Invited paper presented at the 6th African

Conference of Agricultural Economists, September 23-26, 2019, Abuja, Nigeria

Copyright 2019 by [authors]. All rights reserved. Readers may make verbatim copies of this document for non-commercial purposes by any means, provided that this copyright notice appears on all such copies.

Drivers of cattle demand in

Thierry A. Kinkpé* Laboratoire d’Analyse et de Recherche sur les Dynamiques Economiques et Sociales (LARDES), Université de , BP 123 Parakou, Bénin. Email: [email protected]

Cokou P. Kpadé Université Nationale d’Agriculture, BP43 Kétou, Bénin. Email: [email protected]

Rodrigue V. C. Diogo Département des Sciences et Techniques de Productions Animale et Halieutique, Faculté d’Agronomie, Université de Parakou, BP 123 Parakou, Bénin. Email: [email protected]

Jacob A. Yabi Laboratoire d’Analyse et de Recherche sur les Dynamiques Economiques et Sociales (LARDES), Université de Parakou, BP 123 Parakou, Bénin. Email: [email protected]

Luc H. Dossa Ecole des Sciences et Techniques de Production Animale, Faculté des Sciences Agronomiques, Université d’-Calavi, 03 BP 2819 Jéricho, Bénin. Email: [email protected]

*Corresponding author

Acknowledgements The authors would like to thank the responsible of Benin cattle markets and the stakeholders on these markets for their contributions to the success of the data collection. They would like to be also grateful to the enumerators helping them in the data collection. We are also grateful to the German Foundation Volkswagen which supported the data collection.

Abstract This paper analyses cattle demand in Benin by including the characteristics of buyers and the attributes of cattle. Data were collected based on revealed preference approach on 347 real cattle transactions on 14 real cattle markets from 68 cattle buyers. Cattle demand functions, respectively the yearly number of cattle and the yearly total cattle live weight, were estimated using Lancaster theoretical framework. Results showed that buyer requirements about cattle attributes depend on the purchase objective. The Borgou breed was the most demanded. The cattle attributes determining its demand in Benin were the body condition, age, apparent health status and the breed crossed with the purchase objective. The cattle price, the small ruminant price and the buyer income determined also significantly the cattle demand in Benin. The price- elasticities show that cattle was a highly elastic whereas the income-elasticities indicate that cattle was a necessity good in Benin. Cross-price elasticities indicate that cattle and small ruminants were complements. These results suggest that the farming of both cattle and small ruminants should be increased in Benin. The stock farmers could also target a specific segment of the market and orientate their farming to meet the requirements of the buyers. Keywords: attributes, market share, elasticity, cattle, West-Africa

1. Introduction Stock farming provides proteins for equilibrate human diet (Elmadfa & Meyer 2017; FAO 2016a). The norms suggest that the need per capita is about 7.3 kg of protein per year (FAO 2016a). For that great quantity of protein, one can consume 33 kg of meat or 45 kg of fish or 60 kg of egg or 230 kg of milk (FAO 2016a). In some countries, people consume meat largely over the norm. The five most important countries wherein people consume meat the most were the United States of America with an average of 95 kg/capita/year; Australia with 92 kg/capita/year in average; Argentina with 86 kg/capita/year in average, Israel with an average of 81 kg/capita/year and Uruguay with 80 kg/capita/year in average (OECD 2017). There are several other countries in the world wherein the meat consumption is largely over the norm, however, the average meat consumption in the Sub-Saharan African (SSA) was about 9 kg/capita/year (OECD 2017). This quantity was largely lower than the norm. So there was a great imbalance between the meat supply and the need in SSA. In Benin even the low supply of meat was largely dominated by imported meat (Tougan et al. 2013). In 2013 for instance, Benin meat imports was estimated at more than 117 thousand tons (FAOSTAT 2017). Then, Benin was not yet self-sufficient in terms of quantity and quality of meat. However, despite these huge meat imports, domestic stock farming contributes for about 3% to Benin Gross Domestic Product- GDP (Adégbola et al. 2013a). It is then evident that stock farming has a great potential to contribute to Benin development. Furthermore, there is a great potential market to conquer. The stock farming improvement will allow better contribution to economy. However, before suggesting the increasing of the animals’ raising for more meat production, it would be better to elicit the characteristics of the demand for stock farming products in Benin. Cattle are one of the most important proteins provider all over the word in general and in Africa in particular (FAO 2017). Due to this great importance of the cattle, several researches have been carried out on its demand in SSA (Jabbar & Diedhiou 2003; Kassie et al. 2009; Kassie et al. 2011; Fadiga 2013). Jabbar & Diedhiou (2003) reported that the cattle buyers in Nigeria preferred specific breeds for specific purpose. They then demonstrated that the cattle demand is determined by the purchase objective. According to Kassie et al. (2009), the fertility, the resistance to disease, calf vigour and milk production capacity are very important in cattle farmers’ preference for cows in Ethiopia. Moreover, Kassie et al. (2011) demonstrated that season, market location, class of cattle, body size and age are important determinants of its price in Ethiopia. The body condition, the origin, and the category of cattle, highly determine buyer preference in Mali (Fadiga 2013). All these researches contribute to clarify the cattle buyer behaviour about cattle in SSA. However, none did analyse the cattle demand as recommended by Lancaster (1966) consumer theory according to which the utility depends on the product attributes. In Benin, cattle is one of the most important stock farming product. It contributes to the GDP for about 1% (Adégbola et al. 2013b). Despite this great importance of the cattle, there is no deep research for its demand analysis, so that suitable breeding policy can be set up. This paper aims at contributing to fill the gap concerning Benin and the one concerning scientific researches on cattle demand in SSA. This study analyses the characteristics and socioeconomic factors of the cattle demand in Benin. 2. Methods 2.1.Study area The whole Republic of Benin was the area covered by this research. The Republic of Benin lies between latitudes 6° and 13°N, and longitudes 0° and 4°E. Benin population was estimated at 10 million (INSAE 2013) for a cattle population of about 2.2 million (FAO 2016b). The cattle production in the country is based on extensive systems (Alkoiret et al. 2009). Benin Republic can be divided in four regions based on the cattle production capacity: the North-East (Borgou and Alibori counties) was the most important cattle production zone; the North-West (Atacora and Donga counties) was the second most important cattle production zone; the South (Ouémé, Plateau, Atlantique, , Mono and Couffo counties) and the Centre (Collines and Zou counties) were the less important cattle production zones (FAO 2016b). To have a market sample representative of the whole country, two criteria were used to select 14 cattle markets: their geographic position and the cattle production capacity of the region wherein there are located. They are distributed in 9 out of the 12 counties. The North- East and the North-West were represented by 8 cattle markets, i.e. 2 per county. The Centre was represented by 3 cattle markets, 2 in Collines and 1 in Zou, and the South was represented by 3 cattle markets in 3 different counties.

2.2.Lancaster’s theoretical model According to the neoclassic theory, the analysis of the consumer behaviour aims at explaining the level of goods’ demand that someone consumes in order to maximize his utility under the constraints of the prices, his real income and a set of individual characteristics such as the age, the education, the professional status, the geographic environment (Sadoulet & de Janvry 1995). However, Lancaster (1966) argued that the attributes, i.e. the intrinsic characteristics of the goods were omitted in this neoclassic theoretical approach. The attributes of goods determine the reaction of consumers before them. Thus, Lancaster (1966) argued that, the utility is provided to the consumer by the attributes of the goods (taste, colour, aroma, etc.). Based on that, Lancaster (1966) suggested to incorporate the attributes of goods to explain consumer behaviour. As this paper is analysing the demand for cattle in Benin, the Lancaster approach is used. This approach hypothesized that the utility ( ) depends on both the product attributes and consumer characteristics as follow: = ( , ) 𝑈𝑈 (1) where, is the set of the good attributes and is the set of consumer characteristics. The𝑈𝑈 maximization𝑓𝑓 𝑥𝑥 𝑧𝑧 of this function taking into account the budget constraints conducts to a demand𝑥𝑥 function ( ) as follow: 𝑧𝑧 = ( , , , ) (2) where represents𝐷𝐷 a set of prices including the price of the product and the price of its potential substitutes𝐷𝐷 𝑓𝑓 𝑝𝑝 𝑥𝑥; 𝑦𝑦 is𝑧𝑧 the set of the attributes of the good; is the consumer income and represents 𝑝𝑝 𝑥𝑥 𝑦𝑦 𝑧𝑧 the other consumer’s characteristics. The cattle demand in Benin is modelled following the equation (2). 2.3.Data collection Data were collected on real cattle transactions made on real markets in Benin. In each market, the random sampling method was used. In fact, enumerators stood at the entry gate of the market allowing them to meet the majority of those coming in the market. Enumerators approached each of them separately and asked if they came for cattle purchase. To those who replied positively, they asked to participate to the survey after briefly explaining the objective to them. Data were collected on all cattle transactions done by the consenting cattle buyers on the market. The cattle buyer expresses the demand of the cattle on a market. The cattle buyer is considered as an individual with a given cattle purchase objective (cattle farming, domestic trading, export or immediate slaughtering for meat production). Table 1 reports the distribution of the cattle buyers’ sample and the purchased cattle. The characteristics of all purchased cattle were collected in order to assess the demand and the market share of the different cattle attributes. These characteristics were about the mensuration to estimate the weight1. The Body Condition Score (BCS)2, the breed, the approximate age using teeth and horn method, the coat colour and the origin were collected. The price of the purchased cattle was collected as well as the price of other animals with red meat (small ruminants, pig). The socioeconomic characteristic of buyers (income, age, education level and the location of residence) were also collected during the field experiment. Afterward, each selected cattle buyer was asked to give the attributes of the cattle he used to buy and indicate the cattle among those he bought which size was the most close to the average cattle he used to buy (for the mensuration in order to assess the average weight). Cattle buyer was also asked to give the average frequency of the purchase and the average number purchased per frequency in order to estimate his yearly demand. Table 1: Sample distribution and size Region County District Cattle markets Number of buyers Number of purchased cattle North- Alibori Gogounou Petit Paris 4 30 East Guéné 3 30 Borgou Kalalé Dérassi 4 30 Parakou Guèma 11 29 North- Atacora Matéri Matéri 4 30 West Péhunco Péhunco 4 22 Donga Aoro 6 30 Kolokondé 3 25 Centre Collines Ouèssè Ouèssè 7 20 Doumè 6 9 Zou Avokanzoun 4 30 South Littoral Cotonou Selling park 3 30 Mono Comè Comè 3 12 Plateau Kétou Iwoyé 6 20 Total-Benin 68 347 Source: Authors’ tabulation

1 Using the linear body measurements such as chest girth (cm), scapula-ischial length (cm) and height at the withers (cm), the cattle bodyweight (W) was estimated. The formula of Vanvanhossou (2017) was used for taurine subspecies. The formula of Dodo et al. (2001) was used for zebu subspecies while the formulas of Youssao et al. (2013) were used for the crossbreed (Borgou breed). 2 The BCS as suggested by Vall and Bayal (2004) varies from 0 to 5 according to the aspect of the cattle. The score 0 corresponds to the cachectic animal, the score 1 corresponds to emaciated animal, the score 2 corresponds to the animal with a general appearance lean enough, the score 3 corresponds to the animal with a general good appearance, the score 4 corresponds to the animal with a general well covered appearance and the score 5 corresponds to the animal with a general fat and smooth appearance.

2.4.Modelling the cattle demand in Benin Using Lancaster theoretical model, the cattle demand in Benin can be mathematized as follow: = , , , (3) where𝑖𝑖𝑖𝑖 is 𝑗𝑗the 𝑖𝑖quantity𝑖𝑖 of cattle demanded by the cattle buyer ; represents a set of prices including𝑞𝑞 𝑓𝑓�𝑝𝑝 the𝑥𝑥 price𝑦𝑦 𝑧𝑧 �of cattle and the price of its potential substitutes which can be animals 𝑖𝑖𝑖𝑖 providers𝑞𝑞 of other red meats (sheep𝑗𝑗 and goat and pig); is the set𝑖𝑖 of𝑝𝑝 the attributes of the cattle demanded by the buyer ; is the income of the cattle buyer ; represents the other 𝑗𝑗 characteristics of the cattle buyer and is the symbol𝑥𝑥 of function. 𝑖𝑖 𝑖𝑖 In𝑗𝑗 Benin, the pork constitutes𝑖𝑖 𝑦𝑦a forbidden meat for several people especially𝑖𝑖 𝑧𝑧 those practicing the Islam, the Celestial Christianism𝑖𝑖 and𝑓𝑓 several traditional religions. Then, considering pig as substitute of cattle can bias the results. Thus, one can assume that for the cattle buyers, the small ruminants (sheep and goat) were the main substitute of cattle. Based on the previews development, in the cattle demand function in equation (3) represents a set of prices including the price of cattle and the price of its potential substitutes which are small ruminants (sheep and goat). 𝑝𝑝 Some buyers’ socioeconomic characteristics can interact with the good’s characteristics. Based on that, the expression of the equation (3) can be: = + + + + + + + (4) where𝑖𝑖𝑖𝑖 𝑖𝑖𝑖𝑖 is the𝑖𝑖𝑖𝑖 minimal𝑖𝑖𝑖𝑖 𝑖𝑖𝑖𝑖 yearly𝑖𝑖𝑖𝑖 quantity𝑖𝑖 𝑖𝑖 𝑖𝑖𝑖𝑖 of𝑗𝑗 cattle𝑖𝑖 𝑖𝑖 necessary𝑖𝑖𝑖𝑖 𝑗𝑗 𝑖𝑖 for the𝑖𝑖𝑖𝑖 need of the buyer ; is the𝑞𝑞 average𝑐𝑐 price𝛼𝛼 𝑃𝑃 which𝛾𝛾 𝑝𝑝the buyer𝛽𝛽 𝑦𝑦 purchases𝛿𝛿 𝑥𝑥 𝜆𝜆 the𝑍𝑍 cattle𝜏𝜏 𝑥𝑥 ;𝑍𝑍 is𝜖𝜖 the average price which the 𝑖𝑖𝑖𝑖 𝑖𝑖𝑖𝑖 buyer 𝑐𝑐 purchases the small ruminants ; is the𝑗𝑗 yearly income of the buyer ; is the𝑖𝑖 set𝑃𝑃 of 𝑖𝑖𝑖𝑖 attributes of the cattle ; is the set𝑖𝑖 of characteristics (except𝑗𝑗 𝑃𝑃 the income) of the buyer ; 𝑖𝑖 𝑘𝑘 𝑦𝑦𝑖𝑖 𝑖𝑖 𝑥𝑥𝑗𝑗 is the set of interactions between𝑖𝑖 some attributes of the cattle and some characteristics𝑗𝑗 𝑖𝑖 of the buyer ; α, γ, β, δ𝑗𝑗, 𝑧𝑧 and are the parameters to be estimated and is the error terms.𝑖𝑖 𝑥𝑥 𝑧𝑧 𝑥𝑥𝑗𝑗 𝑗𝑗 𝑧𝑧𝑖𝑖 is a positive value of narrow distribution. In this case, there could be 𝑖𝑖𝑖𝑖some econometrical insufficiencies𝑖𝑖 in the estimation𝜆𝜆 𝜏𝜏 (Greene 2008). To increase the robustness𝜖𝜖 of the estimations, 𝑖𝑖𝑖𝑖 𝑞𝑞logarithmic transformation was used to widen this distribution; so that, the coefficients of the metric variables will be elasticities (Greene 2008). Then, the equation (4) becomes: ( ) = + ( ) + ( ) + ( ) + ( ) + + ( ) + + + 𝑚𝑚 𝑚𝑚 𝑑𝑑 𝑑𝑑 𝑚𝑚 𝑚𝑚 (5) 𝑛𝑛𝑛𝑛 𝑞𝑞𝑖𝑖𝑖𝑖 𝐶𝐶𝑖𝑖𝑖𝑖 𝛼𝛼𝑖𝑖𝑖𝑖𝑛𝑛𝑛𝑛 𝑃𝑃𝑖𝑖𝑖𝑖 𝛾𝛾𝑖𝑖𝑖𝑖𝑛𝑛𝑛𝑛 𝑝𝑝𝑖𝑖𝑖𝑖 𝛽𝛽𝑖𝑖𝑛𝑛𝑛𝑛 𝑦𝑦𝑖𝑖 𝛿𝛿𝑖𝑖𝑖𝑖 𝑛𝑛𝑛𝑛 𝑥𝑥𝑗𝑗 𝛿𝛿𝑖𝑖𝑖𝑖𝑥𝑥𝑗𝑗 𝜆𝜆𝑖𝑖 𝑛𝑛𝑛𝑛 𝑍𝑍𝑖𝑖 where represents𝑑𝑑 Napierian𝑑𝑑 Logarithm; is the of ; is the set of metric attributes 𝜆𝜆𝑖𝑖 𝑍𝑍𝑖𝑖 𝜏𝜏𝑖𝑖𝑖𝑖𝑥𝑥𝑗𝑗𝑍𝑍𝑖𝑖 𝜀𝜀𝑖𝑖𝑖𝑖 of the cattle ; is the set of dummy attributes of the cattle𝑚𝑚 ; is the set of metric 𝑛𝑛𝑛𝑛 𝐶𝐶𝑖𝑖𝑖𝑖 𝑛𝑛𝑛𝑛 𝑐𝑐𝑖𝑖𝑖𝑖 𝑥𝑥𝑗𝑗 characteristics of the𝑑𝑑 buyer ; is the set of dummy characteristics of𝑚𝑚 the buyer ; and is 𝑗𝑗 𝑥𝑥𝑗𝑗 𝑗𝑗 𝑧𝑧𝑗𝑗 the error terms. is the price 𝑑𝑑elasticity of the cattle demand; is the cross-price elasticity 𝑖𝑖 𝑧𝑧𝑗𝑗 𝑖𝑖 𝜀𝜀𝑖𝑖𝑖𝑖 of the cattle demand𝑖𝑖𝑖𝑖 relative to small ruminants; is the income𝑖𝑖𝑖𝑖 elasticity of cattle demand; and are 𝛼𝛼respectively the coefficients of metric cattle𝛾𝛾 attributes and dummy cattle 𝑖𝑖 𝑚𝑚 𝑑𝑑 𝛽𝛽 attributes;𝑖𝑖𝑖𝑖 𝑖𝑖𝑖𝑖 and are respectively the coefficients of metric buyer characteristics and dummy𝛿𝛿 buyer𝛿𝛿 𝑚𝑚 characteristics𝑑𝑑 and are the coefficients of the interactions between the cattle 𝑖𝑖 𝑖𝑖 attributes and𝜆𝜆 the buyer𝜆𝜆 characteristics. 𝑖𝑖𝑖𝑖 The cattle demand could be expressed𝜏𝜏 in number of cattle as well as total live weight. Then, two models were specified. The first considered the cattle demand in terms of number of yearly cattle as dependent variable and the second considered the cattle demand in terms of yearly total live weight of cattle. The final list of explanatory variables was retained after intercorrelation tests. The Table 2 shows the details and the expected signs of the explanatory variables included in the equation (5). This equation was estimated through a robust method correcting eventual heteroscedasticity problems (Greene 2008) using the STATA 13 software.

Table 2: Explanatory variables of the cattle demand functions in Benin Variables Expected signs Prices Price of the cattle (FCFA/kg of live weight) a - Price of the small ruminants (FCFA/kg of live weight) a + Attributes of the cattle Body Condition𝑚𝑚 Score, BCS (0-5) a + 𝑗𝑗 Approximate𝑥𝑥 age of the cattle (number of months) a ± Attributes of the cattle 𝑑𝑑 Male cattle 𝑗𝑗(1=Yes, 0=No) ± Nulliparous𝑥𝑥 female cattle (1=Yes,0 =No) ± Coat in the majority white (1=Yes, 0=No) ± Cattle apparently with good health status (1=Yes, 0=No) + Cattle raised in Benin (1=Yes, 0=No) ± Interaction of the cattle attributes and the purchase objective b Cattle of Borgou breed– purchased for farming (1=Yes,0 =No) ± Cattle of Borgou𝑥𝑥𝑗𝑗𝑍𝑍𝑖𝑖 breed– purchased for immediate slaughtering (1=Yes, 0=No) ± Cattle of Zebu Fulani breed– purchased for immediate slaughtering (1=Yes, 0=No) ± Cattle of Borgou breed– purchased for domestic trade (1=Yes, 0=No) ± Cattle of Goudali breed– purchased for domestic trade (1=Yes, 0=No) ± Cattle of White Fulani breed– purchased for domestic trade (1=Yes, 0=No) ± Cattle of Borgou breed– purchased for exportation (1=Yes, 0=No) ± Cattle of Goudali breed– purchased for exportation (1=Yes, 0=No) ± Cattle of Zebu Fulani breed– purchased for exportation (1=Yes, 0=No) ± The buyer income Annual income of the buyer (FCFA) a + 𝑖𝑖 The buyer characteristics𝑦𝑦 a Age of the buyer (number of𝑚𝑚 years) ± 𝑖𝑖 The buyer characteristics 𝑧𝑧 The buyer lives in urban area𝑑𝑑 (1=Yes, =No) ± The buyer is educated (1=Yes,𝑧𝑧𝑖𝑖 =No) ± a Variable expressed in logarithmic form ; b All the interactions were not included because of some multicollinearity problems and/or weak frequency Source: Authors’ tabulation

2.5. Characteristics of cattle buyers in Benin The Table 3 presents the characteristics of the cattle buyers surveyed. All the 68 studied cattle buyers were male. Among them, 38% were from urban area while 32% were educated. The butchers were the most represented in urban areas and the most educated in terms of proportion. The cattle exporters had the most important income (more than FCFA3 15 million/year in average) and were the oldest (more than 43 years old in average) while the cattle farmers had the less important income (less than FCFA 5.5 million/year in average) and were the youngest (about 31 years old in average). The average income of the cattle buyers in Benin was almost FCFA 9.6 million/year and they were almost 40 years old in average.

Table 3: Socioeconomic characteristics of the cattle buyers Buyer characteristics Purchase objectives Immediate Export toward Stock farming slaughtering Domestic trade Pooled West-Africa (meat) Urban buyer (%) 40.00 50.00 34.38 35.29 38.24 Educated buyer (%) 40.00 50.00 28.13 23.53 32.35 Yearly income (FCFA million)a 5.46 (3.67) 8.25 (12.08) 7.71 (6.17) 15.12 (9.85) 9.57 (9.04) Age (number of year)a 31.20 (6.26) 37.43 (6.25) 38.94 (8.56) 43.29 (8.99) 39.15 (8.56) Number of buyers 5 14 32 17 68 a : mean (standard deviation) Source: Authors’ tabulation

3. Results and discussion 3.1.Characteristics of cattle demand in Benin The average BCS, age and weight of the cattle demand are reported in Table 4. The average BCS of the cattle purchased in Benin was relatively high while the weight was about 260 kg/cattle. The age of these cattle was more than 50 months showing that the cattle was kept by the farmers for relatively long time before selling. The exported cattle were in average the oldest, the second heaviest and with the highest BCS. This result shows that the best cattle produced in Benin were exported toward the West-African countries, especially Nigeria. The butchers bought the heaviest cattle with an average good stoutness (second highest BCS) which would provide them the good quality and heavy meat to boost their profit. The cattle farmers bought the youngest and the cattle with the lowest weight. In terms of market share, the male cattle were the most demanded no matter the purchase objective (Figure 1). Globally, the second most demanded cattle category was the cows (about 11%). These results show that the cattle farmers do not like selling the young nulliparous females. They keep them to ensure a good production level of their herds. Moreover, they use to buy nulliparous females (almost 35% of the share of their demand) for the same reason. Figure 2 reports the market share of the cattle breeds in Benin following the purchase objective. Globally, the Borgou breed was the most demanded (almost 50%). This breed constitute 80% of the cattle farmers’ demand, certainly due to its relative rusticity. The White Fulani and Zebu Fulani breeds were globally the second and third most demanded cattle breed. The white colours were the most demanded (Figure 3). The cattle buyers in Benin demanded apparently healthy cattle, however they also bought cattle with minor sicknesses, especially farmers and exporters (Figure 4) if the others attributes highly attract them. Cattle produced in Benin satisfied more than 95% of the cattle market in Benin (Figure 5). Just some exporters demanded cattle from the sub-region (Figure 5).

Table 4: Body condition, age and weight of demanded cattle following the purchase objectives in Benin Attributes Purchase objectivesa Immediate Export toward Stock farming slaughtering Domestic trade Pooled West-Africa (meat) Body Condition 3.46 (0.67) 3.85 (0.55) 3.57 (0.56) 3.95 (0.68) 3.73 (0.64) Score (0-5) Approximate age 39.11 (27.82) 58.46 (18.87) 43.63 (30.36) 68.91 (31.28) 53.79 (31.44) (number of months) Live weight (kg/ 181.30 (91.94) 349.56 (152.80) 203.58 (85.18) 317.46 (131.79) 260.58 (130.97) cattle) Number of 35 48 145 119 347 observations a : mean (standard deviation) Source: Authors’ tabulation

100

80

60

Market share share (%) Market 40

20

0 Stock farming Immediate Domestic trade Export toward Pooled slaughtering West-Africa (meat) Purchase objective Male Young (nulliparous) female Cow

Figure 1: Market share of the cattle categories following the purchase objective in Benin Source: Authors’ results

100

80

60

40 Market share share (%) Market

20

0 Stock farming Immediate Domestic trade Export toward Pooled slaughtering West-Africa (meat) Purchase objective

Somba Ndama Borgou Goudali Zebu-Fulani White Fulani Red Fulani

Figure 2: Market share of the cattle breeds following the purchase objective in Benin Source: Authors’ results 100

80

60

40 Market share share (%) Market 20

0 Stock farming Immediate Domestic trade Export toward Pooled slaughtering West-Africa (meat) Purchase objective Homogeneous white Spoted white Homogeneous fawn Spoted fawn Homogeneous brown Spoted brown Homogeneous black Spoted black

Figure 3: Market share of the cattle coat colour following the purchase objective in Benin Source: Authors’ results

100

80

60 Market share share (%) Market

40

20

0 Stock farming Immédiate Domestic trade Export toward Pooled slaughtering West-Africa (meat) Purchase objective

Apparently sick Apparently healthy

Figure 4: Market share of the cattle apparent health status following the purchase objective in Benin Source: Authors’ results 100

80

60 Market share share (%) Market 40

20

0 Stock farming Immediate Domestic trade Export toward Pooled slaughtering West-Africa (meat) Purchase objective West African sub-region Benin

Figure 5: Market share of the cattle origin following the purchase objective in Benin Source: Authors’ results

3.2.Cattle and small ruminants price in Benin The Table 5 shows the price of the live kg of cattle and small ruminants such as goat and sheep on Benin markets according to the purchase objective. The kg of the cattle demanded cost almost FCFA 7703 in average in Benin while the kg of alive small ruminants cost almost FCFA 950 showing that, in Benin, the kg of alive small ruminants is more expensive than the kg of cattle. Whether it was the cattle or the small ruminants, the price was the highest for the butchers and lowest for the domestic traders and stock farmers. These results are normal based on the animal value channel (Tamini et al. 2014) which showed that the stock farmers were the upstream stakeholders while the butchers were the downstream stakeholders. Hence the butcher supported the highest price and the domestic traders and the breeders supported the lowest price. Table 5: Cattle and small ruminants price following the purchase objectives in Benin Parameters Purchase objectivesa Immediate Export toward Stock farming slaughtering Domestic trade Pooled West-Africa (meat) Cattle price (FCFA/ live 744.12 (199.36) 816.12 (193.14) 730.81 (475.99) 777.63 (251.57) 768.31 (356.57) kg) Small ruminants price 840.00 (155.20) 1432.29 (388.45) 826.78 (157.51) 932.77 (231.91) 948.22 (302.47) (FCFA/ live kg) Number of 35 48 145 119 347 observations a : mean (standard deviation) Source: Authors’ tabulation

3 FCFA 1000 varied from US$ 1.61 to 1.66 for an average of US$ 1.63 during the data collection period. 3.3.Yearly cattle demand in Benin The Table 6 reports the yearly cattle demand per buyer according to the geographic location. The individual buyer demand of cattle was in average about 688 cattle per year, equivalent to more than 170 of tons live weight. In terms of number of cattle, the buyer in North-West were the most important demanders while those in South were the less important demanders. However, considering the total live weight of the demanded cattle, the buyer in North-East were the most important demanders due to the relatively high weight of the cattle in this region. Furthermore, the North-West Benin was the predilection zone of the Somba breed (Dossa & Vanvanhossou 2016) which was a breed of relative low weight (Rege 1999). The exporters and the domestic traders were the most important cattle demanders either in terms of number of cattle or in terms of total live weight (Table 7). Each celebration organiser demanded about one cattle, corresponding to about 150 live kg per year. Each stock farmer demanded in average 301 cattle per year either for fattening or reinforcement of the reproduction in their herd (Table 7). Each butcher demanded in average more than 345 cattle, corresponding to more than 105 tons of live weight per year (Table 7). So they demanded almost one cattle of more than 300 live kg per day. Table 6: Cattle yearly demand per buyer following the regions in Benin Region Number of Total live weigh of cattle per buyer observations Number of cattle per buyer (kg) Standard- Standard- Mean deviation Mean deviation North-East 22 762.55 508.33 219235.42 208236.61 North-West 17 920.71 496.99 153007.25 93814.88 Centre 17 623.53 520.01 175191.32 195106.83 South 12 316.54 376.86 98161.02 136619.28 Total 68 688.63 518.85 170301.28 172186.70 Source: Authors’ tabulation

Table 7: Cattle yearly demand per buyer following the purchase objectives in Benin Purchase objective Number of Number of cattle per Total live weigh of cattle per observations buyer buyer (kg) Standard- Standard- Mean deviation Mean deviation Stock farming 5 301.40 299.23 40586.86 39923.44 Immediate slaughtering 14 271.39 290.48 83018.94 113528.15 Celebration 3 1.17 0.76 152.71 112.28 Butcher’ shop 11 345.09 286.02 105618.81 118883.80 Trade 49 847.35 503.33 208475.26 179717.15 Domestic trade 32 690.75 445.02 141194.99 104408.21 Exportation toward 17 1142.12 484.71 335120.47 222888.35 West-Africa Source: Authors’ tabulation

3.4.Factors affecting cattle demand in Benin The factors determining the cattle demand in Benin are reported in the Table 8. Both the model of yearly number of cattle demand and the model of yearly total live weight of cattle demand per buyer were globally significant at the threshold of 1% showing that the models fit the variables included. Moreover, the variation coefficients R2 were respectively of 73% and 77%. So, about 73% of variations in the yearly number of cattle demand per buyer in Benin were explained by variations of the explanatory variables. Likewise, about 77% of variations of the yearly total weight of cattle demand per buyer were explained by variations of the same explanatory variables. The significant determinants of the cattle demand in Benin were the cattle price, the price of the small ruminants (sheep and goat), the cattle BCS, age, apparent health status, breed crossed with the buyer purchase objective and the yearly income of the buyer (Table 8). The price- elasticities of the cattle demand in Benin were negative and less than -1 for both models indicating that the cattle demand was highly elastic in Benin. In other words, when the cattle price increases for 1%, its demand decreases for more than 1%, Ceteris paribus. Hence, a cattle price increasing will induce a decreasing of the turnover of the stock farmers, domestic traders, exporters and butchers whose activity is linked to cattle. The cross-price elasticity of the cattle demand relative to small ruminants were negative showing that when the small ruminants’ price increases, the cattle demand decreases, Ceteris paribus. This shows that the cattle and the small ruminants were not substitutes in Benin, they were complements. This result is inconsistent with Ehui et al. (2000) who showed that sheep and cattle were independent in Ethiopia. In other words, the increasing of the cattle price did not have significant effect on the sheep demand. This difference could be explained by the fact that Ehui et al. (2000) estimated the cross-price elasticity of the sheep demand relative to cattle while in the present paper the cross-price elasticity of the cattle demand relative to small ruminants was estimated. Moreover, in African societies, cattle and small ruminants play different socio-anthropological and cultic roles. In fact, socio-anthropologically, cattle and small ruminants complete each other. During the celebrations such as weddings and anniversaries or funerals, those who have financial capacities would prefer buying cattle for their prestige. Those with limited capacities would buy small ruminants. Likewise, for example during religious celebrations such as Tabaski, some people with financial capacities purchase cattle to complement their sheep and share the meat with their families and neighbours. It provides prestige to them. In traditional religions, each divinity has its requirement in terms of animal for sacrifice. Hence cattle and small ruminants complete each other for the equilibrium in the society. The presence of Borgou breed on the market increases the demand of the cattle for farming. As the Borgou breed is a crossbreed (Zebu*Taurine), it is adapted for the climatic conditions of all regions in Benin (Youssao et al. 2013). Borgou breed had a good resistance to diseases such as trypanosomiases (characteristics from Taurine subspecies). It had also a relative good reproduction and acceptable milk production capacity (characteristics from Zebu subspecies). So with this breed, the stock farmers would face low challenge in terms of diseases. They would also have relative good productivity of their herd. These characteristics justify the high preference of the cattle farmers for the Borgou. This result is consistent with Ouma et al. (2007), Kassie et al. (2009), Siddo et al. (2015) and Woldu et al. (2016) who demonstrated that the resistance to disease, the reproduction and the milk production capacity of the animal breeds determine the farmers’ preference and demand for them. The presence of Borgou and Zebu Fulani breeds increase the cattle demand for slaughtering. This justifies the good meat yield of the Borgou (Youssao et al. 2013) and Zebu Fulani breeds. The Borgou and Goudali breeds increases the cattle demand for domestic trade and export in Benin. The White Fulani breed influences positively the cattle demand for domestic trade. The cattle apparent good health status increases its demand in Benin. When the BCS of the cattle increases for 1%, their demand in number increases for more than 2.3% while their demand in terms of total live weight increases for about 2.7%. These results show that the cattle buyers in Benin needed cattle with good stoutness level. The age of the cattle did not affect significantly its demand in number, however it affected positively its demand in total live weight. Generally, the cattle weight was positively correlated with its age (Youssao et al. 2013), and that justifies the relationship between the age and the demand in total live weight. The income-elasticity of the cattle demand was positive and less than one for the demand expressed in number of cattle and for the demand expressed in live weight. Accordingly, the cattle was a necessity good in Benin. This result was consistent with Ehui et al. (2000) who showed that the sheep was a necessity good in Ethiopia. Table 8: Factors affecting cattle demand (per buyer) in Benin Variables Number of cattle a b Total live weight of cattle (kg) a b Price of the cattle (FCFA/kg of live weight)a -1.58 (0.87)* -2.51 (0.88)*** Price of the small ruminants (FCFA/kg of live weight) a -2.67 (0.85)*** -2.33 (0.88)** Body Condition Score, BCS (0-5) a 2.33 (1.13)** 2.70 (1.10)** Approximative age of the cattle (number of months) a 0.27 (0.26) 0.83 (0.30)*** Male cattle (1=Yes, =No) -0.22 (0.62) 0.01 (0.64) Nulliparous female cattle (1=Yes, =No) -0.28 (0.63) -0.07 (0.68) Coat in the majority white (1=Yes, =No) -0.90 (0.58) -0.77 (0.59) Cattle apparently with good health status (1=Yes, =No) 2.19 (0.68)*** 1.69 (0.72)** Cattle bred in Benin (1=Yes, =No) -0.88 (0.95) -0.98 (0.93) Cattle of Borgou breed– purchased for breeding (1=Yes, =No) 1.91 (0.74)** 1.86 (0.85)** Cattle of Borgou breed– purchased for immediate slaughtering 1.36 (0.56)** 1.44 (0.59)** (1=Yes, =No) Cattle of Zebu Fulani breed– purchased for immediate 1.67 (0.98)* 2.16 (1.06)** slaughtering (1=Yes, =No) Cattle of Borgou breed– purchased for domestic trade (1=Yes, 2.36 (0.52)*** 2.21 (0.53)*** =No) Cattle of Goudali breed– purchased for domestic trade 1.44 (0.66)** 1.55 (0.67)** (1=Yes, =No) Cattle of White Fulani breed– purchased for domestic trade 1.22 (0.70)* 1.39 (0.73)* (1=Yes, =No) Cattle of Borgou breed– purchased for exportation (1=Yes, 3.03 (0.62)*** 2.97 (0.63)*** =No) Cattle of Goudali breed– purchased for exportation (1=Yes, 1.61 (0.58)*** 1.56 (0.65)** =No) Cattle of Zebu Fulani breed– purchased for exportation 0.90 (0.84) 0.98 (0.89) (1=Yes, =No) Annual income of the buyer (FCFA) a 0.52 (0.29)* 0.60 (0.30)* The buyer lives in urban area (1=Yes, =No) 0.17 (0.34) 0.13 (0.35) The buyer is educated (1=Yes, =No) 0.48 (0.34) 0.41 (0.36) Age of the buyer (number of years) a 0.47 (0.79) 0.21 (0.80) Constant 18.68 (9.29)* 25.26 (9.81)** Number of observations 66 66 F of Fisher (dif1= 22 ; dif2= 43) 10.08*** 5.57*** R2 0.73 0.77 a Variable expressed in logarithmic form ; b The values between brackets are the standard errors ; *= significant at the threshold of 10% ; **= significant at the threshold of 5% ; ***= significant at the threshold of 1% Source: Authors’ tabulation

4. Conclusion This paper has analysed the factors driving cattle demand in Benin, to fill some scientific gaps regarding the cattle attributes and the buyer characteristics. The requirements of the cattle buyers depend on their purchase objectives. Borgou breed was the mostly demanded no matter the purchase objective, however they specifically constitute by far the preference of the farmers. All cattle buyers require apparently healthy cattle. They also require relatively good body condition. However, the exporters and the butchers are the most intransigent on this aspect. The cattle farmers demand relative young cattle either for fattening or to sustain the reproduction capacity of their herd. The Butchers and the exporters require the heavy cattle. Cattle is a necessity good in Benin with high price-elasticities of the demand. Cattle and small ruminants are complements on Benin markets. These results suggest that cattle farming should be improved without forgetting about the small ruminants in order to improve the consumer access to meat for the nutritional security in Benin. This could also allow keeping the cattle price almost stable in order to ensure the demand and then the stability of the turnover of the farmers, domestic traders, exporters and butchers for the sustainability of the cattle contribution to Benin economy. Moreover, cattle farmers could target a segment of the market (farmer, exporter, domestic trader and butcher) so that they can focus their cattle farming on the requirements of this segment.

References Adégbola YP, Oloukoï L, Houessionon P Kinkpé AT, Adjovi G, Bankolé AB & Foundohou A, 2013a. Comptes Economiques de l’Agriculture du Bénin, Volume 1: rapport de synthèse. Ministry of Agriculture, Breeding and Fishing (MAEP), Cotonou, Benin. Adégbola YP, Oloukoï L, Houessionon P Kinkpé AT, Adjovi G, Bankolé AB & Foundohou A, 2013b. Comptes Economiques de l’Agriculture du Bénin, Volume 2: Comptes économiques des branches de l’agriculture. Ministry of Agriculture, Breeding and Fishing (MAEP), Cotonou, Benin. Alkoiret IT, Awohouedji DYG, Akossou AYJ & Bosma RH, 2009. Typologie des systèmes d’élevage bovin de la commune de Gogounou au nord-est du Benin. ASA 12(2): 77-98. Dodo K, Pandey VS & Illiassou MS, 2001. Utilisation de la barymétrie pour l’estimation du poids chez le zébu Azawak au Niger. Rev Elev Med Vet Pays Trop 54(1): 63–68. Dossa LH & Vanvanhossou FUS, 2016. The indigenous Somba cattle of the hilly Atacora region in North-West Benin: threats and opportunities for its sustainable use. Trop Anim Health Prod 48(2): 349-59. Ehui SK, Benin S & Nega G, 2000. Factors affecting urban demand for live sheep: The case of Addis Ababa, Ethiopia. Socio-economics and Policy Research Working Paper 31. ILRI (International Livestock Research Institute), Nairobi, Kenya. Elmadfa I & Meyer AL, 2017. Animal Proteins as Important Contributors to a Healthy Human Diet. Annu Rev Anim Biosci 5: 111-131. Fadiga ML, 2013. Valuation of cattle attributes in the Malian humid and sub-humid zones and implications for a sustainable management of endemic ruminant livestock. Environ Econ 4(1): 39–50. FAO, 2017. Source of meat. http://www.fao.org/ag/againfo/themes/en/meat/backgr_sources.html (Accessed 20 November 2017) FAO, 2016a. Meat consumption. http://www.fao.org/ag/againfo/themes/fr/meat/background.html (Accessed 31 March 2016) FAO, 2016b. Statistic database. http://countrystat.org/home.aspx?c=BEN (Accessed 04 June 2016) FAOSTAT, 2017. Crop and livestock products. http://www.fao.org/faostat/en/#data/TP (Accessed 20 November 2017) Greene WH, 2008. Econometric analysis. Pearson Edition. New Jersey. INSAE, 2013. Quatrième Recensement General de la Population et de l’Habitat (RGPH4). National Institute of Statistics and Economic Analysis. Cotonou, Benin. Jabbar MA & Diedhiou ML, 2003. Does breed matter to cattle farmers and buyers? Evidence from West Africa. Ecol Econ 45: 461–472. Kassie GT, Abdulai A & Wollny C, 2011. Heteroscedastic hedonic price model for cattle in the rural markets of central Ethiopia. Appl Econ 43: 3459–3464. Kassie GT, Abdulai A & Wollny C, 2009. Valuing traits of indigenous cows in central Ethiopia. J Agric Econ 60(2): 386–401. Lancaster JK, 1966. A new approach to consumer theory. J. Political Econ 74(2): 132–157. OECD, 2017. Meat consumption. Organisation for Economic Co-operation and Development: https://data.oecd.org/agroutput/meat-consumption.htm (Accessed 20 November 2017) Ouma E, Abdulai A & Drucker A, 2007. Measuring heterogeneous preferences for cattle traits among cattle-keeping households in East-Africa. Am J Agric Econ 89(4): 1005–1019. Rege JEO, 1999. The state of African cattle genetic resources I. Classification framework and identification of threatened and extinct breeds. Anim Genet Resour 25: 1-25. Sadoulet E & de Janvry A, 1995. Quantitative development policy analysis. The Johns Hopkins University Press Baltimore and London. Siddo S, Moula N, Hamadou I, Issaa M, Marichatou H, Leroy P & Moussiaux NA, 2015. Breeding criteria and willingness to pay for improved Azawak zebu sires in Niger. AAB 58: 251–259. Tamini LD, Fadiga ML & Sorgho Z, 2014. Chaînes de valeur des petits ruminants au Burkina Faso : Analyse de situation. Nairobi: International Livestock Research Institute (ILRI). Tougan U, Dahouda M, Salifou C & Ahounou S, 2013. Variability of carcass traits of local poultry populations of Gallus gallus species of Benin. Int J Poult Sci. 12(8): 473-483. Vall E & Bayala I, 2004. Note d’état corporel des zébus soudaniens. Production animale en Afrique de l’Ouest, Slip N°12. Vanvanhossou FUS, 2017. Relation entre la morphologie et quelques caractéristiques productive des bovins Somba au Nord-Ouest du Bénin. Mémoire de master, Université d’Abomey-Calavi, Bénin. Woldu T, Markemann A, Reiber C, Kassie GT & Valle Zárate A, 2016. Combining revealed and stated preferences to define goat breeding objectives in Ethiopia. Livest Sci 191: 179–186. Youssao AKI, Salifou CFA, Séni Alassane D, Senou S, Yacoubou AM, Touré FZ & Alkoiret TI, 2013. Modélisation des performances pondérales de bovins Borgou élevés sur pâturages naturels en ferme au Bénin. LRRD 25(10). Retrieved March 13, 2018, from http://www.lrrd.org/lrrd25/10/yous25186.htm.