. A review of some possible empirical models for studying technology adoption
was done by Besley and Case (1993). One of the most recent adoption studies using an
econometric approach (Kosarek et al. 2001) analyzes the factors explaining the diffusion
of hybrid maize in Latin America and the Caribbean region (LAC), using cross-sectional
data from 18 countries and a two-limit Tobit model. They conclude that if policy makers
in LAC are to accelerate the diffusion of hybrid maize, they will have to ensure an
environment in which it is not only proﬁtable for producers to adopt improved varieties,

but also proﬁtable for the seed industry to produce and sell high-quality seed.

I3

In the case of Mexico, some of the most important characteristics of the
technology transfer process that affected small farmers were analyzed by Gonzalez
(1988). Other studies have focused on analyzing the farm level and aggregate adoption of
new technologies deve10ped by the INIFAP for the semiarid highlands of Mexico
(Gonzalez 1992, 1996) and evaluating the transfer of technology for dry beans (Pajarito
and Moncayo 2000). Also an econometric approach using logistic curves was used by
Byerlee and Hesse de Polanco (1986) to study the adoption of improved technological
components in Mexico. They suggested that agricultural research and extension programs
in developing countries, rather than following the conventional package approach, should
be designed to take into account the fact that farmers adopt improved technological
components in a stepwise manner.

In this dissertation, the adoption of technology was analyzed at the individual
farmer and aggregate levels by using the methodology suggested by CIMMYT (1993) for
survey design and by using the theoretical framework provided by the studies on
adoption of technology mentioned above, especially Feder et al. (1985). For the
econometric analysis the limited dependent variable probit model is used to investigate
the farmers’ participation in the seed distribution program following standard
econometric procedures (Amemiya 1981, Kennedy 1992, Gujarati 1995, Greene 1997,
and Wooldrige 2000).

2.1.3 Economic Impact Analysis

The main purpose of an impact assessment analysis of agricultural research is to
value and compare the costs and beneﬁts that arise from a situation “without” research,

against an alternative situation “with” research. The difference is the incremental net

14

beneﬁt arising from investing in research (Masters 1996, Gittinger 1982). It is
particularly difﬁcult to observe the impact of agricultural research because the beneﬁts
are diffused over many years and many dispersed producers and consumers. Economic
studies are needed to estimate the economic value of scientiﬁc knowledge or new
technologies—including the distribution of beneﬁts and costs—and to make informed
judgments about the trade-offs in allocating scarce resources (Alston et al. 1998). A
review of a wide range of methods for evaluating agricultural research can be found in
Schuh and Tollini (1979). Also a comprehensive review of the major research techniques
that have been developed to evaluate returns to agricultural research was done by Norton
and Davis (1981b).

Methods for doing an impact assessment are generally divided into three main
groups: a) econometric methods, aimed at estimating the marginal productivity of
research over a long time period and a variety of research activities; b) programming
methods, aimed at identifying one or more optimal technologies or research activities
from a set of options; and c) economic surplus methods, aimed at measuring the
aggregate social beneﬁts of a particular research project. For each method, the most
common approaches used for the analysis can be divided into two categories: ex ante and
ex past studies. The former category is applied for technologies that have not yet been
adopted, while the latter is applied for technologies already being used by farmers
(Masters 1996).

Given that the purpose of this study is to measure the aggregate social beneﬁts
derived from farmer ad0ption of improved bean varieties, the method of economic

surplus was selected. The economic surplus method is the most common approach for

15

analyzing the welfare effects of agricultural research in a partial-equilibrium framework,
and it can be applied to the broadest range of situations. The impact assessment of the
investment in research and extension of bean improved varieties (1V5) reported in this
dissertation falls under the category of ex post impact studies because the varieties are
already being used by the farmers.

Under the theoretical framework of partial equilibrium analysis, the focus is on a
single agricultural market. Farmer adoption of new technologies is expected to shift the
supply curve to the right, which will lead to a higher quantity of output and lower price;
and to changes in the consumer surplus, the producer surplus, and the social surplus
(Oehmke 2000). Implementation of the economic surplus analysis to evaluate agricultural
research can be broken dawn in ﬁve components (Alston et al. 1998): 1. Deﬁning the
problem, 2. Compiling the data, 3. Measuring the research-induced supply shift or
K-factor, 4. Analyzing the data, and 5. Interpreting and using the results. These
procedures were used to guide the economic analysis in this dissertation.

Examples of ex post impact assessment studies of investments in research and
extension for different crops include Sterns' (1993) ex post assessment of investments in
Cameroon's cowpea and sorghum research-extension systems from 1979 to 1987,
Dimithe's (1994) ex post estimation of the rate of return on maize research and extension
in Northern Cameroon from 1979 to 2000, Howard's (1994) economic impact study of
improved maize varieties in Zambia from 1978 to 1991, and Mather’s (2001) economic
assessment of bean research in the Dominican Republic. The last two of these studies

used an explicit economic surplus method for impact assessment.

16

2.2 Methods of Analysis

2.2.1 The Area of Study

The purpose of this dissertation is to assess the economic impact of improved
bean varieties released by INIFAP in the early and mid 19905: Pinto Villa 1990, Flor de
Mayo M38 1994, Pinto Mestizo 1996, Pinto Bayacora 1996, Negro Altiplano 1996, and
Negro Sahuatoba 1996 (Table 2.1). These varieties were developed for the rainfed
conditions of Mexico’s semiarid highlands with low rainfall (less than 450 mm a year),
soils with low level of fertility, and agronomic problems associated with monocropping
beans. In the states of Chihuahua, Durango, and Zacatecas—the three most important
states where beans are grown under rainfed conditions—beans are grown during the
Spring-Summer season (from July to October). The analysis for Chihuahua and Zacatecas
was done using secondary data and a drop-off survey, while the analysis for Durango was
done using secondary data and a face-to-face survey with farmers.

From 1990 to 2000, the total area planted to dry beans in Mexico averaged 2.2
million hectares. Approximately 85% of the total planted area was grown under rainfed
conditions (SAGAR 2000). The states of Chihuahua, Durango, and Zacatecas (Figure
2.1), accounted for about 62% (1.15 million ha) of the total rainfed bean area. Each
state’s share of the total was as follows: Chihuahua 10% (0.19 M ha), Durango 15% (0.28
M ha), and Zacatecas 37% (0.68 M ha). Considering that beans are grown in all 32 states
of Mexico, it is evident the importance of the states that were selected for the study.

In Durango beans are grown in most of the state’s 39 counties. The area of study
was deﬁned ﬁrst by considering the most important counties growing beans, based on the

number of hectares planted to beans under rainfed conditions.

17

doom .888 “3:2“ ”505%
68> team a 26: motorway 9.05 =< .n

.contomow 0.8 35% 05 E wows—2: mean—HEB 253:8 05 com 3:253 05 mo agar—838:0 0% 3:0 .m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«poemsﬁm

08-8... 8 3-2. 52m ﬂew seao seam 2 E 9.32

cam—at?

80-84 8 2:8 6on =85 cacao seem 2 E 832
Beta

. BBQ 88QO

005-com ow mm sew owed him :3on mm 2 8:5
883%

85a 8:82

838 ow 3-8 cerium owe: beam 8.85 mm 3 35m

32m 8033 wmz 992

ooméov oo 2» 82383 8:602 -Eom 8.8 in 2 Ba ow SE
803%

3880a 33m 35m
ooméom mm 3 -Eom 8:622 idem 8,9on on :4 ~25 8:3
ﬂamed M2538 watoaoa 8% Nazca—m $82
22> 8 939 8 when 925 28E 37. team maﬁa Bow .530 Bum cocuoﬁaooom «52.23

 

 

.02on 805.82 mo £822.80 waged 2: com Adan—H7: 3 38359 353:3 :3m 3538— oﬁ mo 8555820 _.N 2an

18

Figure 2.1 The Area of Study in Northern Mexico.

 

Chihuahua

\ .
aoz-z-z- __

Durango

      

 

 

Zacatecas

 

 

 

Most of area and production are concentrated in a small number of counties
located in the region known as “Los Llanos” of Durango. In this region four counties
(Cuencame, Guadalupe Victoria, Panuco de Coronado, and Peﬁon Blanco) account for
about 47% of the state’s total rainfed area planted to beans. The participation of each
county in the total is as follows: Cuencame 16%, Guadalupe Victoria 18%, Panuco de
Coronado 7%, and Peﬁon Blanco 6% (INEGI 2000).

The second step was to consider those counties in which the seed of improved
variety was distributed through the govemment’s Kilo per Kilo program. Information
obtained from the Secretariat of Agriculture in Durango for year 2000 indicated that

improved seed was distributed in the District of Rural Development 03 “Guadalupe

19

Victoria”—which is located in “Los Llanos de Durango” region. Of the total amount of
seed, the following percentages were distributed: Cuencame 19.4%, Guadalupe Victoria
19.0%, Panuco de Coronado 27.6%, Pef'ron Blanco 28.7%, Santa Clara 3.8%, and other
counties 1.5% (SAGAR-Durango 2000).

Based on the previous information, four counties (Cuencame, Guadalupe Victoria,
Panuco de Coronado, and Peﬁon Blanco)—which accounted for 94.8% of the seed
distributed—were selected for the survey in Durango. Within these counties, the quota of
respondents was allocated across villages roughly in proportion to their respective area
planted to beans and in which the Kilo per Kilo program distributed the improved seed to
farmers. In each village, farmers were randomly sampled from a list of individuals who
were growing beans under rainfed conditions during the Spring-Summer season 2001.
For the drop-off survey in Chihuahua and Zacatecas, the counties and villages were
selected in the same way as in Durango, and individuals were selected randomly. The
counties included for the survey in Chihuahua were: Namiquipa, Cuauhtemoc, and
Guerrero. The counties included for the survey in Zacatecas were: Sombrerete, Fresnillo,
and Rio Grande.

2.2.2 Rapid Appraisal

A rapid appraisal or rapid reconnaissance is a ﬂexible data collection procedure
designed to learn about the most important characteristics, opportunities and constraints
of the food system or a subsector in a cost-effective way. Several techniques can be used
when doing a rapid appraisal. The rapid appraisal techniques used for this dissertation
were collection of secondary data and informal interviews with key informants. The rapid

appraisal for collecting the data required for the bean subsector overview, government

20

policies, adoption of technology and economic impact analyses was carried out from
September to December 2001.

Secondary data were collected at the country level and for the states of
Chihuahua, Durango, and Zacatecas mainly from archives, reports, and publications from
several government ofﬁces. The most important government agencies visited were the
Secretariat of Agriculture, Livestock and Rural Development (SAGAR), the National
Research Institute for Agriculture, Forestry and Livestock (INIFAP), the National
Institute of Statistics, Geography, and Information (INEGI), the Trusts Instituted for
Agriculture (F IRA), and Supports and Services for Agricultural Marketing (ASERCA).

The informal interviews with key informants carried out in the state of Durango
focused on eliciting relevant information regarding development of improved bean
varieties, seed distribution, and use of new varieties. The key informants were selected
persons having special knowledge and understanding of bean research, support programs,
and adoption of technology. Accordingly, the interviews were conducted with decision-
makers, government staff, research administrators, bean researchers, and technicians.

2.2.3 Farmers’ Survey

Sampling Method

For the face-to-face survey in Durango, the target population was the set of
farmers growing beans under rainfed conditions during the Spring-Summer season 2001
(July to October) in previously selected villages in the counties of Cuencame, Guadalupe
Victoria, Panuco de Coronado, and Pei‘ion Blanco of the state of Durango. The villages
selected were those villages having the largest area planted to beans and in which the

improved bean seed was distributed through the government’s Kilo per Kilo program.

21

Once the target population was identiﬁed, the next step was to ﬁnd the
appropriate list to be used as the sampling frame. In most lists, certain members of the
target population are omitted, included more than once, or the list includes people that are
not of interest for the study. The aim was to choose a list with a minimum of omissions,
duplicate entries, and inaccuracies. After consulting with key informants, it was
determined that the register of farmers maintained by PROCAMPO (a cash transfer
program) is a reliable list that includes approximately 95% of all farmers in each county.
Therefore, it was chosen as the sampling frame for drawing the sample.

The sample size was determined by using the usual statistical procedures. The
sample size required depends basically on the desired level of conﬁdence (Z), the
maximum error of estimate (E), and how varied the population is with respect to the
characteristics of interest. In this case, the main characteristic of interest is the proportion
of farmers who have adopted an improved variety. To estimate a proportion (p) for a
population the sample size is given by n=(Z2*pq)/E2; where q=1-p is the proportion of the
population not having the characteristic of interest (Hogg and Tanis, 1997).

In the case of Durango, to obtain an estimate with 95 percent level of conﬁdence
(2:1 .96, for two tails), assuming an equal number of adopters and non-adopters
(proportion p=q=0.5), and ten percent sampling error (E=0. 10), the sample size needed to
estimate the proportion of farmers who have adopted improved bean varieties is:

n=(Z"2pq)/E"2=(1.96"2*0.5"2)/O.10"2=(3.84*0.25/0.01)=96 farmers

An initial sample size of 120 farmers (25% more than required) was considered
appropriate in order to ensure a ﬁnal sample size of 96 farmers after removing ineligible

respondents, nonrespondents, and discarding incomplete questionnaires.

22

The sample was distributed among the four counties of interest in proportion to
the area planted to beans in each county (Table 2.2). The number of farmers allocated in
this way was considered to be the minimum number to be interviewed. However, with the
help of enumerators, it was possible to carry out additional interviews. These additional
farmers were selected randomly. Therefore, the actual number of interviews obtained was
higher than the number initially computed. As a result, it was possible to reduce the

sampling error from ten to seven percent.

Table 2.2 Number of Face-to—Face Interviews in Durango, Mexico 2001.

 

 

 

 

 

 

 

County Planted Percentage Quota Percentage Actual Percentage
area (ha) of total minimum of total Number of total

Cuencame 51,908 34. 8% 42 3 5.0% 69 32. 8%

Guadalupe 57,316 38.5% 46 38.3% 73 34.8%

Victoria

Panuco de 22,642 15.2% 18 15.0% 38 18.1%

Coronado

Peﬁon 17,124 11.5% 14 11.7% 30 14.3%

Blanco

Total 148,990 100.0% 120 100.0% 210 100.0%

 

 

 

 

 

 

 

Source: INEGI (2000).

Random sampling was used to select the farmers to be interviewed. The records
of PROCAMPO farmers were obtained ﬁ'om the Secretariat of Agriculture. To draw the
farmers for the list, ﬁrst the lists of farmers in the villages of interest in each county were
combined. Then farmers were selected randomly until the quota required for each county
was ﬁlled. After ﬁlling the quota required per county and by adding the number of
farmers selected for each county, a total number of 120 farmers was obtained. However,

with the help of enumerators the number of interviews was increased to 210 (Table 2.2).

23

 

Questionnaire Design

The questionnaire for conducting the face-to-face survey with bean farmers in
Durango was designed to obtain the information required to carry out an assessment of
improved varieties adoption and economic impact analysis. The questionnaire included
both open-ended and close-ended questions, depending on which format was considered
better suited for providing the kind of information desired (Appendix A).

In order to detect potential problems in the wording of speciﬁc questions, the
questionnaire was pretested twice with the assistance of researchers, surveyors, and
farmers. Based on insights gained during the pretest, the questionnaire was revised by the
researchers and these changes were discussed with the enumerators. Then, the
enumerators used the revised questionnaire to interview an additional group of farmers.
In both cases, the answers were analyzed with the team of enumerators. After revising
individual questions, the ﬁnal version of the questionnaire was obtained. For the drop-off
surveys in Chihuahua and Zacatecas, the questionnaire was simpliﬁed and redesigned to
make it easy for the respondents to complete the questionnaires on their own.

In addition to the basic information to identify the farmer and his/her location,
questions were included to obtain information about the farmers’ land inventory, crops
planted, access to technical information, input use, bean varieties planted, grain yields,
production costs per hectare, and access to credit. The questionnaire also included
questions related to the farmers’ ownership of farming machinery, sales of beans,
assessment of improved bean varieties compared to traditional varieties, and participation
in government programs. Finally, the questionnaire included questions related to the

Farmers’ socioeconomic background, output storage, and transportation costs.

24

To carry out the survey, seven agricultural technicians living in the area of interest
were hired as enumerators. These enumerators were selected, based on the
recommendations of the Secretariat of Agriculture working in the counties covered by the
survey. In September 2001, the enumerators were trained on how to conduct the ﬁeld
interviews and the criteria for selecting the respondents. The survey was implemented

during the Spring-Summer season, during October and November 2001.

25

CHAPTER 3. AN OVERVIEW OF THE MEXICAN BEAN SUBSECTOR

In Mexico, common beans are an important crop, both in terms of consumption
and production. Along with maize, common beans3 are a fundamental component of the
diet of the low-income population in both urban zones and rural areas. In addition, bean
farmers utilize a high percentage of their production for self-consumption—from 18% to
30%, as reported in different studies. In terms of planted area, beans are second in
importance (after maize), averaging 2.2 million hectares during 1990-2000, compared to
approximately 8.5 million hectares for maize. Furthermore approximately 570,000
farmers grow dry beans—making it the second most important cropping activity (after

maize) from the social point of view (F IRA 2001).

3.1 Characteristics of Supply

While dry beans are grown throughout the world, during 1998-2000 the Americas
accounted for approximately 33.8% of the total bean area and 29% of total world
production. During the 19905 the most important bean-producing countries in the
Americas were Brazil (2,283,753 Mt, 38.0%), US. (1,289,186 Mt, 21.4%), and Mexico
(1,067,712 Mt, 17.8%). In terms of area, the most important countries were Brazil
(3,343,370 ha, 45.8%), US. (668,991 ha, 9.2%), and Mexico (1,783,629 ha, 24.4%).
Thus, Mexico is the third largest producer of dry beans in the Americas and has the

second largest planted area (FAO 2002a).

 

3 In Mexico the common bean (Phaseolus vulgaris L.) is the most important bean included in the category
“dry beans”.

26

3.1.1 Area, Production, and Yields

Area Trends

During the period 1995-2000, Mexico’s bean planted area averaged 2.2 million
hectares (with a standard deviation of 0.2 million ha and a coefﬁcient of variation of 7%).
While the total planted area increased 3.4% during 1996-2000, compared to 1990-1995,
the harvested area decreased 2.0%, averaging 1.9 million hectares (with a standard
deviation of 0.3 million ha and a coefﬁcient of variation of 14%). While both the planted
and harvested area have varied considerably from year-to-year (Figure 3.1), the harvested
area has been more variable, due to erratic weather conditions and because approximately

85% percent of the total area is planted under rainfed conditions.

Figure 3.1 Total Planted and Harvested Area for Beans in Mexico, 1990-2000.

 

 

 

 

 

 

 

 

 

3.000
Planted Area

2.500
m v \/
-= Harvested Area
:5 1.500 V
5

1.000

0.500

0.000 I I H I I I I I I I

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Year

Source: Appendix B.

27

Most likely, the area planted to beans will decline in the future, given that
producer prices have been declining. Thus, it is expected that the area planted to
commercial beans will decline, while the area of beans grown for subsistence will remain
roughly constant (FIRA 2001).

Average Yield Trends

Average bean yield has ﬂuctuated over the 1990-2000 period, with slumps in
1992 and 1998, and peaks in 1991, 1993 and 1996 (Figure 3.2). During the decade, yield
averaged 632 kg/ha with a standard deviation of 42 kg and a coefﬁcient of variation of

7%. However, average yield decreased 2.5% in 1996-2000 compared to 1990-1995.

Figure 3.2 Total Bean Production and Average Yield in Mexico, 1990-2000.

L600 800
Production

 

1.400

 

*-750

1.200 /\ M A
1.000 "\ /4 \Y/V i700

 

 

 

 

 

 

 

 

 

E /\ \ / /\ w
c A .c
E Yreld "‘
(1600
\ / \\/ -— 600
(L400 \/
v 550
(L200
0.000 I I I I I I I I I I 500
19901991199219931994199519961997199819992000
Year

Source: Appendix B.

28

Iota] Production Trends

Total production has ﬂuctuated over the 1990-2000 period, and has followed a
pattern very similar to the pattern of yields, with slumps in 1992 and 1997, and peaks in
1991, 1994 and 1996 (Figure 3.2). During the decade, production averaged 1.2 million
Mt with a standard deviation of 0.2 million Mt and a coefﬁcient of variation of 17%. The
greater variability in production (compared to harvested area) is due the ﬂuctuations in
both yields and harvested area. Production declined 4.5% in 1996-2000 compared to
1 990-1 995.

3.1.2 Area, Production, and Yields by Season

In Mexico, beans are grown during two seasons: the spring-summer (SS) season
and the fall-winter (FW) season. Most of Mexico’s bean crop is grown during the SS
season. During 1990-2000, the area harvested in the SS season averaged 1.6 million ha
(84% of total harvested area), production averaged 0.9 million Mt (73% of total annual
production), and yield averaged 546 kg/ha (Table 3.1). Yields are relatively low during
the SS season because beans are grown mainly under rainfed conditions in the semiarid
highlands where rainfall rarely exceeds 450 mm during the growing season. Farmers who
grow beans during this season face considerably risk—given the great variability in
weather due to both low levels of rainfall and the uneven distribution of rainfall over the
growth period. Another weather-related problem is early frosts, which makes production
very risky for farmers who are late in planting their SS bean crop.

Only a small share of Mexico’s beans is produced during the FW season. During
1990-2000, the area harvested averaged 0.3 million ha (16% of total harvested area),

production averaged 0.3 million Mt (27% of total annual production) and yields averaged

29

1,056 kg/ha (Table 3.1). Yields are relatively high during the FW season because most
beans are grown under irrigated conditions. Irrigation reduces production risks and soils
are more fertile in the irrigated area (e. g., Valle del Fuerte, plateau of Sinaloa), which is
mainly oriented to the production of cash crops like vegetables. The fact that yields are
higher during the FW season than during the SS season explains why the FW season
accounts for only 16% of harvested area, but 27% of total annual production. However,
although yields are almost twice as high during the FW season as during the SS season,

overall national average yield is low because most beans are grown during the SS season.

Table 3.1 Area, Yield and Production of Beans by Season in Mexico, 1990-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S ring-Summer Fall-Winter
Year Harvested Yield Production Harvested Yield Production
Area Kg/ha Million Mt Area Kg/ha Million Mt
Million ha Million ha
1990 1.888 573 1.081 0.206 1,000 0.206
1991 1.653 627 1.037 0.336 1,017 0.342
1992 1.080 476 0.514 0.215 949 0.204
1993 1.600 591 0.945 0.274 1,252 0.342
1994 1.775 571 1.012 0.312 1,127 0.352
1995 1.706 532 0.908 0.334 1,086 0.363
1996 1.768 601 1.063 0.280 1,021 0.286
1997 1.315 482 0.633 0.301 1,103 0.332
1998 1.789 517 0.925 0.357 941 0.336
1999 1.285 494 0.636 0.409 1,087 0.445
2000 1.579 537 0.848 0.301 1,034 0.311
Average 1.585 546 0.873 0.302 1,056 0.320

 

 

 

 

 

 

Source: SAGAR (2000).

Comparing trends in the early versus the late part of the decade of the 19905
indicates that average yields have declined over time, both in the SS and the FW season.
However, harvested area and production have decreased for the SS season but increased

for the FW season. When comparing the average in 1996-2000 with the average in 1990-

30

1995, during the SS season harvested area, yields, and production decreased by 6.2%,
6.4%, and 11.1%, respectively. In contrast, the same kind of comparison shows that
during the FW season yields decreased 3.2%, while both area and production increased
by 17.8% and 13.3%, respectively. This indicates that bean production is shifting from
the SS season, mainly under rainfed conditions, to the FW season, where most beans are
grown under irrigated conditions and that the increase in production is due to area
expansion, which has offset decreasing yields and led to a positive increase in production.

3.1.3 Production by Water Condition

In order to better understand trends in bean production, it is necessary to make a
distinction between beans grown under rainfed versus irrigated conditions. During 1990-
2000, an average of 1.6 million ha of the harvested area (84% of total harvested area) was
rainfed and 0.3 million ha (16% of total) was irrigated (Figure 3.3).

Approximately 768,000 Mt of beans (64% of total annual production) were
produced under rainfed conditions, while 432,000 Mt (3 6% of total) were produced
under rainfed conditions. Yields averaged 480 and 1,450 kg/ha under rainfed and
irrigated conditions, respectively.

It is also important to consider production by season and by water condition.
During 1990-2000 in the SS season, approximately 83% of the harvested area was rainfed
and 17% was irrigated. Yields averaged 456 and 1,456 kg/ha respectively, and about 77%
of Mexico’s SS season bean crop was produced under rainfed conditions while the other

23% was produced under irrigated conditions.

31

Figure 3.3 Harvested Area, production, and Yield under Rainfed and Irrigated Conditions
in Mexico, 1990-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

90
1550 kg
80 q ,
R :‘a:i:?;j:r'zi?iisi:3
70 a l
l I'
60 "2223??? --
8: n R
2 50 f ‘i‘. ,
5 e 'I ’ lmgated Rainfed giag-gg'
8 40 d n “353:3jfii‘iiiilj
& f 48° "9
30 . e
""931le1 'd ISLE—ff]
2° ‘
10
O 3‘33’2:2.EIE‘S‘Ef-I.:'E‘E'EJ. T ' . {515125113135if»? 1
Harvested Area Production AverageYield

Source: SAGAR (2000).

In the FW season, approximately 48% of harvested area irrigated and 52%
rainfed. Yields averaged 1,477 and 690 kg/ha, respectively, and about 67% of Mexico’s
FW season bean crop was produced under irrigated conditions while the other 33% was
produced under rainfed conditions (SAGAR 2000).

3.1.4 Impact of Production Seasonality on Bean Marketing

Beans planted during the SS season are harvested during September to December
and beans planted during the FW season are harvested during January to April. Because
84% of annual production is produced during the SS season, beans ﬂood domestic
markets during September-December (Figure 3.4). The abundant amount of beans in the

domestic market during this period has contributed to lower price of domestic beans and

32

making it more difficult for farmers to market their production. In contrast, there is a
relative scarcity of domestic beans in the market during May to August.

Figure 3.4 Seasonality of Bean Production through the Months of the Year in Mexico,
1990-2000.

40%

 

SS Season )1
35%
84% of total/ \

30% / \
20%

FW Season / \
15%

16% of total / \
10% ./\\ / r
5% = / T_
\\,__,__// Imports

 

 

 

 

 

 

 

Percentage of annual production

 

 

 

 

00/0 I I I I m m f I I f I
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month

Source: SAGAR (2000).

During the past few years, Mexico has mainly imported beans during the months
from August to October. Thus, there has been some overlap between the timing of
imports and when domestic production reaches the market.

3.1.5 Bean Imports and Exports

Since Mexico does not produce enough beans to meet its domestic requirements,
it has been importing beans of different market classes to meet part of the domestic
demand (estimated at 1.2 million Mt during 1990-2000). As a result, in the 19905,
Mexico became one of the largest importers of beans in the world. During 1991-2000,

Mexico imported an average of 76,277 Mt of dry beans per year (6% of total production),

33

ranging from a minimum of 2,909 Mt in 1992 to a maximum of 202,005 Mt in 1998
(Figure 3.5). Although there have been a lot of ﬂuctuation in the amount of beans

imported during this period, a clear increasing trend over time can be observed.

Figure 3.5 Total Imports of Dry Beans into Mexico, 1991-2000.

 

250,000

200,000 /\
150,000 / \

100,000 V \

50,000 A
\// \/
0

I T I I I I I I I

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Years

 

 

 

Metn’c ton

 

 

 

 

Source: FAO (2002a).

Regarding market class, most imports are pinto and black beans. The main source
of imports has been the US. (95%), followed by Canada and Argentina (5%). Pinto beans
is the most important market class in the US. During 1990-2000, pinto beans averaged
40% of total US. bean production, while black beans averaged 6% of US. production
(CORECA 1999, ERS 2002a).

During the 19905, Mexico exported only limited quantities of beans because
almost all domestic production was used to meet domestic consumption requirements

(Figure 3.6). During the 1991-2000 period, beans exported averaged only 24,568 Mt per

34

year (2% of total production), ranging from a minimum of 420 Mt in 1991 to a maximum
of 99,870 Mt in 1994 (mostly to Guatemala and Belize) (CORECA 1999). As exports
have ﬂuctuated from 1991 to 2000, there is no well-deﬁned increasing or decreasing
trend over time. In general exports have remained at very low levels, except in 1994 and

1995, when production exceeded domestic requirements.

Figure 3.6 Total Exports of Dry Beans from Mexico, 1991-2000.

 

120,000
100,000 r
80,000 R
60,000 / \
40,000
20,000 / \
0 r r w r r r r 1 1

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Years

 

 

 

Metric ton

 

 

 

 

 

Source: FAO (2002a).

Given the trends in imports and exports during the 19905, Mexico has been a net
importer of beans. During the 1991-2000 period, exports averaged only 24,568 Mt per
year while imports averaged 76,277 Mt (Table 3.2). Thus, net imports—computed as
imports minus exports—averaged 51,709 Mt per year (4% of total production), ranging
from a minimum of -56,810 Mt in 1995 to a maximum of 196,317 Mt in 1998. This

situation has resulted in a loss of foreign exchange and highlights the problem facing

35

domestic producers, who face the challenge of being competitive with imported beans in
terms of price and grain quality. Unfortunately, the prospects for the ﬁrture are ominous.
Starting in 1994, bean import quotas negotiated under the NAFTA have been increasing

by 3% annually and import tariffs have been decreasing.

Table 3.2 Iryorts, Exports, and Net Imports of Dry Beans in Mexico, 1991-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Imports Exports Net imports Percent of total
Mt Mt Mt production
1991 30,080 420 29,660 2%
1992 2,909 17,710 -14,801 -2%
1993 7,571 7,309 262 0%
1994 57,510 99,870 -42,360 -3%
1995 26,062 82,872 -56,810 -4%
1996 130,780 9,017 121,763 9%
1997 90,161 7,256 82,905 9%
1998 202,005 5,688 196,317 16%
1999 128,028 8,446 119,582 11%
2000 87,661 7,091 80,570 7%
Average 76,27 7 24,568 5 1,709 4%

 

Source: Computed with data from FAO (2002a).

Furthermore, from 1994 to 1999 bean imports equaled 100% or more of the
quota. In addition, the governments of the states of Chihuahua and Durango have
complained of illegal entry of dry beans into Mexico via smuggling. These phenomena
have further reduced the price of dry beans in the domestic market and have increased the
problems the farmers face in marketing domestic production. Clearly, these trends
indicate Mexican farmers are becoming less competitive, compared to US. bean growers.

By 2008, tariffs on imported beans will be eliminated among the countries that are
members of NAFTA. However, this is not the only factor that reduces the

competitiveness of Mexican producers. In addition, export subsides to producers in the

36

US, and loans that the US. provide to importing countries make it even more difﬁcult
for Mexican bean farmers to compete with US. producers. Thus, if there are no
improvements in increasing bean productivity and marketing efﬁciency in the future,
Mexico will depend mainly on imports to meet domestic demand, which will have a
negative impact on rural income and employment in the main bean producing regions of

the country—especially in northern Mexico.

3.2 Characteristics of Demand

Dry beans are a traditional staple food in Mexico, especially for low-income
people in rural and urban areas. Several factors play a key role in determining the level of
demand for beans, including the income levels, the degree of urbanization, and relative
prices4. During the past twenty years, Mexico has experienced both rapid urbanization
and transformation from a mainly traditional agricultural society into a modern industrial
society. While around 66% of the country’s population lived in urban areas in 1980, by
2000 this share increased to 74%. Similarly, per capita income has increased from USS
3,029 in 1980 to USS 5,758 in 2000 (IPED 2002).

3.2.1 Changes in Preferences

This trend—and the integration of Mexico into a global economy—has brought
about signiﬁcant changes in the food economy, including consumer’s preferences for dry

beans and value-added bean products.

 

4 In Mexico, consumers in different parts of the country have strong preferences for speciﬁc market classes
of beans (black, light-color, or pinto beans). Thus, producing more beans is not enough to satisfy the
demand by market class. In recent years, there has been an imbalance between the supply and demand for
speciﬁc market classes, which remains an important unsolved problem.

37

First, wealthier and more urbanized consumers are demanding higher quality
beans. For example, it is increasingly important that the beans they purchase are fresh
because the fresher the beans, the less time is required for cooking. In order to satisfy
consumers’ preferences for high quality beans, baggers are adding value by grading for
quality (homogeneous in color and shape, cleaned, polished and recently harvested),
packing their product in plastic bags, and marketing their brand through supermarkets.
However, beans are still commonly sold in bulk in traditional markets and increasingly in
some supermarkets.

Second, dry beans require a relatively long time to cook. With the increasing
participation of women in the labor markets, working women have less time to cook. A5 a
result, women are increasingly looking in the market for “convenient products” (i.e.,
products that are easy to prepare or almost ready for consumption). In response to
changing consumer preferences, processors are developing new and highly convenient
presentations, not only beans in plastic bags but also precooked, canned, and dehydrated
bean products.

Clearly, urbanization and income growth are having an important impact on
consumer preferences for new, more healthy, and convenient products, of standardized
quality. One implication is that increasingly, value is being added to beans after they are
marketed by farmers.

3.2.2 Apparent Consumption
Apparent consumption in Mexico (i. e., computed as production plus imports minus

exports) has ﬂuctuated greatly over the 1980-2000 period (Figure 3.7). This erratic

 

5 The female participation rate increased from 27% in 1980 to 33% in 1999 (World Bank 2002).

38

pattern can be explained mainly by the characteristics of the domestic supply, which
depends primarily on weather conditions. However, apparent consumption is also
affected by changes in the population, the patterns of trade, consumer preferences, and
consumer’s purchasing power.

Figure 3.7 Apparent Consumption of Beans in Mexico, 1980-2000.

2.0
1.8 L

‘11.: ./\ k A A
1.2 ---\/\--.A.T'?”9./..l..r.x/LI_ZL

 

 

 

 

 

 

 

 

 

 

 

 

 

5,0 I \/ \ I \ 7 V ‘"
é ' V \/ \f
20.8 \1 V
0.6
0.4
0.2
0.011111ﬁ11111111111111
1980198219841986198819901992199419961998 2000

Year

Source: Appendix C.

Although Figure 3.7 indicates a slightly downward trend in total consumption,
apparent consumption averaged 1.2 million Mt a year in the 19905, which was about the
same as the average in the 19805. However, although total apparent consumption has
remained essentially unchanged over the twenty-year period (1980-2000), Mexico’s
population has increased from 67.5 million in 1980 to 98.8 million in 2000. Thus, per
capita bean consumption has decreased by 18.5% during the period, as discussed in

section 3.2.3.

39

In years like 1981 and 1990, the high level of the apparent consumption was
primarily due to two factors: the large quantity of beans imported during those years, and
in these years domestic production was higher than the average level. Furthermore, in
both of these years, domestic production in the previous year (e. g., 1979 and 1989) was
far below the average (-40% approximately), which explains the high levels of imports
the subsequent years (Appendix C).

It is important to highlight that during 1990-2000, Mexico experienced a bean
deﬁcit. In seven of the past ten years, total consumption exceeded domestic production.
As a result, during the 19905 the size of the deﬁcit, measured by the relative importance

of bean imports, increased dramatically (Figure 3.8).

Figure 3.8 Net Imports as a Percentage of Apparent Consumption in Mexico, 1991-2000.

16%
14%
12%
10%
8%
6%
4%
2%
0%
-2%
-4%
-6%

Percentage

 

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Year

Source: Computed with data from FAO (2002a).

40

For example, net imports (i.e., computed as total imports minus total exports) as a
percentage of total consumption switched from very low or negative levels in the early
19905, to positive and relatively large percentages of total apparent bean consumption in
the late 19905.

3.2.3 Bean Consumption Per Capita

Among the countries in the Americas, Mexico has one of the higher levels of per
capita bean consumption. However, despite year-to-year variations (Figure 3.9) bean
consumption per capita decreased from and average of 16.2 kg per year during 1980-
1990 to 13.2 kg per year during 1991-2000—which in relative terms means a decrease
equal to 18.5% in bean consumption per capita when comparing the average of the
second period versus the average of the ﬁrst period. Thus, in the 19905, per capita
consumption fell at an annual rate of -3%, while other countries in the Americas were

still experienced positive growth rates6.

 

‘5 In the US, per capita bean consumption averaged seven pounds (1990-2000), and grew at an average
annual rate of 3.6% during the decade of the 19905. Bean consumption per capita is expected to continue to
increase due to the rapid increases in Hispanic population growth rate and because for many immigrants
who are from Central American and Mexico, beans are a important part in their traditional diet. In contrast,
per capita consumption in Brazil averaged 18 kg during the same period and grew at an annual growth rate
of 1% during the 19905 (FIRA 2001).

41

Figure 3.9 Bean Consumption Per Capita in Mexico, 1980-2000.

 

30.0

: /\
. ...\./.\...A.Irf'1"-/.\.. " ‘
v \/ WW

\1

 

 

 

kg per capita
a
O

 

 

 

 

 

0.0 I I If 7 I H I I I I I I I I I I I r I I

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
Year

Source: FAO (2002a).

Most of the decline in the per capita consumption has occurred in urban areas, as
indicated by the fact that in 1999, the estimated consumption per capita for the rural
population (16.8 kg) was 97.6% higher than for urban population (8.5 kg) (SAGAR
2000)

The decreasing share of the rural population in Mexico’s total, as result of the
process of urbanization, and the different preferences of urban consumers compared to
rural consumers are important factors that explain why consumption per capita has been
decreasing in recent years and will likely continue to decline in the future.

In addition, the decrease in bean consumption per capita can be partially
explained by the increase iii the relative price of beans, as a percentage of the daily

minimum wage (Figure 3.10). During 1991-1995, the price of one kilo of beans averaged

42

25.5% of the minimum wage, but during 1996-1999, the relative price of one kilo of

beans averaged 32.8% of the daily wage.

Figure 3.10 Price per Kilo of Beans, as a Percentage of Mexico’s Daily Minimum Wage,
Mexico 1991-1999.

 

 

 

 

 

 

 

 

 

 

 

 

45
40 -
35 //\.\\
a: 30
a)
g 20
°‘ 15
10
5
o r I I I I I I I
1991 1992 1993 1994 1995 1996 1997 1998 1999
Year

Source: FIRA (2001).

Other factors explaining declining per capita bean consumption are related to
changes in consumer preferences (see section 3.2.1), which has led to a reduction in the
share of the consumer’s budget spent on agricultural products and staple food (i.e., beans)
as income per capita increases7. Therefore, while Mexico’s population has increased from

83 to almost 99 million in 1990-2000 (Table 3.3), total apparent consumption has

 

7 This is associated with Engel’s Law, which holds that as income rise, demand for the products of the
agricultural sector will rise, but more slowly than income (households devote a smaller percentage of their
income to food), and with Bennet’s Law, which states that households switch from less to more expensive
calorie consumption, such as from coarse grains to meats and fresh fruits and vegetables, as incomes rise
(value-added foods become a larger part of the household budget than staple foods).

43

remained relatively constant around the mean of 1.2 million Mt per year as a

consequence of the declining per capita consumption.

Table 3.3 Bean Consum tion Per Capita in Mexico, 1990-2000.

 

 

 

 

 

 

 

 

 

 

 

 

Year Population Apparent Consumption
Thousands Consumption per capita
Million Mt kg/ha
1990 83,226 1.618 19.4
1991 84,801 1.408 16.6
1992 86,386 0.704 8.1
1993 87,976 1.288 14.6
1994 89,564 1.322 14.8
1995 91,145 1.214 13.3
1996 92,718 1.47] 15.9
1997 94,281 1.048 11.1
1998 95,831 1.457 15.2
1999 97,365 1.200 12.3
2000 98,872 0.964 9.7

 

 

 

 

 

 

Source: FAO (2002a).

3.2.4 Consumption by Market Class

As noted previously, preferences for different market classes vary greatly between
regions of Mexico. Regional consumption preferences for different bean market classes
were identiﬁed through a survey (n=1,514) conducted by scientists of the INIFAP
(Castellanos 1996). The study divided the country into four regions, plus the Federal
District of Mexico City (Table 3.4). The results indicated that in the North West, 98% of
the surveyed individuals preferred the “azufrado” type (sulphur yellow); in the North
East, 70% preferred “pinto” (beige with brown mottles) and “bayo” (cream) types; in the
South 90% preferred “black” type; in the Central region 34% preferred “Flor de Mayo”

(beige with pink mottles) type (although other commercial classes are also widely

44

consumed); and in the DP. (Mexico City), although there existed diversity in the kind of

beans preferred by consumers, 45% preferred the “black” type.

Table 3.4 Preferences for Bean Commercial Classes by Region in Mexico, 1993-1994.

 

 

 

 

 

 

 

 

 

 

Commercial Northeast Northwest Central South D.F."l
Class 0/0) (W (%) W!) (%I
Black 0.5 0.0 10.9 90.0 44.6
Azufrado 5.4 97.5 13.7 2.1 9.5
Bayo 25.3 0.8 9.5 5.7 10.1
Pinto 44.8 0.4 1.9 0.0 2.7
Flor de Mayo 11.3 0.0 33.7 0.0 12.0
Flor de Junio 0.0 0.0 21.0 0.0 6.8
Other 12.7 1.3 9.3 2.2 14.3
Total 100.0 100.0 100.0 100.0 100.0

 

 

 

 

 

 

 

a. The BF. refers to the Federal District of Mexico City
Source: Castellanos (1996).

A more recent study (F IRA 2001) examined production, consumption, and
balance (i. e., production minus consumption) of dry beans by market class for the 32
states in Mexico. Three market class were deﬁned, based on their color (Table 3.5):
“black” beans (Jamapa, San Luis, Michigan, Nayarit and Altiplano; “light” beans (Flor de
Mayo, Flor de Junio, Garbancillo, Azufrados, Canario, and Amarillo); and “pinto” beans
(Pinto Americano, Pinto Nacional y Ojo de Cabra). This study estimated that whereas
there was a surplus of “light” beans, there also was a large deﬁcit for “pinto” and “black”
beans. As a result, Mexico had an overall average annual net deﬁcit of approximately
28,300 Mt during 1993-1999, which had to be met via imports.

Pinto Beans

In 1993-1999, pinto beans accounted for an average of 25% of national

production. The most important producers of pinto beans were the states of Chihuahua

45

(3 7%), Durango (33%), and Zacatecas (12%), which produce approximately 82% of

domestic production of pinto beans. The center, north, and northeast regions of the

country were the most important consumers of this market class, especially the states of

Mexico (13%) and Chihuahua (12%).

Table 3.5 Annual Apparent Consumption by State and Market Class, Mexico 1993-1999.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

State Consumption (Mt) Production (Mt) Difference (Mt)
Black Light Pinto Black Light Pinto Black Light Pinto
Chiapas 41,957 5,199 5,199 56,196 3,128 3,127 14,239 -2,071 -2,072
Chihuahua 42,783 1 14,833 72,050
D.F 53,921 32,352 21,569 353 -53,568 -32,352 -21,569
Durango 9,832 14,752 15,000 9,000 102,714 15,000 -832 87,962
Guanajuato 6,088 30,436 24,349 1,244 39,331 9,435 —4,844 8,895 -14,914
Jalisco 7,935 47,609 23,804 21,366 4,1 14 -7,935 -26,243 -19,690
Mexico 65,692 49,269 49,269 10,932 3,000 1,500 -54,760 -46,269 -47,769
Nayarit 10,726 2,682 59,983 21,994 876 59,983 11,268 -1,806
Nuevo L. 24,014 24,015 934 2,618 -23,080 -21,397
Sinaloa 28,164 7,042 189,064 1,855 160,900 -5,l87
Veracruz 69,299 8,664 8,664 20,029 3,004 2,223 -49,270 -5,660 -6,441
Zacatecas 20,338 13,560 113,627 176,977 37,876 113,627 156,639 24,316
Subtotal 244,892 266,603 237,688 277,364 467,798 281,171 32,472 201,195 43,483
Other 148,144 233,234 132,221 83,343 98,118 26,603 -64,801 -135,116 -105,618
Total 393,036 499,837 369,909 360,707 565,916 307,774 -32,329 66,079 -62,135

 

Source: FIRA (2001).

Pintos are the market class with the largest annual deﬁcit (62,135 Mt), accounting

for 66% of the total domestic annual gross deﬁcit (94,464 Mt) during 1993-1999. The

states having the major deﬁcit of pinto beans were Mexico, Mexico D.F., Nuevo Leon,

Jalisco, and Guanajuato. In addition to interregional sales, Mexico has consistently

imported pinto beans to meet consumer demand.

Black Beans

During 1993-1999, black beans accounted for an average of 29% of national

production. The most important producers of black beans were the states of Zacatecas

46

 

(3 2%), Nayarit (17%), and Chiapas (16%), which produce approximately 65% of
domestic production of black beans. The center and the south regions of the country were
the most important consumers of this market class, including the states of Veracruz
(18%), Mexico (17%), and Mexico DE. (14%). Black beans are the market class with the
second largest annual deﬁcit (3 2,329 Mt), accounting for 34% of the total domestic
annual gross deﬁcit (94,464 Mt) during 1993-1999. The states having the major deﬁcit of
black beans were Mexico, Mexico DE, and Veracruz. In addition to interregional sales,
Mexico has consistently imported black beans to meet consumer demand.

Light Beans

Light color beans accounted for an average of 46% of national production during
1993-1999. The most important producers of light color beans were the states of Sinaloa
(33%), Zacatecas (31%), and Guanajuato (7%), which produce approximately 71% of
domestic production of light beans. The center and northwestern regions of the country
were the main consumer of this market class, especially the states of Mexico (10%),
Jalisco (9.5%), Mexico D.F. (6.5%), and Guanajuato (6%).

Because demand for these beans is low in the north, northeast, and south of the
country it is difﬁcult for farmers to sell all their light color beans in the domestic market.
As a result, there has been an annual surplus of approximately 66,079 Mt (12% of light
beans production). To market the surplus of this market class—in addition to sales to
other regions in the country and the regulation of the supply over time by storing the
beans—export sales and sales to the food industry have been important channels for

marketing light color beans.

47

The results in the study executed by FIRA (2001) basically agree with the
ﬁndings in the Castellanos study (1996) regarding the preferences for different market
classes of beans (pinto, black, and light color beans) in the most important regions of the

country.

3.3 Characteristics of Prices

Producer and wholesale bean prices vary greatly over time and according to
market class. The highest price is for the ﬁrst-class quality (cleaned, selected, sifted,
polished, and fresh) and the most preferred market classes (e. g., Flor de Mayo, Flor de
Junio, Negro Jamapa or Azufrados). The lowest price is for low quality beans (broken,
stained, mixed color, old hard-to-cook beans), and the nonpreferred market class (e. g.,
Ojo de Cabra).

3.3.1 Real Bean Prices

Rggl Producer Prices

Real producer prices for beans declined drastically during the 19905. From 1990
to 2000, the price fell continuously from MXS 3,307 (base-year=l994) to MXS 1,360 per
Mt (Figure 3.11), which implies a reduction in value of 58.9% in only 10 years. During
the period, the price was lowest in 1999 (MXS 1,279 per Mt). This drop in the real bean
price is associated with price liberalization, which is designed to set prices in the
domestic market to reﬂect the prices in the world market, where bean prices have been

declining.

48

Figure 3.1 1 Real Producer Prices for Beans in Mexico, 1990-2000.
(Base-year==l994)

3500

3000 \

2500 \\
2000 /\ 4.

 

 

 

 

 

 

 

 

 

 

E v
a» \/ x
2 1500 -~ .
\—/.
1000
500
0 T I I I I I I I I I
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Source: Appendix D.

Also, under the NAFTA, the establishment of higher import quotas and lower
tariffs on imported beans has made it possible to import beans at a lower price than
before. The market classes imported beans have been mainly pinto and black beans.

Real Wholesale Prices by Ma_r_ket Clgs_s

In the wholesale market, the decline in real prices has affected all market classes.
For example, in the central market of Mexico D.F., the most important central market in
Mexico, the real wholesale price for beans decreased from 1991 to 2000 for both
domestic and imported beans (base-year=l998). In general, domestic pinto beans have
faced the greatest real price decrease—from MXS 8.50 in 1991 to MXS 4.00 per kg in
2000 (-53%). During the period, the price of imported pinto beans (Pinto Americano) was

approximately 15% higher (on average) than that of domestic pinto beans (Pinto

49

Nacional). In contrast, the price of domestic black beans (the market class with the
smallest price decline) decreased from MXS 9.00 in 1991 to MXS 7.00 per kg in 2000
(-22%). During the period, the price of imported black beans (Negro Michigan) was
approximately 20% higher (on average) than that of domestic black beans (Negro
Nayarit). Domestic light color beans (Azufrado y Peruano) experienced a price decrease
—from MXS 8.00 in 1991 to MXS 6.00 per kg in 2000 (-25%) (FIRA 2001).

3 .3 .2 Wholesale Price Seasonality

By observing the variations of both production and nominal wholesale prices
through the year, price seasonality can be related to ﬂuctuations in domestic production
over the year. During 1990-1999, monthly average prices and production showed an
inverse relationship (Figure 3.12). In general, prices trended upward from April to
October, which is the period of the year when the supply of domestic beans decreases
considerably. In contrast, prices declined during February-March, and in November,
which is the period of the year when there is usually a surplus of domestic beans in the

marketplace.

50

Figure 3.12 Bean Wholesale Price Seasonality in Mexico, 1990-1999.

 

400 8.80

350 /_”'/\ i 8.60
Price r

/ / \ -r 8.20

/ \ - 8.00

x1 / \wo

1— 7.60

./ \\ Production // -_ 7'40
V - 7.20

I I I I l I I I T F T 7.00
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Month

 

 

 

rum

our

00
/

 

MX$I kg

 

Mt (Thousands)

 

.3 .5
O 01
O O
\

I

 

01
O

 

 

 

O

Source: SAGAR (2000).

However, a closer review of monthly production and price statistics shows that
while monthly production is highly variable, prices ﬂuctuate much less. During the 1990-
1999 period, production averaged 101 Mt per month, with a standard deviation of 99.2
Mt and a coefﬁcient of variation of 98%, which indicates a large variability in the amount
produced each month (Table 3.6). On the other hand, prices averaged MXS 8.3 per kg,
with a standard deviation of MXS 0.36/kg and a coefﬁcient of variation of only 4%. This
shows that throughout the year the variation in price is relatively small. For example,
over the year the highest monthly price (MXS 8.70) and the lowest monthly price per kilo
(MXS 7.68) differed only by 13%. This implies that although domestic production is
highly variable, the marketing system has been relatively efﬁcient in terms of
coordinating bean supply and demand through the year and in keeping prices relatively

stable.

51

Table 3.6 Bean Wholesale Price Seasonality in Mexico, 1990-1999.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Month Price Production
MXS/kg 000 Mt
Jannuary 7.96 72
February 7.76 160
March 7.68 137
April 8. 16 3 5
May 8.29 17
June 8.44 14
July 8 . 64 14
August 8.64 24
September 8.66 62
October 8.70 191
November 8.52 343
December 8.56 143
Average 8.33 101

 

 

 

Source: SAGAR (2000).

3 .3 .3 International Prices

The observed price decline for beans in the domestic market is related to the
decreasing trend in international market prices. For example, in the US, which is
Mexico’s most important source of imported beans, prices for beans in the late 19905
were lower than the prices in the in the mid 19905 (Figure 3.13). In contrast, U.S. bean
production was higher in the late 19905 than in the early 19905, as a result of an increase
in yields, while the acreage remained relatively constant (ERS 2002a, 2002b). This
increasing trend in productivity and bean production explains the decreasing trend in
prices.

In addition, ﬂuctuations in prices are related to ﬂuctuations in dry bean
production. Figure 3.13 shows that higher prices are associated with lower levels of
production during the period (e. g., 1993 and 1996) and that lower prices are associated

with higher levels of production during the period (e. g., 1991 and 1999).

52

Figure 3.13 Dry Beans: U.S. Production and Grower Price, 1990-2000.

 

 

 

40 90
35 — 80
30 — 70
25 Pnces _ 60

 

 

W P40

520 M / ”50
(315 \/ .30
10 W\ M

Million cwt

 

 

 

 

 

 

Production i 20
5 ~ 10
0 T I I I I I I I I I O
1990199119921993199419951996199719981999 2000
Year

Source: ERS (2002a).

3 .3 .4 Marketing Margins

Due to high marketing margins, there is a big difference between the price paid by
the consumer and the price received by producer. FIRA (2001) estimated that for
domestic beans, the rural wholesaler gets the largest share (3 0%) of the ﬁnal consumer
price (Figure 3.14), followed by farmers (26%), supermarkets (20%), urban wholesaler
(19%), and packers (5%).

In the case of imported beans (black and pintos), the largest share of the consumer
price goes to the urban wholesaler, which makes it more proﬁtable for wholesalers to buy
imported beans instead domestically-produced beans. For example, FIRA (2001)
estimated that in 1999 the urban wholesaler’s marketing margin for imported beans was
36% versus 19% for domestic beans. The fact that import quotas limit the amounts of

beans that can be imported is an important factor explaining this difference.

53

Figure 3.14 Marketing Margins for Domestic Beans in Mexico, 2000.

Bean packer

5% Rural
' wholesaler
30%

   
  

Bean farmer
26%

Supermarkets

Urban
wholesaler 20%
19%

Source: FIRA (2001).

Both the high concentration of large purchasing ﬁrms in the bean market (see
section 3.4.2), and the development of alliances to facilitate importing beans, have led to
a further reduction in costs and higher proﬁts for these few large ﬁrms. In addition,
according to the rules of the NAFTA, the expectation is that bean imports from the US.
will continue their increasing trend, as a result of scheduled (legally mandated) increases

in import quotas and decreases in import tariffs.

3.4 Bean Marketing System

One of the most important problems facing the bean subsector in Mexico is
marketing. This is related not only to the complexity that arises from the diversity of
demand for different market classes and quality of beans in regional markets, but is also

due to the high level of concentration in the wholesale bean market, which results in large

54

commissions to brokers, which in turn results in lower prices to producers and higher
prices to consumers. In addition, the market also faces the challenge of adjusting to price
liberalization and reduced government intervention.

3.4.1 Bean Marketing System

For many years, the bean marketing system in Mexico has been inefﬁcient, due to
the lack of information about the levels of demand and the preferences of consumers in
regional markets, the low quality of product presentation, and a lack of appropriate
marketing strategies. For example, farmers primarily market their beans in bulk (sacks of
50, 70, and 90 kg) without any previous selection or cleaning. As a result, middleman pay
farmers a low price. In addition, farmers receive low price because they are not organized
to negotiate better price and selling conditions.

For several decades the parastatal CONASUPO played a major role in marketing
many agricultural products including beans. Among other functions, CONASUPO was in
charge of regulating the supply and demand for beans in the domestic market. In addition,
it was solely responsible for importing beans to cover the deﬁcit in domestic production.
During the 19805, CONASUPO marketed approximately 40% of total bean domestic
production. In the late 19905, the process of opening up trade with the US. and Canada
and price liberalization led to the dismantling of CONASUPO. Since the closure of
CONASUPO in 1999, which previously intervened in corn and dry bean markets, the
government has allowed prices to be set by free market forces.

Four factors have constrained bean marketing in the late 19905. First, due to the
long distances between the production areas and the main consumption areas,

transportation costs are high. For instance, for the case of the production areas of Sinaloa

55

and Zacatecas, the distance by railroad to Mexico City is 1,411 and 855 km, respectively,
while the highway distance is 1,266 and 783 km, respectively, but the cheapest means of
transport is the railroad. Second, there are large commissions to brokers that affect both
producers and consumers. A5 a result, farmers receive a lower price for their harvest and
consumers have to pay a higher price for beans. Third, some of the preferred market class
beans have been displaced in the ﬁeld by new improved varieties that are higher yielding
varieties and disease or pest resistant, but these varieties are not preferred by consumers.
Fourth, the government has provided few supports for marketing domestically-produced
beans in the local market. For example, after the elimination of guaranteed prices and the
closure of CONASUPO in the late 19905, farmers had limited access to infrastructure
(mainly post-harvest infrastructure like storehouses and vehicles) and services to support
them in marketing their production, which reduced their ability to compete with bean
private traders—who had more resources available to market beans.

3.4.2 Bean Marketing Channels

The main bean-marketing channels in Mexico and the estimated percentage of the
product that is marketed through each channel are shown in Figure 3.15. Currently, an
estimated 20% of production is utilized for self-consumption, 5% are losses due to
spoilage during storage and transportation, and 5% is saved for seed. Therefore,
approximately 30% of production does not enter formal marketing channels. This share

for self-consumption includes varieties that satisfy very speciﬁc niche markets.

56

 

 

mHMOmNm

 

 

 

a\om
Ommm

.283 SE ”0558

 

 

 

 

 

 

 

20E

-mHMHmHQ

 

 

 

 

 

 

o\om
295397:

 

 

 

 

 

 

 

mHm—MMJNE
.MmA—Dm

 

 

 

 

 

 

‘

 

 

mmEOHm

>mmUOMO

 

 

 

 

3mg:

#00..—

 

 

 

fawn
mmmMoaﬁ

//

mhmog

 

 

 

 

 

 

 

 

$0M mama—(am
$3033

 

/

MmimmAOg
33¢
P

 

 

 

 

 

 

 

 

 

£8 29823
28.5%

 

 

ZOFUDQOME

 

 

 

 

 

gm
mmmmOA

 

 

 

 

UmeEOQ

 

 

 

.82 .882 E 28:20 messes: 8m 2 .m 88E

57

The most important varieties in this category are the “ibes” from Yucatan and
Chiapas, the “espelon” of Yucatan, the “combas” of Guerrero, the “ayacotes” of Morelos
and many other less important bean varieties.

The initial marketing channel for domestic beans are rural wholesalers
(middleman) with a large storage capacity, as well as many low volume buyers who serve
local markets. As small farmers are not organized to market their production and since
bean supply is very dispersed, the middlemen play an important role in gathering small
amounts of production and concentrating them in quantities that are large enough to be
marketed to the urban wholesalers, packers, and processors.

Very few of the middlemen add value to beans (i. e., packing and processing),
although the bulking and spatial concentration of beans is a value-adding activity. Rather,
most of them just sell the beans in bulk, and only in a few cases the beans are packed by
the middlemen. It is primarily the packer who carries out selecting and cleaning.

FIRA (2001) estimates that the middlemen purchase 70% of domestic production,
which they sell to wholesalers, packers, and processors. Processors commonly buy the
beans once that have been cleaned and residuals like stones, dust, straw and other foreign
matter has been removed. About 39% of domestic production is marketed through
wholesalers at central markets, while packers and processors purchase 26% and 5%,
respectively. Both wholesalers and processors import beans, although the former account
for the majority of imports (80%).

There is a high degree of concentration in dry bean marketing in Mexico.
According to FIRA (2001), approximately 60% of the marketed production is purchased

by only 16 firms that procure nationwide (seven ﬁrms purchase 40% of marketed

58

production). This means that the market for beans has the behavior of an oligopsony.
These few buyers have considerable market power, while the many small, dispersed and
unorganized producers have little bargaining power. As a result, bean farmers receive
lower prices and consumers pay higher prices than would be the case under a more
competitive marketing structure.

Consumers buy dry beans in the county market, traditional markets (tianguis),
grocery stores, or at the self-service retail market (e. g., supermarkets). During 1994-1999,
grocery stores sales decreased by 8% in real terms and local and traditional markets sales
declined slightly. In contrast, during this period the supermarkets increased their sales by
20%. This is because supermarkets offer better services to consumers in terms of the
diversity of the product, shopping hours, location, and comfort. In addition, supermarkets
sell beans at the same or even lower prices than other kind of stores. However, according
to consumer surveys, people believe that beans in plastic bags are not as fresh as beans
sold in bulk at traditional markets (FIRA 2001).

3.4.3 Bean Processing Industry

Value is added to beans through packing and the manufacturing of processed
products. The packing industry adds a small amount of value-added to the primary
product by cleaning, selecting, grading, polishing, and packing beans. While these
operations do not significantly modify the product, they generate the kind of product that
is increasingly being demanded by consumers (a clean and homogenous product). The

packing industry is located in the close proximity to the most important central markets

(D.F., Estado de Mexico, Jalisco, and Nuevo Leon). Packers prefer these locations

59

because large amounts of beans are available in these centers and they are near the
packers’ primary sales outlets—mainly supermarkets in densely populated urban areas.
The processing industry, which is concentrated in the states of Nuevo Leon, San
Luis Potosi, Aguascalientes, Estado de Mexico, and D.F., primarily produces canned and
dehydrated beans. A few years ago, few Mexican consumers bought processed beans.
However, since 1994 the production of processed beans has increased rapidly—3,3 14 Mt
in 1994 to 53,515 Mt in 2000 (Figure 3.16). This 16.2 fold increase represents an average 4

annual growth rate of 59% over the period.

Figure 3.16 Production of the Bean Processing Industry in Mexico, 1994-2000.

 

60,000
50,000

40,000 /
30,000 /

20.000 //
10,000

1994 1995 1996 1997 1998 1999 2000
Year

 

 

 

Metric tons

 

 

 

 

 

Source: FIRA (2001).

Over the period 1994-2000, the bean processing industry has been able to reduce

its unit cost of production from MX$ 15/kg to MXS 8/kg in real terms, with 2000 as the

base-year (Figure 3.17). This increase in efficiency has enabled processors to offer the

60

final consumer processed bean products at lower price. Although processed beans
represent only 4.5% of apparent consumption, this percentage is increasing rapidly due to
changes in consumer preferences.

According to FIRA (2001), currently only 5% of the total production of beans is
processed, while 26% of domestic production is cleaned and packed by the packing
industry. In general, processors purchase beans from the wholesalers, although in some
cases they attempt to buy directly from the producers in order to reduce costs of

purchasing and thereby be more competitive.

Figure 3.17 Real Cost of Processing Beans by the Industry in Mexico, 1994-2000.
(Base-year=2000)

 

18
16

\
14
12 \

 

 

 

 

/

 

 

 

 

 

 

:21
e9 \‘
EB -
6
4
2
0 I I I I I I

 

1994 1995 1996 1997 1998 1999 2000
Year

Source: FIRA (2001).

The processing industry is primarily interested in purchasing beans in bulk at a

low price, rather than purchasing high quality beans. In fact, processors prefer low quality

61

beans (old, broken, stained) in order to reduce their input costs, although for some
products whole beans are required. Processors purchase beans in 50 kg sacks, which have
been cleaned and sorted by market class. About 40% of Mexico’s processed bean output
is exported to the US. While the processing industry mainly utilizes bayo and black

beans, it also utilizes small quantities of other market classes.

3.5 Summary of the Chapter

In Mexico, dry beans are the second most important crop (aﬁer maize), both in
terms of consumption and production. Mexico is the third largest producer of dry beans
in the Americas and has the second largest planted area. During 1996-2000, the total
harvested area decreased 2.0% and average yield decreased 2.5% compared to 1990-
1995. As a result, production declined 4.5% in 1996-2000 compared to 1990-1995.

In Mexico, beans are grown during two seasons: the spring-summer (SS) season
and the fall-winter (FW) season. Most of Mexico’s bean crop is grown during the SS
season (84% of area and 73% of production). Yields are relatively low during the SS
season because beans are grown mainly under rainfed conditions.

Bean production is shifting from the SS season (where area, yields, and
production are decreasing) to the FW season, where most beans are grown under irrigated
conditions and where the increase in production is due to area expansion, which has
offset decreasing national yields and led to a positive increase in annual total production.

Beans planted during the SS season are harvested during September to December,

and beans planted during the FW season are harvested during January to April. Thus,

62

beans ﬂood domestic markets during September-December. In contrast, there is a relative
scarcity of domestic beans in the market during May to August.

Since Mexico does not produce enough beans to meet its domestic requirements,
it has been importing beans (mainly U.S. pinto and black beans) to meet part of the
domestic demand. There is an increasing trend in the amount of beans imported, which
has resulted in a loss of foreign exchange and highlights the problem for domestic
producers, who face the challenge of being competitive with imported beans. In addition,
bean import quotas negotiated under the NAFTA have been increasing annually and
import tariffs have been decreasing.

During the past twenty years (1980-2000), Mexico has experienced both rapid
urbanization and transformation from a mainly traditional agricultural society into a
modern industrial society. Urbanization and income growth are having an important
impact on consumer preferences. Increasingly, consumers are demanding new, more
healthy, and convenient bean products, of standardized quality. One implication is that
increasingly, value is being added to beans after they are marketed by farmers.

Although total population increased during 1980-2000, total apparent
consumption remain roughly unchanged. Thus, consumption per capita decreased by
18.5% during the period. This is explained by the process of urbanization, changes in
consumer preferences, and the increase in the price of beans relative to the minimum
wage.

Although Mexico has a surplus of light color beans, it has a deﬁcit of pinto and
black beans. During 1993-1999, pinto beans accounted for an average of 25% of national

production. The most important producers of pinto beans were the states of Chihuahua

63

(3 7%), Durango (33%), and Zacatecas (12%), which produce approximately 82% of
domestic production of pinto beans. Black beans accounted for an average of 29% of
national production.

The most important producers of black beans were the states of Zacatecas (32%),
Nayarit (17%), and Chiapas (16%), which produce approximately 65% of domestic
production of black beans. Light color beans accounted for an average of 46% of national
production. The most important producers of light color beans were the states of Sinaloa
(33%), Zacatecas (31%), and Guanajuato (7%), which produce approximately 71% of
domestic production of light beans.

Real producer prices for beans declined 59% in 1990-2000. This drop in the real
bean price is associated with price liberalization, which is designed to set prices in the
domestic market to reﬂect the prices in the world market, where bean prices have been
declining.

In the wholesale market, the decline in real prices has affected all market classes.
In 1991-1999, domestic pinto beans faced the greatest real price decrease (-50%)
followed by light color beans (-25%) and black beans (-22%).

Price seasonality is related to ﬂuctuations in domestic production over the year.
During 1990-1999, monthly average prices and production showed an inverse
relationship. In general, prices are higher before the SS harvest and lower after the FW
and SS season harvest. In addition, in the late 19905, there was a decreasing trend in the
grower price for dry beans in the US, which is Mexico’s most important foreign source
of imported pinto and black beans. If prices remain low, only some areas irrigated or with

good rainfed conditions in Mexico will be able to produce beans at competitive prices.

64

One of the most important problems facing the bean subsector in Mexico is
marketing. This is related not only to the complexity that arises from the diversity of
demand for different market classes and quality of beans in regional markets, but is also
due to the high level of concentration in the wholesale bean market, which results in large
commissions to brokers, which in turn results in lower prices to producers and higher
prices to consumers.

Value is added to beans through packing and the manufacturing of processed
products. The processing industry primarily produces canned and dehydrated beans.
Since 1994 the production of processed beans has increased rapidly. About 40% of
Mexico’s processed bean output is exported to the US. Although currently only 5% of
the total production of beans is processed and 26% is cleaned and packed, these

percentages are increasing rapidly due to changes in consumer preferences.

65

CHAPTER 4. GOVERNMENT POLICIES AND BEAN SUPPORT PROGRAMS
4.1 Government Agricultural Policies

In order to achieve the national food self-sufﬁciency, for several decades the
Mexican government has implemented different strategies to promote bean production
—including input subsidies and support prices. During the first half of the 19805, bean
production increased greatly as result of the "Sistema Alimentario Mexicano" (SAM)
program, which emphasized government participation in marketing and price regulation
(guaranteed prices). In the early 19903, the process of price liberalization required the
government to develop new strategies to support the agricultural development—mainly
through the "PROCAMPO" and "Alianza para el Campo" (Alliance for the Countryside)
support programs. Furthermore farmers adoption of improved varieties has been
promoted through the Kilo per Kilo program, a component of the Alliance for the
Countryside program.

4.1.1 Support Programs

During the decade of 19903 the Mexican government started several new
programs to support farmers and raise their standard of living. The “Programa de Apoyo
Directo al Campo” (PROCAMPO) provides direct support to farmers, based on their
planted area. The objectives of the program are to help farmers make the transition from a
govemment-regulated to a market-driven agricultural sector by compensating them for
subsidies provided to farmers in other countries, encouraging the development of
farmers’ organizations, and increasing competitiveness in production and marketing. As
the direct support is paid per hectare (Table 4.1), the program requires that the plots need

to be planted for a farmer to receive the support. The program was designed for 15 years,

66

because the government estimated that this period would provide enough time for farmers
to adopt new technologies, get organized, fuse smaller plots, and convert their land to a
more productive use.

PROCAMPO, which started in the Fall-Winter season 1993/1994, initially
supported farmers growing basic crops (i. e., maize, beans, wheat, and rice). From
September 1996 to August 1997, the program supported 13.9 million hectares, paid out
MXS 7,543.7 million (approximately US$ 980 million), and beneﬁted 2.96 million
farmers. Approximately 80% of those farmers planted under rainfed conditions (Spring-
Summer season). At present, as this support does not depend on the kind of crop being
grown, farmers can switch to another crop if it is better suited to the environmental and

market conditions.

Table 4.1 Direct Cash Payments of PROCAMPO in Mexico.

 

 

 

 

 

 

 

 

 

Crop Year MXS/ha Exchange US$/ha
Rate
1994 350 4.22 83.02
1995 440 6.67 65.98
1996 484 7.62 63.51
1997 556 7.96 69.84
1998 626 9.1 l 68.70
1999 708 9.84 71.92
2000 778 9.54 81.53
Mean 563 7.85 72.07

 

 

 

 

 

 

Source: SAGAR (2000).

As a complement to PROCAMPO, in 1996 the Federal Government started
“Alianza para el Campo” (Alliance for the Countryside), a support program for
promoting agricultural production and rural development. Some of the most important

components of Alliance for the Countryside that affect agricultural production are the

67

programs F ertigation, Kilo per Kilo, Rural Equipment, Mechanization, Research and
Technology Transfer, Integrated Rural Training and Extension, and the Elemental
Program for Technical Assistance.

For bean farmers, one of the most relevant components of Alliance for the
Countryside is the Kilo per Kilo program, which has the objective to increase yields by
promoting technical change (under irrigated as well as rainfed conditions) through the
substitution of modern varieties for traditional varieties. Seed of the improved varieties is
delivered to farmers at a price equivalent to commercial grain. This signiﬁcant reduction
in price makes improved seed very affordable to farmers. Since its inception, the Kilo per
Kilo program has shown positive results. During 1996-1997, more than 100,000 hectares
were planted modern bean varieties, resulting in a signiﬁcant increase in yields —up to
100%—in the lowland and the tropic (SAGAR 1998).

4.1.2 Price Policy

Until 1993, bean prices were determined mainly by the strong intervention of the
government through guaranteed prices that gave farmers a minimum revenue which
covered their production costs. This system operated via the “Compania Nacional de
Subsistencias Populares” (CONASUPO), a parastatal that purchased commodities from
farmers. But starting in 1993, the government initiated important changes to promote a
higher level of private sector participation in bean marketing. After these changes,
CONASUPO’s role in the market was relegated to being a buyer of last resort.

Up to the F all-Winter season 1995/1996, CONASUPO maintained an active
participation in the market of grains. Subsequently, it started to reduce its market

participation (as part of the government process of deregulation) and began to use the

68

criterium of “minimum buying price”. Under this policy, the farm-gate price was
established at a level about 40% lower than the price in the central market—a price that
was considered as the “lowest price”. Once the process of deregulation was initiated,
using the “last resort buyer” criterium, CONASUPO’s market share fell to around 6% of
domestic bean production, with the private sector increasing its participation over time in
the marketing process.

Starting in May 1998, CONASUPO stopped buying dry beans and establishing
minimum prices. Since then, bean prices have been determined according to a freer
interaction of supply and demand in the marketplace. In addition, several services that
supported the marketing of agricultural products were taken on by the office “Apoyos y
Servicios a la Comercializacion Agropecuaria” (ASERCA). Finally, on May 24th, 1999 a
decree that called for closing CONASUPO was published in the Federation Oﬂicial
Diary (DOF).

However, under the new policy of price liberalization, growing beans under
rainfed conditions is not sustainable without the direct support of PROCAMPO. In the
period 1993-1999, FIRA-Bank of Mexico carried out an annual survey to measure
proﬁtability8 for farmers growing beans under rainfed conditions in Chihuahua, Durango,
and Zacatecas; and under irrigated conditions in Sinaloa and Nayarit—the most important

states producing beans under each water condition.

 

8 Proﬁtability was measured (without including the direct cash payments) by using the ratio: Net
Revenue/Cost = N/C = (Revenue-Cost)/Cost. The results from the survey show that during that period,
under rainfed conditions, the average N/C ratio (1993-1999) was close to zero. However, it was positive for
Chihuahua (0.11) and Durango (0.10), but negative for Zacatecas (-0.12). Under irrigated conditions, the
average N/C ratio (1993-1999) was slightly positive and higher for Sinaloa (0.34) than for Nayarit (0.27).
When PROCAMPO payments were included in the analysis the average N/C ratio for Zacatecas (under
rainfed conditions) changed from -0. 12 to 0.28 on average over the 1994-1999 period (F [RA 2001).

69

The results of the study show that without the PROCAMPO payment, bean
production is sustainable only under irrigated conditions or in areas with good rainfed
conditions and high yieldsg.

4.1.3 Trade Agreements

While Mexico has signed trade agreements with different countries, the most
important agreement (for most of products including dry beans) has been the North
American Free Trade Agreement (NAFTA) to which Canada, Mexico and the US. are
signatories. Under NAFTA, which came into force in 1994, three different categories
were established: “beans for seed” (code 0713.33.01), “beans except for seed” (code
0713.33.02) for the most important market classes, and “other” (code 0713.33.99) for
other bean market classes. “Beans for seed”, was tariff-exempt before the agreement. The
“other” was made tariff-exempt as soon as the agreement started. For “beans except for
seed”, the agreement established that it would become tariff-exempt over a fourteen-year
period—starting with a tariff of 133.4% ad valorem and a tariff-exempt import quota of
51,500 Mt (50,000 for the US. and 1,500 for Canada). For both Canada and the US, an
annual increase of 3% in the import quota was established. By 2008, dry beans will be
tariff-exempt and they will be imported freely. The phase-out tariﬁ“ schedule and tariff-
exempt quotas for importing dry beans into Mexico, as agreed to under the NAFTA, are

shown in Table 4.2.

 

9 Under rainfed conditions, average cost of production averaged MXS 2,270 per ha in 1999 (around USS
231), whereas yield averaged 420 kg per ha in 1993-1999, which implies a unit cost of USS 550 per Mt
Under irrigated conditions, cost of production averaged MX $4,397 per ha in 1999 (around USS 447), and
yield averaged 1,595 kg/ha in 1993-1999, which implies a unit cost of USS 280 per Mt (FIRA 2001).

70

Table 4.2 Phase-out for Bean Import Tariffs and Quotas under the NAFTA.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Tariff U.S. (Mt) Canada (Mt) Total (Mt)
1994 133.4% 50,000 1,500 51,500
1995 127.8% 51,000 1,545 53,045
1996 122.3% 53,045 1,591 54,636
1997 116.7% 54,636 1,639 56,275
1998 111.2% 56,275 1,688 57,964
1999 105.6% 59,964 1,739 59,703
2000 93.9% 59,703 1,791 6L494
2001 82.1% 61,494 1,845 63,339
2002 70.4% 63,339 1,900 65,239
2003 58.7% 65,239 1,957 67,196
2004 46.9% 67,196 2,016 69,212
2005 35.2% 69,212 2,076 71,288
2006 23.5% 71,288 2,139 73,427
2007 1 1 . 1% 73,427 2,203 75,629
2008 0.0% No quota No quota No quota

 

 

Source: FIRA (2001).

From 1994 to 1999, bean imports from the US. and Canada were only lower than
the NAFTA established quota in 1995. In 1996 and 1998, imports were 123% and 190%
higher than the quota, respectively (Table 4.3). The signiﬁcant increase in the amount
imported in 1996 was due to a shortage of beans in the domestic market, which by
October 1996 led to a price increase of 63.3% for pinto beans and 80% for black beans,

with respect to the price level of January 1996 (SAGAR 1998, 2000).

Table 4.3 Imports from US. and Canada as a Percentage of the NAFTA Import Quota.

 

 

 

 

 

 

 

 

 

 

 

Item 1994 1995 1996 1997 1998 1999 2000 p/
Imports 50,725 25,538 121,687 55,987 168,262 60,998 2L085
Quota 51,500 53,045 54,636 56,275 57,963 59,702 61,494
M/Q 98% 48% 223% 99% 290% 102% 34%

 

 

p/ preliminary data only for January-July 2000.
Source: SAGAR (2000).

71

 

The Mexican government established that the rights for the import-quota would
be assigned through a public bid auction, which is open to the participation of physical
and legal persons (individuals and ﬁrms) established in Mexico that ﬁll the requirements
in the bid announcement. The allocation is awarded to the minimum price bidder. The
auction, which is held annually, is subject to speciﬁc regulations for tariffs and quotas,
under the “Law of Foreign Trade”.

Under an agreement with the World Trade Organization (W TO), Mexico agreed
to annually reduce its initial non-NAF TA tariff of 139 % for beans, starting on January
1995 and up to 2004. This implies that at the end of the period, the tariff will be reduced
to a level of 126%. Mexico has also signed agreements to reduce the tariffs on bean
imports with other countries, including Costa Rica (1994), Bolivia (1995), Chile (1998)
and Nicaragua (1998). The agreement with Costa Rica provides for gradually reducing
tariffs on beans imported into Mexico over a ﬁfteen-year period, starting from a tariff of
135.3% in 1995 to a tariff of 0% in 2009. Costa Rica will also reduce import tariffs to
0%, starting from an initial 55% tariff.

A similar agreement was made with Bolivia, under which Mexico will gradually
reduce tariffs on beans imported into Mexico over a ﬁfteen-year period, from an initial
133.4% in 1995 to 0% by 2009. Under the agreement with Chile, beans imported from
Chile face a tariff of 0%. However, to import to Mexico, the importer must ﬁrst obtain an
import license. In the case of Nicaragua, the agreement grants a tariff-exempt quota of
4,000 Mt for beans imported into Mexico from April to August every year, with a 3%

increase per year in the quota. However, imports from Nicaragua above the quota are

72

assessed a tariff, which starting from an initial 127% will be phased out over a ten-year
period.

4.1.4 Farm Credit

Public credit is administered through BANRURAL, a ﬁrst ﬂoor development
bank that provides loans and other bank services to farmers. During the period 1990-
1999, BANRURAL assigned 68% of its total credit for short-term loans, 22% to long-
term loans and 10% for other purposes (pledge loans). In 1990-1999, dry beans
represented the second largest area receiving short-term loans. On average, 258,000 ha of
beans were planted with BANRURAL credit during that period. About 15% of

BANRURAL’s total loans supported bean production (Table 4.4).

Table 4.4 Area Planted with BANRURAL Credit 1990-2000, Mexico.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Thousands of hectares Percentage of
Total area Bean area total area
1990 1,951 275 14%
1991 1,236 193 16%
1992 1,198 123 10%
1993 1,045 82 8%
1994 1,129 92 8%
1995 1,251 249 20%
1996 1,851 376 20%
1997 2,178 343 16%
1998 2,115 450 21%
1999 2,105 323 15%
2000 p/ 1,846 328 18%
Mean 1,628 258 16%

 

 

 

 

 

p/ preliminary data for 2000 subject to ﬁnal revision.
Source: SAGAR (2000).

From 1990 to 1993, the total crop area covered by credit from BANURAL

showed a decreasing trend—falling to 1,045,000 ha in 1993. After 1993, the total crop

73

area trended upward through 2000. The bean area with credit basically followed the same
pattern, ranging from a minimum of 82,000 ha in 1993 to a maximum of 450,000 ha in
1998.

Considering that the area planted to beans in 1990-2000 averaged 2.2 million
hectares and the area that received credit averaged 25 8,000 ha, the share of the area

receiving ofﬁcial credit represented only 11.7% of total bean area.

4.1.5 Crop Insurance

Beans have been one of the crops with the largest participation in the crop
insurance program. In 1990-1999, the bean area protected with crop insurance was on
average 180,000 ha, while the total area was 1,072,000. The share of beans in the total
area with crop insurance averaged 17% (Table 4.5), which is larger than the share of

beans in total planted area (10.6%).

Table 4.5 Bean Area Protected with Crop Insurance 1990-2000, Mexico.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Thousands of hectares Percentage of
Total area Bean area total area
1990 1,226 239 19%
1991 624 3 8 6%
1992 870 46 5%
1 993 912 72 8%
1994 1,050 78 7%
1995 91 5 96 10%
1996 1,146 215 19%
1997 1,218 191 16%
1998 1,358 263 19%
1999 1,123 350 31%
200M 1,3 50 393 29%
Mean 1,072 180 17%

 

p/ preliminary data for 2000 subject to ﬁnal revision.

Source: SAGAR (2000).

74

 

However, considering that the area planted to beans in 1990-2000 was on average
2.2 million hectares, the share of the area protected with crop insurance to grow beans
represented only 8.2% of total bean area. In the early 19905, both the total and the bean
area with crop insurance decreased and reached a minimum in 1991. From 1992 to 2000,
they both increase to reach a maximum for the total and bean area in 1998 and 2000,
respectively.

Given that approximately 1.9 out of 2.2 million ha planted to beans are grown
under rainfed conditions and that rainfall is highly variable from year-to-year, bean
farmers face considerable risk. Consequently crop insurance plays a key role in
supporting bean farmers. For this reason, the government has made an effort to increase
the bean hectares covered by crop insurance, as well as the percentage of the total bean
area covered. As a result, in 2000 the bean hectares covered by the insurance increased to
393,000 ha and the percentage of covered area, as a percentage of the total bean area (2.2

million ha), increased to 17.8%.

4.2 Research, Technical Assistance, and Seed Production

4.2.1 Research and Technology Transfer

Mexico’s Agricultural Research and Technology Transfer Program (ARTTP), a
component of the Alliance for the Countryside umbrella program, is intended to increase
productivity by providing access of farmers to new technologies and by promoting
agricultural research, diffusion, and technical assistance. The program is administrated by
the “Fundaciones Produce” (Produce Foundations), which are state-level associations

whose members are the leaders of farmer organizations. At the national level, they are

75

associated under the “Coordinadora Nacional de Fundaciones Produce” (COFUPRO)
(National Coordination for Produce Foundations) for the purpose of establishing a
national strategy and to share experiences regarding how to strengthen agricultural
research accelerate technology transfer.

According to the government regulations, the federal and state governments
should allocate seven percent of the resources that they allocated to the Alliance for the
Countryside program in each state to ARTTP. The regulations also recommend that at
least 50 percent of those resources be allocated to projects and activities implemented by
the INIFAP (the national institute for agricultural research), subject to the process of
approval designed by each Produce Foundation (SAGARPAlo 2001a).

ARTTP target agricultural producers and their economic organizations, and other
agents in the agricultural and rural sector requiring speciﬁc projects or research activities,
validation, and transfer of technology. The federal and state governments will provide
support (equally matched) for research projects, and events of validation, demonstration,
diffusion, and specialized training, and also for administration expenditures, equipment,
and infrastructure for projects and events.

To receive resources from ARTTP, farmers and agents must submit an application
to the Produce Foundations in their state. The Produce Foundations should inform the
applicants in writing whether or not their application was approved (see Appendix E).

4.2.2 The National Research Institute for Forestry, Livestock, and Agriculture

Agricultural research in Mexico ofﬁcially began in 1907, with the establishment

of the Central Agricultural Experiment Station of San Jacinto D.F. In the 1930s, the

 

'0 The acronym of SAGAR for the Secretariat of Agriculture was changed to SAGARPA in 2001.

76

government created the Department of Experimental Stations under the Secretariat of
Agriculture, which became the Agricultural Research Institute (IIA) in 1940.

In 1943, an agreement between the Secretariat of Agriculture and the Rockefeller
Foundation created the Special Studies Ofﬁce (OEE), dedicated exclusively to
conducting research basic crops. In 1961, the challenge of providing food to a
continuously increasing population led the Mexican government to create the
Agricultural Research National Institute (INIA).

In 1985, this institute was renamed the “Instituto Nacional de Investigaciones
Forestales, Agricolas y Pecuarias” (INIFAP) (National Research Institute for Forestry,
Livestock, and Agriculture).

INIFAP has a mandate to implement projects and programs that:

“generate technologies that are necessary for increasing forest, agricultural, and livestock
productivity and production in the country considering the interests, requirements, and
socioeconomic conditions of producers, in such a way that the increased production may
be able to meet the food and nutritional needs of an increasing population, the
requirements of the domestic industry, and generate surplus for exporting, contributing to
improve the well-being of peasants and population in general” (SAGARPA 2002).

INIFAP’s strategy for achieving its objective includes the following:

Development of domestic research capacity.

. Recognition and use of farmers’ empirical knowledge.
o Rational exploitation of genetic materials.

. Realization of joint research.

. Exchange of materials and information.

. Validation of research results.

. Diffusion of research results.

Appendix F includes an explanation of these activities.

77

4.2.3 Technical Assistance Program
As part of Mexico’s programs for agricultural development, implemented under
the Alliance for the Countryside umbrella program, a technical assistance program named

“Programa Elemental de Asistencia Tecnica” (PEAT) (Basic Program of Technical

Assistance) was established with the following objectives (SAGARPA 2002):

. Promote the application of technologies by supporting eligible farmers to hire and pay
for technical assistance services from independent contractors in all the stages of the
productive process, including production planning, implementation of appropriate
technological activities, preservation of the physical environment, acquisition and
application of technological inputs, harvesting, storing, marketing, and the promotion
of farmers’ organizations.

. Diffuse technologies validated by research and teaching institutions, including the
advances of innovative farmers, in order to accelerate farmers’ adoption of
technologies to reduce production risks, and increase yields and ﬁnancial returns.

. Guarantee the quality of technical assistance and training by applying the norms and
regulations for the authorization and registration of professionals, so that they are able
to provide the services to programs coordinated by the SAGAR with the registration
of the “Registro Federal de Trarrrites Empresariales (Entrepreneurial Procedures
Federal Register).

The federal and state government will allocate resources to PEAT that can be
used to provide the following types of ﬁnancial support (SAGARPA 2001a):

. Up to 5% of PEAT budget for training farmers.

. Up to 10% for consulting, studies, and projects.

78

. Up to to 4% for promoting the formation of Professional Services Enterprises and
Development Agencies.
. Up to 4% for operation expenditures.
. Up to 10% for the formulation of a Management Program for Training Professionals,
and a monitoring system for the network of technical assistance.
. The remainder will be allocated to pay for professional technical assistance services
provided by the technicians individually or through development agencies.
. Federal support for all purposes will not exceed 75%, and the rest will be covered by
state governments.
Appendix G describes the requirements for eligibility.
4.2.4 The National Service for Seed Inspection and Certiﬁcation
The “Servicio Nacional de Inspeccion y Certiﬁcacion de Semilla” (SNICS)
(National Service for Seed Inspection and Certiﬁcation), an organization of the
Secretariat of Agriculture, is in charge of controlling and supervising the ﬁilﬁllment of
the legal regulations regarding seeds and plant varieties. The Mexican government has
established the following three fundamental activities for the SNICS (SAGARPA 2002):
. Verify and certify the origin and the quality of the seeds.
0 Provide legal protection to the rights of those who generate improved plant varieties.
. Coordinate actions regarding the use of genetic materials for food and agriculture.
Other important activities of the SNICS include promoting soil preservation,
sustainable agriculture, and genetic diversity in crops of economic interest. Therefore, the
activities in educational centers and government organizations related to phytogenetic

resources are coordinated by the SNICS.

79

Technological advances in agriculture have generated many improved varieties.
Certiﬁcation implies the inspection of the origin of the seeds and verifying that improved
varieties were produced under the established strict standards. Only seed that meets the
requirements of high genetic quality, physiologic, physical and phytosanitary are certiﬁed
by SNICS. In that way, SNICS ensures that certiﬁed seed acquired by farmers is of high
yielding varieties.

The generation of improved varieties requires expensive resources. Therefore, to
promote agricultural research and technology transfer, the “Ley Federal de Variedades
Vegetales” (Vegetable Varieties Federal Law) was established to protect the rights of
breeders. Breeders and organizations can obtain legal protection and secure the right to
payment of royalties from those multiplying the variety. SNICS is in charge of the legal
procedure for securing the breeder’s patent and verifying that the improved varieties are
different, uniform, and stable. The requirements to apply for a breeder’s patent are
included in Appendix H.

4.2.5 The National Producer of Seeds

Productora Nacional de Semillas (PRONASE) (National Producer of Seeds), an
organization of the federal government, was created by the “Law of Production,
Certiﬁcation, and Trade of Seeds” in 1960. The institutional purpose of the PRONASE is
to promote the production and utilization of certiﬁed seeds. The methods of genetic
maintenance and technical management by which the seed is produced and marketed by
the PRONASE ensure that its seed have high levels of genetic, physical, physiological,

and phytosanitary quality.

80

The general objectives of the PRONASE are (SAGARPA 2002):

. To contribute to agricultural development and national food self-sufﬁciency by
enhancing the production and utilization of certiﬁed seed of improved plant varieties
that will allow farmers to increase their production and productivity.

- To be a proﬁtable organization that through the best quality, prices, services and
edge-technology increases its market share in certiﬁed seed market.

Since 1996 the PRONASE has participated in the Alliance for the Countryside,
providing certiﬁed seed to the Kilo per Kilo program. PRONASE is charged with
meeting the needs of all of the country’s different agroecological conditions. Therefore,
PRONASE multiplies seed of a wide range of crops (e.g., maize, beans, oats, sorghum,
wheat) and multiplies a diversity of hybrids and varieties that are adapted to different
agroecological conditions. In addition, it multiplies a broad diversity of vegetable species
and varieties that are suitable for mild, hot, cold, and extreme weather. In order to
provide good service, PRONASE has licensed seed dealers throughout the country at
which farmers can purchase the seeds.

PRONASE is required to meet the quality standards established by the SNICS.
Quality control includes supervision of the stages of production, storage, preservation,
and services to dealers. At the national level, at all of its agroindustrial plants, PRONASE
has laboratories equipped for testing the seeds, according to the methods established by
international organizations like the International Seed Testing Association (ISTA) and the
Association of Official Seed Analysts (AOSA). PRONASE also provide external seed

analysis and preservation services to dealers and farmers.

81

Most of PRONASE’s certiﬁed seed are bred by scientists at INIFAP and other
research centers in the country. PRONASE is in charge of multiplying the seed from
small quantities to the commercial levels. To ensure quality, reproducing seed in the
higher categories (basic and registered) is carried out directly in PRONASE’s plots. To
ensure a high level of genetic quality, the basic seed is subjected to a process of genetic
maintenance. The seed obtained through this process is used to produce registered seed,
under strict supervision by PRONASE’s specialized technicians.

Registered seed is used to produce certiﬁed seed through contracts with individual
farmers. According to the contract, PRONASE should offer farmers:

. Production credit, up to harvest.

. Technical supervision from specialized technicians.

. An economic incentive and purchase guarantee, if the seed meets the quality
standards established by the quality control department of the PRONASE and those
of the SNICS.

Seed that meets the quality standards is delivered to PRONASE’s plants for cleaning,
classiﬁcation, and chemical treatment. This preparation process guarantees the physical,
physiological, phytosanitary, and genetic quality that is required for certiﬁed seed. After
SNICS veriﬁes that the seed meets the expected quality standards, it provides labels of
certiﬁcation that must be attached to the seed sacks, which is a guarantee to farmers that

the seed is of high quality.

82

4.3 Seed Distribution in the North of Mexico

4.3.1 National Program for Seed Distribution

In 1996, as component of the Alliance for the Countryside umbrella program, the
Mexican government started a seed distribution program known as the Kilo per Kilo
program. The objective of the program was deﬁned as:

“Promoting the technological change and the shift to more productive crops under both
rainfed and irrigated conditions through the substitution of certiﬁed seed of improved
varieties for traditional varieties, and promoting the introduction of new species that can
take advantage of the different agricultural regions in the country (SAGARPA 2001a)”.

The program targets ejidatarios (small common property), small private farmers,
and farmers’ associations in areas recommended by studies carried out by INIFAP.

The program distributes seed of maize, beans, rice, oats, sorghum, amaranth,
safﬂower, sunﬂower, canola, groundnuts, triticale and chickpeas, as well as other
crops—intended to shift production into more productive crops—approved by the
Directorate General of Agriculture (DGA).

Under the current (2000) program’s regulations, a federal subsidy covers up to
45% of the difference between the price of the improved seed and the price of
commercial grain. In addition, the state governments are allowed to cover up to 55% of
the price difference. The prices of the commercial grain that is used to determine farmers’
contribution to the program is formally published by the “Comite Tecnico del
Fideicomiso” (Technical Committee of the Trust), based on the date when the seed is
delivered to farmers. Computation of these prices is supported by information on market
prices generated by the “Distritos de Desarrollo Rural” (Districts of Rural Development)

and by other government organizations in charge of keeping records of commodity

prices.

83

According to the regulations, a farmer may receive support for a maximum of ﬁve
hectares. The amount of the federal subsidy is based on the quantity of seed per hectare
that is recommended in the technological package. The support is provided per crop
season. The number of years a farmer can receive the subsidy is determined by the
Technical Committee of the Trust.

In order to participate in Kilo per Kilo, farmers must meet the following
requirements:

. Fill out an application for the desired type of improved seed.

. Demonstrate they are agricultural producer by showing written proof from the county
administration or from their farmers’ organization.

. Submit a letter of commitment to make the necessary investments and follow the
complementary practices required by the program.

. Comply with the phytosanitary campaigns in the region and participate in the
technical assistance programs.

. Specify in the application the crop area (maximum of ﬁve hectares) that is located in
a micro-region with the potential for the use of improved varieties where usually
improved varieties have not been adopted before, or is located in a region where a
shift of production into more productive crops is promoted.

The regulations also require that the program distributes certiﬁed seed. The
varieties and planting densities should be the ones recommended in the technological
packages validated by SAGAR and published in the “Boletin de Variedades

Recomendadas” (Recommended Varieties Bulletin), as well as those that have been

84

validated by the INIFAP and meet the “Lineamientos de Evaluacion” (Evaluation
Standards) for the recommendation of varieties.

In exceptional cases, the Directorate General of Agriculture can approve the
utilization of noncertiﬁed seed. In such cases, the state delegation of SAGAR should ask
to the “Servicio Nacional de Inspeccion y Certiﬁcacion de Semillas” (SNICS) (National
Service of Seed Inspection and Certiﬁcation) to perform of germination, purity, and
sanitation tests in order to ensure that the seed meet the required standards.

4.3.2 Seed Distribution in Chihuahua

In Chihuahua 1,252,000 hectares are cultivated (64% under rainfed conditions and
36% irrigated) by approximately 25,000 farmers in the state—although this number is
decreasing due to migration. The main rainfed crops are beans, maize, oats, sorghum,
wheat, and sunﬂower. During 1900-2000, the harvested dry bean area under rainfed
conditions averaged 149,411 ha, with an average production of 70,729 Mt and an average
yield of 473 kg/ha. The harvested area averaged 79% of the planted area, due to erratic
weather conditions. The value of production averaged MX$ 209.9 million and the
average producer price was MXS 2.97 per kilo (INEGI 1992-2001).

During 1996-2000, the Kilo per Kilo program distributed 899 Mt of bean seed to
6,968 farmers, who planted 30,296 ha (Table 4.6). The annual area planted to improved
varieties averaged 3.1% of the total bean area. The planted area averaged 4.3 ha per
farmer and each farmer received an average of 129 kg of seed (30 kg per hectare). The
quantity of seed distributed increased from 1996 to 1998, but decreased from 1998 to
2000, depending on the annual budget assigned to the Kilo per Kilo program. The

number of beneﬁciaries and the area planted to MVs varied with the amount of seed

85

distributed. The total investment of SAGAR in the program during 1996-2000 was MXS

10,644,000 and the cost of the seed averaged MX$ 11.85 per kilo (SAGARPA 2001b).

Table 4.6 Statistics of the Kilo per Kilo Program for Beans in Chihuahua 1996-2000,

 

 

 

 

 

 

 

 

 

 

 

 

Mexico.
Item 1996 1997 1998 1999 2000
Bean seed (Mt) 120 272 339 102 65.5
Beneﬁciaries (farmers) 818 1,993 2,501 1,179 477
Area (hectares) 4,308 8,885 11,302 3,632 2,171
Investment (000 MX$) 1,200 2,720 4,273 1,534 917
Quantity of seed (kg/ha) 27.8 30.6 29.9 28.1 30.1

 

Source: SAGAR (2001a).

In 1996, the Kilo per Kilo program distributed 120 Mt of certiﬁed seed of the
variety Pinto Villa in the state of Chihuahua. In 1997, the program distributed a total of
272 Mt of improved bean seed. Pinto Villa accounted for 96% of the seed distributed and
Pinto Americano accounted for 4% of the total (Table 4.7). Two categories of seed were
distributed: certiﬁed seed and non-certiﬁed seed. In the case of Pinto Villa, 64% was
certiﬁed seed and 36% was non-certiﬁed. In the case of Pinto Americano, 69% was

certiﬁed seed and 31% non-certiﬁed.

Table 4.7 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program 1997,

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Pinto Villa 154,385 Certiﬁed 5,042
Pinto Villa 84,700 Non-certiﬁed 2,766
P. Americano 22,800 Certiﬁed 745
P. Americano 10,161 Non-certiﬁed 332
Total 272,046 N.A. 8,885

 

 

 

 

 

Source: SAGAR (2001a).

86

 

In 1997, the program distributed 272 Mt of improved bean seed, which was
planted on 8,885 ha. SAGAR estimated that due to the distribution of improved seed via
the Kilo per Kilo program, bean yields of participants farming under rainfed conditions
increased from 0.51 to 0.68 Mt/ha (33.3%) and their production increased from 4,089 to
5,488 Mt (34%), compared to non-participants (SAGAR-Chihuahua 1998).

In 1998, the program distributed a 339 Mt of improved bean seed. Pinto Villa
accounted for 82% of the seed distributed and the varieties pinto Bill Z and Pinto
Americano accounted for 18% of the total (Table 4.8). Both certiﬁed seed and non-
certiﬁed seed were distributed. In the case of Pinto Villa, 80% was certiﬁed seed and
20% was non-certiﬁed seed. In the case of Bill 2 and Pinto Americano, all of the seed
was certiﬁed. The average price was MX$ 15.46 per kilo for certiﬁed seed and MXS 5.70

for non-certiﬁed seed.

Table 4.8 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program 1998,

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Pinto Villa 211,900 Certiﬁed 7,063
Pinto Villa 65,150 Non-certiﬁed 2,172
Bill Z 35,000 Certiﬁed 1,167
P. Americano 27,000 Certiﬁed 900
Total 339,050 NA. 11,302

 

 

 

 

 

Source: SAGAR (2001a).

In 1998, total investment in the Kilo per Kilo program for dry beans was MX$
4,273,000. By purchasing the seed, farmers covered 38.5% (MXS 1,644,392) of the total
cost and the government covered the remaining 61.5% (MXS 2,629,216). The budget

assigned to diffusion activities was $ 415,710 (9.7% of total), and the price farmers paid

87

for improved seed averaged MX$ 4.85 per kilo. SAGAR estimated (SAGARPA 2001b)
that participants yields increased by 32% and the increase in annual income per farmer
averaged MX$ 1,087. The increase in the value of production was estimated at MX$
2,719,300 and the increase in bean production at 565 Mt.

In 1999, the Program distributed 102 Mt of improved bean seed. Pinto Mestizo
accounted for 97% of the seed distributed and Pinto Villa accounted for 3% of the total
(Table 4.9). In both cases, all the seed distributed was certiﬁed seed. The cost for the
government of the certiﬁed seed averaged MX$ 15.00 per kilo. The price of commercial
grain of dry beans (about MX$ 3.40/kg) served as reference to determine the price the
farmers had to pay for the certiﬁed seed. The subsidy of the program covered the

difference between these two prices; therefore, the subsidy averaged MX$ 11.60/kg.

Table 4.9 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program 1999,

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Pinto Mestizo 99,249 Certiﬁed 3,524
Pinto Villa 3,034 Certiﬁed 108
Total 102,283 NA. 3,632

 

 

 

 

 

Source: SAGAR (2001a).

In 2000, the Program distributed 65.5 Mt of improved bean seed. Pinto Mestizo
accounted for 75.7% of the seed distributed, Negro Altiplano 19.9%, and Pinto Villa
4.4% of the total (Table 4.10). In all cases, the seed distributed was certiﬁed seed. The
average price for the certiﬁed seed was MX$ 14.00 per kilo (7% lower than in 1999). The
price of commercial grain of dry beans was MX$ 5.00/kg. Thus, the subsidy averaged

MX$ 9.00/kg.

88

Table 4.10 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program
2000, Mexico.

 

 

 

 

 

 

Variety Quantity (kg) Category Area (ha)
Pinto Mestizo 49,554 Certiﬁed 1,643
Negro Altiplano 13,040 Certiﬁed 432
Pinto Villa 2,882 Certiﬁed 96
Total 65,476 N.A. 2,171

 

 

 

 

 

Source: SAGAR (2001a).

The amount of seed distributed in Chihuahua during 1996-2000 is summarized in
Table 4.11. Certiﬁed seed accounted for 82 percent and non-certiﬁed seed accounted for
18 percent of the total. Together, Pinto Villa and Pinto Mestizo accounted for 88 percent
of certiﬁed seed and other varieties for 12 percent. Pinto Villa also accounted for 94
percent of the non-certiﬁed seed and other varieties for 6 percent. Pinto Villa and Pinto
Mestizo accounted for 90 percent of the total seed distributed and other varieties for 10

percent.

Table 4.11 Bean Seed Distribution in Chihuahua through the Kilo per Kilo Program
1996-2000, Mexico.

 

 

 

 

 

 

 

 

Variety Certiﬁed ﬂ») Noncertiﬁed (%) Total (%)

Pinto Villa 492,201 67% 149,850 94% 642,051 73%
Pinto Mestizo 148,803 21% 0 0% 148,803 17%
Pinto Americano 32,961 5% 10,161 6% 43,122 5%
Bill Z 35,000 5% 0 0% 35,000 4%
Negro Altiplano 13,040 2% 0 0% 13,040 1%
Total 722,005 100% 160,01 1 100% 882,016 100%

 

 

 

 

 

 

 

Source: SAGAR (2001a).

4.3.3 Seed Distribution in Durango

In Durango 642,000 hectares are cultivated (79% under rainfed conditions and

21% irrigated) by approximately 80,000 farmers—although this number is decreasing

89

 

due to migration. The main rainfed crops are beans, maize, oats, and wheat. During 1990-
2000, the harvested dry bean area under rainfed conditions averaged 235,497 ha, with an
average production of 109,745 Mt and an average yield of 466 kg/ha. The harvested area
averaged 85% of the planted area—due to erratic weather conditions. The value of
production averaged MX$ 297.2 million and the average producer price was MX$ 2.71
per kilo (INEGI 1992-2001).

During 1996-2000, the Kilo per Kilo program distributed 6,270 Mt of bean seed
to 40,534 farmers, who planted 196,061 ha (Table 4.12). The annual area planted to
improved varieties averaged 14.1% of the total bean area. The planted area averaged 4.0
ha per farmer and each farmer received an average of 127 kg of seed (32 kg per hectare).

The quantity of seed distributed increased from 1996 to 1998, but decreased from
1998 to 2000, depending on the annual budget assigned to the Kilo per Kilo program.
The number of beneﬁciaries and the area planted to improved varieties varied with the
amount of seed distributed. The total investment of SAGAR in the program during 1996-
2000 was MX$ 43,416,000 per year and the cost of the seed averaged MX$ 6.92 per kilo

(SAGARPA 2001c).

Table 4.12 Statistics of the Kilo per Kilo Program for Beans in Durango 1996-2000,

 

 

 

 

 

 

 

 

 

 

 

 

Mexico.
Item 1996 1997 1998 1999 2000
Bean seed (Mt) 479 1,511 2,290 1,353 637
Beneﬁciaries Lfarmers) 4,061 10,517 15,177 14,193 5,586
Area (hectares) 15,966 39,976 76,065 42,952 21,102
Investment (000 MX$) 1,521 5,241 15,388 14,533 6,733
Quantity of seed (kg/ha) 30.0 37.8 30.1 31.5 30.1

 

Source: SAGAR (2001b).

9O

 

In 1996, the Program distributed 479 Mt of Pinto Villa (Table 4.13), of which 3%
was certiﬁed and 97% was non-certiﬁed seed. The certiﬁed variety was Flor de Mayo

and the non-certiﬁed variety was Pinto Villa.

Table 4.13 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1996,

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Flor de Mayo 15,000 Certiﬁed 500
Pinto Villa 464,000 Non-certiﬁed 15,466
Total 479,000 NA. 15,966

 

 

 

 

 

Source: SAGAR (2001b).

In 1997, the Program distributed a total of 1,511 Mt of improved bean seed
(Table 4.14), of which 17% was certiﬁed and 83% was non-certiﬁed seed. The certiﬁed
varieties were Flor de Mayo and Pinto Villa; and the non-certiﬁed varieties were Flor de

Mayo, Negro Jamapa, and Pinto Villa.

Table 4.14 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1997,

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Flor de Mayo 181,320 Non-certiﬁed 4,797
Flor de Mayo 120,880 Certiﬁed 3,198
Negro Jamapa 846,160 Non-certiﬁed 22,387
Pinto Villa 226,650 Non-certiﬁed 5,996
Pinto Villa 135,990 Certiﬁed 3,598
Total 1,511,000 N.A. 39,976

 

 

 

 

 

Source: SAGAR (2001b).

In 1998, the Program distributed a total of 2,290 Mt of improved bean seed (Table

4.15), of which 10% was certiﬁed and 90% was non-certiﬁed seed. The certiﬁed varieties

91

were Pinto Villa, Negro Altiplano, Flor de Mayo, and Negro Sahuatoba; and the non-

certiﬁed varieties were Pinto Nacional and Negro Criollo.

Table 4.15 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1998,

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Pinto Nacional 1,401,480 Non-certiﬁed 46,552
Negro Criollo 659,520 Non-certiﬁed 21,907
Pinto Villa 95,498 Certiﬁed 3,172
Negro Altiplano 20,518 Certiﬁed 682
Flor de Mayo 47,966 Certiﬁed 1,593
Negro Sahuatoba 65,018 Certiﬁed 2,160
Total 2,290,000 N.A. 76,065

 

 

 

 

 

Source: SAGAR (2001b).

In 1999, the Program distributed a total of 1,353 Mt of improved bean seed (Table
4.16), of which 37% was certiﬁed and 63% was non-certiﬁed seed. The certiﬁed varieties
were Pinto Mestizo, Pinto Villa, Pinto Bayacora, Azufrado Namiquipa, and Negro

Altiplano; and the non-certiﬁed seed varieties were Negro Queretaro and Pinto Nacional.

Table 4.16 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1999,

 

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Negro Queretaro 378,840 Non—certiﬁed 12,027
Pinto Nacional 473,550 Non-certiﬁed 15,033
Pinto Mestizo 136,175 Certiﬁed 4,323
Pinto Villa 232,226 Certiﬁed 7,372
Pinto Bayacora 84,047 Certiﬁed 2,668
A. Namiquipa 15,494 Certiﬁed 492
Negro Altiplano 32,669 Certiﬁed 1,037
Total 1,353,000 N.A. 42,952

 

 

 

 

Source: SAGAR (2001b).

92

 

In 2000, the Program in Durango distributed a total of 633 Mt of improved bean
seed (Table 4.17), of which 58% was certiﬁed and 42% was non-certiﬁed seed. The
certiﬁed varieties were Negro Altiplano, Negro Sahuatoba, Pinto Mestizo, and Pinto
Bayacora; and the non-certiﬁed varieties were Negro Altiplano, Negro San Luis, and

Pinto Mestizo.

Table 4.17 Bean Seed Distribution in Durango through the Kilo per Kilo Program 2000,

 

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Negro Altiplano 23,478 Certiﬁed 778
Negro Altiplano 25,539 Non-certiﬁed 846
Negro San Luis 117,680 Non-certiﬁed 3,898
Negro Sahuatoba 24,365 Certiﬁed 807
Pinto Mestizo 273,360 Certiﬁed 9,056
Pinto Mestizo 124,320 Non-certiﬁed 4,118
Pinto Bayacora 48,258 Certiﬁed 1,599
Total 637,000 N.A. 21,102

 

 

 

 

 

Source: SAGAR (2001b).

In 2000, the total investment in the Kilo for Kilo program for beans was MX$
6,733,000, of which 46% was provided by the federal government, 15% by the state
government, and 39% by farmers. SAGAR estimated that the yield of participants
increased an average of 44% and their increase in annual income averaged MX$ 1,465
per hectare (SAGARPA 2001c). The increase in the value of production was estimated at
MX$ 30,924,830 and the increase in bean production at 5,521 Mt.

The amount of seed distributed in Durango during 1996-2000 is summarized in
Table 4.18. Certiﬁed seed accounted for 22 percent and non-certiﬁed seed accounted for
78 percent of the total. Pinto Villa, Pinto Mestizo, Flor de Mayo and Pinto Bayacora

accounted for 87 percent of certiﬁed seed and other varieties for 13 percent. Pinto

93

Nacional, Negro Criollo, Negro Jamapa, and Pinto Villa accounted for 93 percent of the

non-certiﬁed seed and other varieties for 7 percent. Pinto Nacional, Negro Criollo, Pinto

Villa and Negro Jamapa accounted 79 for percent of the total seed distributed and other

varieties for 21 percent.

Table 4.18 Bean Seed Distribution in Durango through the Kilo per Kilo Program 1996-
2000, Mexico.

 

 

 

 

 

 

 

 

 

 

 

 

 

Variety Certiﬁed (%) Noncertiﬁed (%) Total (%)

Pinto Villa 463,714 34% 690,650 14% 1,154,364 18%
Pinto Mestizo 409,535 30% 124,320 3% 533,855 9%
Flor de Mayo 183,846 13% 181,320 4% 365,166 6%
Pinto Bayacora 132,305 10% 0 0% 132,305 2%
Negro Sahuatoba 89,383 7% 0 0% 89,383 1%
Negro Altiplano 76,665 6% 25,539 1% 102,204 2%
Azuf. Namiquipa 15,494 1% 0 0% 15,494 0%
Negro Jamapa 0 0% 846,160 17% 846,160 13%
Pinto Nacional 0 0% 1,875,030 38% 1,87 5,030 30%
Negro Criollo 0 0% 1,156,040 24% 1,156,040 18%
Total 1,370,942 100% 4,899,059 100% 6,270,001 100%

 

 

 

 

 

 

 

Source: SAGAR (2001b).

4.3.4 Seed Distribution in Zacatecas

In Zacatecas 1,343,690 hectares are cultivated (90% under rainfed conditions and

10% irrigated) by approximately 46,000 farmers—although this number is decreasing

due to migration. The main rainfed crops are beans, maize, and oats for fodder. During

1990-2000, the harvested dry bean area under rainfed conditions averaged 577,891 ha,

with an average production of 262,160 Mt and an average yield of 454 kg/ha. The

harvested area averaged 84% of planted area—due to erratic weather conditions. The

value of production averaged MX$ 688.7 million and the average producer price was

MX$ 2.63 per kilo (INEGI 1992-2001).

94

 

During 1996-2000, the Kilo per Kilo program distributed 4,908 Mt of bean seed

to 79,707 farmers, who planted 158,168 ha (Table 4.19). The annual area planted to

improved varieties averaged 4.6% of the total bean area. The planted area averaged 2.0

ha per farmer, and each farmer received an average of 62 kg of seed (31 kg per hectare).

The quantity of seed distributed increased from 1996 to 1998, but decreased from 1998 to

2000, depending on the budget assigned to the Kilo per Kilo program in each year. The

number of beneﬁciaries and the area planted to improved varieties varied with the

amount of seed distributed. The total investment of SAGAR in the Program during 1996-

2000 was MX$ 36,864,000 per year and the cost of the seed averaged MX$ 7.51 per kilo

(SAGARPA 2001a).

Table 4.19 Statistics of the Kilo per Kilo Program for Beans in Zacatecas 1996-2000,

 

 

 

 

 

 

 

 

 

 

 

 

Mexico.
Item 1996 1997 1998 1999 2000
Bean seed (Mt) 641 808 2,473 481 504
Beneﬁciaries 4,750 6,522 21,421 29,997 17,017
Area (hectares) 18,330 29,063 76,277 17,481 17,017
Investment (000 MX$) 3,785 4,844 14,511 6,907 6,817
Quantity of seed (kg/ha) 34.9 27.8 32.4 27.5 29.6

 

Source: SAGAR (2001c).

In 1996, the Program distributed 641 Mt of improved bean seed (Table 4.20), of

which 4% was certiﬁed and 96% was non-certiﬁed seed. The certiﬁed varieties were Flor

de Mayo and Pinto Americano; and the non-certiﬁed varieties were Manzano, Bayo

Blanco, Pinto Nacional, Negro San Luis, and Pinto Villa.

95

 

Table 4.20 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 1996,

 

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Manzano 1 9, 793 Non-certiﬁed 566
Flor de Mayo RMC 15,350 Certiﬁed 439
P. Americano 10,000 Certiﬁed 286
Bayo Blanco 8,000 Non-certiﬁed 229
Pinto Nacional 421,493 Non-certiﬁed 12,053
Negro San Luis 122,274 Non-certiﬁed 3,497
Pinto Villa 44,089 Non-certiﬁed 1,261
Total 640,970 N.A. 18,330

 

 

 

 

 

Source: SAGAR (20010).

In 1997, the Program distributed 808 Mt of improved bean seed (Table 4.21), of
which 14% was certiﬁed and 86% was non-certiﬁed seed. The certiﬁed varieties were
Flor de Mayo RMC, Negro Zacatecas, and Pinto Villa; and the non-certiﬁed varieties

were Pinto Nacional, Flor de Junio, and Negro San Luis.

Table 4.21 Bean Seed distribution in Zacatecas through the Kilo per Kilo Program 1997,

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Flor de Mayo RMC 51,307 Certiﬁed 1,844
Flor de Junio 2,810 Non-certiﬁed 101
Pinto Nacional 539,530 Non-certiﬁed 19,394
Negro Zacatecas 2,475 Certiﬁed 89
Pinto Villa 56,113 Certiﬁed 2,017
Negro San Luis 156,287 Non-certiﬁed 5,618
Total 808,522 N.A. 29,063

 

 

 

 

 

Source: SAGAR (20010).

SAGAR-Zacatecas (1997) estimated that the yields of program participants under

rainfed conditions increased by 34 percent for farmers planting certiﬁed seed, and by 26

percent for farmers who planted non-certiﬁed seed. Considering that 3,950 ha were

96

planted to certiﬁed seed and 25,113 ha to non-certiﬁed seed, the estimated increase in
production is 5,991 Mt with a value of approximately MX$ 25 million.

In 1998, the Program distributed 2,473 Mt of improved bean seed (Table 4.22), of
which 11% was certiﬁed and 89% was non-certiﬁed seed. The certiﬁed varieties were
Flor de Mayo RMC, Negro Zacatecas, and Pinto Villa; and the non-certiﬁed varieties

were Flor de Mayo, Pinto Nacional, Manzano, Flor de Junio, and Negro San Luis.

Table 4.22 Bean seed distribution in Zacatecas through the kilo per kilo program 1998,

 

 

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Flor de Mayo RMC 261,785 Certiﬁed 8,073
Flor de Mayo 160,692 Non-certiﬁed 4,956
Pinto Nacional 1,749,347 Non-certiﬁed 53,947
Manzano 61,201 Non-certiﬁed 1,887
Flor de Junio 69,096 Non-certiﬁed 2,131
Negro Zacatecas 3,075 Certiﬁed 95
Pinto Villa 3,375 Certiﬁed 104
Negro San Luis 164,858 Non-certiﬁed 5,084
Total 2,473,429 NA. 76,277

 

 

 

 

 

Source: SAGAR (2001c).

In 1999, the Program distributed 482 Mt of improved bean seed (Table 4.23), of
which 48% was certiﬁed and 52% was non-certiﬁed seed. The certiﬁed varieties were
Flor de Mayo RMC, Flor de Mayo UAZ”, Flor de Junio UAZ’86, Bayo Zacatecas,
Manzano, Pinto Villa, and Garbancillo Supremo; and the noncertiﬁed seed varieties were

Negro San Luis and Pinto Nacional.

 

n UAZ is the acronym for Universidad Autonoma Zacatecas.

97

Table 4.23 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 1999,

 

 

 

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Flor de Mayo RMC 200,420 Certiﬁed 7,276
Flor Mayo UAZ 4,784 Certiﬁed 174
Flor de Junio UAZ'86 4,640 Certiﬁed 168
Negro San Luis 252,050 Non-certiﬁed 9,150
Bayo Zacatecas 3,800 Certiﬁed 138
Manzano 1, 560 Certiﬁed 57
Pinto Villa 9,909 Certiﬁed 360
Pinto Nacional 1,075 Non-certiﬁed 39
Garbancillo S. 3,300 Certiﬁed 120
Total 481,538 N.A. 17,481

 

 

 

 

 

Source: SAGAR (2001c).

In 2000, the Program distributed 505 Mt of improved bean seed (Table 4.24), of
which 62% was certiﬁed and 38% was non-certiﬁed seed. The certiﬁed varieties were
Flor de Mayo Sol, Negro Zacatecas, Flor de Mayo RMC, and Bayo Durango; and the

non-certiﬁed varieties distributed were Pinto Villa, and Negro San Luis.

Table 4.24 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 2000,

 

 

 

 

 

 

 

 

 

Mexico.
Variety Quantity (kg) Category Area (ha)
Flor de Mayo Sol 170,620 Certiﬁed 5,755
Pinto Villa 147,950 Non-certiﬁed 4,991
Negro Zacatecas 147,950 Certiﬁed 4,458
Negro San Luis 41,250 Non-certiﬁed 1,391
Flor de Mayo RMC 10,500 Certiﬁed 354
Bayo Durango 2,000 Certiﬁed 67
Total 504,490 N.A. 17,017

 

 

 

 

 

Source: SAGAR (2001c).

In 2000, the total investment in the Kilo for Kilo program was MX$ 6,817,470, of
which 46% was provided by the federal government and 54% by the state government.

SAGAR estimated that the yield of program participants increased on averaged by 35%

98

and their increase in annual income averaged MX$ 685 per hectare (SAGARPA 2001d).

The increase in the value of production was estimated at MX$ 11,661,885 and the

increase in bean production at 2,748 Mt.

The amount of seed distributed in Zacatecas during 1996-2000 is summarized in

Table 4.25. Certiﬁed seed accounted for 19 percent and non-certiﬁed seed accounted for

81 percent of the total. Flor de Mayo and Negro Zacatecas accounted for 91 percent of

certiﬁed seed and other varieties for 9 percent. Pinto Nacional and Negro Criollo

accounted for 87 percent of the non-certiﬁed seed and other varieties for 13 percent.

Pinto Nacional, Flor de Mayo, and Negro Criollo accounted for 88 percent of the total

seed distributed and other varieties for 12 percent.

Table 4.25 Bean Seed Distribution in Zacatecas through the Kilo per Kilo Program 1996-
2000, Mexico.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variety Certiﬁed (%) Noncertiﬁed (%) Total (%)

Flor de Mayo 714,766 75% 160,692 4% 875,458 18%
Negro Zacatecas 153,500 16% 0 0% 153,500 3%
Pinto Villa 69,397 7% 192,039 5% 261,436 5%
Pinto Americano 10,000 1% 0 0% 10,000 0%
Flor de Junio 4,640 0% 71,906 2% 76,546 2%
Bayo Durango 2,000 0% 0 0% 2,000 0%
Manzano 1,560 0% 80,994 2% 82,5 54 2%
Pinto Nacional O 0% 2,711,445 68% 2,711,445 55%
Negro Criollo 0 0% 736,719 19% 736,719 15%
Bayo Blanco 0 0% 8,000 0% 8,000 0%
Bayo Zacatecas 0 0% 3,800 0% 3,800 0%
Garbancillo Sup. 0 0% 3,300 0% 3,300 0%
Total 955,863 100% 3,968,895 100% 4,924,758 100%

 

 

 

 

 

 

 

Source: SAGAR (20010).

99

 

4.4 Summary of the Chapter

In the early 19903, the process of price liberalization required the government to
develop new strategies to support agricultural development. Therefore, during the decade
of 19903 the Mexican government, via the Secretariat of Agriculture (SAGAR), started
several new programs to support farmers and raise their standard of living. In 1994, the
government started PROCAMPO, a ﬁfteen-year cash transfer program to help farmers
make the transition from a government-regulated to a market-driven agricultural sector.
As a complement to PROCAMPO, in 1996 the Federal Government started “Alianza para
el Campo” (Alliance for the Countryside), a broad strategy for promoting agricultural
production and rural development. For bean farmers, one of the most relevant
components of Alliance for the Countryside is the Kilo per Kilo program, which has the
objective to increase yields by promoting the substitution of modern varieties for
traditional varieties. Since its inception, Kilo per Kilo program has shown positive
results, as indicated by increased adoption rates and yields. Until 1993, bean prices were
determined mainly by the strong intervention of the government through guaranteed
prices. This system operated via CONASUPO, a parastatal that purchased commodities
from farmers. But starting in 1993, the government initiated important changes to reduce
government intervention and allow prices to be determined according to a freer
interaction of market supply and demand. However, under the new price policy of free
prices, growing beans under rainfed conditions is not sustainable without the direct
support of PROCAMPO. While Mexico has signed trade agreements with different
countries, the most important agreement (for most of the products including dry beans)

has been the North American Free Trade Agreement (NAP TA). Under the NAFTA,

100

import quotas and tariffs for imports of beans from the US. and Canada into Mexico are
being phased out over a fourteen-year period. By 2008, dry beans will be tariff-exempt
and they will be imported freely. From 1994 to 1999 there was a signiﬁcant increase in
the amount of beans imported, mainly from the US. During 1990-2000, bean production
was partially supported by rural credit and crop insurance. BANRURAL provided public
credit to 11.7% of the total bean area and crop insurance covered 8.2% of total bean area.
Several government institutions have been created to increase agricultural
productivity. The Agricultural Research and Technology Transfer Program, a component
of Alliance for the Countryside, is oriented to increase productivity by providing farmers
access to new technologies and by promoting agricultural research (mainly through
INIFAP), diffusion, and technical assistance. INIFAP’s (National Research Institute for
Forestry, Livestock, and Agriculture) objective is to improve the standards of living of
farmers and the population in general by generating technologies to increase agricultural
production and productivity. Among other activities, INIFAP scientists have been
generating improved varieties for the most important crops in Mexico, including beans.
The “Basic Program of Technical Assistance” (PEAT), a component of the Alliance for
the Countryside, promotes the use of technologies generated by research institutions and
improve the quality of technical assistance services. The “National Service for Seed
Inspection and Certiﬁcation” (SNICS), an organization of SAGAR, is in charge of
verifying and certifying the origin and the quality of the seeds generated by research,
provide legal protection to the rights of those who generate improved plant varieties, and
coordinate actions regarding the use of genetic materials for food and agriculture. The

“National Producer of Seeds” (PRONASE), an organization of the federal government,

101

promotes the production and utilization of certiﬁed seeds. PRONASE is in charge of
multiplying the seed from small quantities to the commercial levels. Most of
PRONASE’s certiﬁed seed is bred by scientists at INIFAP and other research centers in
the country. Since 1996 the PRONASE has participated in Alliance for the Countryside
by providing certiﬁed seed to the Kilo per Kilo program, which distributes certiﬁed seed
(but also noncertiﬁed seed) to farmers at a price equivalent to commercial grain for a
maximum of ﬁve hectares. A federal subsidy covers up to 45% of the difference between
the price of the improved seed and the price of commercial grain. In addition, the state
governments are allowed to cover up to 55% of the price difference. This signiﬁcant
reduction in price makes improved seed very affordable to farmers.

During 1996-2000, the Kilo per Kilo program distributed 12,077 Mt of bean seed
to 127,209 farmers, which planted 384,525 ha in Chihuahua, Durango and Zacatecas in
northern Mexico. SAGAR estimated that average yields increased in the range from 26%
to 44%. The quantity of seed distributed increased from 1996 to 1998, but decreased from
1998 to 2000, depending on the annual budget assigned to the Kilo per Kilo program.
SAGAR’s total investment of in the program during 1996-2000 was MX$ 90,924,000
and the cost of the seed averaged MX$ 7.53 per kilo.

Two categories of seed were distributed: certiﬁed seed (25% of the total) and non-
certiﬁed seed (7 5%), although some varieties were distributed under both categories. The
certiﬁed varieties were PintoVilla (34%), Flor de Mayo (29%), Pinto Mestizo (18%), and

other (19%).

102

The non-certiﬁed varieties were Pinto Nacional (51%), Negro Criollo (21%),
Pinto Villa (11%), and other (17%). Pinto Villa and Pinto Mestizo were distributed
mainly in Durango and Chihuahua. Pinto Nacional and Flor de Mayo were distributed
mainly in Zacatecas and Durango. The improved varieties Pinto Villa, Pinto Mestizo, and

Flor de Mayo were generated by INIFAP.

103

CHAPTER 5. IMPROVED VARIETIES ADOPTION ANALYSIS

This chapter presents the analysis of the cross-sectional data obtained ﬁ'om the
farmer surveys in Chihuahua, Durango, and Zacatecas. Sections one to four report
descriptive statistics for the data collected from farmers. These sections also report
hypothesis testing results, using t-test for two means, chi-square for counts or percents,
and analysis of variance (AN OVA) for comparison of several means. Section ﬁve
analyzes the relationship between adoption and participation in the seed distribution
program, and reports the results of the econometric analysis for explaining the factors
associated with the participation of farmers in the Kilo per Kilo program. Section six

presents a summary of the chapter.

5.1 Characteristics of Farmers and Farms in Northern Mexico

5.1.1 Characteristics of the Farmers

For this analysis, the farmers were divided into two groups: non-adopters and
adopters. Non-adopters are deﬁned as those farmers planting only traditional varieties in
Spring-Summer 2001. Adopters are deﬁned as those farmers planting improved
varieties”, including some farmers who planted both improved and traditional varieties.
In Durango, the number observations was 206; non-adopters were 88 (43%), and adopters
were 118 (57%). In Chihuahua, the number of observations was 163; non-adopters were
27 (17%), and adopters were 136 (83%). In Zacatecas, the number of observations was
86; non-adopters were 33 (3 8%), and adopters were 53 (62%). The total number of

observations was 455; adopters were 307 (67%), and non-adopters were 148 (3 3%).

 

‘2 Improved bean varieties included Negro Sahuatoba, pinto Bill Z, Pinto Bayacora, Flor de Mayo M3 8,
Negro Altiplano, Negro Zacatecas, Pinto Mestizo, and Pinto Villa.

104

Table 5.1 reports the mean and the standard deviation for age of the farmers
(years), years of experience as a farmer, years of education, and the percentage of farmers
who farm under private property versus common property (ejido tenure) rights.

Regarding the age of farmers, compared to non-adopters, adopters were younger

in Durango (47 vs. 49 years) and Chihuahua (47 vs. 57 years), and same age in Zacatecas

(51 years). The difference is statistically signiﬁcant (5% level) only for Chihuahua.

Regarding years of experience, adopters have less experience in Durango (26 vs.

31 years), Chihuahua (22 v3. 31 years), and Zacatecas (25 vs. 29 years). The differences

are statistically signiﬁcant for Durango and Chihuahua.

Regarding years of schooling, adopters have attended school longer in Durango (7

vs. 6 years), Chihuahua (7 vs. 5 years), and Zacatecas (6 vs. 5 years). The difference is

statistically signiﬁcant only for Durango.

Table 5.1 Characteristics of Farmers in Durango, Chihuahua, and Zacatecas 2001,

 

 

 

 

 

 

 

Mexico.

Category Durango (n=206) Chihuahua (n=163) Zacatecas (n=86)
Mean Std Dev Mean Std Dev Mean Std Dev

Age (yeaIS) (*)
Nonadopters 49.10 14.29 56.85 13.13 51.00 12.55
Adopters 47.30 14.65 46.94 10.82 51.08 11.61
Experience (years) (*) (*)
Nonadopters 31.01 15.92 31.40 19.14 29.28 14.10
Adopters 26.47 14.96 21.72 12.78 24.52 12.22
Education (years) (*)
Nonadopters 6.06 3.42 5.26 3.85 4.90 2.94
Adopters 7.08 3.80 7.27 6.11 5.88 5.44
Owner (percent)
Nonadopters 47% N/A“ 26% N/A 22% N/A
Adopters 45% N/A 43% N/A 25% N/A

 

 

 

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the t-test.

a. N/A means not applicable.

105

 

Regarding the percentage of landowners, adopters account for a lower percentage

of owners in Durango (45 vs. 47%) and a higher percentage in Chihuahua (43 vs. 26%)

and Zacatecas (25 vs. 22%). However the differences are not statistically signiﬁcant.

5.1.2 Characteristics of the Farms

Table 5.2 reports the mean and the standard deviation for farm size, bean area,

improved variety area, and the percentage of farmers having good soil, based on their

own perception.

Regarding farm size (ha), compared to non-adopters, adopters have smaller farms

in Durango (31 vs. 38 ha) and larger farms in Chihuahua (39 vs. 31 ha) and Zacatecas (31

vs. 22 ha). However the differences are not statistically signiﬁcant.

Regarding the bean area (ha), adopters farm a smaller bean area in Durango (27

vs. 35 ha) and Chihuahua (24 vs. 29 ha) but a larger bean area in Zacatecas (25 vs. 19

ha). However the differences are not statistically signiﬁcant.

Table 5.2 Characteristics of Farms in Duran

o, Chihuahua, and Zacatecas 2001, Mexico.

 

 

 

 

 

 

 

 

 

 

Category Durango (n=206) Chihuahua (n=163) Zacatecas (n=86)
Mean Std Dev Mean Std Dev Mean Std Dev

Farm size (ha)

Nonadopters 3 7.63 42.20 31.05 59.77 22.26 19.80

Adopters 30.64 32.08 39.02 57.93 30.78 38.36

Bean area (ha)

Nonadopters 34.87 38.08 28.77 60.11 19.26 18.43

Adopters 27.03 28.47 23.57 33.27 24.77 35.89

IVs area (ha)

Nonadopters N/A N/A N/A N/A N/A N/A

Adopters 15.17 17.42 14.49 15.67 11.30 16.41

Good soil (percent)

Nonadopters 42% N/A 26% N/A 6% N/A

Adopters 46% N/A 38% N/A 12% N/A

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the chi-square or t-test.

106

 

Regarding the area planted to improved varieties (ha), adopters plant a larger area
to improved varieties (IVs) in Durango (15 ha) than in Chihuahua (14 ha) and Zacatecas
(11 ha).

Regarding the percentage of farmers judging to have good soil, adopters have a
higher percentage in Durango (46 vs. 42%), Chihuahua (38 vs. 26%), and Zacatecas (12
vs. 6%). However the differences are not statistically signiﬁcant.

5.1.3 Sources of Technical Information

For Durango, additional information was collected regarding the means by which
the farmer learned about the existence of the new improved bean varieties (Table 5.3).
Among adopters, their most important source of technical information was SAGAR
(64%) i. e., the Secretariat of Agriculture, followed by other farmers (57%), technicians
(48%), INIFAP (40%), demonstration plots (40%), and brochures containing technical
information (3 5%). Although almost all percentages are higher for adopters than for non-

adopters, these differences are not statistically signiﬁcant.

Table 5.3 Percentage of Farmers by Source of Information, Durango 2001, Mexico.

 

 

 

 

 

 

 

 

 

 

 

 

Source of Information Nonadopters Adopters
Radio/T V 15% 17%
Newspaper 9% 1 1%
Technical brochure 30% 35%
Other farmer 44% 57%
Demonstration plots 37% 40%
SAGAR 52% 64%
Insurance company 0% 1%
Bank 4% 3%
INIFAP 33% 40%
Technicians 38% 48%

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the chi-square test.

107

5.2 Crop Practices, Production Costs, and Capital

5.2.1 Crop Practices and Chemical Inputs

Table 5.4 shows the percentage of non-adopters and adopters who used soil

preparation (plow and harrow), were able to obtain seed of the variety that they wanted to

plant, their mean bean planting date, and their perception of the number of years with bad

weather (out often).

In Durango, the same percentage (94%) of adopters and non-adopters plowed

their ﬁelds, while a higher percentage of adopters in Chihuahua (90 vs. 85%) and

Zacatecas (94 vs. 84%) plowed their ﬁelds. However, the differences are not statistically

signiﬁcant.

Table 5.4 Soil Preparation and Planting Date in Durango, Chihuahua, and Zacatecas 2001

 

 

 

 

 

 

 

 

Mexico.

Category Durango Q=206) Chihuahua (n=163) Zacatecas (n=86)
Mean Std Dev Mean Std Dev Mean Std Dev

Plow (percent)
Nonadopters 94% N/A 85% N/A 84% N/A
Adopters 94% N/A 90% N/A 94% N/A
Harrow (percent)
Nonadopters 99% N/A 96% N/A 84% N/A
Adopters 98% N/A 97% N/A 85% N/A
Got seed (percent) (*)
Nonadopters 79% N/ A 52% N/ A 1 9% N/ A
Adopters 57% N/ A 5 5% N/ A 3 1% N/A
Planting datea
Nonadopters 199.58 9.70 176.00 21.59 189.28 20.07
Adopters 200.84 12.73 180.93 17.53 187.65 20.16
Bad years (percent)
Nonadopters 3.91 2.80 5.63 3.16 4.66 2.29
Adopters 4.60 2.62 5.68 2.40 6.29 3.01

 

 

 

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the chi-square test.
a. Days after January 1, 2001.

108

 

In Durango, a lower percentage of adopters (98 vs. 99%) harrowed their ﬁelds,
whereas a higher percentage of adopters in Chihuahua (97 vs. 96%) and Zacatecas (85 vs.
84%) harrowed. However, the differences are not statistically signiﬁcant.

In Durango, a lower percentage (57 vs. 79%) of adopters were able to get the seed
they wanted, while a higher percentage of adopters in Chihuahua (55 vs. 52%) and
Zacatecas (31 vs. 19%) were able to get the seed they wanted. However, the difference is
statistically signiﬁcant only for Durango.

In Durango, adopters and non-adopters reported the same average planting dates
(Jul 19). In contrast, adopters planted later (Jun 30 vs. Jun 25) in Chihuahua and they
planted earlier in Zacatecas (Jul 6 vs. Jul 8). However the differences are not statistically
signiﬁcant.

While in Durango and Zacatecas adopters perceived that drought had occurred
more frequently in the past ten years (Durango 4.6 vs. 3.9 years; Zacatecas 6.3 vs. 4.7
years), both groups reported the same average in Chihuahua (5.7 years). However, the
differences are not statistically signiﬁcant.

Table 5.5 shows the percentage of adopters and non-adopters who applied
fertilizer, did mechanical weeding (weeding 1 and weeding 2), handweeding, and applied
insecticide.

In Durango (50 vs. 45%) and Chihuahua (88 vs. 85%), a higher percentage of
adopters fertilized their bean crop, whereas in Zacatecas a lower percentage of adopters
(67 vs. 91%) applied fertilizer. However, the difference is statistically signiﬁcant only in

Zacatecas.

109

Table 5.5 Cultural Practices and Chemical Inputs in Durango, Chihuahua, and Zacatecas
2001, Mexico.

 

 

 

 

 

 

 

 

 

 

 

Category Durango (n=206) Chihuahua (n=163) Zacatecas (n=86)
Mean Mean Mean

Fertilizer (percent) (*)

Nonadopters 45% 85% 91%

Adopters 50% 88% 67%

Weeding 1 (percent)

Nonadopters 100% 85% 97%

Adopters 99% 76% 88%

Handweed (percent)

Nonadopters 94% 89% 91%

Adopters 88% 80% 73%

Weeding 2 (percent)

Nonadopters 98% 78% 84%

Adopters 96% 68% 7 1%

Insecticide (percent) (*) (*)

Nonadopters 2% 59% 22%

Adopters 10% 78% 21%

 

(*) Indicates differences at the 5% level of signiﬁcance in the chi-square test.

While ﬁrst weeding was done by a lower percentage of adopters in Durango (99
vs. 100%), Chihuahua (76 vs. 85%), and Zacatecas (88 vs. 97%), the differences are not
statistically signiﬁcant. Handweeding was carried out by a lower percentage of adopters
in Durango (88 vs. 94%), Chihuahua (80 vs. 89%), and Zacatecas (73 vs. 91%), but the
differences are not statistically signiﬁcant. Second weeding was carried out by a lower
percentage of adopters in Durango (96 vs. 98%), Chihuahua (68 vs. 7 8%), and Zacatecas
(71 vs. 84%), but differences are not statistically signiﬁcant.

Insecticide was applied by a higher percentage of adopters in Durango (2 vs.
10%) and Chihuahua (78 vs. 59%), but by a lower percentage of adopters in Zacatecas

(21 vs. 22%). The differences are signiﬁcant in Durango and Chihuahua.

110

 

5.2.2 Production Costs, Credit, and Equipment

Table 5.6 reports adopters’ and non-adopters’ average cost of production per

hectare, the percentage of farmers who received credit, the average amount of credit, and

the percentage of farmers who owned a tractor and a tied-ridger.

Table 5.6 Production Costs, Credit, and Equipment in Durango, Chihuahua, and
Zacatecas 2001, Mexico,

 

 

 

 

 

 

 

 

 

 

 

Category Durango (n=206) Chihuahua (n=163) Zacatecas (n=86)
Mean Std Dev Mean Std Dev Mean Std Dev

Prod Cost (MX$/ha)

Nonadopters 1,714 419 1,366 1,167 1,386 1,236

Adopters 1,634 357 1,608 1,064 1,480 1,189

Credit (percent)

Nonadopters 28% N/A 30% N/A 3 1% N/A

Adopters 34% N/A 21% N/A 21% N/A

Amount (MX$/ha)

Nonadopters 373 1,237 714 1,783 898 2,586

Adopters 314 669 574 2,334 1,255 5,757

Tractor (percent) (*)

Nonadopters 70% N/A 59% N/A 63% N/A

Adopters 75% N/A 82% N/A 62% N/A

Tied-ridger (percent)

Nonadopters 34% N/A 22% N/A 13% N/A

Adopters 46% N/A 26% N/A 1 5% N/A

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the chi-square test.

In Durango, average production costs were lower for adopters (MX$ 1,634 vs.

1,714), but higher for adopters in Chihuahua (MX$ 1,609 vs. 1,366) and Zacatecas (MX$

1,480 vs. 1,386). However, the differences are not statistically signiﬁcant.

In Durango, a higher percentage of adopters received credit (34 vs. 28%), whereas

a lower percentage of adopters received credit in Chihuahua (21 vs. 30%) and Zacatecas

(21 vs. 31%). However, the differences are not statistically signiﬁcant.

lll

 

On average, adopters received less credit in Durango (MX$ 315 vs. 374) and
Chihuahua (MX$574 vs. 714), but larger loans in Zacatecas (MX$ 1,255 vs. 898).
However, the differences are not statistically signiﬁcant.

Tractors were owned by a higher percentage of adopters in Durango (75 vs. 70%)
and Chihuahua (82% vs. 59%), but a lower percentage in Zacatecas (62% vs. 63%). The
difference was statistically signiﬁcant only in Chihuahua.

A tied—ridger was owned by a higher percentage of adopters in Durango (46 vs.
34%), Chihuahua (26% vs. 22%), and Zacatecas (15% vs. 13%), but the differences are
not statistically signiﬁcant.

5.2.3 Family Labor and Hired Labor

Table 5.7 reports the amount of total labor available to adopters and non-adopters
including family and hired labor to grow beans, the number of family members having
off-farm employment, and the number of months per year that the farmer and his wife
work in an off-farm employment.

Less family labor is used by adopters in Durango (2.6 vs. 3.2 persons), but more
is used by adopters in Chihuahua (2.3 vs. 2 persons) and Zacatecas (2.6 vs. 2.2 persons).
The difference is statistically signiﬁcant only for Durango.

More hired labor is used by adopters in Durango (2.7 vs. 1.7 workers) and
Zacatecas (2.1 vs. 1.5 workers), but the same amount is used by both groups in
Chihuahua (0.89 workers). The difference is statistically signiﬁcant only for Durango. In
total, adopters used more labor—family plus hired labor—than non-adopters in the three

states. However, differences are not statistically signiﬁcant.

112

Table 5.7 Family Labor, Hired labor, and Off-farm Employment in Durango, Chihuahua,
and Zacatecas 2001, Mexico.

 

 

 

 

 

 

 

 

 

 

 

 

Category Durango (n=206) Chihuahua Q1=163) Zacatecas (n=86)
Mean Std Dev Mean Std Dev Mean Std Dev

Family labor (No.) (*)

Nonadopters 3 .24 1.92 2.04 1.48 2.22 1.41

Adopters 2.62 1.69 2.31 1.57 2.60 1.82

Hired labor (No.) (*)

Nonadopters 1.78 2.58 0.89 1.80 1.56 3 .47

Adopters 2.77 3.21 0.89 1.51 2.10 3.16

Total labor (No.)

Nonadopters 5.02 3.19 2.93 2.43 3 .78 3 .61

Adopters 5.39 3.66 3.20 2.36 4.69 3.39

Off-farm emp. (No.)

Nonadopters 0.82 1.11 2.44 3.95 1.25 1.74

Adopters 1.10 1.28 1.26 1.81 1.37 1.94

Farmer (months)

Nonadopters 4.17 4.80 8.19 4.80 8.06 4.54

Adopters 5.03 5.55 7.52 4.27 6.42 4.13

Wife (months) (*) (*)

Nonadopters 1.13 3.52 4.52 5.61 5.91 5.68

Adopters 3.00 5.03 3.84 4.99 2.69 4.27

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the t-test.

More family members have off-farm employment in Durango (1.1 vs. 0.8
persons) and Zacatecas (1.3 vs. 1.2 persons), but fewer adopters have off-farm
employment in Chihuahua (1.2 vs. 2.4 persons). However, differences are not statistically
signiﬁcant.

Adopters worked more months in off-farm employment in Durango (5.0 vs. 4.1
months), but fewer months in Chihuahua (7.5 vs. 8.1 months) and Zacatecas (6.4 vs. 8.0
months). However, the differences are not statistically signiﬁcant.

The wives of adopters worked more months in off-farm employment in Durango

(3.0 vs. 1.1 months), but fewer months in Chihuahua (3.8 vs. 4.5 months) and Zacatecas

113

 

(2.6 vs. 5.9 months). The differences are statistically signiﬁcant in Durango and

Zacatecas.

5.3 Program Participation, Storage and Prices of Beans

5.3.1 Participation in Government Programs

The govemment’s Kilo per Kilo seed distribution program is farmers’ primary

source of improved bean seed and PROCAMPO is a cash transfer program. Table 5.8

reports the percentage of adopters and non-adopters (2001) who participated in these

programs and characteristics of their participation.

Of the sampled farmers, the highest percentage of adopters participating in the

Kilo per Kilo program was in Durango (65%), followed by Zacatecas (45%) and

Chihuahua (3 5%).

Table 5.8 Program Participation in Durango, Chihuahua, and Zacatecas 2001, Mexico.

 

 

 

 

 

 

 

 

 

 

 

Category Durango (n=206) Chihuahua (n=163) Zacatecas (n=86)
Mean Std Dev Mean Std Dev Mean Std Dev

Kilo per Kilo (%)

Nonadopters N/A N/A N/A N/A N/A N/A

Adopters 65% N/A 35% N/A 45% N/A

Bean seed (kg)

Nonadopters N/A N/A N/A N/A N/A N/A

Adopters 150.03 225.89 44.44 112.30 30.69 50.75

Seed price (MX$/kg)

Nonadopters N/A N/A N/A N/A N/A N/A

Adopters 4.84 0.96 4.90 1.02 4.87 1.79

PROCAMPO (%)

Nonadopters 100% N/A 92% N/A 97% N/A

Adopters 97% N/A 99% N/A 96% N/A

PROCAMPO (has) (*)

Nonadopters 33.52 35.97 9.38 7.00 16.01 12.35

Adopters 28.17 29.94 26.10 41.46 24.91 36.96

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the t-test.

114

 

Participants in Durango received the largest amount of seed (150 kg), followed by
farmers in Chihuahua (44.4 kg), and Zacatecas (30.7 kg).

In all states, farmers paid about about the same price for a kilo of seed (Durango,
MX$ 4.84; Chihuahua, MX$4.90; and Zacatecas, MX$ 4.87).

While almost all adopters and non-adopters participated in PROCAMPO, a lower
percentage of adopters in Durango (97 vs. 100%) and Zacatecas (96 vs. 97%) participated
in PROCAMPO, and a higher percentage participated in Chihuahua (99 vs. 92%).
However, these differences are not statistically signiﬁcant.

In Durango adopters registered fewer hectares in PROCAMPO than non-adopters
(28 vs. 33 ha), but they registered more hectares in Chihuahua (26 vs. 9 ha) and
Zacatecas (25 vs. 16 ha). However, the differences are statistically signiﬁcant only for
Chihuahua.

5.3.2 Storage of Beans and Farm Location

Table 5.9 reports market-related characteristics of adopters and non-adopters. In
Durango, adopters stored smaller quantities of beans (6 vs. 14 Mt), but they stored more
beans in Chihuahua (9 vs. 3 Mt) and Zacatecas (5 vs. 3 Mt). These differences are
statistically signiﬁcant in Durango and Chihuahua.

Adopters stored their beans for a shorter time in Durango (2.1 vs. 2.2 months) and
Chihuahua (2.0 vs. 2.3 months), but for a longer time in Zacatecas (2.44 vs. 2.40 Mt).
However, these differences are not statistically signiﬁcant.

Adopters’ farms were farther from the city in Durango (29 vs. 21 km) and
Zacatecas (32 vs. 26 km), but closer to the city in Chihuahua (29 vs. 32 km). The

difference is statistically signiﬁcant only for Durango.

115

Table 5.9 Storage and Farm Location in Durango, Chihuahua, and Zacatecas 2001,

 

 

 

 

 

 

 

Mexico.

Category Durango (n=206) Chihuahua (n=163) Zacatecas (n=86)
Mean Std Dev Mean Std Dev Mean Std Dev

Quantity Stor. (Mt) (*) (*)
Nonadopters 14.45 33.67 2.81 6.07 3.17 3.61
Adopters 6.36 13.27 8.98 15.24 5.45 9.06
Stor. Time (months)
Nonadopters 2.25 2.59 2.35 3.46 2.40 3.58
Adopters 2.14 2.49 1.99 1.87 2.44 2.77
Farm distance (km) (*)
Nonadopters 21.15 14.81 31.78 35.46 26.06 19.63
Adopters 29.21 23.56 28.69 44.80 31.92 22.61
Travel time (min)
Nonadopters 34.86 20.83 47.04 41.17 46.31 31.49
Adopters 37.95 17.77 34.77 31.06 51.44 31.19

 

 

 

 

 

 

 

(*) Indicates differences at the 5% level of signiﬁcance in the t-test.

Adopters reported longer travel times to their primary market in Durango (38 vs.

35 min) and Zacatecas (51 vs. 46 min), but shorter travel times in Chihuahua (35 vs. 47

min). However, these differences are not statistically signiﬁcant.

5.3.3 Selling Prices of Beans by Market Class

During the survey, data were collected about the prices at which the farmer sold

their beans (by market class). Following the 2000 harvest, farmers in Chihuahua (n=191)

received the highest average selling price—MX$ 5.18 per kg, with a standard deviation

of MX$ 1.70/kg (Table 5.10). For purposes of comparison, beans were grouped into the

market classes (light color, pinto, and black beans).

116

 

Table 5.10 Selling Prices of Beans by Market Class in Chihuahua Mexico, 2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

Variety name Market Mean N Standard Group}
classa (IVIXS/lg) Deviation

1 Canario L 7.00 2 1.41 a
2 Bayo L 6.75 2 1.77 ab
3 Flor de Mayo L 6.52 4 1.76 abc
4 Bill Z (*) P 6.14 7 0.63 abcd
5 Pinto Mestizo (*) P 5.36 25 1.10 bcd
6 Mantequilla L 5.13 8 1.46 cd
7 Ojo de Cabra P 4.99 20 1.38 cd
8 Pinto Villa (*) P 4.86 100 1.01 d
9 Negro Criollo B 4.83 3 2.02 d
10 Pinto Nacional P 4.71 20 1.12 (1
Total N/A 5.18 191 1.70 N/A

 

 

 

 

 

 

 

(*) Improved bean varieties.
a. L=ligth, P=pinto, B=black.
b. Groups of means in homogenous subsets at the 0.05 level, AN OVA.

In general, farmers received the highest prices for light-colored beans [e.g.,
Canario (MX$ 7.00), Bayo (MX$ 6.75), Flor de Mayo (MX$ 6.52), and Mantequilla
(MX$ 5.13)]. They received an intermediate range of prices for pinto beans [e.g., Bill Z
(MX$ 6.14), Pinto Mestizo (MX$ 5.36), Ojo de Cabra (MX$ 4.99), Pinto Villa (MX$
4.86), and Pinto Nacional (MX$ 4.71)]. Farmers received the lowest prices are for black
beans [e.g., Negro Criollo (MX$ 4.83)]. The analysis of variance (ANOVA) rejected the
null hypothesis of equal means (p-value<0.00005). The multiple comparisons displayed
in the last column of the table indicates that the mean difference is not signiﬁcant among
the varieties from one through four, from two through ﬁve, from three through seven, and
from four through ten, but is signiﬁcant between those varieties that do not belong to the

same group (e. g., variety one and variety eight).

117

Farmers in Durango (n=3 94) received an average selling price of MX$ 4.05 per
kg, with a standard deviation of MX$ 1.20/kg (Table 5.11). For purposes of comparison,

beans were grouped into the market classes (light color, pinto, and black beans).

Table 5.11 Selling Prices of Beans by Market Class in Durango Mexico, 2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

Variety name Market Mean N Standard Groupb
classa (MX$/kg) Deviation

1 Bayo L 5.75 4 1.75 a
2 Flor de Mayo L 4.73 35 0.96 b
3 Pinto Nacional P 4.68 34 1.24 b
4 Canario L 4.61 40 1.39 b
5 Pinto Mestizo (*) P 4.42 38 1.20 b
6 Pinto Bayacora (*) P 4.42 10 0.83 b
7 Pinto Villa (*) P 4.01 120 1.03 bc
8 Negro Sahuatoba (*) B 3.29 10 0.75 c
9 Negro Criollo B 3.27 94 0.92 c
10 Negro Altiplano (*) B 3.24 9 0.90 c
Total N/A 4.05 394 1.20 N/A

 

 

 

 

 

 

 

(*) Improved bean varieties.
a. L=ligth, P=pinto, B=black.
b. Groups of means in homogenous subsets at the 0.05 level, ANOVA.

In general, farmers received the highest prices for light-colored beans [e.g., Bayo
(MX$ 5.75), Flor de Mayo (MX$ 4.73), and Canario (MX$ 4.61)]. They received an
intermediate range of prices for pinto beans [e.g., Pinto Nacional (MX$ 4.68), Pinto
Bayacora (MX$ 4.42), Pinto Mestizo (MX$ 4.42), and Pinto Villa (MX$ 4.01)]. Farmers
received the lowest prices for black beans [e.g., Negro Sahuatoba (MX$ 3.29), Negro
Criollo (MX$ 3.27), and Negro Altiplano (MX$ 3.24)]. The analysis of variance
(ANOVA) rejected the null hypothesis of equal means (p-value<0.00005). The multiple
comparisons displayed in the last column of the table indicates that the mean difference is

not signiﬁcant among the varieties from two through seven and from seven through ten,

118

but is signiﬁcant between the variety one and the varieties from two through ten, and
between the varieties from two through six and the varieties from eight through ten (as
variety seven belongs to groups b and cits mean is only signiﬁcantly different from the
ﬁrst variety).

In Zacatecas (n=119), farmers received an average selling price of MX$ 3.86 per
kg with a standard deviation of MX$ 1.33/kg (Table 5.12). For purposes of comparison,

beans were grouped into the market classes (light color, pinto, and black beans).

Table 5.12 Selling Prices of Beans by Market Class in Zacatecas Mexico, 2000.

 

 

 

 

 

 

 

 

 

 

Variety name Market Mean N Standard Groupb
classall (MX$/kg) Deviation

1 Flor de Junio L 4.57 13 1.89 a

2 Flor de Mayo L 4.43 20 1.63 a

3 Negro Zacatecas (*) B 4.24 5 1.13 ab

4 Bayo L 4.15 13 0.97 ab

5 Canario L 3.88 4 0.75 ab

6 Pinto Nacional P 3.76 8 1.08 ab

7 Negro Criollo B 3.41 56 1.08 b
Total N/A 3.86 119 1.33 N/A

 

 

 

 

 

 

 

(*) Improved bean varieties.
a. L=ligth, P=pinto, B=black.
b. Groups of means in homogenous subsets at the 0.05 level, ANOVA.

In general, farmers received the highest prices for light-colored beans [e.g., Flor
de Junio (MX$ 4.57), Flor de Mayo (MX$ 4.43), Bayo (MX$ 4.15), and Canario (MX$
3.88)]. They received an intermediate range of prices for pinto beans [e.g., Pinto Nacional
(MX$ 3.76). Farmers received the lowest prices for black beans [e.g., Negro Zacatecas
(MX$ 4.24) and Negro Criollo (MX$ 3.41)]. The analysis of variance (ANOVA) rejected
the null hypothesis of equal means (p-value=0.020). The multiple comparisons displayed

in the last column of the table indicates that the mean difference is not signiﬁcant among

119

the varieties from one through six and from three through seven, but is signiﬁcant
between the ﬁrst two varieties and the last variety because they are in different group (a
and b).

As noted previously, the mean prices farmers received varied greatly from state-
to-state (e. g, Chihuahua, MX$ 5.18; Durango, MX$ 4.05; and Zacatecas, MX$ 3.86).
Similarly, the prices that farmers received for speciﬁc varieties varied greatly across
states. For example, the price of Pinto Villa—the most commonly market variety——
averaged MX$ 4.86 in Chihuahua, MX$ 4.01 in Durango, whereas in Zacatecas Negro

Criollo (MX$ 3 .41) was the dominant variety.

5.4 Area Planted to Improved Varieties and Yields

5.4.1 Area Planted to Improved Varieties

The surveyed farmers reported planting several different bean varieties over the
1997-2001 period. In Chihuahua (Table 5.13), modern varieties accounted for 71 percent
of the planted area (ﬁve-year mean). The most widely planted varieties (ﬁve-year mean)
were Pinto Villa (56%), Pinto Mestizo (15%), Ojo de Cabra (9%), Pinto Nacional (8%),
Mantequilla (7%), and other (5%). The varieties that accounted for an increasing share of
planted area during the ﬁve-year period were Pinto Mestizo (+13.5 points), Mantequilla
(+8 points), and other (+1 points). In contrast, the share of the area planted to Pinto Villa
(-17 points), Ojo de Cabra (-5 points), and Pinto Nacional (-0.5 points) declined during
the period. These data suggest that while MVs have maintained their dominant share of

the planted area, farmers are gradually replacing Pinto Villa with Pinto Mestizo.

120

 

Table 5.13 Percentage of Area Planted by Variet in Chihuahua Mexico, 1997-2001.

 

 

 

 

 

 

 

 

 

 

Variety 1997 1998 1999 2000 2001 Mean Changea
Pinto Villa (*) 65 63 58 46 48 56 -17.0
Pinto Mestizo (*) 6 11 13 25 19 15 +135
Ojo de Cabra 11 10 13 4 7 9 -5.0
Pinto Nacional 11 7 5 7 10 8 -0.5
Mantequilla 3 3 7 12 10 7 +8.0
Other 4 6 4 6 6 5 +1.0
Total 100 100 100 100 100 100 ~-
Total hectares 1,948 1,865 2,333 2,520 3,984 12,650 --
Observations 1 18 121 143 167 263 812 --

 

 

 

 

 

 

 

 

(*) Improved bean varieties.

a. Absolute change in 1997-98 mean, versus 2000-01 mean in percent points.

In Durango (Table 5.14), improved varieties accounted for 42 percent of the

planted area (ﬁve—year mean). The most widely planted varieties (ﬁve-year mean) were

Pinto Villa (3 8%), Negro Criollo (25%), Pinto Nacional (12%), Canario (11%), Pinto

Mestizo (4%), and other (10%).

Table 5.14 Percentage of Area Planted by Variety in Durango Mexico, 1997-2001.

 

 

 

 

 

 

 

 

 

 

 

 

Variety 1997 1998 1999 2000 2001 Mean Changea
Pinto Villa (*) 34 43 46 35 33 38 -4.5
Negro Criollo 15 23 31 32 26 25 +100
Pinto Nacional 25 12 7 5 9 12 -11.5
Canario 20 17 7 4 7 11 -13.0
Pinto Mestizo (*) 0 0 4 10 5 4 +7.5
Other 6 5 5 14 20 10 +115
Total 100 100 100 100 100 100 --
Total hectares 573 1,146 4,440 6,451 6,507 19,116 --
Observations 69 103 244 381 418 1,215 --

 

 

 

 

 

 

 

(*) Improved bean varieties.

a. Absolute change in 1997-98 mean, versus 2000-01 mean in percent points.

The varieties that accounted for an increasing share of planted area during the

ﬁve-year period were Negro Criollo (+10 points), Pinto Mestizo (+7.5 points), and other

121

 

 

(+11.5 points). In contrast, the share of the area planted to Pinto Villa (-4.5 points), Pinto
Nacional (-11.5 points), and Canario (-13 points) declined during the period. These data
suggest that while MVs have maintained their dominant share of the planted area, farmers
are gradually replacing Pinto Villa with Pinto Mestizo, and are increasing the area
planted to Negro Criollo (a traditional variety).

In Zacatecas (Table 5.15), modern varieties accounted for only eight percent of
the planted area (ﬁve-year mean). The most widely planted varieties (ﬁve-year mean)
were Negro Criollo (56%), Flor de Mayo (12%), Bayo (10%), Negro Zacatecas (8%),
Pinto Nacional (6%), and other (8%). The varieties that accounted for an increasing share
of planted area during the ﬁve-year period were Flor de Mayo (+5.5 points), Pinto
Nacional (+1 point), and other (+125 points). In contrast, the share of the area planted to
Negro Criollo (-9.5 points), Negro Zacatecas (-5.5 points), and Bayo (—4 points) declined
during the period. These data suggest that while Negro Criollo has maintained its
dominant share of the planted area, farmers are gradually replacing Negro Criollo with

Flor de Mayo.

Table 5.15 Percentaﬁof Area Planted by Variety in Zacatecas Mexico, 1997-2001.

 

 

 

 

 

 

 

 

 

 

 

 

 

Variety 1997 1998 1999 2000 2001 Mean Changem
Negro Criollo 65 53 65 48 51 56 -9.5
Flor de Mayo 7 14 6 14 18 12 +5.5
Bayo 10 16 5 7 11 10 -4.0
Negro Zac. (*) 10 9 11 4 4 8 -5.5
Pinto Nacional 5 4 10 7 4 6 +1.0
Other 3 4 3 20 12 8 +125
Total 100 100 100 100 100 100 --
Total hectares 523 517 1,139 899 2,138 5,215 --
Observations 43 42 67 71 1 56 3 79 --

 

 

 

 

 

 

(*) Improved bean varieties.
a. Absolute change in 1997-98 mean, versus 2000-01 mean in percent points.

122

 

5.4.2 Yields of Traditional and Improved Varieties

The surveyed farmers also reported their yields for each variety that they planted

over the 1997-2001 period. The ﬁve-years of data for Chihuahua, Durango, and

Zacatecas were grouped by variety and analyzed as a pooled data set in order to analyze

the performance of each variety during the period, and the yields of traditional compared

to improved varieties. Farmers’ mean yield (N=2,153)13, averaged 492 kg/ha (Table

5.16), with a standard deviation of 273 kg and a coefﬁcient of variation (CV) of 55%.

Table 5.16 Descriptive Statistics for Bean Yields in Northern Mexico, 1997-2001.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variety name N Mean Std Dev CV. 95% CI. for Mean
(kg/ha) Lower B. Upper B.
Negro Sahuatoba (*) 24 221 143 65% 161 281
Flor de Junio 32 346 151 44% 291 400
Bill Z (*) 21 379 222 59% 277 480
Queretaroa 9 394 1 18 30% 303 485
Bayo Blanco 57 411 230 56% 350 472
Canario 165 414 215 52% 3 80 447
Flor de Mayo 99 440 237 54% 393 487
Pinto Nacional 203 443 251 57% 409 478
Negro Criollo 384 481 245 51% 456 506
Pinto Bayacora (*) 19 482 226 47% 373 591
Flor de Mayo M38 (*) 23 485 161 33% 415 554
FM Media Oreja 60 490 165 34% 447 533
Negro Altiplang’f) 14 491 391 80% 266 717
Negro Zacatecas (*) 20 500 178 36% 417 583
Pinto Mestizo (*) 156 533 327 61% 481 584
Pinto Villa (*) 726 535 282 53% 515 556
Ojo de Cabra 101 561 377 67% 486 635
Mantequilla 40 673 322 48% 570 776
Total 2,153 492 27 3 55% 481 504

 

 

 

 

 

 

 

(*) Improved bean varieties.
a. Also known as bayo Rio Grande.

 

‘3 The sample size is larger than the number of farmers (N=210) because the data set includes data on
yields for the last ﬁve years (1997-2001). Some farmers planted up to three varieties. However, given that

data were collected on a recall basis, there were missing values for earlier years.

123

 

The 95% conﬁdence interval for the mean has a lower bound of 481 and an upper
bound of 504.

For the pinto and black market classes, the yields of improved and traditional
varieties were compared using a t-test. The two improved pintos (Pinto Mestizo and Pinto
Villa) when analyzed together to represent “improved pintos”. Their weighted average
yield (535 kg/ha) was signiﬁcantly higher (p-value<0.00005) than Pinto Nacional (443
kg/ha), with a yield difference of 91.4 kg (20.6%).

In contrast, the two improved black varieties Negro Altiplano and Negro
Zacatecas jointly had a weighted average yield (496 kg/ha) that was not signiﬁcantly
higher (p-value=0.727) than Negro Criollo (481 kg/ha). Furthermore, the improved light
variety Flor de Mayo M38 had a yield (485 kg/ha) that was not signiﬁcantly higher (p-
value=0.388) than Flor de Mayo (440 kg/ha). The other improved varieties were not
analyzed because they had not been widely adopted by the farmers (e. g., Pinto Bayacora),
their yield was lower than traditional varieties (e. g., Negro Sahuatoba), or because they
were not bred by INIFAP (e. g., Bill Z).

These data suggest that while improved pintos have higher yields than traditional
pintos, there is no evidence that improved varieties of other market classes (black and
light-colored beans) have higher yields than their respective traditional variety.

5.4.3 Adopters’ Perception of Improved Varieties Characteristics

The surveyed adopters were asked their perceptions regarding the advantages of
improved varieties compared to traditional varieties (Table 5.17). Based on farmers’
mean response across the total sample (Chihuahua, Durango, and Zacatecas), adopters

perceived that the improved varieties have a better quality for consumption (67%),

124

require less days to maturity (59%), have a better plant type (57%), have higher yields
(54%), have a better market acceptance (51%), are more resistant to diseases (43%), are
more pest resistant (42%), are more drought resistant (41%), are better frost evaders
(31%) (i. e., require fewer days to maturity), and have less threshing problems (29%).
These data suggest that better consumption quality (which is related to market
acceptance), fewer days to maturity (which is related to drought resistance), and higher
yields (which is related to the type of plant) are the most salient characteristics that

inﬂuenced farmers’ decision to adopt improved varieties.

Table 5.17 Percentage of Adopters Considering Improved Better than Traditional
Varieties, Mexico 2001.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Category Durango Chihuahua Zacatecas Mean
(adopters=1 1 8) (adopters=1 36) (adopters=53) (n=3 07)
Yields of grain 52% 59% 51% 54%
Drought resistance 38% 46% 39% 41%
Days to maturity 77% 49% 52% 59%
Market acceptance 47% 59% 47% 51%
Consumption quality 59% 73% 69% 67%
Pest resistance 28% 43% 54% 42%
Frost evasion 24% 33% 36% 31%
Threshing problems 17% 42% 27% 29%
Diseases resistance 43% 40% 47% 43%
Type of plant 49% 66% 55% 57%

 

5.5 Econometric Analysis of Program Participation

As noted previously, the Kilo per Kilo seed distribution program is the driving
force for promoting the adoption of improved bean varieties in northern Mexico. This
section ﬁrst summarizes the relationship between adoption and program participation,

then describes the model and variables used for the econometric analysis, and ﬁnally

125

 

presents the results of the analysis. The analysis is based in the cross-sectional data set for
2001, which was collected during the face-to-face survey in the state of Durango. The
descriptive statistics of the data were previously presented in sections 5.1 to 5.4 of this
chapter.

5.5.1 Adoption of [VS and Program Participation

Farmer adoption of improved bean varieties in northern Mexico is largely
determined by the government’s seed distribution program (Kilo per Kilo), because there
are practically no other sources of improved bean varieties. The program mainly
distributes varieties developed by the INIFAP and other research centers in the country.
Given that beans are “self-pollinated” and farmers do not have to purchase new seed
every year, improved bean varieties have the characteristics of a public good (i.e., a good
that once produced is available to all on a nonexclusive basis). With appropriate
management, farmers can save and share or sell the seed to other farmers, without
experiencing a signiﬁcant reduction in yields from one year to the next. This leads to a
lack of economic incentives, which explains why there is minimal private sector
participation in research to generate improved bean varieties, bean seed production, and
seed marketing—although some farmers make contracts with the government through
PRONASE to produce certiﬁed seed for the Kilo per Kilo program.

Table 5.18 shows that program participation in 2001 explains adoption in 37% of
the cases and that no-program participation explains no-adoption in 43% of the cases,
totaling 80% of the cases. The remaining 20% of the cases are adopters who did not
participate in the program in 2001. However, all of these farmers either save the seed

they planted from the seed that they obtained through the program in a previous year or

126

obtained the seed they planted in 2001 from another farmers who participated in the

program—either in the year of the survey or in previous years.

Table 5.18 Program Participation and Adoption of Improved Varieties in Durango
Mexico, 2001.

 

 

 

 

 

 

 

 

Program participation Improved variety No. of observations Percentage
Yes Yes 77 37%
Yes No 0 0%
No Yes 41 20%
No No 88 43%
Total 206 100%

 

 

 

Source: Survey data.

Considering the crucial role that the program plays in the diffusion of improved
bean varieties, the rest of this section focuses on explaining the factors determining the
participation of farmers in the program.

5.5.2 The Model for the Econometric Analysis

For the econometric analysis of the participation of farmers in the seed
distribution program, a binary dependent variable with yes or no response was regressed
on several independent variables selected from the literature review. Given that the range
of values of the dependent variable is clearly restricted, a limited dependent variable
model was used for the analysis. Limited dependent variable models can be used for time
series and panel data, but they are most often applied to cross-sectional data. The linear
probability model (LPM), logit, and probit models are all appropriate for analyzing data
where the dependent variable is binary. The LPM can be estimated by using Ordinary
Least Squares (OLS) method and the estimates can be interpreted as the changes in the

probability of success, given a one-unit increase in the corresponding explanatory

127

 

variable. The disadvantages of the LPM are that it can produce predicted probabilities
that are less than zero or greater than one, the conditional probabilities increase linearly
with the values of the explanatory variables, and heteroskedasticity (Wooldridge 2000;
Gujarati 1995; Amemiya 1984).

The two nonlinear limited dependent variable models—logit and probit—are
usually estimated by maximum likelihood estimation (MLE). The MLE is consistent,
asymptotically normal, and asymptotically efﬁcient. Both logit and probit models
guarantee that the estimated probabilities lie in the 0-1 interval and that they are
nonlinearly related to the explanatory variables. The disadvantage of nonlinear limited
dependent variable models is that the results are more difﬁcult to interpret because the
change in one independent variable does not affect the probability of success (y=1) in a
linear way. Rather, the change in probability of success for a given change in one
explanatory variable depends both on the change in the variable and on the values of all
the other explanatory variables in the model (Wooldridge 2000; Greene 1997; Kennedy
1992; Gujarati 1995; Maddala 1983).

The probit model can be derived from an underlying latent variable model that
satisﬁes the classical linear model assumptions (W ooldridge 2002). Let y* be an
unobserved, or latent, variable determined by

y*=xB+e, y=1[y*>0], e~Normal (0,1) (1)
where x is a lxK vector with the ﬁrst element equal to unity, and B is a le vector of
parameters. The assumptions are that e is independent of x and that e has a standard
normal distribution. Instead of observing y*, the binary variable y is observed indicating

only the sign of y*. The indicator function 1['] takes on the value of one if the statement

128

in brackets is true and zero otherwise (then y=1 if y*>0 and y=0 if y*=0). From (1) and
the given assumptions, the response probability for y is

P(y=1IX)=P(y*>OIX)=P[e>-XBIX):1@1413]: (1)043) (2)
where <I>(‘) is the standard normal cumulative distribution function.

Considering that the econometric analysis in this dissertation is based on a cross-
sectional data set, that a binary response variable was used for analyzing program
participation, and the advantages of a nonlinear limited dependent variable models, the
probit model was selected as appropriate to analyze farmers participation in the Kilo per
Kilo seed distribution program. The probit model was selected over the logit model
because the probit model’s assumption of normality of the error term is a common
assumption in economics; whereas the logit model assumes a stande logistic
distribution of the error term. However, both models give similar results that are difﬁcult
to distinguish statistically (Wooldridge 2000; Amemiya 1981).

5.5.3 Description of the Variables

To analyze the factors explaining the farmer participation in the Kilo per Kilo
program, the binary dependent variable “program” was deﬁned as equal to one if the
farmer was participating in the seed distribution program in 2001 and zero otherwise. The
explanatory variables selected for the probit model of program particiaption are described
in Table 5.19.

The independent variable “age” measures the age of the farmer in discrete years.
Older farmers may have more experience, greater access to resources, or better social
status—all of which increase their opportunity to participate in the program (CIMMY T

1993). Therefore, the sign on this variable is expected to be positive.

129

The independent variable “bnsarea” measures the total area planted to beans in
hectares. As the program is aimed mainly at small farmers, the sign on this variable is
expected to be negative.

The independent variable “badyears” measures the number of dry years —-out of
the past ten years—according to the farmer’s perception. Farmer’s technology choices are
based on their subjective probabilities (Feder et al. 1985). An increased expectation of
drought may affect positively farmer’s decision to plant improved varieties that are
drought resistant and thereby motivate them to participate in the seed distribution

program. Therefore, the sign on this variable is expected to be positive.

Table 5.19 Identiﬁcation of Variables in the Probit Model for Program Participation in
Durango, Mexico, 2001.

 

 

 

 

 

Variable description Variable name Measurement Exgected sign
Demndent variable
Program participation program Participation in the seed Participation:
distribution program no=0, yes=1
Independent variables
Age of the farmer age Age of the farmer in years (+)
Area planted to beans bnsarea Total bean area in hectares (—)
Weather conditions badyears Dry weather out of ten years (+)
Fertilizer use fertilizer Fertilizer: no=0, yes=1 (+)
Insecticide use pesticid Insecticide: no=0, yes=1 (+)
Tied-ridger use tiedrdgr Tied-ridger: no=0, yes=1 (+)
Credit to grow beans credit Got credit: no=0, yes=1 (+)
Off-farm employment offarme Family members employed (+)
Distance to nearest city distankm Distance in kilometers (+)
Consumption quality consqlty IVs quality: same=0, better=1 Q)

 

The independent variable “fertiliz” equals one if the farmer applies fertilizer and
zero otherwise. The independent variable “pesticid” equals one if farmer applies

insecticide and zero otherwise. Farmers who use modern inputs—like fertilizer and

130

 

insecticide—are more likely to adopt other modern technologies (Feder et al. 1985).
Thus, they may be more likely to participate in the seed distribution program. Therefore,
the sign on these two variables is expected to be positive. Given that fertilizer and
insecticide use depends mainly on weather and environmental conditions, simultaneityl4
was not considered to be a potential problem.

The independent variable “tiedrdgr” equals one if the farmer owns a tied-ridger
and zero otherwise. Farmers who own a tied-ridger may be more likely to adopt a new
technology (CIMMYT 1993), because the program gives preference to enrolling farmers
who use improved soil conservation practices. Therefore, the sign on this variable is
expected to be positive.

The independent variable “credit” equals one if the farmer got credit to grow
beans and zero otherwise. Program participation may be facilitated by credit, especially if
participation requires complementary investments (CIMMYT 1993; Feder et al. 1885).
Therefore, the sign on this variable is expected to be positive. Given that credit is not
provided to buy the improved seed distributed by the program, simultaneity was not
considered to be a potential problem.

The independent variable “offarme” employment measures the number of family
members having an off-farm employment during the year. The program is directed
mainly to support farmers in marginal areas where farmers may have more incentives to
look for off-farm employment to reduce the risk of farming. Therefore, the sign on this

variable is expected to be positive.

 

‘4 Simultaneity means that at least one explanatory variable (e.g. , “fertiliz”) in a multiple regression model
is determined jointly with the dependent variable (e. g., “program”).

131

The independent variable “distankm” measures the distance in kilometers from
the farm to the nearest city (e. g., the county city). The program is aimed mainly at
distributing seed in areas where the IVs have not been adopted before—which are farther
from the cities. Therefore, the sign on this variable is expected to be positive.

The independent variable “consqlty” equals one if the farmer believes that
improved varieties have better consumption quality than traditional varieties and zero
otherwise. Perceiving a better quality for consumption of the improved varieties may
have a positive effect in farmer’s decision to participate in the seed distribution program.
Therefore, the sign on this variable is expected to be positive. Given that consumption
quality depends mainly on preferences in the market, simultaneity was not considered to

be a potential problem.

132

5.5.4 Results of the Analysis

The probit model was estimated by the maximum likelihood estimation (MLE)
method, using the econometric software package Stata 6.0. The results of regressing
“program” on the set of the selected explanatory variables show that ﬁve variables have
coefﬁcients that are statistically signiﬁcant at the 0.05 level (Table 5.20). The coefﬁcient
for “bnsarea” is negative, as expected, indicating that smaller farmers are more likely to
participate in the program than larger farmers.

The coefﬁcients for “badyears”, “pesticid”, “tiedrdgr”, and “distankm” are
positive, as expected, indicating that the farmer’s perception of a higher percentage of dry
years, the use of insecticide, tied-ridger ownership, and a greater distance from the farm
to the city have a positive effect on program participation. The coefﬁcients for the
variables “age”, “fertiliz”, “credit”, “offarme”, and “consqlty” are positive, as expected,

but they are not statistically signiﬁcant at the 0.05 level.

Table 5.20 Estimates of the Probit Model for Participation in the Kilo per Kilo Program
in Durango, Mexico 2001.

 

 

Variables Coefﬁcient Std. Error z-value P>|z]
age 0.0025 0.0075 0.333 0.739
bnsarea -0.0080* 0.0038 -2.080 0.038
badyears 0.1191* 0.0436 2.727 0.006
fertiliz 0.0977 0.2419 0.404 0.686
pesticid 0.8079* 0.3956 2.042 0.041
tiedrdgr 0.4833 * 0.2280 2.120 0.034
credit 0.3799 0.2424 1.567 0.117
offarme 0.1523 0.1046 1.456 0.145
distankm 0.0269* 0.0063 4.246 0.000
consqlty 0.4277 0.2220 1 .926 0.054
constant -2.2140 0.5150 -4.299 0.000

 

 

 

 

 

 

N=184, LR chi2(10)=50.23, Prob>chi2<0.00005, Log likelihood=-95.542146,
Pseudo R2=0.2082

 

* Signiﬁcant at the 0.05 level.

133

 

As a measure of the goodness-of-ﬁt of the regression, the value of the Pseudo R2
is reported. The value of the Pseudo R2 shows that the proportion of the total sample
variation in the dependent variable explained by the regression, corrected for degrees of
freedom, is 0.2082. To test the joint signiﬁcance of the coeﬂicients of the explanatory
variables used in the regression, the p-value for the chi-squared test is also included in
Table 5.20. The p-value shows that the overall level of signiﬁcance for the regression is
less than 0.00005, which implies that the regression is signiﬁcant.

The skewness and kurtosis test for normality was applied on the residuals of the
regression for “program” (pgmres) to evaluate the assumption of normality of the error
term of the probit model and to verify that selecting probit over tobit model was an
appropriate choice. The results of the test (Table 5.21) suggest that there is no problem
with skweness (p-value=0.247) or with kurtosis (p—value=0.43 8). The result of the joint
chi-square test (p-value=0.375) indicates that the null hypothesis of normality of the error

term cannot be rejected.

Table 5.21 Skewness and Kurtosis Test for Normality for Residuals.

 

Variable Pr(Skewness) Pr(Kurtosis) ------ Joint ------
Adj. Chi-sq(2) Pr(chi-sq)

 

 

 

Pgmres 0.247 0.438 1.96 0.375

 

 

 

 

 

 

Therefore the results of the probit model can be considered reliable. Even when
the assumption of normality is not essential for estimation, rejecting the hypothesis of

normality would raise concerns about the validity of hypothesis testing on the coefﬁcients

134

of the model, although the results of the individual and joint hypothesis tests are still
valid asymptotically.

The results of the econometric analysis conﬁrm that in northern Mexico the
participation of farmers in the seed distribution program is associated mainly with farm
size and location. Smaller farmers located in marginal areas are more likely to participate
in the program, which is consistent with the objectives of the Kilo per Kilo program for

the distribution of improved bean varieties.

5.6 Summary of the Chapter

This chapter presents the analysis of the cross-sectional data obtained from the
farmer surveys in Chihuahua, Durango, and Zacatecas. For this analysis, the farmers were
divided into two groups: non-adopters and adopters. Non-adopters are deﬁned as those
farmers planting only traditional varieties in Spring-Summer 2001. Adopters are deﬁned
as those farmers planting improved varieties, including some farmers who planted both
improved and traditional varieties. For Durango (n=206), non-adopters were 43%, and
adopters were 57%. For Chihuahua (n=163 ), non-adopters were 17%, and adopters were
83%. For Zacatecas (n=86), non-adopters were 38%, and adopters were 62%. In total
(n=455), non-adopters were 33% and adopters were 67% in northern Mexico.

When comparing adopters vs. non-adopters, the statistical analysis indicates that
there are few differences regarding age, experience, education, and land tenure. Adopters
are signiﬁcantly younger in Chihuahua and they have more years of education in
Zacatecas. There are not signiﬁcant differences between adopters and non-adopters for

farm size, bean area, area planted to improved varieties, and soil quality. In Durango, the

135

most important sources of technical information are SAGAR (64%), other farmer (5 7%),
technicians (48%), INIFAP (40%), demonstration plots (40%), and technical brochures
(3 5%). No signiﬁcant differences were found between the two groups of farmers. Also
few signiﬁcant differences were found for soil preparation, planting dates, and cultural
labors. Adopters use less fertilizer in Zacatecas and they use more insecticide in
Chihuahua and Durango than non-adopters. There are no signiﬁcant differences for
production costs, credit, and tied-ridger ownership, but adopters have higher percentage
of tractor ownership in Chihuahua. There are no signiﬁcant differences for the number of
family members having an off-farm employment, but adopters use less family labor and
more hired labor than non-adopters in Durango.

Adopters’ participation in the Kilo per Kilo program was 65% in Durango, 35%
in Chihuahua, and 45% in Zacatecas. There were no differences for participation in
PROCAMPO, which was almost 100%.

There are few signiﬁcant differences for quantities and months of storage, and for
farm location (distance to nearest city). Adopters stored smaller quantities of beans in
Durango, but they stored more beans in Chihuahua. Adopters’ farms are farther from the
city in Durango, compared to non-adopters’ farms. Differences in traveling time are not
signiﬁcant.

The analysis of selling price by market class shows that in general, light color
beans have the highest prices, followed by pintos and that the lowest price are for black

beans. However, there were few signiﬁcant differences.

136

The percentage of area planted to improved varieties averaged 71% in Chihuahua
(Pinto Villa 56% and Pinto Mestizo 15%), 42% in Durango (Pinto Villa 38% and Pinto
Mestizo 4%), and 8% in Zacatecas (Negro Zacatecas).

The statistical analysis of the yields indicates that while improved pintos have
higher yields than traditional pintos (20.6%), there is no evidence that improved varieties
of other market classes (black and light-colored beans) have higher yields than their
respective traditional variety.

The surveyed adopters were asked their perceptions regarding the advantages of
improved varieties compared to traditional varieties. The analysis of the data indicates
that better consumption quality, market acceptance, fewer days to maturity (which is
related to drought resistance and frost evasion), and higher yields are the most salient
characteristics that inﬂuenced farmers’ decision to adopt improved varieties.

The descriptive statistics show that the Kilo per Kilo program is the driving force
for the adoption of improved bean varieties in northern Mexico. Therefore, an
econometric analysis of program participation was carried out to investigate the factors
explaining the participation of farmers in the Kilo per Kilo program. A probit model, with
the binary dependent variable “program” to indicate program participation, was estimated
by the maximum likelihood method.

The results of regressing “program” on the set of the selected explanatory
variables indicate that smaller farmers are more likely to participate in the program than
larger farmers. The results also suggest that the farmer’s perception of a higher
percentage of dry years, the use of insecticide, tied-ridger ownership, and a greater

distance from the farm to the city have a positive effect on program participation.

137

The results of the econometric analysis conﬁrm that in northern Mexico the
participation of farmers in the seed distribution program is associated mainly with farm
size and location. Smaller farmers located in marginal areas are more likely to participate
in the program, which is consistent with the objectives of the Kilo per Kilo program for

the distribution of improved bean varieties.

138

CHAPTER 6. BEAN RESEARCH ECONOMIC ANALYSIS

The collaborative research agreement between the INIFAP and the Bean/Cowpea
Collaborative Research Support Program (CRSP) was signed on June I”, 1982, and in
1983 the collaborative breeding program was initiated. During the period 1990-2000,
INIFAP has released several improved bean varieties for rainfed conditions, with the
support of the B/C CRSP. For the purpose of the economic evaluation of bean research,
the relevant improved varieties are Pinto Villa (released in 1990) and Pinto Mestizo
(released in 1996). The statistical analysis of Chapter 5 indicates that in both Chihuahua
and Durango, these two varieties have yields that were signiﬁcantly higher than the
traditional pinto market class variety Pinto Nacional. The other improved varieties
described in Chapter 5 were not considered in the economic assessment because there is
not enough evidence that yields are signiﬁcantly higher than the traditional varieties in
the same market class. Similarly, Zacatecas is not included in the analysis, since the two
improved pintos do not account for a major share of the bean area. However, research
and extension costs for other varieties and for Zacatecas are included because they are
part of the total cost of generating and promoting the improved varieties.

The returns to research investment for improved bean varieties in the north of
Mexico are evaluated using the economic surplus method, which is the most common
approach for analyzing the welfare effects of agricultural research in a partial-equilibrium
framework, i. e. , in a model where the focus is on a single agricultural market and treats
most other economic variables as being exogenous in the analysis. According to Alston et
al. (1998), for most purposes, the partial-equilibrium economic surplus model is the best

available method to evaluate returns to research.

139

6.1 Data for the Analysis

6.1.1 Area, Production and Yields

Considering that the analysis is focused on the improved varieties Pinto Villa and

Pinto Mestizo, only the data for the states of Chihuahua and Durango are used in the
analysis. During the period 1990-2000, while the total area planted to dry beans under
rainfed conditions in Chihuahua and Durango averaged 466,883 ha (CV=14%) (Table
6.1), the area harvested averaged 384,909 ha (CV=24%), which implies a loss of planted
area of 18% per year. During this period, total production averaged 180,474 Mt
(CV=47%) and yields averaged 449 kg/ha (CV=30%). The large year-to-year variations
in area, the high percentage of losses in the planted area, high variability in total

production and yields, and low yields are largely explained by low and variable rainfall.

Table 6.1 Area, Production, and Yields for Dry Beans under Rainfed Conditions in
Chihuahua and Durango Mexico, 1990-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Planted Area Harvested Quantity Mean Yield Area Planted
(ha) Area (ha) (Mt) (kg/ha) to IVs (ha)a
1990 497,819 466,923 185,268 397 --
1991 412,840 392,879 227,933 580 --
1992 324,721 214,841 53,752 250 24,630
1993 47L964 453,722 304,940 672 87,983
1994 497,482 403,293 170,126 422 181,120
1995 505,006 415,247 183,759 443 230,788
1996 524,269 503,383 336,097 668 241,682
1997 490,098 259,226 98,153 379 192,613
1998 537,834 435,090 174,732 402 219,276
1999 479,613 409,142 149,017 364 157,703
2000 394,064 280,249 101,438 362 154,754
Average 466,883 384,909 180,474 469 135,504

 

 

 

 

 

 

 

a. Pinto Villa and Pinto Mestizo represent 86% of the area planted to improved varieties
according to the data collected from the surveys with farmers.
Source: INEGI (1992-2001).

140

Over the last thirty years (1970-2000), annual rainfall in Durango has averaged
459 mm per year (CV=3 0%) and 429 mm in Chihuahua (CV=33%) (CONAGUA 2002).

The last column in Table 6.1 presents the area planted to improved bean varieties
(about 86% are pintos), which shows an increasing trend from 1992 to 1996 and a
decreasing trend from 1996 to 2000. Data from SAGAR were used for area planted to
improved varieties instead of survey data because they include all of the counties where
beans are grown and to avoid an overestimation of the adoption rate by including data
only from those counties where the seed distribution was more concentrated. Data from
seed distribution via Kilo per Kilo were not used to avoid an underestimation of the
adoption rate because the seed that is saved by the farmers and used in subsequent years
is not included. Before the federal government’s Kilo per Kilo program started in 1996,
farmer adoption of improved varieties (1992-1995) is explained by the effort of the state
governments to produce and promote Pinto Villa, which was released in 1990, through
programs speciﬁcally designed to make the improved seed affordable to small farmers. In
fact, those programs provided the background that led to the design of subsequent
programs including the Alliance for the Countryside, and especially Kilo per Kilo.

Data collected from the survey were used to estimate the incremental yields.
Results from Chapter 5 (section 5.4.2) indicate that while there is no evidence that
improved bean varieties of black and light-colored market classes have higher yields than
their respective traditional variety, the yields of improved pintos (Pinto Villa and Pinto
Mestizo) are signiﬁcantly higher than the traditional variety (Pinto Nacional). Based on

the evidence provided by the statistical results (section 5.4.2), the estimate of a 20% yield

141

increase for the improved pintos was used in the analysis. The effect on returns of a
potential reduction in the incremental yield is addressed the sensitivity analysis.

6.1.2 Value of Production and Prices

During the period 1990-2000, the value of production in Durango and Chihuahua
averaged MX$ 507 million (Table 6.2). However, the value of production varied greatly
from year-to-year, due to changes in the quantity produced each year, ranging from a
minimum for 1992 (MX$ 122.92 million) to a maximum in 1996 (MX$ 1,267.13
million). Grower real domestic price averaged MX$ 1,973 per Mt, showed a decreasing
trend over time, and exhibited large annual ﬂuctuations (CV=29%). Considering that
price differences among varieties are not signiﬁcant (Sec. 5.3) and that most of the area is
planted to pinto beans (Sec. 5.4), the average domestic price was used for the ﬁnancial

analysis of the improved pintos.

Table 6.2 Value of Production, Grower Real Domestic Price (1994:100) for Beans in
Chihuahua and Durango Mexico, and FOB price for US. Pinto Beans, 1990-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Production Grower Real FOB for US. FOB for US.
Value (thou— Domestic Price Pinto Beans Pinto Beans
sand MX$) (MX$/Mt) JUS$lcwt) (USS/Mt)
1990 315,272 2,831 36.83 811.23
1991 436,285 2,597 19.76 435.24
1992 122,920 2,684 19.46 428.63
1993 592,117 2,077 27.03 595.37
1994 277,652 1,632 29.02 639.21
1995 381,745 1,539 21.25 468.06
1996 1,267,131 2,078 28.68 631.72
1997 415,116 1,933 25.25 556.17
1998 882,134 1,990 26.31 579.52
1999 462,791 1,050 20.40 449.34
2000 425,469 1,295 19.69 433.70
Average 507,148 1,973 21.04 548.02

 

 

 

 

 

Source: INEGI (1992-2001) and ERS (2002b).

142

 

The FOB price for US. pinto beans required to compute the import parity price

(IPP) —to be used in the economic analysis as the opportunity cost of producing beans

domestically instead of importing them from the U.S.—is also included in Table 6.2. The

FOB price is originally reported in US. dollars per hundredweight (cwt), and was

converted into US. dollars per metric ton for the analysis.

6.1.3 INIFAP and B/C CRSP Research Costs

Research expenditures associated with developing improved bean varieties for

Northern Mexico (under the agreement of the INIFAP with the Bean/Cowpea CRSP)

started in 1983 (Table 6.3). After at least eight years of investing in research and

development, several new varieties were released. For the purpose of the analysis, the

relevant varieties are Pinto Villa (released in 1990) and Pinto Mestizo (released in 1996).

INIFAP’s expenditures in support of its bean research (1983-1996), totaled MX$

14,151,998 (70% of total), while the B/C CRSP’s expenditures totaled MX$ 5,926,378

(3 0% of total).

Table 6.3 Research Cost of Bean Research in Chihuahua, Durango, and Zacatecas

Mexico 1983-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year INIFAP B/C CRSP Year INIFAP B/C CRSP
(MX$) (MX$) (MX$) (MX$)

1983 42,949 7,392 1992 1,206,557 475,242
1984 65,921 14,820 1993 1,304,440 476,657
1985 102,378 28,837 1994 1,395,331 648,220
1986 184,779 76,322 1995 1,672,000 1,025,422
1987 421,895 58,801 1996 2,520,000 1,171,822
1988 713,646 283,492 1997 268,800 122,417
1989 841,660 327,667 1998 375,250 140,121

1990 995,309 371,969 1999 429,750 151,373

1991 1,142,134 399,077 2000 469,200 146,727

 

 

 

Source: INIFAP (2000b) and CRSP (2002).

143

 

After the release of Pinto Mestizo in 1996, only 10% of the expenditures in the
bean research program are included for both the INIFAP and B/C CRSP, which reﬂects
an estimation of the costs required for seed multiplication and for maintaining the quality
of the varieties.

6.1.4 Extension and Adoption Costs

During the period 1996-2000, government investments (SAGAR) in the Kilo per
Kilo seed distribution program in Chihuahua, Durango, and Zacatecas increased from
MX$ 6,506,000 in 1996 to MX$ 34,172,000 in 1998 and then decreased to MX$
14,467,000 in 2000, totaling MX$ 90,924 (Table 6.4). The Kilo per Kilo budget basically
depends on the availability of federal and the state governments resources, the
agreements between the federation and the states, and the govemment’s criteria for
assigning the budgetary resources among the different sectors of the economy and
between programs within the same sector (i.e., according to the priorities established for
PROCAMPO and for the components of Alliance for the Countryside). The budget for
government programs is also affected by the political cycle of the presidency (Zedillo
1994-2000), which usually reﬂects low levels at the beginning and the end of the

administration.

Table 6.4 Budget of the Kilo per Kilo Program in Northern Mexico, 1996-2000.

 

 

 

 

 

 

 

 

Year Thousands of Mexican Pesos (MX$)

1996 1997 1998 1999 2000 Total
Chihuahua 1,200 2,720 4,273 1, 534 917 10,644
Durango 1,521 5,241 15,388 14,533 6,733 43,416
Zacatecas 3,785 4,844 14,51 1 6,907 6,817 36,864
Total 6,506 12,805 34,172 22,974 14,467 90,924

 

 

 

 

 

 

Source: SAGAR (2001a, 2001b, 20010).

144

 

Farm-level adoption costs are the increase in production costs of adopting a new
technology, relative to the unimproved technology. In the case of improved bean
varieties, the new technology involves only a change of variety, with negligible
additional costs compared to other varieties because the price difference between the
improved and the traditional seed is covered by the subsidy of the Kilo per Kilo program.
Therefore, farm-level adoption costs assumed to be zero in the ﬁnancial analysis, but in
the economic analysis, the price subsidy for the improved seed is removed.

6.1.5 Consumer Price Index and Exchange Rate

To convert nominal prices and costs into real terms, the consumer price index
(CPI) is commonly used. Table 6.5 shows that the CPI in Mexico increased considerably
faster in the 19803 than in the 19903. The nominal exchange rate (ER) to convert US.
dollars into Mexican pesos is also included in the Table 6.5, which shows a continuous
increase over the period—raising to more than three times higher in 2000 than in 1992

(when adoption started).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 6.5 Consumer Price Index and Exchan 0 Rate in Mexico, 1983-2000.
Consumer Exchange Consumer Exchange
Year Price Index Rate Year Price Index Rate
(1994:100)“ (MX$/USS) (l 994=100) (MX$/US$)

1983 1.60 0.1221 1992 85.20 3.0907
1984 2.70 0.1700 1993 93.50 3.0999
1985 4.30 0.2830 1994 100.00 4.2157
1986 8.00 0.6585 1995 135.00 6.6688
1987 18.50 1.5941 1996 181.40 7.6209
1988 39.60 2.2489 1997 218.80 7.9613
1989 47.50 2.4755 1998 253.70 9.1127
1990 60.10 2.8102 1999 295.80 9.8444
1991 73.70 3.0150 2000 323.83 9.5423

 

 

a. The base-year for official statistics on CPI in Mexico is currently 1994.
Source: INEGI (1996, 2001).

145

This rapid depreciation in the Mexican peso resulted from Mexico’s decision in
the early 19903 to adopt a ﬂoating exchange rate regime, allowing the ER to ﬂuctuate
freely in response to changing economic conditions.

6.2 Basis of the Economic Analysis

6.2.1 Supply and Demand Elasticities

According to Alston et al. (1998), while total research beneﬁts are relatively
insensitive to elasticities of supply and demand, elasticity assumptions (or estimates) are
much more important in relation to the distribution of beneﬁts. In particular, the more
elastic supply is, relative to demand, the greater the consumer share of total research
beneﬁts (and smaller the producer share) and vice versa. Conclusions obtained using
linear supply and demand functions with a parallel shift apply fairly generally, but in the
case of a pivotal supply shift, whether producers beneﬁt from research depends on the
elasticity of demand—when demand is inelastic producers lose (i. e., all of the beneﬁts go
to consumers). However, Oehmke and Crawford (2002) show that when using the Alston
et al. (1998) approach, returns to research may be very sensitive to the supply elasticity
parameter. In contrast, the returns calculated using the Akino-Hayami (1975) procedure
are relatively invariant to the value of the supply elasticity. Thus, they recommend using
the Alston et al. (1998) approach because of its conceptual and practical advantages, and
to address the sensitivity of research returns to the supply parameter by investing greater
efforts to obtain sound empirical estimates of the supply elasticity.

In this dissertation a value of 0.5 was assumed for the elasticity of supply. This
value is within the range reported by other researchers [e.g., 0.5 for beans by Mather

(1999); 021-035 for agricultural production (Tsakok 1990)], and because most

146

published elasticities of supply for agricultural products fall between 0.1 and 1.0 (Alston
et al. 1998). Also, an absolute value of 0.5 was assumed for the elasticity of demand,
based on the demand elasticity of 0.49 for pulses (beans included) and of 0.45 for coarse
grains (e. g., maize) reported by Tsakok (1990).

6.2.2 Supply and Demand Functional Forms

Linear supply-and-demand curves have been used in the majority of studies that
estimate research beneﬁts. With such curves, the elasticities change as quantity changes
along the curve. When a parallel shift is assumed, the functional form is largely
irrelevant. Thus, a linear model provides a good approximation—regardless of the true
functional form of supply. The main alternative is to use constant elasticity curves that
are typically combined with an assumption of pivotal shift, which is the most important
consequence of the functional form choice. Empirically, the measures of total research
beneﬁts and their distribution between producers and consumers are quite insensitive to
the choices of functional form. They are much more sensitive to the separate choices
concerning the nature of the research-induced supply shift and elasticities (Norton et al.
1998)

According to Masters (1996), linear curves are an appropriate speciﬁcation for
many situations. Following the approaches of Norton et al. (1998) and Masters (1996),
two of the most widely used approaches currently used for returns-to-research
calculations, linear supply-and-demand curves are assumed in this dissertation. In

addition, a pivotal supply shift is assumed as explained in the next section.

147

6.2.3 Research-Induced Supply Shift

The choice of the nature of the research-induced supply shift is crucially
important in the analysis. In contrast, the choices of ﬁinctional forms and elasticities are
not very signiﬁcant (Alston et al. 1998). When using a linear supply function, a parallel
shift almost doubles the total research beneﬁts—compared to a pivotal shift. When the
supply shift is parallel, producers always beneﬁt from research but, as mentioned before,
with a pivotal shift producers beneﬁt only when demand is elastic. When demand is
inelastic, producers necessarily lose from a pivotal supply shift.

Considering that the yield increase resulting from the use of the improved bean
varieties is proportionally the same for different levels of production, then the most
reasonable assumption is that the supply shift must be proportional. Therefore, a pivotal
(or proportional) research-induced supply shift was chosen for the computation of total
research beneﬁts. Furthermore the use of a pivotal shift will always generate conservative
estimates of total economic beneﬁts compared to a parallel shift, but in this case the use
of parallel (nonproportional) research-induced supply is not the most appropriate choice.

6.2.4 The Logistic Adoption Curve

Much of the literature on diffusion assumes that the cumulative proportion of
adoption follows a S-shaped curve in which there is slow initial grth in farmers
adoption of the new technology, followed by a more rapid increase, and then slower
adoption down as the cumulative proportion of adoption approaches its maximum
(Griliches 1957; CIMMYT 1993). The most common function used to represent the

«at-[it

curve is the logistic function given by P=K/(1+e ), where P = the percentage of area

using the new technology at time t, K = the upper bound of adoption (the ceiling), or = a

148

constant, related to the time when adoption begins, and B = a constant, related to the rate
ofadopﬁon.
The equation can be transformed to ln(P/K-P)= or+Bt, which makes it possible to

directly estimate the parameters or and B by ordinary least squares (OLS) regression. The
value of K can be estimated by plotting the data and choosing the level that appears to be
the upper bound of adoption or by running the regression using different values of K and
choosing the one that maximizes R-squared value. In this dissertation, a logistic curve
was used to ﬁt the actual data on adoption and to project future adoption levels for
purposes of the economic evaluation.

6.2.5 Economic Surplus Formulas for a Closed Economy

This section presents the formulas for estimating the social gains from research in
a closed economy model. The closed economy refers to a situation where the price of the
commodity is determined inside the country of interest (Alston et al. 1998). The formulas
and estimations of research beneﬁts for the small open economy model, where the price
is exogenous, are given in section 6.6.

Net economic beneﬁts or net gains are estimated as NG=SG-R-E, where
SG=social gains, R=research costs, and E=extension costs. Social gains for a parallel

shift are estimated as SG=KPQi0.5KPAQ, where K=vertical supply shift, P=observed

product price, Q=observed quantity produced, and AQ=quantity change. The second term
is subtracted when data are observed after adoption (an ex post study) and added if
adoption has not yet ocurred (ex ante study) (Masters 1996). For a pivotal shift, the ﬁrst

term is multiplied by 0.5, as shown below. The supply shift is estimated as K=(J/e)-C,

where J=proportional quantity increase, e=elasticity of supply, and C=proportional cost

149

increase. The proportional quantity increase is estimated as J=(AY/Y)*A where Y=mean
yield, AY=yield increase, and A=adoption rate expressed as proportion of the total area.

The proportional costs are estimated as C=ACA/YP, where AC=adoption costs
(additional costs of the new technology). The quantity change is estimated as
AQ=(QKen)/(e+n), where n=demand elasticity. Plugging AQ into SG=KPQ-_+0.5KPAQ
gives SG=KPQi[0.5PQK28n/(e+n)].

In the case of a pivotal supply shift, the formulas for change in total surplus
(ATS), change in consumer surplus (ACS), and change in producer surplus (APS), are
estimated as: ATS=0.5KPQ(1+Zr|) ACS=ZPQ(1+0.SZn) APS=ATS-ACS
where Z=Ke/(e+n) is the reduction in price, relative to its initial value, due to the supply
shift (Alston et al. 1998). Plugging Z into ATS gives ATS=0.5KPQ+[0.5PQK28n/(e+n)]
for a pivotal shift, which can easily be compared with the formula shown above for a
parallel shift. Multiplying the ﬁrst term by 0.5 leads to a smaller total surplus for a pivotal
shift. In this dissertation, the net economic beneﬁts were computed by deducting the
research and extension costs from the ATS (pivotal shift), where the second term was
subtracted because the analysis is an ex post assessment.

6.2.6 Net Present Value and Rate of Return

There are several alternative methods to evaluate a stream of beneﬁts and costs
associated with a particular research program. Generally, they involve calculating
summary statistics such as the internal rate of return or the net present value. The net
present value (NPV) of a research program investment is calculated as the sum of the
stream of past or ﬁiture beneﬁts (B1) minus the costs (Ct) associated with the program,

discounted at an appropriate rate (r) as follows:

150

 

_ n (Bt-Ct)
NPV—Z"! (1+r)t

where: Bt=beneﬁt in each year; Ct=cost in each year; n=number of years; t=1,2,. . ., n;
and r=discount rate (most often the opportunity cost of capital or OCC).

The internal rate of return (IRR) is calculated as the discount rate at which the

NPV is exactly zero.

n (Bt-Ct) :0
Z"-'1(1+IRR)‘

According to the NPV criterion, any program with a positive NPV is proﬁtable.
According to the [RR criterion, programs are proﬁtable if the IRR is greater than the
opportunity cost of funds. Typically, the results of the calculations for both the IRR and

the NPV are reported to provide broader information for assessing the total returns to

research (Alston et al. 1998; Gittinger 1982).

6.3 Financial Analysis

Financial analysis is done by using the observed market price—the price at which
a good or service is actually exchanged for another good or service or for money
(ﬁnancial price). This section presents the results of the ﬁnancial analysis using the
economic surplus method. After analyzing the costs and beneﬁt streams associated with
the bean research program, the net present value and the internal rate of return are
reported.

6.3.1 Financial Cost Stream

The ﬁnancial cost stream (Table 6.6) includes the the expenditures of the Kilo per

Kilo program (started in 1996), INIFAP’s research cost, and the ﬁnancial contributions of

151

the B/C CRSP from 1983 when collaborative research started to 2000, ten years after

Pinto Villa was released and four years after Pinto Mestizo was released.

Table 6.6 Total Research and Extension Real Costs in Northern Mexico. 1983-2000.

 

 

 

 

 

 

 

 

 

 

 

Year Research Costs Extension Costs Year Research Costs Extension Costs
11994 MX$) (1994 MXSL (1994 MX$) (1994 MX$)
1983 3,146,302 -- 1992 1,973,942 3,677,120
1984 2,990,406 -- 1993 1,904,916 6,768,3 52
1985 3,051,497 -- 1994 2,043,551 16,227,237
1986 3,263,764 -- 1995 1,998,090 12,719,470
1987 2,598,360 -- 1996 2,035,183 3,586,549
1988 2,518,027 -- 1997 178,801 5,852,148
1989 2,461,741 -- 1998 203,142 13,469,350
1990 2,275,006 -- 1999 196,458 7,766,684
1991 2,091,196 -- 2000 190,201 4,467,613

 

 

 

 

 

 

Source: Tables 6.3, 6.4, and 6.5.

The costs of the seed distribution programs for 1992-1995 in northern Mexico
(before the Kilo per Kilo program started), were estimated by multiplying the number of
hectares planted to improved varieties (Table 6.1) times the average cost per hectare
(1996-2000). Total nominal costs were converted into real cost dividing them by the
consumer price index (divided by 100).

6.3.2 Financial Beneﬁt Stream

Table 6.7 shows the gross social gain and the net gain from research in real
ﬁnancial terms from 1983 to 2000 (base-year=1994). Social gain was computed by using
the formulas and assumptions stated in section 6.2, regarding the elasticities and
ﬁinctional forms of supply and demand for ex post analysis with a pivotal research-
induced supply shift. The incremental net gain was computed by subtracting the total real

and extension costs from the social gain. In the ﬁrst ten years, the beneﬁt stream is

152

 

negative because there was a period of eight years of research investment before the

variety Pinto Villa was released in 1990, and it took two more years before adoption

started and turned the stream to positive numbers.

Table 6.7 Social and Net Gains from Research in the North of Mexico, 1983-2000.

 

 

 

 

 

 

 

 

 

 

 

Year Social Gain Net Gain Year Social Gain Net Gain
(1994 MX$) (1994 MX$) (1994 MX$) (1994 MX$)
1983 0.00 -3,l46,302 1992 2,172,032 -3,479,030
1984 0. 00 -2,990,406 1993 23,170,976 14,497,708
1985 0.00 -3,051,497 1994 19,481,128 1,210,340
1986 0.00 -3,263,764 1995 24,664,446 9,946,886
1987 0.00 -2,598,360 1996 61,433,719 55,811,987
1988 0.00 -2,518,027 1997 14,326,600 8,295,651
1989 0.00 -2,461,741 1998 27,196,272 13,523,781
1990 0.00 -2,275,006 1999 9,950,534 1,987,392
1991 0.00 -2,091,196 2000 9,914,161 5,256,348

 

 

 

 

 

 

 

Source: Computed from data on Tables 6.1, 6.2, and 6.5.

6.3.3 Financial Rate of Return

The period from 1983 (when research started) to 2010 was considered as the
appropriate period for evaluating the economic impact of the improved bean varieties.
Because the collaborative research program began in 1983, this was selected as the ﬁrst
year to be included in the analysis. The selection of 2010 as the cut-off year allows for
twenty years of evaluation after the ﬁrst improved pinto bean variety (Pinto Villa) was
released in 1990. For purposes of the evaluation, the costs and beneﬁts streams were
assumed to remain constant in real terms after 2000. The logistic curve used to ﬁt the
data on adoption rate (Figure 6.1) shows that the adoption of improved pintos seems to
have reached the ceiling in a level of approximately 38%. The parameters for the curve

(a=-5.70, B=1.61) were obtained from the transformed equation ln(P/K-P)= 0L+Bt by an

153

OLS regression using K=38. Therefore, it is reasonable to assume no ﬂirther increases in
the percentage of area using the improved pinto varieties. The possibility of changes in
key elements of the costs and the beneﬁts of bean research (e. g. , adoption rate decrease)

is addressed in the sensitivity analysis.

Figure 6.1 Logistic Curve for Estimating the Rate of Adoption in Mexico, 1991-2000.
50

40 Real / \VA
,5 / W
/ / Estimated V
/ /
f /
/ /

 

 

 

 

00
O

 

 

Percentage
N
OI

N
O

 

 

 

 

13 //
, //

 

 

Wﬁ I I I I ﬁ I I

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Year

0
l

 

Source: Table 6.1 for “real”; “estimated” is based on logistic curve estimate.

The results from the ﬁnancial analysis—using a pivotal shift—indicate that the
net present value (NPV) of the investment in bean research is MX$ 17,681,052 (base-
year=1994), when using a discount rate of 10%, which is the estimated cost of
opportunity of capital (OCC) in agricultural activities in Mexico (FIRA 2001). The
estimated internal rate of return (IRR) is 17.5%, which is almost twice as much the OCC.
As mentioned before, according to the NPV criterion, any program with a positive NPV

is considered to be proﬁtable, and, according to the IRR criterion, a program is proﬁtable

154

if the IRR is greater than the opportunity cost of funds. Therefore, since the ﬁnancial
point of view, investing in generating and promoting the use of improved bean varieties
in northern Mexico has been a proﬁtable investment under both the NPV and the IRR
criteria.

Regarding the distribution of the beneﬁts (change in total surplus of MX$
107,051,062), over the period 1992-200015 (i.e., the period when the improved varieties
were adopted), change in consumer surplus was MX$ 204,052,485, and the change in
producer surplus was MX$ -97,001,423. These results indicate that all the beneﬁts from
the change in total surplus were captured by the consumers and that producers actually
have a loss, as result of the demand being inelastic (n=0.5), because this implies that the
percentage reduction in output price is larger than the percentage increase in the quantity

demanded by consumers.

6.4 Economic Analysis

Economic analysis is done by using the economic values. If there are no market
distortions, the ﬁnancial price and the economic price are equal. However, if market
distortions exist (i. e., subsidies, tariffs), the economic price is estimated as the
opportunity cost (shadow price). This section presents the results of the economic
analysis using the economic surplus method. After analyzing the costs and beneﬁts
streams of bean research, the net present value and the internal rate of return are

computed.

 

‘5 This period is different from the period 1983-2010 used to estimate the NPV and IRR because the focus
is on estimating the distribution of research beneﬁts between consumers and producers.

155

6.4.1 Economic Cost Stream

The ﬁnancial and economic analyses are complementary. However, while the
ﬁnancial analysis assesses the ﬁnancial effects of an investment on individual
participants, the economic analysis takes the point of view of the society. In the case of
the cost stream, there was no need to adjust the research and extension costs for market
distortions (e. g., taxes, subsides, or exchange rate).

Therefore, the cost stream used for the economic analysis was the same as the
cost stream for the ﬁnancial analysis.

6.4.2 Economic Beneﬁt Stream

In the case of the economic beneﬁts streams (Table 6.8), two main adjustments
were made. First, the subsidy for the price of the seed distributed through the Kilo per
Kilo program was removed. Second, the import parity price (IPP) at the farm gate level,
in real terms (base-year=1994), was used instead of the domestic price for pinto beans.
There are some caveats regarding the use of the [PP to estimate the opportunity cost of
producing beans domestically instead of importing: a) the analysis is based in the use of
annual average prices for both domestic and IPP, which doesn’t show the month-to
month bean price ﬂuctuation; b) the price of imported beans depends not only on the
market class but also on the origin, the quantity and the season when beans are imported;
and c) the beans imported, even if they are of the same market class (i.e., pintos) may not
exactly match the characteristics of the domestically-produced beans.

To compute the [PP at the farm gate level the FOB price for US. pinto beans
(Table 6.2) was adjusted to approximate the CIF price (which includes freight and

insurance charges to point of import) by using the FAO’s Standard conversion factor of

156

1 12 percent (F AO 2002b). Then, the estimated CIF price was converted into Mexican
pesos by using the official exchange rate (Table 6.5), and ﬁnally adjusted by marketing
and transportation costs to the bean production areas of Chihuahua and Durango to
estimate the [PP at farm gate. The official exchange rate was not adjusted (e. g., for
overvaluation) because in the early 19903 Mexico adopted a ﬂoating exchange rate
regime under which the exchange rate was allowed to ﬂuctuate freely in response to
changing economic conditions (Mankiw 1992). During the research phase (1983-1991),
both the economic social gain and the incremental net gain are the same as in the
ﬁnancial analysis because the research cost stream are the same. After adoption started,
both the economic social gain and the incremental net gain are higher than in the
ﬁnancial analysis because the higher price of imported beans (IPP) more than offsets the

removal of the seed subsidy, which increases the cost of the improved seed.

Table 6.8 Social and Net Gains from Research in Northern Mexico, 1983-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Social Gain Net Gain Year Social Gain Net Gain
(1994 MX$) (1994 MX$) Q994 MX$) (1994 MX$)
1983 0.00 -3,146,302 1992 761,966 -4,889,096
1984 0.00 -2,990,406 1993 22,138,552 13,465,284
1985 0.00 -3,051,497 1994 32,698,863 14,428,075
1986 0.00 -3,263,764 1995 37,535,593 22,818,033
1987 0.00 -2,598,360 1996 81,313,630 75,691,898
1988 0.00 -2,518,027 1997 14,067,838 8,036,888
1989 0.00 -2,461,741 1998 26,982,999 13,310,507
1990 0.00 -2,275,006 1999 13,897,187 5,934,045
1991 0.00 -2,091,l96 2000 8,978,218 4,320,405

 

Source: Computed from data on Tables 6.1, 6.2, and 6.5.
6.4.3 Economic Rate of Return
As in the case of the ﬁnancial analysis, the period from 1983 to 2010 was used for

evaluating the economic impact of the improved bean varieties. The costs and beneﬁts

157

 

streams were assumed to remain constant in real terms after 2000. The possibility of
changes in key elements of the costs and the beneﬁts of bean research and extension is
addressed in the sensitivity analysis.

The results from the economic analysis—using a pivotal shift—indicate that the
net present value (NPV) of the investment in bean research is MX$ 29,420,210 (base-
year=1994), when using a discount rate of 10%, which is the estimated cost of
opportunity of capital in agricultural activities in Mexico. The estimated internal rate of
return (IRR) is 21.4%, which is more than twice as much the discount rate used to
compute the NPV. Therefore, since the economic point of view, investing in generating
and promoting the use of improved bean varieties in northern Mexico has been a
proﬁtable investment under both the NPV and the IRR criteria. Higher values for the
economic net present value and the internal rate of return, compared to the ﬁnancial
analysis estimates, are explained by the higher price level of imported beans (IPP) versus
bean domestic prices.

Regarding the distribution of the beneﬁts (change in total surplus of MX$
153,116,040), over the period 1992-2000 (i. e., the period when the improved varieties
were adopted), change in consumer surplus was MX$ 251,644,325, and the change in
producer surplus was MX$ -98,528,285. These results indicate that all of the beneﬁts
from the change in total surplus were captured by consumers and that producers actually

have a loss, as result of the demand being inelastic (11:05).

158

 

6.5 Sensitivity Analysis

This section presents the sensitivity analysis performed on the results of both the
ﬁnancial and the economic analysis. Sensitivity analysis is used to evaluate the impact of
uncertainty about ﬁiture events and values on the baseline results. A sensitivity analysis is
done by varying the value of one key variable or a combination of key variables, and
determining the impact of that change on the result as measured by the IRR or the NPV
(Gittinger 1982; ADB 1997).

6.5.1 Changes in Adoption and Yields

The ﬁnancial and economic rates of return are very sensitive to changes in the
adoption rate and incremental yields (Table 6.9). The switching value is the percentage
changes in the variables for the NPV to become zero or the IRR to fall to the cut-off rate

(i. e. , the OCC=10 percent).

Table 6.9 Sensitivity Analysis of Decreasing Changes in Adoption Rate and Incremental

 

 

 

 

 

 

 

Yield.
Adoption Financial Economic Incremental Financial Economic
Rate Rate of Rate of Yield Rate of Rate of
Decrease Return Return Decrease Return Return
0% 17.5% 21.4% 0% 17.5% 21.4%
-10% 15.4% 19.4% -10% 15.2% 19.0%
-20% 12.9% 17.0% -20% 12.3% 16.0%
-30% 9.7% 14.1% -30% 8.5% 11.8%
-40% 5.2% 10.2% -40% 1.9% 3.1%

 

 

 

 

 

 

If a 40 percent decrease in the adoption rate is assumed, the ﬁnancial returns

decrease from 17.5 to 5.2 percent (switching value=29% decrease) and the economic

returns decrease from 21.4 to 10.2 percent (switching value=40.5% decrease), which

implies a negative ﬁnancial net present value because the opportunity cost of capital is 10

159

 

percent. However, a large decrease in the adoption rate (e. g., 40%) for the improved pinto
bean varieties is not likely to happen until better improved varieties are released.

If a 40 percent decrease in the incremental yield is assumed (i. e., from 20% to
12%), the ﬁnancial returns decrease from 17.5 to 1.9 percent (switching value=26.5%
decrease) and the economic returns decrease from 21.4 to 3.1 percent (switching
value=33% decrease), which implies negative net present values both ﬁnancial and
economic. In both cases, the economic 'rate of return is still higher than the ﬁnancial rate
of return, because imported beans are more expensive than domestically produced beans.

The 95% conﬁdence interval for the mean yield of the improved pintos (section
5 .42) indicates that the lower bounds are 10% and 4% below the mean for Pinto Mestizo
and Pinto Villa, respectively. Therefore, a large decrease in incremental yields (e. g.,
>26%) is not very likely.

6.5.2 Changes in Prices and Costs

The ﬁnancial and economic rates of return are also very sensitive to changes in
the domestic price and the import parity price, respectively (Table 6.10). If a 40 percent
decrease in the domestic price is assumed, the ﬁnancial rate of return decreases from 17.5
to 4.7 percent (switching value=28.4% decrease), which implies a negative net present
value because the opportunity cost of capital is 10 percent.

A 40 percent decrease in the import parity price reduces the economic rate of
return from 21.4 to 6.7 percent (switching value=3 5% decrease). However, the economic
rate of return is less sensitive to changes in the ofﬁcial exchange rate. If an increase in the

exchange rate (MX$/USS) is assumed, to account for potential overvaluation, making

160

imported beans even more expensive relative to domestically produced beans, the

economic rate of return also increases.

Table 6.10 Sensitivity Analysis for Changes in Prices and in the Ofﬁcial Exchange Rate.

 

 

 

 

 

 

 

Domestic Financial Import Economic Exchange Economic
Pricea Rate of Parity Price Rate of Rate Rate of
Decrease Return Decreaseb Return Increase Return
0% 17.5% 0% 21.4% 0% 21.4%
-10% 15.3% -10% 19.1% 10% 23.4%
-20% 12.7% -20% 16.2% 20% 25.1%
-30% 9.4% -30% 12.5% 30% 26.6%
-40% 4.7% -40% 6.7% 40% 28.0%

 

 

 

 

 

 

a. The change in domestic price affects only ﬁnancial returns.
b. The change in import parity price affects only economic returns.

The ﬁnancial and economic rates of return are less sensitive to changes in
research and extension costs (Table 6.11). However, a large increase in R&E costs would
need to be assumed (40%) in order to decrease the ﬁnancial rate of return from 17.5 to
9.9 percent and make the net present value negative (switching value=39.5% increase).
For a very large increase in R&E costs (70%), the economic rate of return decreases from
21.4 to 9.2 percent (switching value=65.5% increase) and turns the NPV negative. The
ﬁnancial rate of return is not very sensitive to changes in the adoption costs, which
depend basically on the subsidy for the price of the improved seed. Even assuming a
complete elimination of the subsidy—a one hundred percent reduction—the ﬁnancial rate

of return is 14.0 percent, which is higher than the OCC (10 percent).

161

 

Table 6.11 Sensitivity Analgsis for Changes in Research, Extension, and Adoption Costs.

 

 

 

 

 

 

 

 

 

 

Research and Financial Rate Economic Rate Adoption Costs Financial Rate
Extension Costs of Return of Return (Seed Subsidy of Return
Increase Decrease)“

0% 17.5% 21.4% 0% 17.5%
10% 15.6% 19.5% -20% 16.9%
20% 13.6% 17.8% -40% 16.2%
30% 11.8% 16.0% -60% 15.5%
40% 9.9% 14.4% -80% 14.8%
50% 8.0% 12.7% -100% 14.0%
60% 6.0% 1 1.0% -- --
70% 3.9% 9.2% -- --

 

 

 

 

 

a. The change in subsidy level affects only ﬁnancial returns.

6.5.3 Changes in Elasticities

 

The ﬁnancial and economic rates of return are very sensitive to the estimates for
the elasticity of supply (Table 6.12). If a supply elasticity (a) of 0.1 is assumed, the
ﬁnancial and economic returns increase to 44.2 and 48.8 percent, respectively, but if a
supply elasticity of 0.9 is assumed, the ﬁnancial and economic rates of return decrease to
2.2 and 2.6 percent respectively, turning the NPVs negative. Break-even values are

8=0.7O and 8:077 for the ﬁnancial and economic IRR, respectively.

Table 6.12 Sensitivity Analysis for Different Estimates of Supply and Demand Elasticity.

 

 

 

 

 

 

 

 

 

 

 

Supply Financial Economic Demand Financial Economic

Elasticity Rate of Rate of Elasticity Rate of Rate of
Return Return Return Return

0.10 44.2% 48.8% 0.10 18.0% 21.8%
0.20 33.3% 37.7% 0.20 17.9% 21 .7%
0.30 26.7% 31.0% 0.30 17.7% 21.5%
0.40 21.7% 25.8% 0.40 17.6% 21 .5%
0.50 17.5% 21.4% 0.50 17.5% 21 .4%
0.60 13.7% 17.3% 0.60 17.5% 21.3%
0.70 10.1% 13.2% 0.70 17.4% 21.3%
0.80 6.3% 8.7% 0.80 17.3% 21.2%
0.90 2.2% 2.6% 0.90 17.3% 21.2%

 

 

 

 

 

 

162

 

In contrast, the ﬁnancial and economic returns have little sensitivity to changes in

the demand elasticity estimate. If a demand elasticity (n) of 0.1 is assumed, the ﬁnancial

and economic returns only increase to 18.0 and 21.8 percent, respectively, but if a

demand elasticity of 0.9 is assumed, the ﬁnancial and economic returns will only

decrease to 17.3 and 21.2 percent, respectively. In both cases the NPV is still positive for

a large values of the elasticity of demand (e. g., 11:10).

The magnitude of the change in total surplus and the distribution of the beneﬁts

between consumers and producers (1992-2000) are very sensitive to the estimates for the

elasticity of supply (Table 6.13). If a supply elasticity (a) of 0.1 is assumed (and n=0.5),

the ﬁnancial and economic change in total surplus increase to MX$ 850,196,000 and

MX$ 1,186,254,000, respectively and the largest share of the beneﬁts goes to producers.

If a supply elasticity of 0.9 is assumed, the ﬁnancial and economic change in total surplus

decrease to MX$ 22,823,000 and MX$ 35,431, respectively, and all beneﬁts are captured

by the consumers while producers actually lose from a more elastic supply.

Table 6.13 Sensitivity of Beneﬁt Distribution to Different Estimates of Supply Elasticity.

 

 

 

 

 

 

 

 

 

 

 

 

Financial Analysis Economic Analysis
Supply (Thousands of MX$) (Thousands of MX$)

Elasticity Change Change in Change in Change in Change in Change in
in Total Consumer Producer Total Consumer Producer

Surplus Surplus Surplus Surplus Surplus Surplus

0.10 850,196 344,437 505,759 1,186,254 468,145 718,109
0.20 387,128 293,900 93,228 542,983 390,127 152,856

0.30 231,996 256,289 -24,293 327,203 332,121 -4,917
0.40 154,041 227,208 -73,167 218,636 287,306 -68,670
0.50 107,051 204,052 -97,001 153,116 251,644 -98,528
0.60 75,592 185,179 -109,586 109,206 222,592 -113,386
0.70 53,037 169,500 -116,463 77,694 198,468 -120,774
0.80 36,064 156,269 -120,205 53,960 178,116 -124,156
0.90 22,823 144,953 -122,l30 35,431 160,717 -125,286

 

 

 

 

 

 

 

163

 

 

In contrast, the magnitude of the change in total surplus and the distribution of the

beneﬁts between consumers and producers (1992-2000) are not very sensitive to the

estimates for the elasticity of demand (Table 6.14). If a demand elasticity (n) of 0.1 is

assumed (and e=0.5), the ﬁnancial and economic change in total surplus increase only to

MX$ 112,270,000 and MX$ 159,014,000, respectively and all beneﬁts are captured by

the consumers while producers actually lose. If a demand elasticity of 0.9 is assumed the

ﬁnancial and economic change in total surplus decrease only to MX$ 104,814,000 and

MX$ 150,589,000, respectively and all beneﬁts are captured by the consumers while

producers actually lose—but the loss is proportionally smaller than when demand is more

inelastic.

Table 6.14 Sensitivity of Beneﬁt Distribution to Different Estimates of Demand Elasticity

 

Financial Analysis

Economic Analysis

 

 

 

 

 

 

 

 

 

 

 

 

 

Demand (Thousands of MX$) (Thousands of MX$)
elasticity Change Change in Change in Change in Change in Change in
in Total Consumer Producer Total Consumer Producer
Surplus Surglus Surplus Surplus Surplus Surplus
0.10 112,270 335,738 -223,468 159,014 414,493 -255,479
0.20 1 10,406 289,107 -l78,701 156,907 356,784 —199,877
0.30 109,008 253,842 -144,834 155,328 313,173 -157,846
0.40 107,921 226,242 -118,321 154,099 279,059 -124,960
0.50 107,051 204,052 -97,001 153,116 251,644 -98,528
0.60 106,339 185,826 -79,486 152,312 229,133 -76,821
0.70 105,746 170,587 -64,841 151,642 210,318 -58,676
0.80 105,245 157,658 -52,414 151,075 194,358 -43,283
0.90 104,814 146,551 -41,736 150,589 180,649 -30,060

 

 

 

 

 

Previous results are consistent with theory, which states that elasticity estimates

are important in relation to the distribution of beneﬁts. In particular, the more elastic

164

 

 

supply is relative to demand, the greater the consumer share of total research beneﬁts

(and smaller the producer share) and vice versa (Alston et al. 1998).

6.6 The Small Open Economy Model

This section presents the formulas and estimations of research beneﬁts in a small
open economy model. The small open economy refers to a situation where the price is
determined exoneously by the world price of the commodity (Alston et al. 1998). The
small-country assumption may be appropriate for the case of Mexico, considering that
dry beans are tradable and that the country does not have a signiﬁcant inﬂuence on
international prices. The ﬁnancial analysis includes the government subsidy for the
improved bean varieties and the NAFTA tariffs and quotas for importing beans into
Mexico. For the economic analysis both the subsidy for bean seed and the tariffs and
quotas for bean imports (market distortions) were removed.

6.6.1 Financial Analysis

In a small country, a tariff '6 on imports raises the domestic price above the world
price (Pw) to a price equal to the world price plus the tariff rate ([1+t]Pw) and generates
tariff revenue for the government. The research-induced supply shift reduces imports of
the commodity and government revenues from the tariff. While the total beneﬁt from
research is unaffected by the tariff, producer beneﬁts are greater, offsetting reduced

government revenues from tariffs (Alston et al. 1998).

 

‘6 A tariff and a quota are in some senses equivalent In most analyses, the main difference between a tariff
and an import quota is that the tariff generates tariff revenue for the government whereas the import quota
generates quota rents, which are often private beneﬁts (Alston 1998).

165

The formulas used in this study for the ex post assessment of research beneﬁts
from a pivotal supply shift are:
AGS=—thQKe APS=0.5(l+t)PwQK(1-Kc) ACS=0 ATS=APS+AGS
where AGS=change in government revenue, t=ad valorem tariff (rate),
Pw=world price, Q=observed quantity produced, K=vertical supply shift
expressed as a proportion of the product price, c=elasticity of supply,
APS=change in producer surplus, ACS=change in consumer surplus, and

ATS=change in total surplus.

Research and extension costs were subtracted to obtain the net changes in
producer surplus and total surplus.

The results from the ﬁnancial analysis—using a pivotal shift—indicate that the
net present value (NPV) of the investment in bean research for a small open economy
model is MX$ 26,331,431 (base-year=1994), when using a discount rate of 10% (OCC).
The estimated internal rate of return (IRR) is 21.3%, which is more than twice the
opportunity cost of capital.

Given that the NPV is positive and that the IRR is greater than the OCC, from the
ﬁnancial point of view, investing in generating and promoting the use of improved bean
varieties in northern Mexico has been a proﬁtable investment under both the NPV and the
IRR criteria when assuming a small open country.

Regarding the distribution of the beneﬁts (change in total surplus of MX$
85,258,806), over the period 1992-2000 (i.e., the period when the improved varieties
were adopted), the change in consumer surplus was zero, the change in producer surplus

was MX$ 224,372,237, and the change in government revenues was MX$ -139,113,431.

166

These results indicate that all of the beneﬁts from the change in total surplus were
captured by producers. Consumers didn’t beneﬁt from technological change because the
output price is assumed to be exogenous. Additionally, there was a loss in tariff revenues
to the government.

6.6.2 Economic Analysis

In a small open economy without market-distorting policies, all of the research
beneﬁts accrue to the producers in the innovating country (i.e., Mexico), regardless of
whether the country is an exporter or an importer (Alston et al. 1998). Given that a small
country does not have a signiﬁcant inﬂuence in the world price for beans, the research-
induced supply shift does not reduce the domestic price, which remains equal to the
world price. There is no change in consumer surplus because the price is relatively
constant and the change in total surplus is equal to the change in producer surplus that
results from the reduction in the cost per unit of output.

In the case of a small open economy, the formulas for the ex post assessment of
research beneﬁts from a pivotal supply shift are:

ACS=O APS=ATS=0.5PwQK(1-Ke)
where AC S= change in consumer surplus, APS=change in producer surplus,
ATS=change in total surplus, Pw=world price, Q=observed quantity produced, K=vertical
supply shift expressed as a proportion of the product price, and e=elasticity of supply.

This formula applies for both an exporter and an importer country in the absence
of market distorting policies. Research and extension costs were subtracted to obtain the

net changes in producer surplus and total surplus.

167

The results from the economic analysis—using a pivotal shift—indicate that the
net present value (NPV) of the investment in bean research for a small open economy
model is MX$ 24,405,096 (base-year=1994), when using a discount rate of 10% (OCC).
The estimated internal rate of return (IRR) is 20.7%, which is twice as much the
opportunity cost of capital. Given that the NPV is positive and that the IRR is greater
than the OCC, from the economic point of view, investing in generating and promoting
the use of improved bean varieties in northern Mexico has been a proﬁtable investment
under both the NPV and the IRR criteria when assuming a small open country. Lower
values for the economic NPV and IRR, compared to the ﬁnancial analysis estimates, are
explained by the removal of the bean seed subsidy and the beans import tariffs and
quotas. The difference is relatively small because in the ﬁnancial analysis the research-
induced supply shift also reduces the government revenues from the tariff.

Regarding the distribution of the beneﬁts (change in total surplus of MX$
152,848,491), over the period 1992-2000 (i.e., the period when the improved varieties
were adopted), the change in consumer surplus was zero, and the change in producer
surplus was equal to the change in total surplus. These results indicate that all of the
beneﬁts from the change in total surplus were captured by producers. Consumers didn’t
beneﬁt from technological change because the output price is assumed to be exogenous.

6.6.3 Closed versus Small Open Economy

Both the closed and small open economy models are used to analyze the rate of
return to bean research because of the complexity of output pricing for beans in Mexico.

In the closed economy model, the price of the commodity is determined inside the

168

country but in the small open economy model the price is exogenous. In this study, a
closed economy model is justiﬁed by the fact that the domestic price of beans clearly
showed a decreasing trend in 1990-2000, while a decreasing trend in the international
price was not so evident (section 6.1.2). In contrast, the small open economy model is
justiﬁed by the increasing inﬂuence of the world price on the domestic price and the open
economy model takes into account the effect of barriers to trade (i.e., tariffs and quotas).
However, the case of the bean subsector in Mexico does not exactly matches either of
these models. In fact, given that NAFTA evolved during the years considered in this
study, these years represent a period of transition from a relatively closed economy to an
open economy. For example, data from Chapter 3 (section 3.1.5) show that during 1991-
2000, bean imports averaged 6% of total production and although net imports were
positive, they averaged only 4% of total production. In addition, in ﬁve out of ten years
(1991-1995), net imports were negative or nearly zero. Because beans are mostly grown
under rainfed conditions in Mexico (85% of total planted area), the erratic weather
conditions explain the large ﬂuctuations in bean production and imports. Therefore, price
setting for beans in Mexico depends not only on domestic production, but also on the
world bean production (mainly the US), which is also subject to large ﬂuctuations from
year-to—year due to changes in weather and market conditions.

Therefore, comparison of the estimates for the closed vs. small open economy
model is intended to help in addressing this complexity. The results of the analysis
indicate that there is very little difference in the rates of return for the economic analysis
when estimated assuming a closed economy model (21.4%) vis-a-vis the small open

economy model (20.7%). This result is as expected because in both cases the seed

169

subsidy and bean import tariffs were removed, and the world price was used to compute
the rates of return. In the case of the ﬁnancial analysis, the rate of return is higher for the
small open economy model (21.3%) than for the closed economy model (17.5%). In both
cases, the seed subsidy is taken into account so the difference is explained by the use of
the world price for the open economy model, which before the tariff is higher than the
domestic price used for the closed economy model. In the open economy model, the
greater producer beneﬁt after the tariff is offset by the reduction in government revenues
due to the decrease in bean imports.

The most important difference between the closed vs. small open economy model
has to do with the distribution of beneﬁts from the research-induced supply shift. In the
closed economy model, for both the ﬁnancial and economic analysis, all of the beneﬁts
from the change in total surplus are captured by consumers, while producers actually
have a loss—when demand is inelastic—because technological change leads to a lower
domestic price.

In contrast, in the case of the ﬁnancial analysis for the small open economy
model, instead of decreasing, the domestic price remains equal to the world price. Thus,
all of the beneﬁts from the change in total surplus are captured by producers, there is no
change in consumer surplus, and there is a loss in government tariff revenues.

In the case of the economic analysis for the small open economy model, the fact
that the output price is exogenous implies that all of the beneﬁts from the change in total
surplus are captured by producers and there is no change in consumer surplus after

technological change.

170

As mentioned before, in the case of the closed economy model, consumers
capture all of the beneﬁts from research. The development and diffusion of improved
bean varieties directly beneﬁts all consumers by making larger quantities of beans
(supply shift) available at lower prices. However, given that both consumption
preferences and demand response to price changes vary with income, the impact differs
among income classes. Because most staple commodities (e. g., dry beans) are inferior
goods (i. e., have negative income elasticities), poorer people consume a greater quantity
than richer consumers (Alston et al. 1998). Individually, low-income consumers tend to
beneﬁt relatively more from research to generate improved bean varieties than research
on non-staple commodities (e. g., export crops) because low-income consumers spend a
high proportion of their income on beans.

In the case of the small open economy model, producers capture all of the beneﬁts
from research, but the impact among producer groups is different. Producers with
different incomes, with different farm sizes, in different locations, and with diverse
tenure situations can gain or lose, depending on the suitability of the new technology to
their particular circumstances. Recent evidence suggests that neither farm size nor tenure
have been a major impediment to adoption of new biological technologies (Alston et al.
1998). However, in this study, the evidence ﬁom the econometric analysis of program
participation (section 5.5) indicates that in northern Mexico farmer participation in the
seed distribution program is associated mainly with farm size and location. Smaller
farmers located in marginal areas (which were targeted by the program) are more likely
to participate in the program. Therefore, they beneﬁt more from bean research than larger

farmers with a better location (i. e., closer to the city).

171

In addition, the observed decline in the real price of pinto beans (section 3.3.1)
suggest that early adopters beneﬁted more from technological change than late or non-
adopters because at the beginning of the adoption process the price for the improved
pintos was higher than the price of traditional pintos—until improved pintos gradually
ﬂooded the market.

For example, through the 19803 Ojo de Cabra was the dominant variety grown in
Chihuahua. In 1990 CONASUPO (marketing parastatal) announced that in the ﬁiture, it
would no longer purchase this variety, since although preferred by local consumers,
CONASUPO could not sell its surpluses of this variety in other regions of Mexico.
Following this announcement, farmers began looking for a replacement variety, just at
the time when Pinto Villa was being tested. As a result, commercial production of Pinto
Villa, which began in 1992, expanded rapidly.

The incentives for farmers to adopt the new varieties were higher yields and
higher prices (initially) than traditional varieties, due to a better market acceptance. The
improved varieties have higher yields than traditional varieties because they are resistant
to anthracnose—a disease causing serious crop losses—and tolerant to drought, the main
problem in the region. At the beginning of the adoption process, farmers also beneﬁted
from higher prices for the improved pintos versus traditional varieties (e. g., Ojo de
Cabra), but as the relative amount of these beans in the market increased, prices declined.
However, pintos still have better market acceptance. Thus, bean research beneﬁted
producers by generating new varieties that increased or maintained productivity and

farmers’ income.

172

6.7 Summary of the Chapter

In this chapter the returns to research investment for improved bean varieties in
the north of Mexico are evaluated using the economic surplus method. During the period
1990-2000, INIFAP released several improved bean varieties for rainfed conditions, with
the support of the B/C CRSP. For the purpose of the economic evaluation of bean
research, the relevant improved varieties are Pinto Villa and Pinto Mestizo. The statistical
analysis of Chapter 5 indicates that in both Chihuahua and Durango these two varieties
have yields that are signiﬁcantly higher (20.6%) than the traditional pinto market class
variety Pinto Nacional.

Other improved varieties are not considered in the economic assessment because
there is insufficient evidence that their yields are signiﬁcantly higher than traditional
varieties in the same market class. Similarly, Zacatecas is not included in the analysis,
since the two improved pintos do not account for a major share of the state’s bean area.
However, research and extension costs for other varieties and for Zacatecas are included
because they are part of the total cost of generating and promoting the improved varieties.

The ﬁrst section of the chapter presents the data used for the analysis regarding:
a) area, production, and yields; b) value of production and prices; 0) INIFAP and CRSP
research costs; d) extension and adoption costs; and e) the consumer price index and the
exchange rate. Section two explains the assumption used in the analysis, regarding supply
and demand elasticities, functional forms, and the adoption path. It also presents the
formulas for the closed economy model which were used to compute the changes in total
surplus (ATS), consumer surplus (ACS), and producer surplus (APS), the net present

value (NPV), and the internal rate of return (IRR). Sections three and four present the

173

results of the ﬁnancial and economic analysis, assuming a closed economy model.
Section ﬁve reports the results of the sensitivity analysis. Section six presents the results
of the ﬁnancial and economic analysis, assuming a small open economy model and
compares these results with results obtained using the closed economy model.

In the closed economy model, the results from the ﬁnancial analysis indicate that
the NPV of the investment in bean research is MX$ 17,681,052 (base-year=l994), when
using a discount rate of 10%, which is the estimated cost of opportunity of capital (OCC)
in agricultural activities in Mexico, and that the IRR is 17.5%. Therefore, ﬁom the
ﬁnancial point of view, investing in generating and promoting the use of improved bean
varieties in northern Mexico has been a proﬁtable investment under both the NPV and the
IRR criteria. Regarding the distribution of the beneﬁts between consumers and producers
(change in total surplus of MX$ 107,051,062), over the period 1992-2000 (i. e., the period
when the improved varieties were adopted), the change in consumer surplus was MX$
204,052,485, and the change in producer surplus was MX$ -97,001,423. These results
indicate that the consumers captured all the beneﬁt from the change in total surplus and
that producers actually have a loss, as result of the demand being inelastic.

In the closed economy model, the results from the economic analysis indicate that
the NPV of the investment in bean research is MX$ 29,420,210, (OCC=10%), and that
the IRR is 21 .4%. Therefore, from the point of view of the society as a whole, investing
in generating and promoting the use of improved bean varieties in northern Mexico has
been a proﬁtable investment under both the NPV and the IRR criteria. The economic
NPV and IRR, are higher than the ﬁnancial analysis estimates because the higher price of

imported beans (IPP), compared to domestically produced beans, more than offsets the

174

removal of the subsidy to farmers on the price of the improved seed. Regarding the
distribution of the beneﬁts (change in total surplus of MX$ 153,116,040), the change in
consumer surplus was MX$ 251,644,325, and the change in producer surplus was MX$
-98,528,285. These results indicate that the consumers captured all the beneﬁts from the
change in total surplus and that producers actually have a loss, as a result of the demand
being inelastic.

Sensitivity analysis was carried out to assess the sensitivity of the rate of return to
changes in the parameters used in the model. Sensitivity analysis estimated the switching
values, deﬁned as the percentage changes in the value of key variables that results in
NPV to become zero or the IRR to fall to the cut-off rate (i.e., the OCC=10 percent). For
the adoption rate, the switching values are a 29% decrease for ﬁnancial returns and a
40.5% decrease for economic returns. However, a large decrease in the adoption rate
(e. g., 40%) for the improved pinto bean varieties is unlikely to occur until better
improved varieties are released. For the incremental yield, the switching value is a 26.5%
decrease for ﬁnancial returns and a 33% decrease for economic returns. However, the
95% conﬁdence interval for the mean yield of the improved pintos (section 5.4.2)
indicates that the lower bounds are 10% and 4% below the mean for Pinto Mestizo and
Pinto Villa, respectively. Therefore, a large decrease in incremental yields (e. g., >26%) is
not very likely.

For the domestic price, the switching value is a 28.4% decrease for ﬁnancial
returns. For the IPP, the switching value is a 35% decrease for economic returns. In the

case of the exchange rate, if an increase in the exchange rate (MX$/USS) is assumed, to

175

 

account for potential overvaluation, making imported beans even more expensive relative
to domestically produced beans, the economic rate of return also increases.

For research and extension cost the switching values are a 39.5% and a 65.5%
increase for ﬁnancial and economic returns, respectively. Changes in adoption costs
depend basically on the subsidy for the price of the improved seed. Even assuming a
complete elimination of the subsidy, the ﬁnancial rate of return is 14.0 percent, which is
still higher than the OCC.

If a supply elasticity (e) of 0.1 is assumed, the ﬁnancial and economic returns
increase to 44.2 and 48.8 percent, respectively, but if a supply elasticity of 0.9 is assumed
the ﬁnancial and economic rates of return decrease to 2.2 and 2.6 percent respectively,
turning the NPVs negative. Break-even values are e=0.70 and e=0.77 for the ﬁnancial
and economic IRR, respectively. This means that research investment remains proﬁtable
for e<0.70 in the ﬁnancial analysis and for e<0.77 in the economic analysis.

If a demand elasticity (n) of 0.1 is assumed, the ﬁnancial and economic returns
only increase to 18.0 and 21.8 percent, respectively, but if a demand elasticity of 0.9 is
assumed, the ﬁnancial and economic returns will only decrease to 17 .3 and 21.2 percent,
respectively, which are still higher that the OCC. In both cases the NPV is still positive
for a large values of the elasticity of demand (e. g., n=10).

In addition, the sensitivity analysis was used to assess the distribution of the
beneﬁts between consumers and producers. If a supply elasticity (e) is of 0.1 is assumed
(and n=0.5), there is a large increase in total surplus and the largest share of the beneﬁts

is for the producers. If a supply elasticity of 0.9 is assumed (and 11:05), there is a large

176

decrease in total surplus and the consumers capture all beneﬁts while producers actually
lose from a more elastic supply.

If a demand elasticity (n) of 0.1 is assumed (and 8:0.5), there is a small increase
in total surplus and all beneﬁts are captured by the consumers while producers actually
lose. If a demand elasticity of 0.9 is assumed (and e=0.5),there is a small decrease in total
surplus and the consumers capture all beneﬁts while producers actually lose—although
the loss is proportionally smaller than when demand is more inelastic.

In a closed economy model, the price of the commodity is determined inside the
country and in a small open economy model the price is exogenous. In this study, a
closed economy model is justiﬁed by the fact that the domestic price of beans clearly
showed a decreasing trend in 1990-2000, while a decreasing trend in the international
price was not so evident. The small open economy model is justiﬁed by the increasing
inﬂuence of the world price on the domestic price and the open economy model takes
into account the effect of barriers to trade (i.e., tariffs and quotas).

In the small open economy model, the results from the ﬁnancial analysis indicate
that the NPV of the investment in bean research is MX$ 26,331,431 (OCC=10%) and
that the IRR is 21.3%, which implies that bean-research investment is ﬁnancially a
proﬁtable investment under the small open country assumption. Regarding the
distribution of the beneﬁts (change in total surplus of MX$ 85,25 8,806), the change in
consumer surplus was zero, the change in producer surplus was MX$ 224,3 72,23 7, and
the change in government revenues was MX$ -139,113,431. These results indicate that

all of the beneﬁts from the change in total surplus were captured by producers.

177

Consumers didn’t beneﬁt from technological change because the output price is assumed
to be exogenous. Additionally, there was a loss in tariff revenues to the government.
In the small open economy model, the results from the economic analysis indicate

that the NPV of the investment in bean research is MX$ 24,405,096 (OCC=10%) and

 

that the IRR is 20.7%, which implies that bean-research investment is economically a
proﬁtable investment under the small open country assumption. Lower values for the
economic NPV and IRR, compared to the ﬁnancial analysis estimates, are explained by
the removal of the bean seed subsidy and the beans import tariffs and quotas. The
difference is relatively small because in the ﬁnancial analysis the research-induced
supply shift also reduces the government revenues from the tariff. Regarding the
distribution of the beneﬁts (change in total surplus of MX$ 152,848,491), the change in
consumer surplus was zero, and the change in producer surplus was equal to the change
in total surplus. These results indicate that all of the beneﬁts from the change in total
surplus were captured by producers. Consumers didn’t beneﬁt from technological change
because the output price is assumed to be exogenous.

Comparison of the results for the closed vs. small open economy model indicates
that there is very little difference in the rates of return for the economic analysis when
estimated assuming a closed economy model (21.4%) vis-a-vis the small open economy
model (20.7%). This result is as expected because in both cases the seed subsidy and
bean import tariffs were removed, and the world price was used to compute the rates of
return. In the case of the ﬁnancial analysis, the rate of return is higher for the small open

economy model (21.3%) than for the closed economy model (17.5%). The difference is

178

explained by the use of the world price for the open economy model, which before the
tariff is higher than the domestic price used for the closed economy model.

The most important difference between the closed vs. the small open economy
model has to do with the distribution of beneﬁts from the research-induced supply shift.
In the closed economy model, for both the ﬁnancial and economic analysis, all of the
beneﬁts from the change in total surplus are captured by consumers, while producers
actually have a loss—when demand is inelastic—because technological change leads to a
lower domestic price. In contrast, in the case of the ﬁnancial analysis for the small open
economy model, instead of decreasing, the domestic price remains equal to the world
price. Thus, all of the beneﬁts from the change in total surplus are captured by producers,
there is no change in consumer surplus, and there is a loss in government revenues from
tariffs. In the case of the economic analysis for the small open economy model, the fact
that the output price is exogenous implies that all of the beneﬁts from the change in total
surplus are captured by producers and there is no change in consumer surplus after
technological change.

The case of the bean subsector in Mexico does not exactly matches either of these
models. In fact, under the NAFTA the subsector is in a period of transition from a
relatively closed economy to an open economy. Comparison of the estimates for the

closed vs. small open economy model is intended to help in addressing this complexity.

179

CHAPTER 7. SUMMARY AND CONCLUSIONS
7.1 Summary of Findings

7.1.1 Challenges Facing the Bean Subsector

In Mexico dry beans are the second most important crop (after maize), both in
terms of consumption and production. However, a variety of factors threaten this key
agricultural subsector. First, during 1996-2000 the total harvested area decreased by 2.0%
and average yield decreased by 2.5%, compared to 1990-1995. As a result, production
declined 4.5% in 1996-2000, compared to 1990-1995.

Second, in Mexico, beans are grown during two seasons: the spring-summer (SS)
season and the fall-winter (FW) season. However, most of Mexico’s bean crop is grown
during the SS season (84% of area and 73% of production) when yields are relatively low
and highly variable from year-to-year because the crop is mainly grown under rainfed
conditions in the semiarid highlands.

Third, since Mexico does not produce enough beans to meet its domestic
requirements, it has been importing increasingly large quantities of beans (mainly pinto
and black beans) to meet its domestic demand. Mexico has primarily imported beans
from the US, since under the NAFTA import quotas and tariffs for imports of beans
from the US. and Canada into Mexico are being phased out over the period 1994-2008.
This increasing trend in bean imports highlights the problem facing domestic producers
—the challenge of being competitive with imported beans.

Fourth, urbanization and income growth are having an important impact on
consumer preferences. Increasingly, consumers are demanding new, more health-

conscious, and convenient bean products of uniform quality. One implication is that

180

increasingly, value is being added to beans after they are marketed by farmers. Value is
added to beans through packing and the manufacturing of processed products. The
processing industry primarily produces canned and dehydrated beans. During 1994-2000
the production of processed beans increased rapidly (3,314 to 53,515 Mt). However, this
volume represents only about 4.4% of Mexico’s annual bean production. About 40% of
Mexico’s processed bean output is exported to the US. Furthermore, while population
increased by 46% during 1980-2000, total apparent consumption remain roughly
unchanged. Thus, consumption per capita decreased by 18.5% during the period.

Fifth, in recent years the supply of various bean market classes has been poorly
aligned with consumer demand. Although Mexico has a surplus of light color beans, it
has a deﬁcit of pinto and black beans. During 1993-1999, pinto beans accounted for an
average of 25%, black beans for 29%, and light color beans for 46% of national
production. On the other hand, pinto beans accounted for an average of 29%, blacks for
31%, and light color beans for 40% of national demand.

Sixth, until 1993, bean prices were determined mainly by the strong intervention
of the government through guaranteed prices. But starting in 1993, the government
initiated important changes to reduce government intervention in the agricultural sector
and to allow prices to be determined according to a freer interaction of market supply and
demand. As a result, real producer prices for beans declined 59% from 1990 to 2000.
This drop in the real bean price is associated with price liberalization, which is designed
to set prices in the domestic market to reﬂect prices in the world market, where bean
prices have been declining. In the wholesale market, the decline in real prices has

affected all market classes. During 1991-1999, domestic pinto beans faced the greatest

181

real price decrease (-50%), followed by light color beans (-25%) and black beans (-22%).
Furthermore, in the late 19903, prices to US. producer for pinto and black beans

declined, which contributed to the decline in Mexican bean prices since the US. is

 

Mexico’s most important source of imported beans.

Finally, one of the most important problems facing the bean subsector in Mexico
is marketing. This is due not only to the complexity that arises from the diversity of
demand for different market classes and quality of beans in regional markets, but is also
due to the high level of concentration in the wholesale bean market, which results in large
commissions to brokers, which in turn results in lower prices to producers and higher
prices to consumers.

7.1.2 Government Intervention to Increase Productivity

Several institutions and programs have a mandate to serve the needs of the
agricultural sector. First, the National Research Institute for Forestry, Livestock, and
Agriculture (INIFAP) is charged with improving the standards of living of farmers and
the population in general by generating technologies to increase agricultural production
and productivity. Among other activities, INIFAP scientists have generated improved
varieties for the most important crops in Mexico, including beans.

Second, The “Basic Program of Technical Assistance” (PEAT), a component of
the Alliance for the Countryside, is responsible for promoting the use of technologies
generated by research institutions and improving the quality of technical assistance
services.

Third, the “National Producer of Seeds” (PRONASE), an organization of the

federal government, promotes the production and utilization of certiﬁed seeds by

182

multiplying seed from small quantities to commercial levels. Most of PRONASE’s
certiﬁed seed is bred by scientists at INIFAP and other research centers in the country.

In the early 19903, to deal with the threat of the price liberalization, the Mexican
government developed new strategies to support agricultural development. In 1994, the
federal government started PROCAMPO, a ﬁfteen-year cash transfer program to help
farmers make the transition from a government-regulated to a market-driven agricultural
sector. As a complement to PROCAMPO, in 1996 the government started “Alianza para
el Campo” (Alliance for the Countryside), a broad strategy for promoting agricultural
production and rural development. For bean farmers, one of the most relevant
components of Alliance for the Countryside is the Kilo per Kilo program, which has the
objective of increasing yields by encouraging farmers to substitute modern varieties for
traditional varieties.

Since 1996, PRONASE has participated in Alliance for the Countryside by
providing certiﬁed seed to the Kilo per Kilo program, which distributes certiﬁed seed
(but also noncertiﬁed seed) to farmers at a price equivalent to that of commercial grain
for a maximum of ﬁve hectares. A federal subsidy covers up to 45% of the difference
between the price of the improved seed and the price of commercial grain. In addition,
state governments are allowed to cover up to 55% of the price difference. This signiﬁcant
reduction in price makes improved seed very affordable to farmers. During 1996-2000,
the Kilo per Kilo program distributed 12,077 Mt of bean seed to 127,209 farmers, who
planted 384,525 ha in Chihuahua, Durango and Zacatecas in northern Mexico. SAGAR
estimated that yields in the demonstration plots increased from 26% to 44%. The quantity

of seed distributed increased from 1996 to 1998, but decreased from 1998 to 2000,

183

depending on the annual budget assigned to the Kilo per Kilo program. During 1996-
2000, SAGAR’s total investment in the program was MX$ 90,924,000, most of which
(98%) represented the cost of the seed that averaged MX$ 7.53 per kilo. Two categories
of seed were distributed: certiﬁed seed (25%) and non-certiﬁed seed (75%), although
some varieties were distributed under both categories. The certiﬁed varieties were
PintoVilla (34%), Flor de Mayo (29%), Pinto Mestizo (18%), and other (19%). The non-
certiﬁed varieties were Pinto Nacional (51%), Negro Criollo (21%), Pinto Villa (11%),
and other (17%).

7.1.3 Farmers’ Adoption of Improved Bean Varieties

During August-October 2001, a stratiﬁed random sample of bean farmers in
Northern Mexico was selected and the farmers were interviewed. In the analysis, the
farmers were divided into two groups: non-adopters and adopters. For Durango (n=206),
43% of the respondents were non-adopters and 57% were adopters. For Chihuahua
(n=163), 17% were non-ad0pters, and 83% were adopters. For Zacatecas (n=86), 38%
were non-adopters, and 62% were adopters. In total (n=455), 33% of the sampled farmers
were non-adopters and 67% were adopters.

When comparing adopters vs. non-adopters, the statistical analysis indicates that
these groups are quite similar with respect to age, experience, education, land tenure,
farm size, bean area, soil quality, soil preparation, planting dates, cultural practices, and
number of family members having off-farm employment. Adopters’ participation in the

Kilo per Kilo program was 65% in Durango, 35% in Chihuahua, and 45% in Zacatecas.

184

The analysis of farmers’ selling price by market class shows that in general, light
color beans sold for the highest prices, followed by pintos, and black beans—which sold
for the lowest price. However, there were few statistically signiﬁcant differences.

The percentage of the total area planted to improved varieties varied greatly from
state-to-state, ranging from 71% in Chihuahua (Pinto Villa 56% and Pinto Mestizo 15%),
to 42% in Durango (Pinto Villa 38% and Pinto Mestizo 4%), and 8% in Zacatecas (Negro
Zacatecas).

The statistical analysis of farmers’ yields indicates that while improved pintos had
higher yields than traditional pintos (20.6%), there is no evidence that improved varieties
of other market classes (black and light-colored beans) have higher yields than their
respective traditional variety.

Finally, the surveyed adopters were asked about their perceptions regarding the
advantages of improved varieties, compared to traditional varieties. The analysis of these
data indicates that better consumption quality, market acceptance, fewer days to maturity
(which is related to drought resistance), diseases resistance, and higher yields are the
most salient characteristics that inﬂuenced farmers’ decision to adopt improved varieties.

7.1.4 Econometric Analysis of Program Participation

The descriptive statistics show that the Kilo per Kilo program is the driving force
for the adoption of improved bean varieties in northern Mexico. Almost all farmers who
planted improved bean varieties either participated in the program during 2001, or had
participated in an earlier year. Therefore, an econometric analysis of program
participation was carried out to investigate the factors explaining the participation of

farmers in the Kilo per Kilo program. A probit model, with the binary dependent variable

185

“program” to indicate program participation, was estimated using the maximum
likelihood method.

The regression analysis indicates that smaller farmers are more likely to
participate in the program than larger farmers. The results also suggest that the farmer’s
perception of a higher percentage of dry years, the use of insecticide, tied-ridger
ownership, and a greater distance from the farm to the city have a positive effect on
program participation.

The econometric analysis conﬁrms that in northern Mexico the participation of
farmers in the seed distribution program is associated mainly with farm size and location.
Smaller farmers located in marginal areas are more likely to participate in the program
than larger farmers with better location (i.e., closer to the city), which is consistent with
the objectives of the Kilo per Kilo program for the distribution of improved bean
varieties.

7.1.5 Returns to Bean-Research Investment

The returns to the research investment for improved bean varieties in the north of
Mexico are evaluated using the economic surplus method. During the period 1990-2000,
INIFAP released several improved bean varieties for rainfed conditions, with the support
of the Bean/Cowpea CRSP. For the purpose of the economic evaluation of bean research,
the relevant improved varieties are Pinto Villa and Pinto Mestizo. The statistical analysis
indicates that in both Chihuahua and Durango, the yields of these two varieties are
signiﬁcantly higher (20.6%) than the traditional pinto market class variety, Pinto

Nacional.

186

The evaluation of returns to research was carried out using both a closed economy
model and a small Open economy model. In a closed economy model, the price of the
commodity is determined inside the country and in a small open economy model the
price is exogenous. In this study, a closed economy model is justiﬁed by the fact that the
domestic price for beans showed a clear decreasing trend during 1990-2000, while a
decreasing trend in the international price was not so evident. However, the small open
economy model is justiﬁed by the increasing inﬂuence of the world price on the domestic
price and this model take into account the effect of barriers to trade (i.e., tariffs and
quotas).

Closed Economy Model

The results from the ﬁnancial analysis indicate that the NPV of the investment in
bean research is MX$ 17,681,052 (base-year=1994), when using a discount rate of 10%,
which is the estimated cost of opportunity of capital (OCC) in agricultural activities in
Mexico, and that the IRR is 17.5%. Therefore, from the ﬁnancial point of view, investing
in generating and promoting the use of improved bean varieties in northern Mexico has
been a proﬁtable investment under both the NPV and the IRR criteria. Regarding the
distribution of the beneﬁts between consumers and producers (change in total surplus of
MX$ 107,051,062), over the period 1992-2000 (i. e., the period when the improved
varieties were adopted), the change in consumer surplus was MX$ 204,052,485, and the
change in producer surplus was MX$ -97,001,423. These results indicate that consumers
captured all the beneﬁt from the change in total surplus and that producers actually had a

loss, as a result of the demand being inelastic.

187

The results from the economic analysis indicate that the NPV of the investment in
bean research is MX$ 29,420,210, (OCC=10%), and that the [RR is 21 .4%. Therefore,
from the point of view of the society as a whole, investing in generating and promoting
the use of improved bean varieties in northern Mexico has been a proﬁtable investment
under both the NPV and the IRR criteria. The economic NPV and [RR are higher than the
ﬁnancial analysis estimates because the higher price of imported beans (IPP), compared
to domestically produced beans, more than offsets the increase in the price of the
improved seed that resulted from removing the seed subsidy.

Regarding the distribution of the beneﬁts (change in total surplus of MX$
153,116,040), the change in consumer surplus was MX$ 251,644,325, and the change in
producer surplus was MX$ -98,528,285. These results indicate that the consumers
captured all the beneﬁts from the change in total surplus and that producers actually have
a loss, as a result of the demand being inelastic.

Sensitivity analysis was carried out to assess the sensitivity of the rate of return to
the parameters used in the model. In the ﬁnancial analysis, the switching values17 for key
variables were: adoption rate (29% decrease), incremental yield (26% decrease), and
domestic price (28.4% decrease). In the economic analysis the switching values for key
variables were: adoption rate (40.5% decrease), incremental yield (33% decrease), and
import parity price (3 5% decrease). The switching values indicate that the IRR is less
sensitive to changes in the market exchange rate, research and extension cost (seed price

subsidy), and demand elasticity estimates.

 

'7 The switching value is the percentage change in the variable for the NPV to become zero or the IRR to
fall to the cut-off rate (i.e., the OCC=10 percent).

188

The economic rate of return is also sensitive to the estimates for the elasticity of
supply. If a supply elasticity of 0.1 is assumed, the economic IRR increases from 21.4 to
48.8 percent. In contrast, if a supply elasticity of 0.9 is assumed, the economic IRR
decreases from 21.4 to 2.6 percent, which is lower than the opportunity cost of capital
(10%). Break-even values is e=0.77 for the economic IRR. This means that research
investment remains proﬁtable for e<0.7 7 in the economic analysis.

In addition, sensitivity analysis was used to assess the distribution of the beneﬁts
between consumers and producers. If a supply elasticity (c) of 0.1 is assumed (and
n=0.5), there is a large increase in total surplus and the largest share of the beneﬁts go to
the producers. If a supply elasticity of 0.9 is assumed, there is a large decrease in total
surplus and the consumers capture all beneﬁts, while producers actually lose as a result of
a more elastic supply. In contrast, the distribution of beneﬁts between consumers and
producers is not very sensitive to the estimates for the elasticity of demand (1]).

ﬁnial Open Economy Model

The results from the ﬁnancial analysis indicate that the NPV of the investment in
bean research is MX$ 26,331,431 (OCC=10%) and that the IRR is 21.3%, which implies
that bean-research investment is ﬁnancially a proﬁtable investment. However, regarding
the distribution of the beneﬁts (change in total surplus of MX$ 85,25 8,806), the change
in the consumer surplus was zero, the change in the producer surplus was MX$
224,3 72,23 7, and the change in government revenues was MX$ -139,113,431. These
results indicate that all of the beneﬁts from the change in total surplus were captured by

producers. Consumers didn’t beneﬁt from technological change because the output price

189

is assumed to be exogenous. Additionally, there was a loss in tariff revenues to the
government.

The results from the economic analysis indicate that the NPV of the investment in
bean research is MX$ 24,405,096 (OCC=10%) and that the IRR is 20.7%, which implies
that bean-research investment is economically a proﬁtable investment. Lower values for
the economic NPV and IRR, compared to the ﬁnancial analysis estimates, are explained
by the removal of the bean seed subsidy and the beans import tariffs and quotas. The
difference is relatively small because in the ﬁnancial analysis the research-induced
supply shift also reduces the government revenues from the tariff. Regarding the
distribution of the beneﬁts (change in total surplus of MX$ 152,848,491), the change in
the consumer surplus was zero, and the change in the producer surplus was equal to the
change in the total surplus. These results indicate that all of the beneﬁts from the change
in total surplus were captured by producers. Consumers didn’t beneﬁt from technological
change because the output price is assumed to be exogenous.

Closed vs. Sm_aﬂ_()pen Econongr Model

The results for the closed vs. the small open economy model indicate that there is
little difference between the economic rates of return under the assumptions of the closed
economy model (21.4%) vis-a-vis the small open economy model (20.7%). This result is
as expected because in both cases the seed subsidy and bean import tariffs were removed,
and the world price was used to compute the rates of return. In the case of the ﬁnancial
analysis, the rate of return is higher for the small open economy model (21.3%) than for

the closed economy model (17.5%). This difference is explained by the use of the world

190

price in the open economy model, which before the tariff is higher than the domestic
price used in the closed economy model.

The most important difference between the closed vs. the small open economy
model has to do with the distribution of beneﬁts from the research-induced supply shift.
In the closed economy model, for both the ﬁnancial and economic analysis, all of the
beneﬁts from the change in total surplus are captured by consumers. Producers actually
have a loss—when demand is inelastic—because technological change leads to a lower
domestic price.

In contrast, in the case of the ﬁnancial analysis for the small open economy
model, instead of decreasing, the domestic price remains equal to the world price. Thus,
all of the beneﬁts from the change in total surplus are captured by producers, there is no
change in consumer surplus, and there is a loss in government revenues from tariffs. In
the case of the economic analysis for the small open economy model, the fact that the
output price is exogenous implies that all of the beneﬁts from the change in total surplus
are captured by producers and there is no change in consumer surplus after technological
change.

The case of the bean subsector in Mexico does not exactly matches either of these
models. In fact, during the years covered by this study when NAFTA was evolving, the
subsector was in transition from a relatively closed economy to an open economy. Thus,
the estimates for both the closed and the small open economy models are presented to
illustrate the impact of the closed vs. small open economy assumptions on the rate of

return and on the distribution of the beneﬁts from research.

191

7.1.6 Beneﬁts for Consumers and Producers

In the case of the closed economy model, the results from the ﬁnancial and
economic analysis show that consumers are the major beneﬁciaries of bean research,
because they capture all the increase in the total surplus. The development and diffusion
of improved bean varieties directly beneﬁts all consumers by making larger quantities of
beans (supply shift) available at lower prices. However, given that both consumption
preferences and demand response to price changes vary with income, the impact among
income classes differs. Because most staple commodities (e. g., dry beans) are inferior
goods (i. e., have negative income elasticities), poorer people consume a greater quantity
than richer consumers (Alston et al. 1998). Individually, low-income consumers tend to
beneﬁt relatively more from research to generate improved bean varieties than research
on non-staple commodities (e.g., export crops) because low-income consumers spend a
high proportion of their income on beans.

In the case of the small open economy model, the results ﬁom the ﬁnancial and
economic analysis show that producers capture all of the beneﬁts from research, but the
impact among producer groups differs. Producers with different incomes, with different
farm sizes, in different locations, and with diverse tenure situations can gain or lose,
depending on the suitability of the new technology to their particular circumstances.

In this study, the evidence from the econometric analysis of program participation
(section 5.5) indicates that in northern Mexico farmer participation in the seed
distribution program is associated mainly with farm size and location. Smaller farmers
located in marginal areas are more likely to participate in the program. Therefore, they

beneﬁt more from bean research than larger farmers with a better location (i. e. , closer to

192

the city). In addition, the observed decline in the real price of pinto beans (section 3.3. 1)
suggest that early adopters beneﬁted more from technological change than late or non-
adopters. This is because at the beginning of the adoption process the price for improved
pintos was higher than the price of traditional pintos—until improved pintos gradually

ﬂooded the market (section 6.6.3).

7.2 Policy Implications

These results have implications for government decision-makers, bean
researchers, and the extension services—all of whom are important agricultural research
investment stakeholders.

7.2.1 Implications for Government Decision-Makers

It is widely accepted among economists that there is market failure in the
provision of agricultural research by the private sector and that, without action by the
government, there would be an underinvestment in research. Market failures arise when
ﬁrms cannot appropriate all of the beneﬁts of their research investments and as a result
the market does not provide the private sector with sufﬁcient incentives for it to invest at
the level that would be best from society’s point of view.

Given that beans are “self-pollinated”, farmers do not have to purchase new seed
every year. Thus, improved bean varieties have the characteristics of a public good““.
With appropriate management, farmers can save and share or sell the seed to other
farmers, without experiencing a signiﬁcant reduction in yields from one year to the next.

This leads to a lack of ﬁnancial incentives, which explains why there is minimal private

 

'8 A good that once produced is available to all on a nonexclusive basis.

193

sector19 participation in research to generate improved bean varieties, bean seed

production, and seed marketing. The positive results from this economic analysis indicate

that there is a strong case for public investment in research and extension to develop and
distribute improved bean varieties in rainfed areas of northern Mexico.

In northern Mexico, farmer adoption of improved bean varieties has been highly
dependent on government investment (SAGAR) in bean research (ARTTP-INIF AP), seed
production (PRONASE and SNICS), seed distribution (Kilo per Kilo), and technical
assistance (PEAT). Thus, the government should continue to invest in these programs to
ensure that research beneﬁts society, especially poor consumers and small farmers.

7.2.2 Implications for Bean Researchers

The success of the improved pinto bean varieties (i. e., Pinto Villa and Pinto
Mestizo) in Chihuahua and Durango; the success of the collaborative research between
INIFAP and the Bean/Cowpea CRSP, which led the release of these varieties; the limited
impact of improved varieties in Zacatecas; the increasing trend in bean imports; and
changes in consumer preferences have the following implications:

. The bean research program should continue to focus on generating improved varieties
that perform well in farmers’ ﬁelds and are highly acceptable in the market. To
achieve this goal, social scientists must continue to solicit feedback from farmers and
market agents regarding their needs and preferences.

. Further collaboration with international programs (e. g. , the Bean/Cowpea CRSP) that

supports international research partnerships or strategic alliances to increase the

 

’9 Although some farmers make contracts with the government through PRONASE to produce certiﬁed
seed for the Kilo per Kilo program.

194

availability of beans and to increase the scope and the efﬁciency of research
investments should be supported.

. The breeding program needs to strengthen its efforts to develop improved bean
varieties of the dominant market class grown in Zacatecas (black beans) that have
higher yields than existing varieties and other characteristics that are desired by
producers and domestic consumers. This is very important because Zacatecas
accounts for 37% of the total rainfed bean area, which is more than the area of
Chihuahua and Durango together (25%).

. Priority should be given to generating improved bean varieties in the market classes
in which there are large deﬁcits that are currently being met through imports (e. g.,
pinto and black beans) vs. those market classes for which there is a surplus and face
marketing problems (e. g., light color beans).

. Given the decreasing trend in consumption per capita and the increasing demand for
value-added products, closer coordination is needed among researchers, producers,
and processors to insure that future varieties meet the requirements of the industry
and niche export markets.

. Continued collaboration between agronomist and social scientists is needed to
periodically evaluate the on-farm performance of improved bean varieties after they
are released and distributed to farmers.

7.2.3 Implications for Extension Services
The Kilo per Kilo seed distribution program has played a key role in the diffusion
of research results and farmer adoption of new bean technologies. However, the

variability in the operations of the program from year-to-year, from one state to another,

195

and in the type of seed distributed (certiﬁed vs. non-certiﬁed), as well as the scarcity of

seed, implies that the program needs to:

. Identify more stable funding mechanisms that will allow the program to set priorities
and make plans to promote the improved varieties, without there being a lack of
continuity in seed distribution.

. Provide better ﬁnancial incentives20 in the contracts with the growers who multiply
the seed, to ensure that they produce a sufﬁcient quantity and diversity of seed to
meet the farmers’ seed demand for different production regions and market classes.

. Increase the percentage of certiﬁed seed distributed (25%) vs. the noncertiﬁed seed
(75%), given that farmers prefer certiﬁed seed because its physical, physiological,
phytosanitary, and genetic quality are guaranteed.

. Provide farmers training on the appropriate management practices required to
preserve seed quality from season-to-season, so that the program doesn’t have to
provide seed to the same farmers every year and, thereby, more farmers will beneﬁt

from the seed distribution program.

7.3 Limitations of the Study
The most important limitations of this study are:
. The surveys focused on the main bean-producing counties in each state where the

Kilo per Kilo seed distribution program operated. However, there are other counties

 

“0 Interviews with key informants indicated that from 1999 to 2000 the price paid to growers by the
government for certiﬁed seed decreased and that swd growers believe that the price they receive hardly
covers the costs of production of the seed and the risks they have to face.

196

producing beans and where the program operated, which were not included in the
surveys.

. The survey included current data for yields in 2001, but the information used for
1997-2000 was based on farmers’ recall. This historical information was likely less
accurate than the current data and the historical data contained missing values for
earlier years.

. There are no empirically-based estimates of supply and demand elasticities for dry
beans in Mexico, so the estimates for these economic parameters were assumed based
on estimations reported for the same or similar crop in other countries.

. There are intangible costs and beneﬁts (i.e., pollution ﬁom chemical inputs or
preventing deterioration of the environment) associated with the impact of bean
research that were not addressed in this study because, even if they can be identiﬁed,
by deﬁnition they are hard to quantify and are difﬁcult to value in monetary terms.

. A risk analysis was not included because information on the probability of changes in

the key variables was not available to estimate their effect on research beneﬁts.

7.4 Future Research Needs
The most important research issues that need to be addressed by future
socioeconomic research studies are to:
. Apply the economic surplus approach to estimating the returns to research investment
to other important crops and regions in Mexico, and to research results from INIFAP

and other centers in the country for improved varieties and other technologies.

197

Assess options for improving the price that farmers receive for their beans by
marketing collectively to processors, urban wholesalers, and to export markets.
Assess the feasibility of farmers’ groups adding value to their product by cleaning,
sorting, grading, and packing their beans.

Identify alternative institutional arrangements for multiplying and marketing seed of
improved varieties, given that the availability of certiﬁed seed represents a bottleneck
for the diffusion of the improved bean varieties.

Disaggregate the consumer and producer surplus (that results from the research-
induced supply shift for beans) by using various categories of domestic consumers
(e. g., according to income class) and various categories of producers (e. g., according
to farm size, tenure, or location) to identify the impact of technical change for each
group.

Analyze the multiple social objectives of research (e. g., economic efﬁciency, equity,
reduce income risk, and national food security or self-sufﬁciency) to determine
whether or not those objectives can be achieved through public-sector agricultural
research or can be better pursued using alternative policies.

Carry out an institutional analysis to better understand how the institutional setting
affects economic performance. Institutions affect how research and extension
organizations operate, so it is important to analyze how change in institutions (e. g.,
property rights and policies) can affect the magnitude and distribution of research

beneﬁts.

198

APPENDICES

199

Appendix A: Farmer Questionnaires in Spanish and English

ENCUESTA SOBRE ADOPCION DE VARIEDADES MEJORADAS DE FRIJOL EN EL
DISTRITO DE GUADALUPE VICTORIA, DURANGO, MEXICO 2001
INIFAP-SAGARPA

 

 

 

INFORMACION BASICA

Nombre del Agricultor Numero de Cuestionario
Localidad Municipio

Nombre del encuestador Fecha

 

 

DECLARACION DE CONSENTIMIENTO
(Antes de comenzar la entrevistg, lea la siguiente declaracion al agricultor]

Mi hombre es Horacio Gonzalez. Estoy llevando a cabo un estudio sobre la produccion de
frijol en Mexico. Me gustarla hacerle algunas preguntas acerca de sus aaividades de
produccio'n de frijol. La informacidn que usted provea serd usada para entender mejor las
decisiones de produccién de frijol de los agricultores, especialmente su eleccién de 103
variedades que ellos siembran. La informacidn que yo obtenga de usted y de otros productores
de frijol en el drea serd usada para completar mi disertacidn doctoral en el Departamento de
E conomt'a Agricola en Michigan State University, E UA.

Este estudio es ﬁnanciado por el INIFAP, en colaboracidn con el Proyecto de
F inanciamiento e Investigacidn Colaborativa (CRSP) para F riiol/Cowpea de Michigan State
University, E UA.

Su participacién es VOL UNTARIA. El que usted se rehrise 0 se retracte de participar en
el estudio no implica ninguna accidn penal ni la pérdida dc ningrin beneﬁcio. Completar la
entrevista tomard aproximadamente 60 minutos. Usted es Iibre de N0 CON T ES TAR
cualquiera de las preguntas que yo le haré. Sin embargo, espero que usted esté de acuerdo en
contestar todas mis preguntas, ya que sus respuestas me ayudardn a entender mejor el sistema
de cultivo de frijol en Durango y por qué los agricultores preﬁeren diferentes variedades de
frijol. T oda la informacién que usted provea se mantendrd CONFIDENCIAL. Esta significa
que sus respuestas a mis preguntas no serdn mostradas a nadie mds. Nadie conocerd sus
respuestas excepto yo, y su identidad serd protegida en cualquier reporte basado en la
informacidn. Su privacidad serd protegida hasta el llmite mdximo permitido por la ley.

Si usted tiene cualquier pregunta acerca de este estudio, por favor contacte al Profesor
Richard Bernsten, Departamento de Economt'a Agricola, Michigan State University, 211 E
Agriculture Hall, East Lansing, MI 0 (lame al Profesor Bernsten al (517) 355-3449. Si tiene
preguntas sobre sus derechos y su papel como objeto humano de investigacidn, por favor
contacte al Dr. David Wright, Michigan State University, 248 Administration Building, East
Lansing, MI 0 Home al Dr. Wright al (51 3 355-2180.

Estoy de acuerdo en que mi participacién es voluntaria y acepto ser entrevistado.

 

Firma del Agricultor

200

l. INVENTARIO DE TIERRAS

Primero me gustaria preguntarle sobre las parcelas que sembro en el aﬁo 2001.

1.1 gConsiderando todas las parcelas cuantos diferentes cultivos sembro en 2001?

1.2 g,Cuantas hectareas sembro de cada cultivo (incluyendo todas las parcelas)?

1.3 1,C6mo clasiﬁcaria la calidad de la tierra de las parcelas (Bucna; Regular; Mala)?

1.4 g,Qué derechos de propiedad tiene sobre las parcelas (Ejido; Pequeﬁa Propiedad; Aparccria)?

1.5 (Que tipo de semilla sembro de cada cultivo (Certiﬁcada, Mejorada, Criolla)?

 

1.1 Cultivo 1.2 Hectareas 1.3 Calidad 1.4 Tenencia 1.5 Tipo de Semilla

 

l Frijol

 

2

 

3

 

 

 

 

 

 

 

Tipo de semilla: C = certiﬁcada; M = mejorada o grano apto; R = Criolla

II. ACCESO A INFORMACION TECNICA

En seguida, me gustaria preguntarle de que' forma recibe Ud. informacién tocante a mejores
maneras dc producir ﬁijol y acerca de la existencia de nuevas variedades dc frijol.
2.1 En el ultimo aﬁo (,que' asesoria técnica recibio usted para mejorar sus rendimientos de frijol?

(Especiﬁquc)

 

 

En el ultimo aﬁo, usted ha recibido informacion sobre nuevas variedades por medio de:

2.2 Radio/TV(O=No,1=Si)___
2.3 Periodico ........................................................................... (0=No, 1=Si)__
2.4 Folleto dc cxtensién ............................................................. (0=No, 1=Si)__
2.5 SAGARPA ......................................................................... (0=No, 1=Si)___

201

 

2.6 Compaﬁia de seguros ........................................................... (0=No, 1=Si)

2.7 Banco .............................................................................. (0=No, 1=Si)—
2.8 Otro agricultor ................................................................... (0=No,1=Si)___
2.9 Parcelas demostrativas ......................................................... (0=No, 1=Si) __
2.101NIFAP (0=No, 1=Si)—
2.11 Te'cnicos PEAT/SINDER(0=N0,1=Si)__

2.12 Otro (Especiﬁque)

 

USO DE INSUMOS (indique que labores realizé)

2.13 Barbecho (SI/NO) 2.17 Primer escarda (SI/NO)

2.14 Rastreo (SI/NO) 2.18 Deshierbo (SI/NO) __
2.15 Fecha de siembra 2.19 Segunda escarda (S/N) __
2.16 Fertilizo (SI/NO) 2.20 Insecticida (SI/NO) __

III. VARIEDADES SEMBRADAS DE FRIJOL Y COSTO

Ahora quisiera preguntarle sobre las variedades dc frijol que ha scmbrado en los ultimos aﬁos.
3.1 gQué variedades de frijol ha scmbrado en los aﬁos 1997-2001 (Certif, Mejorada, Criolla)?
3.2 (Donde obtuvo la semilla que sembro?

3.3 g,Que' precio pago por la semillla ($/kg)?

3.4 g,Cuantas hectarcas sembro dc cada variedad?

3.5 g, Que’ rendimiento obtuvo de cada variedad (kg/ha)?

 

 

 

 

 

 

 

 

 

 

 

Afro 3.1 Variedad 3.2 Donde obtuvo la 3.3 Precio 3.4 Superﬁcie 3.5 Rendto
(C, M, R) Semilla de frijol (S/kg) (ha) (kg/ha)
l
2001 2
3
(Continua)

202

(Continuacion)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Aﬁo 3.1Variedad 3.2 Donde obtuvo la 3.3 Precio 3.4 Superﬁcie 3.5 Rendto
(C, M, R) semilla dc frijol ($/kg) (ha) (ton/ha)
1
2000 2
3
1
1999 2
3
1
1998 2
3
1
1997 2
3

 

 

 

 

 

 

 

Tipo de semilla: C = certiﬁcada; M = mejorada o grano apto; R = Criolla

3 .6 g,Generalmente puede usted obtener suﬁciente cantidad de semilla de la variedad de frijol que
desea sembrar? ..................................................................... (0=No, 1=Si)

SI LA RESPUESTA ES “NO”
3.7 LCual es el problema? (Especiﬁque)

 

3.8 gDe los pasados 10 aﬁos cuantos fueron aﬁos malos para producir frijol?

 

3 .9 gCuantas personas ocupa cultivando frijol? Familia + Ud. Contrata

3.10 (Cual fue su costo de producccion por hectarea ($/ha)?

 

IV. CREDITO
En seguida, quisiera preguntarle en relacion con su disponibilidad dc credito.

4.1 Durante el aﬁo 2001, recibio algt'rn crédito para ﬁnanciar sus actividades de produccion de

frijol‘7(0=No,1=Si)

203

SI LA RESPUESTA ES “SI”

4.2 (gDe quién recibio el crédito? (Especiﬁque)

 

4.3 g,Cuanto crédito recibio ($/ha)?

 

4.4 [Para qué uso el crédito? (Especiﬁque)

 

v. MAQUINARIA

En seguida, me gustaria preguntarle acerca la maquinaria que usa en la produccion de frijol.

5.1g,Tieneusted tractorpropio?(0=No,1=Si)

SI LA RESPUESTA ES “NO”

5.2 (Como obtiene acceso a un tractor? (Especiﬁque)

 

5.3 g,Tiene usted pileteadorapropia? (0=No, 1=Si)

VI. VENTAS DE FRIJOL
En seguida, me gustaria preguntarle en tomo a sus ventas de ﬁijol.

6.1 g,Cuantas toneladas vendio de cada variedad en el aﬁo 2000?
6.2 (A quién le vendié cada una dc estas variedades?

6.3 (gCual fue el precio que recibio por cada variedad ($/kg)?
6.4 (gCree Ud. que éste es un buen precio para ésta variedad?

6.5 SI LA RESPUESTA ES “NO”: {,Por qué no?

 

 

 

 

 

Nombre de 6.1 Ventas 6.2 A quie'n 6.3 Precio 6.4 g,Buen 6.5 Si “NO”
la Variedad (ton) 1e vendio? ($/kg) precio? gPor qué no?
1

2

3

 

 

 

 

 

 

204

 

6.6 g,Cuantos de su familia trabajan fuera de la parcela (incluye'ndolo a Ud.)?
6.7 g,Cuantos meses a1 aﬁo trabajan? Ud

6.8 g,Usualmente qué hace Ud. con la paja de su cosecha de ﬁijol? (Especiﬁque)

csposa

OtI'OSI a

 

b

 

 

 

6.9 (Cree usted que algunas variedades producen mas paja? .............. (0=No, 1=Si)

6.10 SI LA RESPUESTA ES “SI”: g,Cuales variedades producen mas paja? (Especiﬁque)

 

VII. EVALUACION DE NUEVAS VARIEDADES DE FRIJOL
En seguida, me gustaria preguntarle su opinion sobre las variedades de frijol que el INIFAP ha

liberado durante los ultimos 10 afios, comparadas con su semilla criolla

(Como compara la variedad mejorada con la criolla, en cuanto a:

7.1 Rendimiento (ton/ha)?

7.2 Accptacion del mercado?

7.3 Resistencia a sequia?

7.4 Resistencia a plagas?

7.5 Escape a heladas?

Escala: P = Peor

= Igual M = Mejor N = No opina/No sabe

 

(Nombre de la criolla)

7.9 Dias a madurez?

7.8 Tipo de planta (forma)?

7 . 10 Calidad para consumo?

7.6 Desgrane de vaina a la cosecha?

7.7 Resistencia a enfermedades?

 

Variedad
Mej orada

7.1
RE

7.2
AM

7.3
RS

7.4
RP

7.5
EH

7.6
DC

7.9
DM

7.10
CC

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(Var.Mej. = Pinto Villa, P.Mestizo, P.Bayacora, Negro Altiplano, Negro Sahuatoba, etc.)

205

 

VIII. PARTICIPACION EN PROGRAMAS DEL GOBIERNO

 

 

 

8.1 (gPara cuantas hectareas recibio subsidio de Procarnpo ($829/ha)? ha

8.2 (gCuantos kilos de semilla certiﬁcada recibio del prograrna Kilo por Kilo? kg
8.3 g,Que' variedad recibio de KxK y a qué precio? $/kg
8.4 g,Otro apoyo del gobiemo (Pronasol)? g,Cual? $/ha

 

IX. INFORMACION SOCIOECONOMICA
En seguida, me gustaria hacerle unas preguntas relacionadas con usted.

9.1 g,Qué edad tiene?

 

9.2 g,Hasta que’ aﬁo estudio?

 

9.3 (For cuantos aﬁos ha scmbrado frijol?

 

9.4 (3E3 Ud. miembro de alguna asociacion de agricultores? 1,Cual?

 

9.5 SI LA RESPUESTA ES “SI”: (Esta asociacion ha inﬂuenciado su decision acerca de la
variedad de frijol que ha scmbrado (0=No, 1=Si)

9.6 SI LA RESPUESTA ES “SI”: (,Como? (Especiﬁcar)

 

9.7 g,Usualmente cuantas toneladas dc frijol almacena usted antes de vender? ton

9.8 (For 10 general cuantos meses almacena el ﬁijol antes de vender? meses

 

X. COSTOS DE TRANPORTE

10.1 (,Cual es la distancia km y tiempo hr de su parcela a la ciudad mas cercana?

iGRAClAS POR SU AYUDA!

A. (“Le gustaria hacer algun comentario adicional?

 

 

B. Comentarios adicionales del encuestador

 

 

C. g,Podria proporcionamos una muestra del frijol que cosecho este afro?

206

SURVEY ON ADOPTION OF IMPROVED BEAN VARIETIES IN THE DISTRICT OF
GUADALUPE VICTORIA, DURANGO, MEXICO 2001
INIFAP—SAGARPA

 

 

 

 

 

 

BASIC INFORMATION

Farmer’s Name Questionnaire ID Number
Locality Municipality
Enumerator’s name Date

CONSENT STATEMENT

(Before beginning the interview, read the following statement to the farmer]

My name is Horacio Gonzalez . I am conducting a study of bean production in Maico. I’d
like to ask you some questions about your bean production activities. The information that you
provide will be used to better understand farmers ’ bean production decisions, especially their
choice of the varieties that they plant. The information that I collect from you and other bean
farmers in the area will be used to complete my doctoral dissertation in the Department of
Agricultural Economics at Michigan State University, USA.

This study is sponsored by INIFAP, in collaboration with the Bean/Cowpea Collaborative
Research and Support Project (CRSP) at Michigan State University, East Lansing, Michigan,
USA.

Your participation is VOLUNTARY. Your refusal to participate in or to withdraw from the
study carries no penalty or loss of any ben efits. The interview will take about 60 minutes to
complete. You are free to NOT ANS WE_R any of the questions that I will ask you. However, I
hope that you will agree to answer all of my questions, as your answers will help me to bater
understand the bean farming system in Durango and why farmers prefer different bean
varieties. All of the information that you provide will be kept CONFIDENTIAL This means
that your answers to my questions will not be shown to anyone else. No one will know your
answers except me and your identity will be protected in any report based on the data. Your
privacy will be protected to the maximum extent allowable by law.

If you have any questions about this study, please contact Professor Richard Bernsten,
Department of Agricultural Economics, Michigan State University, 2 I I E Agriculture Hall,
East Lansing, M1 or phone Professor Bernsten at (510 355-3449. If you have any questions
regarding your rights and role as a human subject of research, please contact Dr. David
Wright, Michigan State University, 248 Administration Building, East Lansing, MI or phone
Dr. Wright at (5 I 7) 355-2180.

I agree that my participation is voluntary and accept to be interviewed.

Farmer’s signature

 

207

1. LAND INVENTORY

First, I’d like to ask about each of the ﬁelds that you planted in 2001.

1.1 Including all ﬁelds how many crops did you plant in 2001?

1.2 Including all ﬁelds how many hectares did you plant to each crop?

1.3 How would you assess the quality of the soil of the ﬁelds [Good, Regular, Bad]?
1.4 What is the tenure status f or the ﬁelds [Ejido; Pequeﬁa Propiedad; Aparceria]?

1.5 Which kind of seed did you use for each crop [Certiﬁed, Improved (noncertiﬁed), Landrace]?

 

1.1 Crop 1.2 Hectares 1.3 Soil quality 1.4 Tenure 1.5 Kind of seed

 

l Beans

 

2

 

 

3

 

 

 

 

 

Kind of seed : C = Ceriﬁed; M = Improved (noncertiﬁed); R = Landrace

II. ACCESS TO TECHNICAL INFORMATION
Next, I would like to ask you how you receive information regarding better ways to grow beans
and the availability of new bean varieties.

2.1 Last year what technical advice did you receive to increase your bean yields? (Specify)

 

Last year you received information about new varieties from the following means:

2.2 Radio/TV (O-No, 1=Yes)___
2.3 Newspaper (0=No, 1=Yes)_
2.4Extensionbrochure...........................................................(0=No,1=Yes)__
2.5 SAGARPA (0=No, 1=Yes)_

208

 

2.6 Insurance Company .........
27Thebank
2.8 Anotherfarmer
2.9 Demonstrationplots

2.11 PEAT/SlNDERtechnicians

2.12 Other (specify)

(0=No, 1=Yes) _
(0=No, 1=Yes) __
(0=No, 1=Yes) __
(O=No, 1=Yes) __
.. (0=No, 1=Yes) __

(0=No, 1=Yes)

 

INPUT USE (point out the ones that apply)
2.13 Fallowing (Y ES/NO)
2.14 Harrowing (YES/NO)

2.15 Planting date

 

2.16 Fertilizer (YES/NO)

2.17 First weeding (YES/NO)
2.18 Hand weeding (Y ES/NO)
2.19 Second weeding (Y /N)

2.20 Insecticide (YES/NO)

lII. BEAN VARIETIES PLANTED AND PRODUCTION COST

Next, I’d like to ask you about the bean varieties that you have planted during the past 5 years.
3.1 What bean varieties have you planted from 1997 to 2001 [Certiﬁed; Improved ; Landrace]?

3 .2 Where did you obtain the seed you planted?
3.3 How much did you paid for the bean seed [MX$/kg]?

3.4 How many hectares did you plant to each variety?
3 .5 What was the yield for each bean variety [kg/ha]?

 

 

 

 

 

 

 

 

 

 

 

Year 3.1 Variety name 3.2 Where did you 3.3 Price 3.4 Area 3.5 Yield
(C, M , R) get the seed? (MX$/kg) (has) (Mt/ha)
l
2001 2
3
(Continues)

209

 

(Continued)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year 3.1 Variety name 3.2 Where did you 3.3 Price 3.4 Area 3.5 Yield
(C, M , R) get the seed? (MX$/kg) (has) (Mt/ha)
l
2000 2
3
1
1999 2
3
l
1998 2
3
l
1997 2
3

 

 

 

 

 

 

Kind of sced: C = certiﬁed; M = Improved (noncertiﬁed); R = Landrace

3 .6 Are you generally able to obtain a sufﬁcient quantity of the seed of the bean variety that you
wishto plant? (0=No, 1=Yes)

3.7 IF NO

What is the problem? (Specify)
3 .8 How many of the last ten years have been bad years for bean production

 

 

3.9 How many people do you occupy for bean production? Family L Hired L
3.10 What was your production cost per hectare this year [2001]?

 

IV. CREDIT
Next, I would like to ask you about your access to credit.
4.1 During year 2001, did you receive any credit to ﬁnance your bean production activities?

(O=No, 1=Yes)

210

 

4.2 IF YES

From whom did you receive credit? (Specify)

4.3 How much credit did you receive [$/ha]?

4.4 For what did you use this credit? (Specify)

V. FARMING MACHINERY

 

 

 

Next, I would like to ask you about the machinery you use in growing beans.

5.1 Do you ownatractor?

5.2 IFNO

How do you get access to a tractor? (Specify)

5.3 Do you ownapileteadora?

VI. BEAN SALES

(0=No, 1=Yes)

 

Next 1 would like to ask you about your bean sales.

6.1 How many Mt of each variety did you sell in 2000?

6.2 To whom did you sell each of these varieties?

6.3 What was the price that you received for each/variety [$lkg]?

6.4 Do you think that this is a good price for this variety?

6.5 IF NO
Why not?

(0=No, 1=Yes)

 

Variety 6.1 Sales
Name (Mt)

6.2 To whom
did you sell?

6.3 Price
(MX$/kg)

6.4 Is this a

good price?

6.5 IF NO:
Why not?

 

l

 

2

 

 

 

 

 

 

 

 

211

 

6.6 How many members of your family have an off-fann employment (including you)

 

 

 

6.7 How many months they work a year? You Wife Other: a b c
6.8 Usually what did you do with the chaff from your bean crop? (Specify)
6.9 Do you believe that some varieties produce more chaff? (0=No, 1=Yes)

6.101FYES

Which varieties produce the most chaff? (Specify)

 

 

VII. ASSESSMENT OF NEW BEAN VARIETIES

Next, I would like to ask your opinion about the bean varieties that INIFAP has released during

the past 10 years, compared to traditional variety

How does the new compare to your preferred traditional variety, in terms of:

7.1 grain yield (kg/ha)?
7.2 market acceptance?
7 .3 drought resistance?
7.4 pests resistance?
7.5 frosts evasion?

Scale: P = Worse I = Same M = Better N = No opinion/Don’t know

 

7.6 treshing problems?

7 .7 disease resistance?

7.8 type of plant?

7 .9 days to maturity?

7. 10 consumption quality?

(Name of landrace)

 

Improved
Variety

7.1
GY

7.2
MA

7.3
DR

7.4
PR

7.5
FR

7.6
TP

7.7
DR

7.9
GP

7.10
CQ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(IV = Pinto Villa, P.Mestizo, P.Bayacora, Negro Altiplano, Negro Sahuatoba, etc.)

212

 

V111. PARTICIPATION IN GOVERNMENT PROGRAMS
8.1 For how many hectares did you receive the Procarnpo subsidy [MX$829/ha]?

8.2 How many kilos of certiﬁed seed did you receive from the KxK program?

 

8.3 What variety did you get from KxK and what was the price? $/kg

 

8.4 Other government support [Pronasol?] (Specify) $/ha

IX. SOCIOECONOMIC INFORMATION

Next I would like to ask you some questions about yourself.

9.1 How old are you?

 

 

9.2 Up to what grade did you study?

9.3 How many years have you planted beans?

 

9.4 Are you a member of a farmer association? Which one?

 

9.5 IF YES: Has the association inﬂuenced your decision regarding which bean variety you
planted? (O=No, 1=Yes)

9.6 [F YES: How? (Specify)

 

9.7 Usually how many tons of beans do you store before selling?

 

9.8 Usually how many months do you store the beans for sale?

 

X. TRANSPORTATION COSTS
10.1 What’s the distance/time from your farm to the nearest city? km hrs

THANK YOU FOR YOUR ASSISTANCE!

A. Would you like to make any additional comment?

 

 

B. Additional cormnents from the enumerator

 

 

C. Could you provide us a sample of the beans that you harvested this year?

213

Appendix B: Table 1. Bean area, Production and Yield in Mexico, 1990-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Planted Area Harvest Area Production Yield

Million ha Million ha Million Mt Kg/ha
1990 2.272 2.094 1.287 615
1991 2.199 1.989 1.379 693
1992 1.861 1.296 0.719 555
1993 2.151 1.874 1.288 687
1994 2.386 2.087 1.364 654
1995 2.354 2.040 1.271 623
1996 2.196 2.048 1.349 659
1997 2.320 1.615 0.965 598
1998 2.376 2.146 1.261 587
1999 2.394 1.695 1.081 638
2000 2.114 1.787 1.159 648
Average 2.238 1.879 1.193 632

 

Source: SAGAR (2000).

214

 

Appendix C: Table 2. Apparent Consumption of Beans in Mexico, 1980-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Production Imports Exports Apparent
Mt Mt Mt Consumption“
Million Mt
1980 935,174 444,306 2,138 1.377
1981 1,331,287 482,126 2,006 1.811
1982 979,802 250,044 25,548 1.204
1983 1,285,171 145,848 55,851 1.375
1984 930,692 1 19,125 130,274 0.920
1985 911,908 178,921 122 1.091
1986 1,085,536 178,951 0 1.264
1987 1,023,734 39,470 69 1.063
1988 862,428 41,612 8,696 0.895
1989 593,436 110,370 1,178 0.703
1990 1,287,364 330,471 210 1.618
1991 1,378,519 30,080 420 1.408
1992 718,574 2,909 17,710 0.704
1993 1,287,573 7,571 7,309 1.288
1994 1,364,239 57,510 99,870 1.322
1995 1,270,915 26,062 82,872 1.214
1996 1,349,098 130,780 9,017 1.471
1997 965,056 90,161 7,256 1.048
1998 1,260,658 202,005 5,688 1.457
1999 1,080,631 128,028 8,446 1.200
2000 1,508,801 87,661 7,091 1.239

 

 

 

 

 

a. Apparent Consumption = Production plus imports minus exports.
Source: Computed with data from FAO (2002a) and SAGAR (2000).

215

 

Appendix D: Table 3. Real Producer Prices for Beans in Mexico, 1990-2000.

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Current Price“ CPI Real Price
MX$/Mt (1994:100) MX$/Mt
1990 1,988 60.12 3,307
1991 2,052 73.75 2,782
1992 2,266 85.18 2,660
1993 2,157 93.49 2,307
1994 1,901 100.00 1,901
1995 2,000 135.00 1,481
1996 3,790 181.41 2,089
1997 4,205 218.83 1,921
1998 5,154 253.68 2,032
1999 3,783 295.76 1,279
2000 4,403 323.83 1,360

 

 

 

 

 

a. Average price paid to growers in the bean regions.
Source: FAO (2002a).

216

Appendix E: ARTTP Regulations

1)
2)

3)

4)

5)

6)
7)

8)

The speciﬁc regulations of the ARTTP establish that (SAGARPA 2001a):
The Produce Foundations should make agreements with the COFUPRO and assign
for that purpose up to 3% of the resources allocated to each state.
COFUPRO, in order to assist the Produce Foundations, should design a classiﬁcation
system for the projects.
COFUPRO and the Secretariat of Agriculture should establish a unique format for
applications that should be distributed to the Produce Foundations for its utilization in
all of the country.
The resources provided by the supports of the program can not be utilized for the
acquisition of real estate properties.
In the case of projects for multiplying seed of new varieties or the utilization of that
seed for emergency programs, the approval of the Secretariat of Agriculture
(SAGAR) will be required.
The technical guides and procedure regulations issued by the Secretariat of
Agriculture (SAGAR) will be submitted to the consideration of the COFUPRO.
The Produce Foundations in each state will contract every year an external technical
audit that will be submitted to the Technical Committee of the Trust.
Produce Foundations will be granted, to for general operation expenditures, up to 3%
of those authorized for the programs of the Alliance.

For the purpose of supporting the implementation of research projects or events at

the national level, the Secretariat of Agriculture (SAGAR) could allocate budgetary
resources to coordinate actions with the “Consejo Nacional de Ciencia y Tecnologia”
(CONACYT) (National Council for Science and Technology), or with international
organizations.

217

Appendix F: Strategy of INIFAP

INIFAP’s strategy for achieving its objective includes the following principles
(SAGARPA 2002):

Development of domestic research cageLity. To achieve research of high quality,
INIFAP is committed to developing the research capacity of scientist and technicians,
and the skills of auxiliary personnel. To achieve that goal, the INIFAP has the support of
the National Council of Science and Technology (CONACYT), which provides ﬁrnding
for formal training and education of the researchers. INIFAP is also committed to
developing administrative and operative personnel, as well as the improving its physical
infrastructure.

Recoggtiorﬂnd use of farmers’ empirical knowledge. INIFAP recognizes that
farmers have developed technologies over many years that allow them to produce their
food. These technologies are considered as the base for the development of ﬁxture
technologies. Therefore, high priority is given to documenting the empirical knowledge
and practices of farmers, and collecting native genetic materials.

Rational exploitation of genetic materials. Exploitation of genetic materials of
both native and species introduced and adapted to the country is of crucial importance to
breeders. INIFAP scientists use the different sources of germoplasm to generate new
improved plant varieties and animal species to meet the demand for food with the
appropriate quality.

Realizagtion ofﬁint research. Given that there are other research institutions in the
country and a diversity of international research centers, to avoid duplication and make a
better use of scarce resources, the INIFAP seeks opportunities to implement joint
research projects supported by collaboration agreements with private and public
institutions at the national and international level.

Exchange of materiziaand informatgion. INIFAP scientists participate in research
international programs. Through those programs, the institute exchanges genetic material,
scientiﬁc information, and scientists to take advantage of the scientiﬁc advances in the
rest of the world. Agreements for scientiﬁc collaboration have been established with
organizations like IRRI in Philippines, CIAT in Colombia, CIP en Peru, ICRISAT in
India, and CIMMYT in Mexico.

Valkgttion of research results. One INIFAP’s important activities is testing and
adjusting technological information produced by the different research programs.
Through this process, technologies generated at the experiment stations are adapted to the
speciﬁc conditions of the farmers and to the particular characteristics of regions.

Diffusion of research results. INIFAP promotes the speedy and effective diffusion
of research results to: a) technicians and extensionists, b) research and academic
institutions, 0) agricultural services organizations, (1) formal farmer organizations, e)
individual farmers, and f) the general public. Diffusion is done via the Secretariat of
Agriculture and other institutions of the agricultural sector, not directly to farmers.

218

Appendix G: Requirements for Eligibility of PEAT

To qualify to receive the beneﬁts of PEAT, farmers must meet the following
criteria (SAGARPA 2002):

Support is directed to individual farmers and their economic organizations that
desire to incorporate technologies into their production systems, independent of land
tenure, and farm no more than 20 hectares under rainfed conditions or the equivalent, as
long as their plots are classiﬁed as high potential, medium potential, and low risk.

For hiring individual technicians or a technical ofﬁce, farmers should voluntarily
organize into groups for receiving technical assistance, with an area of 500 to 600 ha in
basic crops or the equivalent in other crops. The group should include 80 to 120 farmers,
according to the characteristics of the area, distribution of crops, and availability of roads.

The farmers should elect one farmer as representative of the group, and a group of
technical assistance and diffusion should be integrated with early-adopter farmers to
establish with them the demonstration plots for technology diffusion. A list with the
names and addresses of these farmers should be submitted.

Other members of the group should receive technology diffusion and technical
assistance for all the stages of the crop ﬁ'om planning to harvest and marketing.

Orientation about the programs and procedures to obtain the beneﬁts of the
programs of the Alliance for the Countryside should be provided to all group members,
so that they can reduce their costs and improve the terms of exchange.

The group’s applications will be evaluated by the “Comision de Desarrollo Rural”
(Rural Development Commission” by using the criteria for competitive funding.

More weight should be given to supporting areas with less relative development
where there are some contribution and co-responsibility of the partners, socially marginal
groups, technological development, shift into the production of more productive crops,
sustainable agriculture, diversiﬁcation of production, generation of rural employment,
and access to input and output markets.

219

Appendix H: Requirements to Apply for a Breeder’s Patent

The requirements to apply for a breeder’s patent are speciﬁed in the Vegetable
Varieties Federal Law (SAGARPA 2002):

1) Breeder’s application in official format.

2) Technical report containing the variety description in accordance with the SNICS
normative guidelines, or following the International Union for the Vegetables
Breeding Protection Procedures.

3) Rights payment receipt (for analyzing and executing the application and, if necessary,
to recognize the priority right.

4) If required, legal document to demonstrate the representative’s identity.

All documents should be submitted to SNICS’ ofﬁce. The fees to pay are detailed
in the regulations established in the Vegetable Varieties Federal Law, which are updated
and published periodically in the “Diario Oﬁcial de la Federacion” (DOF) (Federation
Ofﬁcial Diary).

220

REFERENCES

Acosta, J ., F. Ibarra, R. Rosales, A. Castillo, and P. Fernandez. 2000. "Negro Altiplano y
Negro Sahuatoba: Nuevas Variedades de Frijol para el Altiplano de Mexico".
SAGAR-INIFAP-B/C CRSP, Folleto Tecnico Num. 14, Marzo del 2000.

ADB. 1997. “Guidelines for the Economic Analysis of Projects”. Asian Development
Bank. Economics and Development Resource Center, February 1997.

Akino, M. and Y. Hayami. 1975. “Efficiency and Equity in Public Research: Rice
Breeding in Japan’s Economic Development”. Amer. J. Agr. Econ. 57 (1).

Alston, J ., G. Norton, and P. Pardey. 1998. "Science Under Scarcity: Principles and
Practice for Agricultural Research Evaluation and Priority Setting".Ithaca, NY:
Cornell University Press.

Alston, J .M., C. Chan-Kang, M.C. Marra, P.G. Pardey, and TJ. Wyatt. 1999. "A
Meta-Analysis of Rates of Return to Agricultural R&D". Washington:
International Food Policy Research Institute.

Amemiya, T. 1981. “Qualitative Response Models: A Survey”. Journal of Economic
Literature, 19:1483-1536.

ASERCA. 1997. "El Frijol en Mexico: Relato de una Diversidad". Claridades
Agropecuarias, Revista de Publicacion Mensual, SAGAR, Num. 44 (Abril).

ASERCA. 1998. "Importancia del Apoyo de PROCAMPO". Claridades Agropecuarias,
Revista de Publicacion Mensual, SAGAR, Num. 64 (Diciembre).

Besley, T., and A. Casse. 1993. “Modeling Technology Adoption in Developing
Countries”. American Economic Review 83 (2): 396-402.

Boughton, D., E. Crawford, J. Howard, J. Oehmke, J. Shaffer, and J. Staatz. 1995. "A
Strategic Approach to Agricultural Research Program Planning in Sub-Saharan
Africa". MSU International Development Working Paper No. 49. Department of
Agricultural Economics, Michigan State University, East Lansing MI.

Byerlee, D. and E. Hesse de Polanco. 1986. “Farmers’ Stepwise Adoption of
Technological Packages: Evidence from the Mexican Altiplano”. Amer. J. Agr.
Econ: 519-527.

Calva, J .L. 1992. “La Agricultura Mexicana Frente a1 Tratado Trilateral de Libre
Comercio”. CIESTAAM-UACH, Juan Pablos Editor, Mexico.

Calva, J.L. 1995. "El Modelo Neoliberal Mexicano. Costos Vulnerabilidad y
Alternativas". Mexico, Juan Pablos Editor.

221

 

Calva, J .L. 1997. "El Campo Mexicano: Ajuste Neoliberal y Alternativas". Mexico,
CIESTAAM, Universidad Autonoma Chapingo.

Castellanos, J. 1996. “Research Pinpoints Favorite Bean Flavors in Mexico: Preferential
Habits of Consumers of Common Bean (Phaseolus vulgaris) in Mexico”.
Michigan Dry Bean Digest. Spring 1996, 30 (3).

Chambers, R. 1981. "Rapid Rural Appraisal: Rationale and Repertoire". Public Admini-
stration and Development, Vol. 1, 95-106 (1981).

CIMMYT. 1993. "The Adoption of Agricultural Technology: A Guide for Survey
Design". Mexico, D.F.: CIMMYT, Economics Program.

Cochran, W.G. 1977. “Sampling Techniques”. New York USA: John Wiley & Sons.

CONAGUA.2002. “Precipitacion Media 1941-1996”. CONAGUA, Mexico.
http://smn.cna.gob.mx/productos/normales/precme.html, November 2002.

CORECA. 1999. “E1 Mercado Mundial de Frijol y sus Vinculaciones con el Mercado
Centroamericano”. Consejo Regional (10 Cooperacion Agricola, Costa Rica.

CRSP. 2002.”Obligations to CRSP Components at MSU”. Bean/Cowpea CRSP
Management Ofﬁce, Michigan State University.

Daniels, L., J. Howard, M. Maredia, J.F. Oehmke, and R. Bernsten. 1992. "Assessment of
Agricultural Research: Ex-Post, Ex-Ante, and Needed Methodologies". Staff
Paper No. 92-5 7, Department of Agricultural Economics, Michigan State
University, East Lansing, Michigan.

Dimithe, G. 1994. "Ex-post Rate of Return on Maize Research and Extension in Northern
Cameroon". MS. Thesis, Department of Agricultural Economics, Michigan State
University.

Dixon, R. 1980. “Hybrid Corn Revisited”. Econometrica 48(6, September): 1451-1461.

Echeverria, R.G. 1990. "Assessing the Impact of Agricultural Research". In: ISNAR.
1990. "Methods for Diagnosing Research System Constraints and Assessing the
Impact of Agricultural Research. Volume II: Assessing the Impact of Agricultural
Research". Proceedings of the ISNAR/Rutgers Agricultural Technology
Management Workshop, 6-8 July 1988, Rutgers University, New Jersey, USA.

Eicher, GK, and J. M. Staatz. 1998. "International Agricultural Development". Third
edition, Baltimore: The Johns Hopkins University Press.

ERS. 2002a. http://jan.mannlibcomell.edu/data-sets/specialty/8901 1/.

222

 

ERS. 2002b. http://www.ers.usdagov/data/sdp/view.asp?f=specialty/86003.

Estrada-Valle, Juan. 2001. "A Sub Sector Overview of The Guatemalan Bean Industry:
Constraints and Opportunities for Vertical Coordination and Sustainable Growth".
MS. Thesis, Department of Agricultural Economics, Michigan State University.

Ethridge, D. 1992. "Research Methodology in Applied Economics: Organizing, Planning
and Conducting Economic Research". U.S.A.

FAO. 2002a. Food and Agriculture Organization of the United Nations. FAOSTAT-
Agriculture: http://www.fao.org/waicent/portal/statistics_en.asp.

FAO. 2002b. “FaoStat Trade Notes”. www.fao.org, October 2002. Statistical Databases,
FAOSTAT-Agriculture.

 

Feder, G., R. Just, and D. Zilberman. 1985. “Adoption of Agricultural Innovations in
Developing Countries: A Survey”. Economic Development and Cultural Change
33 (1985): 255-298.

FIRA. 1994. “Elementos de Analisis de las Cadenas Productivas. Frijol”. Documento
Tecnico, Agosto 1994, FIRA-Banco de Mexico.

FIRA. 2001. “El frijol en Mexico. Competitividad y Oportunidades de Desarrollo”.
Boletin lnformativo Num. 316, Vol. XXXIII, FIRA-Banco de Mexico.

Gittinger, J .P. 1982. “Economic Analysis of Agricultural Projects”. Baltimore, MD: The
Johns Hopkins University Press.

Gonzalez R., H. 1988. “Transferencia de Tecnologia en las Unidades de Produccion
Campesinas”. Resumen publicado en: Primera reunion cientiﬁca forestal y
agropecuaria. SARH-INIFAP, octubre de 1998, Durango, Mexico.

Gonzalez R., H. 1992. “Adopcion de Technologia en el Distrito de Guadalupe Victoria,
Durango”. En: Resultados de Investigacion sobre Frijol 1990. Proyecto
colaborativo con MSU, SARH-INIFAP-Durango, Mexico.

Gonzalez R., H. 1996. “Evaluacion de la Tecnologia Agricola Modema Desarrollada por
el INIFAP para la Region de los Llanos de Durango”. Revista Ubamari, Afro XIII,
Num. 37, Instituto Tecnologico de Durango, abril de 1996.

Greene, W.H.. 1997. “Econometric Analysis”. New York: Macmillan Publishing.

Griliches, Z. 1957. “Hybrid Corn: An Exploration in the Economics of Technological
Change”. Econometrica, Volume 25, Issue 4 (Oct, 1957): 501-522.

Griliches, Z. 1980. “Hybrid Corn Revisited: A Reply”. Econometrica 48(6): 1463-1465.

223

Gujarati, D.N. 1995. “Basic Econometrics”. New York: McGraw-Hill, Inc.

Haggblade, S. and Gamser, M. 1991. "A Filed Manual for Subsector Practitioners". Tools
for microenterprise programs: nonfmancial assistance section. DAI/Gemini
Publication Series, Development Alternatives, Inc. Bethesda, MD.

Hogg, RV, and EA. Tanis. 1997. "Probability and Statistical Inference". New Jersey:
Prentice-Hall, Inc., 5th ed.

Holtzman, J .S. 1986. "Rapid Reconnaissance Guidelines for Agricultural Marketing and
Food System Research in Developing Countries". MSU International
Development Papers, Working Paper No. 30, Department of Agricultural
Economics, Michigan State University, East Lansing MI.

Howard, J .A. 1994. "The Economic Impact of Improved Maize Varieties in Zambia".
Ph.D. Dissertation, Department of Agricultural Economics, Michigan State
University.

 

INEGI. 1992-2001. Anuarios Estadisticos de Chihuahua, Durango y Zacatecas. Mexico,
Instituto Nacional de Estadistica, Geograﬁa e Informatica.

INEGI. 1996. "Anuario Estadistico. Estados Unidos Mexicanos. Edicion 1995". Mexico,
Instituto Nacional de Estadistica, Geograﬁa e Informatica.

INEGI. 2000. "Anuario Estadistico de Durango. Edicion 2000". Mexico, Instituto
Nacional de Estadistica, Geograﬁa e Informatica.

INEGI. 2001. "Anuario Estadistico. Estados Unidos Mexicanos. Edicion 2000". Mexico,
Instituto Nacional de Estadistica, Geograﬁa e Informatica.

INEGI. 2002. Instituto Nacional de Estadistica, Geograﬁa e Informatica, Mexico.
http://www.inegi.gob.mx/difusion/espanol/ﬁetab.html.

INIFAP. 2000a. “Tecnologias Para Aumentar la Productividad del Frijol en Durango”.
Mexico, Publicacion Especial Num. 12.

INIFAP. 2000b. "Aportaciones del INIFAP en Durango 1995-2000". INIFAP-SAGAR,
Informe Sexenal, Durango, Mexico.

INIF AP. 2001. “Logros de Investigacion en Mejoramiento Genetico de Frijol”. Mexico,
Desplegable Informativa No. 1, Octubre de12001.

IPED. 2002. Institute for Policy and Economic Development.
http://iped.utep.edu/datacenterecon.asp, May 2002.

224

Jarvis, LS. 1981. "Predicting the Diffusion of Improved Pasture in Uruguay". American
Journal of Agricultural Economics 63 (August 1981): 495-502.

Johnson, G. 1986. "Research Methodology for Economists: Philosophy and Practice".
New York: MacMillan Publishing Co.

Karanja, D. 1996. “An Economic and Institutional Analysis of Maize Research in
Kenya”. MSU International Development Working Paper No. 57. Department of
Agricultural Economics, Michigan State University, East Lansing MI.

Kennedy, P. 1992. “A Guide to Econometrics”. Cambridge, MA: The MIT Press.
Kish, L. 1967. “Survey Sampling”. New York USA: John Wiley & Sons, Inc.

Kosarek, J .L., P. Garcia, and ML. Morris. 2001. “Factors Explaining the Diffusion of
Hybrid Maize in Latin America and the Caribbean Region”. Agricultural
Economics 26 (2001): 267-280.

Mankiw, G. N. 1992. “Macroeconomics”. New York: Worth Publishers, Inc.

Marion, B.W. and the NC-117 Committee. 1986. “The Organization and Performance of
the US. Food System”. Lexington: Lexington Books.

Martel, RV, and RH. Bernsten. 1994. "The Bean Subsector in Honduras: Historical
Developments, Current Status and Policy Issues". Bean/Cowpea CRSP
Socioeconomics Working Paper 94-02.

Martel-Lagos, Pedro V. 1995. "A Socio-Economic Study of the Honduran Bean
Subsector: Production Characteristics, Adoption of Improved Varieties, and
Policy Implications". Ph.D. Dissertation, Department of Agricultural Economics,
Michigan State University.

Masters, W. 1996. "The Economic Impact of Agricultural Research: A Practical Guide".
Department of Agricultural Economics, Purdue University, West Lafayette, IN,
February 1996.

Mather, D.L. 1999. “Bean Supply Response in the Dominican Republic”. Term paper for
EC823: Applied Econometrics, Michigan State University, Fall 1999.

Mather, D.L. 2001. “Economic Assessment of Bean Research in the Dominican
Republic”. MS. Thesis, Department of Agricultural Economics, Michigan State
University.

Miles, T. 1994. "Agribusiness Subsector Assessments". A Guide to Developing

Agricultural Markets and Agro-enterprises. Edited by Daniel Giovannucci,
Development Alternatives, Inc.

225

 

Norton, G., and J. Davis. 1981a. "Evaluation of Agricultural Research". Proceedings of
Symposium Sponsored by NC-148, Minneapolis, May 12-13, 1980. Agricultural
Experimental Station, University of Minnesota, April, 1981.

Norton, G., and J. Davis. 1981b. "Evaluating Returns to Agricultural Research: A
Review". American Journal of Agricultural Economics 63(4): 685-699,
November, 1981.

Norton, G.W., V.G. Ganoza, and C. Pomareda. 1987. "Potential Beneﬁts of Agricultural
Research and Extension in Peru". American Journal of Agricultural Economics 69
(May): 247-257.

North, DC. 1990. “Institutions, Institutional Change and Economic Performance”. New
York: Cambridge University Press.

Oehmke, J .F., P. Anandajayasekeram, and WA. Masters. 1997. "Agricultural
Technology Development Transfer in Africa. Impacts Achieved and Lessons
Learned". Washington, DC: SD Publication Series, Ofﬁce of Sustainable
Development Bureau for Africa, Technical Paper No. 77, November 1997.

Oehmke, J .F. 2000. "Analysis of Projects with Price Effects, and Application to
Innovation and Technical Change: AEC 865 Lecture Notes". Department of
Agricultural Economics, Michigan State University, Staff Paper # 97-53.

Oehmke, J .F. and B.W. Crawford. 2002. “The Sensitivity of Returns to Research
Calculations to Supply Elasticity”. Amer. J. Agr. Econ. 84(2), May 2002.

Ortiz, V.M. 1998. "E1 Frijol en el Estado de Zacatecas". Mexico, Impresora E1 Aguila,
Zacatecas.

Pajarito R., A. and JR. Moncayo. 2000. “Transferencia de Tecnologia de Frijol en el
Estado de Durango”. INIFAP-SAGAR, Publicacion Especial Num. 15, Mexico.

Pray, CE, and CF. Neumeyer. 1990. "Problems of Omitting Private Investments in
Research when Measuring the Impact of Public Research". In: ISNAR. 1990.
"Methods for Diagnosing Research System Constraints and Assessing the Impact
of Agricultural Research. Volume II: Assessing the Impact of Agricultural
Research". Proceedings of the ISNAR/Rutgers Agricultural Technology
Management Workshop, 6-8 July 1988, Rutgers University, New Jersey, USA.

Rauniyar, GP. and FM. Goode. 1992. “Technology Adoption on Small Farms”. World
Development, 20(2): 275-282.

Rinderman, R., M.A. Gomez C., and G. Williams, eds. 1998. "TLC y Agricultura.
Funciona el Experimento?". Mexico, Juan Pablos Ed.

226

Rogers, EM. 1962. “Diffusion of Innovations”. New York: Free Press of Glencoe.

Ruiz de Londono, N, and W. Jenssen. 1990. "Un Caso de Adopcion de Tecnologia: La
Variedad de Frijol Gloriabamba en Peru". Working Paper No. 61, Cali, Colombia:
International Center for Tropical Agriculture (April).

SAGAR-Zacatecas. 1997. “Alianza para el Campo en Zacatecas: Kilo por Kilo”.
Delegacion de la SAGAR en Chihuahua, Mexico.

SAGAR-Chihuahua. 1998. “Alianza para el Campo en Chihuahua: Kilo por Kilo”.
Delegacion de la SAGAR en Zacatecas, Mexico.

SAGAR. 1998. "Situacion Actual y Perspectiva de la Produccion de Frijol en Mexico:
1990-1998". Mexico, SAGAR-CEA, Julio 1998.

SAGAR. 2000. “Situacion Actual y Perspectiva de la Produccion de Frijol en Mexico
1990-2000”. CEA-SAGAR, Mexico.

SAGAR-Durango. 2000. “Informe de Resultados 2000. Programa Kilo Por Kilo: Frijol
Semilla Certiﬁcada”. DDR 03 Guadalupe Victoria, Durango, Mexico.

SAGAR. 2001a. Informes del Programa Kilo por Kilo en Chihuahua 1996-2000.
Delegacion de la SAGAR en Chihuahua, Mexico.

SAGAR. 2001b. Informes del Programa Kilo por Kilo en Durango 1996-2000.
Delegacion de la SAGAR en Durango, Mexico.

SAGAR. 20010. Informes del Programa Kilo por Kilo en Zacatecas 1996-2000.
Delegacion de la SAGAR en Zacatecas, Mexico.

SAGARPA. 2001a. “Reglas de Operacion de la Alianza para el Campo 2001”. Secretaria
de Agricultura, Ganaderia, Desarrollo Rural, Pesca y Alimentacion, Mexico.

SAGARPA. 2001b. “Evaluacion de la Alianza para el Campo 2000 en Chihuahua: Kilo
por Kilo” e Informes de Kilo por Kilo 1996-2000. SAGARPA, Mexico.

SAGARPA. 20010. “Evaluacion de la Alianza para el Campo 2000 en Durango: Kilo por
Kilo” e Informes de Kilo por Kilo 1996-2000. SAGARPA, Mexico.

SAGARPA. 2001d. “Evaluacion de la Alianza para el Campo 2000 en Zacatecas: Kilo
por Kilo” e Informes de Kilo por Kilo 1996-2000. SAGARPA, Mexico.

SAGARPA. 2002. Secretaria de Agricultura, Ganaderia y Desarrollo Rural Pesca y
Alimentacion. Mexico, http://www.sagarpa.gob.mx.

227

Salant, P., and DA. Dillman. 1994. "How to Conduct Your Own Survey". New York:
John Wiley and Sons Inc.

Shaffer, J .D. 1973. “On the Concept of Subsector Studies”. American Journal of
Agricultural Economics, 55(2): 333-336.

Schmid, AA. 1987. “Property, Power, and Public Choice. an Inquiry into Law and
Economics”. New York, Praeger, 2“d edition.

Schmid, AA. 2000. “Institutional and Behavioral Economics”. Michigan State
University, Department of Agricultural Economics, AEC-EC-RD 810 Class
Notes.

Schwartz , L.A., J .A. Sterns, and J .F. Oehmke. 1993. "Economic Returns to Cowpea
Research, Extension and Input Distribution in Senegal". Agricultural Economics 8
(February): 161-171.

Schuh, GE, and H. Tollini. 1979. "Cost and Beneﬁts of Agricultural Research: The State
of the Arts". World Bank Staff Working Paper No. 360, October 1979, IBRD:
International Bank for Reconstruction and Development.

Staatz, J.M. 1997. "Notes on the use of Subsector Analysis as a Diagnostic Tool for
Linking Industry and Agriculture". Paper presented at the UNIDO Expert Group
Meeting, Addis Ababa, Ethiopia, September 1996.

Sterns, J .A. 1993. "Ex post Assessment of Investments in Cameroon's Cowpea and
Sorghum Research-Extension Systems". MS. Thesis, Department of Agricultural
Economics, Michigan State University.

Tellez K., L. 1994. “La Modernizacion del Sector Agropecuario y Forestal”. Fondo de
Cultura Economica, Mexico.

Tsakok, I. 1990. “Agricultural Price Policy: A Practitioner’s Guide to Partial-Equilibrium
Analysis”. New York: Cornell University Press.

Wooldridge, J.M. 2000. “Introductory Econometrics. A Modern Approach”. South
Western College Publishing, Thomson Learning.

Wooldridge, J .M. 2002. “Econometric Analysis of Cross Section and Panel Data”.The
MIT Press, Cambridge, Massachusetts.

World Bank. 2002. “World Bank Development Report 2002: Building Institutions for
Markets”.The World Bank, Washington, DC.

228