Economic Analysis of Poverty Status of Small-Scale Farmers in Bayelsa State, Nigeria

The study analyzed the household poverty status of small scale farmers in Bayelsa State, Nigeria using a multi-stage random sampling technique to sample six hundred farmers. Data were collected using structured questionnaire and were analyzed using descriptive statistics, FGT [1] index and the logistic regression model. The result revealed that the majority of the farmers 80% were females, while 79% of the respondent was married with 46% of them having no formal education. Twenty-seven (27) percent of the crop farmers are poor while thirtyeight (38) of the livestock farmers were poor. Also, the poverty depth and severity of crop farmers were 0.072 and 0.038 respectively whereas they were 0.098 and 0.052 respectively for the livestock farmers. The logistic regression model revealed that age, educational level, household size, farming experience, farm/herd size, household income, household expenditure and membership of cooperative contributed significantly in determining the poverty status of the farmers. This study therefore recommends measures needed to be put in place to encourage and improve the welfare of the farming household towards productive and sustainable agricultural development for poverty reduction.

the 20 poorest countries in the world. Over 70% of the population is classified as poor, with 35% living in absolute poverty. Poverty is especially severe in the rural areas, where social services and infrastructure are limited or non-existent. Majority of those who live in rural areas are poor and depend on the agriculture for food and income.
The concern on the threat posed by poverty has led the Nigerian government over the years to devote considerable attention to alleviating its scourge through various policy projects and programmes which seems not to have stem the ugly situation till date. In view of these, the question about the poverty status of rural dwellers especially the small scale farmers remained unanswered.
It is on this premise that this study was carried out to answer these questions.

Study Area
The study was conducted in Bayelsa State, Nigeria. It is located between latitude 5 0 00 1 to 10 0 30 1 N and longitude 4 0 55 1 to 6 0 00 1 E and covers an estimated land area of 1,810km 2 with a population of about 856,729 thousand [12].

Sampling Procedure and Sample Size
The sampling involved a multistage random sampling technique.
Firstly, six (6) local government areas were randomly selected using the proportionate sampling method at 75% precision level from the purposively selected three (3) agricultural zones according to Agricultural Development Programme (ADP) structure. In the second stage, ten (10) villages were also randomly selected from the six (6) LGAs each making a total of sixty (60) villages. The third stage involved a simple random selection of five (5) crop and five (5) livestock farmers each from the villages using the list provided by ADP from each of the villages. A total of six hundred (600) respondent farmers were used.

Data Collection
Both primary and secondary data were used for the study. The collection of primary data was achieved using a set of structure questionnaire that was administered by the researcher and trained enumerators complemented with oral interview, information that was collected covered the areas of socio-economic characteristics, farming operations, and income and expenditure patterns. the extremely poor (those whose income was less than one-third of MPCI), the moderately poor (those whose income lies between onethird and two-third of the poverty line) and the non-poor (those whose income was above two-third of the poverty line).
Adult equivalents were generated following Nathan and Lawrence [14], thus:

Logit Regression Model
A binary logistic regression model was used to analyze the determinants of poverty. Thus, poverty is the dependent variable and is determined by independent variables such as socioeconomic characteristics of households and access to services. The L i = log of the odd ratio, which is not only linear in Xi but also linear in the parameters, P i = is the probability of being poor and ranges from 0 to 1.
Z i = the function of the explanatory variables (x) which is expressed explicitly as:     Table 2 shows the summary of the poverty incidence (P 0 ), depth  shows that about 87% of the outcome (Likelihood of being poor)

Factors Influencing Poverty Status of the Respondents
can be explained by the selected independent variables captured in the model (Table 3). The coefficient of age of the farmer was significant and negatively related to the probability of a household becoming poor. This implies that the age of the farmers is a causative factor of poverty.
As age of the farmers increase, the likelihood of being non poor is reduced. This conforms to a priori expectations and work by Ayalneh [15], Obiesesan [16], who opined that older households had greater likelihood of being non-poor. This may be attributed to increased experience and exposure to farming operations and management practices as their age increases.
A positive and significant relationship was found between educational qualification and the likelihood of being non-poor, hence, the higher the educational level, the lower the tendency of been poor. The result is in conformity to a priori expectations and work by Ogwunike [16] who found that a positive significant relationship existed between educational level and the probability of being non-poor. The coefficient of household size was negative and was significant at 1% level. This implies that, the higher the household size, the more likely to become poor. Ceteris paribus.
This could be as a result of the fact that the members of such households would have to depend on the limited resources that is available to the household thereby reducing the per capita income of the household. This is in agreement to a priori expectations and work by Khan [5] and Ogwumike [16].
A positive and significant relationship was found between farming experience and the likelihood of being non poor at 5% level. This implies that the higher the years of farming, the higher the probability of being non-poor. This is in conformity to a priori expectations and work by Omonona [10] who stated that exposures and experiences gathered over the years help rural poor people to fight poverty. The author further opined that experience in farming help to reduce losses thereby encouraging proper handling and management of relatively scarce resources. There was a positive and significant relationship between farm/herd size and the likelihood of being non-poor. This implies that as the farm/herd size of the farmer increases, the probability of the household being nonpoor is increased. This finding conforms to a priori expectations and work by Eneyew [17] and Alemu [18] who found that a unit increase in land holding increased the probability of being nonpoor. The coefficient of household income was significant at 1% level and positively related. This implies that as the household income increase, the probability of being non-poor increases. This is in agreement to a priori expectations and work by Alemu [18] who found a positive relationship between household income and the likelihood of being non-poor.
In conformity to a priori expectations, the coefficient of household expenditure was negative and significant at 5% level.
This indicated that, the higher the household expenditure, the lower the likelihood of being non-poor. Ogwumike [19] stated that, excessive expenditure by household head is a pointer to poverty.
The coefficient of membership of cooperative was positive and significant at 5% level. This implies that, if a household head is a member of cooperative, the likelihood of being non-poor increases.
This will not be unconnected with the fact that members of cooperative in the rural settings help their cooperative members in time of needs and also provide incentive and loan facilities to those in need.

Conclusion
The research has shown that, the incidence, depth and severity of poverty were high among the farming households even though some of the farmers fall above the poverty line. The study has also shown that the rate of poverty is relatively higher among livestock farmers compared to crop farmers. Meanwhile, the study has revealed that several factors influences the poverty status of the farming households such as age, educational level, household size, farming experience, farm/herd size, household income, household expenditure and membership of cooperative [20].
Given these findings, therefore, it is recommended that: