.11. 1 . n .1 1 (1.1 rt 1.. £- . . -1 .1 .. .. I1 . 1..1. 3 1.1. .1..(.;. . . .. . fi330fl3L3 11. I1 11... . 111.1.“ 1 .1 1 1 o . )u-1n1.b.111 u... 1.11:1.uh. 1 .1»... .113}. . 1 .1... . . It 1 111 1 1 ApOnl - PI . 1 1111.1.193131- . I...“ v.1 1 L - - rudvyltl- 1| 1- 1 .1. 11.1.1.1! 1 ii . l _ 3 .3 ‘l l 3 1 I 111 1 1 1 3 .313 11 3.3.3.1333... 1 C~F13l111 .- .1 . . .131? 1 111 fl11101 1 31 1 5.11111 1.111.141...”.111J1N1H111111m-111m~..1..:11d1.1.1.1.11 1 . HI1 1 >111 1-H? "1.13....“ .. . 1 .1 3 J . J13. .. 1 1 - 1. r 111111...1113. 1. .1. 4.11.11.11.11....1-:1111111.1111.fiuhac11h1&nn31.3H3 . .11 [MM-3.1. 1111.111. 1 .0! VI 4 A \C I 13‘ 1 1.311 a ~I3OVA3 II. .7131h11114110113 1333.11.31“ .Ifll.:1 1 . 1 . 11111 1 914m+m111311 11 .1 1 1 1.1.k11 1.1 11111 ..11|11.1.113¢1.1.1111 1111.11“. 173%“. . . 1 1 . 1 1 1 111 1....111 1111 1-1 13.1111 11.11.11 33 €1.33 . o '7‘ 3 3.. ...1.w11.t..41 hm1 11- - .11n.1.-11.-..nn..»....u1w 516.13.331.51... ..1 1......nu113311. 113.111.3113 1 11119... . .VA 71. .1 11nd“? Land-mu.- 1 1 .11-1.. .1w.11m11w1.1 .111111111.11111n.111111113u.-.1.1Hv1 ”1131111- .. 111.011.111.41 . 1.135.1- 1W. 1 1th..111..1.31.133.113.114uj1u 1.11. HI! 1 .1 .1: . 11 , 1.11111 1 . . 1.1.1111. . 3 :13 H“ 10. 11: 11 J11 .1314 fight” Ann-3. a. .rfia 1 .1 1.1 1111. . 1 1. 1613111 11.11 11 11! .11 11 11 . 1 3 o l I. «Oh . o .3 .1 14 11 I 3 d" .‘I‘.’~"O ‘ 3 .. IO‘, lfi 1 1119111“:ka . ..1\11 3 . . 1‘1 :11 111.11.”. < 4 1 . 1144.1...1Avu3dun 13 11 I 31h. 1 1 3.1.1 .1333 131363311 I 31.0.1.” 1‘ .1 1 11l 013.11 . 1- . 1 1711.5uv13-p11h11fi1w1u... . . E. .1 .. 11 1. 1. 1 1.3 11 1 1113 . . . 1 1 1 .1-W11vu-1r1. hump-1.1.. W1flfl. “hr-”.1114 "-111" 1 1133!}. )IlbUQ391313 1 {F1 11 123.1. .1 11.1.1 .U-ruu- 31 .13..1.13 311. 1L1J1H| ”41113.11 .11 .11 K1... 1 . . ghnkhhflmwbundif1317 1 “.11.. 33111331113131-1111 1 11311.” . 1.- ”flaunt. ran“ 11.11113. 1 D c 11 11 .3 '3. ’ 1 W113 .118 111.141 11?l.3 .1 1.11 31 Anni .131 1 111 .1 . 6113111111 In“! 11“! 1311113.} 1. 1143133333 11 I111 1-11 1:1 . 1&1- .I.&1fih.fifl3.§333¢%1%1¥131-H11- b 3133 1 1 1. 1111.3 . .. . 1111-11111...» . av”.- .11 11111.3. , L1! .111 131.193.. 1311133 1 11. E13911. . 31.111141315113311“ O 3 .1 1 1 . .x , 1.. 1.1.3.- . 1. 11.hr1u..-v1.1..»111131111..11"111 - 1. 1111. 1.11151 .fifl61ur13111r1131 111.13% 1.1 1.1;"...1 "”331” . £1011 1.1.3 1110111111111 A... .1Hu11n9411031fl3hnl33 .101-I3I1.11 1 . 1-1 . 111. .111F11. 11 .1 11 .1 1 1 1 O 11 o .hjlahw13h331 - 3 3313 .31 1 1 .1131. 133”! 11- 13 1 .. 1J1 . ”Kw-1.111111111111110. - 13.1.1.3}. 1333 3‘11 3 1 1 1 - 1 1 I. p. L133 “31 I . 1 1.1 1... 33 111133131. 1 1 1 I 1.1 1 1 0 1 11., 11 1 11 4 . 1- - 1 11 - . . .11 . 1 13133311331111 1 1 I. 111 K31I\A\1111.31311l3111.131.1333H.\I\31311 1 11‘ 1 31v 3 “P11011333 031n1u1333kb.113 11 11 33303 1 1 I 11.1.11131111J1111113 1111-11131 111111111111331-1311111111113.1131131111311113131111.11 . 1 1 111111 .1211.) 1.11.u1-1-..-.1111111.1.1111...\ .1111 . 314mg.» 1.1 .15 1 9.11 110m MVP-1313.- 1.h1 uni-1111.13.11”- .1-11111111111 11111-31311m1unr1111113. 111 .11 111 . 111111.31..h.1r..1n11111111.1 11.11..11111u.11.11.1..111 ..011.1 “11.1.1... 1 1.. 11- 111 3 11 .3. 1111 .1. .1-H13W1I1113311111311113- 31011.11! 131311 1.1.111. 1 . 31.1.1111... . H11 . 111.11.41.11. - 11311 1. 1 1111111111 .1311 .1 ..11 1 £1-HHH1-v -.11H.1|.r.H1131.n31 "14.331111111131031! 13%| 1 1 11.111313111111331. 1 [$111133 11.3.1..." L11”111.1.1111.1.. .1 111.; 111.111.“.P11 u . u11U11-11- . 11 1 131313.111 “3333111131 I131 1111,1133. 313111. 31...111|r1131.1.31.l11|\13111v«1131 31113311114339.3123 113 13.111111111114113 113111 1311.3 .1. 3M: 31111 1111. 11 .1111 1.1”» 13131 31111111331 1131131 133N333: I 13.31.31! 1 1111-1 1 11 11 1113 11111. . 1 1 1- 1- 11.1.1.1 13113113111111.1131 Jl.l.fl1111.31311.11111 11. 31111111111 1111 1.111111 1-11“ . ...- 1. . 11333 31131131113331.011HN3333131133-11133lull1311“ £13131 33‘3u3lvfiun33n33lnkfivufiiul“ 11331133111131.1131 I 1.131 31 1.11 .1 11 11 11 11.1111 11 1. 111111111 1. .1131 131111 13 133111111X! 1-1 . . 113101111 .3n3 1331“.” 11111 .111 .6111. . 11 1 T 11 11319131131111 1 .- 11131113313911 3131-11. 111133. 3.23113 11 III-13311 1 11111.31 34%|!) 11.111.11.113 1111.11.1111l.1-..1||..1|.111 .11..r.fi .11 1 s . 1 1 y 1131111w1-11n1111- 311-1. 111N1133313133l31l11111133333311111113111l33131131 3l3a.1131\|33|1311..11111.l3333131\1.\31|.1 1 .1 .. 1111113111111ufl11.113.1 (II11. .111 11111 1 1 . 1 .311 11111141111111.1131.- 311111111 1.1.11111u11-11111111111111.1111111111..u|1113111.111.\1...11.1 1-111.311.1111.” . . (1111.. .. 11 11-1. .1 .11 .1 33 - 3 I||III3| .3. 31 3 1 “m3.1313|3u111n1113\11.13-.N1111113p1\11 311119011 1.11.. 11.1.1”.1 1111AV11111 - .1 . 13333131335 1 33 . 1.11 1 3111 113 111 3 3 11 113 1111311 13111 .-11 "111131 F allfl1 “3 1 11111 11.131111 [H11 33313111331313 1.111113% 11 11 1.21.."- 1. H 1 333 113. 1 3\.. . 111- 31‘ 1-1 .1 I 3 3 I II - l I ‘1' Ilia. - 11-- 111. -1. . - ..11.-1 1 111 - 1 1 1 1. 1 ‘ 1 1 .1 1 11110 I 1.1 1 I 1 3 11 1 1 v 11 1 . 1 U 1 1 1 1. . v31 3 3 U 3 11 1. 11.1 1 1 v1 - u 11 Q 3 “1 \I 3 3 'l 1 I 3 3 3 _ 1 1.. 11 I11 1 I I I 1 11 1 a .1 . I111 1 .1 1. 1 1 1 3 1.1. 1 .I 1 3 1 a 1 n. .1. 1 . .11- . 1 1 N11,.“ , 1- 1h 311-1 1.11.11 11 131111 I 1 . 1 1-1. 1. I 1113..)1I3-3 4 1- . 3033133 31 “13.3 33313 33133133 I. 31‘ 1311l11 1331111011 1 11“- gl‘ll‘lllll 3 IE3 3 1311 113111 I3 . I _ II1 3311-1 .313 .111311 1133 33 l3|1h1333113 33131333333113 13 111313. 3313333333 13 v3 3 I33 131101113311 3 11333 33 111 .1 11.19.1131 11 11111113111133 1411.113 111.311 “H.313. .1 1| 1 3 3113. 13131 311.113.1-1..11.1 . 111111. .1 1- .?333 3113 3 11331333. 1. [3131333133311 134.1 11313 3 11.13 11 3 113131 11131 33 11111-1331 3311333I3 3 CI 11 33. |3\1.IL31133\I(333I”.\u11\33|33 11 1113333 3‘1 31111 3111 ‘33 31 3-! 3. 1 11‘. 111111111313 1- -11 - 1 1 11 .1111 1.11 I»! 111111 .1 11 -11 111.. 1113111111. - 1 1H131H11 3131111311131 . 11 1 . 11 11 1u.1.1 .- 13113 no 13 311 11111113111111 1. 1 Pl .1 11 1 111. 1:... "(11 10.111111.“- 1 . 3. 3313133113 11-1 11. 1 13 1 313311.11 “313331113313 11111 11.10411111- 11111111311131 11.11.111.111 1131. “1111111111 1 . .1 - -..1.-11 1 f .. 1111.1 - .1111 IH1-4111111111111r 1111 .1- II\1(13.|1 .3131011 11111111113111111fl11315131113311n113h11111113 11 111.1- inn-11% 1 1 1.. .. 1 1111..“- 11. ....1.11-l1 1111 3 14111. 1 \111 I333: 1. . ,13 3 31313111133 13! I1 311113 1331313 1.1.33 .1 11 11 “1 ll. L13111 .1 . (11111-1- 3n3PNW1131-1111111111P1N111fi111 . . 31-1b-ufl1131310311313 .u1 $1.1”.- 11..111 111111.- -1u 111.1 111-1 -11. 1...1-1-1111113111Pc"11|..-.l113fl.h1.|, 11 1 .111 11.1 11-11111-- 11 11111111111.| 11. 1.111 1111111 1. 11 16.“.- 1 E1111 .1111131111. 1-111»..-- 111. 1111.1 -MUI .1311 1-1.. .1111 1....11 111.13%..113? . -. 333111133.1.U.Hh113\11 1 £3,11- 11. 131111131113 11 . 111 . 13\1111 3 111 1r11-11 an»! 111111111311". ...1111|11--.111 .1 1 13111-111 33333.1...131. no . 11 1 ”113.1111! 1 11111. 111, 11.131311111111311 111111U .111 11331111111111. .11u11.nu “11131111113111.1h..11..111.11 1.. .11....1-1 .1. Luv-11H 1 ,h111 13111313311131 13-.111111 1 1111. g311fik1u31313133- «flu-111mm,?“- .h.1 - . 1-... .-1-11fi....)b1 1 .111. 1111-1 . 151113.111. AWN-133.1111 .131- 15311-11111" 1- (131113.111? 31331333311331.11HNH11. -:.....111v.11.l1...11..h1.1111 .111 333.330.3113 1:111 J43 1-131 . 11133.1 3|. 1 1 “~13. 3“. 3113M3 141111.114MII33 331.1 311.31 M .1-1 1111 11131313331111.- 31 1133“"! 3 1 1 0,1113% .. 1 33131193.! 113.111r 1U11111 :1.11 11... 1 I11- . 1 .1111113111 1111 Quit-3.131131 13333“:3 11.1111 1 33? 3 3333 111 11 - 11311 111.1 1 03.1.1 31. 4W3 13 ”“33 311.3313!) 4:33 11 I 111l 1. 1 1 1 1 1131”! 111WI111114H411411fl3 "13.313133313313331 . I 1 \ .11 1 1 1 1| 1 1 3 13 l 111. 113131 333W3O 311313341 133331th 3131“ 1 1 3| 1 1 113 1 3 ”M1333 5 31 I3 \‘I‘3 313‘ 3| 1 -uunnunde-uwwwnfiulur .1H11111\.111u11\.h.1n111111“|111.w. 11 11 31 11-3 11.. 31331 31139: 311 3 . 1 . 1-- 11111 . . 111-n1u1v111111.1.1111.1111.,1..- 11. 1171111 w 1 1.1U11 111-1-11nu11111mufl. $9.1141113 . 1-11.l1.3111.1 1.1.4.1311 . 11 MI 3111 1.3131." -1111 $111.41 11133111131111.1111 111.1%“? . .Qw11..311u11».11-,m 111 1h31111l1.11-11.1a111...1 111.1H1111111Uh11rsmflu. - . _-1-1J.r1.1uu11111u 13-1 ,1 111-1111331111. 7.111111 . «9.... :1)... aw . 3n1u11 u.“ 1 . 1 3.3 11 1113”” . 1 .11 1 . . . . 1H1... 115. 11»! ”13.111.113.11... n14.11111- . . . | 133310. E THES‘S 7, ‘t' “ . "i Vi "if i ‘t ‘1 5., . Paw ,‘In*."’) { F -" :1 f”; '5 d t3 l g? .— m?! \ .“’_w VUUbJ ‘ i i This is to certify that the dissertation entitled MARKETING OF FOOD CROPS AND INPUTS: THE CASE OF FUNTUA AGRICULTURAL DEVELOPMENT PROJECT IN KADUNA STATE, NIGERIA presented by Salisu Ahmed Ingawa has been accepted towards fulfillment of the requirements for l l PIL DI degree in Agricultural Economics %WA in wall r Major professor "7 Date (it? 2 2 / 9513 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 ____ ._. ._V __ _.V _.._ r--__.—v——— , ,7 ,7 , i . i MSU LlBRARlES .-__. RETURNING MATERIALS: Place in book drop—to remove this checkout from your record. FINES wiIl be charged if book is returned after the date stamped below. ISA a ='I «7‘ If ’ K'— 3’ "N79” . I ,. . é -‘ r .5 $373. ti. :1. ‘ ".I ”J .' "u .- "f‘ .- '. - " \-:".‘./ a“ ‘_J" .3 E V' fl : 3‘36: '5". 2- a. war “I" -_ ”:1 H ‘L:\ it: 13" *N‘ -"'z7 . 'r .. we :fiv‘r ' . .--‘ i" . 7- no .9” It] . ““33"” v ‘ 0- - I. i «‘5‘ 'f 1.“ ”Ii .1 ""1- {1 COPYRIGHT BY SALISU AHMED INGAWA 1983 MARKETING OF FOOD CROPS AND INPUTS: THE CASE OF FUNTUA . AGRICULTURAL DEVELOPMENT PROJECT IN KADUNA STATE, NIGERIA By Salisu Ahmed Ingawa A DISSERTATION Submitted to Hichigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1983 ABSTRACT MARKETING OF FOOD CROPS AND INPUTS: THE CASE OF FUNTUA AGRICULTURAL DEVELOPMENT PROJECT IN KADUNA STATE, NIGERIA By Salisu Ahmed Ingawa The poor performance of the Nigerian agricultural sector in the late 19605 and the first half of the 19705 prompted a large number of policies and programs to revitalize the sector. However, marketing problems have emerged that threaten to check the gains being made through large-scale agricultural development projects. This study evaluated the performance of the marketing systems for food crops and farm inputs in the Funtua Agricultural Development Project (FADP), the first of several large-scale ADPs in Nigeria. The evaluation ' of the input procurement and distribution system also considered per- formance at the state and national levels. Data for the analyses came from Agricultural Project Monitoring, Evaluation and Planning Unit at Kaduna, the FADP evaluation unit at Funtua and from secondary sources. The evaluation of the food crop marketing system was largely based on examination of price behavior over time and space. Results showed that food crop prices declined significantly during the last two years of the project, especially for maize which was a relatively new grain crop in the area. Seasonal price fluctuations were larger than in previous studies and seasonally highest prices were occurring much i..- v" m: H. . Salisu Ahmed Ingawa earlier than previously reported. Comparison of seasonal price increases with storage costs indicated that short-term storage would have been profitable. Correlation analyses showed that farm level prices of food crops were strongly correlated among the various FADP districts. Comparison of price spreads with transportation costs indicated that the price spreads did not depart significantly from transportation costs. It was concluded that the food crop marketing system performed reasonably well especially in terms of price correspondence among the area districts. However, the system did not perform as well in terms of temporal price behavior, particularly for maize. It was recommended that the planning of ADP-type projects include a careful assessment of demand prospects for increased crop output and that storage, processing and market information needs be anticipated and provided for in the pro- ject's operational plan. - The FADP fertilizer distribution system functioned effectively through a network of farm service centers. However, the heavily sub- sidized national fertilizer procurement and distribution system failed to provide adequate quantities and timely product delivery. Decentraliza- tion of the fertilizer procurement function and a revision of subsidy policies are recommended. Dedicated to my father, Alhaji Ahmadu Ingawa and in memory of my late mother, Binta Fatsima ii ‘ ACKNOWLEDGEMENTS I wish to express my sincere gratitude to Professor Harold Riley, my major professor and thesis supervisor, for his guidance, encouragement and constructive criticism. I also wish to thank Dr. Eric Crawford for his helpful criticism and suggestions for inprovement. I gratefully acknowledge the contributions and assistance of the other menbers of my comnittee, Professor Carl Liedholm and Professor T. Schillhorn Vanveen. The support, guidance, and understanding shown by Professors Carl Eicher, David Norman, and Lester Manderschei d, throughout my graduate ' program is deeply appreciated. I wish to thank members of the Department of Agricultural Economics and Rural Sociology, Ahmadu Bello University for friendship and suggest- Ions. Assistance given by Professor George Abalu, Dr. M. S. Krishnaswamy and Dr. M. R. Raza is also appreciated. The encouragements of Dr. Jacob VOh, toward the completion of this study is appreciated. I am grateful to the World Bank, the Agricultural Project Monitor- ing and Evaluation Unit of the Nigerian Federal Ministry of Rural Devel- Opment, and the Funtua Agricultural Development Project for their consent, data provision, and assistance in conducting this study. At FADP I am particularly grateful to Al haji Ahmed Alti, the project manager, and to Al haJ'i Mohmed Aminu, Alhaji Mohmed Abu, and Ibrahim Kwajaffa, members of the Project evaluation unit. At APMEPU I wish to specifically thank Ray Robertson, Derek Poate, Joseph Bivins, M. G. Daplyn and Peter Olorunfemi for their help and patience in explaining and preparing the data. The assistance of Mr. Roger Slade of the World Bank is gratefully acknowledged. I am grateful for financial support from Ford Foundation and Ahmadu Bello University. The help of Paul Nolberg and Chris Wolf in setting up the data for computer analysis, as well as their readiness to help at all times, is deeply appreciated. I also thank Eleanor Noonan for initial typing and useful advice on secretarial assistance. Lois Pierson's enthusiasm in final editing and typing of the final draft is greatly appreciated. Thanks to Barbara Miller for good naturedly typing many of the tables. I benefitted greatly from the developing technology in microcomputing. Without the Osborne I the revisions, deletions, additions, etc. would have been much more painful. Finally, I owe a debt of sincere gratitude to my wife, Hadiza, and our children, Nasir and Nafisa, for their patience and understanding. iv TABLE OF CONTENTS Page LIST or TABLES ....................... ' . . viii LIST OF FIGURES .................. i ...... x LISTOFMAPS ..... ............ xi LIST OF ABBREVIATIONS ..................... xii CHAPTER I: -INTRODUCTION .................... 1 Background ....................... 1 Trends in the Nigerian Economy ............. 2 Gross National Product ............... 2 Population and Urbanization ............ 5 Agricultural Production . . ............ 6 -Food Prices and Imports ............... 6 Earlier Food Marketing Studies in Nigeria ........ 10 Agricultural Development Projects ............ ll The Special Place of Maize Under ADPs .......... 12 Objectives of the Study ................. 14 Organization of the Study ...... . .......... 15 CHAPTER 2: REVIEW or THE LITERATURE .......... '. . . . 17 General Approaches to the Study of Food Marketing . . . . l7 Structure-Conduct-Performance Paradigm ......... 20 Measures of Pricing Efficiency in Food Marketing . .1. . 22 Food Marketing Studies in Nigeria ............ 24 Methods of Analysis ................ 25 Results of the Studies ............... 27 Policy Recommendations ............... 28 Criticisms of Nigerian Food Marketing Studies ...... 29 What Do Correlation Coefficients Indicate? ..... 3l «Ii-9'1 lg. "All. CHAPTER 3: INTEGRATED AGRICULTURAL DEVELOPMENT PROJECTS . . . . 35 ‘ General Discussion on Agricultural Development Projects ....................... 35 Integrated Agricultural Development Projects in Nigeria ........................ 4l Conceptual Framework on Agricultural and Rural Development Projects ................. 42 Government Policies Relating to Agricultural Development Projects . . .. ............ 45 Design Elements of Nigerian ADPs .......... 46 Number and Locations of ADPs in Nigeria ...... 47 Activities of Nigerian ADPs ............ 48 Major Problems Associated with ADPs ........ 50 CHAPTER 4: STUDY AREA, FADP AND THE DATA ........... 52 Study Area and Project Background ............ 52 Funtua Agricultural Development Project Activities . . . 56 The Data and Data Collection Methods .......... 63 Producer and Market Price Surveys ............ 67 'Producer Price Surveys ............... 67 Place of Produce Sale ............... 70 Market Price Surveys ................ 72 CHAPTER 5: INPUT PROCUREMENT AND DISTRIBUTION SYSTEM ‘ ..... 74 Introduction ............. - ......... 74 Channels of Fertilizer Procurement and Distribution 1967-76 and 1976-80 .................. 78 Costs in the Procurement and Distribution of Fertil- izers ......................... 82 Domestic Fertilizer Production ............. 83 Market for FSFC Fertilizer ............... 85 Fertilizer Demand in Nigeria .............. 86 Kaduna State System of Fertilizer Procurement and Distribution, Befbre and After I976 .......... 87 Kaduna State Farmers Supply Company ........... 92 Input Procurement and Distribution at FADP ...... .' 95 Relative Importance of Farm Inputs at FADP ....... 97 Other Farm Inputs .................... 103 vi CHAPTER 6: INTER—TEMPORAL GRAIN PRICE RELATIONSHIPS AT FADP . . 104 Time Series Variations or Movements ........... 105 Trend Analysis ..................... 106 Visual Method ................... 106 Trend Analysis Using Regression Methods ...... 112 Seasonal Analysis of Prices ............... 115 ' Estimation of Seasonal Price Increase Using Regression Procedures ................. 127 Conparison of Seasonal Price Increase with Estimated Cost of Storage .................... 131 Sumnary ......................... 141 CHAPTER 7: SPATIAL PRICE ANALYSIS ............... 143 Spatial Price Analysis . . . . . . . . . . . . ..... 144 Data Problems Relating to Spatial Price Analysis . . 147 Villages in Spatial Price Analyses Studies ..... 148 Spatial Integration of Staple Food Prices within FADP Distri cts .................... 149 Interdistrict Price Spread and Transportation Costs . . . 165 SUIImary ......................... 168 CHAPTER 3: SUMTARY AND CONCLUSIONS .............. 169 ,Data and Limitations of the Study ............ 171 Agricultural Input Distribution System ......... 172 Temporal Price Behavior ................. 176 Spatial Price Behavior ................. 178 Conclusiohs and Reconmendations ........... ' . . 180 Marketing System for Staple Food Crops ....... 180 The Marketing System for Inputs .......... 182 Project Planning and Implementation ........ 184 Suggestions for Further Research .......... 186 APPENDIX A: RESULTS OF ESTIMATING TRENDS IN ACTUAL UNDEFLATED FOOD CROP PRICES AT FADP, 1976-1979 ........ 188 APPENDIX B: STATISTICS ON MAIZE AND OTHER CEREALS ....... 192 APPENDIX C: SEASONAL AND INTERCROP PRICE ANALYSES, 'ZARIA AREA . 201 APPENDIX D: GUARANTEED MINIMUM PRICES FOR FOOD CROPS ..... 202 APPENDIX E: " CONSUMER PRICE INDICES USED TODEFLATE 'POOO EROP ' fl 7‘ PRICES, 1976-1979 ................. 202 APPENDIX F: ACTUAL FOOD CROP PRICES AT FADP, 1976-1979 . . . . 203 APPENDIX G: EXCHANGE RATES, NAIRA PARITY NITH U.S. DOLLAR, 1973-1979 ..................... 207 REFERENCES ........................... 208 ¥ vii .- Table 1.1 1.2 1.3 1.4 4.1 4.2 4.3 5.1 5.2 5.3 5.4 5.5' 5.6 5.7 5.8 6.2 6.3 6.4 6.55 LIST OF TABLES Major Components of GNP (percentages) .......... Area Planted, Production and Yields of Some Major Crops in Nigeria (1971/72-1977/78) .......... Mean, Standard Deviation, and Coefficient of Vari- ation of Total Production, Areas Planted and Yields, 1971-1978 .................. .Food Imports, 1970-1978 (millions of naira) ....... Distribution of Population by Districts, FADP, 1975 . . . Major Surveys Carried Out at FADP ............ Frequency Distribution of Sale Location for Staple Foods , FADP ..................... Fertilizer Consumption by States, Nigeria, 1979 (tons) Fertilizer Distribution in Kaduna State by LGA, 1976-77 to 1979—80 (in 1,000 tons) .......... Funding Allocations by Zones, Kaduna State IRDP, 1981-1985 (millions of naira) ............ Input Sales at FADP, 1976-1980 . . .' .......... Farmer Use of FSCs for Input Purchase, FADP, 1978 . . . . Purpose of FADP as Seen by Farmers ...... - ..... Purchases of Inputs from Farm Service Centers, FADP . . . Fertilizer Purchase by Progressive and Non-Progressive Farmers, FADP .................... Regression Equations for Trend Estimates in Real Producer Prices in FADP Districts, 1976-1979 ......... Further Analysis of Seasonal Price Indices . . .l. . .'. Regression Results Fitted to the Upward Sloping Portions of Average Seasonal Producer Prices for Food Crops at FADP, 1976-1979 (Prices in real 1970 value) ..................... Regression Results Fitted to the Upward Sloping Portions of Average Seasonal Producer Prices of Food Crops at FADP, 1976-1979 (Nominal 1976—79 Prices) ....................... Harvest Prices Plus Hypothetical Storage Costs fOr Food Crops'FADP, 1976-1979 .............. viii Page 4 7 '53 64 71 91 94 97 99 99 100 136 7.1 Correlation Matrix for Monthly Producer Prices of Sorghum)Farfara Among FADP Districts. 1976—1979 n = 41 ....................... 151 7.2 Correlation Matrix for Monthly Producer Prices of Sorghum Kaura Among FADP Districts, 1976-1979 (n = 41) ....................... 152 7.3 Correlation Matrix for Producer Prices of Maize Among FADP Districts, 1976-1979 (n = 41) ....... 153 7.4 Correlation Matrix for Producer Prices of Millet , Among FADP Districts, 1976-1979 (n = 41) ....... 154 7.5 Coefficients of Price Variation fOr Staples, FADP Districts, 1976-1979 (percent) ............ 155 7.6 Comparison of Correlation Coefficients of Absolute Monthly Prices and Monthly Price Differences, - FADP, 1976-1979 ................... 160 7.7 Correlations of Staple Food Producer Prices at FADP ' Calculated on a Yearly Basis ............. 162 7.8 Distribution of Correlation Coefficients of Producer ‘ Prices fOr Staple Crops at FADP by Location, 1976-1979 ...................... 163 7.9 Producer Price Differences between Bakori District and Other Districts, FADP, 1976-1979 (Among less than Bakori district in naira per ton) ........ 166 7.10 Expected Transport Cost between Bakori and Other District Headquarters (Naira per ton) ........ 167 ix 6.7 6.8 6.9 6.10 6.11 LIST OF FIGURES Imports of Fertilizer in Nigeria, 1965-1980 ...... Channels of Fertilizer Distribution, Nigeria, 1967-1975 ..................... Channels of Fertilizer Distribution, 1976-1980 . . . . Farfara Producer Prices at FADP, 1976-1979 ...... Kaura Producer Prices at FADP, 1976—1979 ....... Maize Producer Prices at FADP, 1976-1979 ....... Millet Producer Prices at FADP, 1976-1979 ...... .. Food Crop Prices at Malumfashi, 1976-1979 ....... Seasonal Price Indices fOr SorghUm Farfara. 1977-1979 . . Seasonal Price Indices for Sorghum Kaura, 1977-1979 . . . Seasonal Price Indices for Maize ........... Seasonal Price Indices for Millet, 1977-1979 ..... Seasonal Indices of Producer Prices for Food Crops, Malumfashi District, 1977-1979 .......... . Actual and Expected Prices for Farfara, Malumfashi District, 1976-1979 ................ Actual and Expected Prices for Kaura, Malumfashi District, 1976-1979 ................ Actual and Expected Prices for Maize, Malumfashi District, 1976-1979 ................ Actual and Expected Prices for Millet, Malumfashi District, 1976-1979 ................ Page 77 79 Tilt" 3.1 4.1 LIST OF MAPS Eggs Nigeria: International and State Boundaries ....... 3 Administrative Map of Nigeria: .............. 49 Towns and Villages, FADP ................ 54 xi ABU ADP APMEPU ADC EADF' FFPlJ - FS(Z IAN! KAFscOM ' NAm=rqi New: OFN LIST OF ABBREVIATIONS AND ACRONYMS Ahmadu Bello University Agricultural Development Project Agricultural Projects Monitoring, Evaluation and Planning Unit of the Federal Ministry of Agriculture Agra—Service Center Funtua Agricultural Development Project Federal Fertilizer Procurement Unit 'Farm Service Center Institute for Agricultural Research, Ahmadu Bello Univer- sity, Zaria Kaduna State Farmers Supply Company National Accelerated Food Production Program Nigerian Grains Board Operation Feed the Nation Currency Units 100 Kobo = One Naira Note: 1 Naira.= 1.596 U.S. Dollars See Appendix G for trend in Exchange Rates. xii CHAPTER 1 INTRODUCTION Background Nigeria is implementing agricultural development projects throughout the country as a means of solving its agricultural and rural development problems. The investments in monetary and manpower requirements are imnense. These projects are intended to have substantial effects not only on the agricultural sector, but the. rest of the economy as well. One area such projects will affect substantially is the marketing sSistem for staple food crops. This is due to an euphasis on food crop Production which reflects the. government's concern with large increases in food iiipo‘rts.1 Major concerns include the coordination of project activi ties with those of the private, traditional staple food marketing s.Ystemumarketing inputs for. farm production and the output from the Phojects. The purpose Of this study is to investigate the marketing Of staple 1"(Jed crops and inputs at Funtua Agricultural Development Project (FADP), One of many such agricultural development projects in Nigeria. Funtua Agri cultural Development project is in Kaduna State, one of the 19 i 1Nigeria's food inports increased from less than 60 million naira 1" 1970 to over 1 billion naira in 1978, in current prices. When an ihfiation rate of 17 percent per year is taken into account, the increase 3 much less, but still remarkable. See Table 1.4. 52:95 in i success 119'" pm 533‘ i, t if :f‘eri IN Am i: fist “21".: ""b I i" ‘ ‘W‘igri ' O I“ cl" states in the country (Map 1.1). Productivity of small farmers cannot be successfully increased without an effective marketing system for both their products and the inputs they require. An effective marketing system, which encourages production of the right agricultural products by offering incentive prices, is essential. FADP planners were confident that the traditional, private market- .ing institutions in the area would be able to adequately handle the marketing Of the anticipated increases .in production of staple food Crops. However, this was not the case in input marketing where the Project incorporated elaborate arrangements for the marketing of i nputs.3 Trends in the Nigerian Economy Gross National Product The Nigerian economy has been growing rapidly in terms Of national income'over the 1970's. ‘ The Gross National Product (GNP) grew from ab<>Lat N9 billion in 1971 to about N30 billion in 1980, as measured in "931 1974-75 prices (Table 1.1). A recent World Bank report indicated that Nigeria had a per capita GNP of 1,010 dollars in 1980. The growth I" GNP per capi ta between 1960 and 1980 was an average 4.1 percent per \ 2M. S. 0. Nicholas, Foreward to "The Private Marketing Entrepreneur an¢1 Rural Develo ment." FAO Agricultural Services Bulletin No. 51, ( nne: FAO, 1982 , p. 1. i 3FADP planners pointed out that the system for staple food market- "9 was highly organized but fragmented. Marked spatial and tenporal :F‘i ce irregularities occurred. Based on this, the project was planned ‘3 [gravide market intelligence service, crop assembly, and marketing ng1 ce to help farmers improve their bargaining position. See The World (fink. Appraisal of Funtua Agricultural Development Project, Nigeria, aSihington, D.C.: The World Bank, 1974K Annex 2, p. 9. Page 9.: .2. SEE: "Egg 3 e: =.ow\mnmpim~\¢~m_ “avgomrz mo postage ovumaeoo mmOLw=i.mo*umwumpm we oo_mmo Pmsmcom mmw .a .mnm— .moam4 .uemEgo_o>mn oweocoou to xgpmwcvz Faceted .omapimnmp .cmpa oceanmho>oa paeopumz ugvzp .opgompz.mo o_pa:aom Fondue; "mooszom ._~\o~a. u _~mp .mmvzaxps .oe\m~a_ to uopcoa.u=*u==ooun ago so» ”magnet mu twosome; appasuou one camp Low moczm.u .cmo>_gmu:opmo a gap: acoumwmcoo ac: m? vowgon acpuczooom ogh "ouoz cowauupcaeeou ¢.m p.m m.~ o.~ ¢.~ m.~ —.N _.N m._ m.~ a “soamccgp N.m w.n m.~ N.“ 0.x n.m m.o ¢.n —.m ~.m copuanrgumwo m.pp m.m m.m ¢.~ m.o ~.m e.m n.¢ —.¢ m.n .xzpuagumcou a mcpcpvsm m.o m.o m.o ¢.o ¢.o ¢.o e.o . e.o m.c m.o Loam: a xuwuvsuuopu c.m F.n m.c m.m —.m ~.e m.¢ m.¢ p.¢ o.m wgauun$scmz p.5m F.mm F.p¢ m.~¢ e.¢e m.m¢ —.me. «.me m.mm F.mm mcvxggmao a acwcvz ¢.mp w.mp o.—~ o.- n.NN «.mm ~.¢~ m.- c.~m o.om mszupsupsm< Tow 7mm méu map mé— Y: 72 c.- ~.: ed A232 5:23 . . mzw punch 82 32 39 :2 32 map. :2 «KS 1/ ~22 p3— «5:828 Amommucoogoav azu to mucocogsou some: I use .Igr yin l n 9" '1' 'I - .. Qt. l u. 1‘.” year (World Bank, 1982, p. 110). This places Nigeria in the ranks Of middle-income countries in Africa--a distinction held by only a few. The structure of GNP shows one striking change: agriculture's con- tribution fell from 36.0 percent of GNP in 1971 to 18 percent in 1980. On the other hand, the contribution from petroleum extraction, as shown under Mining and Quarrying, rose from 33 percent to 46 percent in 1975 and then fell to 37 percent by 1980. Building and construction showed a steady growth from 1971 to 1980. Changes shown in other sections of the economy are less significant since the changes are much smaller in relative terms and data is generally 01' questionable quality to allow for differentiating between small changes. The deeline in the manufacturing and distribution categories are undoubt- edly questionable based on casual Observations of the economy. Population and Urbanization By the mid-1980's. Nigeria's population is estimated to be 84.7 million . which converts to an annual growth rate of 2.5 percent. It is p"Skirted that the population will be growing at 3.4 percent per year be- mée" 1980 and the year 2000, giving a projected population Of 119 million I" 1990 and about 170 million in the year 2000.4 The growth in population is also acconpanied by rapid urbanization. Urban Dapulation as a percentage of total population increased from 13 Percent 'in 1963 to 20 percent in 1980. The percentage of urban people ”ring in cities of over 500,000 has grown from 22 percent in 1960 to 58 Percent in 1980. The nunber of cities with over 500,000 inhabitants grew 1“ Wu 1960 to nine in 1930.5 ”Dive The World Bank, World Development Report, 1982. (New York: Oxford ”Sitar Press for the World Bank, 1982), pp. 142-148. 51b? (1. 4'1"" HUI-i If! 6 r Hr r.- II" (I, II. 6 The population growth described, coupled with the rates of urbanization, are putting considerable strain on the domestic food dis- tribution system. In order for the marketing system to continue with the level of efficiency observed by earlier researchers, it has to adapt to the emerging circumstances. Agricultural Production Agricultural production generally stagnated in the 1970's, and in some cases, even declined. Production of the major cereals--sorghum, millet and maizenhas declined over this period. However, more marked decline is evident when considering so-called cash crops--groundnuts and cotton. Pia"ted areas of both cash crops and food, crops have also declined as dePl Cted in Table 1.2. However, yields of all crops have generally i"Plt‘f’ved slightly. Calculation of the coefficientof variation with ”Spect to production, areas planted and yields shows that production was "lore unstable than either areas planted or yields (Table 1.3).6 This “"1136 r-ative instability of total production indicates the flexibility °f Sh'i fting from one crap to another, from year to year, as well as 1“” uences of hostile production environment in terms of the weather and ”9331 onal oUtbreaks of pests and crop diseases. Food Prices and Inports The average annual rate of inflation for the period 1960-69 was 2.6 Per-cant; however. by the next decade Nigeria experienced a dramatic \ , 6These figures should on 1y be regarded as indicative. The quality of n— 3!: data is poor. It is widely believed that production of maize has shOBSed considerably in many areas during the period of the 1970's. This "7“ have been reflected in the national average. Table 1.2 Area Planted, Production and Yields of Some Major Crops in Nigeria* (1971/72-1977/78) Crop 1971/72 1972/73 1973/74 1974/75 1975/76 1976/77 1977/73 I Mi 1 let Production 2835 2391 ' 3794 5554 4737 2893 2579 Area 4788 3692 5651 4787 5478 3939 3090 Yi e1 d 592 648 671 1160 860 736 834 Sorghum . Production 5794 2298 3125 4738 . 3328 2950 3327 Area 5387 » 3792 5516 4653 5721 4842 3480 Yi e1 d 704 606 567 1018 . 581 609 956 Mai ze Production 1274 , 639 - 809 528 1332 1075 758 Area 1197 1050 1130 579 971 892 610 Yield 1064 609 715 912 1372 1205 1243 Groundn uts Production 1381 1350 878 1946 449 459 557 Area 1 796 2032 2076 1796 1472 684 755 Yi e1 d 769 665 -- 423 1084 305 671 737 Cotton Producti on 426 105 85 481 81 294 269 Area 798 236 l 21 478 . 197 384' 278 Yield 533 445 705 1006 411 765 968 *Area in 1000 hectares, production in 1000 tons and yield in kg/ha. Source: A. O. Falusi and L. B. Williams, "Nigeria Fertilizer Sector: Present Situation and Future Prospects." IFDC, 1981. increase in the average annual rate Of inflation--18.2 percent] A num- ber of factors are responsible for this marked increase in inflation rates. Two of these factors are: (1) the rapid increase in incomes, and (2) the decline of agricultural production. Food constitutes a large prOportion of the average household expenditure; therefore, 7The World Bank, op. cit., p. 110. Tab1e 1.3 thean, Standard Deviation, and Coefficient of Variation of Total Production, Areas Planted and Yields, 1971-1978 Production Area Planted Yield Cm” Mean SD cv Mean so cv Mean S0 cv (%) (H (M Mi 1 let 3540 1205 34 4489 949 21 786 192 24 Sorghum 3651 1196 33 4470 864 18 720 188 26 Maize 916 314 34 918 242 26 1017 284 28 Groundnut 1003 573 57 1516 579 38 665 252 38 Cotton 249 165 66 356 228 64 690 240 35 Notes: $0 is standard deviation CV is coefficient of variation and is calculated as the ratio of standard deviation to its mean Source: Calculated from Table 1.2. shortfalls in fbod production translate into increased food prices resulting in higher rates of inflation in the economy. Food prices are , one of the major contributors of a high inflation rate in Nigeria. The country is becoming increasingly dependent on imported food. Many feel that the oil market, which the country depends on for balanc- ing its foreign exchange, is very unstable. And yet the Oil market provides the means of payment for imported foods. Thus it is apparent that-increasing food imports is not the answer to Nigeria's food problem. Food imports made up 12.4 percent of total imports in 1978; this trans- lates into 7 percent of Gross Domestic Product (GDP) for that.year. The recent historical trend in the value of food imports is shown in Table 1.4. Table 1.4 Food I orts, 1970-1978 (mill ons 0f naira) Year Food Imports Food Imports Food as Percent‘ CPI Nominal Naira Real 1970 Naira Total of Imports 1970 57.7 57.7 7.6 100.0 1971 87.9 76.0 8.1 115.6 1972 95.1 79.7 9.6 119.3 1973 126.3 100.5 10.3 125.7 1974 154.8 108.6 8.9 142.6 1975 298.9 156.7 8.0 190.8 1976 440.1 190.4 18.5 231.2 1977 780.7 277.9 10.4 280.9 1978 ‘ 1108.2 385.3 -12.4 287.6 Source: Import figures including percentages from B. U. Ekuerhare, _ "A Theoretical Framework for the Economic Appraisal of the Green Revolution in Nigeria." Paper, First National Seminar on the Green Revolution in Nigeria, ABU Zaria, September 21- 24, 1981; real value figures calculated using general Consumer Price Index from Central Bank of Nigeria. Shown in last column. Increasing food imports and rising foodprices have prompted serious efforts by the government to channel revenues from the Oil sector into food production. Various programs have been initiated and there seems to be some confusion as to the appropriate policies needed to reverse the situation. Several of the government-sponsored programs are: the Opera- tion Feed the Nation (OFN), the National Accelerated Food Production Pro- gram (NAFPP), the Integrated Rural Development Projects, the Guaranteed Minimum Price Scheme for Food Crops, and the latest addition, the Green Revolution Program. The Guaranteed Minimum Prices have not served as 10 incentives to producers so far since they have remained well below the market prices. The case of maize is even more striking (See Appendices D and F). Earlier Food Marketing_Studies in Nigeria A number of food marketing studies were carried out under the guidance of W. 0. Jones of Stanford Research Institute. These studies were part of a larger study researching the staple fOOd marketing systems in Nigeria, Kenya and Sierra Leone. The studies set out to appraise the 'efficiency of the staple food marketing systems and to identify ways in 8 which their effectiveness might be enhanced. The general findings of these studies, and similar studies conducted later, are that the marketing systems for staple food crops are Operating in a competitive manner.9 Markets are characterized by large numbers of retailers, wholesalers and consumers. The activities Of trade associ- ations, even where they existed, did not seem to deter competition. Entry and eXit into and out of the system was found to be free from 8William 0. Jones, "The Structure of Staple Food Marketing in Nigeria as Revealed by Price Analysis." Food Res. Inst. Studies, Vol. 8, N0. 2, 1968, p. 95. 9This'general conclusion can be found in Alan R. Thodey, "Analysis Of Staple Food Price Behavior in Western Nigeria,“ Ph.D. Dissertation, lhfiversity of Illinois, 1969, p. 176; Elon H. Gilbert, "Marketing Of Staple Foods in Northern Nigeria: A Study of the Staple Food Marketing Systems Serving Kano City," Ph.D. Dissertation, Stanford University, 1969. p. 281; Anita Whitney, "Marketing of Staple Foods in Eastern Nigeria," _gricu1tural Economics Report No. 114, Department of Agricul- tural Economics, Michigan State university, 1968, p. 48; H. M. Hays. Jr., Organization of the Staple Food Marketing System in Northern Nigeria," Fh.D. Dissertation, Kansas State University, 1973, pp. 165-166; and 11. Ejiga, "Economic Analyses of Storage, Distribution and Consumption g7§owpea§6gn Northern Nigeria," Ph.D. Dissertation, Cornell University, a p. . 11 obstacles. There was very little government intervention in the operation of the staple food marketing system.10 Policy recbnmendations from the studies centered on suggesting that government should not interfere with the system of staple food marketing. There was optimism that the system was flexible and adaptable enough to handle emerging circumstances. The government was advised to concentrate on the provision of infrastructures like roads and improved market stalls. There were also suggestions for standardization of units of measurement. The studies provided a first-time, comprehensive look at the tradi- tional staple food marketing systems of the major geographical regions of Nigeria. The ability of the food marketing systems to adapt to new circunstances is now being tested under such impacts as agricultural development projects, population growth, and rapid urbanization. Am cultural Develggnent Projects Agricultural development projects like Funtua ADP are regarded as a key to solving Nigeria's food problems. They are also expected to make farming a more renumerative occupation and thus reverse the increas- ing problems of rural-urban migration. Among the expected contributions of ADPs are: (1) increased yields Obtained for most crops, coupled with increased total production of various crops, (2) provision Of a motivated and better-qualified cadre of extension agents, (3) development of an extensive system of rural road networks to improve comunication between various comnuni ties within the ADP areas, and (4) the provision of employment opportunities to vari- ous levels of skilled and unskilled manpower from both within and without the ADP area. \ loAnita Whitney, op. cit., pp. 48-50. 12 These ADPs were also expected to generate a large body of information relating to agriculture in various parts of the country that could be used as an input towards future agricultural policy formulation. Pro- jects, no matter how small, always require monitoring and evaluation as a means of gathering feedback, formulating guidelines for policy changes, and assessing progress as well as new circumstances which might evolve. The Nigerian ADPs are provided with strong monitoring and evaluation 'components. However, the monitoring units have served more in a data- gathering capacity than in analysis and evaluation.11 The Special Place of Maize Under ADPs Maize was not a major grain crop in the northern parts of Nigeria before the advent of agricultural development projects. Production was often restricted to backyard gardens. Most of the maize produced was eaten on the cob after roasting. There was very little conversion into grains, as is the case with sorghum and millet--the major grain Crops of the arealz Maize contributed only 3 percent of the daily caloric intake of cereals in the Zaria area.13 Even though maize was not a major staple in the northern states of Nigeria, it has been an import- ant fbod crop in the southern parts of the country for a long time. 11R. H. Slade, The Monitoringgpf FuntuaggGusau,and Gombe Agricul- turalAngelopment Projects. (Washington, D.C.: The World Bank, n.d.), p. I. .3. 12L. A. Tatum, "Maize as a Grain Crop in the Northern States of Nigeria." Samaru Agricultural Newsletter, Vol. 13, No. 4, October 1971, pp. 7-90. 13E. B. Simmons, “Calorie and Protein Intakes in Three Villages of Northern Zaria Province, May 1970-July 1971." Samaru Miscellaneous Paper No. 55. 13 FADP focused special attention on the production of maize. Maize is the grain crop with the greatest potential in terms of increased yields and total oUtput of all grain staples in the savanna zones of Nigeria. It has the highest responsiveness to fertilizer compared with the majOr competing cereal grains, namely, millet and sorghum. Research has shown that maize production in the savanna areas is tech- nologically feasible and economically profitable. Doubts have, however, been expressed as to the ability of the existing marketing system to handle levels of large-scale production of maize, both at the product level as well as in terms of supplying the required production inputs.14 0n the basis of maize production research at the farmer level, Norman, et a1. (1976), pointed out that the potential for maize produced in the northern states will depend on: 1. The willingness of consumers in the northern states to change their diets by substituting maize for sorghum; 2. The ability to tap the southern market for maize for human consumption; 3. The development of the livestock industry so as to ' create demand for feed grains; and 4. The development of agro-industries, such as starch and oil processing. There are indications that the consumption of maize, particularly among the higher income groups in urban centers, is increasing, and that use of maize fOr meal preparation is not confined to special occasions. There does not seem to be any taste preference for a particular variety ’14Norman, et al., "The Feasibility of Improved Sole Crop Maize Production Technology for the Small-Scale Farmer in the Northern Guinea Savanna Zone of Nigeria.t Samaru Miscellaneous Paper No. 59, Institute for Agric. Research, Admadu E110 University, 1976. 14 of maize in the area. However, the traditional white variety is easier to process into flour, and thus requires less labor by housewives who do the majority of meal preparations. The yellow maize varieties have higher yields but due to processing problems local mills were charging premiums for milling the yellow varieties. Thus, even if people in the area are willing to substitute maize for sorghum, the difficulties of processing maize could slow down the substitution process. As discussed later in this thesis, little has been done to take advantage of the avenues enumerated above in an attempt to promote the production and acceptance of maize. Objectives of the Study The aim of this study is to understand the relationship between the agricultural development projects in Nigeria and the traditional staple food marketing system. The importance of such projects in terms of potential for increased staple food grain production suggests that marketing problems could pose bottlenecks to successful execution of the project plans. This study will concentrate on the linkage between project activi- ties and the traditional staple food marketing system. The study will try to put together elements from the public sector food marketing pro- grams and the private, traditional marketing system. The assumption is that both the private and the public aspects of staple food marketing are needed and each has to take cognizance of the other for a more effi- cient system. This reflects what Lele termed as the "pluralistic approach."15 15Uma Lele, The Desi n of Rural Develo ment: Lessons from Africa. (Baltimore: Johns Hopkins University Press for The World BanE, 1975), p. 100. 15 Thus, this study hopes to extend earlier studies of food marketing in the area by applying similar methods of analysis under the context of an agricultural development project. The study will also add a new dimension to such studies by including a detailed consideration of the input procurement and distribution system. The specific objectives of the study are as fellows: 1. To describe in detail the various activities carried out by agricultural development projects in Nigeria as exemplified by Funtua Agricultural Development Project. 2. To describe, compare, and examine the inter-relationship between the systems of input procurement and distribu- tion at the project, state, and national levels, before and after the re-organization of 1976. 3. To evaluate the performance of the staple fbod marketing system serving Funtua Agricultural Development Project districts through an analysis of staple fOOd prices over space and time. 4. Based on the results Of the evaluation of the staple fOOd marketing system, draw inferences on how well maize is integrated into the marketing system in the area. 5. To draw conclusions and make suggestions as to the specific considerations regarding agricultural market- ing to incorporate in the planning of agricultural development projects like FADP. Organization of the Study Chapter 2 reviews literature on staple fOOd marketing. It starts with a consideration of the approaches to the study of foOd marketing. The chapter then examines the structure-conduct-perfbrmance paradigm, fOllowed by fOOd marketing studies in Nigeria and a criticism of the studies. Chapter 3 discusses integrated agricultural development projects, in terms of conceptual framework, government policies and locations 33’, i." 16, of such projects in Nigeria. Chapter 4 then specifically discusses the Funtua Agricultural Development Project with details involving the area, project, and project activities. Chapter 4 also discusses the data collection method and details of some surveys used in the study. Chapter 5 describes and discusses inter-relationships between systems of input procurement and distribution at the project, state, and national levels. Chapters 6 and 7 present and discuss results of Price analyses--Chapter 6 dealing with temporal price analyses and Chapter 7 discussing spatial price analyses. Chapter 8 sumarizes and concludes the study. CHAPTER 2 REVIEW OF THE LITERATURE This chapters reviews the literature on food marketing studies in Nigeria. It starts with a discussion of approaches to the study of food marketing and the structure-conduct-performance paradigm. General Approaches to the Study of Food Marketing] There are as many ways of studying food marketing as there are daft nitions of marketing. The various methods can be grouped into three "'31 n categories as used by Kohls and Uhl (1980). These categories are: (1 ) the Functional Approach, (2) the Institutional Approach, and (3) the Behavioral Systems Approach. The Functional Approach emphasizes the various functions performed by the marketing system, such as the exchange functions of buying and selling; the physical functions of storage, transportation, packaging, Processing; and the facilitative functions like risk-bearing, financing, grading, and the'provision of market intelligence. Main attention is not focused on who does what, but on what is done irrespective of who performs the service. \ Ma 1This section is based on Richard L. Kohls and Joseph N. Uhl, rketin of A ricultural Products. 5th ed., (New York: MacMillan Pub- IIsfiing E5mpany. WOO}, Chapter 2. 1.7 18 This approach is useful in determining the minimum cost of performing a given marketing function by conparing the costs of'different middlemen performing the sane function.2 Since this approach breaks a complex system into smaller parts, it is amenable to detailed analysis of parts i Of the system. However, unless enough attention is paid to the linkages relating the subsections, the bigger system viewpoint may be lost. The Institutional Approach, on the other hand, emphasizes the insti- tutions and institutional arrangements that are involved in the marketing Pr13<2ess. The approach endeavors to understand operating procedures, sca1es of operations, costs and returns, etc. Studies using this app roach concentrate efforts on institutional arrangements of middlemen-- at retail and wholesale level, brokers and comission men, as well as PPOcessors and supporting institutions that provide facilitative ser- V“ Ces, such as banks, moneylenders and government departments appropriate to marketing. .-_ The institutional approach is useful in analyzing attitudes toward Change and improvement of the marketing system. The acceptance or. "ejection of a proposed innovation can be highlighted by the study of 1:he various institutions involved and how the proposed change affects them in terms of losses and gains. Thus, it can help by showing the 1 ‘1 kely future pattern of new technology adoption. The third approach, the Behavioral System Approach, stipulates that food marketing should be looked at as a nunber of behavioral Systems involved in various kinds of decision-making. It emphasizes \ 2The marketing functions have characteristics among which is the ('1 fficulty of eliminating them. Middleman may be eliminated but their unctions will be done by someone else. Ibid., p. 27. 19 the multi-disciplinary nature of a meaningful study of such a system aggregate. It implicitly considers dynamic elements in the system's aggregate through the inpact of decisions in one system on the behavior of the other interrelated systems. A variant of the Behavioral Systems Approach is the Subsector Approach developed by Shaffer3 in 1968. He defined a subsector as "the vertical set of activities in the production and distribution of a closely related set of comodities.“ Studying agricultural production- distribution systems using the concept makes such studies more manage- able without compromising coverage since subsectors have both a vertical as well as a horizontal dimension that can be delimited based on the Circumstances of a particular situation. . The subsector approach to the StUdy of agricultural marketing is not tied to specific methods of analysis since the concept is only concerned with providing an analyti- ca‘l framework within which available tools could be used for the a"Ia‘lysis.4 The Food System approach, closely related to subsector approach, "as used in a large nuuber of studies carried out by the Latin American Stlacli es Center at Michigan State Uni versity.5 This approach regards p"‘Oduction and distribution as an integrated system that coordinates \ 3James D. Shaffer, “On the Concept of Subsector Studies." % Agr. Econ., 55 (May 1973). pp. 333-335. s 4No single field of study is adequate to handle all aspects of sub- .Fector studies; therefore, such studies draw from diverse fields like an" management, industrial organization, marketing, cooperatives, etc. ems De 5Kelly Harrison, et al., "Improving Food Marketing Systems in NOVETOping Countries: Experiences from Latin America." Research Report \6, Latin American Studies Center, Michigan State University, 1974 20 production, distribution and consumption. It "emphasizes interdependence of related activities and is concerned with the coordination 0f economic activities as a system." This approach is particularly useful in the context of developing countries where numerous, often diverse, factors have to be taken into account in performance of economic activities that form part 'of an over- al 1 tightly interdependent system. Factors are both economic and non- eco nomi c. Structure-Conduct-Performance Paradigm A large nunber of agricultural marketing studies rely on the theo- r‘Et‘ical' foundations laid by the "perfect competition“ model.6 This is Particularly true in studies based on the structure-conduct-performance (S~ C-P) paradigm. The S-C-P paradigm has its origin in the works of Bain] The structure variable refers to nunber and relative sizes of firms as well as the degree of product differentiation and the extent of Vertical integration.8 Market conduct, on the other hand, refers to the behavior of firms relating to pricing practices, i nnovativeness, \ 6The basic assumptions of the perfect competition model are (1) firms produce homogeneous comodity, and consumers are identical from sellers' point of view, (2) both firms and consumers are numerous, and he sales or purchases of each individual unit are small in relation to aQgregate volume of transaction, (3) both firms and consumers possess perfect information about prevailing price and current bids, (4) entry “to and exit from the market is free for firms and consumers in the I:_Ong run. See James M. Henderson and Richard E. Quandt, Microeconomic e h o : A Mathematical Approach. 3rd ed., (New York: McGraw-Hill Co., ISEFL—‘T—‘i', pp. 36- 37. 7 8Product differentiation and vertical integration have a direct connection to conditions of entry and exit in an industry. Joe S. Bain, Industrial Ogganization. (New York: Wiley, 1959). 21 investment behavior and similar matters. Structure and conduct are influenced by “basic conditions“, like location and ownership of raw materials, available technology, price elasticity of demand for pro- ducts, laws and government policies within which the firms operate.9 Performance is the consequence of structure and conduct. It is a multi-dimensional concept whose criteria for evaluation needs to be set out clearly, especially where one is not using the p.c. model as the norm. Often used criteria for performance evaluation, besides the standard norm of the p.c. model, include elimination or minimization of waste, technological innovativeness, full or conmensurate employment 01“ resources, and equitable distribution of income. Some of the latter Chi teria are already inplicitly considered in the p.c. model under its I"equiirements of operational and pricing efficiency conditions. However, for employment and income distribution issues additional analysis is Often needed beyond strict adherence to the p.c. model. The criteria listed above, unlike the p.c. model, have received 1 ‘1 ttle serious attention. Part of the problem results from the inde- terminance _of the S-C-P model when the p.c. model is no longer the .Ya rdstick. The structure of the market influences but does not totally Control the conduct of firms. Thus, the system is not determinate with respect to conduct and therefore not determinate with respect to per- 1{"Siilr‘mance. No given performance can be attributed to a given structure “'1 thout considering conduct, with is i ndetermi nate.10 \ ' 9F. M. Scherer, Industrial Market Stucture and Economic Performance. (Chicago: Rand NcNaTT—TTy, 9 O). prI-‘fi 10Harold F. Breimyer, Economics of the Product Markets of A ricul- %. (Ames, Iowa: Iowa State University Press, 1976), pp. 80- . 22 Measures of Pricing Efficiency in Food Marketing 'Pricing efficiencyn refers to how well prices across space, time, and variety reflect expected prices based on asSumptions of competitive - conditions. In spatial price analysis, prices in two locations for the same comodity should differ only by transfer cost, provided competi- tive conditions prevail.12 The analysis is simplified when there are few central markets to which produce is sent and thus the pattern of flow is easily mapped. Where the pattern of sources of supply and des- ti nations are many, the mapping and interpretation of spatial price Va riation needs to be approached more cautiously. There are problems in determining the levels of transfer costs-- they are not a linear function of distance. There is a distance- ‘independent cost component that reflects the cost of loading and unload- 1 fig . - In temporal price analySis a price series is studied to reveal the Coniponents of the series like the trend, the seasonal, and if the series 1 3 long enough, some cyclical variations. Trend is the long-run move- "Ent in the price level, with long-run being dependent on the type of Product (for example. two to three years in grain crops and up to ten \ 11Two major criteria of efficiency used in agricultural marketing are pricing and operational efficiency. Operational efficiency is aChieved when the maximum amount of marketing services are achieved with a minimum of inputs. It is the achievement of maximum output/input r‘atio. See Breimyer, op. cit., pp. 132-134. 12Transfer costs include loading and handling besides transporta- ‘31 on charges. High value of transfer costs in relation to the value of aSlr'icultural products may result in big price differentials between loca- t‘l ons. See William G. Tomek and Kenneth L. Robinson, A ricultural Pro- W. (Ithaca, New YOrk: Cornell University PrEg'sTT9'72TTTE—T43. 23 twelve years in cattle). Trend can be estimated using simple linear regression of the form: Y = a + bx + e where, b is the trend coefficient Y is the estimated price X is the time variable of the series, and e is the error term The seasonal component is usually estimated using the method of moving averages. Actual prices are expressed as a percentage of their moving average. Seasonal indices are calculated similarly from the moving averages. Seasonal analysis is important in showing periods of Pe1 ative low and high prices over the season and could thus help pro- dulcers and merchants in the process of selling and purchasing decisions. Estimation of cyclical price variation requires deseasonalizing the data by dividing it with corresponding seasonal indices and then remov- 1 ng trend by dividing further with trend values. What is left after the i“Mo-step division is the cyclical component and irregular movements. The cyclical component indicates long-term oscillations about a trend 1 Tue. The oscillations are sometimes periodic. Analysis of pricing efficiency over space and grade makes consider- able use of correlation theory. Correlation coefficients are calculated between different grades to see how well they substitute for one a"other. The higher the correlation between different grades of a pro- Cluct, the stronger is the substitution relationship between them. L‘ikewise, over space, high correlation coefficients that are positive Indicate that prices are unified under a cannon system. Price spreads 24 across space or grades are also very useful when coupled with transportation rates and estimates of premium prices across grades. Food Marketing Studies in Nigeria W. 0. Jones and a team of researchers carried out a number of food marketing studies under the auspices of Stanford Food Research Insti- tzuize.13 These studies were part of a larger study of staple food mar- keting in Tropical Africa which included the countries of Sierre Leone and Kenya, as well as Nigeria. As pointed out by Jones, these studies were carried out to overcome deficiencies in information about internal marketing of food stuffs in tropical African countries. Two important food marketing studies in Northern Nigeria, carried out after the. Stanford studies, were by Hays (1973) and Ejiga (1977). These studies are discussed together with the earlier studies since they all employed $1 milar methodologies. Data for some of the analyses came from government price data col- 1 ected by the Nigerian Federal Office of Statistics and data reported in the Northern Nigerian government's "Crop and Weather Reports." These data were supplemented with field work carried out by the researchers. ‘_ 13The studies of greatest concern here are: William 0. Jones, "The Structure of Staple Food Marketing in Nigeria as Revealed by Price Analy- sis." Food Res. Inst. Studies, Vol. 8, No. 2 (1968), pp. 95-123; Elon H. Gilbert, "Marketing of Staple Foods in Northern Nigeria: A Study of the rketing Systems Serving Kano City." Ph.D. Dissertation, Stanford Uni- Versity, 1969; Alan R. Thodey, "Analysis of Staple Food Price Behavior in Western Nigeria." Ph.D. Dissertation, University of Illinois, 1969; ANita Whitney, "Marketing of Staple Foods in Eastern Nigeria.“ Agricul- &al EconomicsReport No. 114, 1968. 25 Methods of Analysis Each of the studies has examined, in considerable detail, the operational and pricing efficiency of the systems of food marketing under investigation. Emphasis was on the structure, conduct, and per- formance of the systems. For example, in order to determine the oper- ational efficiency of the marketing system for sorghum and millet, Hays examined in detail the market channels, institutions, intermediaries, and other functionaries involved in the marketing process, and evaluated costs in relation to the marketing services provided, including a bud- geting of the incomes of the intermediaries at each stage of the market- ing channel.14 The general research framework for Hays' study is based on d,15 with modifications taken into consideration of the econ- Pritchar omic, technological, and social constraints in the developing country's environment. In general, each of the studies devoted considerable attention to the analysis of temporal price behavior in order to test the allegation of depressed food Crop prices during the immediate post-harvest period as well as claims of an excessive rise in prices during the off-season period. Seasonal price indexes were calculated fOr various locations and crops. These indices were then analyzed to reveal the pattern of seasonal variations. Seasonal increases in price were then compared with costs of storage. ‘ 14Henry M. Hays, Jr., "The Organization of Staple Food Grain Mar- keting Systems in Northern Nigeria: A Study of Efficiency of the Rural- Urban Link." Ph.D. Dissertation, Kansas State University, 1973. 15N. T. Pritchard, "A Framework for Analysis of Agricultural Marketing S.YStems in Developing Countries," Agri.‘Econ. Res., 21: 78-85. ‘ 26 The studies have looked at the level of spatial integration of markets in order to see how well the structure compares with what it would have been if the assumptions of p.c. model applied. The important- variables are the cost of transfer of‘commodities between locations and the level of information flow between the locations. It is important to know which locations trade with one another. Bivariate correlation coefficients were estimated for various crops and pairs of locations. Price differences between locations are also estimated and then evaluated against the cost of transportation and handling between locations. Bivariate correlation coefficient between two series of prices is defined as: 12:“,- - THY,- -‘T) r _ E" (x1 - 715mg: 0. - 7121} "2 1.th 1 = th observation of price series X Y1 = i observation of price series Y N = number of observations X = mean of price series X Y = mean of price series Y The degree of correlation as expressed by the value of the correla- tion coefficient is taken as an indicator of the extent to which two markets or locations are integrated.16 Under conditions of perfect competition the correlation coefficient between prices in two locations will be 1.00. This is not achieved in the real world due to less than perfect conditions in information flow, homogeneity of products, and Physical mobility.‘7 ¥ 16Uma J. Lele, Food Grain Marketing in India. (Ithaca, New York: Cornell University Press, 1971), p. 23. ”Ibid. 27 There are many problems in using correlation coefficients as indicators of market integration. Results of such analysis have indi- cated unexpected outcomes. High correlations have been found between markets that do not trade with one another. It is also common to find significant negative correlations between locations which do not make. sense. Additionally, it is difficult to set up the cutoff points show- ing what levels of correlation indicate what level of market integration. Perfect competition only calls for a correlation of 1.00, anything less has to be explained by other means. Results of the Studies Results of seasonal price analyses showed that urban locations tended to have less price variation compared to more rural locations, suggesting that some storage might be taking place in the urban centers. However, this was not verified by interviews with traders. The studies indicated that traders held stocks no larger than were needed for cur- h.]8 rent needs, lasting about a week to a mont The seasonal price analyses also showed that prices "appear to be subject to more irregular fluctuations than might be expected."19 . The results from the market integration studies in Nigeria indicate that there is only weak integration of staple food markets as indicated by the low levels of correlation coefficients between markets. However, cowpeas showed the existence of a well-integrated system of markets.20 ‘ 18Thodey, op. cit., p. 108; Gilbert, op. cit., p. 95; Hays, op. cit” p. 163; and N. O Ejiga, "Economic Analyses of Storage, Distribution and Consumption of Cowpeas in Northern Nigeria." Ph.D. Dissertation, Cornell University, 1977. pp. 134-135. 19 20 Jones, op. cit., p. 110. Thodey, op. cit., p. 118; Ejiga, op. cit., p. 296. 28 Many reasons were suggested to explain the reasons for the low levels of the correlation coefficients observed in food crops other than cowpeas. These included the poor quality of the data used and problems associated with converting to standardized measures from the local measures used in transactions.21 Other possible explanations of poor market integration are poor information flow about prices and the complete absence of even rudimen- tary market news reports. Even where available, such reports were only for limited official circulation. Quality differences were also hypo- thesized to have contributed towards some of the low correlation coef- ficients observed.22 The studies have achieved the overall goal of providing detailed information on the structure and functiOning of the staple food market- ing systems in Nigeria. The summary findings of these studies regarding the operational and pricing efficiency of the Nigerian staple food mar- keting system is that the systems are operating in a competitive manner except in a few "Unimportant" cases.23 Policy Recommendations Policy recommendations were similar in all the studies. These recommendations suggested that government would do better by not inter- fering with the system of staple food marketing. There was optimism 21Local measures show significant variations from one location to another and in some cases from one season to another, reflecting supply conditions. ' 22Jones, op. cit., p. 155. 23Thodey, op. cit., p. 176; Gilbert, op. cit., p. 286; Whitney, op. cit., p. 48; Hays, op. cit., pp. 165-174. 29 that the system was flexible and adaptable enough to handle emerging circumstances. The government was advised to concentrate on the pro- vision of infrastructures like roads and improved market stalls. There were also suggestions for standardization of units of measurement used - in marketing transactions. Featured prominently was also a call for the establishment of an organized market intelligence system to serve pro- ducers, consumers, and merchants. In conclusion, these studies provided a comprehensive look at the traditional staple food marketing system that represented the major geographical regions of Nigeria. Inter-regional comparisons can be made, since the basic research framework used was very similar in all the studies. Even where conclusions from the analyses were disputed, the researchers' well-documented field accounts are very valuable. The studies have also served greatly in stimulating interest in students and government officials about the complexities of the Nigerian staple foOd marketing system. Other studies of the staple food marketing system in Nigeria not discussed as part of the general overview include studies by Anthonio, Welsch, Olayemi and Okereke.24 Criticisms of Nigerian Food Marketing_Studies Criticisms of these studies is, in general, criticism of the structure-conduct-performance paradigm--and its various derivations. 240. B. O. Anthonio, "The Marketing of Staple Foodstuffs in Nigeria: A Study of Pricing Efficiency." Ph.D. Dissertation, University of London; 0. E. Helsch, ”Rice Marketing in Eastern Nigeria." Food Res. Inst. Studies, 6, (1966), pp. 329-352; J. K. Olayemi, "FoodTMarEeting and str bution in Nigeria: Problems and Prospects." Nigerian Institute for Social and Economic Research, University of Ibadan, 1974. 3O Harriss divided the criticism into two parts: one part relating to the data and the methodology, and the other dealing with relevance of data in the conclusions of the studies. On both accounts Harriss pointed out_ deficiencies in the studies.25 The quality of the secondary data used for the Nigerian studies was generally poor. This is even pointed out by the researchers who used the data. It is a common problem in developing countries; it reflects the poor conditions of data-gathering institutions, particularly those under government ministries. On methods of analysis, the shortcomings of correlation analysis have been discussed earlier. Those shortcomings coupled with poor data fUrther weaken the strength of the analyses. 0n the use of margin analysis, problems include the necessity of using value judgement to explain what a "fair" margin is, explanation of reasons fOr losses if the costs exceed returns, and an explanation of profits at some point and losses at other times.26 Other problems _ according to Harriss included the lack of clear statement in the studies as to what measure of seasonal price variability was being used, inade- quate explanations as to the selected intereSt rates in calculating storage costs, and the use of urban estimates for such costs while most of the storage is carried out at the village level.27 The use of unidirectional flow of product model, from the rural areas to the urban centers, has also been questioned since it ignores 25Ibid., p. 199. 261bid.. P- 204- 27Ibid., p. 205. 31 the possibility of price reversals at certain periods of the Season as 28 evidenced in Indonesia. However, this is a weak argument in the case of the major Nigerian staples which more reasonably conform to unidirec-fi tional flow than Timmer's other models.29 A recent study on staple fbod marketing from Stanford indicated that the criticisms have been noted, but the same methods were used as in earlier studies.30 What Do Correlation Coefficients Indicate? Although a bigpart of Harriss'critique is based on the improper use of correlation coefficients, she made some of the same mistakes for which she was criticizing others. Harriss stated that "high coeffici- ents indicate stable margins or stable prices . . ." She went on, " . . . Since in India and West Africa they (correlation coefficients) obviously do not indicate stable prices, they must indicate stable margins." High correlation coefficients between prices do not neces- sarily indicate only "stable margins or stable priceSt Prices and margins could be rising while coefficients are still high. A simple illustration of this is shown below. ' Assume a series of prices in each of three markets (A, B, C) for ' a commodity. Market A is the reference market whose prices are corre- lated with other markets. The price series in market B were constructed to reflect a constant margin of .02 over prices in market A. Likewise, 28C. Peter Timmer, "A Model of Rice Marketing Margins in Indonesia." Food ResearchpInst. Studies, Vol. 13, No. 2, (1974), pp. 99-143. 29Bidirectional flow models may be useful in the case of those pro- ducts that have a large import component, like rice in recent years. 30Van Roy Southworth, "Food Crop Marketing in Atebubu District, Ghana." Ph.D. Dissertation, Stanford University, 1981. 32 ILLUSTRATIVE DATA Price of Commodity X (Naira/kg) Period Market A Market 8 Market C l .25 .27 .27 2 .26 .28 .30 3 .28 .30 .33 4 .27 . .29 .33 5 .29 .31 .36 Notes: 8 - A = constant . C - A<>constant Corr. AB = .99 Corr. AC = .97 prices in market C were constructed to reflect a rising margin over the series. Computation of correlation coefficients for markets B and C as related to market A shows that both coefficients are above .90, the generally accepted level above which such coefficients are considered high. Correlation coefficient is a measure of the covariation between two variables--it measures the degree of linear association between them. It is dimensionless and can range from —l to +l. For most pur- poses attention is given to positive coefficients. When they are high, approaching unity, this indicates that the two prices move together in the same direction. Increases in price in one series corresponds with an increase in prices in the other series. Likewise, if the price in one declines, the price in the other market also declines. A negative. correlation coefficient simply indicates that as the price in one market increases, the price in the other market decreases. It is difficult to 33 explain in rational terms how two markets can operate in this fashion. When the correlation coefficient between two markets is zero or close to zero, the two markets are said to be unrelated. Correlation coefficients do not imply causation. The two variables involved in the simple correlation coefficient are at the same level of importance. The correlation between A and B, AB, is the same as the correlation coefficient between B and A, BA. There is no independent or dependent variable, unlike the case of regression. These are discussed in most statistics text books.31 Correlation coefficients could be used to indicate patterns of spatial relationships provided precautions are taken in the choice of locations, appropriate prices, and in the interpretation of the results. For high correlations to signify monopolistic situations in a system of markets, a very high degree of market control must exist. Most studies indicate this condition is rare in African food marketing systems.32 In summary, this chapter has examined the literature on food mar- keting with emphasis on studies done in Nigeria. It started with a general review of the framework for food marketing studies. 'It then looked at Nigerian food marketing studies in terms of methodology and findings, and concluded with a critique of the studies. The review indicated that much research has been conducted in the area of food 31(3eorge W. Snedecor and William G. Chochran, Statistical Methods. 6th ed., (Ames, Iowa: Iowa State University Press, 1967); John Meter and William Wasserman, Applied Linear Statistical Models. (Homewood, Illinois: Richard D. Irwin, 1974). 32An exception to this could be in the case of controls imposed by governments as in the case of OPAM operation in Mali, but even there it is indicated such controls fail to be effective. See Elliot Berg, “Reforming Grain Marketing Systems in West Africa." Center for Research on Economic Development, University of Michigan, 1979. 34 marketing in Nigeria, but there is a gap in terms of relating such studies to agricultural development projects. There are also problems in terms of some of the methodologies used in the studies. CHAPTER 3 INTEGRATED AGRICULTURAL DEVELOPMENT PROJECTS This chapter examines agricultural and rural development projects, particularly those labelled as integrated agricultural development projects. The first section deals with a general overview of such projects, including those in East Africa. The chapter then considers ADPs in_Nigeria--their conceptual framework, government policies, num- ber and location, and their activities.. The chapter wraps up with a discussion of some of their problems. General Discussion onAgricultural Development Projects With the demise of the Community Development approach and the inconspicuousness of small-scale agricultural development projects, increasing attention has been given to large-scale Integrated Agricul- tural Development Projects in developing countries. Of special importance in the turn towards Integrated Projects is the availability of funding fbr such projects from foreign aid donors and lenders who seem to have accepted the approach, e.g., the World Bank. Most of these large-scale projects operated as autonomous or semi- autonomous organizations removed from normal established bureaucracy. This detachment from the traditional organization is thought to bring dynamism into the operations of the projects. But this same detachment 35 36 has often been the cause of interorganizational conflicts inimical to the achievement of project goals.1 Integrated Agricultural Development projects are often deliberately_ made large and conspicuous. They are established in geographical regions most conducive to achievement of rapid increases in production and well served with infrastructural facilities. Establishment of such projects is often accompanied by political campaigns in support of the projects. The political support fbr such projects enable the projects access to an unusually large share of available qualified personnel and other resources. The projects become isolated enclaves of concentrated re- source use in comparison to the surrounding areas. - Although the major aim of these projects is to increase the incomes of the farmers via increases in.the output of one or more farm products, other components are invariably included. That is the essence of the term "integrated“ in such prdjects. The projects differ in how much of other components are included. The most often included components are credit, enhanced agricultural extension and Supervision, marketing of outputs, provision of physical inputs such as fertilizer, and, in some cases, the provision of social services like health care facilities, schools, etc. Although there is a large volume of literature on Integrated Agri- cultural Development Projects, there are very few studies that specifi- cally examine the role of marketing in such projects. For Africa, Lele 1This is more likely to occur where personnel from the governmental organization are moved to the new project organization and promoted, wifli greatly improved.conditions of service. The colleagues left in the old organization tend to cause problems, particularly when their cooperation is requested. 37 has comprehensively assembled and critically examined the available evidence.2 The projects examined dealt with various issues including agricultural marketing. The growth of such projects after l975 has been mainly in West Africa, particularly in Nigeria. With few excep- tions, the new projects in West Africa have a lot in common with their East African predecessors and, hence, a look at the experience on pro- jects in East Africa could help in understanding the West African pro- jects. Of special interest are the Chilalo Agricultural Development Unit (CADU) and the Wolamo Agricultural Development Unit (WADU), both in Ethiopia and the Lilongwe Rural Development Program in Malawi. These projects will be reviewed based on Lele, with special emphasis on the marketing component. . . The similarity of the projects in eastern Africa to the Nigerian projects is striking in terms of the components of the two groups of projects. Both are based on the assumption that a "critical minimum effort“ is necessary to make a noticeable impact on the target popula- tion in a relatively short period of time. Both sets of projects pro- vided very similar services made.up of soil conservation, roads, general agricultural extension, credit, marketing services, training and, in the case of the eastern Africa projects, health clinics and nutrition educa- tion.3 2Uma Lele, The Desi n of Rural Developmegt. (Baltimore and London: Johns Hopkins University Press for the World Bank, 1975). 3Similarities in the services provided by the East African projects as compared to the Nigerian ADPs is not surprising since the experience gained by the World Bank in the older projects influenced the structure of the ADPs in Nigeria. The details of the East African projects' acti- vities are described in Lele, Ibid., pp. l4-Zl. 38 Lele's examination of the East African Projects revealed that there was a tendency to set up formal marketing institutions in such projects. There was a serious neglect of the traditional marketing institutions existing in the project areas. The result of such an approach led to higher marketing costs compared to the traditional marketing institutions existing in the area. Lele argued for a pluralistic approach in the marketing of project outputs. and inputs under such projects. The pluralistic approach allows the participation of multiple institutions, both fbrmal and nonformal, parallel to one another which may provide new alternatives fOr producers and enhance overall efficiency of the system. The suggestion that the pluralistic approach is a better choice should, however, be treated with caution since it is possible to theo- retically conceive of a single system that could be equally efficient. However, in practice and based on a number of projects examined, the single formal institution approach has not fared well. Price incentives form a major rationale for including marketing components in Agricultural Development Projects. It is claimed that markets in areas where these projects operate are small and fragmented and the marketing system is prone to various kinds of inefficiencies with middlemen exploiting producers. Other reasons offered include the need to reduce defaults in credit programs.4 The reasons offered as rationale for including formal marketing programs in integrated projects are generally not substantiated by 41bid., Chapter VI, pp. l00-ll5; The World Bank, Appraisal of Funtua Agricultural Development Project, Nigeria. (Washington,TD.C.: The World Bank, l974), Annex l2, pp. l-lB. 39 documented evidence. Cohen,5 for example, wrote, ”There is a lack of competition and much collusion in Chilalo markets. This results in wide marketing margins, lower price to farmers and erratic seasonal price fluctuations due to speculation." This may be true in Chilalo but the conditions described by Cohen may arise from other causes besides lack of "competition" due to collusion. Both Ethiopian projects, CADU and WADU, carried out purchasing of crops. The arrangement to purchase crops generally starts with one crop but later gets extended to other crops. Different crops pose different management problems to projects and formal marketing institutions. Export crops are easier to manage than fOOd crops. Export crops have a centralized system of marketing with unidirectional flow of the products to "central" collection markets from*where they are transported to the ports. In addition, there is Usually only a single organization in charge of the crop. This is close to the position of wheat in CADU. CADU management fbund it relatively easy to handle wheat marketing since the wheat goes to a few big flour mills. Most of the other crops, how- ever, had to go to local markets that deal in small quantities of sales and purchases.6 The small quantities dealt with in the marketing of staple fOOd crops have similarly posed problems in Malawi's LLDP where the marketing of inputs and outputs was entrusted to the Agricultural Development and Marketing Corporation (ADMARC). Although ADMARC handled export crops 5John M. Cohen, "Effects of Green Revolution Strategies on Tenants and Small-Scale Landowners in the Chilalo Region of Ethiopia." Jour. Developing Areas, Vol. 9, (April l975). 6Uma J..Lele, Ibid., p. 103. 4o fairly easily, the organization could not handle staple fbod crops. In the case of maize, LLDP project management had to introduce a scheme to purchase maize directly from the farmers. As a result of the LLDP management pricing differential (they paid more than ADMARC), there was conflict and bitterness from ADMARC administrators that ended up hurting the farmers and increasing the cost of LLDP marketing opera- tions.7 Findings also showed that the involvement of project management in direct crop purchases led to financial problems. CADU, WADU and LLDP have suffered through their participation in schemes to stabilize pro- duct prices with no idea as to the level which prices should be stabilized. More problems are created than solved in attempts to stabilize prices if yield variability is larger than pricevariability. If price variability is due to yield variation, it may be unwise to institute stabilization schemes without a detailed analysis of the price structure and yields over a number of crop seasons. There is no indication that this was done in any of the projects discussed. Intervention in the marketing of inputs and outputs in these pro- jects also neglected to consider the possibilities of modifying the existing marketing institutions, both traditional and formal, so as to handle the marketing aspects of the projects. This completely ignores the advantages to be gained from Lele's "pluralistic approach." Although the problems in Nigeria are similar to those found in East Africa, there are also some major differences. While in East African projects over-centralization was a major problem, this has not been the 70p. cit., pp. lDS-l06. 41 case so far in Nigeria, at least in staple food marketing. It is, however, an important issue in the case of input procurement and distri- bution system. In the case of staple fbod marketing the local marketing- institutions were left to take care of this aspect. There was an informal arrangement for the purchase of some project output of staples via government channels, but the arrangement did not turn out as planned. Inadequate background research on the capacity of the local traditional marketing system to handle the envisaged quantities of maize led to problems in the disposal of the crop at prices expected by the farmers.8' The problems in Nigerian ADPs included lack of vertical coordination of production and marketing activities which led to economic losses to both producers and consumers. There were no provisions for agricultural product processing activities under the ADP arrangements. Complaints were made regarding late opening of buying stations for statutory crops ‘ like cotton and groundnuts.9 - The concept of agricultural development projects as applied to the Nigerian situation is dealt with in the following sections. Integrated Agricultural Development Projects in Nigeria The concept of an Integrated Rural Development Project (IRDP) is not a new one in Nigeria, but it got major support at federal government level when it was explicitly considered in the third National Development 8FADP, Quarterly Report, January-March, 1980. 9F. S. Idachaba, "Concepts and Strategies of Integrated Rural Development: Lessons from Nigeria." Food Policy Technical Research Paper No. 1, Department of Agricultural Economics, University of Ibadan, Nigeria, p. 30. 42 Plan of l975-l980.10 According to the plan document, it was the policy of the Nigerian government to: . . promote a new strategy whereby available extension personnel be redeployed to permit concentrated efforts in selected compact areas. Taken together with an appropriate institutional set up, such as farmer's groups and coopera- tives, this strategy will ensure that extension, input supply, supporting services, such as marketing and equipment hiring are integrated at the village level. The above essentially makes up the Nigerian government's concept of Integrated Rural Development. Further, in the same paragraph it was mentioned that the IRDPs and National Accelerated Food Production Pro- gram depended on the above policy guidelines. Development Prpjects There is considerable ambiguity in the discussion relating to the Integrated Rural Development Projects in developing countries and its 1] Often many take Integrated Rural relation to development objectives. Development to be synonymous with Integrated Agricultural Development '(IAD). 'The latter was the case in the third National Development Plan document. This was probably due to the great weight of agriculture in most rural areas of developing countries. However, IRD is much more broadly based than IAD and includes components generally considered to be non-agricultural, like health care services, educational facilities and programs, and development of awareness in the political process. It 10Ibid., p. l. It should be noted that the difference between IRDPs, IADPs and ADPs is often lost, particularly in government documents. The difference, though important, is not strictly adhered to. nTekola Dejene, "Integrated Rural Development in Africa: Planning and Evaluation." Masters Thesis, Michigan State University, l973. 43 is clear that these issues are obviously all inter-related and hence the futility of attempts to draw demarcations as to where agricultural devel- opment stops and rural development starts. The problem is not just a matter of semantics, particularly if one i looks at what gets carried out from the planning documents. The planning and execution, as well as the results obtained so far from the first three IADPs in Nigeria, do not support the claim of these being integrated in the sense of IADs let alone as IRDs. The Nigerian ADPs cannot be categorized as Integrated Rural Development projects since they are mainly production-oriented ADPs. For the term IRDP to be used there is need for projects to include components that reflect greater concern with rural welfare, like adult literacy, health care components, etc. ‘Even as integrated agricultural projects there is very little coordin- ation, let alone integration, of project activities as revealed by recent studies.12 -. . 0f the five models presented by Idachaba, Model I, "Integrated Supply of Farm Inputs and Marketing Facilities", probably came closest to the situation in Nigerian ADPs. As pointed out by Idachaba, the basis for this model rests on the premise that agriculture is the pre- dominant occupation in rural areas and any attempts to raise productivity must thus consider the sector. Another premise is that raising produc- tivity has a number of necessary prerequisites. The necessary prerequisites include: l. Timely provision of the right inputs at the right places 2. An effective extension system ”Idachaba, op. cit., p. 36; D'Silva and Raza, pp. 282-297. 44 An integrated extension-research-training system 3 4. An extensive network of feeder roads 5 Supportive farm credit 6 Supportive institutions Another premise of the model is that gains in productivity can be lost if not complemented with development of an efficient marketing system to handle issues relating to pricing, storage, transportation, processing, and others that might arise as a result of project activi- ties.13 The other fOur models included various components in addition to the above, like equity, rural non-farm production activities of the small-scale industry types, social amenities relating to health and education, and political awareness. Attention will now focus on Model I since even the requirements of this minimal model for the term of inte- grated development have not been met. Although all the components of Model I are often present in Nigerian ADPs, the coordination needed to consider them as a single unit is absent. It is also clear that the form of these components is strik-. ingly different from the form required for an integrated system. For example, the marketing component does not include the processing and marketing of staple food crops, the leading justification for the pro- jects. The case of maize marketing is particularly illustrative and results presented later tend to confirm the lack of integrated planning and execution of the projects in general, and in the FACP in particular. 131pm. pp. 3-5. 45 Government Policies Relating to Agricultural Development Projects The origin of the policies relating to the ADPs was the increasing concern with low productivity of agricultural production and the conse- ‘ quent inability of agriculture to feed the growing population of the country, which is becoming more urbanized and non-agricultural. The Federal government discussed with the World Bank and various state governments the ways to arrest the problem.- What emerged was a policy directive on the establishment of agri- cultural development projects, particularly in the more northern parts of the country. The geographical coverage has since been extended to cover most of the zones of the country. The agricultural development projects were planned to provide a "short-term“, quick way of improving farm production and incomes. The projects would provide a concentration of support services to areas of reasonable potential and dense farm population. The guidelines assumed that in such areas farm sizes cannot increase and that farmers are faced with the task of maintaining fertility of existing lands to maintain present production levels. The government and the bank felt that con- ditions would permit the rapid acceptance of new improvements offered to the farmers.14 Since the ADPs are planned to provide only short-term solutions to problems of farm production and incomes, the government had other plans for a more long-term solution based on the utilization of unused parcels of land in the tsetse-infested middle belt areas of the country. Thus the failure or success of the current ADPs will probably influence con- sideration of the long-term option. 14The World Bank (1974), op. cit., p. 4. 46 Design Elements of Nigerian ADPs The specific elements of ADP design for successful operation included (1) the careful selection of location for the project, ensur- ing that the soil is fertile and extending the availability of higher yielding, tested-crop production packages to farmers, (2) the project_ should focus on farm inputs, rural roads, water resources and improved extension, (3) the projects had to be large and prominent to attract attention to researchers and farmers, and away from administrators, (4) an appropriate incentive structure, based on farmers' estimates of finanCial profitability, had to be built in to encourage voluntary farmer participation. Subsidies on inputs were very generous and the management structure allowed for the use of foreign expertise to make up for deficiencies in local personnel. There was also provision to incorporate the-training of local manpower for the projects. ' The above factors accounted far the reported relative success of the agricultural development strategy based on ADPs in Nigeria.‘5 There are also a number of criticisms of the policies. One such criticism focused on the large investments involved and the lack of concern with cost recovery. The government's subsidy provisions are too generous to allow for the replication of the success of such projects elsewhere with- out financial capability for similar subsidies.16 The heavy reliance on foreign personnel is also seen as a negative factor since upon termination of their contracts local personnel of 15The World Bank, Accelerated Development in Sgt-Saharan Africg, An Agenda for Action. (Washington, D.C.: The WorldlBank, l9Bll, p. 53. 161bid., p. 16. 47 equivalent training and experience are often unavailable to manage the projects. There is need for an integration of local personnel in all key areas of project decision-making, extending from the project prepara- tion stage to the total transfer of the project to local management. Al-Sudear appropriately comments that ". . . planners in the developing countries must themselves become more fully involved in charting develop- ment strategies and investment projects appropriate to their specific "17 But to get the statements realized there is a country requirements. need for greater involvement of local personnel in the day-to—day running of projects. Number and Locations of ADPs in Nigeria The number of ADPs in Nigeria has increased steadily since the first three were initiated in 1974-75. Six more had been added by the end of l980, and a number of the original projects had been extended to cover wider geographical areas. The first set of projects included Funtua, Susan and Gombe ADPs. The second batch of projects focused on Lafia, Ilorin, Ayangba and Bida ADPs. Other ongoing ADPs include the Oyo North ADP and the Ekiti- Akoko ADP. There are a number of other ADPs in preparation, some of which were at appraisal stage in 1981 (See Figure 3.l). Funtua, Gusau and Gombe ADPs are in Kaduna, Sokoto, and Bauchi states, respectively. Lafia ADP is in Plateau state, Ayangba ADP in Benue state, Ilorin ADP in Kwara state, Bida ADP in Niger state, Oyo North ADP in Oyo state, and Ekiti-Akoko ADP in Ondo state. - 17Abdelmuhsin M. Al-Sudeary, Forward to Investment Projects in Agriculture by McDonald P. Benjamin, (Harlow, Essex: Longman, l98l5, p. xv. 48 The project activities were monitored by the Agricultural Project Monitoring, Evaluation, and Planning Unit (APMEPU) at Kaduna. Recently two more APMEPUs have been added--Benin and Enugu--located in Bendel and Anambra states, respectively. There is also an Agricultural Rural Management and Training Institute at Ilorin in Kwara state as part of the system (See Map 3.1).. Activities of Nigerian Apps]8 The ADPs have multiple functions that included Rural Infrastruc- tures, Farm Service Centers, Farm Inputs and Farm Support Services. Under the Rural Infrastructures component the projects undertake the construction of feeder roads, earth dams, ponds and soil conservation schemes. . Farm Input component, one of the most important components of the projects, handled seed multiplication activities, supply of fertilizers, insecticides, tractors, sprayers, ox carts, ox ploughs and credit facil- ities to allow the purchase of some of the inputs. Farm support services dealt with the provision of extension services and had the objective of greatly reducing the high farmer/extension agent ratios existing in the project areas. Project marketing activities, a part of farm support services, center on the activities of the commercial sections which run the Farm Service Centers. These FSCs sell inputs to the farmers and provide management infbrmation at locations easily accessible to the farmers. 18For a detailed description of the activities of the projects see The World Bank (l974), 0p. cit., pp. 7-l0, and McDonald P. Benjamin, op. cit., pp. 186-188. 49 Map 3.1 Location of Agricultural Development Projects in Nigeria ADMINISTRATIVE MAP of NIGERIA Shaded to show ADP: and marks to show location of APMEPU Kaduna, Benin, Enugu, ARMTI Ilorin. WM [:1 Ongoinu proiocts Im--r““ APPIQIM pmioct. A luau-— 0-. nun..— u nau- nan- - Under preparation - No project plans (a of 1931) Source: Federal Ministry of Agriculture, Department of Rural Development. 50 The details of project activities will be illustrated using the case of Funtua Agricultural Development Project in the next chapter. Major Problems Associated with ADPs Nigerian ADPs are providing a greatly improved system of farm input distribution, including extension and infrastructural services--parti- cularly rural feeder roads. However, attention has not been given to the impacts of increased use of production inputs on the broader food system. There is very little in terms of relating project activities to_the rest of the economy beyond the immediate project areas. Integra- 'tion in terms of Nigerian ADPs can only mean that the projects have multiple components. However, these components are not integrated and in a number of cases are not even well coordinated. ‘ There are problems of poorly addressed equity issues as well. Some of the operational procedures used in project execution have been found to be inequitable and anti-small farmer.19 Other economy-wide issues include the likely impact of the ADPs on the traditional marketing system for farm products and the underlying conditions of supply and demand for the various crops with which the projects are working. Relationship of project activities to government food policies and relationship to the agro-industrial sector have received relative neglect. These issues are by no means easy to grapple with much less successfully incorporate within the ADPs, particularly at the planning stages whereithis is most needed. Nevertheless, given the size of investments involved and the resulting non-marginal mature of 19D'Silva, Brian C. and M. Rafique Raza, "Integrated Rural Development in Nigeria - The Funtua Project," Food Policy, Vol. 5, No. 4, November 1980, pp 282-297. 51 the projects, one would have expected more precautions built in the project design to reflect concern with the issues discussed above. Projects, no matter how small, always require monitoring and evalu-- ation to provide feedback fer guiding policy changes as well as to reflect progress and identify new circumstances. The Nigerian ADPs are provided with a strong monitoring and evaluation component. However, these units have served more in the area of data-gathering than in analysis and evaluation. The monitoring and evaluation unit has so far not had much impact in redirecting policy relating to managing ADPs. This is a serious shortcoming, even though admittedly the unit had a number of serious obstacles initially.20 In the first three years of the ADP's existence there were a number of serious disagreements between the Project Management Unit and the Project Monitoring and Evaluation Unit at Kaduna regarding the validity of some results of the initial data evaluation. Things have improved, however, even though data pro- cessing still lags behind data collection capability. However, these problems should not detract analysts from consider- ing the successful aspects of ADP operations, including the success of. APMEPU. 20R. H. Slade, "The Monitoring of Funtua, Gasua, and Gombe Agricul- tural Development Projects." (Washington, D.C.: The World Bank, n.d.), pp. Ij.B. 25-1. B. 3] . . CHAPTER 4 STUDY AREA, FADP AND THE DATA This chapter describes the Funtua Agricultural Development Project area andthe major activities of the project. The chapter also presents the various FADP/APMEPU surveys from which data was obtained for the analyses carried out in Chapters 6 and 7, as well as the supporting information with regard to the input distribution system in Chapter 5. Study Area and Project Background Funtua Agricultural Development Project (FADP) is located in Kaduna 1 The state.is bordered by the State, one of Nigeria's l9 states. Republic of Niger to the north; Kano, Bauchi and Plateau states to the east; the new Federal Capital Territory and Plateau State to the south; and by Sokoto and Niger states to the west.” The state has an estimated population of about 7,000,0002 and lies between latitudes 9-13° north and 6-9° east and is comprised of approximately 70,000 square kilometers} Rainfall in the state varies from over l,250 millimeters in the southern parts of the state, to less than 750 millimeters in the extreme 1At the onset of the'project, Kaduna State was called North-Central State. There are calls to increase the number of Nigerian States to much more than the current l9. 2Thistestimation was based on the state population of 5.5 million in 1974 and an assumed growth rate of 2.7 percent per annum. , 3World Bank, "Appraisal of Funtua Agricultural Development Project, Nigeria." (Washington, D.C.: The Werld Bank, l974), p. 2. . 53 north. Most of the rainfall is concentrated in the months of May to September, with a high coefficient of variation in its distribution within and between seasons. There is a long dry season between October and April during which little farm work is done and unemployment and - underemployment is a serious problem. FADP area covered the five southernmost districts of the former Katsina Province (See Map 4.l). Total population of the project area, based on the I963 census, was estimated at 500,000 for 1975 based on a 2.5 percent growth rate. Breakdown of the population by district is presented in Table 4.l. _ . _ Table 4.l Distribution of Population by Districts, FADP, l975 !. District . _ ‘Number of Villages Population Funtua " 16 126,5l9. Bakori l9 l28,730 Malumfashi 22 125,967 Kankara 1 l2 73,204 Faskari 8 ' 45,580 ' Total 77 500,000 Source: FADP, "Guide for Project Staff." FADP, 1975, pp. 243. Note: Original total population estimate was 905,000. FADP area covered 7,500 square kilometers or about l0 percent of Kaduna State's total area. Average farm size in the area was less than four hectares but there are many large farms in the area as well. A survey carried out for the period of l979/80 revealed that 36 percent of the farms in the area were less than 2 hectares, 74 percent were less 54 Map 4.1 Towns and Villages, FADP FUNTUA AGRICULTURAL DEVELOPMENT PROJECT 55 than 4 hectares and 93 percent were less than 6 hectares with a Gini Coefficient of .55.4 Average farm family size in the area was six per- sons per household, but this fluctuates from period to period due to sudden arrivals or departures of relatives.5 Crops grown in the project area include: sorghum and millet, which have been the basic staple food grains in the region for a long period; cotton and groundnuts, which in the past provided important sources of cash income for farmers; and others like maize, cowpeas, vegetables, sugarcane and some rice. Most crops are grown intercropped in mixtures. The prevalence and nature of growing crops in mixtures is well studied by Norman and others.6 Finally the project area is well served by 4APMEPU, "Funtua Agricultural Development Project Completion Reportfl APMEPU, Federal Department of Rural Development, Kaduna, 1982, p. 67. Gini coefficient measures the degree of equity in the distribution of a resource among classes of a population. It varies from zero (maximum equity) to one (all the resources owned by one individual or class of the population). The value of .55 in the distribution of land is in the middle range of the scale. This is somewhat higher than what Norman reported for other areas of northern Nigeria.(See Norman, et al., "Tech- nical Change and the Small Farmer in Hausaland, Northern Nigeria." Afri- can Rural Economy Paper No. 21, Department of Agricultural Economics, Michigan State University, East Lansing, 1979, p. 123.) Other sources regarding incomes and their distribution in northern Nigeria include Peter J. Hatlon, "Income Distribution Among Farmers in Northern Nigeria: Empirical Results and Policy Implications." African Rural Economy Paper No. 18, Michigan State University, East Lansing, 1979; Eric W. Crawford, "A Simulation Study of Constraints on Traditional Farming Systems in Northern Nigeria." MSU International Development Paper No. 2, Michigan State University, East Lansing,-1982; and James 0. Olukosi, "The Dis- tribution of Personal Incomes Among African Farmers--A Two Period ?3;;ysis." Ph.D. Dissertation, Michigan State University, East Lansing, 5R. H. Slade, "The Monitoring and Evaluation of the Funtua, Gusau and Gombe Agricultural Development Projects." (Washington, D.C.: The World Bank), n.d., p. 6See fbr example Norman, et al., 1979, pp. 56-64; E. F. 1. Baker and Y. Yusuf, "Mixed Cropping Research at the Institute for Agricultural Research, Samaru, Nigeria." In Intercroppin in Semfi-Arid Areas, J. H. Monyo, A. D. R. Ker and M. Campbell, (eds.)§(Ottawazlnternational Development Research Center). ' 56 good first-class roads linking the major towns but with a poor rural road network connecting villages to one another and to major towns. The FADP area, like other areas of the country, was experiencing problems related to rural-urban migration and stagnant agricultural production. These were a part of the reasons fbr establishing FADP in the area. The project was the first and largest of three pioneering agricultural development projects in northern parts of Nigeria. The other two were the Gusau ADP, which was contiguous to FADP area but located in the then North-Western state, and the Bombe ADP in Bauchi state. Funtua Agricultural Develppment Project Activities According to the project appraisal report, the FADP was estimated to cost 37.9 million naira (U.S. $57.6 million) of which 51 percent was made up of foreign exchange loan to the Nigerian Federal Government repayable at 8 percent over 20 years with a five-year grace period for the principal.7 Part of the World Bank loan was then on-lent to the Kaduna State government for the operation of the FADP. Purchase of farm inputs by farmers was to make up about 5 percent of the project cost. The project was planned to participate in the following activities: 1. ‘Agricultural Road Development: Construction of 1,500 kilome- ters of rural roads to allow light traffic during the wet season and heavy trucks during the dry season. This would improve the efficiency of farm produce evacuation and the supply of farm inputs as well as other possible benefits. 7World Bank, op. cit., p. ii. 57 2. Water Development: Construction of 85 small- and medium-size dams each with a minimum capacity of 100,000 cubic meters. Besides the dams 160 ponds were also included in the construction plans. All these _ water development schemes would provide water for human and livestock consumption. . 3. Soil Conservation: This involved construction of cutoff ditches and contour ridging for the protection of an estimated 2,700 square kilometer area. 4. Building Development: Construction of 350 houses, project office, 5 development center offices, a railhead store, 77 farm service centers, and the improvement of 3 market depots for cotton in the area. Each farm service center was planned to have a storage capacity of 500 tons of farm products or inputs. 5. Seed Multiplication Farms: Improvement and expansion of exist- ing seed multiplication farms at Kaudawa and Malumfashi. 6. Training Facilities: Expansion of boarding facilities at Daudawa training center from 20 to 40 trainees and the setting up of another training center at Malumfashi for another 40 residential trainees . 7. Farmer Support Services: Reduce the estimated farmer extension agent ratio from 2,440:l to a more manageable 240:1. Extension agents would increase from 41 to 420--over a tenfold increase. Types of exten- sion personnel planned included basic extension agents and personnel specialized in farm management, livestock husbandry, seed multiplica- tion, farmer training and farm equipment. The project also planned to phase out the state-run tractor hiring service and replace it with pri- vate operators. A major element of the farmer support program was the 58 credit and marketing services which planned to employ over 260 people, most of them to be trained by the project. 8. Farm Inputs: Under this component 56,000 tons of fertilizer, 4,000 tons of improved seeds, 2,000 tons of insecticides, 47,000 ULV8- sprayers, 10,000 ox carts, 10,000 ox ploughs and 100 tractors would be made available to farmers for cash and credit. Farm service centers would serve as the outlets for these sales. 9. Project Monitoring and Evaluation: This would be established with two sections--one dealing with the review of technical and financial records of the project, and the other dealing with evaluation and analy- sis to supply much needed data for agricultural and rural planning. 10. Post Project Development: This terms refers to plans of project activity continuation under various local institutions once the project investment period is completed. The project headquarters is located in Funtua. The project was deliberately intended to be a large-scale project both in terms of geo- graphical coverage and in terms of the size of investments involved. It covered the domainaof two-loca.1 governments conposedof the districts of Funtua, Bakori, Malumfashi, Kankara and Faskari. One of the areas of emphasis in the extension component is the assistance given to maize growers; special attention was needed since maize was not a major crop in the area prior to the establishment of the project. Detailed information on farm production activities were .8Ultra-Low-Volume sprayers have been used by IAR, Zaria,for its cotton spraying trials for quite some time. 59 supplied to the farmers, with emphasis given to the group of farmers the project identified as “progressive farmers".9 As can be seen from the above description of project activities, the project marketing component revolved around activities of the comp mercial section which dealt with the running of the Farm Service Centers. These FSCs sell inputs to the farmers and provde management information at locations which are easily accessible to the farmers. Other direct marketing activities of the project include the con- struction and running of scheduled crop buying centers in the project area fbr cotton and groundnuts, in cooperation with the Cotton and Groundnut Boards. The project, however, had very little to do with the marketing of staple crops except under abnormal circumstances, like when the market for maize in the project area became saturated in 1979/80 and the project had to step in and purchase the product to avoid catastrophic declines in the crop prices.' About 3,000 tons of maize were purchased by the project management--a small amount relative to total production for the season, a reported 57,254 tons.10 Besides the above mentioned direct involvements with marketing activities, the project was also indirectly involved. The construction of rural roads had a direct connection to the opening of additional mar- ket outlets fbr farm products from remote regions of the project. The project was also involved in the training of marketing personnel that managed the FSCs. Other activities included the collection of price infbrmation as part of the main project surveys. ‘ 9These are farmers who have adopted at least some of the project recommended practices. 10FADP, Quarterly Report, January-March, 1980. 60 The FADP waS'officially terminated after its five-year investment phase. It is now part of the Kaduna State Integrated Rural Development Authority projects under Zone II. Now it is useful to reflect on the successes and shortcomings of the project durings its operation over the 1975-1980 period. Such reflection regarding project intentions could serve as a guideline to assess which goals were actually met by. the project. To a large extent the project achieved most of the goals it set out to achieve. However, there were areas where the set targets could not be achieved and cases where substantial benefits became apparent when originally none were expected. With respect to road construction, the project could not achieve its set target of 1,500 kilometers of rural roads. By the end of the project's five-year period, only 507 kilometers of rural roads were constructed. It was thought; however, that the original target was overambitious.n In terms of water development targets the project constructed 43 dams although 85 were anticipated in the appraisal report. Of the 160 ponds targeted, none were constructed. Similarly no soil conservation scheme was undertaken except the necessary ones around the constructed dams. The project did very well in achieving its building projects, par- ticularly the farm service centers (FSCs) in which 71 out of the targeuai 77 were built. A large proportion of the housing units, including the dormitories for trainees at Daudawa and Malumfashi, were built. However, IIAPMEPU, op. cit., 1982, p. 5. 61 the quality of the buildings, constructed by indigenous contractors, was not satisfactory. The project had succeeded in operating the two seed multiplication farms they started, although the actual capacity was lower than antici- pation projections. This was partly due to an overestimate of the actual demand fbr improved seeds by farmers. The project implementation process revealed farmer preference to keep their own seeds fer the next season should have been given more serious consideration. The project has done very well in its training program for the extension staff. The goal of reducing the high farmer extension agent ratio from 2,400:l to 300:1 was achieved. This ratio converts to a concentration level of extension staff at FADP which is ten times higher than the rest of the state.12 I In terms of other farmer sUpport services little success was evi- dent. The credit scheme for-farm input purchase was scrapped due to high cost of distributing "small amounts" of money (about 20 naira per farmer) and also due to the heavy subsidy on both fertilizer (80 percent) and crop protection chemicals (50 percent).13 The marketing support services, which were mainly geared to cotton marketing, could not succeed due to the government placing unfavorable price controls on the crop. The constructed cotton markets were, however, useful to farmers. As a licensed buying agent for cotton, the project also tried some innovative ways of bulk-transporting cotton from buying stations to the IIAPMEPU, op. cit., 1982, p. 5. 12Ibid., p. 10. 131bid., p. 17 62 ginnery. The project was able to purchase about 75 percent of the cotton produced in the FADP area. Although the project had no plans to get involved with fbod crop marketing, it was forced into such action in the case of maize during the 1979/80 season in order to ease gluts in local I markets. ApproXimately 3,000 tons of maize were purchased. By far the most successful component of the project operations was the supply of farm inputs. The operation had its own peculiar problems, but on the whole it was an undoubted success. _The details of this component will be discussed as the major topic in Chapter 5. In terms of achieving expected crap targets the project revealed unexpected results. The anticipated increase of maize was not achieved to the extent planned. The expected decline in the acreage of sorghum due to growth in the acreage of maize did not come about either. Acre- age and production of sorghum increased beyond projection and actually helped in contributing substantial project benefits. Expected increases in cotton instead turned out to be a decline below the pre-project estimates. Millet, which was completely ignored by the project, also made substantial contribution. Thus even though the project crop pro- duction activities can be said to be a success, the success came from unexpected and unplanned directions. In summary one can say that on the whole the project was a success. It also served as a learning experience that should be utilized when planning new projects. An indication of its acceptance as a success is the extension of the project on a state-wide basis and also the imple- mentation of the input distribution system, developed during the project, in the new Kaduna State Farmers Supply Company. This topic will be discussed further in the next chapter. 63 At this juncture it would be useful to discuss the data which are used in the analyses reported in Chapters 6 and 7. The Data and Data Collection Methods Most of the data used in this study was collected as part of the monitoring and evaluation activity of the project. The majority of this data has not been analyzed beyond simple aggregate statistics (Slade, 1981). The data is also a part of a large data base collected under the same arrangements for the three pioneer ADPs at Funtua, Gusau and Gombe in northern Nigeria. The survey methods used are those reported by Slade and APMEPU, as well as reports obtained directly from FADP.14 The first problem encountered by the planners of the surveys was a total lack of basic data on which to base their survey sampling frames.15 This made it necessary for the project to carry out a basic listing ’survey in conjunction with a baseline survey which provided the infor- mation needed for a proper sampling frame. The baseline survey was made up of two components, a listing com- ponent and a socio-economic component. The listing component provided information on the number of families, hamlets (ungunni), and villages in the FADP area in 1976. The results indicated the presence of about 100,000 families in 693 hamlets. This helped in setting up the sampling frame for future surveys. The estimates have been revised downward to 14Slade, op. cit., n.d.; C. D. Poate and P. F. Daplyn, "Farm Sur- veys and Project Evaluation, A Methodology Manual." APMEPU, 1982. See also Jean C. Balcet and Wilfred Candler. "Farm Technology Adoption in Northern Nigeria: Summary and Conclusions." World Bank Research Project, RPO 671-88, 1981. 15Slade, op. cit., Part II. 64 about 84,000 families and a total populati0n_of 500,000 (See APMEPU, Project Completion Report, p. 1). Some of the major surveys carried out at FADP over the 1976-1979 period are listed in Table 4.2. Table 4.2 Major Surveys Carried Out at FADP Survey Period Villages Households Baseline '1975/76 . n.a. 5,103 Mainline 1976/77 24 576 Punchline 1977/78 23 276 . Deadline 1978/79 15 180 Sources: Slade, n.d.; and FADP/APMEPU.’ There are major differences in the procedures used for the various surveys. Some surveys received more planning and supervision than others and are hence believed to be‘more reliable. The surveys are briefly discussed in turn. The baseline survey questionnaire was designed for computer pro- cessing and hence there was need for a detailed coding manual for tran- scription of field data into computer-readable codes. It took a year from the inception of this survey to the time when initial results of the analysis were obtained. ,The actual field data collection period only lasted fbr six weeks. The baseline survey was fOllowed immediately by the mainline survey which was carried out in 24 villages and involved 24 households per village. Coding and transcription for this survey was done at APMEPU 65 headquarters after initial checking at the monitoring and evaluation unit in Funtua. The survey collected information on labor use, farm expendi- tures, farm income, non-farm income and expenditures, and household expenditures. Interviews were conducted on a weekly basis from the first week of May 1976 to March 1977. Lack of transportation encouraged the use of enumerators who were asked to reside in the survey villages for the duration of the mainline survey (Slade, 1981). The method of data collection required the division of the sample so that when one part of the sample was being interviewed the other was left to rest so as to reduce boredom. How- ever, enumerators were occupied throughout the survey period. I Data collection for the mainline survey ended in March 1977, but coding and transcription lasted until August 1977 and computer valida- tion could not start until November 1977 (Slade, 1981). Slade also noted that the survey missed an important part of the growing season thus necessitating a similar survey the next year. The new survey was to rectify the shortcomings of the mainline survey. It was called punchline survey and was carried out during the 1977/78 season. The punchline survey used a smaller sample size--a sub-sample of. the mainline survey (see Table 4.2). The important difference, however, is that the punchline survey received more detailed preparation and was planned to facilitate rapid computer processing. Thus, even though it dealt with the same kind of infbrmation collected during the mainline survey, the format of entering the information gave the punchline survey a decided advantage in speed of processing. The final main survey for which data was available for analysis was the deadline survey which was carried out during the 1978/79 season. 66 Unfortunately the data from this survey is believed to be of questionable . reliability due to a number of factors among which are lack of adequate preparation and lack of supervision during collection. The senior evaluation officer in charge of the survey for FADP was called to take charge of the preparatidn of the second phase of FADP which involved statewide coverage. Thus supervision of the deadline survey execution took a secondary position in priorities. The discussion so far has described the data sets for this study. However, most of the data needed far this study came from what the prof ject considered supplementary surveys. These surveys were carried out in conjunction with the main surveys described above. They included such surveys as the agronomic surveys, the producer price surveys, mar- ket surveys, and extension surveys. For this study the results of the producer price surveys from 1976 to 1979 were the most important. The market survey and a few other supplementary surveys were also used. The main surveys were also used to fill in infbrmation gaps and to obtain general infbrmation relating to the project. By noting kinds of main surveys carried out, it is evident that the project was mainly concerned with production while marketing issues were considered secondary. Thus problems arise for anyone trying to study the marketing aspects of the project. The data collected does not provide means of carrying out a complete study of the issues since such issues were not considered in the data collection. A lot of the infor- mation one would expect to be included in the price and market surveys unfortunately were left out. This is a limitation of this study and the position is-taken that the analysis is still worthwhile and should 67 proceed using the available information rather than postponing the analysis pending the availability of more complete information. The next section describes the price data collected by FADP. Producer and Market Price Surveys It was mentioned earlier that the price collection surveys were carried out as part of the main surveys discussed above. The earliest price survey provided data during the month of August 1976. The price data collection was carried out starting with FADP and Gusau ADP fol- lowed by other projects. In all cases the data collection was part of the farm management surveys. TWo types of prices were collected over the period of 1976 to 1979. (Note: Price collection is still going on.) These two price types were called the producer prices and the market prices. Market prices were also called commodity prices:— Producer Price Surveys The term producer prices often connotes the idea of a fixed price offered to farmers for their crops by a price fixing agency in develop- ing countries. This has been the case in Nigeria for a long time when referring to the term in relation to cocoa, groundnuts, cotton and other so-called cash crops. .However, the use of this term for the data col- lected at FADP was a misnomer in that it has nothing to do with fixed prices. Producer price here refers to the price at which a farmer can sell his crop at his farm or household or at the nearest market location. It is more like the concept of farmgate price than that of a fixed price 68 since the crops concerned are sold in the open market and there are multiple outlets for the crops rather than a monopoly purchasing agency. Nonetheless, producer price here is not precisely the same as farmgate price since there are various locations involved at which the price was quoted and it is not possible to convert all the prices to a farmgate basis. Suffice it to state here that the producer prices used in this study only serve as very rough proxy for farmgate prices. The survey procedure for collecting farmgate or producer prices involved selecting a sub-sample of two houses per village from those households participating in the main farm management surveys. Infbrma- tion about the prices was obtained from the household head who is often in charge of most household decision-making, particularly with affairs that relate to issues outside the household. The prices collected from the household head were those which he expected to receive if the particular crop was sold that day or the actual price received that day if a sale had actually been made (Hesling, 1980). This manner of questioning has probably introduced some error in that the prices of more commonly traded goods were likely to have been more accurately reported. In fact, upon examining the data, it became apparent that the more common crops of the area such as millet, sorghum, groundnuts, and cotton were more adequately and completely reported than other crops, including maize. All prices were collected in local units and later converted to standardized units of Kobo per kilogram fOr the crop items and to Naira per unit for livestock items. The price information was further con- verted to Naira per kilogram via a computer program. 69 This series of conversions is not unusual for price data collected in developing countries and various examples can be cited (Hays, 1973; Ejiga, 1977; Thodey, 1969; and Gilbert, 1969). Even though conversions from local units to standardized units are common, there are no stand- ardized ways of carrying out such conversions since the units of local measure vary from one area to another and in some cases from one farmer to another. Such conversions, no matter how carefully carried out, probably tend to introduce some error in the data. The more care taken in the canversion process the greater the likelihood of reducing such errors. A table giving some idea of the conversion factors used in the data collection for this study is attached in the appendix section. In general a number of ways were used for the weight conversions: (1) using standard mean weights derived from market price surveys, (2) using mean weights derived from agronomic surveys, and (3) using weight from items sold by the farmer. - The producer price survey locations varied from one survey year to another since the households sampled are drawn from the sample of house- holds participating in the larger farm management surveys, which vary their sample households from one survey year to another. The survey procedures made it necessary to group the village data by district so as to obtain a more continuous time series over the years of the survey. However obtaining a time series was not the only ration- ale for the district level aggregation in this study. It was felt that~ . _ _' ' ' . K; g will-5 k" w ”-1335“; 70 in most cases the district is the logical unit for analysis of the producer price data even though information is lost relating to intra- district price structure. Place of Produce Sale Besides collecting price information the producer price surveys also recorded the location of sale under six categories. The six cate- gories are shown below. A large proportion of the data recorded in the survey relates to prices at the farmer's local market or the market nearest to the farmer (under 5 kilometers). Place of Produce Sale On respondent's farm In respondent's household In respondent's local market In nearest market under 5 kilometers In nearest market between 5 and 10 kilometers In amarket over 10 kilometers away 9‘ 9‘ f' 9’ 9’ 7‘ Of all the sales made between 1976 and 1979, over 47 percent were in the farmer's own local market. An additional 18 percent were in a market less than 5 kilometers away. Less than 2 percent of the farm produce was sold directly on the farm. House trade accounted fbr over 13 percent of sales confirming an earlier study by Hill16 that empha- sized the importance of such trade. The distribution of sales (Table 4.3) greatly point out the importance of rural village markets as a first point of sale for the producers. Less than 20 percent of the 16Polly Hill, Rural Hausa: A Villa e and a Setting. (Cambridge: Cambridge University Press, 1972). 71 .mmupea Fpeuoe pecueon m anonc use mmuwen emoauosa acousmm em «zone so a: out: w. mane woven mvsh "muoz .mzo>e=m more; F_uamg\emuauoea 39¢; soap mama pmcwupeo ”mugzom am.mm _e~.m_. peace mm.mm _m._P use.P sage Ex at peso posse: .e mo.mm em.» amo.P ,Ex op ease who; totem: .m ~_.pm am.m_ omm.~ e; m cage who; Hosea: .a mm.~e we.aa a Npm.~ gotta: Fades at .m Am.¢_ m_.m_ mac.~ eesoaeoo ezc em .N so._ so._ hem same ego co ._ Mvcmuemav mwcmugmgv. . ozmacmem ocmzcmcm zucmacmeu m>pum~=e=o o>pumpm¢ manpoma< copumuog moeoo 555: 555.55 .~ .5555305 «555. 555:5.53 uuaueoiocoeguoo 5:55: 55551.55» 5:5.»5305 ..< ._ .5055: .555. acoucoa o— a. acau.u.cu.m + —5>o— acougoa m an acuu.e_=a.m a .555. 5555555 — a. u:55.m.=u.m as "55: 1'13 .555.5. . 555.5 5 555.. .55..- 555.555 ..:\5. 55.55 55...: 5555555 .5. ..55.5. 555.5 5 555.. 555.5 555.... ..:\5. 55.55 55...: .55555. .5. .555.5. 555.5 5 555.. .55.- 555.55. ..=\z. 55.55 55...: .55555 .5. .555.5. 55..5 5 .55.. .55.5 .55.55. ..:\=. 55.55 55...: .555.e=.5= .5. .5.5.5. 55555.5. 555. ..5.. .5555.5. 5.5.555 ..:\z. 55.55 55.5: 5555555 .5. .55..5. .5.5.5 555. .55.. .555..- 555.55. ..x\z. 55.55 55.5: .555555 ... .5.5.~. 55555.5 55.. 5.5.. 55555.5- .55~.5~5 ..x... 55.55 55.5: .55555 .5. .555.5. 55.5..5 55.. .55.. ...55.5- , 555.555 ..x... 55.55 55.5: .555555.5: .5 +33. 3 5.5.. 555. 555.5 555..- 555.55. ..x\z. 55.55 55555 5555555 .5 ...5.5. 55..5 5 .55.5 .55.5- ..5.... ..x... 55.55 .5555 .555... .. .55.... _ 555.. .55. 55..5 55...- 5.5.5.. ..x\z. 55.55 555.5 .55555 .5 .5.5... 5.5.5 555. 555.. 45.5.5- 555.55. ..x\x. 55.55 .5555 .555.e=.5= .5 .55.... .55.. .55. 555.5 5.5..- 555.55. ..:\=. 55.55 555.555 5555555 .5 .55..5. 555.5 5.5. 555.. 555.5- 55..55. ..x\=. 55.55 55.5555 .555555 .5 .555... 555.. 555. 555.. +555..- .55.... ..xxx. 55.55 55.5555 .55555 .5 .555... 555.. 5.5. 55... 555.5- 5.5.55. ..z\z. 55.55 5555555 .55aces.5x .. 55.55.. a. 5: 3 3.5... 55853... 535...; 55.6 55:55.5 555555555 5.5.-5.5. .555.emmmo 5555 5. 555.55 555555Hm .555 5. 55555.555 5555. 555 555.55555 55.5555555 ..5 5.55. 114 or better. However, only four were significant at the 5 percent level. All feur significant equations‘were for maize. The three other equations, those significant only at the 10 percent level, were for sorghum farfara (2) ans sorghum kaura (1). All the seven significant equations showed negative trend in prices. Only two equations showed a positive trend-- millet in Malumfashi district as well as in Faskari district. However, the positive trend coefficients were not significant. It is clear that trend is not an important explanatory variable in f00d crop prices in the area when it is estimated over the whole period. This is supported by the low value of the coefficient of determination (R2)4 associated with the significant equations. The rising trend in the first half of the data is cancelled by a declining trend in the second half. I 4The coefficient of determination, R2, is defined as: 2 5 2 - -2 R l--[ut (Pt P)l Where, P 5 price in month t t P' 5 the overall mean monthly price of the series Ut = error term A perfect fit of the estimated regression equation will give an R2 of 1 since U = 0. When the mean price will do equally well in predicting the dependent variable, then R2 = 0. R2 simply gives the proportion of the variation in the dependent variabl explained by the regression equation. It is cautioned here that R is sample-specific as well as model specific. See Eric A. Hanushek and John E. Jackson, Statistical Methods for Social Scientists. (New York: Academic Press, 1977). PD. 58- 59. 'R is simply R2 that is adjusted to take into account the number of estimated variables in the regression equation. R2 = l-Hn-RZ) Where, n = sample size k = number of estimated coefficients including the constant term Thus'II'2 is generally lower than R2. While R2 can not be negative, R? can. All reported R2 in this study are adjusted and are hence R25. 115 The significance of these results is that for all crops considered, the trend coefficients are mostly negative and will be much more so for periods including months 20 to 40. Under the assumptions of the perfect competition model prices over time vary only by the cost of storage. The cost of storage thus serves as a means of evaluating such price variations. Factors influencing storage decisions are of paramount importance in the analysis. GilbertS in his discussions of factors affecting storage decisions by farmers in the northern parts of Nigeria mentioned the following as the most important factors: 1. The timing of harvest of each crop in relation to the harvest of other crops, 2. The importance of the subsistence component of each crop, 3. The timing of the need for cash on the part of producers, 4. The timing and importance of cash income from the sale of other non-staple food crops and from secondary occupatidns, 5. The expectations regarding the timing and size of the new harvest. The above factors were found to affect the storage of food crops in the fellowing manner: the timing of harvest of a crop in relation to other crops is important in that if the crop is the first to be harvested in the season, then there is very little tendency for storing it since it is probably needed for immediate consumption to cover possible short- ages in diet over the long dry season. There is also anticipation of. other crops coming soon. 5Elon H. Gilbert, "Marketing of Staple Foods in Northern Nigeria: A Study of the Staple Food Marketing Systems Serving Kano City," Ph.D. Dissertation, Stanford University, 1969, pp. 218-224. 116 The indication of the eagerness with which farmers welcome the ripening of new millet is reflected in the tradition of tumu.6 A sample of the heads of the just ripened millet are roasted on fire and served to family members before the period of full-scale harvest. This proce- dure is followed in the case of maize as well, although the term tgmg is not applied. Maize is also an early crop like millet. Thus, other things being equal, one would expect less storage of millet and maize when compared with late crops like sorghum. The importance of a crop in the subsistence of the farm family affects the level of storage undertaken. The more important the crop in subsistence, the higher the amount that will be stored for later use. The reverse is the case with a non-subsistence crop. This could account for the differences in the storage pattern of sorghum as compared with groundnuts or cotton. - The timing of cash receipts from other sources is important in storage decisions. rThe more the availability of income from other sources, the higher the likelihood of storing the f00d crop for later use. 4 Finally, the expectation with regard to yield prospects of crops in the field could increase or decrease the amount of staple food set aside for storage. Gilbert pointed out that the number of factors that influence storme decisions are many and introduce a lot of uncertainty regarding the process.7 élgpg refers to the just-ripened heads of millet as well as the roasting tradition. 7Gilbert, op. cit., p. 222. 117 With this as a background the results of seasonal price analysis of producer prices in FADP districts are presented in Table 6.2. The sea- sonal indices were calculated using the method of moving averages. Figures 8 The 6.6-6.9 give a visual representation of the seasonal indices. figures show that fer sorghum farfara and sorghum kaura price indices were lowest in the month of December, which is the month of harvest for these crops. Prices during this period average about 75 to 95 percent of their yearly average, depending on district. Prices of sorghum, both kaura and farfara, show a smaller decline in seasonal prices at harvest in Kankara than in other districts. From December to February prices show a strong upward movement for sorghum while after February they show a gradual decline until the month of May. From May to August, prices increase again. From August to December the prices of sorghum decline to their season low. Thus, there is a strong correlation between the timing of the low prices for sorghum and the harvest period. Similar strong correlation of the timing of high prices with the period just before harvest is n0t evident. There appears to be two distinct peaks in the price of sorghum over the season, with the first peak just two months after harvest and the second one about four months befbre the harvest of new crops. Gilbert observed only the second peak and thus maintained that timing of highs and lows in prices fellowed closely the period just before harvest and immediately after it. A likely explanation of the seasonal price behavior observed may be related to increased production of staple food crops (See Appendix B) which might have led to accumulation of local stocks as the project devel- oped. If this was the case then the realization of the existence of 8Calculation of the seasonal indices was based on nominal prices. 118 .anm:£<\ma9. hue mum 03¢ 55555: p p — p F _ ...... 2:... >5. ...—c me: am..— as... u-¢=arya 5.3:. 91.. $55. I cash—«c.5115. V gaping. 5.5.135 2555.5 3: 2.6 9535. 555.9 5555 .55 555.55 55555555 55 555.55. .5555555 2. on Do.— a: can on— DVu HQ'Un-UOUD WQMWOCMP 126 percent per month compared to 6 percent for sorghum kaura. 0n the other hand, millet had close to a 9 percent increase per month while maize had just over a 7 percent average increase in price per month. The 8 percent increase in the case of farfara is due to the occurrence of the high prices in February instead of the normal period in August and thus eXaggerated the increase in price on a monthly basis. The actual in- crease is probably closer to 6 percent as shown in the case of sorghum karua (see Table 6.2). The variation in the timing of the high price points appears to be due to either poor data quality or due to the complexity of storage decisions and the timing of the release of the crops in small village markets in the FADP districts. Further study on this is required before definite conclusions can be made regarding the seasonal patterns observed in this study. A second calculation of the seasonal indices based on real (deflated prices) resulted in a more consistent interval between the low and the high indices. The timing of the low indices is still consistent with the harvest period but the high indices occur much earlier than reported by Gilbert. The high indices for sorghum farfara occurred in February and March rather than in June to October. Similarly the high indices for millet were also in February and March rather than 9 These earlier high prices in June and July, as reported by Gilbert. could be due to localized phenomena which are not clear at this point. This seems to be the case since the prices decline after the initial high and then rise again to the normal or expected period of high prices in June, July or August. 9 Ibid., PP. 231-233. 127 The early increase in price could be due to initial withholding of the crop from the markets by farmers since other sources of income are generally available to them at the time. Other income sources include the sale of groundnuts, cotton. and cowpeas. Since sorghum is the most important staple crop in the areas, it is stored fbr later consumption provided cash needs are met from other sources. Estimation_of Seasonal Price Increase Using Regression Procedures Regression equations fitted to the rising portions of seasonal prices can be fitted to estimate the extent of price increase over a season or an average rise per season based on a number of seasons put together. This procedure has been used by Ejiga.8 The simple regres- sion equation is given as: Pt = a + bT + u Hhere, Pt = the average seasonal price at time t T- = time in months starting from the lowest average seasonal price a,b == constants for intercept and slope, respectively u = time error term A summary of fitting this equation to data covering the period of 1976-1979 for various crops and FADP districts is shown in Tables 6.3 and 6.4. Results in Table 6.3 were obtained using deflated prices (1970 = base). The period of rising prices varies from crop to crop and so the periods used for the estimation is indicated for each crop. 8N. 0. Ejiga, "Economic Analyses of Storage, Distribution and Con- sumption of Cowpeas in Northern Nigeria," Ph.D. Dissertation, Cornell University, 1977, p. 285. 128 .....a... ...... 53...... 3355-22555 5...... 53.-.... .833... ... .N .53qu 5o .0»: a... «5 2.3.5.5.. 05- 28—5 .5 5..- unoosouc. .5 53 «...-07» 3‘ .— 33.! .... a... .. ... ... ...... ...-.58 .5... ...... 2.3.... .... .5... 5... ... 5...- 5..... 61...... ...... :3. =35 .... s... 8.. 5.. 2..- ....5. .118. .5... . :3. . ...... ...- .... .... ... 2..- ...8. .3.-.8. .5... :3. 5.5-...... 5..- ...: .... ... .5..- 8.... ...-.8. .5... ...-5.... ...-...... .... ...... .... ... .....- .n.... ...-.8. .5... ...-...: to... .. .- ...... 3.. ... a..- ... ... .3.-...... .5... ...-...... 23.-...... 5... .... .... 5.. .5..- ..z... 61...... .5... :3. ...-...... ... .5... .... ... .... ...... .3.-...... .5... .....5 5.5-...... ... a... .... ... 8.. ...... 1.3-.... .5... ...... ...-...... 5.... ...... .... ... .... ...... 82.-.... .5... ...... to... ... ...... .... ... ...: ...... {...-.8 .5... ...... ...-3.... .... .... 8.. 5.. ...... .... ....1-68 .5... 2.5. . ...... ... a... 8.. ... 5... ...... .3.-...... .5... :3. ...-51...: ... 2... 8.. ... 8.. ...: . .3.-.... .5... ...-5.... ...-5...... 3. ...-9.3... 8....-. .... .— ..... .823... 3.2.8 8.8. .25 8...... ...u nun 3.... ...... .3... 5... .8. .. =2... .3.-.... . . = .525 .8. .... .3.... .332. .288... ...-.2... ... .83.... .8... 2...... .... 3 .3... 3...... 8.3:... m6 535.. 129 ......5. 55... 5.55.... 555555....55555 5e... 5.5...55. 555...... ... .5 .5555... 55 .5...— u... 55 55.5.5.5: .5. 5.55.. 55 5... 555553... .5 55. .55—55-5 .3 .. "5.55.. ..5- ......5. 55.5 5... 8.55- .....5. 53.555 .25... 5...... .53.... 5. .- 5. ... 55.5 .5. .....- ...5.. .53.... .25... ...... ......5. ..5- 55.. ...5 5.. .5..- ...55. ..:.-.5.5 ....... ..... .55... 5. .- 5...... 5... 55. .5..- ...... .3.-.85 .25... 3.5.... ......5. .. .- .5... 55.5 5.. .5..- 55.... 52...... .25... 3...... ...... .. .- .5... ...... .5. .5..- 5..... 5..... 5.5 .25... .35.... 25.3.... ... 5.... 55.5 ... 55.. 55.8. 5.1-.555 .5... 5...... .53.... 5.5 5..... 2.5 5.. .5. ...... ..2-555 .25... 5...... .55... 5.5 5.... .... ... ...5. 8.... 5.1-.555 .25... .....5 5.5.3.... ... 2.55 55.5 ... 5... 2.555 .55-.555 .25... ...... ......5. 5.. 55.5. 55.5 .5. 5.5. ...... .55-.555 .25... ...... ...... 5. .. 5.. .5. 5... 5.. 5.... 5. .5. 2.5-.85 .25... ...... 5.35.53. 5.. 55.. .... ... 5... 55.5.. 52...... .25... ...... ......5. 5.. 5... .... 5.. 5... 55.5. ......5.5 ....... ..s.. ......5 ... .... .... ... 55.5 .5. ... 5.1-.... .25... 3...... .53.... 5.. ....5 .... ... .... ....5. .3.-.... .25... .3.... .53.... 5.5 5... 5... ... 55.5 .5... 5.1-.... . .25... 9...... ...... .... 3.2.5... 8.5..-. B u ...... 55.5.8... 5...... 8.... .25 52.5.3 .5” ......un 5 5...... 8.. ......5... . .59.... .. 5.855.. .... .85.... ...5... ....l... 5555 .5 85.... 5.535... .3558... ...5 .5 .35.. 55...... 8.3.5... 5 5. .5555 .u 55!... 5.5 535.. 130 The results show a striking inconsistency in price increase from one year to another. For the 1976/77 season all slopes were positive and range from about 3 naira per ton per month to over ll naira per ton per month: Significant results for 1977/78 and l978/79 all had negative slopes indicating a decline in real prices for the food crops over-those seasons. The extent of the decline ranges from 3 to about 7 naira per ton per month, excluding the 12.77 naira per ton per month recorded for millet in Kankara district. 'If the intercepts represent the beginning low prices for the crops, then the ratio of the slopes to the intercepts give the average increase in prices per month over the season.10 The results of calculating the increases per month from this method are shown in the last columns of Tables 6.3 and 6.4. They show that in the l976/77 season farfara prices increased an average of 4.5 percent, kaura prices 4.8 percent, millet prices 6.8 percent and maize prices 6.l percent per month. In the 1977/78 and 1978/79 seasons, however, all prices declined in real terms. The ratios show that kaura prices declined approximately l.75 percent - per month,.farfara prices 1.90 percent and millet prices 2.l percent per month. During these seasons no significant decline in the price of maize was recorded. . These results indicate that only during the 1976/77 season were there significant seasonal price increases. The two subsequent seasons showed a significant drop in seasonal prices for sorghums and millet, 10See Nathaniel 0. Ejiga, “Economic Analyses of Storage, Distribu- tion and Consumption of Cowpeas in Northern Nigeria." Ph.D. Dissertation, Cornell University, l977. 13] but not fbr maize. This raises the issue as to whether the complaints about the saturation of maize markets in the area were genuine. In our results maize fared better than the other fppd crops. Even though maize prices did not increase significantly in 1977/78 and 1973/79, they did not decline significantly either. However other food crop prices declined as reported above. The results also show how volatile any decisions regarding specula- tive storage and hoarding could be from one season to another. while storage was potentially successful in the 1976/77 season, there was no potential fOr profitable storage in the two subsequent seasons, 1977/78 and 1978/79. The estimated seasonal price increases can be compared with estimated storage costs to indicate the extent of possible profit in the 1976/77 season. Comparison of Seasonal Price Increase with Estimated Cost of Storage CompariSon of seasonal price increase with the cost of storage is often used to get some idea as to how well a marketing system is allocat- ing resources over time. The comparison is useful in clarifying accu- sations pointed at middlemen as the underlying cause of price increases. Although there was no case of a price doubling over a single season in the data in this study, it is often claimed that this doubling of price occurs between harvests. The general indication is that if the seasonal increase in price is higher than the cost of storage, then there is an indication that traders 132 or other participants in the marketing system could make abnormal profits from the storage process.n Unfortunately the cost of storage is an area that is greatly under- researched and thus very little concrete information is available in this regard. This lack of information weakens the strength of seasonal price variation studies. The shortage of information with regard to storage is even more acute at the producer level in northern Nigeria, even though it has been shown that most of the storage undertaken in the area is generally done by the farm producers rather than urban or rural traders.12 In the FADP survey data no storage costs were collected. This is not very surprising since the surveys had other points of emphasis. A solution to this problem is to use estimates from other studies. Hays (1973)13 estimated the major components of storage cost based on data collected from traders in the urbanmarket at Zaria and second- ]4 The major cost components were rent, ary information from Giles. interest, labor cost for guarding the stored food grains, and amount of losses over time. Rent was estimated at about 15’shillings per hundred nH. M. Hays, "Organization of the Staple Food Grain Marketing Sys- tem in Northern Nigeria." Ph.D. Dissertation, Kansas State University, 1973, p. 156. 12Gilbert, op. cit., p. 272; Hays, op. cit., p. 163; and Ejiga, op. cit., p. 219. See also Barbara Harriss, "There Is Method in My Madness: Or Is It Vice Versa? Measuring Agricultural Market PerformanceJ' Food Res. Inst. Studies, Vol. XVII, No. 2, 1979, p. 205. 13Hays, op. cit., pp. 156-164 14P. H. Giles, "The Storage of Cereals by Farmers in Northern Nigeria." Samaru Research Bulletin, 42, Institute for Agricultural Research, Ahmadu BelloUniversity, Zaria, 1965. 133 sacks of sorghum or millet. Interest was taken to be the current rate at which private commercial banks were making short-term loans--l per- cent per month or l2 percent annually. Storage loss for sorghum and millet was estimated to be 5 percent per year. These estimates were then converted to the basis of storage cost per ton per month. Rent facilities: Stall fees 1.5 naira per 100 sacks of grain--sorghum or millet Guard service 2.0 naira per month Subtotal 3.5 naira per ton per month Interest charges 1 percent per month Storage losses _5 percent per year (0.42 percent per month) Depreciation of sacks 0.15_naira per ton per month Grand Total 0.5 naira/ton/month + 1.42 percent of price/mo. The calculations done by Hays regarding the cost of storage are ambiguous in many instances and it is not clear how the final figures were computed. For example, the final figure regarding rent of facili- ties was not clearly explained. The same is true in the case of how the percentages relating to interest charges and storage losses were used.15 Hays (1975)16 calculated storage costs in rural areas based on the use of rugby, which is the common storage structure in rural areas. His total annual estimated storage costs for gumbo use are about 3.48 naira per metric ton. Ejiga‘7'has also done same work in trying to assess a ”Ibid., p. 37. I6Hays, op. cit., pp. 156-l64. 17Ejiga, op. cit., PP. 246-263. 134 reasonable estimate of the costs of storage for cowpeas,but has met with little success. His estimates varied from less than one-half of a naira to 49 naira, depending on assumptions involved as well as location of the area. Ejiga speculated that the wide divergence could indicate either enjoyment of monopoly conditions by owners of storage facilities or could be due to the availability of poor data. The costs of storage used in these studies are very crude and have to be regarded as indica- tive pending better estimates. Given this divergence in the estimates of storage cost an approach is taken here to_use a basic rate of 0.5 naira per ton per month as the cost of rent for facilities plus/minus depreciation and losses. In addition, an interest cost of 24 percent per year is assumed to cover the cost of invested capital. The interest rate is divided into 12 equal rates of 2 percent per month. But since the monthly rates are compounded this came to an effective interest rate of 26.82 percent per annum. This rate is not unrealistic given the wide divergence of the sources of funds to farmers and varying rates associated with the different sources. A better way of arriving at an appropriate rate would have been to get the various interest rates and amounts borrowed at each rate and then weight them accordingly to arrive at a weighted interest rate which can then be used fbr the analysis. This is not possible here since there are no figures on sources of credit. Further justification for the interest charge is provided by FADP which charged l0 percent for its loans plus the intangible cost of having to get certification from village leaders before the loan is approved. The FADP loan was a loan in kind for the purchase of inputs and some farm implements. 135 The harvest prices plus hypothetical storage costs for food crops at Malumfashi for three seasons are shown in Table 6.5 and graphed with .actual food crop prices in Figures 6.ll-6.l4. ‘ The results indicate that for the two sorghums in two of the three seasons examined, there is an opportunity for carrying out profitable . storage activity. The near perfect substitute relationship between the crops is shown in the pattern of the charts which follow identical paths fOr both crops. The biggest margin between actual and expected price was in 1977/78 with a smaller marginin l976/77. For l978/79, there was a positive margin fbr only the first few months of the season and a loss for the rest of the season. The timing of the period of maximum margin varies but generally it comes toward the end of the crop season4-just before a new harvest. This was not the case for farfara in the 1977/78 season when the maximum margin occurred much earlier in the season (See Figure 6.ll). For maize and millet all seasons showed a margin of actual prices over expected prices. For maize the largest margin was in the l976/77 season, while for millet it was in the middle of the latest season, 1978/79. . The margins shown in the charts are more than the actual margins in real terms since nominal prices were used. Inflation rate in the economy at the rate of 17 percent per annum reduces the margins considerably. 17 As pointed out by Hays the presence of excess after subtracting actual price from expected price only indicates the possibility of 17H. M. Hays, "The Marketing and Storage of Food Grains in Northern Nigeria." Samaru Miscellaneous Pa er, 50, Institute for Agric. Research, Ahmadu BelWUniversity, Zaria, l 5, p. 89. 136 Table 6.5 Harvest Prices Plus Hypothetical Storage Costs for Food Crops FADP, 1976-1979 Maize Kaura Farfara Millet Naire er ton Aug. 1976 - ( - p ) - - Sept. 154.0* - - 115* Oct. 157.58 - - 117.8 Nov. 161.23 - 120.66 Dec. ' 164.96 115* 124* 123.57 Jan. 1977 168.76 117.8 126.98 126.54 Feb. 172.63 120.66 130.02 129.57 Mar. 176.59 123.57 133.12 132.66 Apr. 180.62 126.54 136.28 135.82 May. 184.73 129.57 139.51 139.03 June 188.93 132.66 142.80 142.31 July 193.20 135.82 146.15 145.66 Aug. 197.57 139.03 149.58 149.07 Sept 241* 142.31 153.07 174* Oct. 246.32 145.66 156.63 177.98 Nov. 251.75 149.07 . 160.26 182.04 Dec. 257.28 163* 170* 186.18 Jan. 1978 262.93 166.76 173.9 190.40 Feb. 268.69 170.60 177.88 194.71 Mar. 274.56 174.51 181.94 199.11 Apr. 280.55 178.50 186.07 203.59 May 286.66 182.57 190.30 '208.16 June 292.89 186.72 194.60 212.82 July 299.25 190.95 198.99 217.58 Aug. 305.74 195.27 203.47 222.43 Sept. 312.35 199.68 208.04 227.38 Oct. 178* 204.17 212.70 223* Nov. 182.06 208.75 217.46 227.96 Dec. 186.20 182* 196* 233.02 Jan. 1979 190.43 186.14 200.42 238.18 Feb. 194.73 190.36 204.93 243.44 Mar. 199.13 194.67 209.53 248.81 Apr. 203.61 199.06 214.22 254.29 May 208.18 203.54 219.00 259.87 June 212.85 208.12 223.88 265.57 July 217.60 212.78 228.86 271.38 Aug. 222.46 217.53 233.94 277.31 Sept. 227.40 222.38 239.12 283.36 Oct. - 227.33 244.40 - Nov. - 232.38 249.79 - Note: * Actual harvest price for the season 137 ¢N\mm u mmumm "mu\ns n mmump mnm\msmp u mpupv mgucoz h» an . nu .— owhuumxu 3-.- 49:9: .11 memp-m~m_ .pu_eum.o .gmace=_az .azacaaa Lac mou_aa eaaumaxu tea _aaou< pp.o usamwm 8— emu can 2: Q0!!- ZQ’HS-or-U l—O: ZMOI-LIU 138 Amk\m~ "_Nm-m~ “mm\- m m~-m_ m-\m~m. u m_-_v “gage: on n— 2: cu— new PC: Q0!- 2043qu zCU-v-S-OU aubuumxu 1-... 4.5-.06 ll. as m~m_-m~a. .uupzum.o .gmacsapaz .azaag Lac mau_ca emuaaaxm new Paapu< 2.8 2.5: 139 o Amu\mn u.~mnm~ “ms\sn n mmump mh~\cnmp u mPu—V usage: on mu 3 . . \s\\.ii .xx. m w . .\\\\.o \\ umaomaxu --- posuu< III. msmpumsm— .po—gum.o _cmmwsapax .u~.oz so» mauve; umuuogxu ecu paauu< 2.... 23: 8.. can 2: 09¢ 0.03. 20498090 I—O: ZMPLM 140 Amsxmm u munum mm~\~n u mmump usm\onmp w mpu—v mcucoz hm mu mu — cwbuumxw ..... 42:0: 1.1 mumpummmp .aurgpmpo —=mawsspmz .uu—ppz 50$ mac's; umuumaxm ¢—.m mgaavm no. i—O: om— DON ZOHLw-U 0mm com 0.08- 0mm cc. ZlUw-I-M omv Dom new Pasau< 141 profits and does not show whether they were actually attained by traders. In fact his study of traders indicated that they do not engage in any substantial storage beyond a turnaround stock of approximately one month. Thus, if there is any profit to be made from seasonal price increase, it is likely to be made by the farm population who carry out storage. Summary The analyses carried out in this chapter indicated that fbod crop prices at FADP area increased only during the 1976/77 season but there- after declined as the project developed. The decline in food crop prices are even more prominent when real prices are considered. Analy- sis of seasonal price increases indicate that significant increases in prices were recorded only during the 1976/77 season. Thereafter prices declined 1.8 percent per month for sorghum farfara, sorghum kaura and millet. ‘Hithin individual seasons maize prices did not decline signi- ficantly. Comparison of seasonal price increases with estimated storage costs did not show the existence of excessive profits for sorghum or millet. Estimated seasonal price increases in real terms were positive only during the 1976/77 season. The increases were 3.80 naira per ton per- month fbr farfara, 3.88 naira per ton per month for kaura and 8.13 naira per ton per month for maize. The estimated storage cost per month (on the basis of 27 percent interest for 8 months and storage overhead costs of 0.5 naira) came to about 4.00 naira per month per ton for sorghum and .nnllet and 6.5 naira per ton per month for maize. Thus in the 1976/77 season maize showed an excess of about 1.63 maira per ton increase in price over storage costs. This suggests a potential fbr profit in 142 maize storage for that season. The subsequent two seasons indicated clear losses relating to storage of sorghums and millet, and possible losses for maize. Since the above results were based on real prices, taking into account an inflation rate of about 17 percent, another method of com- parison was done based on nominal prices. These are plotted with expected prices given estimated storage costs. The plots showed ample room for profiting from storage of foodcrops in all three seasons except for sorghums in the 1978/79 season. However, these possible gains would be less than depicted by the charts when considered in real terms. CHAPTER 7 SPATIAL PRICE ANALYSIS This chapter discusses and analyzes the pattern of staple food price behavior across space within the project area of FADP. The chap- ter also looks at inter-commodity price relationships. The intention is to test the hypothesis of spatial pricing efficiency within the FADP districts. The results of such analyses could serve in pointing out weaknesses of the system and hence indicate a possible avenue for remedial action. If on the other hand the system turns out to be operating efficiently, given the criteria of the analysis, this could help in dispelling some of the allegations of inefficiency often directed at the system.1 Efficiency is not an end in itself. Although a marketing system is deemed efficient, it may still operate at high costs. Given the pre- vailing conditions, it would be considered efficient since no other com- binations of resources will lower costs. Therefbre efficiency in this context means that marketing services are being provided at the lowest current costs.2 Similarly a system is efficient in terms of pricing 1See S. M. Essang, "The Middlemen in the Domestic Marketing of Palm Oil: Asset or Liability." Bull. Rur. Econ. andJSoc., Vol. 3, No. l, 1968. Essang came to the conclusion thatéfli . . far from proving harmful to the producers, the distributors' functions are essential to continued 'and increased production of palm oil." See also The World Bank, "Apprai- sal of Funtua Agricultural Development Project, Nigeria," (Washington, D.C.: The Horld Bank, 1974), Annex 12, pp. 12-13. 2Van Roy Southworth,"Food Crop Marketing in Atebubu District, Ghana," Ph.D. Dissertation, Stanford University, 1981, p. 115. 143 144 only given the current conditions. Thus there is more in the analysis than the final decision as to whether the system is efficient or not. There is need to examine ways of changing current conditions in such a way that further reductions in costs of marketing services become possi- ble. There may be cases in fact where marketing costs may have to rise in order to satisfy certain preferences of the consumers. Thus consumer satisfaction is also a legitimate concern of marketing systems analysis. Given this short discussion it can be seen that it is possible to have multiple and conflicting goals in analyzing market performance. Spatial pricing efficiency is one of the important factors in marketing system performance. The system should be able to allocate commodities across space at minimum cost. If it does not then there is a need to look at the sources of the problem and examine ways of lower- ing the transfer costs. If the system does act efficiently under the circumstances being considered, then further examination of new circum- ‘stances is needed in an attempt to identify an even more efficient operation of the marketing system. Spatial Price Analysis Since the prices being analyzed are essentially village-level prices, then there is a need to look at some of the problems alleged to be associated with village-level pricing mechanisms. It is claimed that village-level pricing does not reffiect the forces of supply and demand. It is also alleged that cultivators receive prices that are much lower than in town markets. The low prices are said to be due to monopolistic practices of village traders, farmer cash requirements at harvest which necessitates selling most of the crop in the immediate post-harvest 145 period, activities of village moneylenders, and ignorance regarding prices.3 Under the conditions of perfect competition the price spread between two locations that trade with one another cannot exceed the cost of transfer of the product between the locations. If the price temporarily exceeds the cost of transfer, then the commodity will be shipped from the location with the lower price to the one with the higher price until the two prices become the same, thus resulting in no incentive fbr the prOduct movement. This is the ideal situation and is rarely met in the real world where the model assumptions are impossible to meet. However, it is still widely used as a guideline in measuring efficiency over space. These analyses use correlation coefficients between market locations as indicators of spatial market integration.4 Lele explained the concept of market integrationwith reference to the work of Cochrane5 which stated that markets of agricultural commo- dities in developing countries are closely interrelated in the sense that price formation in one market is related to the prices in another- 3Uma J. Lele, Food Grain Marketingoin India: Private Performance and Public Polic . (Ithaca,*New York: rnell University Press), 1971, pp. 21-22. . 4Muhammad Osman Farruk, "Structure and Performance of the Rice Marketing System in East Pakistan," Cornell International Agricultural Development Bulletin 23, Cornell University, 1972; Lele, op. cit.; H. M. Hays, Jr., "Organization of the Staple Food Marketing System in Northern' Nigeria," Ph.D. Dissertation, Kansas State University, 1973; N. 0. 0. Ejiga, "Economic Analyses of Storage, Distribution and Consumption of Cowpeas in Northern Nigeria," Ph.D. Dissertation, Cornell University, 1977: and Southworth, op. cit. .5w. w. Cochrane, "Markets as a Unit of Enquiry in Agricultural Economics Research.“ Journal of Farm Economics, XXXIX, February 1957. 146 market. The interrelation in price movement between two markets is what is called market integration.6 The extent of the influence of the pricing process in one market on another is measured by the correlation coefficient of wholesale prices in the two markets for a given commodity. Under conditions of perfect: competition the value of the correlation coefficient will be 1.00. How- ever, the assumptions of perfect competition are never completely met in real situations and as such the correlation coefficient is generally always below l.00--often well below the ideal. Lele suggested the fol- lowing reasons to account for low correlation coefficients of prices between markets:7 ' d . Lack of perfect mobility due to transport costs 2. Existence of transport bottlenecks 3.. Uncertainty on the duration of price difference between markets 4. Lack of scientific grading of produce with the result that prices do not refer to equivalent grades in the two markets 5. Poor dissemination of information regarding market con- ditions The above factors are expected to have substantial effect on the value of correlation coefficients calculated fbr the FADP districts. Transportation problems, particularly in rural villages, have been iden- tified as a major constraint. Similarly, grading of staple food crops is rarely performed except by visual, on-the-spot inspection by purchas- ing consumers. Although project activities included the broadcast of _— 6Lele, op. cit., p. 22. 71bid., pp. 23-25. 147 market conditions over the radio network, fbrmal price information dissemination to producers, traders, and consumers is minimal. Given these factors correlation coefficients are not expected to come close to the ideal value of 1.00. Data Problems Relating to Spatial Price Analysis The major limitation of the data set used for the analysis in this chapter is the grouping of the data on the basis of the five districts of-FADP. Since the analysis is based on district level data, the pre- cise location of the prices is taken to be the district headquarter of each of the districts. This is a major limitation; however, the group- ing is necessitated by the lack of complete series data fbr any village. Since most of the locations with useable data were generally located close to the district headquarters, the aggregation conditions nay not pose serious problems. More remote areas tend to have incom- plete data and a lot of them had to be rejected for-that reason. Neverr theless, it has to be pointed out that the results as depicted cannot be said to apply to any one specific location in the district, but rather are an average for the district as a whole. Another point is that the prices used are producer prices--an approximation of farmgate prices--not wholesale prices which are pre- ferred in such analyses.8 Finally, the sampling methods used in the selection of villages for the survey were based on the needs of agronomic surveys and have little to do with the marketing of staple foods. As such the villages selected 8Hays, op. cit., p. 127. 148 cannot be easily categorized in relation to their importance as market centers for staple food crop marketing. There is a big difference in terms of the amount of data collected from individual district villages. The most complete data set came from Malumfashi and Bakori districts while Faskari and Kankara had less representation. In fact Funtua district is excluded from the analyses for lack of complete data. In addition, the villages included in the survey kept changing from one year to the next, thereby necessitating the aggregation of the price data based on districts rather than on individual villages as the ideal situation would have required. 0n the other hand the collected data has taken into account varie- tal differences in the case of sorghum since data was available for the two distinct sorghum varieties in the area (i.e., sorghum farfara, the 'white Sorghum variety, and kaura, the yellowish sorghum variety). This distinction was never tried before in the reported studies. The two are easily distinguishable from one another and have some characteristics that result in subtle differences in the taste of prepared foods. This differentiation between the two sorghums has not been taken into account in earlier studies partly due to lack of detailed differentiated data as well as the apparent high substitution between the two sorghums. Villages in Spatial Price Analyses Studies Due to the village selection procedure used in the surveys a large number of villages were included, many of which had complete data records fbr less than a year. In Funtua district major villages with useful data included Tafbki, Mahuta, Ruwan Godiya, Mairuwa, Goya, and Dandume. In 149 Malumfashi district the villages included Yaribori, Dandarai, Karfi, Ruwan Sanyi, Dankanjiba, Yargoje and Borin Dawa. Faskari villages in- cluded Damari, Sabua, and Yankara. Bakori district included Jiba, Kabomo, Kakumi and Kurami; while Kankara district included Kukasheka, Gatakawa and Zango (See Map 4.1, p. 54). All the villages listed above, with the exception of district head- quarters, are Strictly rural in that each is made up of less than 20,000 inhabitants. The villages vary in terms of their proximity to the head- quarters as well as the quality of main access road that links the vil- lage to the headquarters. Some villages, like Karfi, Kurami and Yankara. are on a main, all-weather, hard surface road; others have portions of the access road that are untarred, dust roads; while in some further cases access is possible only by footpaths, particularly during the rainy season. Distances between the district headquarters, although generally small, can be up to 65 miles, as in the case of the distance between Faskari and Sabua (See Map 4.1). Spatial Integration of Staple Fbod Prices within FADP Districts .A visual indication of the pattern of spatial price behavior can ' be discerned from Figures 6.01 through 6.05. The figures show that 9 price movements for sorghum and millet are more in consonance with one another across the districts than the prices of maize. The figures also 9Allen R. Thodey, "Analysis of Staple FOod Price Behavior in Hest- ern Nigeria." Ph.D. Dissertation, University of Illinois, 1969, pp. 27- 29. 150 indicate that there is considerable difference in price consonance fbr any of the crops as the time periods change. For example, in the case of millet (Figure 6.04) the prices in different districts were more con- sonant with one another in the first 30 months of the data indicating that the earlier prices will have a higher correlation coefficient than the last year of price data. . In order to examine the level of interbdistrict integration in terms of staple food crop pricing, correlation coefficients were calcu- lated for sorghums, millet and maize prices fbr the FADP districts. As suggested by Jones and many researchers who used the method, the levels of the calculated bivariate correlation coefficients serve as indicators of the level of spatial integration. The coefficients reflect the level-of information flow between the two locations involved and measure trading connections (Hays, 1973). The results of the calculations are shown in Tables 7.1-7.5. The results are for the entire period of 1976-1979. The number of months for the actual calculation was 41. With this quantity of observations fbr a correlation coefficient to be significant at the 5 percent level, the correlation coefficient only has to have a value of .30 or higher. If, however, the number of observations are limited to 12, then the correlation coefficient has to be higher than .576 fer the same level 10 of significance. Thus all the correlation coefficients reported here are significant at the 5 percent level. All correlations fbr sorghum and millet are over .80 except for the sorghum farfara correlation between Faskari and Kankara district loEjiga, op. cit., p. 290. 151 .uopcumvu on» to; mcopau>smmno 3mm co» ou mac uzo Hemp mp uuwgumwu mauczm “muoz .3auz¢<\aooo move; . ngmucmum mmmsm>< :15 22-32 .82me .35. 292 «gauge; Eanmgom mo mmuwgm gmuauogm a—gucoz toe xpgumz zowumpmggoo F.~ mPnMP .uuwgum_o mg» com mcorum>gmmno 3mm ooa ou use use pump mp uupcumvu capes; "muoz .=auza<\aomo mu_ca ugmucmum wmmgm>< A_e..=v mum—-asmp .mppeaom_o aogmmno guy so» on mss use new” my auwgumps «spasm “mpoz .samza<\asmo moves sgmscmum mmucm>< ape u es mus.-a~m_ .asp_tumes aogmmao saw oou on mss use use. mp puwcpmws asucsu ”muoz .samza<\aomo moves sgsscmum mmmgm>< A_e u av asap-a~ms.amppsasm_s as< mn.- so.mm mm.- m—.o~ mm.- cgmxcmx oo.n~ om.~m mm.mm mm.m~ mm.¢~ _swxmou mm.o~ mm.pm NP.NN me.om mm.¢~ Pgoxsm mm.m~ ~c.om om.m~ em.m~ um.mm wnmmmsspmz mmssm>< ump_wz mnwsz sessx Esgmgom seamen; Essmgom uuwcumwo Aucmugmmv mumpnmmmp .mHUwLumwo ao move; we mucmwummmmou m.n m—nmp 156 which was .79. Correlation coefficients for maize were lower than for sorghums and millet but all were above .70. While sorghUms and millet showed a large number of coefficients in the range of .90 and above, maize had none in this category. The pattern of trading linkages revealed by the correlation analysis indicates that Malumfashi district is more closely linked to other districts than any other one. Malumfashi producer prices are most closely related to Bakori prices and least related to those of Faskari district. This pat- tern closely agrees with the pattern of roads in the FADP area. Faskari is more isolated as a district than all others: on the other hand, Bakori, Malumfashi and Kankara are connected to one another via an all-weather, tarred road. ’ ' The correlation coefficients are much higher than those obtained by Hays and Gilbert in their studies. For example in the case of Hays' study none of the correlation coefficients were above .90. In fact only 1 percent of the correlations fbr millet and 1 percent for sorghum were recorded by Hays at .80 or above for the period of 1958-65.11 However, when considering the Zaria area Hays used personally collected data and found that correlation coefficients were approximately .90 fbr sorghum ‘2 Thus while competitiveness while those fer millet were all above .70. may be low when markets are Spatially separated by long distances, local conditions within individual areas are highly competitive. This conclu- sion from Hays is partly supported by the results presented in this study with respect to the high degree of interdistrict price integration. 1IHays, op. cit., pp. 130-132. 121bid., p. 132. 157 However there is no corresponding data for a wider geographical calculation to examine the integration of locations at that level. The results from Gilbert's13 study also revealed a similar pattern of generally low correlation coefficients with regard to regionally dispersed markets, except for cowpeas which had a well integrated market Aboth regionally as well as nationally in southern markets, particularly Ibadan. This exceptional performance of the cowpea market integration has been further confirmed by Ejiga's study.14 There are also indications that in general the level of locational integration is increasing over time. Analysis fOr the earlier periods tended to have lower correlation coefficients than more recent periods. Part of this is due to the improvement in price data collection methods as well as the improvement in infrastructural facilities available in developing countries which allow for more rapid price information dis— semination. . The reanns for these relatively high levels of correlation coeffi— cients could be attributed to a number of factors. First, the data quality is probably much better than earlier data from similar studies in the area. .This improvement in data quality over time has also been observed by Hays in his study where he fOund that the correlation coefe ficients from data he collected in the early 1970's were higher than those from crop and weather reports of the 1960's. 13Elon H. Gilbert, "Marketing of Staple Foods in Northern Nigeria: A Study of the Staple Food Marketing Systems Serving Kano City." Ph.D. Dissertation, Stanford University, 1969. ‘ 1”’Ejiga, op. cit. 158 Second, the high correlation coefficients are an indication of integrated local rural markets which does not necessarily imply overall integration at a regional level. The region of the FADP location in- volved only contiguous districts that attend more or less the same rural and town markets. This could lead to improved trading communications, especially when contrasted to what the case would be if markets outside the project area had been included. Thus this limited geographical coverage likely improved the values of the correlation coefficients obtained. A A recent study by Southworth in Atebubu district of Ghana also reported higher correlation coefficients than those reported by earlier studies in Nigeria by Jones, Hays and Thodey. Southworth also used a wide range of locations across Ghana and thus had a diverse group of markets. Southworth used data collected by the Ghana government which' was sinfilar to the crop and weather report used by Hays, Gilbert and Ejiga in their northern Nigerian studies. In most cases the correlation coefficients.calculated using the. data collected by the researchers themselves.turned out to show higher correlation coefficients than the data from the government sources. This' indicates that quality of the data is of prime importance in the study of market integration. The fact that the newer data tend to give higher correlations than the secondary data from government sources is an indi- cation that future efforts will require more carefully collected data before one can reaffirm some of the earlier conclusions made regarding the efficiency with which the staple food marketing systems operate. Data problems in earlier studies has led to criticisms of such studies in terms of the relevance of the data used as related to the 159 conclusions drawn (Harriss, 1979). One other objection expressed by Harriss is that in carrying out market integration analysis one should not use absolute prices but a number of other alternatives, including residuals after trend and residuals of polynomials which minimize resid- ual elements. Here a method of correlating first differences of the monthly average prices is tried in order to test the stability_of the high correlation coefficients obtained based on the use of absolute prices. This reflects the method used by Southworth.15 The results alongside the absolute price methods are presented in Table 7.6. The results show a big drop in the levels of the correlation coefficients indicating a lower level of interdistrict integration than found when using the absolute prices. The reduction is, however, not similar among crops or among the districts. While there are still coeffiCients in the .70 and .80 levels for sorghum, there are none at that level for millet or maize. All correlation coefficients using the price difference method were less than .90. For sorghum farfara the highest correlation coefficient was .88, between Malumfashi and Bakori districts. For sorghum kaura it was .83, also connecting Malumfashi and Bakori. FOr millet the highest cor~ relation coefficient was .68, between Bakori and the adjacent Faskari district. Maize has the lowest coefficients—-the highest being only .51 between Malumfashi and Kankara districts. The results of price difference correlation analysis, although still supporting the efficiency with which the sorghum marketing system 15Southworth, op. cit., p. 135. 160 .sau=a<\aa9 moves >0 mops; >u , moves >9 moves uupgumvo poppvz m~.m: asses Esnmgom mgmmsmm Essmcom Ago» cos ms_m: cw “upgum_s .goxsm saga mmmp ucsoe