SUMMARY REPORT OF THESIS Publisher: Asian Economic and Social Society ISSN: 2224-4433 Volume 2 No. 2 June 2012 Acknowledgement: The authors acknowledge the support of the Bean/Cowpea CRSP Project, University of Ghana and University of Georgia, USA Consumer Preference for Processed Cowpea Products in Selected Communities of the Coastal Regions of Ghana Nimoh, F. (Department of Agricultural Economics, Agribusiness and Extension, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana) Asuming-Brempong, S. (Department of Agricultural Economics & Agribusiness, University of Ghana, P. O. Box LG 68, Legon, Accra, Ghana) Sarpong, D. B. (Department of Agricultural Economics & Agribusiness, University of Ghana, P. O. Box LG 68, Legon, Accra, Ghana) Citation: Nimoh, F., Asuming-Brempong, S. and Sarpong, D. B. (2012): “Consumer Preference for Processed Cowpea Products in Selected Communities of the Coastal Regions of Ghana”, Asian Journal of Agriculture and Rural Development, Vol. 2, No. 2, pp. 113-119. Asian Journal of Agriculture and Rural Development, 2(2), pp. 113-119 114 Author(s) Nimoh, F. Department of Agricultural Economics, Agribusiness and Extension, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana Email: frediemoh@yahoo.com Asuming Brempong, S. Department of Agricultural Economics & Agribusiness, University of Ghana, Legon, Accra, Ghana Sarpong, D. B. Department of Agricultural Economics & Agribusiness, University of Ghana, Legon, Accra, Ghana Consumer Preference for Processed Cowpea Products in Selected Communities of the Coastal Regions of Ghana Abstract The nutritive value of cowpea as an essential source of protein to supplement carbohydrate diets has long been recognized. Its role as a subsidiary crop to be relied on during the “hungry season” and during times of food shortages, drought, inflation and the subsequent erosion of the consumer’s purchasing power, particularly among the urban poor, makes it a crop of choice by housewives who look for nutritious but cheaper sources of food. This paper sought to investigate consumer preference for processed cowpea-based products, such as, boiled cowpea with cereals, fried cowpea paste, and cowpea fortified maize dough in selected communities of the coastal regions of Ghana. Using descriptive statistics, Kendall’s Coefficient of Concordance, and Logit Model, it was found that there was high preference for processed cowpea-based products in all the communities studied; and that processing cowpea into various food types was relatively profitable. Key socio- economic factors and consumer characteristics that influence preference include gender, marital status, income, education, product taste, sustainability of products (satisfying) and product availability. The production of gas (flatulence) after consumption of the products was the most pressing factor that influences preference. Unavailability of the products was identified as the least pressing factor. The researchers recommend that the production and utilization of cowpea in the study area and in other parts of Ghana should be encouraged as it would help to both improve the nutritional status of consumers and also help generate income to producers and processors. There should also be further research into the disliking intrinsic characteristics of the products considered. Keywords: Consumer Preference, Cowpea Products, Coastal Regions, Ghana Introduction Among the most difficult problems confronting the world communities since the history of humankind have been those of food shortages and diet deficits. It is estimated that, more than 800 million people in developing countries, including Ghana, are undernourished and the total gram of protein consumption per day is low in these countries as compared to the developed countries. For example, Ghana consumes 49.6 g/day, as compared to the USA (112.5g/day), (FAO, 1996). Protein energy malnutrition as assessed by physical growth and body measurement is still widespread throughout Ghana (NNS, 1986). Many consumers spend a greater proportion of their income on food, but they still remain malnourished, lacking in the required protein-calorie level needed for effective development and proper growth of the body. Cowpeas have been identified as containing adequate levels of protein to help curb the problem of protein malnutrition. Because of their nutritive value, cowpeas have been identified as good vehicles for combating protein calorie malnutrition in Ghana (Sefa-Dedeh, 1993). It has been documented that over 80% of mothers in Ghana use cowpea to prepare infant foods and family diets (Mensa-Wilmot et al., 2001). Nevertheless, the preference for cowpeas is highly variable among consumers. Amegatse (1995) identified some of the mitigating factors against increased utilization of cowpea as: the hard-to-cook defect, high fuel consumption, prolonged cooking time, and characteristics of beany flavor and indigestion. Cowpeas can be processed into many local low-cost, but highly nutritious ready-to-eat or partially made foods such as: fried cowpea paste (Akla / Kosai); boiled cowpea with gari (Bobo / Yo’kε’ Gari); boiled cowpea mailto:frediemoh@yahoo.com Asian Journal of Agriculture and Rural Development, 2(2), pp. 113-119 114 and rice (Waakye); steamed cowpea paste (Alele / Tubani), and cowpea fortified maize dough (CFMD). Such products if acceptable and affordable by a broader section of the population will be an important contribution to improving the lives and health of the people of Ghana and elsewhere. The main objective of the study therefore was to investigate consumer preferences for cowpea-based products in selected communities of the coastal regions of Ghana. The specific objectives of the study were to: determine the size, distribution, and level of preference for cowpea/cowpea-based products; identify the factors that influence the preference for these products; analyze the effect of the factors that influence the preference for the products; and evaluate the financial benefit of processing cowpeas in the study area. Methodology Study Area Six communities, namely; Akatsi (Volta region–V/R), Nima and Madina (Greater Accra region-GAR), Winneba (Central region–C/R), and Abuadze and Takoradi (Western region–W/R) of Ghana were the study area. The choice of these locations was based primarily on the fact that, the Bean/Cowpea Collaborative Research Support Project (CRSP), of which the study was part, undertook some preliminary studies at the locations and additional social-economic information was needed for further studies in those localities. The communities are densely populated and many inhabitants are migrants who are engaged in commercial activities. Majority of the people have their daily meals outside the home (food-away-from-home). The importance of cowpea, with its high nutritive value, has therefore, made it possible for cowpea-based food processors/sellers to occupy a sizeable portion of the food market in the study area. Type and Source of Data Both primary and secondary data were used for the study. Two types of questionnaires (for consumers and cowpea processors) were used to obtain data for the study. A total of 166 consumers and 135 cowpea processors were interviewed. The purposive and simple random sampling techniques were used to select the respondents. Analytical Methods Descriptive statistics such as frequency tables and percentages were used to analyze the size, distribution, and level of preferences for cowpea and cowpea-based products in the communities of the study area. Kendall’s Coefficient of Concordance (W) analysis was used to rank the items identified as constraints/problems against the preference for the cowpea-based products into the most pressing to the least pressing (Mattson, 1986). The degree of agreement of the rankings by the consumers was then measured. W ranges from 0 to 1. In deriving W, let T, represents the sum of ranks for each constraint/problem being ranked (socio-economic factors and product characteristics). The variance of the sum of ranks is given by:   n nTT VarT    / 22 (1) The maximum variance of T is given by:   12/122 nm (2) Where m is the number of consumers and n is the number of constraints/problems. The formula for (W) is then given by:     12/1 /)/( 22 22      nm nnTT W (3) This simplifies to the computational formula for W as: W =     1 /12 22 22    nnm nTT (4) In this study, n includes elements like product taste, product scent, gas production (flatulence), product price, unavailability of the products, mode of presentation, and ignorance about nutritive value of products. Hypothesis: H0: there is no agreement between the rankings of the influencing factors, and H1: there is agreement between the rankings of the influencing factors. W was tested for significance in terms of the F distribution, given by: ((m - 1)*Wc) / (1 – Wc), with (n- 1) - (2/m) degrees of freedom for the numerator and (m- 1)*((n-1)-(2/m)) degrees of freedom for the denominator (Edwards 1964). The Logit model, based on the cumulative logistic probability function was used to analyze the effect of factors that influence the consumer’s preference for the cowpea products (Pindyck and Rubinfeld, 1991). The dependent variable is dichotomous or binary choice, having two options of preference and non-preference for processed cowpea products. The Logit model was specified as follows: Pi = F (Zi) = F ( + Xi) = Zie11 = )(11 Xie  (5) Asian Journal of Agriculture and Rural Development, 2(2), pp. 113-119 115 where, e is the base of natural logarithms, which is approximately equal to 2.718. Pi is the probability that the consumer will make a certain choice, given Xi, the products characteristics and socio-economic characteris- tics of the consumer. Estimation of the Logit model Both sides of equation (5) are multiplied by Zie1 to get:  Zie1 Pi = 1 (6) Equation (6) is divided by Pi, and then 1 subtracted from it to get: Pi e Zi 1  - = 1 – Pi / Pi (7) By definition, however, ZiZi ee 1 so that: Pi Pi eZi   1 (8) By taking the natural logarithm of both sides we get: Pi Pi Zi   1 log Or XiZi Pi Pi   1 log (from equation (5) (9) The dependent variable in the regression equation is the logarithm of the odds that, a particular choice would be made. Since Pi is the probability of consumer preference for processed cowpea-based products, (1-Pi) is the probability of non-preference for processed cowpea- based products. Pi / (1-Pi) is the odds ratio in favor of preference for processed cowpea-based products. Socio- economic characteristics of the consumer (sex, age, household head, household size, income, educational level, marital status, and occupation) and product characteristics (scent, taste, nutritive and economic value, sanitary and health related issues) served as the explanatory variables. From equation (9), the Logit model for this study is given by: Log (Pi/I-Pi) = β0 + β1SEXi + β2AGEi + β3HHHDi + β4HHSZi + β5INCOMi + β6EDUCi + β7MSTi + β8OCCUi + β9Ppi + β10NUTRi + β11SUSTi + β12TASTEi + β13SCENTi + β14CHPDTi + β15GASi +Ui (10) Subscript i is the ith observation, Pi is the probability that the ith consumer has a preference for the product given Xi. SEXi is Gender of the ith Consumer (male = 1, female = 0); AGEi is Age in years; HHHDi is Household Head (head = 1, otherwise = 0); HHSZi is Household Size; INCOMi is Income level in Ghanaian Cedis/month; EDUCi is Educational Level (Dummy: Primary = 1, otherwise = 0; J.S.S. = 1, otherwise = 0, etc.); MSTi is Marital Status (Dummy: Married = 1, otherwise = 0; Single = 1, otherwise = 0; etc.); OCCUi is Occupation (Dummy: Self employed/businessperson = 1, otherwise = 0; etc); Ppi is Price of Products in Ghanaian Cedis; NUTRi is Nutritive Value of Product (Nutritious (protein content) = 1, otherwise = 0); SUSTi is Sustainability of Product (sustainable =1, otherwise =0); TASTEi is Product Taste (tasty/delicious = 1, otherwise = 0); SCENTi is Product Scent (like scent = 1, otherwise = 0); CHPDTi is Affordability of Product (cheap = 1, otherwise = 0); GASi is Gas content/flatulence (gaseous = 1, otherwise = 0); βi = the vector of estimated coefficients; and Ui = the error term. The a-prior expectations were: β1 > 0, β2 < 0 β3 < 0, β4 < 0, β5 < 0, β6 < 0, β7 > 0, β8 > 0, β9 < 0, β10 > 0, β11 > 0, β12 > 0, β13 > 0, β14 > 0, β15 < 0. The Z statistics was used to measure the level of significance for each of the estimated coefficients. The goodness of fit statistics given is the Mc-Fadden R- squared. The likelihood-ratio (LR) test was computed to determine the joint significance of the independent variables in the model. The LR test statistics followed a standard chi-square (X 2 ) distribution with the degrees of freedom equal to the number of independent variables used in the model. The higher the percentage of the prediction, the greater is the predictive power of the model. The discussion results were based on the log- odds ratio. The profitability of processing cowpea into various products was evaluated using profit margin analysis. The analysis, a financial ratio, was aimed at measuring how efficient the cowpea processors use their assets in their operations (Ross et al. 2001). This is given by the net income divided by sales as follows: Profit Margin = Net Income / Sales (11) The net income from processing cowpea was computed by finding the difference between the total revenue generated and the total cost of operation per olonka (bowl of cowpea of approximately 3kg) per day. Mathematically, the net return was computed as follows: NR = TR – TC (12) iiQPTR  and    n iXi XPTC 01 , where TR = total revenue generated from sales per day, Pi = the price of product, Qi = quantity of product produced and sold in a day, TC = total cost of operation in a day, Pxi = price of inputs, and Xi = ith input. Asian Journal of Agriculture and Rural Development, 2(2), pp. 113-119 116 All other things being equal, a relatively high profit margin is desirable. This would imply a low expense. Lowering sales price will not usually increase unit volume, but will normally cause profit margins to shrink. Total profit (or more importantly, operating cash flow), may go up or down, so the fact that margins are smaller isn’t necessary bad, because, prices may be so low that there may be a loss on everything sold but this is recovered in volume. A positive margin, however, is an indication of profit whereas a negative margin indicates a loss. Results and Discussion Results of the study showed that, of the 166 consumers interviewed, approximately 49%, prefer already processed cowpea (ready-to-eat or partial products) and 51% prefer unprocessed cowpea (make-at home products), (Table 1). The preference for cowpea in the various communities showed that majority of consumers in Akatsi (V/R), and Abuadze and Takoradi (W/R) prefer unprocessed cowpea to processed cowpea, but in Nima and Madina (GAR) and Winneba (C/R) the preference is processed cowpea to unprocessed cowpea (Table 2). About the preference for the various cowpea- based products, the results of the study showed that there is a high consumer preference for Akla/Kosai, Bobo/Yo ke Gari, and Waakye compared to CFMD, Alele/Tubani (Table 3). The frequency of consumption of the various cowpea-based products was found to be twice per week (Table 4). Table 1: Preference for Cowpea: Processed and Unprocessed (All Communities) Processed Cowpea Unprocessed Cowpea Total Frequency 82 84 166 Valid % 49.40 50.60 100 Source: Field Survey, 2004 Table 2: Preference for Cowpea: Processed and Unprocessed (Separate Communities) Akatsi (VR) Nima / Madina (GAR) Winneba (CR) Abuadze / Takoradi (WR) Pro Unpr Total Pro Unpr Total Pro Unpr Total Pro Unpr Total Freq. 16 23 39 23 19 42 24 16 40 19 26 45 V. % 41 59 100 55 45 100 60 40 100 42 58 100 Note: Pro. = Processed, Unpr. = Unprocessed, V = Valid. Source: Field Survey, 2004 Table 3: Preference for Cowpea-Based Products Product Frequency Valid % CFMD 29 17.47 Akla / Kosai 137 82.53 Bobo / Yo’kε’ Gari 161 96.99 Waakye 161 96.99 Alele/Tubani 56 33.73 Source: Field Survey, 2004 Table 4: Number of Times of Consumption of Cowpea Products per Week ONE TWO THREE FOUR FIVE SIX SEVEN Total Freq 30 39 35 23 16 3 14 160 Valid % 18.75 24.38 21.88 14.38 10.0 1.88 8.75 100 Source: Field Survey, 2004 Table 5 presents the results of the rankings and the degrees of agreement of the rankings, W, of the factors that influence the consumer’s preference for cowpea and cowpea products. In Akatsi (V/R), the most pressing factor was found to be the presence of foreign matter (such as stones) in the products, and the least pressing factor was testa color of the beans. At Nima and Madina (GAR), the most pressing factor was the presence of foreign matter (such as stones) in the products, and the least pressing factor was ignorance about the nutritive value of the products. At Winneba (C/R), the production of gas (flatulence) was the most pressing factor, and ignorance about the nutritive value of the products was the least pressing factor. At Abuadze and Takoradi (W/R), the most pressing factor was the production of gas (flatulence) and the least pressing factor was unavailability of the products. In aggregate (thus, for all the communities), the production of gas (flatulence) was the most pressing factor, and unavailability of the products was the least pressing factor. The tests of significance in terms of F distribution of the degree of agreement or concordance (W) between the rankings of Asian Journal of Agriculture and Rural Development, 2(2), pp. 113-119 117 the influencing factors showed fairly low (<50%) degrees of agreement between the rankings for the separate communities and also for the aggregate. Table 5: Identifying and Ranking of Factors that Influence Preference for Products Influencing Factors (Problem/Constraint) RANKING (Separate Communities) Influencing Factors (Problem/Constraint) RANKING (All Communities) Akatsi (V/R) Nima/ Madina (GAR) Winneba (C/R) Abuadzi/Takoradi (W/R) Foreign matter (stones) 1 1 2 2 Gas (flatulence) 1 Damage level (weevils) 4 3 3 3 Foreign matter (stones) 2 Gas (flatulence) 3 2 1 1 Damage level (weevils) 3 Lengthy cooking time 2 4 5 5 Lengthy cooking time 4 Mode of presentation 8 5 4 4 Mode of presentation 5 Price of products 5 6 6 6 Price of products 6 Products scent 6 7 7 7 Products scent 7 Products taste 7 9 11 8 Products taste 8 Size of beans 10 10 8 10 Size of beans 9 Testa color 12 8 9 9 Testa color 10 Ignorance about prodt 8 12 12 11 Ignorance about product 11 Unavail. of product 10 11 10 12 Unavail of product 12 No. of Consumers 39 42 40 45 No. of Consumers 166 Coefficient of Concordance (W) 0.1375 13.75% 0.1868 18.68% 0.2035 20.35% 0.2880 28.80% Coefficient of Concordance (W) 0.1885 (18.85%) Source: Authors Computations The result of the Logit analysis showed that, among the various socio-economic factors of the consumer and products characteristics, sex (gender), income, education, marital status, sustainability (satisfying) of product and taste, influence the consumer’s preference for the cowpea products. Sex (males), marital status (single - bachelor/spinster), sustainability and taste exert positive influence on preference for the products, while income and education exert negative influence on the preference for the products (Table 6). Table 6: Logit Analysis on Factors Influencing the Preference for Cowpea Products Variable Coefficient Std. Error z-Statistic Prob. C -1.109004 1.286893 -0.861769 0.3888 Sex 1.203198*** 0.432085 2.784629 0.0054 Income -6.44E-07* 3.63E-07 -1.772922 0.0762 Essec -1.450965** 0.674059 -2.152579 0.0314 Etert -2.593890*** 0.748859 -3.463791 0.0005 Msing 1.908392*** 0.633784 3.011106 0.0026 Sust 1.304174* 0.736504 1.770763 0.0766 Taste 1.348201* 0.721530 1.868530 0.0617 Mean dependent var = 0.475904 S.D. dependent var = 0.500930 LR statistic (15 df) = 66.21596 Probability (LR stat) = 2.09E-08 McFadden R-squared = 0.288222; Obs. with Dep=0: 87 Obs. with Dep=1: 79 Total obs. = 166 Note: *, **, ***, denotes significance at 10%, 5%, and 1%, respectively. ESSEC = education at secondary level, ETERT = education at tertiary level, MSING = single for marital status, SUST = sustainability (satisfying) of products. Processing cowpea into the various food types was found to be relatively profitable in the study area. On the average, for every olonka of cowpea processed into Kosai/Akla, processors in all the communities generated ¢26.31 profit for every ¢100.00 sales in a day; ¢36.72 for yo ke gari processors; and ¢30.95 for waakye processors. However, there were disparities in profits among the various communities (Table 7). Asian Journal of Agriculture and Rural Development, 2(2), pp. 113-119 118 Table 7: Financial Benefit of Processing Cowpea Akatsi (V/R) Winneba (C/R) Abuadze & Takoradi (W/R) Nima & Madina (GAR) Average (ALL Communities) Akla / Kosai TR/olonka/day TC/olonka/day NR/olonka/day Amount (¢) 43,000.00 35,000.00 8,000.00 Amount (¢) 54,049.30 41,725.35 12,323.95 Amount (¢) 49,295.75 35,211.27 14,084.51 Amount (¢) 43,173.96 27,716.98 15,456.98 Amount(¢) 47,379.75 34,913.40 12,466.36 PM/olonka/day 0.1860 (18.60%) 0.2280 (22.80%) 0.2857 (28.57%) 0.3580 (35.80%) 0.2631 (26.31%) Yo ke Gari TR/ olonka/day TC/ olonka/day NR/olonka/day Amount (¢) 52,971.00 36,374.93 16,596.08 Amount (¢) 57,937.21 33,538.73 24,398.48 Amount (¢) 72,237.88 47,370.89 24,866.99 Amount (¢) 40,646.25 24,322.82 16,323.43 Amount(¢) 55,948.09 35,401.84 20,546.25 PM/olonka/day 0.3133 (31.33%) 0.4211 (42.11%) 0.3442 (34.42%) 0.4016 (40.16%) 0.3672 (36.72%) Waakye TR/ olonka/day TC/ olonka/day NR/ olonka/day Amount (¢) 139,601.14 87,172.13 52,429.01 Amount (¢) 79,086.54 60,010.52 19,076.02 Amount (¢) 92,862.82 62,784.09 30,078.13 Amount (¢) 63,090.91 48,705.05 14,385.86 Amount(¢) 93,660.35 64,667.95 28,962.26 PM/olonka/day 0.3756 (37.56%) 0.2412 (24.12%) 0.3239 (32.39%) 0.2280 (22.80%) 0.3096 (30.96%) Note: TR = Total Revenue, TC = Total Cost, NR = Net Return, PM = Profit Margin. Olonka of cowpea weighs approximately 3kg. Exchange rate: 1 US Dollar = ¢8,775.40 Ghanaian Cedi. Conclusion and Recommendations Based on the results of the study, it is concluded that consumers in the study area have different levels and forms of preference for cowpeas. The characteristics of the cowpea products and socio-economic features of consumers also have influence on the level of preference. Key socio-economic factors and consumer characteristics that influence consumer preferences include gender, marital status, income, education, product taste and sustainability (satisfying) of products. The processing cowpea into various food types is also relatively profitable. Recommendations drawn from the results of the study are that: (1) the production and utilization of cowpea in the study area and in other parts of Ghana should be encouraged as it would help both improve the nutritional status of consumers and also help generate income to cowpea producers and processors; (2) there should also be more research into the disliking intrinsic characteristics of cowpea; and (3) to solve the problem of malnutrition, particularly protein malnutrition, and also improve the living standards of people in the study area and Ghana at large and elsewhere, there is the need for structural development support such as the establishment of an effective producer-processor linkage and an efficient food research and standardization, instead of symptomatic nutrition intervention activities. References Amegatse W. K. (1995) “Assessment of Cowpea (Vigna Unguiculata Walp) Production, Consumption and Nutritional Evaluation of Cowpea–Fortified Fermented Maize in Ga District”, M. Phil Thesis submitted to the Dept. of Nutrition and Food Science, University of Ghana, Legon. Edwards A. L. (1964) “Statistical Methods for the Behavioral Sciences”, Holt, Rinehart and Winston, New York, pp. 402, 410. FAO Paper (1996) “Food and Nutrition Division and Statistics Division”, FAO, Rome, Italy IITA (1983) “Research Highlights”, Ibadan, Nigeria, IITA. Mattson D. E., (1986) “Statistics–Difficult Concepts, Understandable Explanations”, Bolchazy - Carducci Publishers, Inc. pp. 281-283, 361,423. Mensa-Willmot Y., Dixon, P. R. and Sefa-Dedeh, S. (2001) “Acceptability of Extrusion Cooked Cereal / Legume Weaning Food Supplements to Ghanaian Mothers”, International Journal of Food Science & Nutrition, Vol. 52(1), pp. 83. Asian Journal of Agriculture and Rural Development, 2(2), pp. 113-119 119 NNS (1986) Report on NNS by National Food and Nutrition Board, Accra, Ghana. Onwueme, I. C. and Sinha, T. D. (1991) “Field Crop Production in Tropical Africa”, Technical Centre for Agric and Rural Co-operation (ACP), pp. 289, 292, 298. Pindyck, R. S. and Rubinfeld, D. C. (1991) “Econometric Models and Economic Forecasts”, Third Edition, McGraw - Hill Inc. pp. 248. Rachie K. O. (1985) “Cowpea Research, Production and Utilization, A Wiley –Inter science Publication. Ross, S. A., Westerfield, R. W. and Jordan, B. D. (2001) “Essentials of Corporate Finance”, Third Edition, McGraw-Hill/Irwin Co. Inc. pp. 59. Sefa-Dedeh S. (1993) “Development of High Protein Energy Foods from Grain Legumes”, Proceedings of an International Workshop, University of Ghana, Legon, Ghana, Feb. 5-7,1991, Association of African Universities, Food and Nutrition Project, 1993.