Economic performance of smallholder soya bean production in Kwara State, Nigeria

1.1 ODDA scheme

ODDA is an intervention programme of the Kwara State established in 2017. It is formulated to drive employment in agriculture, boost food production, alleviate poverty, and improve standards of living while also diversifying the state’s economy. The scheme provides seedlings, fertilizers, chemicals, and other farming tools and also matches farmers with off-takers. The programme is targeted at farmers who engage in the production of soya bean, cassava, rice, and maize. ODDA aims to make farming attractive to the youths, and it has disbursed about N1billion to over 1,000 farmers in the state and provided access to 2,550 hectares of land across the state.

1.2 Theoretical framework

In economic theory, the term “production efficiency” refers to the highest output that may be achieved given a combination of inputs and available technologies. The ratio of actual output to stochastic frontier output is regarded as production efficiency. A producer can produce the highest output (at the frontier) with a set of factors and technology [21]. However, individual production may differ due to personal, environmental, or institutional attributes, which may fall below the frontier.

Production efficiency could be assessed using either the parametric technique of stochastic frontier analysis (SFA) or the nonparametric approach of data envelopment analysis (DEA) from a theoretical and empirical standpoint. SFA accounts for random shocks in production activities [22]. DEA method, on the other hand, ignores measurement flaws and other types of statistical noise. Owing to this intrinsic advantage, and also taking into cognisance, smallholder farmers’ production is prone to numerous random errors arising from factor use [23]. Furthermore, SFA is based on the unpredictable nature of agricultural production, which allows factors driving farmer technical inefficiency to be included in the model; hence, SFA was chosen for this study.

A general SFA form is as follows:

The output ( y i ) by an ith producer is a function of input factor, vector of the associated coefficients (β), and the error term ( ε i ), which is expressed as follows:

ε i is further discomposed into v i and u i .

where v i is the symmetric random error accounting that controls for noise effects, and u i is expressed as a function of external factors Z i that influence the production efficiency.

where v i is independent and uniformly distributed across observations with a mean of zero and variance σ v 2 . Also, u i is independent and identically distributed with truncation at zero (N + (μ, σ ² μ)). The non-negativity assumption of μ _i implies that the observations cannot be above the production possibility frontier and u i follows the distribution F.

Equation (1) can be further expanded as:

where i = 1, 2, 3 … n and j = 1, 2, 3 … n are the number of producers, In is the natural logarithm, y i is the output, and x i is the vector of factors of production.

Equation (6) depicts the functional form of technical inefficiency, which is influenced by a variety of socioeconomic and farm-specific factors:

where u i = ( 1 − TE i ) , which is the technical inefficiency of the ith producer, and z j i is the set of exogenous factors, which are vector inefficiencies such as age, sex access to credit, education level, and household size.

SFA is estimated using equation (5) and the maximum-likelihood estimation (MLE) technique, where the likelihood function is expressed in terms of the variance parameter [24,25,26]:

where σ ² s represents the output variance owing to changes in random shock and inefficiency. If the value of y = 1, there is complete inefficiency, and if the value of y = 0, there is no technical inefficiency [26].

2.1 Area of study

This research was conducted in Kwara State, Nigeria. Kwara State is located in Nigeria’s north-central geopolitical region. It covers approximately 32,500 km² and has a population of 2,371,089 inhabitants [27]. The state is bordered by Niger State to the north, Osun and Ondo States to the south, Kogi State to the east, and Oyo State to the west. Kwara State shares an internal boundary with the Republic of Benin [28]. More than 80% of its population is engaged in farming. The state’s main mineral resources are limestone, feldspar, kaolin, clay, granite, quartz, and tantalite [29].

2.2 Sampling procedure and sample size

The study respondents were chosen using a purposive sampling methodology, which involved the selection of two local government areas (LGAs) from Zone C agricultural development project zones based on the predominance of soya bean farmers in the two LGAs. The two LGAs chosen are Moro and Ilorin East LGAs. Five villages were selected each from the two LGAs: Oke-Ose, Onikoko, Onimomo, okeoyi, and Oshin from Ilorin East LGA, while Logun, Malete, Marafa, Obalaja, and Obale from Moro LGA. Finally, ten farmers that benefited from the intervention were selected at random from each village, giving a total of 100 respondents.

2.3 Data collection

Primary data were used for this study. The data were obtained by administering a well-structured questionnaire coupled with a scheduled interview to collect requisite information from the respondent. Farmers’ socioeconomic characteristics (age, household size, educational status, amount of credit received, number of extension contacts, and number of years spent as cooperative members) and production characteristics data (level of inputs and output) among others were collected.

2.4 Analytical techniques

The data gathered were analysed using descriptive statistics, profitability analysis, and SFA.

2.5 Profitability analysis

Following previous studies [30,31,32,33,34,35], the net farm income (NFI) was used to estimate the profitability of soya bean farmers. Below is the formula for net farm profits:

where NFI is the net farm income (₦), TR is the total revenue (₦), and TC is the total cost of production (₦).

where TVC is the total variable cost (₦), and TFC is the total fixed cost (₦).

where RI is the return to investment.

Decision rules:

When RI > 1, it implies farmers are making profit.

When RI = 1, the farmers are at breakeven.

When RI < 1, the farmers are at a loss.

2.6 The SFA model

To assess the technical efficiency of soya bean farmers, the MLE technique was used to calculate the Cobb–Douglas output frontier function. SFA is defined as follows:

where ln represents the natural logarithm, y i indicates the output of soya beans (kg), β 0 is the constant, β 1 – β 5 are the regression coefficients, x 1 indicates the farm size (hectares), x 2 represents the quantity of seeds (kg), x 3 represents the fertilizer (kg), x 4 indicates the total labour employed (man hours), x 5 indicates the agrochemical volume (litres), and ε i is the composed error term.

To estimate the determinants of technical efficiency, socioeconomic characteristics were incorporated into equation (6) to estimate the determinants of the farmers’ technical inefficiency.

where U i is the ith farm’s technical inefficiency, Z 1 represents the farmer’s educational level in years, Z 2 represents the household size (number), Z 3 indicates the farmer’s age in years, Z 4 represents the years of farming experience, Z 5 represents the membership in cooperative society, ∂ 0 is the constant, and ∂ 1 – ∂ 5 are the parameters to be estimated.

3.1 Costs and returns analysis

The survival of a business is a function of the amount of profit realized at a given time. Profit is calculated as the difference between the economic value of goods produced and the expense incurred on the resources utilized to make them. The sum of revenue generated and the running costs of a business enterprise determine how much profit or loss the enterprise can make over time. The profitability analysis, which was used to evaluate the profitability of soya bean production, is presented in Table 2.

The average cost of soya bean seeds used per hectare is ₦5,750, accounting for about 8% of the overall production costs. The estimated cost of fertilizer used is ₦10848.49, which stands at 15% of the overall production costs. The average cost of agrochemicals constitutes 8% of the total cost of production. The labour costs consist of the cost of land preparation, planting, fertilizer and insecticide application, weeding, harvesting, and processing. Labour is obtained from both relatives and employed sources. The average labour cost per hectare is ₦46985.96, accounting for 65% of overall production costs, while the total cost of fixed inputs (land rental and tool depreciation) incurred on soya bean production is ₦2,256, accounting for 3% of total costs. Labour cost constitutes the largest chunk of the cost of production, while fixed input cost accounts for just a meagre part of the total cost of production. This result corroborates the findings of Ogunjinmi et al. [44] that soya bean production requires a lot of labour.

NFI is calculated by the product of average yield (kg) and the selling price in naira, which is 130/kg.

3.2 Return on investment in soya bean production

The total revenue is ₦107663.40k, while the total cost (TVC + TFC) is ₦72165.48k, according to the results in Table 2. As a result, the NFI is ₦37753.92k. The average rate of return on investment (return on investment per naira invested) is 1.49, which means that for every ₦1 invested in soya bean production in the study zone, a profit of 49 kobo was made. Consequently, it can be inferred that soya bean production is lucrative in the study region with a return of 49 kobo for every naira invested. The results concur with the findings of previous studies [16,44,45,46,47]; these studies were undertaken in Nigeria and other West African countries. Their results revealed that soya bean production by smallholder producers was a profitable enterprise.

3.3 Technical efficiency

Table 3 shows the MLE of the Cobb–Douglas stochastic frontier production function of soya bean farmers in the study area. Results in the table show that the estimated sigma-squared (σ ²) is 0.052 and is statistically significant and different from zero at 0.01 level. This indicates a good fit of the model and the correctness of the specified distributional assumptions of the composite error term. The estimated gamma (γ) is 0.99 and is significant at 1% range, which means that inefficiency effects exist. The results suggest that about 99% of the variation in soya bean output among the farmers in the study area is due to the differences in their technical efficiencies.

3.4 Determinants of technical efficiency

The determinants of technical efficiency (factors influencing technical efficiency) as shown in Table 3 are discussed in the following paragraphs.

Farm size (X ₁): The estimated coefficient of farm size is positive and significant at 1%, which implies that there is a positive relationship between the farm size and technical efficiency. It follows that a 1% improvement in farm size results in a 99% increase in soya bean output. This may be because larger farm sizes promote the adoption of innovations that can result in higher yields. This result supports the findings of previous studies [15,17,44,27,48,49] that farm size has a positive and significant relationship with soya bean yield, but contradicts with those of previous studies [10,45,50,51].

Quantity of seed (X ₂): The coefficient of the quantity of seed is negative and significant at 1%. This implies that the quantity of seed used has an inverse relationship with the technical efficiency of production; this also means that the greater the number of seeds used by the farmers, the lower the output as presented in Table 3. This may be due to inappropriate use of seed per hole, spacing, and quality seeds among other factors. The output will be poor even if other inputs are in sufficient or adequate proportions, while the quality of the seeds is poor. This is supported by Yegon et al. [15]; however, the quantity of seed was not significant in their study. This finding is not in agreement with those of previous studies [17,45,46,52,53], which reported seed as a booster of soya bean efficiency.

Labour used (X ₄): The coefficient of labour used is positive and significant at 1%. This shows that the quantity of labour used has a positive impact on soya bean performance. As the quantity of labour used in soya bean farming increases, so does the output of soya beans. This finding is in tandem with the findings of previous studies [18,19,44] that a positive significant relationship existed between labour and the output of soya beans. However, it contradicts with findings of previous studies [15,49], which showed that labour used to lessen the technical efficiency of soya bean farmers.

Technical inefficiency: As seen in Table 3, the coefficients of age and cooperative membership are positive and statistically significant at the 1% level. The aforementioned findings show that the increase in age and cooperative membership tend to result in a corresponding decrease in soya beans farmers’ performance. This could be because the older a farmer becomes, the less productive the farmer. Farmers’ decisions are also influenced by their age because older farmers tend to be more risk-averse than younger farmers who are risk-takers. Physical strength is essential for the agricultural production process, which is also on the decrease among older farmers. The decline in strength after middle age necessitates a greater investment of labour in the same production activity. According to the study by Guo et al. [54], increasing age is not favourable for improving agricultural output. This is in agreement with the studies by Biam et al. [10] and Yegon et al. [15] but contradicts with the studies by Moses [17], Nyongesa et al. [46], and Shalma [55].

Cooperative membership leading to a decrease in soya bean production could be because being in a cooperative requires some commitment such as time and finance. This could affect farmers’ productivity one way or the other. This result agrees with the findings of Moses [17] and Ugbabe et al. [45]; however, this contradicts the findings of previous studies [10,15,18,56] that farmers who are members of a cooperative are more likely to be more efficient. Household size and education level are negative and significant at 5 and 10%, respectively. However, years of farming experience are positive but not significant. The increase in household size and level of education tend to result in a corresponding increase in farmers’ technical efficiency. An increase in the number of individuals in a household contributes to an increase in farmer technical production by the use of the family members as labour, thereby cutting on labour and improving productivity. As a result, those with higher household sizes are more technically efficient than those with smaller household sizes. This result is affirmed by that of Asodina et al. [19] but opposed to the results by Biam et al. [10] and Moses [17], who reported that household size limits the level of efficiency of soya bean farmers. An increase in educational level could lead to an increase in technical efficiency. This could be because education creates an edge for an individual by exposing such an individual to knowledge and application of the knowledge. This is consistent with the findings of previous studies [10,15,18,19,44,57] that an increase in years of schooling contributes to the technical efficiency of the farmers because they can easily adapt to innovation and improved technology.

3.5 Efficiency scores

Data in Table 4 revealed that the mean technical efficiency was 56%; according to previous studies [17,18,19], the mean technical efficiency recorded by soya bean farmers were 66, 73, and 59%, respectively. This suggests that farmers under this scheme are not as efficient as they ought to be and there is still room to increase their technical performance by 44%. This means that there is an efficiency gap that needs to be filled. This could be bridged by training programmes, effective extension-farmer relationships, and so on.