Ex-Ante Analysis of the Demand for new value added pulse products: A case of Precooked Beans in Uganda Paul Aseete, Enid Katungi, Jackie Bonabana, Michael Ugen and Eliud Birachi
Background Common bean is the most important grain legume (produced and consumed) in east and central Africa Its consumption is however constrained by long cooking time and high fuel requirements. While breeding has produced fast cooking bean varieties, the reduction in cooking time has not been significant Through a Public private partnership, efforts are ongoing to explore industrial processing options to reduce cooking time This will lead to the production of fast cooking beans - precooked beans
Background Precooked beans are beans processed by cooking under high temperature and pressure and packed in air tight and weather proof sachets that can preserve them for about six month Once precooked one will only need to cook beans for 10-15 minutes Key benefits will be; adapt to changing lifestyles of growing urban population, reduce cost of fuel and environmental conservation. Little is known about how benefits from precooked beans will translate into demand and acceptability The objective of this study was to determine consumer acceptability and demand for precooked beans and to segment the market.
Methods A choice experiment was used to elicit consumer demand of precooked bean based on product attributes The main attributes included: cooking time, nutritional enhancement, fuel saving, price, and water requirement These were combined, based one their levels, into choice sets (Figure 1) and subjected to respondents. Data was collected from 558 households in five districts of central Uganda - Kampala, Wakiso, Mukono, Buikwe and Luweero Respondents were blocked in 3 blocks and each respondent was subjected to 7 choice sets
Methods Figure 1: Sample of choice set showing attribute combinations.
Methods The latent class model was used to assess consumer valuation and to segment potential precooked beqn consumers. The model has the capability of estimating weights attached to each attribute and classifying consumer into homogenous groups base on their preferences. The unconditional probability of choosing alternative j in i th situation is PP rr(tt(nn)) = exp (λszn) II ( λszn) jj=1 TT(nn) X tt (nn) exp( ββ 1 ssxx nnnnnn ) II exp( ββ 1 jj=1 ss XX nnnnnn )
Results Variable (Description) Analysis Sample Statistic Will buy precooked beans % Yes 89.1 Purchase frequency No of meal in a week 4.1 Av. Expected Purchase Quantity (For a Households) Kg 0.96 Best Price Ush (USD) 2,695 (0.84) Consumer location % urban 59.6 Gender of the household head % male 72.7 Sauces eaten in a week (ate more than 4 sauces) % 22.8 Source of beans for consumption % own production 30.2 Source of beans for consumption % from market 66.1 Education of the household head (Years ) Mean 10.19 (5.056) No. of dependents (Children below 15 years) Mean 3.33 (2.670) Quantity of beans consumed in a week (Kg) Mean 3.28 (3.535) Distance to the bean market (Km) Mean 0.58 (1.270) Household monthly income Mean 821,000 (1874225)
Results Table 2: Consumer valuation of precooked beans attributes Attribute Parameters Multinomial ASC β ASC 0.0617 (0.0875) logit TIME β TIME -0.0082*** (0.0017) FUEL β FUEL 0.0051*** (0.0010) NUTRI β NUTRI 1.6614*** (0.0515) WATER β WATER 0.1580*** (0.0478) PRICE β PRICE -0.0002*** (0.2409D-4) Latent class logit Class 1 Class 2 Class 3 Utility coefficients of parameters in latent class model Coefficients (Standards Errors) -0.3628** 1.6936*** (0.1519) (0.2015) -0.0087*** -0.0104*** (0.0031) (0.0033) 0.0097*** 0.0042** (0.0019) (0.0019) 2.9218*** 0.4522 *** (0.0789) (0.1163) 0.1214 0.3774*** (0.0817) (0.0980) -0.0002*** -0.0003*** (0.3856D-04) (0.4109D-04) LL= -2784.77, AIC=5649.5, BIC=2950.2, ρ2 = 0.35105 Observations =3906, Sample = 558-1.6766*** (0.5137) -0.0100 (0.0105) 0.014388* (0.0054) 0.7430*** (0.2363) 0.2083 (0.3934) -0.0011*** (0.0002)
Results Table 3: Determinants of Latent Class membership Household characteristic Class 1 Class 2 3 Constants 5.6273*** (2.1154) 6.8917*** (2.1936) - Consumer location (1=Urban, 0=otherwise) 0.1853 (0.4257) -0.0975 (0.4711) - Gender of the household head (1= Male, 0=female) -0.2321 (0.4212) -0.6507 (0.4787) - Education of the household head (Years) 0.0976** (0.0380) 0.0404 (0.0424) - Number of dependents (Number less 15 years) 0.0564 (0.0761) 0.0479 (0.0855) - Quantity of beans consumed in a week (Kg) -0.1255** (0.0615) -0.0430 (0.0681) - Household monthly income (Ln income) -0.3077** (0.1461) -0.3472** (0.1461) - Distance to the bean market (Km) -0.483D-04 (0.0004) 0.0393*** (0.0018) - Processing benefits (PB index) 0.6538***(0.2067) 0.5823** (0.2405) - Sauce Diversity (1=More than 4 sauces, 0=Less than 4 sauces ) Source of beans for consumption (1= Market, 0=Otherwise) -0.2308 (0.5157) 0.5022 (0.5521) - 0.0021** (0.4279) -0.0363*** (0.0022) -
Results Class 1 has significantly more educated household heads, consume less beans, have low incomes, like benefits and rely on bean from the market We can say these are - Fairly low income urbanites. Class 2 has low income earners, face longer distance to the market, like benefits in precooked beans and rely less on bean from the market. We can say these are Rural low income households Though class 3 is normalized, it has close characteristics as class 1.
Conclusions Attributes that will influence the demand for precooked beans include; nutritional enhancement, water saving and fuel saving these can form Unique selling points The principle consumers are likely to be low income urban households that rely on the market as a source of bean for consumption - - - 63% of the market. The precooked bean product should be produced and promoted as a product with diverse benefits Further research needs to be done on the actual product once on the market to elicit actual demand and make demand projections.
ACKNOWLEDGEMENTS
THANK YOU