search
for
 About Bioline  All Journals  Testimonials  Membership  News  Donations


African Crop Science Journal
African Crop Science Society
ISSN: 1021-9730 EISSN: 2072-6589
Vol. 10, Num. 3, 2002, pp. 263-270

African Crop Science Journal, Vol. 10. No. 3,  2002, pp. 263-270

AN EMPIRICAL ANALYSIS OF VARIETY PRICE PREMIUM ATTRIBUTES, SPATIAL AND TEMPORAL PRICING PATTERNS FOR COWPEAS IN OGUN STATE, NIGERIA

C.A. Afolami

Department of Agricultural Economics and Farm Management, University of Agriculture, PMB 2240, Abeokuta, Nigeria

(Received 13 March 2002; accepted, 27 June 2002)

Code Number: cs02026

ABSRACT

The paper examined variety price premium attributes, spatial and temporal pricing patterns of cowpeas in Ogun State Nigeria. Monthly retail cowpea prices were collected over the period of January 1989 to December 1999, for peu (drum) and sokoto varieties. Four markets located in four zones were used for the study.  Choice of the markets was based on the volume of cowpea sales and their location from the state zonal headquarters. Analysis of covariance (ANCOVA) was used to process data collected. Retail price in a given month was regressed on variety, location, season and the covariate, trend. The model had a good fit, explaining 46 percent of total variation in commodity price. The estimated parameters of variety dummy, location dummies and the covariate, trend - were all significant. Six out of the estimated 11parameters of the seasonal dummies were significant while five were not, thus stratifying seasonal price into low commodity price months and high commodity price months. Location price difference was inversely related to distance from the central commodity market, and the seasonal price difference was attributed to storage technique. These show imperfect competitive market behaviour. Peu/drum with characteristics of brown colour, rough skin and large grain size had a price premium than sokoto, which is white, rough skinned and of smaller grains. Spatial and temporal price differences observed can be reduced through a better road network, improved storage techniques and adequate market information. The identified cowpea attributes of large grain, rough skin and brown colour are recommended attributes in breeding programmes.

Key Words: Commodity price, market,  retail dummy variables, storage technique

RÉSUMÉ

Cet article examine les attributs des premiers prix des variétés, les tendances spatiales et temporelles des prix du niébé dans l’Etat d’Ogun au Nigeria. Le prix menstuel du niébé en détail était collecté au delà de la période allant de Janvier 1989 et Décembre 1989 pour les variétés de  peu (drum) et de sokoto.  Quatre marchés situés dans quatre zones étaient utilisés dans cette étude. Le choix du marché était basé sur le volume du niébé vendu et leur position par rapport siège de l’Etat dans la zone.  L’analyse de covariance (ANCOVA) était utilisée pour traiter les données.  Le prix de détail pour un mois donné était régressé avec la variété, la location, la saison et la covariante, et la tendance. Le modèle s’accordait bien, expliquant 46% de la variation totale de prix des commodités. Les paramètres estimés de la variété virtuelle, de la location virtuelle et la covariante, la tendance étaient  significatifs. Six de 11 paramètres estimés de saison dummies étaient significatifs alors que cinq ne l’étaient pas. Par conséquent le prix saisonier a été startifié pour le prix menstuel des commodités à bas prix, et à prix élévé. La différence des prix par location était inversément liée à la distance du marché central, et la différence de prix saisonniers était attribuée à la technique de stockage. Ceci montre un comportement competitive imparfait du marché. Le peu avec la couleur brune, une peau rêche, et largeur de graine a un prix premium que le sokoto, qui était  blanc, une peau rêche et petites graines. Les différences spatialles et temporelles observées en prix peuvent être réduites avec un bon réseau des routes, des techniques améliorées de stockage et une information adéquate sur les marchés. Les attributs du niébé de large graines, la peau rêche et la couleur brune sont recommendés dans les programmes de culture.

Mots Clés: Prix de commodité, marché, variables dummy détailées, technique de stockage

INTRODUCTION

Cowpea, Vigna unguiculata  (L.) Walp is an important grain legume in Nigeria for subsistence and generation of income. The Sudan and Sahel zones in the north dominate in cowpea production though the crop is gradually attaining economic importance in the southern states. Faye et al. (2000) also observed that since the 1980s cowpea has become an alternative cash crop in northern Senegal. Cowpea grain contains between 20-25 percent protein (McFarlane, 1983) and 64 percent carbohydrate (Bressani, 1985). It therefore has a tremendous potential to contribute to the alleviation of malnutrition among poor families. The crop serves to bridge the hunger gap between planting and harvesting periods of main food crops. It is an inexpensive source of protein for both rural poor and urban consumers. In addition, cowpea contributes to the sustainability of cropping systems and soil fertility improvements in marginal lands by providing ground cover and plant residues, fixing nitrogen and suppressing weeds (Inaizumi et al., 1999). They help to maintain yield of agricultural crops in areas where fertilisers are hard to obtain or are not affordable (Golop et al., 1996). Inaizumi et al. (1999) also observed that some cowpea varieties cause suicidal germination of Striga hermonthica, which is a devastating, parasitic weed of cereals. Cowpea has been described as an ideal crop for the semi arid regions of the tropics where other food legumes may not perform well.

In Nigeria, the production trend of cowpeas has experienced about a 441 percent increase in area planted and 410 percent increase in yield from 1961 to 1995 (Ortiz, 1998). According to Inaizumi et al. (1999), several factors account for the impressive increase in cowpea production among which is the significant advances made by International Institute of Tropical Agriculture (IITA) over the last two decades in improving cowpea productivity in sub-Saharan Africa. Singh et al. (1997) also indicated that a number of varieties have been developed which combine diverse plant type, different maturity periods, and resistance to several diseases, insect pests and parasitic weeds and they possess good agronomic traits.

Despite the many advantages of cowpeas to consumers and producers, a major problem of cowpea production is the wide price differential of the commodity over space and time among varieties. Price variation over space in excess of transportation cost could be the result of inadequate market information, poor road networks and inadequate market infrastructure. Price variation over time in excess of storage cost could be the result of inefficient storage facilities, while price variation among varieties could be due to variations in consumers’ preference.

This paper therefore focuses on providing information on the pricing pattern of cowpeas in Ogun State through empirical analyses with a view to improving the marketing of the commodity. This is important because it is through an efficient marketing of farmers’ produce, that the diverse advantages provided by the crop would optimize producers and consumers utilities. The specific objectives of the paper therefore are i) identifying cowpea characteristics which command premium price and ii) examining if there are spatial and temporal price differences and whether they are justified.

MATERIALS AND METHODS

Study location. The study was conducted in  Ogun State, Nigeria. Ogun State Agricultural Development Programme (OGADEP) delineated the state into four ecological/geopolitical zones. The zones are Abeokuta, Ijebu-Ode, Ilaro and Ikenne with zonal headquarters at Abeokuta, Ijebu-Ode, Ilaro and Sagamu, respectively.

Abeokuta is the state capital and the most populated headquarters. It has a population of 0.44 million people (OSBLPP, 1999). It has more amenities than the other zonal headquarters and a higher concentration of government workers because it is the administrative capital of Ogun State. It has many farming families in its surrounding villages. Ijebu-Ode is a semi-urban town with a population of 0.17 million people (OSBLPP, 1999) and it also has many farming families. Ilaro is the least developed of the four zonal headquarters and it is relatively more of a rural setting. It has a population of 0.15million people. Sagamu is also a semi-urban town with a population of 0.18 million people. It is the closest of all the zonal headquarters to the Lagos- Ibadan express way through Ogun state in which goods coming from the northern states pass to Lagos, the commercial centre of Nigeria.

Cowpea varieties. The commodity under study, cowpea, has many varieties. The most commonly cultivated varieties in Ogun State included:  IT 90K-76, IT 90K-59, IT 90K-277-2, IT 87D-941, IT 89KD-88, IT 98KD-88,IAR-48 and Ife brown. However when they reach the markets it becomes difficult to identify them by their code variety names. Traders in the state however, generally sell five basic types of the commodity, which they have categorised in line with physical features and their price premium. These are locally dubbed peu/drum, sokoto, mala,  oloyin (flat and large)’ and olo. However, because of the dearth of appropriate data, this study has focused on two types (i.e., peu/drum and sokoto) on which secondary price data is available over an appreciable period which is required for the study. The two have rough skin and are black-eyed grains. Peu/drum however, has dark brown large grains while sokoto is white with relatively smaller grains.

Data sources. Monthly time series retail commodity price data (in kobo/kilogram) between January 1989 and December 1999, collected by Ogun State Ministry of Finance and Economic Planning from four markets of Abeokuta, Ijebu-Ode, Ilaro and Sagamu (in Ikenne zone) were used. The markets were chosen according to their location and volume of cowpea sales. Additionally, primary data collected in 1999 using checklists on the features of the two-cowpea types and the characteristics of the locations were used.

Analytical technique. Cowpeas are available all the year round in the market but with wide price variation. At harvest, when the grain is abundant in the market, price is usually low but increases with time. The price usually peaks shortly before another harvest. With traders incurring the cost of storage over time, the price of the commodity rises with time. This therefore makes temporal consideration germane in the commodity-pricing pattern. Furthermore, increasing price over the years makes trend consideration inevitable.

Given the focus of the work on variety price premium attributes, spatial and temporal pricing pattern, the general form of the model is specified as:

Pt = f (V, L, S, T, u)

Where,

Pt  = price at time t,

V = variety,

L = location,

S = season (measured in months),

T = trend  (measured over the years),

and  u is the random error term.

The error term u, is assumed to be normally distributed with a mean of zero and a constant variance, s2.

Analysis of covariance (ANCOVA) was used to estimate the relationship between the commodity’s monthly retail price between 1989 and 1999 and the factors of variety, location, season and the covariate- trend, because of the presence of factors and metric variable in our model as regressors. The data matrix is of the dimension 1,056 by 17. The 1,056 cases, 132 are made up of two varieties by four locations by (11 years by 12 months) observations, that is (2 by 4 by 132) = 1,056 cases. The 17 columns are the retail price,(one column); variety dummy (which is number of varieties less one), one column; location dummies (which is number of locations less one) three columns; seasonal dummies (which is number of months less one) eleven columns  and the covariate, trend, giving a total of 17 variables made up of one regress and, 15 non-metric and one metric regressors . The classes of each of the categorical variable have to be less by one to avoid perfect collinearity of the regressors in order to make the model estimable. Given the non-metric regressors represented by dummy variables (dichotomous variables), base variables were chosen for each of the factors as shown in Table 1.

The base variables were chosen to make them coincide with classes, which have the lowest prices among or between their categories. For example, the price of sokoto is generally lower than that of peu/drum, thus sokoto variety was considered  as base class. Prices of cowpea varieties were generally lowest in Sagamu perhaps because of its proximity to the Ibadan-Lagos express way, the route through which cowpea is moved to Lagos from the north where the commodity is produced. Sagamu was therefore, considered as the relative central market for the commodity in Ogun State and thus was chosen as the base location.

With respect to season, January was chosen as base month because cowpea price as observed from our data is lowest then. The choice of these classes of dummy variables as base variables is important, because it allows for positive values of the regression coefficients for ease of interpretation of the results. The base year of the analysis is the year 1989.

A wide variety of functional forms can be employed ranging from a simple linear regression form to quadratic Box-Cox models (Lansford et al., 1995). The appropriate functional form cannot in general be specified on theoretical grounds (Halvorsen and Pellakowski, 1981). Care must be taken therefore in functional form specification. Two functional forms which allow for easy results interpretation were however, used. These being the form where, LN (PRICE) is a function of variety, location, season (measured in months) and the covariate, trend (measured by year of the price observation). The second simply stated as; PRICE  was a function of variety, location, season and trend, that is the linear functional form. Each of the models contained 15 non-metric and one metric variable as regressors. The variables used in the models are listed in Table 2, for comprehension.

The nature of the data suggests the potential statistical problem of heteroscedasticity and autocorrelation. Generalised Least Squares (GLS) was therefore used to correct for correlation of random error with the regressors and the temporal error correlation.

Estimation of the model and addressing potential data problems was followed by an examination of the marginal values for each factor considered. For the non-metric variables (i.e., dichotomous or yes/no variable such as whether the variety is peu/drum or it is not), the marginal or incremental value is the difference between the predicted price of cowpea if the cowpea is peu/drum and the predicted price of cowpea if it is  sokoto, in linear functional form. The marginal prices may aid producers with choice of variety to plant. Marginal price indicates how much value a buyer places on another unit of a given characteristic, given the set of initial characteristics (i.e., those of the base periods). The marginal prices of a model are dependent upon the value of each and every characteristic (or variable) included in the model. Hence, the marginal price of a particular characteristic, such as variety, will vary as the total set of other factors of the commodity changes.

RESULTS AND DISCUSSION

The estimated models fited the data reasonably well given the variables used. The R-square values for the LN (PRICE) and the PRICE models are 0.41 and 0.46, respectively indicating that the models explained 41 and 46 percent of the variations in the monthly retail price of cowpeas  LN (PRICE) and PRICE, respectively (Tables 3 and 4). The results are consistent with that of Commer (1990) and that of Lansford et al.  (1995) who worked on explaining the prices of race-bred yearling quarter horses, and obtained R-square value of 0.38 and R-square value of 0.40,  respectively. However, given the better performance of the linear model in terms of fit and the logical consistency of the estimated parameters, it was adopted for use in the subsequent discussion.

Given the objectives of the study, the factors in the model were focused upon because the significance of the parameters of the various factor classes, lent support to the fact that the model intercepts are shifted, thus implying that there were vatietal and temporal price differences.

For example, the variety estimated parameter value of 469.40 kobo /kg, is statistically significant at the 5% level of probability and has the expected sign. With respect to location, all the three parameter estimates were statistically significant at the 5% level of probability and had the expected positive sign. In terms of season, which showed temporal price pattern, the model stratified the results into two categories. Six out of the eleven estimated parameters were statistically significant at the 5% level of probability. These were the parameters for the months of   June, July, August, September, October and November. Those of February, March, April, May and December were however, not statistically significant at the 5% probability level. Generally, the marginal parameter estimates measured  the estimated price difference between the attribute level and its base period value. The interpretation of the seasonal parameter estimates for this latter group was the base price of cowpea, which is January price for the season dummy variable, and was not statistically significant from those of February, March, April, May and December.

This was expected because during this period (or months), the volume of harvested cowpeas increased and lead  to increase in market supply, thus prices during these months were low and  not statistically different from the base period price, that is the low commodity price in January. However, the story was different for the commodity price in the months of June, July, August, September, October and November. Commodity prices during these months were statistically different from that of the base period of January. This too was expected because commodity supply becomes lean in the markets, and traders engage in temporal arbitrage during these months. Much of the commodity is stored for future sale and consumption, thereby incurring storage costs and thus increasing commodity price.

The analytical technique thus, made it possible to stratify the commodity price into two, the months during which prices were statistically not different from the base period and those that are. Thus, the price pattern exhibited by the data is that, the low price months are December through May while the high price months are June to November. The price difference for these months and January ranges between 448-690 kobo/kg. Commodity price difference over season is a measure of storage cost and efficiency of storage technique (Durojaye and Aihonsu, 1988). Price increases of between 448- 690 kobo /kg. /month is definitely on the high side. Efficient storage system may reduce storage cost and consequently reduce price increases.

In terms of spatial price differences our analysis shows that commodity prices in Abeokuta, Ijebu-Ode and Ilaro were statistically different from those of Sagamu at the 5% level of probability. The lowest price of the commodity was observed at Sagamu. This is attested to by the positive values of the estimated parameters of the location dummies. Recall that a particular location parameter estimate is the predicted price difference between the particular location and the base location when temporal effects have been accounted for. Sagamu market is the closest market to the Ibadan-Lagos expressway in Ogun state on which trucks hauling cowpeas from north to south travel. The commodity price difference at Abeokuta, Ijebu-Ode and Ilaro compared to the base location, Sagamu were estimated at  394, 475 and 321 kobo/kg, respectively. Price was highest in Ijebu-Ode, followed by Abeokuta and Ilaro for the study period. If transportation cost is solely responsible for the spatial price difference and Sagamu is regarded as the central market for cowpeas in Ogun State, given the kilometer separation of 63, 38 and 125 of Sagamu from Abeokuta, Ijebu-Ode and Ilaro, respectively, the sequence of price difference expected starting from the largest to the lowest is Ilaro (the farthest town to Sagamu) followed by Abeokuta and then Ijebu-Ode, which is the closest to Sagamu.

Non-conformity to the sequence shows that there are other factors at play. This could be the result of inefficient infrastructure such as the road network, transportation, inadequate market information as well as the effect of local production to supplement cowpea supply from the north. However, all things being equal, under competitive pricing policy, spatial price difference should equal the cost of transportation, (Bressler, 1975; Ejiga, 1988; Afolami, 2000). If this price relationship were to be upheld, at least there should have been a direct correlation between locational price differences and their distance from Sagamu.

In respect of variety price difference, the estimated parameter for the variety dummy variable is 470 kobo/kg. This was significantly different from zero at the 5% level of probability and has the expected positive sign.  This means that on the average, the difference between the predicted price for peu/drum and sokoto for the period of the study was  470 kobo/kg. This  show that peu/drum received a price premium relative to sokoto. Thus, it is recommended that the attributes of peu/drum cowpea be fostered in breeding programmes. Since peu/drum is large, rough skinned and has dark brown grains, while sokoto is small, rough skinned with white grains  cowpea price premium attributes seem to be large grains and brown colour in the locations investigated.

Other attributes of cowpeas which have not been investigated empirically in this work, but which consumers accord premium are taste, cookability and cooking time, given that energy and time have costs and can be scarce. Another important factor considered by producers and consumers is the resistance of cowpeas to bruchids- Callosobruchus maculatus, a storage pest. Golob et al. (1998), reported weight loss resulting from infestation of about 9% after six months of storage for the crop in Ghana, if not properly stored. These and other organoleptic characteristics are subjects of future empirical work.

There are variety price premium, spatial and seasonal price differences for cowpeas in Ogun State, Nigeria. Spatial price differences cannot be attributed only to transportation cost and seasonal price differences may be high enough to justify a more effective storage system.

Cowpea price was found to be highest in Ijebu-Ode and July was the month in which highest price was recorded for peu/drum variety, while cowpea price was found to be lowest in Sagamu in  January. Spatial and temporal price difference may be reduced through an improved transportation system and better storage facilities while variety characteristics of large grains, rough skin and brown colour should be accorded priority in cowpea breeding programmes.

REFERENCES

  • Afolami,C.A. 2000. Inter-temporal and spatial pricing efficiency in maize  marketing in Nigeria. Moor Journal of Agricultural Research 1(1):77-85.
  • Bressani, R. 1985. Nutritive value of cowpea. In Cowpea Research, Production and Utilization, edited by B.B. Singh and K.O. Rachie. John Wiley and Sons,  N.Y, U.S.A. pp. 353-360.
  • Bressler, R.G. and King, R.A. 1970. Markets, Prices and International Trade. John Wiley and son Inc., N.Y. U.S.A.
  • OSBLPP, 1999. Ogun State Bureau for Lands and Physical Planning   Abeokuta, Ogun State.
  • Commer Malcom Jr., 1990. Price determination of thorough breed yearlings - Effects of non-phenotypic data. Selected Paper, Annual Meeting, American Agricultural Economics Association.
  • Durojaiye, B.O.and Aihonsu, J.O.Y. 1988. Market integration and seasonal prices of  staple food-stuffs: A case study of Ogun State, Nigeria. Food Policy.Butter worth & Co (Publishers) Ltd. pp. 375-382
  • Ejiga, N.O.O. 1988. Pricing efficiency for cowpeas in Northern Nigeria. Readings in  Agricultural Marketing. Adekanye, T.O.  (Ed.), pp. 157-167. Longman Nigeria.
  • Faye Mbene, M. Ndiaye, J.L. Deboer, A. 2000. Identifying cowpea characteristics which command price premium in Senegal markets. World Cowpea Research Conference III, IITA, Ibadan Sept. 4-7.
  • Golob P.,  Gudrups, I. and Moss, C.  1996. The storage and marketing of cowpea in  Ghana.
  • Golob P., Andan, H.F., Atarigiya, J. and Tran B.M.D. 1998. On-farm Storage losses of  cowpea in Northern Ghana. Research project funded by UK Department of  International Department  (DFID) R6503 Crop Post Harvest Research Programme.
  • Halvorsen, R. and Pollakowski, H.O.  1981. Choice of functional form for hedonic price equations. Journal of Urban Economy 10:37-49.
  • Inaizumi, H., Singh, B.B.,  Sanginga, P.C.,  Manyong, V.M.,  Adesina, A.A. and Tarawali,  S. 1999. Adoption and impact of dry season dual purpose cowpea in the semi arid zone of Nigeria”. IITA and Meg.Comm. Network 14pp.
  • Lansford, N., Topliff, D. Freeman, D.  and  Walker, O. 1995. Explaining Prices of race-bred yearling quarter horses produced by quarter horse sires and dams. Oklahoma Current Farm Economics 68:25-37
  • McFarlane, J.A. 1983. In: Food Storage Manual. WFP/FAO/HMSO U.S. Publication  number Q 1547. 263pp.
  • Ortiz R. 1998. Cowpeas from Nigeria: A silent food revolution. Outlook on Agriculture 27: 125-128.
  • Singh, B.B., Chambliss, O.L. and Sharma, B.  1992. Recent advances in Cowpea breeding. In: Advances in Cowpea Research. Singh, B.B., Mohan Raj, D.R., Dashiell, K.E. and  Jackai, L.E.N. (Eds.), pp. 30-39. IITA and Japan International Research Centre for Agricultural Sciences  (JIRCAS) Co-publication. New York.

©2002, African Crop Science Society


The following images related to this document are available:

Photo images

[cs02026t2.jpg] [cs02026t4.jpg] [cs02026t3.jpg] [cs02026t1.jpg]
Home Faq Resources Email Bioline
© Bioline International, 1989 - 2022, Site last up-dated on 11-May-2022.
Site created and maintained by the Reference Center on Environmental Information, CRIA, Brazil
System hosted by the Internet Data Center of Rede Nacional de Ensino e Pesquisa, RNP, Brazil