Oil Production and the Poverty Situation in the Niger Delta Region of Nigeria: A Case of Selected Communities

When the oil producing communities relinquished their land for oil production, they did so expecting direct benefits, which may be summarized as welfare improvement. But unfortunately, rather than celebrate the arrival of these benefits, there were complaints from the communities about damage to the environment, which impacted their means of livelihood negatively. It is therefore to investigate these claims that this study examines the effects of oil production on household income generation and well-being in selected oil producing communities in the Niger Delta Region of Nigeria. To accomplish this task, survey approach was employed on account of which the three (3) participating States; four (4) Local Government Areas; and five (5) communities were selected with the aid of purposive sampling technique. Also, the six hundred and fifty-eight (658) households, which constituted the sampled objects of study were drawn with the employment of random sampling technique. The generated data was therefore analyzed with the aid of descriptive and logistic regression tools. The result of the analysis revealed that oil exploration and exploitation activities, truly impacted negatively on the income capacity and the well-being of households in the oil-bearing communities. Hence, something significant needs to be done in order to reverse the trend. The study recommends that policies that would ensure immediate clean-up of pollution sites, and adequate compensation plan both in the case of pollution and land take-over be put in place. These would ensure sociallyoptimal operations by the firms that would guarantee the protection of community dwellers who have always been at the receiving end of the negative effect of oil production.


Introduction
Since oil production commenced in 1958, Nigeria, over the years has risen from the pre-1956 state of non-oil producing nation to the highest-ranking oil producing nation in Africa and the thirteenth oil producing nation in the world (see Tables 1 and 2) with the production of 2.525 million barrels per day. Prior to the discovery and exploitation of oil, agriculture was the main contributor to the country's Gross Domestic Product (GDP). With the bullish entry of oil into the international market, coupled with increasing demand, the price of oil experienced unprecedented rise. Oil earnings now overtook that of agriculture, forcing government to divert its investment attention to oil. This was how oil gradually took over from agriculture as the major contributor to government revenue earning. However, the experience of the Niger Delta region does not seem to resemble the story of Nigeria given the fact that the wealth from oil is yet to have significant effect on the well-being of the people. In fact, abject poverty seems to be the identity of the region. It is revealed that the incidence of poverty has remained high for the Niger Delta region despite its oil wealth.
The findings by the World Bank (2000) that out of the world's 6.0 billion people, 2.8 billion live on less than $2 a day and 1.2 billion on less than $1 a day strongly support the position that poverty is a global problem, hence the need for everybody to be involved in tackling it. But the poverty problem is more touching when a set of people living under the condition are perceived to be suffering in error (or suffering in the midst of plenty). This is generally the case of Nigeria and that of the Niger Delta Region, in particular, hence the need for this study. According to Thomas and Canagarajah, (2002), Nigeria had the largest export income (after South Africa) in 1994 thus showing the wealth of the

The Study Area
This study covered selected communities involved in the production of oil from the inception of oil discovery at Oloibiri in the present day Bayelsa State of Nigeria. As a result, core communities involved in oil production were examined irrespective of whether they still have oil or not. For instance, oil is no longer being produced in Oloibiri, given the fact that the oil wells in the community dried up already. But the impact of oil production such as oil spill and acid rain are still being felt just like in many other non-oil producing communities in the region. Thus, knowing what became of such communities after producing oil for a reasonable period of time would go a long way in influencing the way forward in managing present and future resourcerich communities.
Following from this position, therefore, the study cover: Ibekwe Clan

Literature Review
Poverty is a condition that is unacceptable and so needs to be aggressively addressed for the purpose of improving the people's welfare. This was the view of the Asian Development Bank (1999) when it asserted that: Poverty is an unacceptable human condition that can and must be eliminated by public policy and action. Poverty is a deprivation of minimum essential assets and opportunities to which every human being is entitled. Everyone should have access to basic education and primary health services. Poor households have the right to sustain themselves by their labour, and be reasonably rewarded, and be afforded some protection from external shocks.
The implication of this submission is that poverty is a menace that should be confronted by all. But given the fact that poverty is multi-dimensional and multi-facetted, the strategy adoptable in addressing it per time would depend upon the depth of understanding of its root causes. These causes are identifiable through the various poverty theories.
From the works of Bradshaw (2006), five theories of poverty were identified. They include: (i) individual deficiency theory, (ii) cultural belief system theory, (iii) economic, political, and social distortions or discrimination theory, (iv) Geographical disparity theory and (v) Cumulative and cyclical interdependency theory. Understanding these theories provides the basic information required in tackling the ensuing poverty.
(a) Individual Deficiency Theory The origin of this theory, which views poverty as a result of the inability of man to meet his own basic needs, is traceable to the work of Maslow (1943) where he maintains that until the most basic level of needs are met; human beings would lack the motivation to move to higher level of needs. In essence, if an individual could work harder to meet the basic needs, he would possess the required drive to address the higher level of needs. But if an individual could not meet the basic needs he would remain at the same level.
With this submission, the theorists hold the view that individuals are responsible for their poverty situation. They claim that the poor individuals could have avoided the situation in which they are found if they had worked harder. This was the position of Asen (2002) when he opined that any individual can succeed by skills and hard work. Hence, it is believed that individuals fail because they do not try hard enough (Bettinger, 1977). Although it is possible to dismiss the individual deficiency theory as an apology for social inequality (Fischer et al 1996) we can easily see how it is embraced in anti-poverty policy, which suggests that penalties and incentives can change behavior (Bradshaw, 2006). Addressing poverty, therefore, requires a focus on how people are able to put to use the natural endowment within and around them. This is what makes this relevant and important to this study. In actual fact, the poverty situation in oil producing communities may be traceable to their incapacitation due to the prevailing property right regime that may not allow them control over their own resources.
(b) Cultural Belief System Theory Culture, which is a way of life of certain people, represents their beliefs, values and skills which are perpetuated over time. It is the system by which human beings perceive and understand the world in which they live (see Gertz, 1973). But the cultural theory of poverty, which portrays the creation and transmission of poverty over generations, is traceable to the work of Lewis (1966). According to Lewis (1969) "this culture emerged when populations that were socially and economically marginalized from a capitalist society developed patterns of behavior to deal with their low status. This behavior was characterized by low aspirations, political apathy, helplessness, disorganization, provincialism, and the disparagement of so-called middle-class values." He further argued that once this culture was in place, mechanisms for its perpetuation were immediately developed, changes in structural conditions notwithstanding (Ryan, 1976).
There have however been some varying levels of confusion on the relationship between culture and poverty. This confusion emanated from the work of Lewis, who himself happened to be among the pioneer scholars on the subject. As presented by Kartz (1989), Lewis's popularization of the culture of poverty concept tolled an intellectual death knell to the optimistic idealism of the mid-1960s that advocated poverty eradication through direct state intervention. So, for Lewis to support the advocates of state intervention as a panacea for poverty eradication negates the principles that culture of poverty upholds (see Lewis 1966b and1967). But the confusion notwithstanding, there has been significant effort (mainly by cultural sociologists) at establishing the link between culture and poverty. These efforts have yielded some position on how cultural factors do affect and are affected by poverty. For instance, scholars like Snow and Benford (1992) and Harding (2005) viewed culture as frames and how they influence the response to poverty. The frames are interpretive schema that simplify and condense the world out there by selectively punctuating and encoding objects, situations, events, experiences, and sequences of actions within one's present or past environment. Their influence on poverty is determined by how individuals frame their neighborhoods. For instance, Harding (focusing on how adolescents framed the idea of pregnancy) found out those adolescents in disadvantaged neighborhoods in Boston exhibit greater heterogeneity of frames than those in other neighborhoods. This is so because perception is a function of the group one belongs to.
In her presentation Swidler (1986) conceived culture as a repertoire (tool kit) of practices, beliefs, and attitudes from which strategies of action are constructed whenever the need arises. According to her, culture is a heterogeneous set of attributes rather than a single coherent system, thus allowing for cultural differentiation and contradictions within a group. Thus, following the attributes in each individual kit of tools, it is easy to understand how they respond to their aspirations. This may help to explain, for instance, why although the poor and the middle class may share common aspirations about education, their overall pattern of behavior in pursuing the goal differs. While the works of Wilson (1996), and Hays (2003) supported and extended the poverty tool kit idea, Lamont (1992) argues that the tool-kit perspective presents problems in the sense that it failed to explain why some certain repertoire choices are made, among other alternatives. But the criticism notwithstanding, Swidler's submission was able to deal with the issue of heterogeneity or contradiction, thus pushing the cause of poverty explanation further.
Consequently, the prevalence of resource crisis is a pointer to the fact that the people have developed the culture of nonacceptance of the status quo, hence the perpetuation of agitation over time for resource control. Where one hundred percent control of the resources may not be possible, they have agitated for a fair share of the resources in order to make life better for the people.
(c) Economic, Political, and Social Distortions Theory The development of this theory is traceable to the work of Karl Marx and the 19 th century social intellectuals as cited in the work of Bradshaw (2006). Unlike the earlier presented theories where individuals are considered to be the source of poverty, this one considers the economic, political and the social system to be responsible for poverty. Theorists here are of the opinion that it is the system that restricts or confines the people to having limited opportunities and resources and hence making it difficult for them to achieve required level of income for minimum well-being. This was the position that was summarized in Bradshaw (2006) when he submitted that The 19 th century social intellectuals developed a full attack on the individual theory of poverty by exploring how social and economic systems overrode and created individual poverty situations. For example, Marx showed how the economic system of capitalism created the "reserve army of the unemployed" as a conscientious strategy to keep wages low. Later Durkheim showed that even the most personal of actions (suicide) was in fact mediated by social systems. Discrimination was separated from skill in one after another area, defining opportunity as socially mediated. Taken to an extreme, radical thinkers argued that the system was flawed and should be radically transformed.
As observed and submitted by the author, much of the literature on poverty suggests that the economic system is structured in such a way that poor people always lag behind regardless of how competent they may be.
From the presentations above, the problem of workers, which help in perpetuating poverty, can be viewed as wage problem. This was supported by Jencks (1996) when he affirmed the problem to be the inability of the minimum wages allowing families to be self-sufficient. It is further compounded by the inability of families to get better jobs since available jobs are limited (see Tobin 1994). From this perspective, therefore, poverty problem can be addressed by positively tuning the economy in such a way that better conditions of service and access to resource base would be available to enable workers/families to enjoy some level of comfort (at least minimum). The resource ownership/control problems in the Niger Delta could equally be viewed and addressed within this context.
(d) Geographical Disparities Theory This theory is traceable to the works of Rostow (1960, and1962) where he identified the five stages of growth and development. According to him, all countries are at different stages of development due to a number of factors such as: i. Natural resource endowment ii. High level of indebtedness iii. Corruption, and iv. Exploitation of poorer countries by More Developed Countries (MDCs), which makes it impossible for them to develop and thereby perpetuating their poverty These factors deal with spatial characterization of poverty, that is, the spread of poverty across geographical planes such as rural poverty, urban poverty, and third world poverty, among others. The theory calls attention to the fact that people, institutions, and cultures in certain areas lack the objective resources needed to generate well-being and income, and that they lack the power to claim redistribution (Bradshaw, 2006). According to Morrill and Wohlenberg (1971) some of the factors responsible for spatial poverty could include disinvestment, proximity to natural resources, density, and diffusion of innovation. Also, Goldsmith and Blakely (1992), in their work on the link between development and poverty in urban contexts, argue that the joint processes of movement of households and jobs away from poor areas in central cities and rural regions creates a separation of work, residence, economic, social and political life. These processes are multiplied by racism and political indifference of the localities in which they flourish (see Bradshaw, 2006).
But efforts at explaining the intensity of poverty in certain areas have been on over time and a number of scholars have attempted to offer some explanation. However, it is important to note that the explanation is better done by utilizing input from other theories. Those theories that have been found helpful therefore include: economic agglomeration theory; human ecology theory; and selective out-migration theory.
Economic agglomeration theory is used to explain the emergence of strong industrial clusters, which eventually draw resources away from poor areas, thereby making them poorer. This is explained by Bradshaw, King, and Wahlstrom (1998), and colligated in Bradshaw (2006) that Agglomeration shows how propinquity of similar firms attracts supportive services and markets, which further attracts more firms. In reverse, the propinquity of poverty and the conditions leading to poverty or the consequences of poverty (crime and inadequate social services) generate more poverty, while competitive areas attract business clusters, drawing away from impoverished communities. Low housing prices in such locations may attract more poor persons, for example, leading to housing disinvestment by building owners. In a world in which the criteria for investment are "location, location, location," it is not unreasonable to track investment going to neighborhoods, communities and regions in which there is already substantial investment, while leaving less attractive areas.
Human ecology theorists hold that technological development would flourish faster in urban areas given the availability of infrastructural facilities, which aid the development of human resources (see Hansen, 1970; Rural Sociological Society Task Force on Persistent Poverty, 1990; and Lyson and Falk, 1992), thus explaining why the rural areas would remain poor while the urban areas are getting better.
On their own part, the selective out-migration theorists have opined that those who migrated from rural areas are the best materials in terms of education, skill, and widest world view. As such, the exit of these people only keeps the areas more impoverished, while improving the lots of the urban areas (Wilson, 1987;and Bradshaw and Muller, 2003).
These presentations are relevant in examining the situation in resource-rich communities, which may be suffering high level poverty despite the fact that they are naturally endowed with natural resources from which their nation derives her wealth.
(e) Cumulative and Cyclical Interdependencies Theory This theory, which stems from the work of Myrdal (1957) on "interlocking, circular, interdependence within a process of cumulative causation", is more complex than the ones earlier discussed as it builds on components of each of the other theories. It is concerned with individuals and their communities as caught in a spiral of opportunity and problems, and that once problems dominate, other opportunities are shut-out, thereby creating cumulative set of problems that make any positive intervention almost impossible (Bradshaw, 2000). The cyclical nature is as a result of the fact that the individuals and the communities are mutually dependent in the sense that when the community is economically weak, the individuals that are members of the community are equally weak, and vice versa. In actual fact, economically strong community members would be able to contribute reasonably, either in the form of tax or other form of investment, for the well-being of the community. Thus, the moment a cycle of decline is ignited, the poverty engendered by the interdependence of factors grows at a terrific progression.

Method of Study
Given the focus of this study, which is about investigating the extent oil producing activities in the Niger Delta region have affected the well-being of selected oil producing communities, survey approach based on participatory research methods is adopted. This is following the works of Hall (1975); Freire (1982); Horton and Freire (1990); and Chambers (2008). Using the tools of direct observation, questionnaires were administered on sampled communities and household in those communities. To this end, while purposive sampling method was employed to determine the communities included in the study, simple random sampling techniques were adopted in the administration of the structured questionnaires on the respondents from the selected communities. This was to ensure the study is focused and the generated data are adequately representative of the population. The ensuing data was analyzed using descriptive approach involving averages/charts and Logit models. Detailed methodology is presented as follows:

Research Design
Given the nature of the study, while some secondary information from various sources were of immense benefit to the success of this work, the main pool of information was procured directly from the field. As a result, a combination of survey and observation techniques was adopted. By the survey technique, structured questionnaire was administered on selected household and community leaders across the communities identified for this study

Population of Study
For the prosecution of this work it is imperative to clarify the issue of study area population and that of target population/sampling frame. While the study area population represents the total number of male and female (including adult and children) residing within the geographical area, the target population/sampling frame represents a group of listed households in the selected communities from which samples are drawn. But a household is a family unit consisting of a man and his wife/wives, their children and other dependent relatives living together under one roof. So, in determining target population/sampling frame household numbering/listing exercise was conducted. It is on the number of listed/numbered households that the sample size is computed and study samples generated.
Following from above, it would become clear that the study/target population, which is the same as the sample frame is the population of households (that is, the number of listed households). Focusing on the household/family units is considered important for this study because in a simple macroeconomic model, the workings of an economy is Region of Nigeria: A Case of Selected Communities measured through the interaction between the household and the business sectors. In fact, the household sector from the traditional to the modern economic system has remained the most important sector that supplies the most basic input required for daily economic operations. Hence, the welfare of the sector is crucial for the welfare of the macro economy. This is why studying the household units is fundamental and very important in studying the performance of the economy. This is the reason for focusing on the household units in order to measure the well-being impact of oil production in the Niger Delta region.

Sample and Sampling Procedure
A combination of purposive and random sampling techniques were used to select the samples studied. Since the purpose of the study is to investigate the socio-economic impact of oil production on selected communities, it is imperative that communities within the Niger Delta region that are involved one way or the other in the production of oil are selected. On account of this, purposive sampling technique was adopted in selecting the five communities involved in the study. Thereafter, random sampling technique was employed to select the households identified for survey administration.
Following from the above, after numbering/listing the households, the sample size was determined based on the work of Krejcie and Morgan (1970), which provided for the computation of the sample size thus: Where: n = sample size = Chi-Square for the specified confidence level at 1 degree of freedom N = Population size P = Population Proportion (.50 for Table in "Appendix B") ME = Desired Margin of Error (expressed as a proportion) When the above formula was applied, the sample size was computed to be 658 households. This computation is presented in Table 3: Table 3. Study sample size. Kpean  94  80  Oloibiri  117  108  Mkpanak  169  132  Upenekang  294  169  Ibekwe  285  169  Total  959  658 The survey administration therefore involved completion of 658 household questionnaires.

Method of Analysis and Model Specification
Given the nature of data generated from the survey exercise, the core analytical tool was logit/probit regression . Generally, the logit/probit models are developed following the works of Cornfield (1951 and1956), and Stigler (1986). It may be presented generally as * = ∑ + Where y* is an unobserved, continuous, underlying tendency behind the observed ordinal response (rating). The Xk represent the independent variables, while the βk represent the associated parameters. The error term (εk) captures stochastic (unobserved) variation, assumed to be distributed logistically for logit models and distributed normally for probit models.
Specifically, therefore, the logit model is presented as: while the probit model is presented as: The logit/probit model is very useful in handling dichotomous and polytomous data, which is the nature of data we generated from the survey. Following from this position therefore, the analytical models employable in this study are logit and they are specified as follows: i) Living Condition Analysis The living condition of the people was investigated based on the model presented below ' ( ! = ) = * + * +,-+ * . /0, + 1 ! (5) Where the equation expresses the fact that living house type (the dependent variable) is influenced by the type of occupation and income being earned by the households in the communities.
ii) Life Sustainability Analysis: The study also investigated how the distance to health facility (DTHF) and drinking water quality (DWQ) determines the sustainability of lives in the communities by employing Logit/Probit model thus: Where the dependent variable is a dichotomous variable stating whether the households have lost family members or not in the communities. So, the model investigated the extent to which losing of family members could have been influenced by the distance to health facility and the quality of drinking water.

Stylized Facts on Oil Production in Nigeria
Detailed information on the oil driving capacity of the Nigerian economy vis-à-vis the welfare status of the oil producing communities are presented here. It is important to note that following the discovery and production of oil in commercial quantities, oil revenue displaced agriculture and took the prime position as the major contributor to government revenue earning as presented in Table 4 and Figure 2. According to the information elicited from the

Preliminaries
To assess the impact of oil production on the welfare of oil producing communities, how the income generating capacity of the people has been affected was examined, thus leading to being able to explain their living condition. Availability of basic life support infrastructure such as health facility and portable water was also examined to help explain the vulnerability of the people. The result of these analyses are presented here:

Living Condition Analysis
Analyzing the well-being of the people involves an assessment of their capacity to maintain a basic standard of living. Before the advent of oil, the people of the Niger Delta region were predominantly farmers and fisherfolks, the trade that had supported them adequately well over the years. But with the coming of oil production and its negative effect on land and water resources, farming and fishing started losing their attraction as people began to queue for government jobs and other means of income generation. A look at Table 5 and Figure 3 confirms this as it reveals that only 60 and 46 households representing approximately 9% and 7% of our sample are still involved in farming and fishing during the period under study. The rest are civil servants (25.99%), artisans (7.14%), motorcyclists (8.96%), taxi drivers (8.36%), traders (19%), and other workers (14.44%).
Having examined the occupational distribution, the income and expenditure pattern of the people were considered with a view to determining the expenditure and income poverty line. This helped to determine the relative percentage of the population in poverty. Figures 4 to 7 showed the distribution of income and expenditure in the communities. The distribution pattern shows that a reasonable proportion of the population clustered around the low-income region. This is explained by the peakness of the curves and its longer tail to the right. But the expenditure distribution is more normal given the fact that some of the people spent more than they earn given the fact that they have access to more funds from credit granting such as cooperative societies, among others.      The implication from this presentation is a depiction of high income inequality, which is buttressed in Table 6 and Figure 8  The Gini index of 0.409441 for the selected communities (as computed) revealed that there was income inequality of 81.89% among the households in the communities under study. This value of Gini Coefficient indicated a wider gap between the low and the high-income households in the communities. It implies that a smaller proportion of the population controls larger proportion of the available resources.
The decomposition from the Lorenz curve analysis (as presented in Table 7) revealed that while the lower 50% of the population controlled 20% of total income earned by the households in the communities (amounting to 0.4% share by 1% of the population), the upper 50% controlled 80% of total income accruable to all the households (amounting to 1.6% share by 1% of the population) in the communities. Also, while the lower 10% and 20% controlled 1% and 4% of total income respectively (amounting to 0.1% and 0.2% share by 1% of the population), the upper 10% and 20% controlled 29% and 45% respectively (amounting to 2.9% and 2.25% share by 1% of the population). The inequality may not have been that high if under the derivation principle there was provision to compensate households in the oil producing communities who were displaced from their original and traditional sources of income following the production of oil, whose situations were compounded by the sufferings inflicted on them generally by the negative effects of oil production. But because this was not done their earning capacity grossly dwindled, which impacted negatively on their well-being. This is buttressed by the poverty threshold information presented in Table 8 and 9. As revealed by Table 8, while the core poverty threshold (median) for the communities under study was N142560 per annum, the core poverty threshold (mean) was N207962.73. On the other hand, the moderate poverty thresholds for the communities are N285120 (Median) and N 415925.47 (Mean). Following from these positions, while the core population in poverty was 23.86% (Median) and 29.94% (Mean); the moderate population in poverty was 13.68% (Median) and 17.02% (Mean). The total population in poverty for the communities under study therefore summed up to 37.54 (Median) and 46.96% (Mean).
But while the overall statistics shows that the population in poverty is generally below 50% when the Mean Poverty threshold is employed, a look at the specific statistics for each of the communities revealed varying degrees of poverty. For instance, while for Ibeno LGA the poverty rate is 36.51% (Median) and 62.24% (Mean); it is 18.44% (Median) and 53.9% (Mean) for Kpean LGA. It is also 40.71% (Median) and 49.07% (Mean) for Ogbia LGA.
Although the National Bureau of Statistics (NBS) uses the Mean threshold to determine poverty rate (see NBS, 2010: 10), the most popular practice now is the employment of the Median threshold (see Gordon, 2006). Given the fact that the Median is the point that divides the distribution into two equal parts, any income award aimed at lifting the poor out of poverty would not raise the Median but would definitely raise the Mean of the distribution, implying that the threshold would also keep changing. As a result, it would be difficult to measure the effect of any intervention aimed at raising people out of poverty, since the poverty level itself would be changing with the Mean threshold poverty rate determinant.
But poverty is not only measured from income perspective, it is also measured from expenditure perspective. To this end, therefore, the expenditure threshold is presented in Table 9. From the result presented in Table 9, while the overall core expenditure poverty threshold for the communities was N412.27 and N484.24 per day for Median and Mean threshold respectively, it was N824.55 and 968.49 for the Moderate Expenditure poverty threshold. The implication was that, on the average those who could not afford to spend the threshold amount per day are in poverty. Following from this analysis, 13.37% and 16.72% were core poor from the Median Mean perspective while 17.48% and 19.15% were moderately poor, bringing the total percentage of population in expenditure poverty to 30.85% and 35.87% respectively.
Specifically, however, the analysis revealed that the poverty rate was higher in Ibeno and Kpean LGAs when compared with other areas. This is given the fact that from the Median and the Mean perspective as presented, while the poverty rate for Ibeno was 34.44% and 43.15%, it was 25.53% and 43.26% for Kpean. Consequent upon the presentation of both the income and expenditure thresholds, it is evident that the poverty level in those communities is relatively high. Exploring the living conditions further, the result of living house type and life sustainability analysis are presented as follows: (a) Living House Type (LHOT) Analysis: The multinomial logit result is presented in Tables 10 and 11.  Before conducting a detailed discussion of the result, it is important to underscore the fact that multinomial logistic regression, on a general note employs maximum likelihood estimation approach, which is an iterative procedure. While the first iteration is called iteration 0 because it is the maximum likelihood of the empty or null model with no predictors, the next iteration has the predictors included in the model. At each iteration the maximum likelihood diminishes in order to fulfill the goal, which is the minimization of the log likelihood. The model is said to have converged when the difference between successive iterations becomes smallest. It is at this point the iteration stops and results are displayed.
From the results displayed therefore, the log likelihood converged at the point -408.79961 with the LR Chi Square (2) = 30.49 being the likelihood ratio Chi Square test indicating that for both equations, mud house and thatch house relative to cement house, at least one of the predictors is not equal to zero. This is confirmed by the probability of the Chi Square. The small value of Prob>Chi 2 = 0.0000 would, in fact, lead support the conclusion that at least one of the regression coefficients in the model is not equal to zero.
From the results, the Living House Type (LHOT) is the response variable in the multinomial logistic regression, and underneath it are two replicates of the predictor variables representing the two models that are estimated: mud house relative to cement house, and thatch house relative to cement house. So, in this analysis, cement house is the base outcome. There are two levels of reference points because every multinomial logit models estimates k-1 models setting one as the reference point, where k is the number of levels of dependent variable. It would be observed that the dependent variable has three levels, which are: Mud house, Thatch house, and Cement house. So, by default, the program chose cement house as the base outcome, hence the estimation of a model for mud house relative to cement house and another model for thatch house relative to cement house.
According to the parameter estimates in the mud house model, the occupation coefficient of -0.791 indicated that if a better change of job occurred, which would result in households moving one point up, the multinomial log-odds for mud house relative to cement house would be expected to decrease by 0.791 units while holding other variables in the model constant. Although testing at 5% significance level it would be concluded that the coefficient is not statistically different from zero in this model, its sign is a strong point that holds the future strength of the variable. But in any case, if the significance level were to be set at 10%, for instance, our decision would have been different given the fact that its p>/z/ is 0.071. Looking at the income coefficient of -7.85e-07, it implied that with an increase in income by one unit the multinomial log-odds for mud house would drop by 7.85e-07 units while all other variables in the model are held constant. What this means in summary is that with better job and enhanced income the households on the average would prefer to live in a cement house instead of mud house. When tested at the 5% significance level, with z of -2.91 and p > /z/ of 0.004, it was concluded that the coefficient is statistically different from zero. So, income is a strong variable determining the type of house people would like to live in.
The next item is the multinomial logit estimate for mud house relative to cement house when the predictor variables are evaluated at zero (the constant). The coefficient (-0.818) of this variable revealed that the log-odds of preferring mud house to cement house would drop even without occupation and income in the model. This implied that ordinarily, the people would prefer to live in cement house even when they have neither job nor income. Thus, their living in mud house is merely a condition they cannot help at the moment. Testing this at the 5% significance level, the z of -3.27 and p >/z/ of 0.001 indicated that the coefficient of the variable is statistically different from zero, meaning that, all things being equal, the people would prefer to live in a cement instead of mud house.
Looking at the thatch house model, with the coefficient of -0.034, it was clear that the availability of a better job would result in households moving one point up in their desire for cement house. This is because the multinomial log-odds for thatch house relative to cement house is expected to decrease by 0.034 units while holding other variables in the model constant. Also, the income coefficient of -2.13e-06 connotes that with an increase in income by one unit the multinomial log-odds for thatch house would drop by 2.13e-06 while all other variables are held constant in the model. However, when tested at 5% significance level the conclusion is drawn that while the coefficient of occupation was not statistically different from zero (given z = -0.48 and p > /z/ = 0.631) in this model, the coefficient of income is statistically different from zero (given z = -3.11 and p > /z/ = 0.002). This implied that while the coefficient of occupation in this model is not statistically significant to drive living house preference, the coefficient of income is statistically significant. But on the whole going by the sign of the coefficients, the two variables have the potentials at driving the preference between thatch house and cement house. The implication of this finding was that with better job and enhanced income the households on the average would prefer to live in a cement house instead of thatch house.
Considering the multinomial logit estimate for thatch house relative to cement house when all the predictor variables are evaluated at zero (the constant). Its coefficient (-1.684) indicates that the log-odds of preferring thatch house to cement house would drop even without occupation and income in the model. This implied that ordinarily, the people would prefer to live in cement house even when they have neither job nor income. So, their living in thatch house was merely a condition they could not help at the moment. Testing this at the 5% significance level, the z of -3.87 and p >/z/ of 0.000 indicated that the coefficient of the variable was statistically different from zero, meaning that, all things being equal, the people would prefer to live in a cement instead of a thatch house.
One very important component of the logit analysis is the relative risk ratio (r r r), which is computed by exponentiating the multinomial logit coefficients (e coeff ). The main aim behind this is to measure the relative risk for preferring one variable to another. To this end, for the living house type (LHOT) analysis, the relative risk associated with preferring mud house to cement house on the one hand; and thatch house to cement house on the other, were obtained. This result is as presented in Tables 12. Looking at the Mud House Model, a relative risk ratio of 0.924 associated with occupation indicates that with a one unit change in occupation (or if occupation were to get better) the relative risk for mud house relative to cement house is expected to rise by a factor of 0.924 given that other variables in the model are held constant. This implied that if occupation were to get better, the people are likely to prefer a cement house to a mud house. Also, a relative risk ratio of approximately 1.0 associated with income implied that with a one unit change in income, the relative risk for mud house compared to cement house is expected to rise by a factor of 1.0 (approx.) given that other variables in the model are held constant. This means that if income gets better, the people are likely to prefer a cement house to a mud house.
With the Thatch House Model, the situation is approximately the same with the Mud House Model. This was because while the risk factor associated with occupation is 0.966, the factor associated with income is approximately 1.0. Hence, on a general note, the relative risk ratio implied that households in the communities under study would prefer cement house to a thatch house if occupation or income should change positively.
(b) Life Sustainability Analysis: The logit result for life sustainability is presented here in Tables 13 to 15   As could be seen from the results above, there were two effects in our model, which were Distance to Health Facility (DTHF), and Drinking Water Quality (DWQ). Looking at the goodness of fit tests of the model (such as LR chi 2 (2) = 65. 27 and Prob >chi 2 = 0.000), it is visible that the model is well fitted to our data. This implied that one or more of the two effects in the model is important for predicting the probability of not losing a family member. The tests for parameters suggested that each of the effects in the model is significant at the 5% significance level. Given the fitted model presented as equation 7.
)89 : : = −3.627 + 0.00152345 + 0.881267 (7) We can say that for a one unit change in the distance to health facilities the log odds of losing a family member is 0.0015 with other variables in the model held constant. In other words, as the distance to health facility increases, for instance, the log odds of losing a family member rises by 0.0015. Also, the coefficient of drinking water quality variable (which was ranked from Excellent = 1 down to Poor = 4) indicated that with a one unit change in drinking water quality along the scale where moving up the scale connotes a drop in the quality of drinking water, the log odds of not losing a family member is 0.881 with other variables in the model held constant.
From the information on the odds ratio, it is evident, for instance, that with a one unit change in the distance to health facility (DTHF), the log odds of losing a family member would change by 1.001473. On the other hand, the odd ratio of the second predictor variable -drinking water quality (DWQ) indicates that a one unit change in the quality of drinking water would change the log odds of losing a family member by 2.414. These findings further revealed that while the two predictors have the capacity to predict changes in the log odds of losing a family member, drinking water quality is a stronger predictor of changes in the dependent variable.

Summary and Recommendations
As shown from the presentation, the growth of the Nigerian economy is attributable to the growth of oil earnings given the higher contribution of oil to total revenue. Ironically, however, the poor condition of the Niger Delta people can also be greatly attributed to oil production. This is because following the deprivation of their means of livelihood due to oil production, their living condition, which is determined by their occupation and income generating capacity including access to basic infrastructural facilities such as health and portable water has been negatively affected. This has exacerbated the problem of poverty and inequality in the communities and region as a whole, as reflected in the outcome of the analysis. Thus, the people are living in thatch and mud houses at the moment, for instance, because of their income level, which is determined by what their occupation is able to offer them. A better alternative would be preferred if accessible.
Following the findings in this study, therefore, the following recommendations are offered: 1. That given the fact that the poverty situation in the Niger Delta is traceable to the erosion of the people's means of livelihood following the degradation of the environment as a result of oil production, thereby impacting negatively on their income generating activities, a modification of natural resource exploration and exploitation law is essential. This is to ensure adequate compensation for land take-over, spills and general pollution arising from oil production. It is expected that when this is in place, the arising poverty problem stemming from losses of rivers and farmlands on which the income generating activities of the people depend would have been addressed to a reasonable extent. 2. That given the fact that the availability of basic infrastructures such as potable water and health facilities has been grossly inadequate, both government and the oil companies should take practical steps to ensure the availability of potable water and reasonably equipped hospitals in the communities with the right staffing. This would enable the people to have access to health facilities whenever the need arises. This was one of their cardinal expectations while welcoming the production of oil in their communities.

Conclusion
The finding in this study (which extracts substantially from the PhD thesis by Orebiyi, 2016) justifies the fact that the people's poverty condition has been perpetuated because after their natural source of livelihood (farming and fishing) was rendered highly unproductive, there was no alternative provision to cushion the effect of the deprivation. This is the core of the Niger Delta problem and it requires proactive attention from all concerned stakeholders.