CI. ,
4! b
o

:3 t
. 31x5
.mafifi

 

 

 

 

 

J‘Jauhzin i. .iafiflfi £1...

 

it. .'.-

LIBRARY

” “I, ‘6 , Michigan State
é b 0 “U 9’2 ‘3 University

This is to certify that the
thesis entitled

Improving Maize Marketing and Trade Policies to Promote
Household Food Security in Southern Mozambique

presented by
Danilo Carimo Abdula

has been accepted towards fulfillment
of the requirements for the

MS. degree in Agricultural Economics

 

 

”//'

-4-"

//f’ ’
4 / jLic/RC AICLQ

 

‘" jor Professor's Signature

9

// N5 fi/r/vé ( 52):? 3

6‘26 /

MSU is an Affinnative Action/Equal Opportunity Institution

 

PLACE IN RETURN Box to remove this ched<out 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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 cz/CIR . 15

 

IMPROVING MAIZE MARKETING AND TRADE
POLICIES TO PROMOTE HOUSEHOLD FOOD
SECURITY IN SOUTHERN MOZAMBIQUE

By

Danilo Carimo Abdula

A THESIS
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of
MASTER OF SCIENCE
Department of Agricultural Economics

2005

ABSTRACT

IMPROVING MAIZE MARKETING AND TRADE POLICIES TO PROMOTE
HOUSEHOLD FOOD SECURITY IN SOUTHERN MOZAMBIQUE

By

Danilo Carimo Abdula

This research is aimed at helping to understand and resolve a number of perceived
problems in the maize marketing system in Mozambique, including relatively high prices
of industrial maize meal especially for lower income consumers in Southern
Mozambique, disappearance of whole maize meal in Maputo urban markets, limited
availability of maize grain in the markets, difficulties of import from South Africa, and
perhaps reduction of maize flow from the Center to the South of Mozambique.

Findings reveal that maize grain is widely produced throughout the country, but,
together with sales, it is very concentrated. Analysis of consumers’ expenditures tells that
expenditures on maize grain did not change in 2002 compared with 1.996, although there
were some differences when the analysis is done by regions. Most of the maize grain
produced by rural households in 2002 was retained at home, and informal traders are the
most important link to whom farmers sell their production. Foreign countries are
competing with the South for maize grain produced in the Central region. The analysis of
the impact of VAT on maize imports shows that it would have frequently been profitable
to bring in maize grain from South Africa if no VAT was charged. Results of the market
integration analysis show that Central-South trade in Mozambique is still somewhat

weak.

Copyright by
Danilo Carimo Abdula

2005

DEDICATION

To my dear late grandmother, F atamo Issofo.
Part of the delightful images that I still have from when I was a little kid, your home is in.
I, Macandene, the sweetest name that you used to call me, am very thankful for all your

love, and to take care of me when I was a child.

To my dear late grandmother, Georgina Taide.
Alone, you took care of your big family with sacrifices. Your spirit of sacrifice and
persistence is an inspiration for me. I, Macabega, the sweetest name that you used to call

me, am very thankful for all your love.

To my dear late grandfather Juma Abdula.
You thought me that anything is reachable with hard work, dedication, integrity and
honesty. Thank you so much to teach me what a Muslim religion really is. With your

lessons, you became one of the most influential people in my life.
To my parents, Omar J uma Abdula and Rossana Abdul Karimo.

Thank you for everything. Thank you for investing unconditionally on my education.

Part of what I am today, I definitely owe it to you. May Allah bless you all.

iv

ACKNOWLEDGMENTS

I would like to acknowledge the support of several people and institutions that
contributed to the completion of my Master Program at the Michigan State University.

I am especially thankful to my committee members. Professor David Tschirley,
my major professor and at the same time thesis advisor, played an important role on the
choosing of the topic. His guidance throughout the whole program and especially during
the thesis writing was extremely important. His time and effort devoted to revise my
innumerous drafts was really appreciated. Professor Mike Weber also played an
important role during the writing process. Through his excellent teaching skills,
meticulous observations and thoughts, he helped me to understand the importance of
maize grain for rural households in Mozambique. I am also very thankful to Professor
Rick Ward to accept to be part of my committee members. To all of you, I express my
gratitude f0r all critically constructive comments and suggestions. 0

From the Department of Agricultural Economics at Michigan State University,
especial thanks goes to Dr. Kelly Raper, Dr. Cynthia Donovan, and Debbie Conway. Dr.
Kelly gave me useful comments regarding the time series analysis, and Dr. Cynthia gave
me useful comments and suggestion regarding the impact of VAT on maize imports. Mrs.
Debbie Conway was always available to help me whenever I had any administrative
problem.

Support for this research has been provided by the USAID/Mozambique Mission,

by the Food Security III Cooperative Agreement (CDG-A-00-000021-00) between

Michigan State University and the United States Agency for International Development,
through the Bureau for Economic Growth, Agriculture and Trade's Office of Agriculture
and Food Security (EGAT/AGR) with supplemental funding from the Africa Bureau's
Office of Sustainable Development and by the Rockefeller Foundation, through its
funding to MSU for "Toward Improved Maize Marketing and Trade Policies to Promote
Household Food Security in Southern Africa".

Many thanks are due to several people that made this thesis possible. Mr. Camilo
Camal gave me an important explanation regarding the functioning of VAT, and how this
tariff is charged. Mr. Vitor Matavele shared with me his knowledge regarding some costs
involved on maize imports from South Africa to Mozambique. Mr. Belemo Lalgy, Mr.
Andre Vonk, Mr. Veeresh Hiremath, and Mr. Stephan Scheafer, shared with me part of
their knowledge regarding the operation of the milling industry in Mozambique. Ms
Ellen Payongayong gave me her advice whenever I had problems either with the
statistical software used for data analysis or with TIA and CAP data. To all these people,
my deeply gratitude and recognition. Without all of you, this work would not be possible.

Third, I am grateful to my colleagues from Market Information System (SIMA) in
Mozambique that helped me during my fieldwork. Mr. Pedro Arlindo was with me
during part of the visits made to maize processors and during the visit of several
companies dealing with maize grain in South Africa. Ms Anabela Mabota was with me
during the field work throughout the Central Region of Mozambique. Mr. Francisco
Monteiro, Mr. Abel Frechaut, Mr. Simao Nhane, and Mr. Inocéncio Banze were with me

during the interviews with maize grain traders. From the bottom of my heart, thank you

vi

all.

Fourth, many thanks are due to special people in East Lansing with whom I had
the privilege to spend amazing moments during the whole program. To Mr. Rui Benfica
and his family, Mrs. Marcia Benfica and little Ruy Allan Benfica, to Mr. Feliciano
Mazuze and Mr. Raul Pitoro, my compatriots and colleagues during the program, I say
thanks for everything. I will not forget those days that we spent together at East Lansing.

Many other people made my presence in East Lansing so special that I cannot
forget to mention. Elliot Mgheniye, my roommate during the first two years of the
program is one of them. I am quite sure that I will not have another roommate as you.
Special thanks for support and friendship also go to my friends at Cook Hall (thanks god,
they are so many!), and to all members of the soccer team Los Intocables.

Finally, I am grateful for the support of my family. My greatest debt is to my wife
Nilza. Because of the distance, we went through difficult times, but thanks to God, we
overcome everything. Thank you for being so patience and to tolerate what seems

intolerable. Thank you for your love and endurance.

vii

TABLE OF CONTENTS

LIST OF TABLES ..................................................... xiv
LISTOFFIGURES ................ xviii
CHAPTER ONE: INTRODUCTION AND RESEARCH CONTEXT ......... 1
1.1 Overview of Problem and Objectives ............................ 1
1.2 Specific Problem Statement .................................... 2
1.3 Specific Research Questions ................................... 6
1.4 Overview of the Methodology and Data .......................... 8
1.5 Research Instruments ......................................... 9
1.6 Thesis Organization .......................................... 9

CHAPTER TWO: MAIZE GRAIN IN MOZAMBIQUE HOUSEHOLDS’

2.1

2.2

2.3

2.4

2.5

ECONOMY .............................. . .......... 11
Introduction ............................................... 11
Structure of Maize Production ................................. 12
National Production Trends ................................... 13

Smallholder’ Maize Production and Sales by Geographical Region . . . . 17

Maize Grain in Smallholder Livelihoods ......................... 19
2.5.1 Northern Region ...................................... 20
2.5.2 Central Region ....................................... 25
2.5.3 Southern Region ...................................... 30

viii

2.6 Maize Expenditure Patterns ................................... 34
2.6.1 Population of Mozambique ............................. 34
2.6.2 Consumer Expenditure Patterns: IAF Results ............... 35
2.6.3 IAF Expenditure Disaggregation: Rural Areas .............. 37
2.6.4 IAF Expenditure Disaggregation: Urban Areas .............. 40
2.6.5 IAF Expenditure Disaggregation: Capital Cities ............. 41

2.6.6 Proportion of Total Food Expenditure Allocated to Food Item
Categories by Tercile of Total Expenditure ................. 43
2.7 Chapter Summary .......................................... 46

CHAPTER THREE: OVERVIEW OF THE MAIZE SUBSECTOR IN

MOZAMBIQUE .............................. 49
3. 1 Introduction ............................................... 49
3.2 Industrial Organization ...................................... 50
3.2.1 Structure-Conduct-Performance Paradigm ....... A .......... 50
3.3 Subsector Approach ......................................... 52
3.3.1 Vertical Coordination .................................. 52
3.3.2 Marketing Channels ................................... 53
3.3.3 Market Participants ................................... 54
3.4 Marketing at producer Level .................................. 55
3.4.1 Informal Traders ..................................... 58
3.4.2 Neighbors ........................................... 60
3.4.3 Informal Rural Retailers ................................ 60

ix

3.5

3.6

3.7

3.4.4 Rural Shops and Formal Traders ......................... 61

Marketing at Wholesale Level ................................. 61
3.5.1 Profile of Informal Assembler/Wholesaler Traders ........... 62
3.5.2 MaizeFlows....................... .................. 63
3.5.3 Volume of Bags Marketed per Week ...................... 73
3.5.4 Suppliers and Demanders of Maize Grain .................. 74
3.5.5 Quality of Maize Grain ................................ 77
3.5.6 Transportation ....................................... 78
3.5.7 Credit .............................................. 79
3.5.8 Access to Market Information ........................... 79
Marketing at the Milling Industry Level ......................... 80
3.6.1 Attributes of Maize Meal Consumed in Mozambique ......... 80
3.6.2 Small Maize Millers ................................... 83
3.6.3 Industrial Maize Millers ................................ 85
Chapter Summary .......................................... 87

CHAPTER FOUR: THE IMPACT OF VAT ON MAIZE GRAIN IMPORTS . . 90

4.1

4.2

4.3

4.4

4.5

Introduction ............................................... 90
The Code of Value-Added Tax (VAT) .......................... 91
VAT and Price of Maize Meal in Maputo ........................ 94
Assessing the Impact of VAT Charged on Imported Maize Grain ..... 98
Regulation to Import Maize Grain in Mozambique ................ 104

4.6

4.7

4.8

The South African Maize Marketing System ..................... 106
Incentives to the Milling Industry ............................. 108

Chapter Summary ......................................... 108

CHAPTER FIVE: ANALYSIS OF VERTICAL AND SPATIAL PRICE

5.1

5.2

5.3

5.4

5.5

5.6

5.7

RELATIONSHIP AMONG GRAIN MARKETS IN

MOZAMBIQUE ................................... 110
Introduction .............................................. 1 10
Market Performance ........................................ 1 11
Indicators of Market Performance ............................. 112
5.3.1 Market integration ................................... 112

5.3.1.1 Importance of Market Integration ................. 113
5.3.2 Marketing Margins ................................... 115
Analytical Framework for Analysis of Market Performance ......... 116
5.4.1 Marketing Margins .......................... i ......... 116
5.4.2 Unit Root and Cointegration Analysis .................... 1 17
5.4.3 Granger Causality Test ................................ 120
5.4.4 Dynamic Adjustments ................................ 123
Markets, Products and Price Data ............................. 124
Results of Descriptive Statistics ............................... 125
5.6.1 Price Pattern at Retail Level ............................ 125
5.6.2 Price Pattern and Wholesale Level ...................... 131
Results of Marketing Margin Analysis ......................... 132

xi

5.8 Econometric Results for Market Integration Analysis .............. 137

5.8.1 Correlation Coefficients ............................... 137

5.8.2 Unit Root Test ...................................... 138

5.8.3 GrangerCausalityTest............... ................. 139

5.8.4 Spatial Market Integration ............................. 140

5.8.5 Vertical Market Integration ............................ 142

5.8.6 Dynamic Adjustments ................................ 143

5.9 Chapter Summary ......................................... 146
CHAPTER SIX: CONCLUSIONS AND POLICY IMPLICATIONS ....... 148
6. 1 Introduction .............................................. 148
6.2 Major Findings ............................................ 149
6.2.1 Maize Production and Marketing Pattern ................. 149

6.2.2 Maize Expenditures .................................. 150

6.2.3 Maize Subsector in Southern and Central Mozambique ...... 153

6.2.4 Impact of VAT on Maize Imports ....................... 155

6.2.5 Performance of Maize Subsector ........................ 156

6.3 Policy Implications ........................................ 157

APPENDIX 1: Rural Household Asset Indicators, and Other Selected Characteristics
by Category of Maize Production and Sales ..................... 162

APPENDIX 2: Questionnaire for Maize Processors ........................... 166

xii

APPENDIX 3: Structure of Maize Milling Industry in Mozambique .............. 179

REFERENCES ....................................................... 194

xiii

Table 1.2

Table 2.2a

Table 2.2b

Table 2.4

Table 2.5.1a

Table 2.5. lb

Table 2.5.2a

Table 2.5.2b

Table 2.5.3a

Table 2.5.3b

Table 2.6.1

Table 2.6.2a

LIST OF TABLES

Maize Grain and Maize Meal Prices, January l996-August 1998 ....... 3
Number of Farms Producing Cereals and Cassava ................. 12

Small Farms Producing Cereals and Cassava, and Patterns of Land
Distribution (Percentage of National Total) ....................... 15

Maize Production and Sales Indicators by Region of Country ........ 17

Characteristics of Production and Marketing by Category of Maize
Production and Sales Behavior Among Households of Mozambique:
Northern Region ............................................ 22

Rural Household Income, Asset Indicators, and Selected Demographic
Characteristics by Category of Maize Production and Sales: Northern
Region ................................................... 24

Characteristics of Production and Marketing by Category of Maize
Production and Sales Behavior Among Households of Mozambique:
Central Region ............................................. 27

Rural Household Income, Asset Indicators, and Selected Demographic
Characteristics by Category of Maize Production and Sales: Central
Region ................................................... 29

Characteristics of Production and Marketing by Category of Maize
Production and Sales Behavior Among Households of Mozambique:
Southern Region ............................................ 31

Rural Household Income, Asset Indicators, and Selected Demographic
Characteristics by Category of Maize Production and Sales: Central

Region ................................................... 33
Population of Mozambique per Province ........................ 34
Percentage of Total Expenditure Allocated to Food Items in 1996 ..... 35

xiv

Table 2.6.2b

Table 2.6.2c

Table 2.6.3

' Table 2.6.4

Table 2.6.5

Table 2.6.6a

Table 2.6.6b

Table 3.5.1

Table 3.5.2a

Table 3.5.2b

Table 3.5.2c

Table 3.5.2d

Table 3.5.2c

Table 3.5.3

Table 3.5.4a

Percentage of Total Expenditure Allocated to Cereals, Tubers, and Roots
in Rural and Urban Areas as Well as in All Country in 1996 ......... 36

Percentage of Total Expenditure Allocated to Cereals, 1996 and 2002 . 37

Percentage of Total Food Expenditure Allocated to Food Items;
Rural Areas ............................................... 39

Percentage of Total Food Expenditure Allocated to Food Items;
Urban Areas ............................................... 42

Percentage of Total Food Expenditure Allocated to Cereals; Provincial
Capital Cities .............................................. 43

Percent of Total Food Expenditure Allocated to Food Item Categories by
Tercile of Total Expenditure and by Rural Regions of Mozambique . . . 44

Percent of Total Food Expenditure Allocated to Food Item Categories by
Tercile of Total Expenditure and by Urban Regions of Mozambique . . . 46

Description of Activities Performed by Informal Traders ............ 64

Source of Maize Grain in Informal Wholesale Markets in Maputo,
July 2004 ................................................. 68

Source of Maize Grain in Informal Retail Markets in Maputo,
July 2004 ................................................. 68

Source of Maize Grain Sold in Maputo Informal Wholesale Market per
Marketing Year, 1998/99-2004/2005 ............................ 70

Source of Maize Grain Sold in Maputo per Marketing Month; Marketing
Years 2002/03-2003/04 ...................................... 72

Source of Maize Grain Sold in Xai Xai Informal Wholesale Market per
Marketing Year, 1998/99-2004/05 .............................. 73

Volumes Marketed by Informal Assemblers/Wholesalers per Week . . . 74

Most Important Supplier of Maize for Informal Traders Selling in
Wholesale Markets .......................................... 75

XV

Table 3.5.4b

Table 3.5.4c

Table 3.5.4d

Table 3.5.5
Table 3.6.]
Table 3.6.2
Table 3.6.3

Table 4.4a

Table 5.6.1

Table 5.6.2

Table 5.7a
Table 5.7b

Table 5.7c

Table 5.8.1a

Table 5.8.1b

Table 5.8.2

Table 5.8.3

Table 5.8.4

Table 5.8.5

Most Important Supplier of Maize for Informal Traders Selling in Retail

Markets .................................................. 75
Most Important Demander for Maize Among Informal Traders in
Wholesale Markets .......................................... 76
Most Important Demander for Maize Among Informal Traders in Retail
Markets .................................................. 77

Problems Faced by Informal Maize Traders During Buying and Selling 78

Attributes of Maize Meals Consumed in Mozambique .............. 81
Percentage of Villages Having Maize and Cassava Mills ............ 85
National Market Share of Maize Millers Interviewed in the Study ..... 87

IPP in Maputo for Maize Grain Imported from Six Different Location of
South Africa by Rail, June 20041 ............................. 100

Descriptive Statistics for Maize Grain at Retail Level (Real Prices) . . . 130

Table 5.6.2 Descriptive Statistics for Maize Grain at Wholesale Level

(Real Prices) .............................................. 131
Real Monthly Average Price of Maize Grain and Price Spreads ...... 132
Stability Test of the Retail Margin Over Time ................... 133
Net Margins for Informal Wholesaler in the Maputo Market ........ 134
Correlation Coefficients Among Maize Retail Markets ............ 137
Road Distances Between Maize Markets (Kilometers) ............. 138
Results of Unit Root Test for Maize Grain at Retail Level .......... 139
Results of Granger Causality Test for Pairs of Retail Markets ....... 140
Results of Cointegration Test for Maize Grain at Retail Level ....... 141
Results of Cointegration Test for Maize Grain at Vertical Levels . . . . 142

xvi

Table 5.8.6a Validation of the Autoregressive Model in Maize Grain Markets ..... 144

Table 5.8.6b Dynamic Price Adjustment in Maize Grain Markets ............... 145

xvii

LIST OF FIGURES

Figure 1.3 Real Retail Prices for Maize Grain, Industrial Maize Meal, Wheat and

Rice; Maputo, January 1993-December 2004 . . . .................... 4
Figure 2.2 Map of Mozambique ........................................ 14
Figure 2.3 Production and Trend for Maize, Cereals, and Equivalent Dried Cassava
for 1992/93 up to 2003/04 .................................... 16
Figure 3.4a Distribution Channel for Maize Grain in the Southern Region of
Mozambique .............................................. 56
Figure 3.4b Distribution Channel for Maize Grain in the Central Region of
Mozambique .............................................. 57
Figure 3.5.2a Flow of Maize Grain from the Central Region to Maputo Informal
Wholesale Market .......................................... 66
Figure 3.5.2b Flows of Maize from the Central Region to the Southern Region ...... 67

Figure 4.3a Real Prices of Maize Grain, Refined Maize Meal, and Rice; Maputo . . . 95

Figure 4.3b Price of Maize Grain Relative to Maize Meal ...................... 96
Figure 4.3c Price of Maize Meal Relative to Rice ........................... 97
Figure 4.4a Scheme for Import Parity Price Calculation ...................... 99

Figure 4.4b Price of Maize Grain from Central Region Versus IPP; With VAT . . . 102
Figure 4.4c Price of Maize Grain from Central region Versus IPP; No VAT ..... 103
Figure 5.6.1a Real Price of Maize Grain at Wholesale and Retail Level; Maputo . . . 126
Figure 5.6.1b Real Price of Maize Grain at Wholesale and Retail Level; Xai Xai . . . 127
Figure 5.6.1c Real Price of Maize Grain at Wholesale and Retail Level; Beira ...... 128

Figure 5.6.1d Real Price of Maize Grain at Wholesale and Retail Level; Chimoio . . . 129

xviii

Figure 5.7. Percentage of Each Marketing Cost in the Total Cost, Excluding Purchase
Cost ....................................................... 136

xix

CHAPTER ONE

INTRODUCTION AND RESEARCH CONTEXT

“An important element in the development of a more eflicient marketing system is the removal of

rigidities and distortions resulting from government policy itself... ” (Ruttan, V. and Hayami, Y.)

1.1 Overview of Problem and Objectives

The broad objective of this research is to gain a better understanding of the maize
marketing system, especially in Southern Mozambique, to contribute to more effective
food security policies and market growth for the benefit of farmers and consumers. This
research is aimed at helping to understand and resolve a number of perceived problems in
the industry, including relatively high prices of industrial maize meal especially for lower
income consumers in Southern Mozambique, disappearance of whole maize meal in
Maputo urban markets, limited availability of maize grain in the markets, difficulties of
import from South Africa, and perhaps reduction of maize flow from the Center to the

South of Mozambique.

1.2 Specific Problem Statement
Mozambique is situated on the south eastern coast of Africa, with an area of
799,380 square kilometers, 2,515 kilometers of coastline running North-South, and 10

provinces divided into three regions, namely Northern, Central, and Southern regions.

Maize grain, which along with cassava is the key domestically produced staple, is
produced primarily in the Northern and Central regions of the country. The South
contains the main urban population center of the country (Maputo and Matola cities) but
is a marginal agricultural production area. Most farmers in this region are net maize
buyers.

In such a long country, and in one with poor road, rail, and sea links, the cost of
transporting maize from where it can be competitively grown to the South is relatively
high.

Due to these conditions, there are regular imports of maize grain to Southern
Mozambique from South Africa to supply established commercial maize mills. For other
users, some maize grain also flows from Central regions to the South to supply informal
wholesale markets that serve rural and urban consumers.

The poverty rate is very high throughout the country, including in Maputo and
Matola cities, with a population around 2 millions of habitants. More than any other city
in the country, the population of these two cities relies on markets to purchase their basic
needs in order to satisfy their food security. Therefore, the cost of purchased food in the
urban and rural areas has a strong impact on the real income of poor households.

Historically, maize grain and rice have been the basic staples in the Southern
region, consumed by both rural and urban households. Table 1 below shows that,
between January 1996 and August 1998, the city of Maputo had the second highest
wholesale price for maize grain among five countries of Southern and Eastern Africa.

Furthermore, the city of Maputo registered the second highest price for industrialized

roller maize roller meal.

Table 1.2 Maize Grain and Maize Meal Prices, Janeiro 1996 4 August 1998

 

Prices (SUS/T on) Kenya Zambia Zimbabwe RSA Mozambique
Producer— Grain 190 133 109 113 101
Wholesale at Capital-Grain 241 174 120 133 217
Roller Meal-Industrial Mill 390 285 172 443 424
Whole Meal-Hammer Mill 272 204 124 - 254

 

During the first three months of 2004, the price of industrial maize meal registered
a mean of $USS43 per metric ton in Maputo city, which may be the highest in all
Southern and Eastern Africa. Moreover, while the prices of industrial maize meal have
been increasing in Southern Mozambique, whole maize meal sold at retail, which is much
cheaper, has disappeared from some of the urban markets. Furthermore, there is some
indication that the availability of maize grain to be purchased by urban as well as rural
consumers have decreased.

In addition, analysis of the real price pattern of rice (Figure 1.3) shows that the
price has decreased throughout the Southern region since the second half of the 19908.
Furthermore, the price of rice is lower than the price of industrial maize meal, and the

premium paid for rice over maize grain in Maputo has decreased.

30> can 5:22

 

 

 

 

 

m. ow N N 0 o. .e .o b m.
chose @stch @99ch com octagov doe oeMowv «cox o949®©q®¢®49® see
9 A Q 6 Q Q 9 Q 9 Q 6 Q 9
«see,» P$Qe§$ee twee/v9 ayes/ea evasive >5
00mm D D D D P D D D P o
in: 0.25» is.
I I a \ \( a . I \a
2 x s a s a 7 e I :o ..
=32 woe—mom 1 .. ..1: 1... .e , Xe. .. .e .. u ...7 x . . u .. ‘11.. ooom
l .... . . .. ./ .... ... 2.. u .... . mm
580 0582 .:. . u- ni" . . ”If; . x , ...— n W
. . ...! .6 R . )7 9.3., ...m .(t ‘1‘ . w . d
- i u i ) , ... . . u. 4
1‘3? : a...“ 88. m
an.“ ... \II
. . M
. .. M
. new
-
_ Doom?
800m

mgu-ma3. baa—3.. 6.3—«E ”35— can “no; .302 via: Evans—E— .520 93a: .3.— moctm Ency— Eoy— MA 0.53...—

The decreasing price of rice has been, without any doubt, an important contributor I
to maintaining households’ real purchasing power in the South. However, the absence of
whole maize meal sold at retail from urban markets in the South, together with the
dramatic increase in prices for industrialized maize meal and limited availability of maize
grain (supported by evidences from Agricultural Market Information System, SIMA), has
resulted in the concern that the cost of the basic food is very high for low income
consumers.

While industrial maize meal has registered high prices and the availability of
maize grain has decreased, there are no records of any imports of maize grain from South
Africa, the only reliable year-round source of maize in the region, to be sold in the
markets as grain to fill the gap. Indeed, it seems that import of maize grain are only done
by the industrial milling industry as a way to supply their mills.

These facts show important potential implications for consumers. When the
availability of local maize grain is reduced, consumers in both rural and urban areas of the
South of Mozambique have to purchase maize flour produced by large industrial mills.
Since the cost to produce maize flour is high in large industrial mills in comparison with
the cost through hand pounding or in small-scale mills, they explicitly have to pay high
prices for maize flour produced industrially. Rural consumers in the South still use a
hand method for processing maize meal, assuming they can source maize grain. Under
these conditions, buy potential or real scarcity of maize grain may threaten the food
security of the poorer consumers.

This lack of imports of maize grain to be sold in local markets as grain has

persisted throughout the years, although there is some evidence that in some periods over
the last six years, such imports might be profitable.

These tendencies - high prices of industrial maize meal especially for lower
income consumers in Southern Mozambique, the disappearance of whole grain maize
meal in Maputo urban markets, the seemingly limited availability of maize grain in the
markets, the difficulties of maize grain imports from South Africa, and perhaps the
reduction of maize flow from the Center to the South of Mozambique — raise important
policy questions. Therefore, in order to address these questions, research on the maize
marketing system is needed to generate the information required for policy and program

decisions to safeguard households’ food security.

1.3 Specific Research Questions
The general purpose of this research is to gain a better understanding of the maize
marketing system in Mozambique in order to contribute to more effective policies to
induce the growth of food security of the poor, as well as the growth of markets in
general. .
The specific research questions of the study are:
1. What are the maize production, consumption and marketing patterns among the
rural population of Mozambique?
2. Are consumers in the Southern region of Mozambique, especially poor urban
consumers and net buying farmer households in rural areas, utilizing less maize

grain at relatively higher prices than in the past? If yes, what is the cause?

How is the maize marketing system structured in Mozambique, especially in the
Central and Southern Region? Who are the major participants in the maize
marketing system in Mozambique, and what are their roles?

Would it be competitive to import more maize grain from South Africa for use in
a broader array of maize meals and for sale as grain in markets in urban and rural
areas of Southern Mozambique, and if so, why is there not more maize grain
imported from South Africa for these purposes?

What are the costs of trade between Central and Southern Mozambique, and how
well integrated over space are maize markets in these two regions?

What might be the component that helps explain what appears to be happening?
a) The value-added Tax (VAT) that is unrecoverable by some traders on imported
maize grain from South Africa?

b) The transport costs for maize grain from South Africa and Central
Mozambique?

c) Competing uses (markets in Zimbabwe and Malawi) for maize grain produced
in Central and Northern Mozambique?

d) Lack of competition among maize meal manufactures in Southern
Mozambique?

e) Cheap rice imports?

f) Others?

1.4 Overview of the Methodology and Data

To describe the maize subsector in Mozambique, the subsector framework was
used. The use of this framework helps in the identification of constraints in the marketing
system, and then in the identification of critical points of intervention necessary to
alleviate those constraints.

Some criticisms have been pointed out about the complexity of the subsector
approach. Indeed, Scott (1995) pointed out that studies using this approach require
enormous quantities of data that took time and had high costs.

To avoid this problem and to get answers to the specific questions pointed above,
the study was first delimited in geographical regions, being the Central and Southern
Regions, the main focus of the study.

Second, secondary data sources were reviewed. This included historical
production data such as area and maize production, prices, household level data, etc.

Third, as information regarding the maize traders operation as well as millers
operation was not available, the decision was made to get this information through the
use of rapid appraisal (RA). Thus, interviews with key participants directly or indirectly
involved in maize subsector were conducted to generate valuable insights related to their
operation. These key participants were selected from among formal and informal
wholesalers, retailers, maize processors, and private companies. Furthermore, direct
observations in some stages of the whole maize chain were made.

The geographic area of the interviews included three provinces, namely Maputo

and Gaza in the Southern region, and Manica in the Central region. More explicitly, the

interviews with those key participants were conducted in the provincial capital cities
because most of the maize trade is done there (most of the informal wholesale markets
are located there). Furthermore, the main processors, importers,and the majority of

private companies dealing with maize trade are located in the capital cities.

1.5 Research Instruments

Two interview guides were designed, being one for informal maize traders and
another one to managers of maize mills. The first interview guide was used to interview
29 informal maize traders. Twenty-two of them were interviewed at informal wholesale
markets located in Maputo, and the remaining seven at the Xai Xai informal wholesale
market.

The second interview guide was used to interview the four principal medium/large
maize millers in the study region. The idea was to include all principal maize millers in
the sample since they are not so many, but was not possible to do it; two of the main

millers were not included in the study.

1.6 Thesis Organization

The present thesis will comprise six chapters. The introductory chapter presents
an introduction to the topic, as well the problem statement and the objectives. Chapter
two will present the role of maize grain in Mozambican rural households’ livelihoods,
describing in more details the structure of maize production, marketing and consumption

throughout the country, and the role of maize grain in smallholder livelihoods. In this

chapter, the first, and second specific research questions will be answered. Chapter three
will present the structure of maize marketing system in Mozambique based on the
subsector approach and insights from taking a marketing channel approach and
examining related questions of industrial organization. In this chapter, the third specific
research question will be answered.

In chapter four, the role of duties and taxes on maize marketing will be presented.
This will help address the fourth specific research question. In chapter five, to assess the
performance of the maize markets, an analysis of the market integration among vertical
marketing stages and across spatially differentiated markets will be conducted, as well as
an analysis of marketing margins. The fifth specific question will be answered.

Finally, in the chapter six, the main findings of the study will be summarized, and

some recommendations for policies will be presented.

10

CHAPTER TWO

MAIZE GRAIN IN MOZAMBIQUE HOUSEHOLDS’ ECONOMY

2.1 Introduction

This chapter presents the role of maize grain in Mozambique’s smallholder
household economy, describing in more details the structure of maize production
throughout the country, and the role of maize grain in smallholder livelihoods.

The next two sections of this chapter summarize maize production trends from a
national perspective, using data from the last Agricultural and Livestock Census (CAP,
1999-2000), and from the National Early Warning System (SNAP), which has
responsibility for generating official agricultural production data. In the fourth section,
data from the 2002 National Agricultural Household Survey (TIA 2002) is used to present
informatiOn on smallholder’s maize production and sales patterns by geographical region.
The fifth section discusses the role of maize grain in smallholder livelihoods. Section six
presents information on maize consumption patterns in the country. One main source of
data was used to compute consumption results, namely data from the First and Second
National Assessment of Poverty (IAF, 1996 and 1997). Finally, the last section presents a

chapter summary.

11

2.2 Structure of Maize Production

Maize production in Mozambique is affected by several conditions that determine
the structure of marketing in the country and thus the range of issues this process faces.

Technology — specifically the widespread absence of modern technology - is a
key determinant of the structure of production and sales. The structure of maize
production in Mozambique is dominated by smallholder farms where the majority of
them use rudimentary technologies. Data from CAP show that nearly 2.4 million
smallholder households and only 8,900 medium sized farm households grew maize grain
at the time of this survey. Thus, over 99 percent of the maize producers in Mozambique
are smallholder farms. Table 2.2a below shows the number of farms producing maize,

other cereals, and cassava by farm size.

Table 2.2a Number of Farms ProducingCereals and Cassava

 

 

 

Commodity Number of Farms Producing Cereals and Cassava
Farm Size Category
Small1 Medium Large Total

Maize 2,398,169 8,933 206 2,407,308
Rice 632,305 906 44 633,255
Sorghum 820,803 2,565 51 823,419
Millet 140,999 1,478 29 142,506
Cassava 1,935,154 2,984 57 1,938,195

 

Data Source: CAP

 

1

For discussion on how the farm size was classified, see Apresentacao Suma’ria dos Resultados do Censo
AgroPecudrioI999-2000. Instituto Nacional de Estatistica, Mocambique, 2002.

12

Another factor that affects maize production is agro-ecological conditions. The
majority of maize producers are concentrated in the Northern provinces
(Figure 2.2), namely Niassa, Cabo Delgado, and Nampula, that have a better rainfall
distribution as well as better soil fertility. Together, these three provinces account for
about 40 percent of area planted to maize and nearly half of national maize production.
Table 2.2b shows the percentage distribution, by province, of the number of small

farms as well as the total area allocated to maize grain, other cereals, and cassava.

2.3 National Production Trends

After the peace agreement of October 1992, maize production increased
dramatically, especially in the first three years. This growth was not due primarily to
increased productivity. Rather, it was a result of the peaceful conditions that allowed
access to more land by the population that was displaced and unable to practice
agriculture during the war period. Also, exports to Malawi provided a good market for
maize grain produced in the Northern Region, and therefore, contributed to the growth of
maize production. Between 1996 and 1999, production rose approximately 10 percent a
year, a pattern that no other food crop has shown (Figure 2.3).

The variability of maize output is due essentially to weather factors. For instance,
in the 1999-2000 crop season, maize production declined 18 percent, due primarily to
floods that devastated large areas of the Center and South of the country. Furthermore,
droughts as the one observed in 1995 and 2002, are another issues that influence the

production.

13

Figure 2.2

Map of Mozambique

 

L \\ 42
. TANZAN
\
I Law
' INF/(15a
ZAMBIA 8
-1: m
t) ”m”
l l .Lichingl
MALAWI?"

lbw

Tote \‘

(3"; ..
my),

\.
\Qu ll
\, e m

Chimoio

ZIMBABWE /

c;
“.1

Naivémia

“‘N'

lnhambane
/

“from

\

swaz
SOUTH g/
AFRICA /

 

/
\\

/ o 100
I-—*—r—‘——\
_. 0

IA 3‘;

Pemba

Cicada

de Naomij
Nampula. j

/

an ,/

Y

INDIAN
OCEAN

zoo Im

 

me 200 '"' 3:

 

.20 ”ouSom Sea

 

.5— 2: 2: 2: 2: 2: 2: 2: 2: 2: Each.
0; . 06 _.o od fin 0N _.o _.o od 0N 8:932
pd Ta 5; ad Nag mg. QN 06 #2 mm g0
_.N~ 0d oé o.m 0.x v.0 Nam QM 0N 0.x o§0§€~
_.~ VAN 03 5.0— 5.x 0d 0.2 5.: w.: 0.0 fiflom
NA 5: _.m_ w.m Num— hN 0.: 0.2 0.m fix 3:82
_.~ w.wm N6 06 0.0— NN v. _N 0.0 0.0 0.0— 89—:
0&— nd 0.3 hfim m4; 0. K v.5 0.2 v. 3 ma" «Eons
mHm 0.0 “mom 5.2 Wm can Nam _ v.0m YNN Kuw— 2:952
0.2 5.0 0.“: o.m_ Nd QS 5o Q: 0.2 ~.: ovum—ca .U
0.m ON 5.0 , 5N 5w new 0.0 Wm wd m8 «332

 

139—. 12.332 .3 cunt—998.—
35.5.5

 

«>330 .052 Esawuem 35— 3?: «>330 .052 Esawuom 33 “in:

 

3 03332 ...—«A .53 mafia..— uo eon-.52
dutch. ...—.232 he 2.33% act—523.5 155 no 3.38am can £2330 can n—uouoUfinmosgm ...—Eur.— =aEm nufi 033—.

 

15

Figure 2.3 Production and Trend for Maize, Other Cereals, and Dried Cassava for the Last Twelve Marketing Seasons

 

Cassava

ID.

 

 

 

 

 

WWW

 

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

 

 

.

 

 

 

 

 

WWW
"—-

l

W

 

 

:\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

 

 

—
s\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ \\\\\\
_

 

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

—_

WW

h
\\

 

\\\\\\\\\\\\\\\\\\\\\\\\

 

 

 

 

 

2.50

8 $2 8
N . .

—1 fl

0 50
0.00 ‘

(mo; 30 uomrw) uopanpord

16

2002- 2GB-
2003

2000- 2001-

2001

1999-
2000

1998-

1

1993- 1994- 1995- 1996-
1995 1997

1992-

2002

1996 1998 1999
Marketing Season

1994

1993

Data Source: SNAP

2.4 Smallholder Maize Production and Sales by Geographical Region

Among cereals, maize grain in Mozambique is the basic staple food for most of
the poor. Indeed, data from TIA surveys showed that, in average, 79 percent of all
households were engaged in its production in 1996, and almost the same percentage (80
percent) in 2002.

Table 2.4 presents the percentage of households producing and selling maize
during the 2002 harvest year, by geographical region, as well the average quantity of

maize produced and sold.

Table 2.4 Maize Production and Sales Indicators by Region of Country (2002

 

 

 

Harvest Year)
Region
North Center South Total

% of National Production 50.5 39.8 9.6 ‘ 100
% of National Sales 59.1 38.5 2.4 100
Household Level by Region
% Planting Maize 73.5 93.7 89.8 -
% Producing Maize 72.5 90 77.5 -
Mean Quantity Produced Among 406 749 250 -
Those Producing (Mean Kg)
% Selling Maize 24 23.4 3.8 -
Mean Quantity Sold Among 154 299 137 -
Those Selling (Mean Kg)

 

 

Data Source: TIA 2002

17

Despite the fact that nearly all households in the Center and Southern region try to
grow maize, the percentage of households that actually realized maize production in 2002
was lower, due to pests and other unfavorable weather conditions, especially in the
Southern region. As such, the percentage of households with maize production as
compared to the number planting decreased in all regions, with more emphasis in the
Southern region where from 90 percent of households planting maize grain, only 78
percent in fact had some production.

The fact that about 25 percent of households did not try to produce maize grain in
the Northern region may be explained by several factors. First, households in this region
have more favorable climatic conditions and as such, they can diversify their production.
Thereby, they take advantage by allocating their available land to alternative crops, with
especial emphasis to cassava that occupies a significant portion of land. Second, the level
of cash crOp production other than maize2 is much higher in the Northern region, so that
available family labor has to be split between food and cash crops. Therefore, since the
monetary value earned is much higher with other cash crops, there is less labor available
to enable the household to be engaged in maize production.

Among those households producing maize, the Central region, which
encompasses the provinces of Sofala, Manica, and Tete, stands out from the other two
regions. In this region, the maize quantity produced per household is on average much
higher than in the other two regions. This pattern has existed for several years, and may

be explained by several factors such as the amount of land allocated for maize grain as

 

2 Maize is a cash crop for some households.

18

shown on Table 2.2b.

In terms of maize sales, the proportion of households selling maize is much higher
in the Northern and Central Region (24 percent) than in the South, even though on
average the households in the Northern region sell less compared to households in the
Central region, that is, 154 Kg against 300 Kg on average. The major reason for Northern
and Central regions selling more maize compared to the Southern Region is probably due
to the agro-ecological conditions that allow greater production per household. Therefore,
households that do sell in the Northern and Central regions have larger production bases
from which to sell. This has also driven exports to neighboring countries, and investment
in maize processing mills.

In summary, the Central Region has the highest percentage of households
producing maize, and the highest mean household production and sales among those
producing. However, the Northern region, due to its higher population, predominates in
total national production and sales. The Southern region has the lowest mean production
among those producing, the lowest mean sales among those selling, as well as the lowest
percentage of households selling. Total production and sales in this region are also the

lowest.

2.5 Maize Grain in Smallholder Livelihoods
For purposes of the analysis, in this section, we divide all households into six
mutually-exclusive categories with respect to maize production and sales, and then we

examine selected characteristics of each of these household groups.

19

The categories are:

1. Households with no maize grain production.
2. Households with maize grain production but no sales.
3. Four groups of sellers, arrayed from those selling the least (quartile one) to those

selling the most (quartile 4). The regional importance of maize production and
sales by each of these groups is computed, and shown in the rows “% of maize
produced in the region”and “% of maize sold in the region”. The households in
the lowest production and sales quartile produced and sold the least amount of
maize grain, while those in the highest quartile produced and sold the largest

amount.

2.5.1 Northern Region

In the Northern region, the results of the household-level distribution of maize
production and sales show that 28 percent of rural households do not produce maize
(Table 2.5.la). In addition, 48 percent do produce, but do not sell.

Among households that both produced and sold maize in 2002 (about 25 percent
of the population), production and, especially, sales are highly concentrated. The top sales
group (highest quartile), representing only 6 percent of the rural households, accounted
for 24 percent of all production and 71 percent of all sales in the Northern region. They
produced on average over one ton of maize grain and sold nearly half a ton.

Among those households without any production of maize grain (28 percent),

nearly three-quarters purchased some quantity of either maize grain or maize meal. This

20

share drops to around one-half for those who produce but sell nothing. This means that at
least 48 percent (0.275*0.726 + 0.477*0.585) of all rural households in the North were
net buyers of maize in 2002.

For the largest sales group, the percentage of households that bought some
quantity of maize grain or maize meal falls to about one-third. The key finding here is
that purchases of maize grain or maize meal are widespread in rural areas, are not
confined to one single type of household, and are related to the amount of maize
produced as well as sold. The majority of households do not produce enough maize grain
to eat, something that may be explained by low maize area cultivated and the level of
technology expressed by low yields (Table 2.5.1a).

Rural households’ income share information indicates the contribution of different
types of economic activities to total household income (Table 2.5.1b). Income includes
the calculated monetary value gained from consuming own production and selling
livestock, cereals and other commodities, as well as income generated from wage
earnings and own business activities. Overall, mean households income ranged from
nearly five million meticais (SUSZSO) for households with no production to nearly ten
million meticais (SUSSOO) for those households in the highest sales quartile.

Among those households without maize production, the main source of income
was commodity production for own consumption plus sales other than cereals,
contributing with 59 percent of the total income. This pattern was observed in all other
production and sale categories except for the category of highest maize sales where cereal

sales accounted for the highest proportion of total rural household income (42 percent).

21

Table 2.5.la Characteristics of Production and Marketing by Category of Maize
Production and Sales Behavior Among Rural Households of
Mozambique: Northern Region

Households Households
With No With

Household With Both Production
and Sales by Quartile

Characteristics
of Maize

 

Production

Households

Production

But No
Sales

1. Lowest
Quartile

2.

3.

4. Highest
Quartile

 

% of HH in
Region

% of Maize
Produced in
Region

% Maize Sold in
Region

Mean Quantity
Produced per HH
in Kg (Median)

Quantity Sold &
Intended Sales
per HH by Those
Selling (Kg)

Average Percent
of Production
Sold (Median)

% of HH Buying
Maize or Maize
Meal (%)

Total Area
Cultivated per
HH (Ha)

HH Area per
Capita
(Ha/person)

Area in Maize
per HH (Ha)

Maize Yield

27.5

(0.0)

(0.0)

72.6

1.01

0.37

0.02

47.7

54.2

334
(193)

(0.0)

58.5

1.16

0.34

0.36

1,439

6.3

3.3

233
(116)

23

20.8

(15.0)

49

1.02

0.32

0.30

1,080

6.8

8.3

329
(227)

59

30.5

(25.8)

42.3

1.18

0.35

0.37

1,583

6.2

10.6

17.2

503
(350)

118

36.5

(30.0)

50

1.35

0.37

0.50

1,410

6.3

23.5

71.2

1 , 102
(706)

483

54.1

(50.0)

33.7

2.00

0.71

0.69

2,212

see/Ha)
Data Source: TIA

Notes:

HH stands for households

22

The key point here is that the value of commodities consumed and sold are the .
crucial contributors to overall household income, with shares ranging from 71 percent in
the first household production and sale category (households without any maize
production) to 83 percent in the fifth household production and sale category (households
in third quartile of sales).

Households’ assets are also examined in Table 2.5. lb in each household
production and sale category. The most important pattern that emerges is that the
households in the largest quartile of production and sales are better off than other
households. Indeed, the percentage of households having a bicycle and radio is higher in
this household category. Furthermore, the average household income is also significantly
higher in this category, and drawing on results from Table 2.5.1a, households in this
group have total area cultivated roughly twice as large as a majority of other smallholder
in the region, and they likewise have about twice the amount of land per person in their
households, as compared to most others in the no production or production with no sales
category.

Also in Table 2.5.1b, selected demographic characteristics of rural households are
examined. The results show that at least nearly three quarters of households in each
household category are headed by males, with households in the largest quartile of sales
appearing less likely to be headed by females. Furthermore, the level of education is
higher in this household category. With the lowest percentage of households whose

heads are illiterate (29 percent).

23

Table 2.5.lb Rural Household Income, Asset Indicators, and Selected Demographic
Characteristics by Category of Maize Production and Sales: North

 

 

Characteristics Households Households Household With Both Production
of Maize With No With and Sales by Quartile
Households Production Production ‘
But No 1. Lowest 2. 3. 4. Highest
Sales Quartile Quartile
Average HI-I 4,671 5,200 3,356 4,870 5,751 10,305
Income (000 Mt)
Income per 1,166 1,204 914 1,187 1,230 2,392
Capita (000 Mt)

 

 

Source of Income - Percent of Income Coming from this Source

Value Cereals "‘ 12.1 29.3 29.2 28.9 37.8 42.4
Value Other 59.0 47.4 52.2 53.2 44.8 34.6
Crops “

Salary 6.6 6.6 7.2 3.6 4.6 2.3

Own Business 16.3 10.9 8.2 10.5 10.1 15.6
Activities

Other" 6.0 5.8 3.2 3.8 2.7 5.1

 

 

Assets Owned - Percent of Household with this Asset

Bicycle 13.5 28.6 23.6 25 48.5 53.5
Radio 35.2 48.9 44 49.4 55.9 62.6
Animal Traction 0 0 0 O 0 0.1
Cattle 1-9 Heads 0 0 0 0 1.1 0.1
Goat 1-19 Heads 15.3 16.8 18.6 17.5 20.4 23.1

Household Characteristics — Percent of Households with this Characteristic-----------

l-II-I Headed by 26.4 22.6 26.5 15.6 16.6 10.4
Female

Head Illiterate 48.0 34.8 38.4 43.5 38.7 28.9
Age ofHH Head 39 39 37 35 38 37
(Median)

Household Size - Average Number of Persons in Each Household

 

HH Size 4.2 4.8 4.5 4.5 5.0 5.1

Data Source: TIA
" Includes value of both production that is consumed on farm and value of sales
** Include value of livestock production consumed and sold, remittances, pensions, and land rent

24

In addition, 50 percent of the households in this category have more than 3 years
of schooling. Recall that this household category has the highest percentage of maize

production sold (54 percent), and also the highest concentrationof maize sales (71).

2.5.2 Central Region

In the Central Region, the results of the household-level distribution of maize
production and sales show that 10 percent of rural households do not produce maize
(Table 2.5.2a). An additional 66 percent do produce, but do not sell any maize.

The concentration of production and sales in the Center is remarkable similar to
that in the North. Actual levels of production and sales, however, are much higher in the
Center.

Among households that both produce and sell maize (about 24 percent of the
population), production and especially sales are highly concentrated. The top sales group
(highest quartile), representing only 6 percent of the rural households, accounts for 22
percent of all production and 72 percent of all sales in the Central Region. They produce
on average at the household-level over two tons of maize and sell nearly one ton. Both
figures are about double those in the North.

Among households without maize production (10 percent), results in Table 2.5.2a
show their strong use of the market for purchasing. Indeed, the results in the column
“Percentage of households buying maize or maize meal” indicates that 9.2 percent of
households without production are net buyers (60 percent including those with production

and no sales) as in the Northern region. As we move into the household production and

25

sale categories, the results indicate that in the first group of households, over 80 percent
of the households bought some maize, while in the fourth category (highest producers and
sellers) some 43 percent bought maize. Therefore, as in the Northern region, purchases of
maize grain or maize meal are widespread in rural areas, but even more so in the category
of households without maize production.

Area in maize per household is significantly higher in the Central Region
(compared to the other regions) with the exception of the highest production and sales
quartile in the Southern region. Both maize area cultivated and yields are relatively high
for those households with the most interaction on the sales side of the market
(Table 2.5.2a). This group has total area cultivated almost twice as large as a majority of
other smallholder in the region. They also have about 53 percent more land per person in
their households, as compared to the large category of households with production but no
sales.

In Table 2.5.2b we present income and asset information for the Central region by
the same six mutually exclusive household groupings. Among those households without
maize production, the main source of income was formal salary, contributing with nearly
30 percent of the total income.

In all other household categories, contrary to what was observed in the Northern
Region, in the Central Region cereals production accounts for the highest share of total
income, and represents almost 50 percent of the total income in households with maize
sales. This difference may be explained by the higher average quantities of maize

produced, consumed, and sold per households in the Central Region.

26

Table 2.5.2a Characteristics of Production and Marketing by Category of Maize
Production and Sales Behavior Among Rural Households of
Mozambique: Central Re ion

Characteristics Households Households Household With Both Production
of Maize With No With and Sales by Quartile

 

Households Production Production
But No 1. Lowest 2. 3. 4. Highest

Sales Quartile Quartile

 

% ofHH in 10 65.7 6.1 5.8 6.1 6.3
Region

% of Maize 0 53.2 6 7.1 1 1.4 22.4
Produced in
Region

% of Maize Sold 0 0 2.8 8 17.5 71.6
in Region

Mean Quantity 0 545 662 827 1,251 2,397
Produced per HH (0.0) (346) (500) (639) (850) (1,751)
in Kg (Median)

Quantity Sold & 0 0 44 131 269 1071
Intended Sales

per HH by Those

Selling (Kg)

Average Percent 0 0 10.7 25.9 36.3 52.2
of Production (0.0) (0.0) (7.5) (20.0) (26.7) (50.0)
Sold (Median)

% of HH Buying 92.4 77.6 80.2 68.3 68.9 43.4
Maize or Maize
Meal

Total Area 1.54 1.73 1.73 1.79 2.02 2.95
Cultivated per
HH (Ha)

HH Area per 0.39 0.45 0.46 0.42 0.51 0.69
Capita
(Ha/person)

Area in Maize 0.25 0.89 1.06 1.01 1.20 1.64
per HH (Ha)

Maize Yield 0 900 1,067 1,355 1,177 2,565
reg/Ha)
Data Source: TIA
Notes:
HH stands for households

27

The key point is the same as in the case of the Northern Region, that is,
commodities produced for own consumption and sales, are the crucial contributors to
overall household income. The difference is that in the Northern Region the major share
comes from commodities other than cereal, while in the Central Region this is the
reverse.

Households’ assets are also examined in Table 2.5.2b in each household
production and sale category. The most important pattern that emerges is that the
households in the largest quartile of production and sales are better off that other
households, as in the Northern region. The percentage of households having bicycles,
radios, animal traction, heads of cattle, and the highest average household income is
higher in this category. The only difference noticed between both regions is that the
percentage of households having animal traction, and between one and nine heads of
cattle, is much higher in the Central than in the Northern region.

Also in Table 2.5.2b, selected demographic characteristics of rural households in
the region are examined. The results show that households in highest two quartile of
production and sales appear less likely to be headed by females. Furthermore, the level of
education is higher in these last two quartile of production and sales. In fact, this
category has the lowest percentage of households with heads that are illiterate (less than

30 percent), and nearly 50 percent of the households have more than 3 years of schooling.

28

Table 2.5.2b Rural Household Income, Asset Indicators, and Selected Demographic
Characteristics by Caggory of Maize Production and Sales: Center

 

 

Characteristics Households Households Household With Both Production
of Maize With No With and Sales by Quartile
Households Production Production '
But No 1. Lowest 2. 3. 4. Highest
Sales Quartile Quartile
Average HH 6,443 7,937 4,812 7,149 9,250 13,452
Income (000 Mt)
Income per 1,332 1,613 1,108 1,367 1,882 2,904
Capita (000 Mt)

 

Source of Income - Percent of Income Coming from this Source

 

 

 

 

Value Cereals "' 22.9 29.8 45 49.9 48 55.7
Value Other 19.4 29.1 27.4 24.1 28.4 22
Crops ‘
Salary 29.8 13.6 6 11.6 7.4 3.4
Own Business 7.7 17.9 14 7.9 11 9.3
Activities
Other" 20.2 9.6 7.6 6.5 5.2 9.6
Assets Owned - Percent of Household with this Asset
Bicycle 15.1 25.2 17.8 18.3 38.1 48.1
Radio 53.7 55 38.8 54.8 62.5 70.1
Animal Traction 23.3 17 10.2 14.3 27.4 33.6
Cattle 1-9 Heads 12 5.8 5.9 6.2 8.4 14.2
Goat 1-19 Heads 39.5 44.4 42.1 41.8 53.3 63.2
Household Characteristics - Percent of Households with this Characteristic
HH Headed by 28.5 25.3 31.1 29.5 8.2 9.6
Female
Head Illiterate 40.6 41.3 47.4 41.3 26.4 30.1
Age of HH Head 38 42 38 38 37 41
(Median)
Household Size - Average Number of Persons in Each Household
HH Size 5.0 5.6 5.1 5.3 5.4 5.7

 

Data Source: TIA
* Includes value of both production that is consumed on farm and value of sales

** Include value of livestock production consumed and sold, remittances, pensions, and land rent

29

2.5.3 Southern Region

In the Southern region, the results of the distribution of maize production and
sales show that 23 percent of rural households did not produce maize in 2002
(Table 2.5.3a). In addition, 73 percent did produce, but did not manage to sell any maize.
These households accounted for slightly more than three-fourths of all maize production
in the region, so in the South as compared to the Center and North, a much higher
percentage of maize producing households retained for own consumption what they can
manage to produce.

Among households that sell maize (about 4 percent of the population), production
and especially sales are also highly concentrated. The top production and sales group
(highest quartile), representing only 1 percent of the rural households in the region,
accounts for 14 percent of all production and 77 percent of all sales. They produce on
average over two tons of maize and sell nearly half a ton.

As in the other two regions, households with no maize production are also
significant net buyers, with 82 percent of the households buying some maize grain or
meals. This means that at least 18.5 percent (0.225*0.822) of rural households with no
maize production were net buyers of maize in 2002, and that percentage increases to 76
percent when households with production and no sales were included
(0.225*0.822 + 0.731*0.79). This pattern is also observable in all other household
categories, especially among the large majority of households producing but not selling,

where average production per household was only about 200 Kg.

30

Table 2.5.3a Characteristics of Production and Marketing by Category of Maize
Production and Sales Behavior Among Rural Households of
Mozambique: Southern Region

Households Households
With No With

Characteristics
of Maize

Household With Both Production
and Sales by Quartile

 

Production Production
But No 1. Lowest 2. 3.

Sales Quartile

Households
4. Highest

Quartile

 

% of HH in
Region

% of Maize
Produced in
Region

% of Maize Sold
in Region

Mean Quantity
Produced per HH
in Kg (Median)

Quantity Sold &
Intended Sales
per HH by Those
Selling (Kg)

Average Percent
of Production
Sold (Median)

% of HH Buying
Maize or Maize
Meal (%)

Total Area
Cultivated per
HH (Ha)

HH Area per
Capita
(Ha/person)

Area in Maize
per HH (Ha)

Maize Yield

22.5

«10)

«10)

82.2

1.02

0.49

0.17

73.1

76.7

203
(37)

«10)

79

1.38

0.45

0.64

685

1.1

2.5

2.5

438
(104)

15

12.2

(11.5)

94.1

1.29

0.60

0.80

1,344

0.8

2.1

5.2

512
(289)

44

17.2

(15.5)

63.2

1.25

0.58

0.68

1,824

1.4

4.9

15.3

670
(482)

74

16.5

(13.5)

84.1

2.02

0.67

0.72

2,290

1.1

13.7

77

:1459
(L386)

488

33.5

(1650

57.2

2.64

1.00

1.86

3,095

_<K.g/Ha>

Data Source: TIA

Notes:

HH stands for households

31

Area in maize cultivated per household is significantly higher in the highest
production and sales quartile in the Southern region, being the highest among all regions.
This group has total area cultivated almost twice as large as a majority of other
smallholder in the region. They also have about 122 percent more land per person in their
households, as compared to the large category of households with production but no
sales.

Among those households without maize production, the main source of income
was coming from commodities produced for consumption plus sales for other than
cereals, contributing with 56 percent of the total rural households income (Table 2.5.3b).
This pattern was also observed in the first household maize production and sale
categories, while in the second, third, and fourth category, cereals consumed and sales
accounted for the highest share of total rural household income.

The percentage of households having bikes is not widespread in this region,
except in the last quartile of production and sales where 41 percent of households have
bicycles. However, the percentage of households having radio is comparable with figures
for the other two regions.

Another aspect that distinguishes this region from the Central and the Northern
Region is the percentage of households having animal traction, which is much higher in
the former region. The Southern Region has a long tradition of cattle raising compared
with other regions, a factor explained by its marginal agro-ecological conditions for

agricultural practices.

32

Table 2.5.3b Rural Household Income, Asset Indicators, and Selected Demographic
Characteristics by Category of Maize Production and Sales: South

 

 

Characteristics Households Households Household With Both Production
of Maize With No With and Sales by Quartile
Households Production Production
But No 1. Lowest 2. 3. 4. Highest
Sales Quartile Quartile
Average HH 11,537 11,528 7,755 6,636 16,166 38,824
Income (000 Mt)
Income per 3,171 2,443 1,593 1,166 3,370 4,454
Capita (000 Mt)

 

Source of Income - Percent of Income Coming from this Source

 

Value Cereals * 0.0 12.7 22.3 51.7 31.9 28.3
Value Other 55.9 44.2 39 30.6 31.7 19.7
Crops *

Salary 18.1 14.9 9.1 13.5 21.5 17.5
Own Business 17.4 15.2 18.9 2.5 3.6 14.8
Activities

Other" 9.4 13 10.7 1.7 11.3 19.7

Assets Owned - Percent of Household with this Asset

 

 

Bicycle 3.4 7.4 1.9 22.9 7.1 41

Radio 50.3 54.2 28.8 49.4 50.7 75.1
Animal Traction 38 42.4 34.1 41.2 45.7 22.4
Cattle 1-9 Heads 8.3 10.1 3.2 19.1 14 34.5
Goat 1-19 Heads 27.3 36.3 42.4 36.8 33.7 43.4

Household Characteristics - Percent of Households with this Characteristic

HH Headed by 36.1 29.5 46.3 43.4 31.5 33
Female

Head Illiterate 49.8 40.1 ' 39.9 42.7 49.9 25.6
Age of HH Head 50 48 46 36 52 42
(Median)

 

Household Size - Average Number of Persons in Each Household

HH Size 4.9 5.6 4.4 4.4 5.7 8.3

Data Source: TIA
" Includes value of both production that is consumed on farm and value of sales
** Include value of livestock production consumed and sold, remittances, pensions, and land rent

33

2.6 Maize Consumption Patterns

2.6.1 Population of Mozambique

According to the National Population Census (1997), projections for 2005

indicate that the total population of Mozambique is nearly 19.5 million, with 71 percent

living in rural area. In the Southern region, nearly 2.5 million live in rural areas, almost

13 percent of the total population. Table 2.6.1 presents the total population of

Mozambique per province, as well as population living in rural and urban areas.

 

 

 

 

 

 

Table 2.6.1 Population of Mozambique per Province
Region Province Population Urban Rural Urban Total
(Habitants) % %
Niassa 999,332 230,005 769,327 23.0 5.1
€ Cabo 1,617,165 272,457 1,344,708 16.8 8.3
2 Delgado
Nampula 3,676,003 919,001 2,757,002 25.0 18.9
Zambezia 3,710,01 1 500,313 3,209,698 13.5 19.1
E Tete 1,511,832 223,144 1,288,688 14.8 7.8
g Manica 1,320,232 371,400 948,832 28.1 6.8
Sofala 1,637,821 675,966 961,855 41.3 8.4
Inhambane 1,381,023 270,548 1,1 10,475 19.6 7.1
5 Gaza 1,304,798 321,899 982,899 24.7 6.7
g Maputo 1,044,946 656,008 388,938 62.8 5.4
Maputo 1,216,873 1,216,873 0 100.0 6.3
City
Total 19,420,036 5,657,614 13,762,422 - 100

 

Data Source: IN E 2005
Notes: The rural and urban breakdown was obtained by using as a base the 1997 population census for
urban and rural population, and then applying this to the 2005 estimates

34

2.6.2 Consumer Expenditure Patterns: IAF Results

In 1996, a collaborative research effort between Ministry of Planning and Finance
(MPF) of Mozambique and International Food Policy Research Institute (IFPRI)
produced the first national assessment report on poverty and well being in Mozambique.
This assessment was based on household survey data collected in 1996. Data from that
household expenditure survey (IAF, 1996) suggest that cereals, tubers, and roots
accounted for about one-half of the household food expenditure and one-third of total
household expenditure. The definition of expenditures includes an estimate of the value
of own produced commodities that were consumed, plus the value of purchases in the
market.

Results in Table 2.6.2a reveal that the proportion of total household expenditure
allocated to cereals was 23 percent nationally, with almost no difference between rural

and urban. Furthermore, expenditure on roots and tubers was much higher in rural areas.

Table 2.6.2a Percentage of Total Expenditure Allocated to Food Items in 1996

 

 

Rural Urban National
Commodity
Percent
All foods 67.6 51.6 61.9
Cereals 22.8 22.7 22.8
Roots and tubers 11.0 4.1 9.6
Beans 5.8 2.9 5.2

 

Data Source: IAF 1996

35

Among cereals, results in Table 2.6.2b show that both rural and urban households
in 1996 spent more on maize and maize derivatives than in any other cereal or root and
tuber. Furthermore, in rural areas cassava was the second most consumed staple, while in

urban areas rice and wheat products stood in second and third place, respectively.

Table 2.6.2b Percentage of Total Expenditure Allocated to Cereals, Tubers,
and Roots in Rural and Urban Areas as Well as in All Country in

 

 

 

1996
Rural Urban National
Commodity
Percent

Wheat and derivatives 0.74 5.28 1.66
Maize and derivatives 16.68 10.56 15 .44
Rice 3.09 6.77 3.84
Sorghum and Millet 2.31 0.08 1.85
Cassava 9.54 2.93 ' 8.19
Other roots and tubers 0.13 0.62 0.23

 

Data Source: [AF 1996

In 2002, the Ministry of Planning and Finance (MPF) of Mozambique produced
the second national assessment report on poverty and well being in Mozambique. Data

from this last expenditure survey are compared with the 1996 assessment, and the results

are reported in Table 2.6.20.

36

Table 2.6.2c Percentage of Total Expenditure Allocated to Cereals; National

 

 

Cereal IAF 1996 IAF 2002
Percent
Rice 3.84 i 3.76
Maize and Derivatives 15.44 15.45
Wheat and Derivatives 1.66 1.31

 

Data Source: IAF 1996 and 2002
Notes: 2002 results computed by the authors

As shown in Table 2.6.2c, there is basically no change in national average
percentage of total households’ expenditures allocated to cereals. In 1996, households
spent on maize as much as in 2002, registering a small decrease on total expenditures for

wheat and rice.

2.6.3 IAF Expenditure Disaggregation: Rural Areas

The national averages tend to hide important regional and other differences.
Results in Table 2.6.3 show that, in 2002, households in the rural Province of Maputo
spent more in rice (nearly double of 1996), followed by expenditures on maize that
clearly decreased from 16 to 9 percent. Expenditures on wheat increased slightly
compared with 1996.

In Gaza and Inhambane, rural households spent more on maize and its derivatives
than on other staples during both years, but the share allocated to maize fell sharply in
2002; households in these provinces spent more on rice and wheat together in 2002 than

they did on maize and derivatives. While expenditures on maize decreased relative to

37

1996, they nearly doubled for rice. In the case of wheat, the expenditure did not change
significantly.

In summary, it seems that throughout rural South, expenditures on maize and its
derivatives decreased at the expenses of rice and wheat among cereals. However, rural
households’ expenditure increased dramatically for fish, leaves, and fruit throughout the
South.

In rural patterns of Central Region, in 2002, households were still spending more
on maize and its derivatives, and even more than in 1996. The expenditures on rice
increased slightly in Manica and Tete, remaining almost the same in Sofala and
Zambezia. However, expenditures on wheat increased slightly in all rural Central as in
rural South, particularly in Maputo.

In Sofala and Zambezia, the expenditure on cassava increased significantly, from
14 to 18 percent, being the second most important commodity in terms of households’
expenditure. Also, households’ expenditure for sorghum and millet increased from 3 to 5
percent, a pattern that is not observable in other provinces in the Central and in all
Southern provinces.

In summary, in the rural Central, among cereals, households were spending more
on maize and its derivatives, followed by rice in Manica and Tete. In Sofala and

Zambezia, expenditures on cassava were occupying the second place.

38

890.3 05 3 03.688 $.32 83 ”882
50.5 use .95.. «o 53:03 :2: 8 3:288 .38 v.3 03— ...—5 603cm Sea

 

 

 

 

 

 

 

39

 

 

 

 

ca 8.: 3 8.2 2: S: a: can 4.: ..2 8.: 4.9.. :8... sec
3 mm a... S 3 3 8.88m;
2 8.: 2: a: 3. a.“ 2 .2 2 3 mm 3 88m
3 3 n: 2 3 ..n «.... 2 a: as S. 2 as;
3. to S. a... E 3. 3 w: «.2 as S: 4.2 8.53
3 3. 2 ..8 a.” _.N_ S S "a Z 2. E ..E
2 an 3” 3. an as am a... 2 2 ....m ....N 58m
2. E ...." 2 2 Z 3 2. 3 2 S 3 an:
3 2. 3 3 So 3 no 2 ..c 2. 3 3 .8888 as:
ea:
5. 3 n... 3 ...... 3 2 as 8... a... 3 o... 2a 5.5.3
2: 4.2 3». .2 a: an 2 2 3 n: S. 3 «>880
.... .3 3. 2 S 2 .... 3 3 an E G a»;
S S 2 we 3 3 3 2 ms 2 ...: S 8e.
an». 3N as 2: EN 2: 3: SN «.2 EN 3 4.2 8a:

2.3.5.—
82 ea. 88 82 ‘ N8" 2...: N8" :2 N8" 32 «Sn X:
8.35 age-N 2.35....— is. :8...
3.0 E: 332 2.35.2 can «Eom 89,—. v... «3.32 E... 50 8.3a!—
_a.:.~— 388— vcoh 8 083274 chameaoflnm cook 33,—. no uneasy—om n0." 03am.

2.6.4 IAF Expenditure Disaggregation: Urban Areas

In Maputo Province, urban households, as in 1996, were still spending the most
on wheat and its derivatives, although less than in 1996 (Table 2.6.4). Expenditures on
rice were still in second place as in 1996, despite the fact that expenditures on it have also
decreased from 1996 to nearly half in 2002. A partial explanation for the small rise in the
maize expenditure share and the fall in rice, is that the real price of maize meal rose
sharply between the two years, while the rice price fell substantially. Apparently,
expenditures on cereals from urban households in Maputo decreased between 1996 and
2002, and from Table 2.6.4, it is clear that expenditures on fish, leaves, and fruit went up
significantly to compensate for the decrease in cereals. The rise of households’ incomes
may have driven the households’ expenditures to these food categories, but this issue
needs further research.

In Gaza and Inhambane, urban households’ expenditures on maize increased
significantly compared to 1996, and households in these provinces were spending more
on it, followed by rice and wheat. In these two provinces, expenditures on cassava
increased slightly, and as in Maputo province, expenditures on leaves and fruit increased
significantly.

In the Central Region, in 2002, urban households spent even more on maize and
its derivatives (the budget shares for maize rose from 23 to 48 percent in Manica and
Tete, and from 19 to 27 percent in Sofala and Zambezia) at expenses of wheat and its
derivatives, whose expenditures on it decreased half in all region, and also rice, which

decreased slightly in Manica and Tete. As in the South, part of the explanation for the

40

increased maize share is that real prices of meal and also grain were substantially higher
in 2002 than in 1996.

Expenditures on sorghum and millet increase significantly in all urban Central
Region, from O to 2 and 5 percent in Manica and Tete, and Sofala and Zambezia,

respectively.

2.6.5 IAF Expenditure Disaggregation: Provincial Capital Cities

Looking only at 2002 households’ food expenditures for just the Cities of Maputo
and Xai Xai in the South, and Beira in the Center, results in Table 2.6.5 show that
expenditures on rice were most important in Xai Xai and Beira, while in Maputo were on
wheat. The weight of rice in households’ expenditures, increase as we move from
Maputo to Beira. Clearly, in all Southern Cities (Maputo and Xai Xai), expenditures on
maize and its derivatives were not that important compared with Beira City in the Central

Region.

41

£053 05 .3 @3388 3.32 8cm “882
535 v5 =22 we 35.53 :05 8 wEEOoou .83 28 ea. "—5 ”330m San

 

42

 

 

 

 

 

 

 

 

 

 

 

 

 

ca ”.mm a: ..NN a6. a.aN m.m_ a. _ m 9: mam 0.2 a.mm nee..— 35¢
9m m._ Wm fin ms ad mo_aSowo>
as N: n.m ad a.m «4. a6 .2. En NM a.~ Nu 230m
fin a. _ flm M: Wm mé md N; v.3 fin e...— NN mam
3. m.m "an mé fin m4 :6 _.v :6— od fin— Em 828‘—
ma 3: «an 5.: 2: 9: Na v.2 ed ha «.3 a6 5E
e." A... A: N. in a. _ m... Wm 9m 2 a.m w. m FEE
a; 9m Nd Nd _.N ad 9m m6 w; a; fin Em “no:
2. x... #6 to Nd ma Nd m._ _.o ed ad M: 8850a 55
3:22

n." _.o m._ m6 Qm ad a; ad ad ad ad od c5 5.5me
m... m..— e.an a. 2 3m _._ ad ad fin a.N m._ 9m «>830
m.~ aé v6 m.n «4. m.» a.~ a.n ed 03 On— agm 80:3
Qm 9m a6 m.n N.a a.» vé fie «a NB 3. Q2 38
adv hon v; _ o. a mfin va— a.am odm m4; 2: v." E 832

. 2.3.3?
«can oaa— «can oaa— nee" oaa— «can oaa— «can caa— «can aa—
3-209 sauna-N 2.2.52...—
BaU E:— nuufiz ...—aauz E... ...—com 39,—. can 3.5.: E:— anaU Sana—2 use: ace..—
53.5 388— eeeh 3 633e=< 23:30:3— ueeh .308 no cyanogen— véfi «Bah.

Table 2.6.5 Percentage of Total Food Expenditure Allocated to Cereals in 2002;

 

 

 

 

Capital Cities
Product Maputo City Xai Xai City Beira City
1996 2002 '
Percent
Rice 18.1 7.1 13.7 15.9
Maize and 5.2 1.6 4.7 10.9
Derivatives
Wheat and 15.0 17.5 8.8 9.3
Derivatives
Other Food Items 61.7 73.8 72.8 63.9

 

Data Source: IAF 1996 and 2002, according to their definition of rural and urban
Notes:

2002 results computed by the authors

1996 data for Xai Xai and Beira cities were not available

2.6.6 Proportion of Total Food Expenditure Allocated to Food Item
Categories by Tercile of Total Expenditure; Rural and Urban Areas

Table 2.6.6a presents information on the percentage of total food. expenditure
allocated to selected food items categories by tercile of total household expenditure.
Maize clearly remains the dominant staple in rural areas, although in the Southern
Region, the cassava share exceeds maize in the lowest tercile while rice equals maize.
The convenience of rice in terms of cooking time, together with its availability
throughout the year and much lower price relative to the predominated types of industrial
maize meal than in past years, turn rice (as well as maize grain and cassava) as potentially

important for the lowest income consumers.

43

Table 2.6.6a Percent of Total Food Expenditure Allocated to Food Item Categories
by Tercile of Total Expenditure and by Rural Regions of

 

 

 

 

 

 

 

Mozambique
Food Items North Center South
Categories Total Expenditure Total Expenditure Total Expenditure
Tercile Tercile Tercile
1 2 3 1 2 3 1 2 3
Percent of Total Food Expenditure
Maize and 16.5 21.7 33.1 30.1 52.6 52.3 8.4 12.7 16.1
Derivatives
Rice 4.4 5.3 5.6 1.2 2.4 4.3 8.5 9.9 10.8
Wheat and 0.3 0.5 1.2 0.3 0.4 1.6 0.9 3.1 6.4
Derivatives
Cassava 28.8 23.0 15.8 1.5 0.7 0.7 12.1 8.0 4.1
Leaves 6.7 5.1 2.8 11.9 5.1 3.0 29.2 16.9 11.6
Other 43.3 44.4 41.5 55 38.8 38.1 40.9 49.4 51

 

 

 

 

Data Source: IAF 2002

In urban areas (Table 2.6.6b), rice and wheat have gained major market shares in

urban consumption habits of the South. Even in the lowest tercile, their combined share

is nearly five times the maize share. This pattern is likely due to several factors. First,

rice and especially bread (wheat) are more convenient ways to consumer cereals, and in

many countries where these are grown and/or imported at relatively low costs, one sees a

movement towards such staples as urbanization proceeds, as incomes rise and consumers

seek less time consuming way to prepare basic foods. So in some respects, Mozambique

may be experiencing similar patterns. But a second likely reason for the patterns can be

related to policies that change the relative availability and cost of goods.

44

Indeed, import and tax policy in Mozambique affects maize grain prices and
availability, and also rice and wheat availability. Business policies also affect the nature
of commercial milling investments. And the price ratio between rice and maize meal has
dropped dramatically over the past several years, due in part to the structure of the maize
milling industry and the extremely high prices for most of the industrial maize meals in
Mozambique. Second, limited availability of maize grain during the hungry season in the
South of the country means that fewer consumers have the option of purchasing cheap
grain and processing it themselves (or by buying milling services from small hammer
mills) into much lower cost maize meal. Finally, rice is available year round through
competitive and large scale formal imports, and there is no VAT assessed on rice.

Policy changes might bring down the price of maize meal and make maize grain
more available throughout the year. But it is difficult to totally reverse the patterns of
shift towards rice and bread observed in the urban South. However, if policies can be
modified, they would almost certainly benefit low income urban consumers. The same
policy issues apply in the rural areas, only much more so, because the shares of maize
grain and meal in rural areas are much higher. In the South especially, the maize
production levels per household are also quite low. Less expensive maize meal (or a
wider range of commercial maize meal types with significant price differentiation) and
year-round availability of maize grain would offer real options and potential benefits,

especially to low income consumers.

45

Table 2.6.6b Percent of Total Food Expenditure Allocated to Food Item Categories
by Tercile of Total Expenditure and by Urban Regions of

 

 

 

 

 

 

 

Mozambique
Food Items North Center South
Categories Total Expenditure Total Expenditure Total Expenditure
Tercile Tercile Tercile
l 2 3 l 2 3 l 2 3
Percent of Total Food Expenditure
Maize and 13.8 21.6 14.3 17.8 28.4 20.1 4.2 3.7 1.9
Derivatives
Rice 4.9 9.9 9.0 9.6 9.6 7.0 11.2 10.4 5.8
Wheat and 1.6 4.9 9.2 3.7 6.2 10.3 8.6 14.0 18.0
Derivatives
Cassava 34.5 11.0 2.5 1.0 0.9 0.8 6.6 2.4 0.7
Leaves 3.5 2.4 1.9 8.8 3.8 3.1 17.8 12.2 8.4
Other 41.7 50.2 63.1 59.1 51.1 58.7 51.6 57.3 65.2

 

 

 

 

Data Source: IAF 2002

2.7 Chapter Summary

The analysis of maize grain in Mozambique smallholder livelihoods showed that

maize grain is actively produced in all regions of the country. Large majorities of rural

households (96 percent in the South, about 75 percent in the Center) either do not produce

maize grain or produce it solely for home consumption. This was explicitly shown with

the concentration and levels of maize production. In all regions, between 50 and 77

percent of the production of maize grain is concentrated among household who produce

but do not sell.

The analysis also shows that maize sales are very concentrated. In all regions,

46

between 71 and 77 percent of the volume of maize sales was accounted for by the top 25
percent of sellers, representing only 6 percent of all households in the Center and North,
and only 1 percent in the South.

The top 25 percent of sellers are clearly better off as compared to other rural
households in all categories studied. Overall, the top producers and seller have the largest
percentage of marketed production, have maize yields more than double the average,
plant more than double the area, and have about double the amount of land per person in
their households. Their incomes are also about double those of other rural households,
they have significantly more assets and they are more literate, and have better access to
price information. In addition, these households are less likely to be headed by females.

Based on IAF results, there was no change in national average percentage of total
households’ expenditures (budget shares) allocated to maize and its derivatives between
1996 and 2002. In 1996, households spent on maize and its derivatives as much as in
2002 (15 percent), registering generally similar patterns on total expenditures for wheat
and rice over the period.

In general, the disaggregated picture from the IAF expenditure/budget share
analysis reveals important changes from 1996 to 2002. There is a pattern of shifts into
maize products in all areas studied, with the exception of the rural South where maize
products lost the strong dominant budget share position held in 1996. And in the rural
South, it appears that households tried to compensate with shifts into fruits and leaves,
and also to rice to a lesser extent.

The key findings of this chapter for the rest of the thesis are that the large majority

47

of rural households do not produce enough maize grain to satisfy their consumption needs
from their own production. Purchases of maize grain or maize meal are widespread in
rural areas, and are not confined to one single type of household. Overall, in the South, at
least 76 percent of rural households are net buyers of maize; based on production levels
and the IAF budget share data, it is likely that many of these households are purchasing
substantial quantities of maize or maize meal. Thus, the performance of rural maize
markets for purchases - prices charged and availability of grain and various types of meals
over the course of the year - will have more influence on the welfare of rural households,
especially in the South, than will the market for sales. In the next chapter, the subsector

analytical approach is applied to begin analyzing this performance.

48

CHAPTER THREE

OVERVIEW OF THE MAIZE SUBSECTOR IN MOZAMBIQUE

3.1 Introduction

As stated by Rubey (1995), an inefficient food marketing system is among the
main causes of hunger. Good food system performance is also very important to meet the
objectives of the participants in the food system such as remunerative producer prices and
accessible retail food prices for the poor, each of which may reduce food insecurity.

Several approaches may be adopted to conduct a marketing study, but the most
applied ones are: Industrial Organization (institutional) and the commodity subsector
approaches. In this chapter, the subsector analytical approach was applied to provide a
descriptive analysis of the maize grain marketing system at the producer level, passing by
the wholesale and the milling industry level, evaluating opportunities for market
expansion and improvements in vertical coordination. Particularly, in this chapter the
specific question number three will be answered, that is, who are the major participants in
the maize marketing system in Mozambique, what are their roles, and how is the system
organized?

In the next section of this chapter, the two approaches mentioned above are briefly
reviewed to enable the reader to link the economic theory to the issues addressed in the
chapter. Then, the rest of the chapter is dedicated to the description of the maize

marketing at the producer, informal wholesale, and the milling level.

49

3.2 Industrial Organization

According to Staatz (2003), Industrial Organization (I-O) is the study of the way
in which the organization of sellers of a particular product affects the performance of the
market and hence, the nation’s economic welfare. In other words, I-O is the study of how
productive activities are brought into harmony with society’ demands for goods and
services through some organizing mechanism such as the market, and how variations and
imperfections in the organizing mechanism affect the degree of success achieved by
producers in satisfying society’ needs.

As stated by Marion and Mueller (1983), although individual firms may be the
unit of observation in some I—O studies, the emphasis in industrial organization is mainly
on understanding behavior of groups of firms that act as competitors (either selling or
buying industry) or interact as suppliers and customers (Marion and Mueller, in Industrial

Organization, Economic Power, and the Food System, 1983).

3.2.1 Structure-Conduct-Performance Paradigm

As a branch of applied price theory, the basic paradigm of 1-0 which was
popularized by Bain in late 1950s, holds that market structure influence the competitive
conduct of firms in the market, which in turn influences market performance. Therefore,
structure, conduct and performance (S-C-P) is the basic framework of analysis in the
theory of 1-0.

In this framework, market structure is defined as characteristics of the

organization of a market that seem to influence strategically the nature of the competition

50

and pricing behavior within the market (Bain, 1968, cited by Pomeroy and Trinidad,
1995). The three salient aspects of market structure include the degree of seller and buyer
concentration, the degree of product differentiation, and the conditions of entry and exit.
The greater the degree of concentration, the greater is the possibility of non-competitive
behavior in the market. In addition, by creating high entry barriers or by attempting to
drive out existing firms, firms can attempt to alter the structure of the market. The larger
the deviation of these elements from perfect competition, the more imperfectly
competitive is the market.

Market conduct was defined by Bain (1968) and cited by Staatz (2003) as a
pattern of behavior that enterprises follow in adapting or adjusting to the markets in
which they sell or buy. On the other hand, Caves (1992) defined conduct as a firm’s
policies regarding the product market and moves by its rivals. In the case of maize
marketing in Mozambique, market conduct will define the conditions that make possible
exploitative relationships between traders and buyers, especially in the milling industry.
This may be done by aggressive tactics such as predatory pricing and exclusionary
strategies with the only purpose to gain advantage over or eliminate rivals, and then
discourage new entries in the market.

Market performance was defined by Brandow and cited by Staatz (2003) as how
well an industry does what society might reasonably expect it to do. A distribution
system displaying acceptable performance is one that, among other things, utilizes
resources efficiently, and transmits prices that reflect costs. Since structure and conduct

are instruments by which performance is influenced, the results of a specific structure as

51

well as aggressive tactics may result in poor performance, with levels of profits above

marketing costs, for instance.

3.3 Subsector Approach

A subsector is defined as the vertical set of activities in the production and
distribution of a closely related set of commodities (Shaffer, 1968 cited by Staatz, 2003).
In another definition, Marion et a1 (1983) defined a subsector as an interdependent array
of organizations, resources, laws, and institutions involved in producing, processing, and
distributing an agricultural product.

Essentially, this approach is an extension of the S-C-P paradigm, but it differs in
its reach and emphasis. Unlike the S-C-P paradigm that is focused on horizontal slices of
the economy (one layer of transactions), the subsector approach looks at more than one
stage of production, presenting therefore the processes and participants in the whole
chain, from producers up to consumers.

Hence, the subsector approach is just a vertical way of looking at food system
coordination, and captures the interdependencies between stages of production,
recognizing that they have to be coordinated in order for the system to perform

efficiently. Therefore, vertical coordination is the focus of subsector studies.

3.3.1 Vertical Coordination
As stated by Morris (1995), the inclusion of the vertical dimension is important

because problems in the food system can frequently be attributed to poor coordination

52

between successive stages of economic activity. Vertical coordination is defined as all
the ways (a process) of harmonizing the vertical stages of production and marketing
(Mighell and Jones, 1963, cited by Staatz, 2003).

The harmonization that Mighell and Jones mentioned in their definition can take
place through various market coordination mechanisms or institutional arrangements.
For instance, in a competitive market, vertical coordination may take place through spot
markets, but there are other mechanisms of coordination ranging from various types of
contracting through full vertical integration. The desired outcomes of vertical
coordination that is guided by some of these coordination mechanisms include effective
market integration and provision of a desired product at prices that consumers are willing
to pay.

Since the objective of subsector analysis is to describe the organization and
functioning of the maize grain subsector, one of the main elements in the study of maize

marketing in Mozambique will be marketing channels.

3.3.2 Marketing Channels

Coughlan et al (2001) stated that the originator of goods or services gains access
to a market through marketing channels. Marketing channels are defined as a set of
interdependent organizations involved in the process of making a product or service
available for use or consumption (Coughlan et al, 2001). In simple words, marketing
channels are the different routes that a specific product takes from producers to

consumers. Thus, marketing channels provide a framework to clarify the factors

53

influencing the organization of maize marketing in Mozambique.

Even though this definition makes clear what a marketing channel is, it bears
some explanation. First, it points out that a marketing channel is not just one firm doing
its best in the market. Rather, many entities are involved, and each depends on the others
to do their jobs. Second, the definition points out that the purpose of a marketing channel
is to satisfy the end-users in the market.

In the context of this study, the term “marketing channel” will be used as a
description of the set of firms or activities that add place, time, form or possession utility
to a product as it is transformed from a raw material or intermediate product into one that
is purchased by another firm or final consumers. In this way, there are two important
functions carried out between producers and consumers. The first is the marketing
function that includes exchange functions (buying and selling) as well as physical
functions (transportation, storage, processing). The second function is the facilitating
function which includes standardization, finance, information and risk bearing.
Therefore, marketing channels through which maize move from producers to consumers
will be identified, as well as the functions performed by each participant. In summary,
the use of the marketing channel concept enables the discussions of which intermediaries

(or participants) act in maize marketing and the functions carried out each.

3.3.3 Market Participants
Market participants are defined as marketing and facilitating intermediaries who

perform one or more market functions and use institutional arrangements such as spot

54

markets, contracts, or vertical integration to coordinate their activities. In the case of
maize grain in Mozambique, marketing participants are informal and formal traders (both
wholesale and retail), millers, and owners of rural and urban shops, while facilitating

participants are basically the transporters and government institutions.

3.4 Marketing at Producer Level

Several characteristics of the production level such as structure of maize
production and major zones of production, were already described in chapter two.
Therefore, the focus of the analysis in this section is on marketing channels at the
producer level.

TIA data showed in Chapter Two that production and especially marketing was
relatively concentrated among a small group of farmers. In the South (Maputo, Gaza, and
Inhambane provinces), we showed that about 1% of all farmers accounted for more than
three-quarters of all maize sales. In the Center (Tete, Manica, and Sofala provinces),
about 6% of farmers accounted for slightly less than three-quarters of all sales. TIA data
also show that production in each region is geographically dispersed: In the South, no
district had more than 12% of total regional sales, while in the Center no district exceeded
a 15% share. These results make it clear that the large producers and sellers are spread
over a large geographic area, implying high costs for assembling maize grain.

The relative importance of each type of “first buyer” is shown in figure 3.4.a for

the Southern Region and 3.4.b for the Central Region.

55

NooN 5... ”3.50m San.

 

 

 

 

 

 

 

 

 

 

 

£64. £6. ..x.~.o~. 3.3 c Q. 3. 38.8.
..m — o 30 «mm am am». _.
sauna... .afiaoh 5...:an 5.0.8... 52.5.02 «3.32 .3537. neam =23— .ocaah. ...—n.5,...—
- _ , - . - . _ . _ - . ,
. . ,
£68 £15 $8.3.
54m oww.~ 53.8 P
Sew 2. 3 .515...— uuzaaE :. ...em 0:5: .a 3:233. 6
. _ 5
co... mv F Kc P
5.325...— .39..
.533. ...—2.23m

2.5.3—use: ..e .833— Eofisem o... ... 59.0 3.3). ...... ...-32.0 net—.....cma swam 9.33%

8cm <5. “Shaw 8.5

 

 

 

 

 

 

 

 

 

£39 388 38. $8.8 38.8 .83.
whoa” v2.9. m 3N Fund wmm ommdw
Lough. ...—Far.— .mbusenv 5.0.3,.— ..eaaw.oz «8.53). .3537. 9am 3.5M 50.3.; 33.3....—
T . . . . . .
an 3:3 5 @038
movdw mo —. .mv www.mwm
Sew 2. 8 606:8:— uoxuaz :. Sew «Sem— 2. 62:53—
4 w - _ . .7.
cot. .. :14
5:935...— .30...
.533— 1.5.3.0

9:53:33: ..e .333— .abaoU a... a. £26 3.3). ...... .oaaaanv act—5...??— avd 9.5»...—

Based on the distribution channels for maize grain in the two regions, five
important features are seen. First, 97 percent of the total household production in the
South, and about 90 percent in the Center, was retained at home for consumption. Thus,
maize markets in both regions can be characterized as very thin. As a consequence of
these thin markets, large buyers or sellers may be able to influence maize prices, affecting
therefore the performance of the sector.

Second, due to lack of data, only the first link in the marketing chain is shown,
that is, the movement of maize grain from producers to the first buyer. Third, the most
important first buyer for producers in both regions is the informal trader
(assembler/wholesaler), with almost 60 percent of all purchases in the South and 55% in
the Center. With low levels of invested capital, relatively low opportunity cost of time,
and the ability to move as needed to reach dispersed production, these traders are well
adapted to the structure of production in the country. Fourth, in the Center, the second
most common sales destination is a foreign country. This reflects primarily sales from
Tete Province into Malawi. Last, there are no reported sales from producers directly to

the milling industry in either region.

3.4.1 Informal Traders

In theory, there is a clear distinction between a wholesaler and retailer of maize
grain: A wholesaler is defined as a trader engaged primarily in buying products from
other traders in relatively large quantities, and then reselling the products in smaller

quantities to retailers, industrial or other institutional concerns

58

(Rosenbloom, 1999). The key point in this definition is that wholesalers purchase a
product (maize grain in this case) from another trader in reasonable quantities, and do not
sell any significant amount of maize grain to consumers.

In the maize subsector in Mozambique, it is not accurate to talk about the pure
wholesaling function embodied in Rosenbloom’s definition. Indeed, informal traders
may sell the whole product in bulk to another buyer, but they may also sell maize grain in
small amounts to consumers, acting therefore as retailers.

Therefore, informal traders of maize grain in Mozambique can be defined as those
traders buying maize grain in bulk and selling either in bulk or at retail. If these informal
traders buy maize grain directly from producers as shown in Figure 3.4a and 3.4b above,
then they are also acting as assemblers. So they can mix assembly, wholesale, and retail
all in one person. Because retailing is less common than the other two functions among
these traders, the term informal assembler/wholesaler will be used in this study to refer to
informal traders.

This particularity of one trader being an assembler, wholesalers, and perhaps
retailer at the same time is not exclusive to Mozambique. In her analysis of the structure
of the wholesale cereals market in Mali, Mehta (1989) divided the wholesalers in two
categories, namely, the wholesaler of wholesalers and semi-wholesalers. The main
difference between the two categories was the sales clientele as well as the average
volume of total sales. The latter category (semi-wholesaler) is quite similar to informal
traders in Mozambique, since their clientele can include consumers and trading volumes

are small.

59

The term “informal” is used to categorize these participants in the maize
marketing system because, although they may have to pay some daily tax to operate in the
physical market (stalls), they have no permanent physical infrastructure, nor do they have
. a unique tax identification number (NU IT). Formal traders, who do have physical
infrastructures and are registered with the tax authorities, have to follow and pay all fiscal
rules and taxes while informal traders only pay a daily tax by occupying a small stall in

the maize market.

3.4.2 Neighbors

Rural producers in the Southern Region make 20.2 percent of their sales to
neighbors, which makes this group the second most important buyer in this region after
informal traders. The importance of this channel likely reflects the very small scale of
sales in this region compared to the Center. In the Central Region, only 6.3 percent of
maize production is sold to this group of actors. The destination of maize grain
purchased from this group is unknown. Members of this group may be a deficit producer
purchasing maize grain to supplement their home consumption, or another rural retailer

purchasing maize to sell in the nearby market.

3.4.3 Informal Rural Retailers
Usually, informal rural retailers are local people, mainly women, who buy maize
grain from local producers or informal assemblers/wholesalers in small quantities to sell

in the nearby retail market. These retailers acquire 12.3 percent of the total production

60

sold by producers in the Southern Region, and 8.2 percent in the Central Region.

Based on conversation with informal rural traders and market observations, on
average the quantity purchased each time by rural retailers does not exceed a 70-Kg bag.
Commonly, rural retailers sell less than a kilogram per transaction, using non-standard
measures such as small cans. Besides maize grain, they also sell other agricultural

commodities such as beans and maize meals.

3.4.4 Rural Shops and Formal Traders

Rural retail shops are another key actor in the maize marketing system in
Mozambique. Although this segment buys only 3.1 percent of the production sold by
producers in the Southern Region, their presence in rural areas plays an important role
because they are ofien the first source of supply of consumer goods in rural areas. In
addition, they buy maize grain from local producers and transport it to consumers’
markets. However, capital is a major constraint for them because many of their shops
were destroyed during the civil war, and reconstruction has been difficult. In addition, it
has been difficult to compete with the informal traders, since this last group does not pay

taxes.

3.5 Marketing at Wholesale Level
As the structure of a subsector affects the conduct that also affects the
performance (S-C-P Paradigm), the nature of informal assemblers/wholesalers (the most

important link for producers) must be understood in order to design workable policies for

61

improving the performance of the maize subsector.

Hence, a survey with a structured questionnaire among informal retailers and
assemblers/wholesalers was carried out in two major cities in the Southern Region,
namely Maputo and Xai Xai. Twenty-nine interviews were conducted during July and
August 2004, 22 in Maputo City, and 7 in Xai Xai City. Information regarding the
traders’ experience in the trade, types of products they transact, maize grain flows,
transaction costs, risk, quality, transport, profitability, and access to credit were collected.

The results are presented below.

3.5.1 Profile of Informal Assembler/Wholesaler Traders

Results from Table 3.5.1 show that nearly 100 percent of the total informal
assemblers/wholesalers interviewed own and operate their businesses independently, for
an average of 6 (Maputo) and 5 years (Xai Xai) in the business.

In addition, 43 percent in Xai Xai and 50 percent in Maputo indicated that they
have only worked with maize grain. In Xai Xai, 57 percent of informal
assemblers/wholesalers sell beans and groundnuts in addition to maize. However, due to
the smaller sample included in this analysis, lack of specialization must be seen with
caution.

The main reason for them to work with these products is because they are more
profitable than maize, together with lack of credit to invest in other commodities. Only
one of the seven traders interviewed in Xai Xai mentioned that they have experience

dealing with products other than maize, beans and groundnut. When compared to other

62

commodities, maize trade is considered by two-thirds of traders to be less profitable.

In both cities, the majority of informal assemblers/wholesalers buy maize grain
from producers in 20-Liter cans, and about 40 percent buy it Strictly in 70-Kg bags. Five
of the seven traders in Xai Xai sell their maize grain exclusively in 20-Liter cans, while in
Maputo nearly 50 percent sell either in cans or bags. The frequent use of 20-Liter cans in
sales suggest that some share of sales is being made directly to consumers, not other

traders.

3.5.2 Maize Flows

Most informal traders in the South and Center are women. SIMA data shows that
in Maputo, from October 1998 to June 2004, 67 percent of all traders were women, while
in Xai Xai the percentage reached 96 percent. The dominance of women may be the
result of men’s immigration to South Africa to work in the mining industry.

From the selling points (terminal markets), these women use public transportation
to reach the point of rural assembly. Once there, they establish a small tent and start
buying maize in bulk from producers or other traders. Usually, maize sellers take maize
grain to the assembly points, carrying it in small bags using a bicycle or by foot.

During the period of purchases, informal assemblers/wholesalers keep the product
under the sun, spread on lonas (large tarp) to reduce the level of moisture. When the
desired amount of maize grain is obtained (function of each trader’s purchasing power),
they start packing maize grain in bags, whose capacity varies from 70-Kg to IOO-Kg.

Since transporters charge a fixed price per bag not based on the bag weight, they prefer to

63

use large bags to reduce the costs. It is common to see informal assemblers/wholesalers

tailoring a standard bag to increase their weight capacity.

Table 3.5.1 Description of the Activities Performed by Informal Traders

 

 

 

 

 

 

 

 

Item Responses Maputo Xai Xai
Years in business of Number of years in the business 7(6) 8(5)
agricultural commodities (Median)
1.- ...- ,4 ..-- _ 4.. __ _, ___ __ _____' _ ., _‘_ - .. Percentage
Business ownership Own business 96 100
Maize plus beans, rice, or 40 57
Products sold groundnut
Maize bran 10 0
Only maize grain 50 43
Not enough money to invest in 50 43
other products
Reason to only sell these , , ,
products Uncertainties With other 0 14
products and heritage
More profit with these products 50 43
Profitability of maize Less profitable 55 100
compared with other
products
Small cans (20-liter or 5-liter) 50 57
How maize grain is
bought Bags (70 Kg) 41 43
Both 9 0
Small cans (20-liter or 5-liter) 24 71
How maize grain is sold
Bags (70 Kg) 28 0
Both 48 29

 

Data Source: Author Informal Wholesaler Survey Results

64

Transport back to the destination market depends largely on backhaul. As a
result, most of the assembling points are located along the main roads. The relative lack
of medium or large scale processing and manufacturing outside of major cities means that
many trucks transporting consumer goods to rural areas from cities risk returning empty.
As a result, informal assemblers/wholesalers are typically able to get very low transport
costs for their goods back to cities in which they intend to sell, though this low rate is
ofien at the cost (non-financial) of waiting several days for a truck to appear. Actual
transportation costs per bag vary among regions. For instance, in May 2004, from
Nhamatanda (a key collection point in the Center) to Maputo, transport of one bag cost
40,000 Meticais (nearly SUSD 2), while transport from Nampula in the Northern Region
to Maputo cost 120,000 Meticais (SIMA database).

In Maputo, at the time of these interviews, all maize grain available in the biggest
wholesale market (Xiquelene) was purchased in Nhamatanda, district of Sofala Province,
in the Central Region of the country. From this biggest informal wholesale market, small
informal retailers buy and carry maize grain out to small informal retail markets, namely
Xipamanine, Wancacana, Compound, and others. Small urban shops (Mercearias) and
the increasingly common Contentores (rail cars adapted to sell products at retail, typically
with a smaller selection than Mercearias but much larger than any market seller) almost
never sell maize grain. In three smaller informal wholesale markets covered in this study,
namely Santos, Bazuca, and Benfica, all informal assemblers/wholesalers had purchased
their maize grain directly in the Central Region, more concretely in districts of

Nhamatanda (Sofala Province) and Manica (Manica Province).

65

Figure 3.5.2a and Figure 3.5.2b depicts the flow of maize grain from the Central
Region to Maputo and from Central Region to the Maputo informal wholesale market and

to consumers, respectively, while Table 3.5.2a and 3.5.2b presents the main sources of

maize grain for all informal traders interviewed in this study.

Figure 3.5.2a Flows of Maize from the Central Region to the Southern Region

  

Tete Province:
Changara,

Angénia
Gaza Province:

‘. Chokwe, Xai-Xai,
Chibuto
Maputo City

66

Manica Province:
Guro

     

Manica Province:
Manica, Chimoio

 

 

 

 

 

 

Sofala Province:
Nhamatanda,
Gorongoza, Caia

 

Boa 02:... 2.839.: .88 2.. 28:2 «30...: :25. ”8.02

 

 

3:25:50 .:::u:o::>>

32:55.0 A . 65:23.56 33:: :_
$02.32 :33. .258:— =:.:m

 

 

 

 

 

 

 

 

 

 

€2.33—
.me.::m 595:6 3.3:: :.
23.22;? _:.:..e..:_ =:.:m

:2me
_:...:o0 :_ «2::

3:22:50 3.5:: 5 $2.32
232259 35.8.2:— .mowwmm

67

 

 

 

 

 

 

 

 

 

 

 

553300 3 ...... 302...: 2322.3 .258:— eusaaz 0.: 3 :29“ _:.....00 2: 8...... £20 92:2 .... B2..— amfid 9...»...—

Table 3.5.2a Source of Maize Grain in Informal Wholesale Markets in Maputo
and Xai Xai, July 2004

Destination Point of Origin

 

Market

 

(Wholesale) Maputo Xai Xai Gorongoza Nhamatanda Manica Tete

Percentage of Traders in Destination Market Obtaining Their
Maize from These Sources

 

Xiquelene O 0 0 100 0 0
Santos 0 0 0 0 l 00 O
Benfica 0 0 0 1 00 0 0
Bazuca 0 0 O 1 00 0 O
Xai Xai 0 43 29 14 14 0

 

Data Source: Authors Informal Wholesaler Survey Results

Table 3.5.2b Source of Maize Grain in Informal Retail Markets in Maputo,
July 2004

Destination Point of Origin

 

Market . . . .
(Retail) Maputo Xai Xai Gorongoza Nhamatanda Manica Tete

 

Percentage of Traders in Destination Market Obtaining Their
Maize from These Sources

 

Wancacana 100 0 0 0 0 0
Compound 1 00 0 0 0 O 0
Xipamanine 75 0 0 0 0 25

 

Data Source: Authors Informal Wholesaler Survey Results

In Xai Xai on the other hand, 43 percent of traders had obtained their maize from
within the city. This means that in Xai Xai, nearly half of informal traders were buying
maize grain brought by other informal traders from the surrounding districts of Gaza

Province or from the Central Region. Although they were buying considerable volumes,

68

maize grain was being sold mainly in cans as reported in Table 3.5.1 above, meaning that
at least some sales were at retail.

When the SIMA data from October 1998 until June 2004 are used to analyze the
flows of maize grain into Maputo’s biggest informal wholesale market (Xiquelene) and
into Xai Xai, the pattern is not different from what informal maize grain traders have told.

Maputo in the past was mainly supplied with maize grain from districts of the
province of Manica located in the Central Region, especially from the district of
Chimoio. However, preliminary examination of SIMA data suggests that this maize grain
flow has decreased in recent years. Indeed, as shown in Table 3.5.2c below, in the last
three marketing years (the marketing season in Mozambique begins in April when the
harvest time of maize grain starts, and ends in March of the next year) the main source of
maize grain for the Maputo informal wholesale market shifted from the districts of
Manica Province to the districts of Sofala Province (especially Nhamatanda). Table
3.5.2c reports the source of maize grain sold in the Maputo informal wholesale market

(Xiquelene) per marketing year.

69

Table 3.5.2c Source of Maize Grain Sold in Maputo Informal Wholesale Market
per Marketinglear, 1998/99 - 2004/05

 

 

Marketing Sources of Maize Grain to Maputo Market
Year , , , ‘ ,
Gaza Provmce Manica Provmce Sofala Provmce
Chokwe, Xai Chimoio All Nhamatanda All
Xai, Chibuto Province Province

 

Percentage of Traders Obtaining Maize Grain From These Sources

 

 

1998-1999 24.7 35.6 35.6 0 0
1999-2000 22.3 45.8 56.3 0 0
2000-2001 30.7 6.7 6.7 0 0
2001-2002 22.2 31.3 35.5 0.0 14.6
2002-2003 13.4 14.0 18.0 54.0 60.7
2003-2004 0 6.0 45.8 34.4 35.8
2004-2005 0 13.5 29.7 62.2 67.6

Data Source: SIMA

Notes:

1. Data represent % of traders indicating each area as the source of their supply, from three trader
interviews per week; .

2. The percentage does not sum up to 100 percent because the rest of maize comes from other districts from
the Central Region that account for small percentages;

3. 2004-2005 is only the first three months (April, May, and June).

From Table 3.5.2c, three important features in every marketing year stand out.
First, since 1999-2000, well over 50 percent of maize grain sold in the Maputo informal
wholesale market has come from the Central Region, either Manica or Sofala Province.
The only exception was in the 2000-2001 market season due to the floods that devastated
the Southern Region and halted the traffic between the Southern and the Central Region.
Second, Chimoio as an assembly point has fallen in importance over the past 7 years.

Despite the reduction in Chimoio’s role as an assembly point, Manica province as a

70

source of supply has more or less “held its own” (its share in 03/04 was the second-
highest of the past 7 years). Third, Sofala Province as a source of supply, and
Nhamatanda as an assembly point, has grown greatly in importance.

Regarding the decreasing importance of the Chimoio area, among other reasons,
informal traders interviewed in this study mentioned that it is easier to get transport fi'om
Nhamatanda to Maputo because there are more empty trucks returning to Maputo City
from Beira City. In addition, the informal traders argued that they may buy maize grain at
much lower prices in the production areas around Nhamatanda, compared to Manica and
Chimoio. Before, maize grain was bought in Chimoio City through intermediaries who
used to buy maize grain inside the production areas. Last, few maize traders in Chimoio
and Manica said that there are some informal exports to neighbor countries such as
Zimbabwe. Hence, the competition for maize grain produced within the Manica province
may be high, a situation that may increase the prices during the harvest. season, forcing
therefore, informal traders to be concentrated in Nhamatanda.

However, the informal traders in the survey mentioned that when the availability
of maize decreases in Nhamatanda, they still bring maize grain from Manica and
Chimoio, as well as from other districts from the Central Region. SIMA data support this
statement, showing that during most months of the 2002/03 and 2003/04 marketing
season, both Chimoio and Nhamatanda were important suppliers to Maputo (Table

3.5.2d).

71

<25. ”858 an:

 

 

 

 

 

c c 8 c c 2: 8. o o c 532
o o 2: c c ...a E... 2 m... o $2.0...
o o 5.. o c S. :8 .... 3 o Sea.
3 .... 9:. o c 2: 8. o c c .3580:
m... S: 2. 92 c :8 a... ... o o .3582
m: 3 ...... c o 9% 3.: can a... o .3800
:8 :8 m... o o 3. :3 a... a... o 5.53%:
e... m... 3 o a 3m 3 :8 3 o .33..
a... 9% 9% 9% c 98 3.. 9o: :8 c .3.
c2 98 a... a... o o o 3. N... 98 25.
c... a... can a... o I. o 3. o ...R a:
8. 8. c o c o o c o :8 .23
89.55 3.2; EB... 5:..0 02:2 ”25:30 9.2.9.... .... 09359.9.—
._..8 ... .....a 8.5.2 32...... «Bow ... «...... 8...... note...
acre... 2.2.: ... 33...... 38.30 595:8 395:: .522 ... note... 2255 52.2.8 .32
23> «33...: 48. - n8. .3. ”5.3.52 8... - N8. Lucia:
VQMSN . mENgN m.:oWLwM.=3_..:S .55.: mate—...: .2— 3.E:E ... 2cm 5:..0 05:2 .... 3......m ..Nfid 2a.;

72

In the case of Xai Xai, what informal traders have told in the survey is also
consistent with data from SIMA, which show that there is some maize grain coming from
districts from Gaza Province, and much more from the Central Region. Table 3.5.2c

below presents the sources of maize grain sold in the Xai Xai market.

Table 3.5.2c Source of Maize Grain Sold in Xai Xai Informal Wholesale Market
per MarketifiYear, 1998/99 - 2004/05

 

 

Marketing Sources of Maize Grain to Xai Xai Market
Year
Gaza Province Manica Province Sofala Province
Chokwe, Xai Xai, Chimoio All Nhamatanda All
Chibuto Province Province

 

Percent of Traders Obtaining Maize Grain From These Sources

 

 

2001-2002 12.2 64.6 82.9 0 0
2002-2003 24.7 39.8 43.0 0.0 9.7
2003-2004 1.3 33.3 41.3 25.3 46.6
2004-2005 4.2 8.3 45.8 25.0 25.0
Data Source: SIMA
Notes:

1. Data represent % of traders indicating each area as the source of their supply, from three trader
interviews per week;

2. The percentage does not sum up to 100 percent because the rest of maize comes from other districts from
the Central Region that account for small percentages.

3. 2004-2005 is only the first three months (April, May, and June).

3.5.3 Volume of Bags Marketed per Week

The scale of operation of informal assemblers/wholesalers is typically very small.
In both Maputo and Xai Xai, more than half reported selling between 1 and 10 bags of
70-Kg of maize grain per week. The smallest traders indicated that they are selling more

volume now compared with three years ago whereas those with sales above 20 bags per

73

week believed that they were selling less. Table 3.5.3a summarizes these results.

Table 3.5.3a Volumes Marketed by Informal Assemblers/Wholesalers per Week

 

 

Number of Number of Maputo Xai Xai
Bags Sold Observations
(70-Kg) Percentage
l - 10 17 57 71
Volumes Marketed
per Week 10-20 5 19 14
> 20 6 24 14

 

Data Source: Authors Informal Trader Survey Results

3.5.4 Suppliers and Demanders of Maize Grain

When the informal assemblers/wholesalers within Maputo and Xai Xai were
asked from whom they were buying maize grain, the most important suppliers for all of
them were producers. Table 3.5.4a and 3.5.4b presents the results indicating the most
important supplier for maize grain sold by informal traders included in the survey, for

wholesale and retail market respectively.

74

Table 3.5.4a Most Important Supplier of Maize for Informal Traders Selling in
Wholesale Markets

Province Wholesale Number of Most Important Suppliers

 

M k t I t ' '
ar e n erviews Percentage Indicating These as the Main

 

 

Suppliers
Producer Informal
Assembler/Wholesaler

Xiquelene 5 100 O

Maputo Bazuca 1 100 O
Santos 3 100 0

Benfica 3 100 O

Gaza Xai Xai 7 57 43

 

Data Source: Authors Informal Wholesaler Survey Results

Table 3.5.4b Most Important Supplier of Maize for Informal Traders Selling in

 

 

 

 

Retail Markets in Maputo
Province Retail Number of Most Important Suppliers
' Market Interviews . , ,
Percentage Indicating These as the Mam
Suppliers
Producer Informal
Assembler/Wholesaler

Maputo Wancacana 3 O 100
Compound 3 O 100
Xipamanine 4 25 75

 

Data Source: Authors Informal Wholesaler Survey Results

In the case of demanders, consumers were indicated as the main demander for
maize grain among informal assemblers/wholesalers. Some informal retailers in

Compound and Wancacana Market have reported that they bought maize in Xiquelene,

75

which means that informal assemblers/wholesalers in Xiquelene are selling maize to
retailers. However, at the time of the interviews none of the informal trader at Xiquelene
mentioned that they were selling to retailers.

Nevertheless, in some markets within the Maputo area, assemblers/wholesalers
sell their maize to retailers as in case of Bazuca Market. This informal wholesale market
supplies the closest retail market (Xipamanine) with maize grain. Despite the
Xipamanine being historically a retail market, one informal trader operating in this
market acquired maize in the Central Region, and started to sell to other small retailers in
that market. This probably is a new emerging pattern and reduces the transaction costs to
small retailers operating in that market. Table 3.5.4c and 3.5.4d presents the results
indicating the most important demander for maize grain sold by informal traders, for

wholesalers and retailers respectively.

Table 3.5.4c Most Important Demander for Maize Among Informal Traders in

 

 

 

 

Wholesale Markets
Province Wholesale Number of Most Important Demanders
Market Interviews , _
Percentage Indicating These as the
Main Demanders
Consumer Retailer
Xiquelene 5 100 O
Maputo Bazuca 1 0 100
Santos 3 100 0
Benfica 3 100 O
Gaza Xai Xai 7 100 O

 

Data Source: Authors Informal Wholesaler Survey Results

76

Table 3.5.4d Most Important Demander for Maize Among Informal Traders in

 

 

 

 

Retail Markets
Province Retail Number of Most Important Demanders
Market Interviews ' , ,
Percentage Indicating These as the
Main Demanders
Consumer Retailer
Maputo Wancacana 3 100 0
Compound 3 100 0
Xipamanine 4 75 25

 

Data Source: Authors Informal Wholesaler Survey Results

3.5.5 Quality of Maize Grain

When the informal assemblers/wholesalers buy maize grain, they identify the
maize quality through visual inspection, focusing on the level of insects and the mixture
of colors.

Usually, the majority do not pay any premium for good quality of maize nor
receive any premium when they sell it. But when they do receive a premium during the
selling process, it is because the level of insects is very low and maize grain is uniform in

color.

77

Table 3.5.5 Most Important Quality Factor for Informal Traders

 

 

 

 

 

 

 

Do You: 3 Percentage Saying Yes
Pay attention to Maize Quality During Purchases? 97
Pay a Premium for Maize Quality During Purchases? 38
For Which Quality Do You Pay a Premium During Percentage

Purchasing? b

Low Level of Insects 36
Homogeneous Color 64

Do You: ° Percentage Saying Yes
Receive a premium for Maize Quality During Selling? 21

For Which Quality Do You Receive a Premium Percentage
During Selling? d

Low Level of Insects 22
Homogeneous Color 78

 

Data Source: Authors Informal Wholesaler Survey Results
Note:
a, b, c, (1. Direct quotation from the questionnaire.

3.5.6 Transportation

According to the informal maize traders, transportation is the biggest cost that
they have to pay on their operations. The whole country has only 19 percent of its road
surface paved, and only 27 percent in gravel (INE, 2002). The precarious conditions of
the roads, as well as the high costs of diesel fuel (the real price increased 610 percent
between February 1995 and December 2003), are among the main reasons pointed out by
transporters.

Among the informal maize traders interviewed in this study, none owned their

own transport, but the majority (65 percent) confirmed that the best way to transport

78

maize grain was by truck. Only 22 percent mentioned that there are some losses during

transport, especially torn bags, and spoilage during the rainy period.

3.5.7 Credit

No interviewed traders indicated that they had access to credit to perform their
operations, but 35 percent of them give credit to some maize suppliers. The inability to
access credit by the informal assemblers/wholesalers is a huge barrier, and determines
largely the volumes of purchases. Therefore, the degree of specialization is directly

related to their financial capacity.

3.5.8 Access to Market Information

At the wholesale level, the access to market information is a crucial element in the
maize marketing system, at least for 55 percent of the informal traders. Interview results
reveal that personal contacts are the most important source of information (87 percent),
followed by the weekly price bulletins provided by SIMA (13 percent). Most informal
traders now use mobile phones to communicate and share market information within their
group. For instance, at the time of one interview, in Xai Xai, one informal trader was in
contact by cell phone with his partner who was in the purchase area acquiring maize
grain. At the end of the contact, the informal trader already knew when the new maize

grain shipment would arrive, as well as the number of bags and the purchase price.

79

3.6 Marketing at the Milling Industry Level

Maize millers may be divided in two groups, namely the industrial millers serving
mostly the urban areas and producing mainly two types of refined maize meal for sale at
retail, and small scale or hammer mills, serving the peri urban and rural areas and
producing whole meal. This meal can be purchased in markets, or “produced” by
households taking grain to the hammer mill. However, some households with access to
maize grain prefer to hand-pound it themselves into maize meal at household level. This
home processing is preferable over maize meal from both small scale and industrial

millers, and carries a price premium when sold in the market.

3.6.1 Attributes of Maize Meal Consumed in Mozambique

Maize grain in Mozambique is mainly consumed as processed maize meal. Both
refined and whole meal can be eaten as porridge or as a stiff mass known as chima or
uxua. Characteristics and some attributes of the various maize meals in Mozambique are
reported in Table 3.6.1. The types of maize meal can vary by the product extraction rate
in processing, which generally varies from 65 percent to almost 100 percent. In a low
extraction rate maize meal, a large portion of the maize germ and pericarp is removed in
processing. A hundred or very high extraction rate is obtained by hand pounding the
whole grain, or by sending whole maize grain to a small hammer mill where the final
maize meal contains virtually 100 percent of the whole maize grain sub components.
Maize meals produced industrially (highest extraction rate) are the more expensive in the

urban markets of the South, especially the first two in Table 3.6.1.

80

83. o... 380 .3... wag—Son .08... 8.... 883?. do...“ :80 ..

 

$8.: - 8m.»
25.” - EEK
can.» - 8c.”

@0508 athSMB: .3 mafia 080.. 8 .38 8.88 82....
83808 08.nov, .85 .08. .8..ow 2689. 0. 3.8.5.. .88.
£80.. ... Emacs .0 08.60.... m. 880 382 .038 0803

$.33: 2: 9.3: .38 :30 838
5... £80. ... Emacs .o @3300... m_ Efiw 382 .038 0803

«3.832 8 8305. :03 8?. $858

8.. mac—5 282. 0:20 $5.85.. can: ..o 08w»... :0 £82.20
3.... 82.888 2:. ...—man; m. 00.28 8.58 805 .88.

8 Ewaon 8 000.608 m. 88% 0882 82:8 .0882. =..8m

«383.2 mm 8305. :03 :89. 8 8m 82:8 .8888 =...8m
:82 ... ham 82:8 3.8.65
:39. ... 8m 82:8 3.5.65

:89. ... .95 82:8 .3530:—

”8.02
$.80 W
.82 3...: 323...... a a
w
$2-8. .82 m
8...: 2.5.. ...... M.
H
D...
m
exceé... «saws: .. n
m
.88.. «Emma: 0
£5... .36.. m xzo>
....ch 855.2 v 0.2m

$3-8. 2.2.... m 0082

 

8m... - 8o... ...... .. b... 8...... 35.6... $3 2.5%.). m 2.0

8...: - 8...: ...... .. a... 80...... .258... .38. 28m 80 . 2.0
Aggy Acts—00¢. 00858.00 9.259 nova—2 ...—E 85— 85088”..—

.. 3.....— Ecfluouoauaau .08 «2:5...— 08.. Z 52

 

23.8.82 ... 85.5.50 ...82 8...: .. 8.2.5.. .....m 2...;

81

Since the majority of maize consumed in Mozambique is consumed by the
households growing it, the method of processing maize in rural areas (method 6, 7, 8 and
9 in Table 3.6.1) is the most common in the country, depending on the Region. Method
6, where consumers purchase “mugaiwa” made by small retailers who sent grain to small
hammer millers, has virtually disappeared in the Maputo market, but is still present in
areas of Central and Northern Mozambique.

The price of maize meal varies per brand, being highest for the most refined
meals. In Table 1.2 in Chapter 1, the price of the top selling refined meal (Top Score) in
Mozambique between January 1996 and August 1998 was the second highest in Southern
Africa, and in late 2004, SIMA data show that the price was even much higher (over
US$600 per metric Ton).

The price ratio between refined meal and maize grain is very high in Maputo, as
seen in Figure 3.6.1, and this ratio is rising to historically high levels. This high price
ratio may be an indication that production of maize meal may provide high profits for

millers, and this evidence provides a room for new entrants into the milling industry.

82

Figure 3.6.1 Price Ratio Between Refined Meal and Maize Grain in Maputo, with a 12
Month Prior Moving Average (March 1998 - February 2005)

 

3.5

 

 

 

 

 

 

   

 

 

’5

X

z

E

.9

8

CE

12 Month CMA

' Ratio
07¢V%:°Le

,3
90 “9000000 079 047

Month and Year

A simple description of the main characteristics of small and large maize millers

is presented below.

3.6.2 Small Maize Millers
The technology used by these millers does not separate the bran in the maize

grain, but rather it grinds the whole grain. The role of these small scale millers is quite

83

different fi'om that of industrial millers. While the industrial millers usually buy maize
grain, process it to flour, and then sell it to retailers, the small scale millers typically act as
custom millers, processing maize grain owned by small retailers and consumers and
receiving a payment for this service.

TIA data (2002) show that small maize mills are quite common in villages of the
various districts of Mozambique (Table 3.6.2). Some rural villages in all regions of the
country have small mills, but the concentration is clearly low in the South, high in the
Center (where production and consumption is more oriented towards maize), and medium
in the North. Given that maize grain is broadly consumed in rural areas, mostly as meal
of some sort, overall, the existence of small hammer mills and hand pounding of maize
grain seems well alive in the rural areas (Table 3.6.2).

The reason for not seeing small scale or hammer millers entering in the market
with a branded product is unclear. However, some plausible reasons may be embodied in
some characteristics of this group of millers. First, they can be characterized as small
business enterprises employing no more than one or two employees. In addition, they are
also characterized by high turn over rates because of ease of entry and exit. Indeed,
during the period of distribution of maize grain as food aid in the early 1990s, the number
of small scale millers operating especially in Maputo was large, but numbers have fallen
substantially since that time. Third, they do not have any transport to put the product into
the marketing chain. Fourth, straight-run meal is relatively perishable, lasting only a few
days on the shelves. Fifth, if they try to invest in order to produce a better product with

more quality, they would be competing directly with large millers, who are much better

84

capitalized and have much more marketing power.

Table 3.6.2 Percentage of Villages Having Maize and. Cassava Mills

 

 

Province Percentage of Villages Having Mills for
Maize N Cassava N
Niassa 45.7 77 0.0 77
Cabo Delgado 27.0 63 1.6 63
Nampula 18.0 77 0.0 77
Zambezia 33.0 91 0.0 91
Tete 56.4 55 0.0 55
Manica 67.3 55 0.0 55
Sofala 69.4 49 0.0 49
Inhambane 18.7 49 2.1 48
Gaza 16.3 48 0.0 49
Maputo 28.6 35 2.9 . 35

 

Data Source: TIA

3.6.3 Industrial Maize Millers

Among industrial maize millers, Mobeira and V & M located in the Central
Region, and CIM, Merec Industries, and Sociedade Moageira do Chibuto in the Southern
Region, are the most important in terms of final output produced. They produce either a
super refined or a refined meal, whose extraction rate is around 65 and 80 percent,
respectively.

To understand the structure of maize milling in Mozambique, face-to-face

interviews with managers and/or directors were conducted with four millers. The

85

questionnaire, which is presented in appendix 2, was divided in seven sections and it was
designed as simply as possible in order to be easy to obtain answers from managers
and/or directors. It covered aspects such as maize purchases from national production
versus imports, sales strategy, quality problems, needing for fiirther research as well as
opportunities for the milling industry in general. Annex 3 presents a write-up of
interview results for each of the four millers; here we present a more summary discussion.

The four millers interviewed vary substantially in terms of production capacity,
although the largest two millers (CIM and MEREC) are nearly equivalent in terms of
capacity. These two largest millers are located near Maputo area in the South. CIM (the
largest) is operating in the market since 1997. Top Score, the brand name of the most
refined meal produced by CIM (and the one that they market most aggressively) is found
throughout the country, but the Southern Region is the main target area. This pattern is
similar with Mpupu, the brand name of maize meal produced by MEREC, which is more
available in the South.

In the last five years, three more industrial mills entered in the market, namely
SMC, Inacio de Sousa, and Vonk. Two of these mills are located outside of Maputo in
the South (SMC and Inacio de Sousa) and the other one in the Center. These millers are
producing a less refined maize meal compared to that produced by CIM and MEREC, and
their main targets are the markets where they are operating. SMC and Inacio de Sousa,
for example, do not market aggressively in Maputo, sticking primarily to outlying
markets in the South. Nevertheless, Vonk is shipping maize meal to the South,

particularly to Maputo. This suggests that the very high prices for refined meal in the

86

country have created space for competition from new companies, even after paying
substantial transport costs from the Center of the country.

The national concentration ratio of top 4 millers (CR4) is very high, about 92
percent, with CIM and MEREC together having 72 percent of the market share, as shown

in Table 3.6.3.

Table 3.6.3 National Market Share of Maize Millers Interviewed in the Study
Name of Starting Name of Current Milling Location National

 

 

 

Miller Year of Meal Capacity Market
Operation Produced (T on/Month) Share (%)
CIM 1997 Top Score, 2350 South 38
Matabicho
Merec 1999 Mpupu 2 100 South 34
MOBEIRA" 1997 ? 900 Center 14
Vonk 2003/04 Power 400 Center 6
SMC 2000/2001 Mananga 240 South 4
Inacio de 2002 Palmeira 240 South 4
Sousa“
Total - 6,230 100
Data Source: Authors miller survey results
Notes:
"' Not included in the Study

3.7 Chapter Summary
As shown in chapter two, the Central and Southern Region of Mozambique
produce together 49 percent of the total maize production. In the Southern Region, after

retaining 97 percent of the total production at home for household consumption, maize

87

producers sell their remaining harvest to informal assemblers/wholesalers (60%), directly
to individuals at nearby markets (12%), neighbors (20%), and the remaining to other
small participants.

The flow of maize grain from producers to consumers is mainly done by informal
traders. They are the principals responsible to stabilize maize prices in consumer
markets, besides the fact that they are the principals “engines” that create incentives for
production.

The key role of informal traders within the producer marketing system is
explained by the relative dispersion of the production. These informal traders are the
only market participants willing to bear all costs of collecting the disperse maize grain
production.

Nevertheless, the growth of this sector is constrained by limited access to formal
credit, which hinder informal traders from specializing and increasing the scale of their
operation. Absence of market information widely and timely available as well as high
transportation costs are other constraint that impedes the grth of these participants.

Districts from the Central Region are still the major suppliers of maize grain to
the informal wholesale markets in the South during most of the year, regardless of
competition from neighbor countries. Despite the increasing importance of the district of
Nhamatanda in supplying Maputo with maize grain in the last three marketing years, this
market is still supplied with maize grain from districts of Manica Province, inclusive
Chimoio. In fact, between February and May, the major suppliers are the districts from

the Southern Region. However, is during this period where the quantity of maize grain

88

coming from Nhamatanda exceeds by far the quantity coming from Chimoio. In the
remaining months, most of the maize grain comes from the districts of Manica Province.

The percentage of villages with small hammer mills is very high in the rural areas
of Central Region, followed by Northern, and Southern Regions, respectively.

The price ratio between refined meal and maize grain is increasing over time, and
this increases may be seen as a potential opportunity window for new entrants in the
milling industry, and one would expect to see more entries into the milling market, at
least until the price ratio between refined meal and maize grain starts coming down.

Based on the number of industrial maize millers (does not exceed six) operating
and producing branded products in the Central and Southern Region of Mozambique, it
seems that there is a significant market power, reflected in a high concentration on the
milling industry since the largest two millers (with production capacity above 2,000 Ton

per month) are holding 72 percent of the market.

89

CHAPTER FOUR

THE IMPACT OF VAT ON MAIZE GRAIN IMPORTS

4.1 Introduction

This chapter presents an analysis of the impact of the value-added tax (VAT)
imposed by the Mozambican government on maize grain trade, either transacted locally
or imported. More specifically, the specific research question number three will be
answered.

The chapter is divided in eight sections. In the next section, a brief description of
the VAT code is presented. This section summarizes the main points included in the
code. It serves as a guide to link the reader to the material covered in other sections. In
section three, the behavior of price of maize meal in Maputo is presented. In this section,
the evOlution of maize meal prices is shown, covering the period before the introduction
of VAT until more recent days. In addition, prices of maize grain are plotted together
with rice prices and maize meal prices in order to assess the magnitude of price
differences among them. Section four presents an assessment of the impact of VAT
charged on imported maize grain. In this section, import parity prices for maize grain in
Maputo were computed for two scenarios, one with VAT charged on imports, and one
without charging VAT. Both scenarios are then compared with prices of maize grain
landed in Maputo from the Central regions. In section five, the regulation to import
maize grain in Mozambique is presented. This is important to get insights about the

problems faced by anyone intending to import maize grain, especially informal traders.

90

In section six, a brief overview of the South Afiican maize marketing system is presented,
emphasizing the potential problems faced by any maize importer. Section seven
summarizes the fiscal incentives given by the Mozambican authorities and possible

impacts of them for the milling industry. The last section presents the chapter summary.

4.2 The Code of Value Added Tax (VAT)

On 29 September 1998, the Council of Minister of Mozambique approved the
code of Value-Added Tax, which replaced previously existing “circulation” and sales
taxes. The VAT took effect five months later, in April 1“, 1999, and its value was fixed
at 17 percent.

The VAT code states that every transmission of goods or services carried out by a
passive person inside the national territory is subject to VAT. In addition, the code states
that all imports of goods are also subject to VAT charges, unless specifically excluded.

A passive person3 in the VAT code is defined as a person, either single or
collective, resident, with a stable establishment or representation in the national territory,
performing production, business, or services activities, including extractive activities,
cattle raising, fishing, or agriculture. In addition, a passive person is also defined as a
person, either single or collective, that even not performing any activity, is carrying out
any operation susceptible to be taxed since it follows the presupposes of real incidences

of industrial tax or tax above income work. Furthermore, a person, either single or

 

3

For a more complete definition of a passive person, see Codigo do 1 VA e do Imposto
sobre Consumos Especificos, article 2"“, on page 18.

91

collective that is carrying out imports of goods, is also defined as a passive person.

How is the VAT charged? Apparently, the mechanism is not that complicated. If
the VAT amounts that a company paid to acquire any product in a specific month is
greater than the amount of VAT charged selling the same product, the difference between
these two amounts is the total value that the government has to reimburse to that
company. On the other hand, if the amount of VAT paid on purchases is smaller that the
amount of VAT charged on sales, the difference is the amount that the company has to
pay to the government.

As an example, lets supposed that in a specific month a company spent
$US400,000 on local purchases. If the products purchased by that company is not exempt
of VAT, the amount of VAT that has to be paid is 17 percent of that amount, that is,
$US68,000. If that company also imported certain products with VAT payments of
US$7,000, the total amount of VAT paid by the company is US$75,000.

If the total sales of that company in the same month were US$1,000,000, the
company will have charged US$170,000 in VAT. Then, as the company paid
US$75,000 on purchases and charged US$170,000 on sales, the company has to pay to
the government US$95,000. In summary, this company recovered the VAT they paid on
their purchases (U S87 5,000), and generated an additional US$95,000 for the government.
This is the motivation for the government to introduce the VAT in Mozambique, that is,

generate revenues.

92

On the other hand, if the same company only sold the equivalent of US$200,000
instead of US$1,000,000 as before, the amount of VAT charged is US$34,000. In this
case, the company invested US$75,000 and generated only US$34,000. Therefore, in this
case the government has to reimburse US$41 ,000 to that company.

Maize meal is exempt from VAT, along with rice and bread. Maize grain is not
exempt. The recovery mechanism just described means that maize grain imported and
then processed into maize meal will in practice be exempt: If a maize miller imports
maize grain and pays US$75,000 in VAT, it will not charge any VAT on maize meal
sales because this product is exempt. Therefore, the government should reimburse the
US$75,000 incurred by the company during the imports. Any trader importing maize
grain for sale as grain will have no right to VAT reimbursement on the purchase, because
he or she must legally charge VAT on the grain sales.

To receive reimbursements from the government, the company has two options.
In the first option, the company may prefer to keep the amount to be available for the next
month as a credit. In the second option, the company may request the reimbursement,
something that the National Directorate of Tax and Auditing should do within three
months after the appropriate request from the company side. After this period of time, the
amount due to the company incurs interest. In practice, however, reimbursements are
reported to take at least six months and sometimes more, and payment of interest is not

assured.

93

4.3 VAT and Price of Maize Meal in Maputo

As mentioned in chapter three, due to problems regarding the quality of local
maize grain, large industrial millers rely mostly on imports of maize grain to satisfy their
necessities, either from USA. or South Afiica. To import maize grain, the millers have
to pay officially a duty of 2.5 percent over the CIF price for maize grain, plus a 0.5
percent for other small duties. Over this total value, the millers also have to pay the VAT
of 17 percent.

The price of maize meal processed by large industrial millers has shown increases
over the years. Figure 4.3a presents the real price of maize grain produced locally, maize
meal processed by local mills, as well as the price of imported rice in the Maputo retail
market. Figure 4.3b and 4.3c shows, respectively, the price of grain relative to maize
meal, and the price of maize meal relative to rice. Three patterns are apparent in these
graphs. First, real maize meal prices have risen dramatically since late 2001. Second, the
real price of maize meal relative to rice has risen dramatically from the same time. After
trading around par from 1997 through late 2001, the relative price of maize meal has risen
dramatically. Third, the price of maize grain relative to maize meal has fallen
dramatically since about the same time: In late 2001 maize meal prices rose, whereas

maize grain prices have fell slightly.

94

80> .08. 5:02

 

 

 

 

 

...
soothes «wavy 9.... age
Ax 1‘ Q Ax 4%. 0’0 A?
v». 0w. v». xv 9 \.
8.x . . . ooom
. . . . . - 88
.82 .0863. _ s»
81 <3 A. - 88
......0 8...: _ 2 < .
.l .l . \ .. _- 88 m.
.. ...... . a
. . .....- 88. u. m.
m... . H ”is 8 —. m
.. ' I “.9. 4' )
.... ...... ...... ._ 88. W
n n ._ 8...: 8
u m (\
.... - 88.
...H t 8...:
m
..88

 

 

G8" 5:538. 83...... 32...: ... 88 ...... ..8: 8...: 883. ....20 8...: .... 8...... =3... .5. an... 9...»...

80> ...... 5:2).

 

 

 

 

 

 

 

b fifinvnvnrkaQQQvamvfl ’0
oomWQwWoweoomWofloemo oowvooweoow cameos... cameos”. geeqeege«q®.q®wg® 90.0.9...
Ox
... or «v 00 «V 0.0 Q: ow. ow. 0.? 0.. .039? Q? .... 00 «V 90»? 0.0.?»
N.
.l m.
I V. d
u.
o
9
j 1 m m %
O
u m. m
.. @
...

 

 

$8. 3:578... 52...... 32...: ... .3: 3...: ... 3.2.3. ......G 3...: .... ...... 8.... =3... .5. an... 2......

30> E... 5.82

 

 

comuooweoow com com nwownwom con. cowawow coon. seem geogoveqe 99 «......x «99 99 ......v
o 4 4 o o. x. « A o 0
or v... 00 «v 90 yaw. ow. ow. oeeox. ...? 0v v... 00 u... «.0 w») 0.0+?
0:0 D LI I D I b L P Duo
2.3. 3....

.Il...| um.

V >x\// > \823/ WK) o.
<< {KL <f/x

 

 

 

d
u.
0
a
m 7
u. 9
O
)
- 3 W
x. a
(
- o.~
m...

 

GEE Pusan—rmaa— mug—5.3 859.2 ... one 3 9523— .32 03:2 6253— a: 3.3— 035 :33— ...ow— and. 9...»...—

4.4 Assessing the Impact of VAT Charged on Imported Maize Grain

To assess the impact of VAT charged on imported maize grain, import parity
price (IPP) for maize grain imported from six different locations of South Africa
(Gauteng province) were computed and compared with maize prices landed in Maputo
from the Central region of Mozambique.

To compute the IPP price, prices from the South Africa Futures Exchange
(SAFEX) were used as a simple proxy for the price in one silo. From this price, rail road
transport and insurance costs were added to obtain the Cost Insurance and Freight (CIF)
price at Ressano Garcia Border (RGB). Then, the official duty of 2.5 percent plus 0.5
percent charged to maize grain was added to the CIF price. After the duties have been
added to the CIF price, 17 percent of VAT is added to the RGB CIF plus the duties
charged.

Rail transport costs and insurance charged by Mozambican Rail Road Company
(CF M) to transport maize grain from RGB to Maputo are then added, as well as the VAT
charged for any internal transport transaction. The result is the IPP in Maputo.

A simple scheme of this calculation is shown below in Figure 4.4a, and IPP
results for maize grain from all six origins in Maputo (railroad) during June 2004 are
presented in Table 4.4a. VAT is included in these calculations because informal traders
do not recover the VAT. Therefore, including VAT on IPP calculation gives the price at

which maize grain will land in the market if imported by the informal sector.

98

Figure 4.4a Scheme for Import Parity Price Calculation

 

I SAFEX Price . I

| Add Transport cost and

 

insurance to RC. border

I Equal to CIF price I

Add 3% of duties I

Add 17% of VATover
CIF plus duties

Equal to CIF price plus duties
plus VAT

Add Transport cost and
insurance to Maputo

l Add 17% of VAT over

 

 

 

.—.

 

 

 

 

 

 

transport cost and insurance

[Equal to IPP in Maputo I

 

 

 

99

 

.8232 ...

89m .3...: 8.... 8...: 2.88....»

8.... SS. 8...... 8m... 8...... 5.... 8...: 38.: ... ....
38.8.... 88.8.... 8.8.... «8.8.... ...... . m... 85...... 8...: .38.: ... ....
.....mm... 28.. 8...... 38.. ...... 3.8.. 8...... 98.: .. ....
...m ...m ...m ...m ...m ...m 8...... .8 .o .8555
8.8 8.8 8.8 8.8 8.8 8.8 8...... .68 .8. :88...
.....m. ...... 8.8m. ...Nmm. ..8: .....8. 8b.... .55 + 8n. + ......
N... 8...... N8. ...: ...m: 8...: 8:3. 8.... + ...... ... $2.55
3.2. 38. 5.. .. 5.8. 38. 2.8. 8:3. 8.5 + ....u
...... ...... 3m .8 3. 8.8 8b.... .....o... ...... ......
.. .8. $8. .8... ...mm... 9.8 3.8 8.5... ......
.. .... 8.8. .8. ...mm. 98 ...8 8,3... ...8 28.8..
.38 .....8 .....8 .....8 .....8 .....8 88... . .8... mem

 

.15 93.32 ...—:2 33— 5853;. segues—.8: ..om 2:85—

 

Aouigum «5.5.0. 3...: 5.5m ... ...wto ue «3.5 .15 E8— .80

 

25m 2.3. £5— »: 32:. 55.5 .... 553:1— »:ouobmfl Mum :8... 68.555 59.0 3?: .....— 33—«2 ... A:— avé 033—.

100

Notes:

‘ SAF EX prices in July 2004.

2 Transport cost from the departing point to Mozambican Border (Ressano Garcia),
charged by Spoornet. .

3 Transport cost from Mozambican Border (Ressano Garcia) to Maputo City (88 Km),
charged by CFM.

‘ Exchange rate Rand to Metical on August, 2004: 3,450 Mt.

5 Average price of maize grain landed in Maputo from the Central region in June 2004.

An IPP time series was also computed based on a silo at Ermelo. During the
authors’ interviews in Johannesburg, South Afi'ica, with traders involved on maize trade,
Ermelo was cited as a logical spot from which to export to Mozambique. Also, it was
emphasized that a good proxy for cash price at this silo is SAFEX price minus a
difference representing the transport cost between this silo and SAFEX, plus 30 Rand per
Ton. The transport cost differential is computed by SAFEX once a year based on a
weighted road transport cost (60%) and rails (40%). Therefore, in this analysis is
assumed that the maize grain is purchased at one silo located at Errnelo that is located 312
Km aside from RGB, at a cash price of SAFEX minus transport cost differential plus 30
Rand per Ton.

SAFEX price minus transport differential plus 30 Rand creates a floor on cash
prices in a region, and if a silo offers a price below this floor, it is better for producer to
deliver to SAFEX Silo because they will receive a better price. If the market is at all
efficient, that is, the participants have the correct information and know how to deliver to

SAFEX, then the market price should not go under that level.

101

We calculate two scenarios, imports with and without VAT. In both scenarios,
the IPP series are graphed together with the prevailing wholesale maize price in Maputo,
which largely reflects landed cost from the Central region ‘of Mozambique.

In Figure 4.4b (with VAT), the evidence shows that first, with VAT charged, it
was usually not profitable to import (in 48 percent of the weeks), but second, there were
some extended periods (February to November 2000; September 2001 to January 2002;
February 2003 to November 2003; March 2004 to Apr 2004; September 2004 to January

2005) when it was still profitable, but trade did not occur.

Figure 4.4b Price of Maize Grain From Central Region Versus IPP; With VAT

IPP with VAT, White Maize

 

 

 

 

0000
0
7000 - h“ ,
A
00 ' \
§ 6000 - l | ’ '
5 A '1". ‘ I
,8 5000‘ t ' ‘1 ‘I ’ '
DE ¢.\ . I'\ I | ’-
§ 4000 - J , ’ I \l
3000 5; Price in Maputo
2000 ' . , , . r . . . . . , . . . . Real IPP
(9 v
441,. 6‘ fl?;:&j¢?;:%:¢49;:‘bfirfipjeojefkfefb
7 .9 0 o 0 0 0 0 0 0
90,9 9,9 00 000 00 07 00; 07 09 009 009 05, 000 000 q, 007 007 00‘

Month and Year

Railroad

102

On the other hand, Figure 4.4c which shows the IPP from the second scenario
(without VAT), the evidence shows that it would have frequently been profitable to bring
in maize grain if no VAT were charged. In other words, IPP without VAT was smaller

than price in Maputo from Central Region in 66 percent of the weeks.

Figure 4.4c Price of Maize Grain From Central Region Versus IPP; No VAT

IPP without VAT, White Maize

 

8000
7000 « “A“
6000 -
5000 -

4000 -

Real Price (Mt/Kg)

 

3000 - Price in Maputo

Real IPP

 

2000

 

I

 

487363094!49+06‘A‘Q’1’4’Y1—‘9‘A‘4’v4‘71—06‘A‘4’444713963944933‘34'4/

z 79099 0‘99 09990999 9""
‘99 90,9 00 000 000 07 00; 007 09 00-? 00v" 00 000 000%" 007 007003

Month and Year

Railroad

103

In summary, although elimination of VAT would have made it more profitable to
import, profit opportunities even with VAT in place were foregone — traders never
imported even when there was a potential profit.

If maize imports from South Africa done by local traders other than millers
would be doable, the competition in the maize subsector would thus increase. The first
consequence of this increase on maize competition would be a decrease on maize prices
in the local markets, especially in Maputo, as a result of increases on the supply side. The

consumers would then be better off with an increase on the consumers’ surplus.

4.5 Regulation to Import Maize Grain in Mozambique

Act number 206/98 signed by the Ministry of Planning and Finance of
Mozambique on November 25‘“, 1998 establishes the main guidelines for importing
maize grain into Mozambique.

Under this regulation, the importer first of all has to fill and present a document
called pre-declaration to the customs, together with the following documents:
1. The importer’s identification card.
2.Quotation showing:

a. Name, complete address, country, phone, and fax number of the supplier.

b. Name and complete address of the importer.

c. Date of the emission of the quotation.

d. Country of origin of the commodity.

e. Boarding port.

104

f. Landing port.

g. Complete designation of the commodity, including technical specification.

h. Quantities imported, total weight, volume or metric units.

i. FOB price of each imported commodity in the currency shown in the quotation.
j. FOB price of the total commodity in the currency shown in the quotation.

k. Value of the total freight in the currency shown in the quotation.

1. Value of the insurance in the currency shown in the quotation.

m. Time limit of delivery.

n. Expiry date of the quotation.

0. Form of payment.

If the value of freight and/or insurance is not available at the time of the pre-
declaration filling, a premium of 12 percent over the FOB value (10% for freight and 2%
for insurance) must be used to calculate the CIF price.

The pre-declaration must be presented to the customs before the commodity
leaves the country of origin. It takes one day for customs to process and return it to the
importer. If everything is in order, the pre-declaration is accepted, and the importer has to
pay a deposit amount within 10 business days.

When the commodity reaches the port of entry, the importer must complete a
declaration form called Dacumento Unico (DU). The DU must be accompanied by the
pre-declaration properly certified by the customs, plus a document proving that the

commodity belongs to the importer, as well as the final receipt. If all is in order, the

105

importer pays the amount that corresponds to the official duty charged for that
commodity, plus the VAT if that commodity is not exempt.

It is clear that, under the new regulation, it will be difficult to an informal trader to
import maize grain from South Africa. First, they have to have an importer card, which is
not easy to obtain for an informal trader. Second, to get a quotation fi'om the maize
supplier, an informal trader would need some detailed knowledge of how the South
Africa maize trade is done. We’ll return to this issue in the next section.

Act 206/98 provides for a simplified system to import any good. Any importer
with goods whose FOB value does not exceed US$500, and who has not imported during
the previous 30 days, may have a simplified dispatch emitted at the port of entry. In this
case, the importer does not have to complete the pre declaration. This simplified system
to import seems to be a good option only for the small traders, since with US$500 it is
possible to import from South Africa only three tons of maize grain, equivalent to forty-
three bags of 70-Kg each. Our data show that 40 percent of the traders involved in maize
trade between the Center and South of Mozambique trade at least this much per month.
Thus, current regulations on imports would seem to create significant barriers for
informal traders wishing to engage in maize trade between the two countries in anything

other than very small scale.

4.6 The South African Maize Marketing System
Before 1995, maize marketing in South Africa was done under a one-channel

market run by the government through a maize board. Maize sales and purchases had to

106

be made through this board, which also was responsible for providing market information
and other services.

Since 1995, maize producers are no longer forced to sell to the maize board. The
risk faced previously by the maize board shifted to maize producers. To sell their maize
grain, producers in South Africa have several options. For instance, they can sell to the
nearest silo, which is now a private company, though some are cooperatively owned. On-
farm storage is not widespread because before, maize price was determined by the maize
board, eliminating therefore any reason to store with the purpose of receiving better
prices. Another option that a maize producer has is to sell to any buyer that directly
approaches them.

At the time that we were in South Africa trying to learn and understand more
about the marketing system there, their high level of formality came up. Unlike what is
observable in Mozambique, where all maize transaction is done in spot markets, in South
Africa it is almost impossible to buy maize grain in wholesale outlets. In fact, we found
no evidence of any source of maize besides that in silos that informal traders from
Mozambique could access for import. The high level of formality reduces the probability
of informal traders of Mozambique being well succeed while buying maize grain there.
For instance, besides the language barriers, no maize grain will be loaded for
Mozambique until the importer presents a letter of credit issued by a reputable South
Afiican Bank. This bank will only issue the letter after the importer proves that has

capital.

107

4.7 Incentives to the Milling Industry
The government has shown some flexibility on light of milling industry
complaints regarding the impact of duties and VAT on their imports of maize grain. In
August 2003, the government approved a special act for the industry with the main
objective of creating more employment, and reducing the price of processed products to
consumers. The act says that every processing industry‘1 will not pay any duty for
imported raw material if and only if:
1. They have annual gross revenue of at least six billion Meticais.
2. The value added to the final product is at least 20 percent. This valued added was
defined as the valued of final products divided by (remunerations to the employers

plus financial costs plus taxes plus amortization, plus results of the exploration).

However, at the time of the interviews conducted in this study, none of the millers
interviewed had already got any benefit from this act, apparently due to heavy

bureaucracy.

4.8 Chapter Summary
The main finding in this chapter is that, while the VAT clearly made maize grain

imports less profitable, imports did not occur even when such trade would still have been

 

4

Industries included in the amendment include the agricultural processors, textiles and
shoes, graphic industry, mechanic industry, as well as chemistry, plastic, and rubber
industry.

108

profitable. This suggests that, while the unequal application of the VAT (due to the VAT
exemption on maize meal, which results in an exemption on grain imports for millers but
not traders) is one relevant issue, it is not the only reason for the poor state of trade in
maize grain between Mozambique and South Africa. Indeed, the regulation to import
maize grain is very heavy for an informal trader, contributing therefore for the poor state
of trade between these two countries. Also, the high level of formality within the South
Africa system contributes for this situation.

However, this topic does not end here. Further research should be done to explore
other options for this trade. For instance, use of the road rather than rail transportation

should be investigated.

109

CHAPTER FIVE
ANALYSIS OF VERTICAL AND SPATIAL PRICE RELATIONSHIP AMONG

GRAIN MARKETS IN MOZAMBIQUE

5.1 . Introduction

In this chapter, time series models will be used to evaluate the performance of
spatially differentiated maize markets and vertical maize marketing levels in
Mozambique. The evaluation contributes to the understanding of the functioning of
maize markets in Mozambique by testing for spatial market efficiency among four maize
grain markets (Maputo, Xai Xai, Chimoio, and Beira), and two marketing levels
(wholesale and retailer).

The chapter is organized as follows. The idea of market performance within the
industrial organization framework is presented in section two. In section three, the
concept of market integration and marketing margins and their importance as measures of
market performance is discussed. In section four, the conceptual framework to test
market integration is presented. In section five, the scope of the study is presented, and in
section six, as a preparation to the econometric tests, a preliminary data analysis is used to
assess the efficiency of the vertical and spatial integration market. Brandow (1976)
mentioned that systems analyses are required to identify the ways of moving foods from
the raw material to consumers’ bags at the least possible cost. Efficiency in this sense
requires the use of modern technology, achievement of economies of scale, the effective

use of transportation facilities, and so on. In the study of maize market performance in

110

Mozambique, operational efficiency is concerned whether maize grain is moved from
producer to consumer at the lowest possible marketing cost. In this preliminary analysis,
food price evolution are analyzed, as well as price volatility and spreads. In section
seven, the econometric tests for market integration are presented, and last, the chapter

summary is presented in section eight.

5.2 Market Performance

Marketing performance is defined as how well the marketing system performs
what society and the market participants expect of it. It is the outcome that results from
the group of firms in an industry pursuing their respective lines of conduct.

Three steps are involved in evaluating market performance. The first step is
defining the performance dimensions, that is, what types of outcomes are most important
for the society. In the second step, it is important to define the perfOrmance norms, that
is, what is the standard of performance against which performance can be measured. The
last step is defining performance measures, that is, what metric or criteria will be used to
measure actual performance against the norms.

Price coordination requires that price signal from the food industry to upstream
producers or downstream consumers be appropriate for guiding their production and
consumption decisions. One aspect of price coordination is how closely producer and
consumer prices move together, as well as price differences between points in time.
Hence, to analyze how the maize markets in Mozambique are performing, two

performance measures will be used, namely the degree of integration between markets

111

and levels of marketing, and marketing margins.

5.3 Indicators of Market Performance

5.3.1 Market Integration

The role of markets in assuring economic welfare and food security is well
known. Deficient functioning of the marketing system is one of many reasons for poor
levels of production and productivity, low earnings and food insecurity. Therefore, the
contribution of agriculture to the economic growth of Mozambique, reducing income
inequalities, and poverty, depends in part on the performance of agricultural markets.

These markets in developing countries frequently suffer from many structural
deficiencies which leads to spatial market inefficiency, that is, inefficient flows of
information and trade among spatially distinct markets. Hence, one of the indicators
commonly used in the analyses of market performance is the level cf market integration.

In a competitive environment, arbitrage will ensure that prices of a homogeneous
product such as refined maize meal, will be linked in two different markets. Hence, two
markets are said to be integrated if prices in the importing market equal the prices in the
exporting market plus marketing costs (Tomek and Robinson, 1990). This definition
implies that, first, there is some trade between markets, and second, the price differentials
between them cannot exceed the marketing costs necessary to move the product from one
market to another.

Market integration can be seen in three dimensions, namely, spatial market

integration (location), vertical market integration (product form), and temporal market

112

integration (time). The first case reflects the effect of a price change in one market
location on the price of the same commodity in another market location. If there is no
linkage between two market prices, then markets are said to be separated.

Vertical market integration reflects the passage of a price change across steps in
the marketing chain. A price relationship between raw and processed products is a good
example of this vertical integration. In this case, the movement of a product is combined
with some form of processing such as in the case of maize grain as a raw product and
maize meal as a processed product. Another example of this vertical integration is a
movement of a product from one level to another level without changing its form. For
example, a movement of maize grain from wholesale to retail is a good example of this
kind of integration.

Last, temporal market integration reflects the effects of a present price change on
future prices. This kind of integration will not be analyzed in this study because it is out

of the purpose of the study.

5.3.1.1 Importance of Market Integration

In the context of Mozambique, there are several reasons to analyze the
performance of maize markets. First, price is a product of market performance. If a
shock occurs in a market, it is expected that price will adjust in other markets to reflect
the changing conditions imposed by that shock. The change in price is a signal that then
facilitates market adjustment of quantities, etc. For instance, if there is a maize shortage

in the Southern region of Mozambique, well-integrated markets will quickly reflect this

113

shortage through appropriate price relationship, creating signals for the imports of maize
grain to Maputo and other Southern markets.

When spatially markets are not integrated, price signals among markets will be
transmitted imperfectly and with delays. One of the main consequences of this poor price
transmission is high price volatility that weakens the food security of both farmers and
consumers.

Second, in poor countries such as Mozambique, market interventions are
constrained by lack of financial resources. Knowing that markets are integrated and will
therefore efficiently transmit information and guide trade flows between surplus and
deficit areas can make it easier for governments to allow markets to work, while
concentrating their scarce resources on investments that will reduce marketing costs or
target needy households in a way that does not disrupt markets.

Third, as stated by Goletti et al (1995), knowing the relationship among spatial
market prices makes forecasting analysis more doable. For instance, knowing the
direction of price signals between integrated markets, enables prediction changes in food
security among farmers and consumers in one market as a result of changes in another
market.

In summary, prices in different markets are important in the decision on where to
buy and sell. In other words, “regulates” trade flows. Therefore, the arbitrage activity of
traders connects spatially separated markets, and market integration analysis provides a
better understanding of the dynamic interaction of prices and the degree by which

physically separated markets are connected.

114

5.3.2 Marketing Margins

Another performance measure that will be used in this study is marketing margins.
A marketing margin is the price spread between the price paid for a commodity and the
price at which it is sold (Tomek and Robinson, 1990). It represents the value of the
marketing services (transport, storage, processing, and others) that are performed on a
commodity, including a normal profit. Therefore, the margin includes costs incurred by
traders plus a mark-up that is added by the trader as a return to their investment,
management, and risk.

Ideally, an efficient marketing system should provide the right good, at the right
place and time, in the right quantity, and at an affordable price. In addition, profit levels
should be reasonable for all participants in the chain. Hence, the analysis of marketing
margins provides useful insights on how marketing costs have changed through time, and
whether they are reasonable or not.

Stable margins are consistent with efficiency in the market, but non constant
margins per se do not reflect lack of efficiency in the market. Brandow (1976) mentioned
that a single industry should not be expected to have stable prices, output, or employment
when changes in the economy impose instability upon the industry. This is true because
marketing costs do not remain the same throughout the time, and therefore, neither can
the final price. Thus, increasing costs in the maize marketing system increase marketing
margins, and changes on it may affect market participants in different ways, depending on

their respective elasticities of supply and demand.

115

5.4 Analytical Framework for Analysis of Market Performance
5.4.1 Marketing Margins
We focus on the retail-wholesale margin in this study, defined as the simple

difference between the retail and wholesale prices of maize grain in a single market.

PMR“ = (PRit - PW“) / PWit (5.4.1a)
where PMRit is the margin gained by retailer over the wholesaler at market i at time t, and
PR“ and PWit are retail and wholesale price of maize grain at market i at time t,

respectively.

Using PMR“, a stability test is performed to determine statistically the magnitude

of retailers’ margin over time. This test is performed running the following regression:

PMRit = or +01“ + p, (5.4.1b)
where T is the time trend,
or and B are the coefficients to be determined, and

u is the error term.

[3 coefficient should be interpreted as follow: One percent increase in T increases
PMR by [3 units.
In addition, marketing costs budgets collected in July 2004 will be presented with

a structure of all costs and net margins incurred by informal maize traders in Maputo.

116

This analysis will be useful to analyze the competitiveness of maize grain transportation

from the Central to the Southern region.

5.4.2 Unit Root Test and Cointegration Analysis

In this study, time series price models will be used to evaluate the performance of
spatially differentiated maize markets and vertical maize marketing levels in
Mozambique. More specifically, a cointegration model is the base of analysis.

Most of economic time series such as prices, behave in a nonstationary manner,
that is, the mean is not constant and the variance is time dependent (Enders, 1995). A
time series is said to be stationary if its mean fluctuates around a constant long-run mean
and the variance is finite.

The assumption of stationarity for modeling is important because, when data
mean and variance are not constant, it means that observations come from different
distribution over time, which causes statistical problems.

If a time series is non stationary, in econometric language, it is said that the series
suffers from a unit root problem. In this case, the standard regression models such as
Ordinary Least Square (OLS), cannot be used to perform regression analysis because
under the presence of a unit root, statistical inferences are biased and inefficient (Pyndick
and Rubinfeld, 1997). Thus, alternative models should be used, one of them being the
cointegration models.

The idea of cointegration is that although each of two or more price series may be

individually nonstationary, a linear combination of them may have the trend terms

117

mutually cancel out so that it becomes stationary (Intriligator et a1, 1996). Basically, the
cointegration analysis identifies whether or not there exists a long-run equilibrium
between two or more price series. Therefore, the presence of cointegration is indicative
of interdependence, and its absence indicates market segmentation (Golleti and Tsigas,
1995)

The attractiveness of cointegration analysis is that it does not require that price
series be transformed into stationary series. Furthermore, there is no need to check for
autocorrelation and heteroskedasticity because OLS is consistent (Tomek and Myers,
1993)

To test for market integration using a cointegration approach, the Engle—Granger
methodology is applied. This methodology is composed of two steps. In the first step,
price series must be tested for order of integration using the Dickey-Fuller Test (DF) or
Augmented Dickey Fuller Test (ADF). The Dickey-Fuller test, which tests for the

presence or not of a unit root, is performed by running the regression model specified as:

Pt = [3, PM + e, (5.4.3a)
where Pt is the time series price for maize grain at time t,

PM is the lagged time series price,

[5, is the coefficient on the lagged term, and

r:t is the error term.

Subtracting PM from both sides of this equation gives:

118

Pr ' PM = Bl PM ’ PM + 8r = AP,=(Bl-1)PH+ 8! =’ APr = 6PM + 8! (5-4-3b)
where AP, is the price first difference (P, - PM), and

6 is equal to (B1 - 1).

The DF Test can also be used for testing a unit root in the presence of a drift

and/or trend. In the presence of drift and no trend, the regression model is specified as:

AP, = p, + 6PM + e, (5.4.3c)

where B0 is the intercept term (drift).

On the other hand, when trend term is included in model 5.4.3c to make sure that
the apparent lack of stationarity is not due to the presence of a deterministic trend

(Bopape, 2002), the regression model is specified as:

AP, = [so + m + 6PM + e, (5.4.3a)

where BI is the coefficient of the trend term, T.

In these two last cases as well as in the first case (without drift and trend), the
coefficient of interest is 5. If 6 is equal to 0, the price series is nonstationary. In other
words, the price series contains a unit root.

The number of times that is necessary to differentiate each price series to turn it

stationary gives the order of integration of the series. For instance, if a price series is

119

differentiated once to turn it stationary, this series is integrated of order 1 and is
symbolized as 1(1).

In the second step of Bugle-Granger methodology, the long-run relationship
between markets is determined. Based on the idea of cointegration, if two series 1(1) can
be combined and the series resulted from this combination are 1(0), then the series are
cointegrated and exhibit a long-run relationship. If P’, denote the price at market j at time
t, and P", denote the price in market k at time t, the coefficient B, in the regression model
5.4.3c below gives the long-run relationship between these two markets only if the error

term (8,) in the same regression model is stationary.

P1,= a, + B, P“, + e, (5.4.3e)

More simple, in the second step, the OLS residual from equation 5.4.3c is taken
and then tested for a unit root using the DF test. The procedure is the same as testing for
a unit root in a single series. The only difference is that the distribution of the test
statistic is different because now the residual is used instead of the observable variable

(Wooldridge, 1999).

5.4.3 Granger Causality
In order to assess the nature of maize price transmission across markets (and
through different marketing levels), the Granger causality test is used. Basically, Granger

Causality Test is another approach to test market integration. A time series prices P", is

120

said to “Granger cause”another time series price P’, if current and lagged values of P",
improve prediction 0ij, (Gujarati, 2003). In other simple words, causality is basically a
measure of the predictability of prices, that is, price movements in one market can be
used to forecast price changes in other markets (Minten and Mendonza, 1998).

The Granger test involves the estimation of the following regressions

(Gujarati, 2003):
Pi! = 61 + Xai PM + EB, Pkt-i + 8n (5433)
Pk: = 62 + ZYi PM + 261 Pkt—i + 82: (5-4-3b)

where 6, and 62 are the intercept terms,
a, , B, , y, , 6, are the coefficients on the lagged values of both P’, and P", ,
i is the lag length used for both price series, and

8,, and 82, are the error terms.

Causation can occur in two ways, unidirectional- where shocks in one market
affect another market but not the reverse — and bidirectionally where shocks in one
individual market are transmitted both ways. Therefore, based on models 5.4.3a and
5.4.3b, four hypotheses of causality can be tested. They are:

1. Unidirectional causality from P“, to P, if the coefficients B, are statistically different
from zero and the coefficients y, are not statistically different from zero.

2. Unidirectional causality from P’, to P", if coefficients B, are not statistically different

121

from zero and the coefficients y, are statistically different from zero

3. Bilateral causality (both P’, and P", cause each other) if all coefficients a, , B, , y, , and
6, are statistically different from zero.

4. Independent causality (both P’, and P", do not cause each other) if all coefficients (1,, B,,

7,, and 6, are not statistically different from zero.

To test these hypotheses of Granger causality, a F -test is used. The test is
conducted first regressing P’, on their own lagged values. This is called the restricted
regression. Second, P1, is regressed on their own lagged values and the lagged values of
P", . This is called the unrestricted regression. Then, the F -test that follow the F

distribution with m and n-k degrees of freedom is computed as:

F = [(SQR, - SQRW) / m] / [SQRW / (II-10] (5430)
where SQR is the sum square of the regression,

r stands for restricted,

ur for unrestricted,

m is the number of lagged values in P“, , and

k is the number of parameters estimated on the unrestricted regression.

The number of lags to include in the models is determined by Akaike Information
Criterion, AIC. The AIC model states that different lag length is used to compute the

AIC value, and the lag yielding the smallest value of AIC is the one chosen.

122

5.4.4 Dynamic Adjustments

Cointegration analysis is very important in understanding if a long-run
relationship exists between two markets. However, the analysis does not answer two
important questions. First, if there is a shock in one market, how much of that shock will
be transmitted to the other market? Second, how long does it take for a price shock in
one market to be transmitted to another market?

The magnitude of the influence between markets and the speed of adjustment to
shocks can be measured to determine the strength (or weakness) of the market linkage,
and the rapidity (or sluggishness) by which markets responds and adjust to shocks.
Where markets are perfectly integrated, price changes in one market are exactly translated
as price changes in another market. In this case, the long-run multiplier or the
contemporaneous price effect that measures the extent of the market linkage will be equal
to unitary, while market isolation is indicated by contemporaneous price effects equal to
zero (Bopape, 2002; Mendonza and Rosegrant, 1995; Minten and Mendonza, 1998).

An autoregressive model representing a marketing system is used to study the
dynamics of price transmission. This autoregressive model is specified as (Mendoza and

Rosegrant, 1995):

AP], = 6, + 20c, AP),k + [[3, AP“,_, + 8, (5.4.4)

where AP’, is the contemporaneous price changes in market j,

AP", is the contemporaneous price changes in market k, j not equal to k,

123

subscripts t-k and H are the historical price changes in each market, and

a, and B, are the parameters to be estimated.

Price changes in equation 5.4.4 are calculated by obtaining the first difference of
the price logarithms. The use of first price differences is desirable because the problem
of autocorrelation is eliminated (Mendoza and Rosegrant, 1995), while the uses of prices
in logarithms form can be interpreted as the intertemporal changes in prices in one market
in response to price changes in another market (Mendoza and Rosegrant, 1995;

Goletti and Tsigas, 1995).

5.5 Markets, Products and Price Data
. In chapter two and three, central Mozambique was identified as the surplus region
while the southern region was identified as deficit.

We postulated that maize price formation is centered around Maputo City which
is by far the largest in terms of population size and maize demand, and is the location of
the largest maize millers and feed producers.

For the analysis of market integration among spatially differentiated maize
markets and vertically integrated marketing levels, four markets were chosen, namely
Chimoio and Beira in the Central region, and Xai Xai and Maputo in the Southern region.
Chimoio is chosen because of its importance as a supply source for Beira and for the

Southern regions. Beira is supplied by surplus from the Central provinces of Sofala and

124

Manica, and is the main consumption center in the Central region. Xai Xai is the second
consumer center in the Southern region, and also an important center of distribution for
those districts located in the interior, not only for maize but for other food products.

All four markets have a long and continuous price series available at retail and
wholesale level. The data consist in weekly retail and wholesale maize prices collected
from October 1998 to June 2004 by SIMA. SIMA collects the producer, wholesale, and
retail prices of several commodities, including maize grain and its derivatives, tubers such
cassava, and commercial commodities such as sugar, at 22 different markets scattered
throughout the country. Prices are collected every Monday, and after verification and
validation, the price information is released to SIMA users every Wednesday afternoon.

All prices used in this analysis were deflated by the consrnner price index for

Maputo (CPI, 1998 = 100).

5.6 Results of Descriptive Statistics

5.6.1 Price Pattern at Retail Level

Tomek and Robinson (1990) state that graphical methods provide a simple means
of identifying trends, cycles or seasonal patterns of behavior in time series data. Figure
5.6.1a to 5.6.1d depicts the real monthly retail and wholesale prices for maize in Maputo,
Xai Xai, Beira, and Chimoio, respectively. All price series exhibit typical annual
variability very common for agricultural commodities. At the harvest period, prices tend
to drop due to high level of supply and then rise over the remainder of the season because

of shortages on supply.

125

80> Ea 5:02

 

h5$6$¢¢®-x900999$
octMotMoOootNotW0affirms/xoctNotWOsotNotWWfiWMac/x
os/wxoo osowxoo osowxoo osowxoo 03056.00 03056.00
:53. 1,.232:
29.205.5' II
‘ k
-
\ § .r \
. 5 - L ~
I! .1
-—~. .11 a. h A“
a £4 a . n
v 3 n (w
‘ 1 — —
‘ — -
...I

 

 

 

8:32 5 £80 382 we mouth

Vgfl 0:55 I ”g.— hOn-Qwoo Ava-uh“: £9,01— =50M~ 1H! 0—88—6‘3 a“ Huge 0362 H5 00th —60~— Bfio¢om Ohflwmrfl

ooow
ooom

coon

ooov

ooom

0000

0005

Doom

Doom

(SN/IN) 9011.1 [998

126

v u m. 6 no av W «W nv nv nv N N
\v 6 Q V 9 ,« xv Q «
ox. ®.®..oe.ov9.os ®.o~<.,wo.oo.ov.o>.Qv.,wx.®o.oo.ov
:33.
28225“ I
~ i
.... I , \ 4 ~ I
K I I . \x! w r s.
/\ f \ N
4 ’ .I ~
\
o ’ ~
:

60> was 5:02

 

 

 

 

ex ex a saw 0%: mo 825

coca

oooN

ooom

ooov

coon

coco

coon

(3mm) 9011.1 1893

127

25a 05:. . «aa— aoacuoO Aux max "—95:— =30¢ Gnu 0.36.25» 3 EEG 0532 we 08.5 13¢ n—dd 0.53%

30> fig £602

 

b h m. ..a 6 av av W N N N 0 Q Q Q. Q. o. %
can, com cow 99% 90¢ 99¢ 90% com com com 90% com 09% 09% aqx 90V 99 gov
N N 9 N 9 N 9 N 9 N
eosnmwx ooeosfimwx 00 $03 ow 00$. ow 00$» ow 00 Nos ow 00
=53. . . _ _ . . . . . b _ r L . . . 82
23225" I x. _ fl .. OOON
, h
p . ~I . I 82 m
I. ~ ~ m
s N i « Ix - 83 M. m
‘ I m 1
< , ~ . 8% m
. - 88 am
a u (x
— .
v. 1 coon
. . coo»

 

 

 

85m 5 580 as: ,8 $25

3.3 2:... - :2 .335 5.5 “.23 :33. ...: 2322.3 .a 529 332 ... 3.5 .8: 3.3 9...»:

30> 23 5:02

 

u u m. 5 6 av nv nv N N N 0 Q Q a. 9 o. %
0N. .9vi 00 0) ow! 00.03 INK 00 03.0W 00 Ox. [$.00 0N. IN<.OO
=Som coo—
282055 ‘. fir. — u OOON
II t . \\. ~ ; - 8%
~ ~ —_ 1 - mm
~ ~ . 7N 88 m.
. )1 C \ h v M. m
.K . . - 8% N
W
4 ‘ - 88
w
I r 83 (
=
A . coow
coca

 

 

 

22:20 E £80 362 .3 335

can 05:. .. mag 52.390 63530 £951— :301 05. 0139—259 «a £20 03:: u: 3m...— _ao~— e—éfi 95w:—

The highest real retail price for maize grain across markets was observed in
Maputo with a mean of 5,483 Mt per Kg (Table 5.6.1). In addition, a visual inspection
suggests that prices in all retail markets seem to be somewhat volatile, with market prices
in the Southern region (Maputo and Xai Xai) showing less volatility than market prices
in the Central region (Beira and Chimoio). This situation is perhaps explained by a more
stable maize supply, since the Southern markets are supplied by districts from the Central

region, from their own region, and also from imports.

Table 5.6.1 Descriptive Statistics for Maize Grain at Retail Level (Real Prices)

 

Market and Mean Price Standard Maximum Minimum CV

Observation (Mt/Kg) Deviation Price Price (percent)
Maputo (N=300) 5,483 1,359 9,775 2,312 25
XaiXai (N=226) 4,215 1,071 9,353 2,379 25
Beira (N=257) 3,635 1,366 8,271 1,500 38
Chimoio (N=292) 3,754 1,335 9,018 1,674 36

 

Data Source: SIMA

The results from Table 5.6.1 also show that the real retail prices in Beira and
Chimoio seem to be very similar between each other, with almost the same mean and
coefficient of variation. This may be an indication of integration between them.
However, in some periods such as the last months of 2003, prices in Chimoio were higher
than those in Beira as shown in Figure 5.6.10 and 5.6.1d, which may reflect the absence
of trade between these markets. As mention in chapter 3, more recently the Beira market

has been supplied by districts other than those from the province of Manica as in the past,

130

a fact that may explain lower prices in Beira as a result of low marketing costs (transport

costs).

5.6.2 Price Pattern at Wholesale Level

At the informal wholesale level, the highest real wholesale price for maize grain
across markets was observed in Maputo with a mean of 4,969 Mt per Kg (Table 5.6.2).
Like in the case of retail markets, prices in all wholesale markets also seem to be volatile,

with Maputo and Xai Xai prices showing less volatility than prices in Beira and Chimoio.

Table 5.6.2 Descriptive Statistics for Maize Grain at Wholesale Level

 

 

(Real Prices)
Markets Mean Price Standard Maximum Minimum CV
(Observations) (Mt/Kg) Deviation Price Price (percent)
Maputo (N=299) 4969 1048 7551 2887 21
Xai Xai (N=113) 4360 734 6034 2631 17
Beira (N=250) 3335 1277 7463 1381 38
Chimoio (N=289) 3206 1232 7264 871 38

 

Data Source: SIMA

As shown in Figure 5.6.1c and 5.6.1d above, wholesale prices in Chimoio and
Beira move together, with almost no difference between them. In addition, prices in
Maputo follow the same pattern as in Chimoio and Beira, that is, when prices in these
two markets increase, prices in Maputo also show increases, with the difference between

them representing the marketing costs.

131

5.7 Results of Marketing Margins Analysis

The maize grain price spreads between retailer and wholesaler were computed for
individual markets, and the results are reported in Table 5.7a. Surprisingly, Chimoio has
the highest price spread. This spread in Chimoio was expected to be low because, as
shown in chapter 3, most of the maize grain sold at the wholesaler in Chimoio goes to the
local retail market, implying therefore few marketing costs to transfer maize grain from
wholesaler to retailer. As a result of high retail price spread in this market, the wholesale
share of retail price is low (86 percent), and therefore, the retail spread as of retail price is
much higher than in other markets (14 percent). One potential explanation for this result
may be the risks associated with moving maize grain from wholesale to retail level,
particularly in Chimoio that is a surplus area.

In Xai Xai, the retail price spread is negative, which means that probably the
wholesale market drifts apart of retail level. In other words, the retailers may have been

supplied with maize grain from other marketing channel rather than the local wholesaler.

Table 5.7a Real Monthly Average Price of Maize Grain and Price Sireads
Market Price Level (Mt/Kg) Retail Spread Wholesale Retail Spread

 

 

(Mt/Kg) Share of as percent of
Wholesale Retail Retail Price Retail Price
Maputo 4,969 5,483 514 92.2 7.8
Xai Xai 4,360 4,215 -145 103.4 -3.4
Beira 3,334 3,635 306 92.2 7 .8
Chimoio 3,205 3,754 548 85.7 14.3

 

Data Source: SIMA

132

The results of the stability test for the magnitude of the retail margin over time
reveals that the retail margin as a percentage of the wholesale price is increasing
significantly in Maputo and Chimoio, while in Xai Xai they are decreasing significantly.
In Beira, the margin is also increasing, but the effect is not statistically significant

(Table 5.7b).

Table 5.7b Stability Test of the Retail Margin (Mt/Kg) Over Time

 

Retail Market Constant Coefficient t-statistic p-value
Maputo -O.l3 0.0012 3.834 0000*"
Xai Xai 0.55 -0.0024 -1 1.783 0000*"
Beira 0.06 0.0002 0.553 0.581

Chimoio -0.1 0.0016 2.485 0.014“ *"'

 

**" Significant at 1 percent level of significance.

Table 5.7c presents the results of the net margins calculated for informal traders in
Maputo. Marketing costs were collected during the month of July (2004) from single
interviews with informal traders, and thus they reflect the costs and earnings from the last

purchase. All maize grain at the time of the interviews was coming from Nhamatanda.

133

Table 5.7c

Parameters

Net Margins for Informal Wholesaler in the Maputo Market
Trader 1 Trader 2 Trader 3 Trader 4 Trader 5

 

Sales location Maputo

Purchase location

Nmnber of bags purchased 100

Number of days in 18
purchase location ‘

Cost Item (Mt/bag)

Transport of trader to 4,000
purchase location

Living expenses in 3,060
purchase location

Purchase price 192,000
Transport of trader back to 0
selling location

Loading, transport, 45,500
unloading of bags

Storage and other costs in 2,000
selling location

Total Cost per Bag 246,560
Selling Price 260,000
Net Margin 13,440
Return to Capital 0.054
(net margin/total cost)

Return per Day of Labor 42 - 54
(000 Mt)

Maputo

180
19

2,222

2,111

192,000
0

45,500

2,000

243,833
260,000
16,167
0.066

88-112

Nhamatanda Nhamatanda Nhamatanda Nhamatanda Nhamatanda

Maputo Maputo Maputo
100 100 30
25 21 17

4,000 3,500 1 1,666
3,750 2,100 9,633
192,000 1 84,000 1 84,000
0 0 0
45,500 46,000 46,000
2,000 2,000 2,000

247,250 237,600 253,300

260,000 250,000 250,000

12,750 12,400 (3,300)

0.051 0.052 -0.013

33 - 40 35 - 44

 

Data Source: Author Informal Wholesaler Survey Results

134

Table 5.7c (cont’d). Net Margins for Informal Wholesaler in the Maputo Market

 

Parameters Trader 6 Trader 7 Trader 8 Trader 9 Trader 10
Sales location Maputo Maputo Maputo Maputo Maputo
Purchase location Nhamatanda Nhamatanda Nhamatanda Nhamatanda Nhamatanda
Number of bags purchased 50 70 80 500 25
Number of days in 10 24 17 33 10
purchase location

Cost Item (Mt/bag)

Transport of trader to 8,000 5,714 5,000 900 16,000
purchase location

Living expenses in 3,000 3,429 2,125 990 4,800
purchase location

Purchase price 184,000 192,000 160,000 160,000 160,000
Transport of trader back to 0 0 0 0 0
selling location

Loading, transport, 46,000 46,000 46,000 40,500 44,500

unloading of bags

Storage and other costs in 2,000 2,000 2,000 2,000 2,000
selling location

Total Cost per Bag 243,000 249,143 215,125 204,390 227,300
Selling Price 250,000 250,000 250,000 250,000 250,000
Net Margin 7,000 857 34,875 45,610 22,700
Return to Capital 0.029 0.003 0.162 0.223 0.099
(net margin/total cost)

Return perDay ofLabor 15-21 2-2 90-116 485-570 24-33
(000 Mt)

Data Source: Author Informal Wholesaler Survey Results

135

Three important points stand out from these results. First, the net margin for these
traders was always greater than zero but one case, and range between 800 Meticais and
46,000 Meticais per bag of 70 Kg. Second, these results suggest that the size of the net
margin is a function of the volume of products that the informal trader acquires
(economies of scale). The smaller the number of bags purchased, the lower the net
margin. Third, the transportation cost of maize is the major contributor on the total

marketing costs, as shown in Figure 5.7.

Figure 5.7 Percentage of Each Marketing Cost in the Total Cost, Excluding
Purchase Cost

 

1:) Trader Transport Cost

~ D Handling and Bag Transport Cost

 

 

 

136

5.8 Econometric Results for Market Integration Analysis

5.8.1 Correlation Coefficients

The correlation coefficients among selected retail maize markets in Mozambique

are presented in Table 5.8.1a using nominal and real prices. These results suggest that the
correlation coefficients are always highest on nominal prices. This was already expected
because of the effects of the inflation in the economy.

Although the use of real prices may also indicate a presence of integration when in
fact there is not (seasonality problems), the correlation coefficient based on them suggest
that three of the observed relationships are weak, namely Xai Xai and Chimoio (0.46),
followed by Xai Xai and Beira (0.49), and Maputo and Xai Xai (0.51).

Chimoio and Beira show the highest correlation (0.84), followed by Beira and
Maputo (0.69) and then by Chimoio and Maputo (0.68). Chimoio and Beira, and also
Chimoio and Maputo are direct linked in the maize trade, so it was anticipated that they

would have among the highest correlations.

Table 5.8.la Correlation Coefficients AmongMaize Retail Markets
Markets Maputo Xai Xai Beira Chimoio

 

Prices (Nominal and Real)

Nom. Real Nom. Real Nom. Real Nom. Real

 

Maputo 1 1 0.81 0.51 0.81 0.69 0.83 0.68
Xai Xai 1 1 0.74 0.49 0.72 0.46
Beira l 1 0.92 0.84
Chimoio 1 1

 

Data Source: SIMA

137

Table 5.8.lb Road Distances Between Maize Markets (Kilometers)

 

Maputo Xai Xai Beira Chimoio
Maputo l 210 1200 1000
Xai Xai 1 990 790
Beira 1 200
Chimoio l

 

Data Source: Mozambique Map Distance.

5.8.2 Unit Root Test

Before testing for market integration, each price series was submitted to a unit
root test (Table 5.8.2). The results of unit root test for maize grain at retail level using
weekly maize prices show that the null hypothesis’ of unit root would not be rejected at
the 0.1 level of significance for every maize price series in the case of no drift and no
trend test. The p-value is greater than 0.1 for every case, and thus, it can be concluded
that the maize price series for Maputo, Xai Xai, Maxixe, Chimoio, and Beira suffer from
a unit root at retail levels. However, in the case of drifi and no trend test, three price
series (Xai Xai, Beira e Chimoio) are stationary at levels in lower levels of significance.

When all price series are differentiated once, the results of the unit root test
indicate that the null hypothesis of a unit root can be rejeCted at 0.01 level of significance.
Indeed, the low probability values indicate that the null hypothesis is strongly rejected.

Therefore, at first difference, all series of prices are stationary and thus 1(1).

 

5

Null hypothesis: the maize price series is not stationary, that is, [3, in equation 5.4.3a is
equal to 1 or 6 in equation 2 is equal to 0.
Alternative hypothesis: the maize price series is stationary.

138

Table 5.8.2 Results of Unit Root Test for Maize Grain at Retail Level

 

 

 

 

Market No Drift, no Trend Drift, no Trend
Level 1" difference Level 1" difference
t-statistic Lag t-statistic Lag t-statistic Lag t-statistic Lag
(p-value) (p-value) (p-value) (p-value)
Maputo -0.301 4 -10.39"* 3 -1.665 4 -10.37"* 3
(0.5756) (0.0000) (0.4491) (0.0000)
Xai Xai -0.686 5 -9.43"'M 4 -3.026** 5 -9.41 *** 4
(0.4184) (0.0000) (0.0325) (0.0000)
Beira -0.983 4 -9.76*** 3 -2.806* 3 -9.74*"'* 3
(0.2948) (0.0000) (0.0573) (0.0000)
Chimoio -1 .077 4 -8.69*** 3 -2.821* 6 -8.67*** 3
(0.2543) (0.0000) (0.0553) (0.0000)
"‘ Significant at 1 percent level of significance.
" Significant at 5 percent level of significance.
"‘ Significant at 10 percent level of significance.
5.8.3 Granger Causality Test

The results of the Granger Causality Tests are reported in Table 5.8.3. The results
reveal that, at a probability level of 1 percent, only Chimoio and Beira show bidirectional
causality. This result is consistent with the known fact that this market pair sometimes
exhibits price inversions, with price levels in Chimoio exceeding those in Beira, and that
the normal flow from Chimoio to Beira therefore stops at some times and is known on
rare occasions to have been reversed.

Maputo-Chimoio and Maputo-Xai-Xai are clearly unidirectional, with Maputo
Granger causing the price in the other market. Maputo-Beira and Xai-Xai-Beira are also
plausibly unidirectional, with Beira being Granger caused by the other market in each
case. The fact that Maputo always Granger causes any market with which it is paired is

139

consistent with Maputo’s dominant position in price formation in the South and Center of

the country. Chimoio and XaiXai is the only clear market pair with independent

causality. This result too is consistent with expectations, as trade flows do not typically

connect these markets.

Table 5.8.3 Results of GranflCausality Test for Pairs of Retail Markets

 

 

Null Hypothesis F-test Probability

Chimoio price does not Granger Cause Beira price 9.6932*** 0.0000
Beira price does not Granger Cause Chimoio price 8.8496*** 0.0000
Maputo price does not Granger Cause Beira price 6.5541 *** 0.0000
Beira price does not Granger Cause Maputo price 2.3487* 0.0544
Xai Xai price does not Granger Cause Beira price 2.7272" 0.0295
Beira price does not Granger Cause Xai Xai price 2.0149* 0.0923
Maputo price does not Granger Cause Chimoio price 6.6974*** 0.0000
Chimoio price does not Granger Cause Maputo price 1.3675 0.2451
Xai Xai price does not Granger Cause Chimoio price 1.7623 ‘ 0.1364
Chimoio price does not Granger Cause Xai Xai price 1.7267 0.1440
Xai Xai price does not Granger Cause Maputo price 1.3957 0.2354
Maputo price does not Granger Cause Xai Xai price 5.2819*** 0.0000

“‘ Significant at 1 percent level of significance.

*" Significant at 5 percent level of significance.

"' Significant at 10 percent level of significance.

5.8.4 Spatial Market Integration

As mentioned before, two or more nonstationary time series are said to be

cointegrated if a linear combination of them is 1(0) even though individually they are each
l(l). Thus, the hypothesis of cointegration consists of two parts: The first part consists of

testing for 1(1) of the individual series, and then testing for 1(0) of a linear combination of

140

the series.

The results of cointegration test for maize grain markets at retail level are
presented in Table 5.8.4. The specification of the equation for the cointegration analysis
followed the results of Granger Causality Test. That is, if the direction of the causality is

from j to k, then k=f(j), while in the case of bidirectional, we also estimate j=f(k).

Table 5.8.4 Results of Cointegration Test for Maize Grain at Retail Level

 

 

Market Pair 6 t-statistic (p-value) Number of Lags
k=f(j)
Xai Xai ........ Maputo -0.43 -5.2261 (0.000)*** 3
Beira .......... Maputo -0.69 -5.6l33 (0.000)*** 3
Chimoio ....... Maputo -0.67 -4.9876 (0.000)*** 7
Beira ......... Chimoio -0.8 -4.4876 (0.0058)*** 6
Chimoio ......... Beira -0.81 -5.9646 (0.0000)*** 3
"* 1 Significant at 1 percent level of significance.

The results of the test statistic reported on Table 5.8.4 shows that the null
hypothesis of a unit root in the cointegration regression can be rejected at 0.01 level of
significance. Therefore, it can be concluded that the linear combination of any retail
market price series with the Maputo retail price series is 1(0), and thus, markets exhibit
some level of integration in the long run.

Among all pair of markets, the results reveal that there is high degree of market
integration or long-run relationship between Beira and Chimoio. This was already

expected based on evidences from the correlation coefficient, and also from the maize

141

trade pattern, since the Beira terminal market is often supplied with maize grain from the

districts surrounding Chimoio area.

5.8.5 Vertical Market Integration

The cointegration test was also performed between retail and informal wholesale
levels within a given market. The results reported in Table 5.8.5 show that there is
Long-rim price relationship between retail and wholesale maize grain markets. In the
case of Maputo, the retail market completely assimilates wholesale market information
within four weeks, while in Beira and Chimoio, the results are almost the same although
with some different time lags. Therefore, the level of market integration between
vertically-related marketing levels is high in Maputo, Beira, and Chimoio, while in Xai
Xai is very small. This result in Xai Xai was already expected, based on the results of
marketing margins presented above. Indeed, the wholesale-retail margin in Xai Xai is
negative, and the result of the cointegration analysis suggests that there is a lack of

coordination between retail and wholesale levels.

Table 5.8.5 Results of Cointegration Test for Maize Grain at Vertical Levels

 

 

Wholesale . . . Retail 6 t-statistic (p-value) Number of Lags
Maputo ..... Maputo -1 -5.0801 (0.000)*** 4
Xai Xai ..... Xai Xai -0.27 -3.9540 (0.031)" 6
Maxixe ..... Maxixe -0.69 -4.6350 (0.003)*** 3
Beira ......... Beira -O.99 -4.4762 (0.006)*** 9
Chimoio . . . Chimoio -0.95 -4.4777 (0.006)*** 7
m Significant at 1 percent level of significance; ** Significant at 5 percent level of significance

142

5.8.6 Dynamic Adjustments

Results of estimation of equation 5.4.5, which followed the results of Granger
Causality Test, are reported in Tables 5.8.6a and 5.8.6b. In Table 5.8.6a, results for
Durbin-Watson statistic (indication for serial autocorrelation) are presented, as well as the
adjusted R2 (overall performance of the estimated equations). Results for Durbin-Watson
statistic indicate that there is no serial correlation in each estimated regression, which
implies that the lag length based on AIC is good enough to capture irregularities in prices.

However, the adjusted R2 is quite low, which means that current changes in maize
prices in each market could be explained only partially by changes in its own past prices
as well as current and historical price changes in other markets. Therefore, in his
dynamic model of spatial price differential, Ravallion (1986) included a vector X as
another exogenous variable to represent other influences such as seasonality, for instance.

Proceeding as Ravallion did, the overall performance of estimation is improved slightly.

143

Table 5.8.6a Validation of the Autoregressive Model in Maize Grain Markets

Endogenous Exogenous Variable Exogenous Variable in
Variable Ravallion Model

 

Maputo Beira Chimoio Maputo Beira Chimoio

 

Durbin-Watson Statistic Durbin-Watson Statistic
Xai Xai 2.01 - - 2
Beira 2.02 - 2 2 1.99
Chimoio 2.01 2.03 - 2.05 2.05

Adjusted R2 Adjusted R2

Xai Xai 0.25 - - 0.27
Beira 0.21 - 0.29 0.29 0.34
Chimoio 0.06 0.09 - 0.12 0.11

 

In Table 5.86b, the results of the dynamic price analysis for maize grain markets
in Mozambique are presented. The results show that there is no perfect price matching
between maize markets in Central and Southern Mozambique, since the contemporaneous
price effect is far from the unitary. A close examination of the contemporaneous price
effect shows that the impact of Maputo seems to be larger for markets that supply it with
maize grain. In fact, the analysis of maize flows to Maputo presented in Chapter 3 had
shown that Chimoio was one of the major suppliers of maize to Maputo, while the role of
Xai Xai was secondary. Furthermore, the same analysis of maize flows pointed out that
there is no role for Beira in supplying Maputo. All these results are reflected in the

dynamic price adjustments whose results are presented below.

144

Table 5.8.6b Dynamic Price Adjustment in Maize Grain Markets

 

 

Pair of Markets Contemporaneous Lagged Price Effect
P ' Eff t
rice ec Magnitude Lag Length (Weeks)
Xai Xai ..... Maputo 0.1146* 0.3048 1
Beira ....... Maputo 0.0445 0.4753 1
Chimoio . . . . Maputo 0.2258"”'”'I 0.546 2
Beira ...... Chimoio 0.1013" 0.5823 3
Chimoio ...... Beira 0.1245" 0.8 3
*" Significant at 1 percent level of significance.
" Significant at 10 percent level of significance.
"‘ Significant at 15 percent level of significance.

The role of Maputo in the process of price formation is also supported by the
magnitude of the lagged price effect. Indeed, past price changes in Maputo exerted a
large influence on contemporaneous price change in all other markets, especially in
Chimoio. In other words, if a market shock occurs in Maputo, 55 percent of that shock
will be transmitted to Chimoio within two weeks, while 30 percent will be transmitted to
Xai Xai within 1 week but very little thereafter. Other small markets located between
Maputo and Chimoio could be responsible for absorbing some of the price shocks,
resulting in an incomplete adjustment following price shocks.

Since the maize grain markets supplying Maputo cannot completely assimilate
market information from Maputo, the speed of response is relatively slow. For instance,
shocks in Maputo take two weeks to cause an effect in Chimoio, inducing therefore low
pricing efficiency. This is even worst in the case of Beira and Chimoio, where a shock

Beira may take three weeks to be felt in Chimoio.

145

5.9 Chapter Summary

The major objective of this study was to assess the performance of maize markets
in Southern and Central Mozambique at the individual market levels and for markets
linked spatially. Two analytical techniques were used to assess that performance, namely
the descriptive assessment of the levels and volatilities of maize prices (and price spreads
too), and also the econometric tests of price transmission across spatial markets and
within levels of transaction.

By examining the nature of relationships among wholesale and retail prices in
individual markets and among different markets, the analysis provided useful
information, important to improve our understanding of the maize marketing system.
This information may promote further debates among all participants in the maize
marketing system toward the formulation of effective maize marketing policies.

’ In general, our descriptive results reveal that maize price (either at retail or
wholesale) in Southern and Central Mozambique exhibits some level of volatility,
completely normal based on the strong seasonality observed in maize marketing. Overall,
prices are much higher in Southern markets (Maputo and Xai Xai), but are more volatile
in Central markets, represented in this analysis by Beira and Chimoio. Recall that the
Central Region is the surplus area while the Southern is the deficit, being supplied by the
Central Region as well as by the producers’ districts from the Southern Region.

The retail spread is always positive in all markets, except in Xai Xai. In general,
the wholesale share as retail price is between 85 percent and 92 percent. In other words,

retailers are adding between 8 and 15 percent over the wholesale price. However, the

146

retail margin as a percentage of the wholesale price is increasing significantly in Maputo
and Chimoio, while in Xai Xai they are decreasing significantly. In Beira, is also
increasing, but not statistically significant.

The results of marketing margins show that in average, a trader buying maize
grain in the Central Region and transporting to Maputo has a net return varying 900 Mt
($US 0.05) and 46,000 Mt ($US 2.3), which gives a minimum return per day of labor of
2,000 Mt ($US 0.10) and a maximum of 570,000 Mt ($US 28.5), depending on scale of
operation of maize traders.

The result of Granger causality test position Maputo market as the price leader,
“Granger causing” prices in Xai-Xai and Chimoio. Our econometric results reveal that
price’s responses are non-unitary, suggesting that price changes are not perfectly matched
across markets, and that price adjustment to new information originating from another
market is slightly delayed. In average, if a price shock occurs in one market, an average
of 54.2 percent will be transmitted to the other markets, a process that takes in average 2

weeks.

147

 

6.1

CHAPTER SIX

CONCLUSIONS AND IMPLICATIONS

Introduction

In this thesis, it is argued that well functioning of food markets is a prerequisite

for attaining food security. As discussed in Chapter one, Section 1.3, this thesis aims at

reaching one broad objective that is to gain a better understanding of the maize marketing

system, especially in Southern Mozambique, to contribute to more effective food security

policies and market growth for the benefit of farmers and consumers. For this purpose,

the preceding chapters sought to answers the five research questions presented in Chapter

1, namely:

1.

What are the maize production, consumption and marketing patterns among the
rural population of Mozambique?

Are consumers in the Southern region of Mozambique, especially poor urban
consumers and net buying farmer households in rural areas, utilizing less maize
grain at relatively higher prices than in the past? If yes, what is the cause?

How is the maize marketing system structured in Mozambique, especially in the
Central and Southern Region? Who are the major participants in the maize
marketing system in Mozambique, and what are their roles?

Would it be competitive to import more maize grain from South Africa for use in
a broader array of maize meals and for sale as grain in markets in urban and rural

areas of Southern Mozambique, and if so, why is there not more maize grain

148

imported from South Africa for these purposes?
5. What are the costs of trade between Central and Southern Mozambique, and how

well integrated over space are maize markets in theSe two regions?

This last chapter will summarize the major findings of the study, and also suggest

policy conclusions which can be drawn.

6.2 Major Findings

6.2.1 Maize Production and Marketing Pattern

The analysis of maize grain in Mozambique smallholder livelihoods showed that
maize grain is actively produced in all regions of the country. A majority of rural
households produce maize grain primarily for home consumption. This was explicitly
shoWn with the concentration and levels of maize production. In all regions, between 50
and 77 percent of the production of maize grain is concentrated among household who
produce but do not sell.

The analysis also shows that maize sales are very concentrated. In all regions,
between 71 and 77 percent of the volume of maize sales was accounted for by the top 25
percent of sellers, representing only 6 percent of all households in the Center and North,
and only 1 percent in the South.

The top 25 percent of sellers are clearly better off as compared to other rural
households in all categories studied. Overall, the top producers and seller have the largest

percentage of marketed production, have maize yields more than double the average,

149

plant more than double the area, and have about double the amount of land per person in
their households. Their incomes are also about double those of other rural households,
they have significantly more assets and they are more literate, and have better access to
price information. In addition, these households are less likely to be headed by females.
Purchases of maize grain or maize meal are widespread in rural areas, and are not
confined to one single type of household. In the Southern Region for instance, some 75
percent of rural households either do not produce or produce on average only 200 Kg, and
these households are the ones who indicated they buy more from the market. This
situation is also aggravated by the inability of emerging maize millers to get sufficient
maize grain from domestic markets during the hungry season, thus reducing maize meal
availability during this season. Perhaps more importantly, this inability of millers to
source grain domestically is an indicator that households may not be able to purchase as
much grain as they want. This is important because, when compared to maize meal,
maize grain is the cheapest option, and would therefore be chosen by many poor
households if they had the option. Therefore, if the market for maize grain and maize
meal, as well as for other products, does not work well enough, a large ntunber of rural

households will be affected, especially during the hungry seasons.

6.2.2 Maize Expenditure
Based on IAF results, there was no change in national average percentage of total
households’ expenditures (budget shares) allocated to maize and its derivatives between

1996 and 2002. In 1996, households spent on maize and its derivatives as much as in

150

2002 (15 percent), registering generally similar patterns on total expenditures for wheat
and rice over the period.

However, disaggregating the national picture, rural and urban expenditure patterns
by regions of the country shifted significantly, and in patterns that reinforce the
importance of the present research about the degree of availability and cost of maize
products in Southern Mozambique. For example, in 2002, households in the rural South
(Maputo, Gaza and Inhambane) were still relying significantly on maize and its
derivatives, but its relative importance fell approximately in half compared to 1996. The
budget share for cassava products also fell significantly, while expenditures on rice rose
from 6 to 11 percent in rural Maputo, and from 6 to 9 percent in rural Gaza and
Inhambane. The expenditures on wheat and its derivatives held about the same over the
two periods in rural areas.

In contrast, households in rural areas of Central Mozambique (Sofala, Manica,
Tete and Zambezia) expanded the share of maize and derivatives to much higher levels
(almost doubling their already high share). Meanwhile, the importance of rice and wheat
held about even, while the role of cassava products increased modestly, and were about
half as important as maize in rural Sofala and Zambezia.

Turning to the urban areas, in 2002, the percentage of total food expenditure
allocated to maize products went up in both the South and Center Regions, and while the
absolute role was still modest in Maputo, it became much more important in the other
parts of the urban South and Center (in urban areas of Gaza and Inhambane, maize

surpassed rice in importance between 1996 and 2002). Maize products also became more

151

important in the urban as compared to rural areas of Gaza and Inhambane. Wheat
products went down modestly or retained their relative position of 1996, and still
dominated the basic commodity groups only in the South.

In urban areas, rice actually went down significantly in the South and remained
about the same in the Center. Also, rice’s relative position is strongest in Beira, lower in
Xai Xai, and lower still in Maputo. Overall the share of basic cereals and cassava for
urban households in Maputo (South) decreased from 1996 to 2002.

In general, the disaggregated picture from the IAF expenditure/budget share
analysis reveals important changes from 1996 to 2002. There is a pattern of shifts into
maize products in all areas studied, with the exception of the rural South where maize
products lost the strong dominant budget share position held in 1996. And in the rural
South, it appears that households may have tried to compensate with shifts into fruits and
leaves, and to also a lesser extent into rice.

Nevertheless, the analysis of expenditure shares by tercile of total household
expenditure pointed out that maize grain clearly remains the dominant staple in rural
areas of Mozambique, although rice (as well as cassava) in the South is potentially
important for the lowest income consumers. In urban areas, rice and wheat have gained
major market shares in consumption habits of the South. Besides the specific
characteristics of these products (convenience, for instance), a likely reason for this
patterns can be related to policies that change the relative availability and cost of these

goods.

152

6.2.3 Maize Subsector in Southern and Central Mozambique

In Chapter III, we analyzed the structure of maize market transactions in Central
and Southern Mozambique. TIA data show that in the South (Maputo, Gaza, and
Inhambane provinces), about 1% of all farmers accounted for more than three-quarters of
all maize sales. In the Center (Tete, Manica, and Sofala provinces) and North (Zambezia,
Nampula, Cabo Delgado and Niassa), about 6% of farmers accounted for about less than
three-quarters of all sales. TIA data also show that production in each region is
geographically dispersed: In the South, no district had more than 12% of total regional
sales, while in the Center no district exceeded a 15% share. These results make it clear
that the large producers and sellers are spread over a large geographic area, implying very
high costs of assembly.

According to the marketing channel approach, in the Southern region, 97 percent
of the total production in the region is retained at home, while the remaining is sold or
intended to be sold in the market. In the Central region, the situation is also similar, with
86 percent of the regional production being retained at home and the remaining sold in
the market.

In both regions, maize grain moves from producers to several buyers such as
formal traders and owners of rural shops, but the most important ones are informal
traders. Indeed, in the Southern Region and also in the Central Region, they are
responsible for more than half of maize grain bought from maize producers. Therefore,
informal traders are the more important participants willing to bear all costs of collecting

the disperse maize grain production. Also, they are responsible to bring maize from

153

production areas to consumer markets.

It is interesting to note that in the Central Region, the second most important flow
of maize grain from producers is to neighboring countries. This is important in the sense
that these countries are competing with the Southern Regions for maize grain produced in
the Central Region. The analysis of SIMA data as well as the results of maize traders
surveys show that Central Region of Mozambique is the major supplier of maize grain for
the markets in the Southern Region. Maize grain is still flowing from districts of the
Manica and Sofala provinces to markets in the South, and this flow has been continuing
since the end of the civil war in 1992 despite the pressure from neighboring countries.

The costs the informal traders incurred for bringing maize from Central to
Southern Region can be classified in transport and other small marketing costs. This last
cost includes the daily living expenses, and also handling costs.

Several problems in the maize marketing in Mozambique that impedes the growth
of the market participants are related to lack of credit, absence of market information
widely and timely available, and underdevelopment of tertiary roads. Because
information on trade opportunities is difficult to obtain, it is important for these traders to
set up an expensive network of assemblers who know local market conditions. Maize
traders often spend several days waiting for transport, which is most of the times
backhaul.

Medium and large maize millers are other important participants in the maize
marketing system. However, contrary to informal traders that purchase maize grain from

producers, large scale millers mainly rely on imports of maize grain to satisfy their

154

demand, while the medirun scale depend primarily on local purchases, complemented by
imports from South Africa, especially during the hungry season. The need for good
quality grain as well as regular supplies throughout the year drives these imports.

The concentration is high in the milling industry, with the largest two millers
holding about 70 percent of the market. The emergence of small to medium scale
commercial millers increases the competition in the milling sector. These small to
medium commercial millers are trying to carve out market niches for a quality of maize
meal that is made primarily from national production, and that is priced to compete well
with the highly refined maize meals made mostly from imported white maize grain from

South Africa and the USA by large industrialized millers.

6.2.4 Impact of VAT on Maize Imports

Maize grain, on the contrary of rice, edible oil, bread, and maize meal, is not
exempt of VAT payments during imports. But if it is processed into maize meal, the total
amount paid on VAT during the process is reimbursed to the miller. This process of
charging VAT on maize imports affects mainly those traders interested in importing
maize grain to be sold as grain in the market, since they cannot get back the money
invested on VAT payments.

The analysis of the impact of VAT on maize imports shows that it would have
frequently been profitable to bring in maize grain from South Afi'ica if no VAT was
charged. With VAT being charged, the analysis shows that usually it was not profitable

to import maize grain, even though there were some extended periods when it was still

155

profitable, but imports did not occur.

The VAT charges may be a potential disincentive to the emergence of more
informal as well as formal maize trade from neighboring countries, especially from South
Africa, to millers and consumers in the South. The fact that the VAT is reimbursable
only if the grain is processed into meal creates an immediate disincentive for the
importation of maize grain for sale as grain. Since there is evidence that maize grain
becomes seasonally scarce in the South, this unlevel playing field for maize imports has
serious implications for the welfare of rural households, most of whom are net buyers of
maize.

The unequal application of the VAT is, however, not the only reason for the poor
state of trade in maize grain between Mozambique and South Africa“. The regulation to
import maize grain is very heavy for an informal trader, contributing therefore to the poor
state of trade between these two countries. Also, the high level of formality, large scale
of operation, and lack of easy availability of maize grain at what might be termed the

“petty wholesale level” within the South African system contributes to this situation.

6.2.5 Performance of Maize Subsector
The results of net marketing margins analyzed at one point in the system (informal
assemblers/wholesalers in Maputo) suggest that maize trade between Central and

Southern Region is profitable when the destination market is Maputo, but that this

 

6

This trade works well for large and some medium scale millers, but quite poorly for smaller importers

156

depends greatly on the scale of operation of the trader.

Results of the market integration analysis show that Central-South trade in
Mozambique is still somewhat weak. Maize traders’ responses to inter-market shocks are
fragmented and incomplete, as indicated by the contemporaneous price effect being less
than unitary, suggesting that market signals are inaccurately read and translated.

Reaction time to unanticipated market news is typically sluggish, exceeding one
week, suggesting a break-down in the transmission and processing of new information.
However, based on the structure of the maize subsector in Mozambique (majority of
households do not sell any maize, disperse production, lack of credit, lack of access to
market information, problems with transports, for instance), a price reaction that takes a

week or two seems good, and may not be judged as a poor performance.

6.3 . Policy Implications

The picture which emerges from our findings is that most rural households in the
South and even in the Center purchase more maize and maize meal than they sell, that
they are paying the highest maize meal prices in the Southern Afiica region despite the
recent emergence of medium-scale millers, and that they probably have difficulty
obtaining maize grain — always the cheapest option — in markets during the hungry
season, due in part to policy, regulatory, and other difficulties involved in importing
maize grain for sale as grain.

Based on these findings, it is possible to formulate some recommendations that

might be initiated by policy makers or development institutes. Nevertheless, it is

157

important to emphasize that implementing only a single improvement in the maize
marketing system will not have the wished effects.

The policy implications of the above findings for thehouseholds’ food security
are all important, especially for households in the Southern Region. First, because most
of the rural households are net buyers of maize grain, policies and investments to make
maize grain more available in the South at prices much lower should be put in place to
mitigate the high costs of maize meal.

Such policies may be the relaxation of some requisites to import maize grain by
informal traders wishing to import FOB values above $500USD, for instance. If this
seems so important for maize traders (not just informal), training them in issues related to
import procedures (and perhaps issues related to maize marketing system in South Africa)
is also crucial because part of the problem is that they just do not understand the
procedures.

Another policy implication of the findings is related to VAT. Although it does
not generate any revenue to the government, the government may be hesitant to remove it
for any maize transaction. Removing VAT for maize grain may open a window to
importers of other goods susceptible of VAT payments to also ask for VAT removals.
Yet the government did remove permanently VAT during the last trimester of 2004 for
edible oil, and sugar; also, rice, bread, and maize meal has been exempt from VAT since
its introduction in 1999. Given this situation, and the way in which the VAT inhibits
imports of maize grain for sale as grain (which would be the cheapest option for poor

households), we believe that the government should seriously consider removing it for

158

maize grain.

Increased competition in the milling sector is desirable, and if it is true that this
may happen through an “invisible hand” as small and medium commercial millers
expand, it is also true that the government role may not be static. For example,
government may reduce import taxes for milling machinery, especially medium and small
scale machinery. This action may provide an incentive for new investments or even
expansion of small and medium scale millers.

The Mozambican policy makers and deve10pment institutions should make a
commitment to create and support local initiatives to improve maize production in rural
areas, and also to strength the extension services, creating for instance, extension
programs to make all participants in the maize marketing system sensible to the problem
of maize grain quality.

Food security in rural areas should also be promoted by improving the production
of cereals in general. This is reflected by the importance of the value of cereals consumed
and sold on farms, especially in the Central Region. Improving households incomes by
increasing sales of commodities with high monetary value only, will not have the desired
effects on household food security.

Expansion of market information systems into more remote areas of the South,
perhaps in collaboration with Early Warning and Vulnerability Assessment units, should
be considered. More information on availability and prices of basic commodities in these
areas may be useful to monitor how the current situation of food security hurts rural

households, and then to inform policy makers. Timeliness of widespread dissemination

159

l

for this information should be stressed: If market participants have accurate and timely
information on market conditions, maize markets may be able to respond more quickly to
market shocks, and market linkages can efficiently and effectively distribute maize grain

from surplus to deficit areas.

160

 

APPENDICES

161

 

APPENDIX 1

162

APPENDIX 1: Rural Hausehold Asset Indicators, and Other Selected
Characteristics by Category of Maize Production and Sales

 

 

A. Northern Region
Characteristics Households Households Household With Both Production
of Maize With No With and Sales by Quartile
Households Production Production
But No 1. Lowest 2. 3. 4. Highest
Sales Quartile Quartile

 

Assets Owned - Percent of Household with this Asset

 

 

Using Irrigation l 3.1 2.8 5 3.3 3.2
Receiving Price 34.3 41.3 45.8 40.6 41.0 54.2
Information

Access to 10.3 10.6 3.9 8.0 6.6 8.0
Formal or

Informal Credit

Household Characteristics - Percent of Households with this Characteristic-----------

 

 

Selling Cash 4.5 10.0 8.6 11.4 10.7 13.6
Crops
Selling Cereals 6.7 3.1 100.0 100.0 100.0 100.0

 

 

Data Source: TIA

163

B. Central Region

 

Characteristics Households Households Household With Both Production

of Maize With No With and Sales by Quartile

Households Production Production
But No 1. Lowest 2. 3. 4. Highest
Sales Quartile A Quartile

 

 

 

Assets Owned - Percent of Household with this Asset

Using Irrigation 15.7 20.2 12.2 27.1 31.6 21.1
Receiving Price 24.7 36.3 32.8 32.6 49.1 50.3
Information

Access to 2.8 9.5 11.0 12.7 8.2 12.3
Formal or

Informal Credit

Household Characteristics - Percent of Households with this Characteristic-m- -------

Selling Cash 0.3 11.9 4.5 11.0 12.9 21.7
Crops
Selling Cereals 0.1 1.0 100.0 100.0 100.0 100.0

 

 

Data Source: TIA

164

C. Southern Region

 

Characteristics Households Households Household With Both Production

of Maize With No With and Sales by Quartile

Households Production Production
But No 1. Lowest . 2. 3. 4. Highest
Sales Quartile Quartile

 

Assets Owned - Percent of Household with this Asset

 

 

Using Irrigation 25.1 28.3 71.3 34.6 27.6 44.7
Receiving Price 1 1.9 12.4 0.3 0.7 8.2 6.7
Information

Access to 11.4 14.1 19.2 47.7 32.1 47.7
Formal or

Informal Credit

Household Characteristics - Percent of Households with this Characteristic

Selling Cash 0.2 0.5 0.0 0.0 0.0 0.4
Crops
Selling Cereals 0.1 0.5 100.0 100.0 100.0 100.0

 

Data Source: TIA

165

APPENDIX 2

166

 

 

APPENDIX 2: Questionnaire for Maize Processors

I. GENERAL INFORMATION

Mills’ Name:

 

Location

1. Province
2. City

 

 

Person Interviewed

1. Name
2. Function

 

 

Types of maize meals and other products produced at the mill and relative
proportions '

 

 

Plant Daily Production Capacity

1. Potential
2. Actual

 

 

Days of Operation per Year

 

167

II. MAIZE’ PURCHASES - FROM DOMESTIC PRODUCTION
1. Did your company purchase maize from national production last year?

1. Yes (Fill Table Below)
2. No (Go to Question 13)

 

Channel What Proportion From Purchases in Each Channel, What
or Your Total Proportion Came from the Following
Purchases of Sellers?

 

Domestic Came

from this Channel Farmer Formal Informal Other

Trader Trader (Specify)

 

Factory Door

 

Contracts

 

Buying Station
in Rural Areas

 

Other
(Specify)

 

 

 

 

 

 

2. If YES, what if any problems did your company experience in local maize
purchasing?

 

 

 

 

 

168

3. When the Purchases Were Made Outside the Plant, What Was the Proportion
Purchased in the Following Provinces

 

 

 

0. Maputo 6. Zambezia

1. Gaza 7. NMpula

2. Inharnbane____ 8. Cabo Delgado
3. Sofala 9. Niassa

4. Manica

5. Tete

 

4. What Was the Company’s Strategy to Guarantee a Regular Supply Throughout
the Year?

 

 

 

 

 

5. Does Your Company Have Any Agreement With Any Individual to Supply Your
Mill Regularly?

1. Yes
2. No (Go to Question 9)

6. If YES, Do These Agreements Specify Quantity and Quality?

Quantity Quality
1. Yes 1. Yes
2. No 2. No

169

 

7. Does Your Company Pay Any Price Premium for High Quality of Maize Grain or
Give a Price Discount for Poor Quality of Maize Grain? Explain.

 

 

 

 

 

8. Does Your Company Pay Any Price Premium for Large Volumes of Maize Grain
or Give a Price Discount for Small Volumes of Maize Grain? Explain.

 

 

 

 

 

9. How Does Your Company Usually Pays for the Purchases? Check all Options.

 

Payment Producers Formal Traders Informal Traders Others
Cash
Check
Other

 

 

 

 

 

 

 

 

10. Is This the Same Way That You Pay the Purchases Under the Agreement?

1. Yes
2. No

170

11. Are the Agreements Written?

1. Yes
2. No

12. What may happen if your company has any problem with the agreements?

 

 

 

 

 

III. IMPORTS

13. If Your Company Buys Imported Maize, Is There Any Especial Reason for It?

 

 

 

 

 

14. From Where Does Your Company Import Maize Grain? In Which Proportion?

 

Source Proportion

 

 

 

 

 

 

171

15. How does the Majority of Maize Grain arrive at Your Mill?

1. Train
2. Trucks
3. Ships
4. Other

 

 

16. In This Case of Imports, What Is the Strategy to Guarantee a Regular Supply
Throughout the Year?

 

 

 

 

 

 

17. Does Your Company Pay Any Tax on Maize Grain Imports?

1. Yes
2. No

18. If YES, Which Are They?

 

 

 

 

 

172

19. How Long Does It Takes to Recovery VAT? What Is the Cost of Credit as They
Wait for Recovery? How Big a Problem Is VAT on Maize Grain for Your Business?

 

 

 

 

 

20. Do You Face Any Other Bureaucratic and Logistic Difficulties in Purchasing Maize
Grain?

 

Difficulty Comments

 

Licenses

 

In the transport
system

 

In port / border

 

Other

 

 

173

IV.

SALES

21. What Kind of Maize Products Does Your Company Produce to Sell?

 

 

 

 

 

22. What Is the Sales’ Percentage for Each Product?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Product Percentage
23. What was the total of sales in the last 12 months?
M Product M Product
0 o
n n
t t
h Percent h Percent
Jan Jul
Feb Aug
Mar Sep
Apr Oct
May Nov
Jun Dez

 

 

 

 

 

 

 

 

 

 

 

 

174

 

 

 

 

24. To Whom Does Your Company Sell the’Bulk of Production?

1. Directly to Consumers
2. Wholesalers

3. Retailers

4. Others

25. What Is the Coverage of the Sales and the Respective Percentage?
1. National market

2. International market

26. In the Case of National Coverage, What Is the Proportion of Sales in Each
Province in 2000 and 2004?

 

 

 

0. Maputo 6. Zambezia

l. Gaza 7. Nampula

2. Inhambane_ 8. Cabo Delgado
3. Sofala 9. Niassa

4. Manica

5. Tete

 

27. Does Your Company Transport the Product Outside the Province Where the
Mill Is Established?

1. Yes
2. No
28. Does Your Company Have Agreements to Supply Some Buyers by Specific

Quantities and Prices?

1. Yes
2. No (Go to Question 31)

175

29. If YES, With Whom Does Your Company Have That Kind of Agreement?
1. Wholesaler

2. Retailers
3. Others

30. How Is Your Company Paid Under the Agreement?

 

Payment Wholesaler Retailer Others

 

Cash

 

Check

 

 

 

 

Other

 

v. QUALITY

31. What Are the Specific Categories of Quality That Are Important for Your
Company? (Check All Possibilities)

1. Size

2. Color
3. Humidity
4. Presence of Foreign Obects (Cleanliness)

5. Other (Specify)

32. How Does Your Company Test for Quality?
1. Own a Laboratory for Analysis

2. Pay for Analysis
3. Other

176

33. If the Company First Conducts an Analysis of Quality, Is at the Time That the
Maize Grain Arrives in the Plant or Before?

 

 

 

 

 

34. What May Happen if the Maize Grain That Arrives at the Mill Does Not Meet
the Company’ Quality Standards?

 

 

 

 

 

35. What Does Your Company Do With the Product That Does Not Meet the
Requirements in Terms of Quality After the Company’ Production?

 

 

 

 

 

177

VI. RESEARCH AND EXTENSION

36. What Is Your Evaluation of the Following Issues for the Milling Industry?
1. A Need to Improve the Overall Quality of LocaltMaize Grain
2. A Need to Evaluate the Efficiency of Maize Marketing System

3. A Need of Identification of Varieties That Will Be More Profitable

1. Very Important; 2. Important; 3. Not Important

VII. OPPORTUNITIES FOR THE MILLING INDUSTRY

37. Which of the Following Technical Improvements Would You Support as Some
Possible Ways to Strength the Milling Industry?

1. Modernization of Maize Production Technology in Order to Increase the
Overall Production

2. Modernization on Storage Technology

3. Develop and Enforce Grades for Local Maize Grain to Increase Competition
between Local and Imported Maize Grain '

1. Strongly support; 2. Moderate support; 3. Not necessary

178

APPENDIX 3

179

APPENDIX 3: Structure of Maize Milling Industry in Mozambique

To understand the structure of maize milling in Mozambique, face-to-face
interviews with managers and/or directors were conducted with four millers. The
questionnaire, which is presented in annex 1, was divided in seven sections and it was

designed as simply as possible in order to be easy to obtain answers from managers

 

and/or directors. It covered aspects such as maize purchases from national production

versus imports, sales strategy, quality problems, needing for further research as well as

 

opportunities for the milling industry in general. The next four sections presents the
results of the interviews conducted with four millers, namely SMC, CIM, VONK, and

MEREC.

SOCIEDADE MOAGEIRA DO CHIBUTO (SMC)

Miller one (SMC) is located in the Southern region, nearly 250 Km North of

 

Maputo, with private ownership. The capacity of this miller is 720 tons per month based
on 24 hours of operation during 7 days, but actual production is 240 tons per month. This
miller operates 280 days per year, employing seven people.

This miller relies entirely on maize grain fiom domestic production for raw
material, but has experienced problems related with the quality of domestic maize, and is
therefore, thinking of importing maize grain fiom South Africa. All maize is purchased
through rural buying stations in the province of Manica. Informal traders are the main

suppliers. When the desired capacity of maize grain is reached, a miller’s truck carries the

180

product directly to the plant.

During the purchases at buying stations, payment is done by check. The amounts
involved are quite high, and due to security reasons, the miller’s policy does not allow
payments in cash.

This miller has small silos (500 Ton) installed in their plant, allowing it to store

maize purchased during the harvest period. The miller does not have any kind of

 

agreement with any individual to supply them regularly. Also, the miller does not pay any

 

premium for maize grain with high quality nor pay any premium for large volumes of
maize.

Miller one sells basically maize meal and maize sub products distributed as 78
percent maize meals, 15 percent grits, and 7 percent dust. This 22 percent extraction rate
qualifies this meal as “refined” as opposed to “super refined”. The total sales of maize

meal in the last 12 months were not revealed. All meal is sold in the domestic market.

 

The major demander for maize meal from this miller is the wholesale and retail
sector, with no direct sales to consumers. However, the miller does not have any
agreement to supply some of the demanders by specific quantities and prices.

Quality of domestic maize grain is a huge constraint for this miller. As mentioned
before, the absence of grades and standards in local production is forcing them to plan to
import maize grain to supply the mill. Up to the time of the interview, the miller did not
have a laboratory nor did it pay for quality analysis. The quality inspection was done by
visual means during purchase. If the maize grain arriving in the plant does not meet the

miller’s requirement, it is processed for animal feed, but the losses were indicated to be

181

about 7 percent.

COMPANHIA INDUSTRIAL DA MATOLA (CIM)

Miller two (CIM) is also located in the Southern region, within 10-15 Km from
central Maputo. The capacity of this miller is 3,000 tons per month, with an actual
production around 2,350 tons per month. This miller operates 7,000 hours per year, which
is equivalent to nearly 292 days per year, 24 hours per day. This miller has some foreign
investors that are the most important owners.

Miller two supplies only about 5 percent of its maize grain needs through
domestic purchases; 95% is imported from South Africa. High domestic prices, absence
of continuous supply, and poor quality, are the main reasons for this miller to rely mostly
on imports.

When maize grain is purchased locally, they never armounce publicly that they
intend to buy maize grain domestically. A hundred percent of it is done through contracts
established with formal traders with good reputations. However, these contracts are not
signed in order to the miller being supplied regularly, that is, the miller does not have any
agreement with any individual to supply them regularly. The miller only buys from
formal traders when they foresee advantages. For instance, if the miller ship maize meal
to the Northern provinces, the containers return with maize grain.

To guarantee a regular supply through imports, the miller uses future contracts
from South Afiican Future Exchange (SAFEX). Seventy percent of imported maize is

from South Africa, while 25 percent is from USA, and 5 percent from other sources.

182

From South Africa, maize grain arrives in the plant through the railroad, while from other
sources, especially from USA, it arrives in ships until the port of Maputo, and then
through railroad to the plant. In both cases maize grain is in bulk.

Besides the fact that this miller is relying mainly in the railroad to transport maize
grain from South Africa to their plant, there are some problems faced by them. For

instance, sometimes there are no enough wagons to load, long and inexplicable delays

 

"’3

from Mozambican Border to the miller’s plant, and sometimes briberies among the
border officials.

Road transportation from South Africa is not used because it is very expensive. At
the time of the interview, the road transport cost from a silo near to Johannesburg, South
Africa, to Maputo was nearly 240 Rand per ton, equivalent to nearly $US3 8, compared to
nearly $USl9 for rail transport.

When maize grain is imported, the miller has to pay all duties, which is divided in
3 percent in official import duty, and 17 percent in Value Added Tax (VAT).

This miller produces two types of maize meal, namely refined and roller meal,

 

sold in 1 Kg, 2.5 Kg, 5 Kg, for both types, and 50 Kg for roller meal. The first type of
meal represents 40 percent of the sales, while the second one 60 percent. The main
demanders of these two are wholesalers and retailers, with no reported sales directly to
consumers.

The principal geographical target of this miller is the national market, where
ninety percent of sales are done in Maputo, and the remaining 10 percent in the rest of the

country. To the clientele operating in the same region where the plant is installed, this

183

miller put the product at the client’s door. Also, miller number 2 transports the product
outside the province where the plant is established, using road transportation to the
Southern Region, and via maritime to the Northern Region.

Use of agreements to supply some demanders by specific quantities and prices, is
a common practice to this miller. These agreements are signed mainly with wholesalers,
using check as the method of payment. This practice, together with the transportation
services, enables the miller to get a high market share within the meal industry.

This miller takes the quality of raw material very seriously. They have two
laboratories where maize grain is tested for moisture. This is done at the time the product
arrives in the plant, and the product is only delivered to the silos after been approved in
the quality tests. If the product does not meet the quality standards, it is then processed

into animal feed.

VONK

Miller three (VONK) is located in the Central region of the country, and it is
private ownership. The actual production of this mill is 400 tons per month, operating
during all year. At the time of the interview, the company had in operation eight small
roller mills, giving the actual miller’s monthly capacity.

Maize grain for this miller is 100 percent purchased in the domestic market
through buying stations in rural areas as well as at the plant door. When the product is not
bought at the plant doors, Sofala and Manica provinces are the principal sources. At the

time of the interview, the price at buying stations was nearly 1,500 Mt per Kg, equivalent

184

 

 

 

to $US0.06, while at the factor door the price was slight higher, that is, 2,000 Mt per Kg,
which is equivalent to $US0.09.

To guarantee a regular supply through the year, this miller buys and stores in
warehouses, fumigating to avoid spoilage. The strategy to establish a buying station
varies from year to year, but most of the times the miller’s manager conducts a wind-
shield survey before the harvest period. When the decision to establish a buying station is
done, small teams equipped with scales, humidity measuring equipment, bags, and so on,
starts buying maize grain, paying in cash if the amount to be paid is less than a million of
Meticais, otherwise in check. No premium for large volumes of good quality is paid.

Miller number three produce basically roller meal packed in 1 Kg, 2.5 Kg, 5 Kg,
12.5 Kg, and 50 Kg, besides the sub-products. Nearly 100 percent of maize meal is sold
in the province where the plant is established, distributed to wholesalers within the area
of millers’ operation as well as to wholesalers located up to 200 Km away. The miller
also sells to local retailers, but it is possible to find retailers 50 Km away from the
millers’ operation selling this product. With few wholesalers, this miller has agreements
to supply them by specific quantities and prices. Under these agreements, the miller is
paid by check.

Quality concerns, especially humidity level, cleanliness, and color, force this
miller to own a small laboratory, where the maize quality is analyzed, especially humidity
level. The quality analysis is conducted first at the buying stations and then at the plant.
Any maize grain arriving at the plant and that does not meet the miller’s standards, is

processed into animal feed.

185

 

 

MEREC INDUSTRIES

Miller number four (MEREC) is located in the Southern region, with a grinding
capacity around 3,000 tons per month operating 24 hoursduring seven days a week. The
manager of this mill believes that the market for maize meals is not large enough to
operate the mill at its full capacity, and therefore, the actual production is around 2,100
tons per month, operating 6 days a week or 312 days per year. It was not possible to get
the exactly number of employees, but official data from Ministry of Industry and
Commerce (MIC) indicate 54 people. This miller has some foreign investors which are
the most important owners.

Even though the domestic maize grain does not guarantee a final product with
good quality, this miller still buys some of it to supply the plant. When the manager
where asked in which situations the miller buys domestic maize, the answer was vague.
However, it seems that the miller never releases a price for domestic maize, but if a
private trader contact them in order to sell maize, they purchase it if the quality, price, and
total quantities are reasonable. For instance, in 2002, this miller bought maize in the
Northern region, and transported via maritime to the South. Therefore, this miller does
not have a strategy to guarantee a regular supply to its plant based on domestic
production.

Due to quality and volume problems with domestic maize grain, miller four
imports maize grain to guarantee a regular supply. In this case, 95 percent of maize grain
is imported from South Africa and 5 percent from USA. In the former case, the miller

pays the contract based on SAFEX prices to receive the product at the plant’s doors,

186

 

 

while in the latter case is through ships until the port of Maputo, and then trucks to the
plant.
As in the case of miller two, this miller also has to. pay all duties, which is divided

in 3 percent in official import duty, and 17 percent in VAT.

Forty percent of sales of this miller are due to sales of refined maize meal, sold in
packs of 1 Kg. The less refined maize meal is sold in packs of 12.5 Kg and 50 Kg. The
bulk of sales is made to wholesalers and retailers, covering only the domestic market

where Maputo retains the largest proportion (70 percent) of the total sales.

Summary

All four millers interviewed in this study are private owned. Two have 100
percent Mozambican capital, while the other two have majority foreign ownership.

The four millers interviewed vary substantially in terms of production capacity,
although the largest two millers are equivalent in terms of capacity. These two largest
millers are located around Maputo area in South, and together have 76 percent of the
market share, as shown in Table 3.6.4.5. The concentration ratio of top 4 millers (CR4) is
very high, about 92 percent.

Regarding the procurement practices, at the time of the survey, only two millers
were relying completely on maize grain from national production as the main source to
supply their mill. Not surprisingly, these two millers were the smallest in terms of

production capacity and market shares. In both cases, almost 100 percent of the purchases

187

“.1 .347. .4
'.
A.

 

were made at buying stations in rural areas, either from producers or informal traders. The
payment is made by check unless the amount due is less than one million Meticais (about
$US40), in which case cash is paid.

For millers purchasing local maize grain, all of them were bought in the province
of Manica, even though in one case some purchases were made at the province of Sofala.
Neither of these two millers had any agreement with somebody else to supply them
regularly. In addition they also do not pay any premium either for high quality or large
volumes of maize grain.

The other two millers interviewed (the large ones in terms of production capacity)
were mostly relying on imports of maize grain to supply their milling industry. The poor
quality of maize produced locally as well as the small volumes offered by national
producers were pointed out as the main causes for importing maize grain. Some other
inefficiencies in the local marketing system were also pointed out, with special emphasis
on the maritime ports which increases the costs to bring down maize grain from the
surplus areas to their plant. For instance, two millers mentioned that the cost per ton to
bring maize grain from the Northern region to the South is double the price paid for
bringing maize grain from South Afiica. In addition, the inefficiencies at the Nacala port
in the Northern region increase the amount of time that the product stay retained there,
which causes deterioration. Also, all maize grain purchased locally is in bags, which adds
costs during handling.

When maize grain is imported, South Africa is the main source of imports,

accounting for at least 70 percent of the total imports. The remaining 30 percent of the

188

 

 

 

total imports come mainly from USA, arriving at the Port of Maputo. The decision to
import from South Africa or USA. depends on the market condition. For example, in
2002, the price of maize in the region was very high due to imports from Zimbabwe. In
addition, the South African railway company decreased the number of wagons to
Mozambique. At the same time, they allocated more wagons to transport food aid to
Zimbabwe. Therefore, due to high prices in the region as well as the scarcity of wagons
from South Africa to Mozambique, one miller took the decision to import maize from
USA.

Throughout the year, both millers rely on future contracts in South Africa (a
standardized contract to take delivery of a standardized quality and quantity of maize
grain in a future date specified in the contract) as a way to guarantee a regular supply of
maize grain.

Vertical coordination between millers and traders is not widespread: Almost no
maize is obtained through long term contracts or integrated operation. However, all
millers agree that one motivation for the establishment of contractual relations should be
the acquisition of maize grain with desirable quality.

The lack of such coordination is explained by several reasons, with more
emphasis on small scale operation of national producers, and dispersed production, which
increases their transaction costs. Unless the scenario of small scale operation and
dispersed production could be reversed, any chance of long term relationship between
millers and maize producers will be unviable.

Almost all millers interviewed in this study are producing two types of maize

189

 

 

 

fl .__ ..
meal, namely refined and roller meal. Low demand by consumers for whole meal was
pointed out as the main reason for not producing it. This is explained by the cost
differential between the final price of whole and refined meal (or roller meal). Since the
cost differential is not significant between them, the miller’s managers believe that most
consumers prefer to buy refined meal (or roller meal) because most consumers in terms of
palatability prefer the later one.
i“

In addition, two millers mentioned that the lifetime of the whole meal is low in

the shelves, and thus, another reason for them for not producing it.

if!"
1

Furthermore, one miller criticized the relative “protection” given to rice. In fact, in
his point of view, if one household has to decide between whole meal and rice, the
decision will go to rice because its retail price is unbeatable as a result of economies of
scale on imports together with exemption on payments of certain taXes that maize grain

and derivatives have to pay. However, none of the millers interviewed could provide any

 

data to get more insights on this issue.

When the millers were asked about the destination that they give to the bran and
maize grain that does not meet their standards, three of them answered animal feed. Thus,
besides maize meals, millers tend to maximize the use of maize grain producing animal
feed. Probably, the observed growth of the poultry subsector turns the production of
chicken feed more profitable compared to the production of whole meal, this being
another reason to not see millers producing whole meal. However, in the absence of any

evidence of this fact, this issue needs further research.

190

None of the millers interviewed in this study sell their products directly to
consumers (no retail activity), being the wholesalers and retailers the main demanders.
However, only two millers have agreements to supply some of their demanders, mainly
formal wholesalers, by specific regular quantities and price.

To avoid an artificial increase on prices done by formal wholesalers, some millers
have established some warehouses in major populated areas to supply directly the
retailers and small informal traders, eliminating therefore the middlemen between millers
and retailers. This enables the retailers to avoid the high prices charged by formal

wholesalers, and thus, the consumers pay a relative lower price for maize meal.

Out of four millers, only one mentioned that there is some demand from
international organizations such as World Food Program (WFP). This organization buy
maize meal and distributes it into the vulnerable areas. The WFP’ requirements are very
high, something that may explain the low percentage of millers supplying them.

None of the rrrillers interviewed export maize meal. In addition, even though all of
them transport part of their production to other provinces (using roads or maritime
freight), the main target on sales is the province where the plant is functioning.

In overall, seasonality is a variable observed in the millers’ sales. The period of
high sales is between October and January because from February and August/ September,
the availability of maize grain is high (harvest period).

Quality of raw material is very important for all millers interviewed. Among

specific categories that are important for them, the maize color, level of humidity,

191

 

 

 

cleanliness, and the size of maize grain were pointed out as the most important.

Three out of four millers (miller number 2, 3, and 4) have their own laboratory
where they conduct all quality analysis, especially the humidity level. These analyses are
conducted mainly at the time that the product arrives in the plant, does not matter if maize
grain arrives there with a quality guarantee. In those cases where maize grain is purchased
locally, the analysis is also conducted at the time of purchases, that is, in the production
areas.

If the product arriving at the plant does not meet the quality standards for the
miller, it is mixed for animal feed. This activity is common in three millers, but never
happened in one of them (miller 4).

Regarding the needs for further research and extension services, the first point
mentioned by all millers was the need to improve the overall quality of local maize grain.
All millers believe that, with maize grain pursuing grades and standards, they can
increase the share of local maize on the total purchases. The second point was the needing
to evaluate the efficiency of maize marketing. For instance, the manager of one nriller
mentioned that sometimes the miller needs to buy maize grain locally, but they do not
know where to go to get it. Another manager mentioned that the millers’ margins are low,
but the investment is very high. Finally, only one miller mentioned a needing to identify
varieties that will be more profitable for the milling industry.

In terms of technical improvements to strengthen the milling industry, the first
one strongly supported by all millers was the need to adopt new production technology in

order to increase the production. The second was the need to develop and enforce grades

192

 

 

 

for local maize in order to increase competition between local and imported maize.

 

 

193

 

REFERENCES

Barrett, C. B. 2000. Measuring Integration and Efficiency in International Agp'cultural
Markets. Review of Agricultural Economics, 23 (1): 19-32.

 

Bopape, L. E. 2002. Analysis of Price Relationships Among Spatially Differentiated
Potato Markets in South Africa. Unpublished Master Thesis. Cornell University.

 

Brandow, G. E. 1977. Appraising the Economic Performance of the Food Indusg.
Lectures in Agricultural Economics. ERS/USDA.

 

Brester, G. W. and Goodwin, B. K. 1993. Vertical and Horizontal Price Linkages and
Market Concentration in the US. Wheat Milling Indusfl. Review of Agricultural

Economics. Volume Fifteen, Number Three.

Caves, R. 1992. American Indusg: Structure, Conduct, Performance. Foundation of

Modern Economic Series. Seventh Edition. Prentice-Hall.

Coughlan, A. T., Anderson, E., Stern, L. W., and El-Ansary, A. I. 2001. Marketing
Channel. Sixth Edition. Prentice-Hall. NJ.

 

Enders, W. 1995. Applied Econometric Time Series. First Edition. John Wiley & Sons,
INC.

Golleti, F. and Christina-Tsigas. 1995. Analyzing Market Integration. In: Gregory J. Scott
(ed). Prices, Product, and People. 1995. Published in Cooperation with the International
Potato Center.

Gujarati, D. N. (2003). Basic Econometrics. Fourth Edition. International Edition.
McGraw-Hill.

194

Holtzman, J. S. 1986. Rapid Reconnaissance Guidelines for Agg'cultural Marketing and
Food System Research in Developing Countries. MSU International Development Papers

No. 30. Department of Agricultural Economics. Michigan State University.

Imprensa Nacional de Mocambique. August 2003. Boletirn da Republica. I Serie -
Numero 33. Mozambique.

INE.1997. National Population Census. National Institute of Statistics. Mozambique.

INE. 2002. Agp'cultural and Livestock Census. National Institute of Statistics.
Mozambique.

INE. 2002. Statistical Year Book. National Institute of Statistics. Mozambique.

Intriligator, M. D., Bodkin, R. G., and Hsiao, C. 1996. Econometric Modele, Teehnics,
and Applications. Second Edition. Prentice-Hall. NJ.

Kinnucan, H. W. and Nelson, R. G. 1993. Verticel Control and the Farm-Retail Price
Spread for Eggs. Review of Agricultural Economics. Volume Fifteen, Number Three.

Marion, B. W. and Mueller, W. F. 1983. Industrial Organization, Economic Power, and

the Food System. In Future Frpntiers in Agg'cultural Marketing. Edited by Paul Farris.
Ames: Iowa State University Press.

Ministério do Plano e Financas. November 1998. Regplamento do Despacho de
Mercadorias. Diploma Ministerial # 206/98. Mozambique.

Mehta, M. 1989. An Analysis of the Stru_cture of the Wholefle Cereals Market in Mali.
Unpublished Master Thesis (Plan B). Department of Agricultural Economics.
Michigan State University.

 

195

Mendoza, M. S. and Rosegrant, M. W. 1995. Pricing Behavior in Philippine Corn
Markets: Implications for Market Efficiency. Research Report No. 101. International
Food Policy Research Institute (IF PRI).

Mendoza, G. 1995. A Primer on Marketing Channels and Margins. In: Gregory J. Scott
(ed). Prices, Product, and People. 1995. Published in Cooperation with the International
Potato Center.

Minten, B. and Mendonza, M. S. (1998). Price Behevior and Market Integgation. In:

Structure and Conduct of Major Agg'cultural Input and Output Markets and Resmnse to
Reforms by Rural Households in Madagascar. Part 3. Final Report. IFPRI.

MOA/MSU Research Team. 1993. The Organization, Behavior, and Performance of the

Informal Food Marketing System in Maputo. Working Paper #10. Ministry of
Agriculture. Mozambique.

Muwanga, G. S. and Snyder, D. L. 1997. Market Integgation and the Law of One Price:
Case Study of Selected Feeder Cattle Markets. Economic Research Institute. Department
of Economics. Utah State University.

Mocambique Editora. 2000. Codigo do IVA e do Imposto Sobre Consumos Especificos.
Primeira Edicao. Mocambique Editora, Lda.

Morris, M. L. 1995. Rapid Reconnaissance Methods for Diagposis of Sub-Sector
Limitations: Maize in Paragu_ay. In: Gregory J. Scott (ed). Prices, Produce and People.

1995. Published in Cooperation with the International Potato Center.

Muendo, K., Tschirley, D. L. and Weber, M. T. 2004. Imp roving Kenya’s Domestic
Horticultural Production and Marketing System: Current Competitiveness, Forces of
Change, and Challenges for the Future. Working Paper 08/2004. Tegemeo Institute of
Agricultural Policy and Development. Egerton University.

Pomeroy. RS. and Trinidad, A. C. 1995. Industrial Organization and Market Analysis:
Fish Marketing. In: Gregory J. Scott (ed). Prices. Produch and People. 1995. Published in
Cooperation with the International Potato Center.

196

Pyndick, R. S. and Rubinfeld, D. L. 1997. Econometric Models and Economic Forecasts.
Fourth Edition. Irwin Mcgraw-Hill.

Ravallion, M. 1986. Testing Market Integgation. American Journal of Agricultural
Economics, 68:102-19.

Rosenbloom, B.l999. Marketing Channels: A Management View. Sixth Edition. The
Dryden Press.

Rubey, L. M. 1995. Maize Market Reform in Zimbabwe: Linkages Between Consumer

Preferences, Small-Scale Enterprise Development and Alternative Marketing Channels.
PhD Thesis. Department of Agricultural Economics. Michigan State University.

Scott, G. J. 1995. Agg'cultural Marketing Research in Developing Countries: Old Tasks

and New Challenges. In: Gregory J. Scott (ed). Prices, Product, and People. 1995.
Published in Cooperation with the International Potato Center.

Staatz, J. 2003. Handouts for Analysis of Food Svstem Organization and Performance
Course (AEC 841). Department of Agricultural Economics. Michigan State University.

Tomek, W. G. and Myers, R. J. 1993. Empirical Analysis of Agg'cultural Commodig

Prices: A View Point. Review of Agricultural Economics. Volume Fifteen, Number One.

Tomek, W. G. and Robinson, K. L. 1990. Agp'cultural Product Prices. Third Edition.
Cornell University Press.

Tschirley, D. L. and Santos, P. 1999. The Effects of Maize Trade with Mal_awi on Price

Levels in Mozambique: Implications for Trade and Development Policy.
Working Paper #34. Ministry of Agriculture and Fisheries. Mozambique.

Wooldridge, J. M. 1999. Introductogg Econometrics: A Modern Approach. First Edition.
South-Westem College Publishing.

197

‘Y

   

11111111111111111113111