Econometric Analysis of Factors Determining Market Outlets Choice of Cultured Fish Farming Entrepreneurship among Nigerian Youths from the Blue Economy

Objectives of the study

• To identify the cultured fish marketers in the study area.
• To examine the factors that determine the market outlet choice decision of cultured fish marketers in the study area. This was achieved with the use of a market outlet choice model using Multinomial Logit Regression.
• To investigate the determinants of the revenue of cultured fish marketers in the study area. This was achieved with the use of Multiple Regression.

Methodology

Sampling procedure

A multistage random sampling was used to select 320 respondents from three states (Ogun, Osun, and Oyo), South West, Nigeria. Purposive sampling was used to select three states, simple random sampling was used to select two local governments from each state, while stratified random sampling was used to select respondents from each local government [1]. The final selection was 76 farmers, 73 wholesalers, 79 retailers, and 92 processors. The selection was based on the number of registered cultured fish farmers and marketers in the ADP. Twenty questionnaires were discarded because they were not well filled out.

Market outlet choice model

A multinomial logit (MNL) model was applied to explain inter-household variation in the choice of a specific marketing outlet. This study assumed that a cultured fish marketer’s decision was generated based on its utility maximisation. This implies that each alternative marketing outlet choice entails different private costs and benefits, and hence different utility, to a household decision maker. The analytical model was constructed as follows. Suppose that the utility to a household of alternative j is Uij, where j = 0, 1, 2…. From the decision maker’s perspective, the best alternative is simply the one that maximises net private benefit at the margin. In other words, household i will choose marketing outlet j if and only if Uij>Uik. It is important to note that a household’s utility cannot be observed in practice. What a researcher observes are the factors influencing the household’s utility, such as household and personal characteristics and attributes of the choice set experienced by the household [28,29]. Based on McFadden [30], a household’s utility function from using alternative j can then be expressed as follows:

U (Choice of j for household i) = Uij = Vij + εij

where,

Uij is the overall utility,

Vij is an indirect utility function and

εij is a random error term.

The probability that household i selects alternative j can be specified as: Pij = Pr (Vij + εij>Vik + εik)

Pij = Pr (εik<εij + Vij – Vik ,)

Assuming that the error terms are identically and independently distributed with a type I extreme value distribution, the probability that a household chooses alternative j can be explained by a multinomial logit model [21,31] as follows:

$Pij= \frac{\text{exp}(\beta jXij)}{{{\displaystyle \sum }}_{j=0}^J\text{exp}(\beta jXij)}$

where,

Xij is a vector of household of the ith respondent facing alternative j

βj is a vector of regression parameter estimates associated with alternative j.

Following the equation above, we can adapt the MNL model fitting to this study as follows:

$P(CHOICEij=j= \frac{\text{exp}(\beta jXij)}{{{\displaystyle \sum }}_{j=0}^J\text{exp}(\beta jXij)}$

where,

i represents the ith cultured fish marketer household, and I = 1,2,3,…,300.

j represents different marketing outlets, j = 0 for sale to wholesalers, j = 1 for sale to processors, j = 2 for sale to retailers, and j=3 for sale to consumers

P represents the probability of cultured fish marketing outlet j to be chosen by cultured fish marketer i;

CHOICEij = j means that cultured fish marketing outlet j is chosen by cultured fish marketer i;

Xi is an independent variable [1,32].

It is a common practice in econometric specification of the MNL model to normalise the equation by one of the response categories such that βj = 0. In this regard, the MNL model can alternatively be specified as follows:

$Pij= \frac{\text{exp}(\beta jXij)}{{{\displaystyle \sum }}_{j=1}^{J-1}\text{exp}(\beta jXij)}$

The coefficients of explanatory variables on the omitted or base category are assumed to be zero. The probability that a base category will be chosen can be calculated as follows:

$Pij= \frac{1}{1+{{\displaystyle \sum }}_{j=1}^{J-1}\text{exp}(\beta jXij)}$

The marginal effects of the attributes on the probability of choice are determined by differentiating equation (11):

$\delta j= \frac{\partial Pj}{\partial Xj}=Pj=Pj [\beta j- {\displaystyle \sum }_{j=0}^J(Pj)(\beta j)$

for j= 1, 2….J

where,

Pj is the probability that farmers choose market outlet j

βj is a vector of regression parameter estimates associated with alternative j.

Independent variables:

These include all variables that are associated with factors that determine the outlet choice decisions of the cultured fish marketers [33].

Descriptions, Measurement, and Expected Signs of Variables for Multinomial Logit Regression Analysis on Outlet Choice Decisions.

Y 0 - Wholesalers

1 – Processors = Dependent Variables

2 – Retailers

3 – Consumers

Results and discussion

Multinomial Logit Regression (MLR) of market outlets of the actors in the market chain

Table 1 presents the results of the Multinomial logit model used to investigate outlet choice decisions in the cultured fish market chain. Four different market choices were used as the outcome, and these are wholesalers, processors, retailers, and consumers. Wholesalers are used as the reference group. A total of 10 explanatory variables were considered in the econometric model, from which 4 were found to be statistically significant at various levels under the processors and retailers market outlet, while 3 were statistically significant under the consumers’ outlet. The significant variables are the gender of the respondent, the cost of fish sold, the quantity of fish for sale, and access to market information. The Wald chi-square (W_T) ratio of 499.90 with a p-value of 0.0000 reveals that the model as a whole is statistically significant. The model has a log likelihood of -267.81136.

Processors market outlet relative to wholesalers

According to Table 1, gender was negative and statistically significant at a 1 percent level of significance. The result showed that the probability of males relative to females decreased by 2.4351 for choosing processors as market outlet relative to wholesalers while holding all other variables in the model constant. The cost of fish sold was positive and statistically significant at a 10% level of significance. This result explains that for a naira increase in the cost of fish sold, the probability of choosing a processor’s market outlet relative to wholesalers increased by 0.0071, given all other variables are held constant. The output price of fish serves as a motivation for marketers as well as to determine the choice of marketing channels. This finding is also consistent with that of Zivenge and Karavina [34], S Edoge [35], and Ololade, et al. [1]. The quantity of fish for sale is negative and statistically significant at the 1 per cent level of significance. This explains that an increase in the quantity of fish sold by one kilogram reduced the likelihood of choosing a processor’s market outlet relative to wholesalers by 0.0003 while other variables are held constant. Access to market information is negative but statistically significant at the 1 per cent level of significance. The result showed that the probability of fish marketers who had access to market information was 12.9741 lower for choosing a processor’s market outlet relative to wholesalers, giving all other variables constant. According to Edoge [35], access to information measured in terms of ownership of cell phone, radio, or television was a significant positive determinant of channel choice at 1% significance level. Farmers with these facilities have better access to market information. A one per cent increase in the number of fish farmers with access to market information increases the probability of deciding on channel choice. This finding is consistent with the report of Zivenge and Karavina [34].

Retailers market outlet relative to wholesalers

According to Table 1, gender was negative and statistically significant at a 1 percent level of significance. The result explains that the probability of males relative to females was 1.9603lower for choosing retailers relative to wholesalers as market outlet while holding all other variables in the model constant. The cost of fish sold was positive and statistically significant at a 1 per cent level of significance. The result shows that for a naira increase in the cost of fish sold, the likelihood of choosing retailers’ market outlet relative to wholesalers increased by 0.0120, given all other variables in the model constant. Marketers are more responsive to market price relative to the transaction cost. The output price of fish serves as a motivation for marketers to determine the choice of marketing channel. This finding is also consistent with that of Zivenge and Karavina [34]; Edoge [35], and Ololade, et al. [1]. The quantity of fish for sale is negative and statistically significant at the 1 per cent level of significance. The result explained that an increase in the quantity of fish sold by one kilogram will reduce the probability of choosing retailers relative to wholesalers as market outlet by 0.0003 while other variables are held constant. Access to market information is negative but statistically significant at the 1 per cent level of significance. The result showed that the probability of fish marketers who had access to market information relative to those who do not have access is 10.9403 lower for choosing retailers’ market outlet relative to wholesalers, while holding all other variables in the model constant. According to Edoge [35], access to information measured in terms of ownership of cell phone, radio, or television was a significant positive determinant of channel choice at 1% significance level. Farmers with these facilities have better access to market information. A one per cent increase in the number of fish farmers with access to market information increases the probability of deciding on channel choice. This finding is consistent with those of Zivenge and Karavina [34] and Amaya and Alwayng 2011.

Consumers market outlet relative to wholesalers

According to Table 1, gender was negative and statistically significant at a 1 percent level of significance. This showed that the likelihood for males relative to females is 3.2753lower for choosing consumers as market outlet relative to wholesalers while holding all other variables constant. This is contrary to the findings of Umoinyang 2014. that there are more males in the fish marketing in Akwa-Ibom state, Nigeria. The cost of fish sold was positive and statistically significant at a 10 per cent level of significance. The result showed that for a naira increase in the cost of fish sold, the probability of choosing consumers relative to wholesalers as market outlet increased by 0.0148 while holding all other variables constant. The output price of fish serves as a motivation for marketers as well as to determine the choice of marketing channels. This finding is also consistent with that of Zivenge and Karavina [34] and Edoge [35]. Access to market information was negative but statistically significant at the 1 per cent level of significance. This explained that the probability of fish marketers who had access to market information relative to those who do not have access was 12.4268 lower for choosing consumers’ market outlet relative to wholesalers, while holding all other variables in the model constant.

MaOriginal effect from multinomial logit on the choice of marketing outlets

It is important to note that the marginal effect from the multinomial logit on the choice of marketing outlets explains the change in the predicted probability of choosing a specific market outlet for a one-unit change in an independent variable.

Processors market outlet relative to wholesalers

As shown in Table 2, the cost of fish sold was negative and statistically significant at the 5 per cent level of significance. This showed that for a naira decrease in the cost of fish sold, the probability of choosing a processor market outlet relative to wholesalers decreased by 0.003045, holding all other variables in the model constant. This is contrary to the findings of FAO [36]. Fish prices are therefore influenced by demand in those areas, which enables them to meet their transport costs and production costs.

Access to market information is negative but statistically significant at the 1 per cent level of significance. The result explained that the probability of fish marketers who had no access to market information relative to those who had access was 0.3301764 lower for choosing a processor market outlet relative to wholesalers, while holding all other variables in the model constant. This, in terms, goes with the findings of Geoffrey [37] that 36%, 65% and 75% of the market participants who sold at the farm gate, local market and urban market, respectively, had access to price information, which implies that the majority of market participants who sold at the urban market had access to price information. Again, business decisions are based on dynamic information such as consumer needs and market trends [38]. According to Edoge [35], access to information measured in terms of ownership of a cell phone, radio, or television was a significant positive determinant of channel choice at 1% significance level. Farmers with these facilities have better access to market information. A one percent increase in the number of fish farmers with access to market information increases the probability of deciding on a one percent increase in the number of fish farmers with access to market information increases the probability of deciding on channel choice. This finding is consistent with those of Zivenge and Karavina [34] and Amaya and Alwayng 2011.

Retailers market outlet relative to wholesalers

According to Table 2, sex was positive and statistically significant at a 10 per cent level of significance. This implies that the probability for males relative to females is 0.1046578 higher for choosing retailers as market outlets relative to wholesalers while holding all other variables in the model constant. This is contrary to the findings of Umoinyang 2014. that there are more males in the fish marketing in Akwa-Ibom state, Nigeria. Marketing experience was positive and statistically significant at a 10 per cent level of significance. This implies that the probability of fish marketers who have more years of marketing experience is 0.006 higher for choosing retailers as market outlets relative to wholesalers, while holding all other variables in the model constant.

Consumers market outlet relative to wholesalers

Sex was negative and statistically significant at a 10 per cent level of significance. This implies that the probability for males relative to females is -0.2160306 lower for choosing consumers as market outlet relative to wholesalers while holding all other variables in the model constant (Table 2). The cost of fish sold was positive and statistically significant at the 5 per cent level of significance. This gives the implication that for a naira increase in the cost of fish sold, the probability of choosing a processor market outlet relative to wholesalers is expected to increase by 0.007115, given all other variables in the model are held constant. Marketers are more responsive to market price relative to the transaction cost. The output price of fish serves as a motivation for marketers to determine the choice of marketing channel. This finding is also consistent with that of Zivenge and Karavina [34]; Edoge [35]. Marketing experience was negative and statistically significant at a 1 per cent level of significance. This implies that the probability of fish marketers who have fewer years of marketing experience is -0.0168259 lower for choosing consumers as market outlet relative to wholesalers while holding all other variables in the model constant. This implies that as marketing experience and education increase, net income also increases. This result corroborates the findings of Offor and Nse-Nelson [39], who opined that marketing experience has a significant influence on net income and marketing efficiency.

Determinants of revenue of the marketers in the fish market chain in the study area

Determinants of revenue of the marketers in the fish market chain in the study area are presented in Table 3. Multiple regression coefficients of catfish marketing are summarised and also presented in Table 3. The exponential functional form was chosen as the lead equation based on the statistical and econometric consideration of the R² (coefficient of determination) value, which shows the explanatory power of the independent variables; the F-ratio, which shows the goodness of fit of the specified model and the conformity of the significant variables with a priori expectation. The R² value of 0.5293 was obtained, which implies that 52.93% of the total variation in the dependent variable was accounted for by variation in the independent variables. The F-ratio was significant at 1% probability level. This indicates the goodness of fit of the model and the relevance of the variables to the model. The significant variables include sex, years of education, total market cost and distance to the market.

Sex: Sex was significant at 1% and is positively related to the revenue of the marketers. A plausible explanation for this is that male marketer tends to take risks, thus they are capable of searching markets in distant and competitive places for their produce. Conversely, female tends to be confined at home by household chores, hence hindering them from attending the marketplaces (Table 3). The finding concurs with that of Morrison, et al. 2007, who found that female farmers are faced with gender specific constraints like a time burden that limits them from accessing the best market for their output, which will definitely increase their income.

Years of education: Years of education were significant at 5% and were positively related to marketers’ revenue. This, therefore, explains that improving the quality of education or individual years of education will positively influence the revenue of marketers (Table 3). This is supported by the findings of Palla, et al. [40], who discovered in their findings that formal education is a significant variable for the marketer’s margins, an obvious indication that formal education enhances the performance of business practitioners. In addition, a possible explanation for the positive and significant nature of the years of education variable may be that education is significant in the life of an individual; it helps to shape attitudes, values and behaviour, promoting inquisitiveness and innovations in the process. It grooms the mind and makes it receptive to technological innovation and managerial skills [41].

Total market cost: Total market cost was high at 1% and is positively related to marketers’ revenue. The market cost comprises the transport cost, labour cost and the cost of marketing equipment (Table 3). This implies that an increase in the cost of transportation decreases the revenue of the marketers. Kumar, et al. [42] also observed that fish marketers at the landing site were more or less fishermen who do not transport or store their produce but sell to marketers coming from the environs. They also observed a negative and significant coefficient in the transportation of fish marketers. Also, as the cost of labour of the cultured fish marketers increases, their revenue reduces. Cost of labour is a very important factor in revenue determination because if the cost of labour incurred in a marketing system by a marketer is high, it will have a negative and significant effect on his/her revenue.

Distance to the market: Distance to the market was significant at 1% and was positively related to the revenue of marketers. Considering the distance to market by a marketer is a very important factor because as the distance to market increases, the cost of transportation will definitely increase, and this will reduce the quantity of fish available for sale, thereby reducing the marketer’s revenue (Table 3). This is related to the result by Gurmis and Melese [43] on analysing the factors that influence market participation among avocado producers in the Kaffa Zone of South-Western Ethiopia [44,45].

Conclusion

The Multinomial Logit Regression (MLR) model revealed that negative but significant relationships existed between sex, quantity of fish for sale, access to market information and processors’ market outlet relative to wholesalers, while a positive but significant relationship existed between cost of fish sold and processors’ market outlet relative to wholesalers. The MLR model also revealed that negative but significant relationships existed between sex, quantity of fish for sale, access to market information and processors’ market outlet relative to wholesalers, while a positive but significant relationship existed between cost of fish sold and processors’ market outlet relative to wholesalers. The MLR model further showed negative but significant relationships between sex, access to market information and processors’ market outlet relative to wholesalers, while a positive but significant relationship between costs of fish sold and processors’ market outlet relative to wholesalers. Sex, years of education, total market cost and distance to the market had a positive and significant relationship with the revenue of the cultured fish marketers in the study area. Therefore, due to the significant relationship between some demographic characteristics of the marketers (sex, years of education, total market cost, distance to the market) and the revenue of marketers in the study area, the null hypothesis was rejected and the alternative accepted. It is recommended that the Nigerian government should pay special attention to market access facilities, especially good roads, to enhance the maximum profitability of the cultured fish business in the study area. Credit should also be made available for cultured fish producers to encourage them to increase their production, since marketing starts from the producers. It is also suggested that government extension workers should encourage more cultured fish youth producers, marketers and processors to register under ADP, so as to be beneficiaries of different incentives or programs that may be organised by the government, since marketers need to be educated on how to maximise their profit.

Data availability statement

Data supporting the findings of this study are available within the article.

Authors contribution

RAO: Conceptualisation, Design, Data collection, Analysis, Writing of manuscript.

FIO: Data collection, Analysis, and Writing of the manuscript.

TJA: Data collection, Analysis, and Writing of the manuscript.