Economic Inequality and Poverty Dynamics: What does Literature tell us?


1.1. Study Background

The dynamics of economic inequality and poverty is a concept with a timeline attached (Fosu, 2017; Marrero and Serven, 2018). Inequality and poverty are significant international development challenges and social challenges that affect many countries. Economic inequality is understood as the income and wealth disparity between the rich and the poor. While some literature report that greater inequality is detrimental to poverty, others say that inequality helps reduce poverty's incidence since inequality and poverty often move together. These are evidence that the relationship between poverty and inequality is not direct (Beteille, 2003) but two distinct concepts independent of each other. Inequality presents itself in the form of financial disparities, gender disparities, unequal opportunities, and living standards, such as health and energy disparities (Brunori et al., 2019; Ramosa et al., 2020). In Nigeria, poverty and inequality remain a mystery (Kolawole and Omobitan, 2015). The author explains this to imply that there is poverty in the midst of plenty and inequality in the face of economic prosperity. Many authors, among others, have used household consumption per capita as a representation of poverty (Garza-Rodriguez, 2018; Ho and Iyke, 2018). There has been further development of existing methods in the literature (Bigsten and Shimeles, 2008; Adepoju, 2012; Katsushi and Jing, 2014; Akerele et al., 2017; Obayelu et al., 2021) on the application of dynamic models to the study of the dynamics of poverty and inequality.

1.2. Objectives of the study

The primary goal of this study is to review the literature on the relationship between economic inequality and poverty dynamics. Specifically, it identifies various approaches to measure inequality and poverty dynamics, data requirements and limitations, factors driving inequality, the process in which a household enters, exits or remains in poverty for a prolonged period of time, provides empirical evidence of the power and stakeholders’ mapping matrix that influence inequality and poverty transition using the case of Nigeria.

1.3. Motivation and significance of the study

A reliable measurement of inequality and poverty dynamics and their drivers is necessary for proper decision-making. It should give a clear picture of the situation for decision-makers to be able to pinpoint essential intervention areas. These measurements must be technically sound and compatible with practical issues and policy recommendations. However, there is still a knowledge gap on the nature of this phenomenon.

of the relationship between poverty and economic inequality. Due to a lack of longitudinal data that tracks and collects valuable information from individuals and households, very few studies, such as Krishna (2004), Bigsten, and Shimeles (2008), were found to have been undertaken on inequality or poverty dynamics in developing nations, particularly in Sub-Saharan Africa.

One cannot move beyond an investigation of the correlates of poverty using static observations (without a time dimension). The study of poverty dynamics allows researchers to have an improved understanding of the causes of poverty. Poverty rates vary from year to year, and even though overall poverty rates remain the same, the proportion of the population that is susceptible to poverty or who is only poor on occasion could be significantly higher. According to Adeleye et al. (2020), income inequality is a critical factor in poverty.

Furthermore, previous empirical studies such as Stoyanova and Tonkin (2018) focused solely on the association between economic growth and poverty reduction, growth and inequality, and the causal association concerning poverty dynamics and economic growth. Based on the high level of income mobility in many developing countries, such as Nigeria, the knowledge gap on poverty dynamics caused by insufficient dynamic data is particularly concerning. On multiple levels, our work adds to the existing empirical literature. The findings of this study will contribute uniquely to the argument on growth, poverty, and inequality about how to 'end poverty in all its forms everywhere' and how to 'reduce inequality within and between countries based on the first and tenth Sustainable Development Goals for the year 2030. Some previous related studies have also wrongly used cross-sectional data to analyze poverty dynamics and inequality. This study comprehensively examines the measurements and correlation between economic inequality and poverty dynamics in the literature to better understand the empirical applications.


2.1. Conceptual Framework

Inequality and poverty have a complicated relationship (Beker, 2020). Simon Smith Kuznets' work is the most common and significant theory used to describe the relationship between poverty, income inequality, and economic growth. According to Kuznets (1955), inequality rises with rising per capita income and then declines in later stages of development. The terms "inequality" and "poor" are frequently used interchangeably in research on income distribution. Inequality is interpreted as poverty, while poverty is interpreted as inequality (Beker, 2020). According to Beker, poverty alleviation is hampered by inequality since it has a negative impact on poverty's growth elasticity (Beker, 2020). According to Besley and Burgess (2003), a positive and significant link exists between inequality and poverty levels within a country. The higher the proportion of national income taken by the rich, the lower the income available for the rest, thus resulting in a higher number of poor people. Inequality is concerned with the entire distribution of well-being, whereas poverty is only concerned with those who fall below the poverty line (McKay, 2002).

Hassan et al. (2015) establish a triangular link between poverty, income inequality, and economic growth (Figure 1), indicating the negative relationship between inequality, growth, and poverty. The incidence of poverty degenerates as the income gap increases, while growth decreases the rising poverty rate, according to the intermediary link between the three. Income redistribution and average income growth will reduce the movement into poverty and increase the movement out of poverty over time. Basu (2005) defined poverty-minimization inequality as the level of disparity society should allow to reduce poverty. He came to the conclusion that a society with absolute equality would be exceedingly impoverished (Basu, 2005). Alvaredo and Gasparini (2013), in their own case, found a weak relationship between poverty and inequality.

Figure 1.Nexus in Poverty-Growth-InequalitySource: Bourguignon (2004)


This study involved bibliographic research to obtain information on the links between economic inequality and the dynamics of poverty. The methods of collecting data and information involved the use of the internet and the grey literature. All documents were evaluated for information on theoretical and conceptual relationships between economic inequality and poverty dynamics, methods of measuring economic inequality and poverty dynamics, advantages and disadvantages of the methods of measurement, empirical studies on the links between economic inequality and poverty dynamics, and data requirements.


4.1. Approaches to Poverty Dynamics

The use of only one dimension, such as Foster, Greer, and Thorbecke's (1984) alpha poverty measurements, continues to dominate poverty measurement. Foster et al. (1984) used the FGT poverty measures to assess households' current poverty status based on their mean household per capita spending or income, neglecting poverty dynamics. These poverty measurements cannot determine whether high poverty rates are due to 'structural poverty' (lack of resources) or 'poverty risk' (large uninsured income volatility), which is crucial information for policymakers. The Multidimensional Poverty Index, established by Alkire and Foster (2011), is another effort to assess poverty that encompasses multiple factors (MPI). This MPI is an attempt to operationalize Amartya Sen's capability approach (Ivanov and Kagin, 2014). A multidimensional concept is described as a deficiency of basic necessities such as food, shelter, education, health care, safe drinking water, and sanitation that allow a person to live a decent, normal, and effective life. As a national MPI, MPI enables each country to select its own poverty indicators and dimensions (Alkire and Foster, 2011).

To investigate poverty dynamics, transition tables between succeeding poverty states are consistently utilized (Bane and Ellwood, 1986). Due to a possible measurement error, a poverty state that incorrectly indicates a classification, such as poor/non-poor, may appear to be only an imperfect measurement of "always poor."

Using transition matrices is one of the best direct techniques to determine to what extent households move out of and back into poverty. In the transition matrices, however, households are categorized as poor or not poor based on whether their incomes (or expenditures) surpass or drop below a predetermined poverty level (which may or may not vary between survey years). Consequently, transition matrices do not ensure :(i) how deprived or comfortable a household is; and (ii) if incomes are estimated with fault, making some households to be erroneously categorized.

Other approaches that have been used to analyze the dynamics of poverty are based on the central emphasis of the study and on the accessibility of data. Some of these approaches include (i) components-of-variance models (Devicienti, 2001); (ii) hazard rate models (Aassve et al., 2006; Stevens, 199); (iii) Latent Markov chain transition models; (iv) dynamic discrete choice models (Weon, and Rothwell, 2020); and (v) decomposition methods (Dickens and Ellwood, 2001). Each of these techniques has its advantages and shortcomings.

The "component" and "spells approach" are two methods for differentiating between chronic and transitory poverty groups (Foster, 2017). The'component approach' determines whether a household is chronically poor or not by looking at its average income over time. The spell method determines how many times a household has been in poverty (Martinez, 2016).

Hazard models and random-effects probit models are used to analyze the transition between different poverty states when more surveys are conducted regularly. The hazard model is appropriate for exploring the issue of duration dependence, whereas the random effects model is appropriate for investigating the issue of randomness. The protobit model is frequently used when a researcher is interested in state dependence.

According to the concept of a latent Markov chain, the true poverty states act like a Markov chain. Response matrices connect perceived poverty levels to actual poverty states. These response matrices represent the odds of observing manifest poverty states for various real (or latent) conditions. The response matrices are equivalent to the unit matrix if actual poverty states act like a Markov chain and no measurement mistake exists (Rendtel et al., 1998). Some studies have used this method to analyze empirical data (Baulch and McCulloch; Adepoju, 2012; Obayelu et al., 2021).

4.2. Approaches to Measuring Economic Inequality

Inequality metrics are used to determine how much economic disparity exists in a country or region and how that disparity varies over time and location. Decomposability is a fundamental issue in the literature on inequality measures. That is, knowing the contribution of each source to the overall inequality. Two methods of solving the problem are (i) the use of 'Shapley value' from the Shapley decomposition introduced by Shorrocks (1999) and (ii) the 'balance of inequality' (BOI) approach proposed by Di Maio and Landoni (2017). The difference in the inequality index value corresponding to the observed circumstance and the reference condition, where the income does not change with the factor, is measured by the Shapley decomposition.

Comprehensive treatments of the theory of inequality measurement were developed in the 1970s by Anthony Atkinson (Atkinson,1970) and Sen (1973), with the latter updated in 1999 by Foster and Sen (1999). Indices such as the Atkinson index, Theil index, Kuznets ratio, Palma index, and Gini index are common measures used to measure inequality. The Kuznets ratio is obtained by dividing the top fifth of the population's income by thetwo-fifths of the population's income. The 50:10 ratio measures income inequality between the middle (the median income, where half have more and half have less) and the bottom half, resulting in a measure of inequality in the bottom half. The '90:50 ratio' is a measure of income disparity between those at the top and those in the middle.

The Palma index is calculated by dividing the total income earned by households in the top decile (the top 10%) by the total income obtained by 40% of the most disadvantaged households. This indicator is based on the idea that economic dynamics at the extremes of the resource distribution are substantially to blame for inequality (earned income, accumulated savings). The 90:10 ratio is obtained by multiplying the 90:50 and 50:10 ratios together (the income at the 90th percentile divided by the income at the 10th percentile income). The income concentration among a small group of the wealthiest people is estimated by the percentage of income flowing to the top 1%. The 99 percent represents the disparity between the wealthiest at the top and the rest of the population. The variance between the weight of an individual (or group) in the population and the weight of his or her income in total income is measured by the 'Theil index'. An index of 0 indicates absolute equality, an index of 0.5 indicates inequality represented by a society where 74% of people have 26% of the resources and 26% of people have 74% of the resources, and an index of 1 indicates inequality represented by a society where 82.4 percent of people have 17.6% of the resources and 17.6% of people have 8.4% of the resources.

The “Gini index” is a recognized inequality metric, which calibrates the proportion of income sharing between individuals in a country compared to the whole population (Adeleye et al., 2020). This coefficient is established on the Lorenz curve, which is a cumulative frequency curve that displays inequality by likening the distribution of a given variable (for example, income) to the population. Higher readings of the Gini index confirm more inequality and vice versa. Measure the average or expected difference between pairs of incomes in the distribution relative to the distribution size. It is linked to the well-known Lorenz curve. A Gini coefficient of zero shows that everyone has the same income, while a Gini coefficient of one suggests that one person has all of the income and everyone else has nothing (or 100 percent). Because of its strong properties, the Gini coefficient is a valuable pointer of inequality: (1) the index would not change if all incomes were doubled; (2) the index would not change if the population size changed, but the concentration remained persistent.; (3) the index would not change if two individuals swapped incomes; (4) If a high-income individual gives money to a low-income individual, the index would drop. Furthermore, this coefficient is straightforward to apply and understand. However, while this measure of inequality is decomposable, it is notsubgroup-consistent. If poverty declines in one subgroup but remains unchanged in another and both subgroups have the same population size, the total poverty level must also fall (Foster and Shorrocks, 1991). When the income ranges of the subgroup distributions overlap, the Gini coefficient becomes a concern. In that instance, the effect of a given distributional change on subgroup inequality may be the polar opposite of its influence on overall inequality (World Bank, 2013). The Gini coefficient is divided into within-group, between-group, and overlap. The overlap term can override the within-group effect, resulting in subgroup discrepancies.

The Atkinson index is a number that ranges from 0 to 1, where 0 represents complete equality, and 1 indicates perfect inequality. The Atkinson index provides an answer to the following question: "What fraction of income would society be ready to give up if it could achieve a truly equal distribution of income?" The Atkinson index is this percentage. All of these indices have their merits and demerits and can be used individually or collectively to identify inequality between and within groups based on their applicability.

4.3. Empirical Review of Literature

Empirical studies on poverty dynamics are restricted owing to the dearth of enough panel data at the macro-level and microscale level in most developing countries. First-time studies have mostly been inhibited by the absence of panel data that would permit one to trail poverty transitions of people over time. Until recently, the dynamics of poverty have long been taken up at the national level and not at the household level. Panel data, which are formerly deficient in most developing countries, are admirable for analyzing poverty overtime at the micro level, although analysis at the national level is possible with cross-sectional data (Gunther, 2017). Some of the previous studies relied on cross-sectional living standard measurement surveys (LSMSs) that were widely available.

Greater income disparity is related to higher rates of poverty (Payne et al., 2017), and increases in income inequality are associated with increases in poverty, according to empirical estimates (McKnight, 2019). Empirical analyzes confirm that changes in work status, rather than changes in family arrangements, are more typically related to moves into and out of poverty. Public taxes and transfers are strongly linked to poverty transitions and persistence in EU member nations, but less so in the United States.

Luttmer (2021), in his study, measured poverty dynamics and inequality in transition economies by employing instrumental variable approaches and decomposing income into transient and permanent components. One of their findings shows that, after accounting for transient shocks, more than 80% of the poor in both Russia and Poland stay poor for at least one year.

Imimole (2021) evaluated the relationship between economic growth, population growth, poverty, and inequality. Annual time series from 1980 to 2019 were analyzed using the "Granger Causality test," and the existence of long-run relationships across variables was tested using the "Autoregressive Distributed Lag" (ARDL) bound test to determine causality direction, while the short-run dynamics were analyzed using the Error Correction Mechanism. The study's findings showed that there is a one-way causality relationship between poverty and inequality and between GDP growth and population growth. Similarly, there is a lasting positive association among these variables, with inequality impacting poverty and GDP growth positively affecting poverty in the short run. It was recommended that policies should aim to reduce inequality, particularly through poverty reduction measures and expanded access to reasonably priced healthcare services.

Between 2008 and 2017, Zizzamia et al. (2019) applied all five waves of data from the National Income Dynamics Study to comprehensively analyze poverty dynamics and inequality in South Africa. Ewubare and Okpani (2018) examined the relationship between poverty and income inequality in Nigeria in the 1980-2017 period. Adopting the Ordinary Least Square, unit root, cointegration, error correction model and Granger causality test to analyze the data from the Central Bank of Nigeria (CBN) Statistical Bulletin and World Bank. Their findings show that the national poverty index was positively related to inequality but statistically not significant.

Ajibola, Loto, and Enilolobo (2018) empirically examined poverty and inequality in Nigeria regarding its implications for inclusive growth from 1980 to 2013. Their findings reveal that inequality has a negative impact on poverty. As the economy grows, the government spends more money on health care, inequality increases, and poverty decreases. This is in contrast to Khan et al. (2014), who found that income inequality increases poverty, whereas growth reduces poverty. Ogbeide-Osaretin (2018) investigated the link between poverty and growth inequality between poverty and growth inequality and the existence and direction of causality in Nigeria, as well as the likelihood of a long-term relationship. Inequality was revealed to be positively significantly related to poverty using data from a five-year panel framework. From 1981 to 2005, Gries and Redlin (2010) used the Gaussian mixture model (GMM) method of estimation to view the dynamics of growth, inequality, and poverty in a panel of 114 developing countries and six geographical subpanels. Positive bidirectional causation exists between growth and inequality, as well as inequality and poverty, and negative bidirectional causality exists between growth and poverty, according to the findings.

In a multivariate approach, Apergis et al. (2011) analyzed a panel data set of 50 US states from 1980 to 2004 to explore the causality between income inequality and poverty. Both in the short- and long-term, the findings discovered a bidirectional connection between poverty and income inequality. In Abia State, Mbanasor et al. (2013) examined income inequality and poverty dynamics among rural agricultural households; the findings reveal a considerable level of inequality, with per capita income falling below the national minimum wage in operation.

In a study on 'poverty dynamics and income inequality in the eastern Brazilian Amazon' by Guedes et al. (2012) using household-level data from a representative sample of small farmers between 1997 and 2005, they discovered a direct link between poverty and inequality among small farmers, despite the fact that poverty reduction was more pronounced among new owners and inequality reduction was greater among original settlers.

Alazzawi (2010) provided a detailed analysis of the dynamics of moving into and out of poverty and inequality in Egypt, using a nationally representative panel survey between 1998 and 2006. Oxfam International's report on inequality in Nigeria in May 2017 showed that economic inequality had reached exciting heights. Regional inequality was found to be high in Nigeria, with higher poverty rates in the northwestern states of the country. According to Oxfam’s calculations, the amount of money that the richest Nigerian man can earn annually from his wealth is sufficient to lift 2 million people from 6 riches for one year.

A new report from the National Bureau of Statistics (NBS) was just released and was tagged '2019 Poverty and Inequality in Nigeria' using NLSS data. According to the report, about 83 million people, or 40% of the entire population, live below the country's poverty threshold of N137,430 (381.75 USD) each year. Cingano (2014) investigated the effect of inequality in human capital on economic growth. Increased financial gaps reduce skill development among people with weaker parental education backgrounds, according to an analysis based on microdata from the Adult Skills Survey (PIAAC), both in terms of the quantity and quality of education received (such as skill proficiency).

In Nigeria, Onwuka (2021) investigated the link between poverty, income disparity, and economic growth with time series data from the National Bureau of Statistics (NBS) and the Central Bank of Nigeria (CBN) Statistical Bulletin between 1981 and 2019. The results revealed that income inequality negatively affected the country's economic growth, but poverty has a favorable relationship with growth.

In addition, Nuruddeen and Ibrahim (2014) used secondary time series data from 2000 to 2012 to examine the relationship between poverty, inequality, and economic growth in Nigeria. The findings revealed a bidirectional causal relationship between literacy inequality and poverty.


5.1. Determinants of Inequality

Econometric modeling of the drivers of inequality and 'a priori' is based on the theoretical growth-inequality-poverty triangle model (Dhrifi, 2015). Some of the factors identified to affect income inequality from the literature include Per capita GDP, per capita health spending, tax pressure, the poverty rate, size of the government budget and the portion devoted to subsidies and transfers, share of the agricultural sector in total labor force, literacy rate,as well as human and land resources endowment. In their study, Aghion et al. (1999) identified technological change as the most important factor driving increasing inequality. Milanovic (2016) states that Kuznets waves (inequality rises, falls, and then rises again) are driven mostly by a technological revolution. This micro study recommends the variables indicated in Table 1 for operationalization when modeling drivers of inequality.

Definition of a variable Unit of measurement A priori Expectation
Inequality (Income/ health) Within and between different groups in society, there is an unequal distribution of income and opportunity. 'Inequality index' = 0 (implies total equality) and the index =1 (implies total inequality)
Household educational level Years of schooling Year -
Employment Level -
Gender Whether the household head is a male or female A dummy variable (1 if the respondent is male; otherwise, 0) ±
Age Age of the Head of Household year +
Age-square Nonlinear relationship of age year +
Household Size Member of the household making decisions and eating and living under the same roof. Number of persons ±
Poverty Rate Poverty Status Dummy (1 if poor, 0if non-poor) ±
Intervening variables
Government policy Tax rate If not available directly from the taxpayer, revenues as a share of GDP can be used. -
Cost of governance The total amount of budget per year Naira value ±
Macroeconomic Indicators Definition of a variable Unit of measurement A priori Expectation
GDP per capita GDP/Population Size Naira -
Price rise level Proportion of surge or reduction in prices during a definite time Percentage +
Real exchange rate The exchange rate of one country's currency against that of another. Percentage +
Investment Level Number of productive assets +
Table 1.Drivers of Inequality and A priori ExpectationsSource: Computed by the authors based on the theory and empirical literature in 2022

Source: Computed by the authors based on the theory and empirical literature in 2022

5.2. Poverty Dynamics' Drivers and Modeling

Poverty began as a result of a lack of social protection and care, as well as a lack of assets. Understanding factors that lift families out of poverty and what could have caused them to fall into poverty is informative for policies aimed at eradicating poverty nationally. In a well-targeted anti-poverty policy, characteristics of households and their members (such as household size and composition, educational attainment, and so on) may affect poverty dynamics.

Although the multinomial logit model (with examples of applications by Bigsten et al., 2008; Obayelu et al., 2021) has turned out to be one of the most widely used models for analyzing poverty dynamics, it is not the only model that can be used for this purpose. There are three flaws in the MNL model: 1) the assumption 'Independence of Irrelevant Alternative (IIA)' ensures that the odds ratio is not affected by other factors; 2) the assumption assumption assumption Independently and Identically Distributed (IID) eliminates heterogeneity that causes variance and covariance, and 3) the ordernature of its results. MNL entails creating a polytomous variable for individuals who have risen out of poverty and those who have remained in poverty, as well as looking at their correlates. To overcome these limitations, the 'sequential logit model' gives the dynamics of the an of poverty anadditional structure in addition to the unordered categories of the multinomial logit model, although the 'nested logit model' allows more variability invariance and covariance of results. Other models that loosen the IIA assumption, such as the sequential model, could be used to circumvent this model's drawback (Baulch and Vu, 2011), ordered logit (or probit), and stereotype logistic models (Baulch, 2011). The multinomial logit model is appropriate when the focus is on the mobility of the poor and only a short panel of survey data is available. The application is typical in developing countries, where panel data consists of only two waves of household data (Nguyen et al., 2018).

Akerele et al. (2017) analyzed the transition of the transition of poverty and its determinants in rural Nigeria using panel data collected during the post-planting season of 2010 and the post-harvest period of 2011 by the World Bank. Findings using Markov chain and probit regression revealed that higher educational attainment and involvement in crop-livestock production substantially reduce the likelihood of a household entering or remaining in poverty, while the increase in the number of adolescents enhanced it, and sole reliance on crop production and geospatial factors substantially depress the chance of escaping poverty. The transition of poverty was shown as a function of variations of economic and demographic factors affecting households by Aassve et al. (2006) and Burgess and Propper (1998).

Justino and Litchfield (2003) used the Markov model of poverty transitions and the multinomial logistic regression model to model poverty dynamics and drivers. Their study found four poverty transition scenarios that differ from each other: (i) both times were poor, (ii) in the first case, it was not poor, but in the second case, it was poor, (iii) in the first case, poor, but not in the second, and (v) in both times, nonpoor. Furthermore, Beccaria et al. (2011) calculated the 'Relative Risk Ratios (RRR),' which denotes the livelihood of a household experiencing one of four jointly exclusive outcomes and normalized by dividing these likelihoods by the chance of experiencing any other state. Although education and asset sizes played an important role in the prevalence and severity of multidimensional poverty among households, Adepoju (2012) found that “educational status, household size, number of assets owned and ownership of land influenced transient poverty, while marital status, household size, land ownership and number of assets owned influenced chronic poverty”. The MNL model used categorical outcomes that were not in any particular order, ignoring the natural order of poverty transitions.

Teguh and Nurkholis (2011) used the 'ordered logit model' to assess the status of poverty in households in Indonesia and looked at factors of poverty dynamics: poverty, temporary poor (-), transient poor (+), and non-poor. The estimated results show that poverty dynamics in Indonesia issignificantly affected by educational achievement, household size, physical assets, employment status, health shocks, access to electricity, and changes in the working sector and the microcredit program. Bokosi (2007) examined the determinants of poverty dynamics among Malawian households between 1998 and 2002 using a bivariate probit model with endogenous selection to address the initial conditions problem. The findings show that the education of the head of the household, the per capita acreage cultivated, and changes in household size are significantly related to the probability of being poor in 2002, despite the poverty status in 1998. In a study by the OECD (2001), changes in work status are more frequently connected with movements into and out of poverty than changes in family structures, despite the fact that the two are closely related.

5.3. Development of software to measure inequality and poverty

Software for estimating inequality and poverty estimates and associated sampling variances is now much more widely available. There are free standalone programs available to researchers, the most notable of which is the “Distributive Analysis Stata Package (DASP 2.1)” by Araar and Duclos (2009), which is included in “STATA 10”. There are numerous free software packages that can be used in combination with general-purpose statistical software like Stata (Jenkins, 2006).

5.4. Measurement Issues in Inequality and Poverty Dynamics

Measurement of poverty and inequality requires a large number of methodological decisions (Pizzolitto, 2005). Some of these choices are based on theoretical considerations. Choosing a dimension to define poverty (income/consumption, basic requirements, endowments), a poverty line, and an aggregate approach, such as a poverty index, are necessary to measure poverty. To measure inequality, similar judgments need to be made. The extensive literature on poverty and inequality indicators reflects the complexity of the problem. Practical issues are addressed at the second level of methodological decisions. Even after the decision on measurements of poverty and inequality, the empirical implementation is usually not trivial. The statistics are then exposed to potential biases due to income nonresponse and under-reporting. Another issue arises from the fact that the things included in household income vary by country and sometimes even by time within a country.

Regarding income distribution and measures of inequality, such as the Gini coefficient, spurious fluctuations in income have been a prominent point of interest (Slemrod, 1992). Within the context of income inequality, spurious changes are expected to generate a higher mobility of earnings and, as a consequence, a higher inequality. In a discrete poverty state context, the latent Markov chain approach seems to be a useful tool to treat spurious changes.

5.5. Data Requirements and Limitations

Poverty dynamics is characterized by frequent in-and-out moves. To appropriately examine the dynamic aspects of poverty, lengthy panel data on income and consumption would be needed. But for many developing countries, lengthy panel data do not exist, and cross-sectional surveys (or sometimes panels with two or three waves), with either income or consumption data, are the only data available. Panel data provide for a more thorough examination of poverty entry and departure points (Suppa, 2017).

Panel data could provide essential information on whether or not poverty-eradication measures are effective, for whom they are effective, and why. Panel survey data may be used to identify and measure trajectories into and out of poverty and determine what drives household moves into and out of poverty over time. There may also be a subset of persons who are perpetually impoverished, unable to escape poverty year after year. Panel data isessential to determine whether this is the case and, if so, who is left behind. The use of household panel data can reveal crucial information on the causes of poverty.

A study in Albania used a synthetic panel methodology to investigate household transitions into and out of poverty due to the scarcity of panel data (UNDP, 2017). The Household Budget Survey was used to generate synthetic panel (HBS) data in 2008 and 2014. The process for the generation of synthetic panels is obtained, according to Dang et al. (2014), by estimating household consumption in a future year (or in a previous year) and transforming two or more cross-sections of household survey data into a panel dataset.

5.6. Identification of relevant power and stakeholders that influence inequality and poverty dynamics in Nigeria

The results of this study have identified that the following actors could play crucial roles in solving the problem of inequality and the transition frompoverty of households in Nigeria. These actors are government, judiciary, politicians, media, non-governmental organizations, civil society organizations (CSOs), investors, financial institutions, local or community leaders and members of households, technology managers, and academia (Table 2). It is obvious to a great extent that all parties could individually or collectively influence inequality and poverty incidences at the micro and macro levels.

Stakeholders Groups Key Concerns Interest Influence and expectations
Government at all levels (legislative, executive members) (i) Legislative framework on project funding and policy regulation, such as taxes (i) Redistribution from high-income to low-income.(ii) Increase economic inclusion (i) Social protection policies, including progressive taxation.(iii)Policy certainty and stability.
Judiciary (i)Enforcement of antidiscrimination laws (i)Enforcement of the rule of law. (i) Interprets laws, applies them to specific circumstances, and evaluates whether they are in conflict with the Constitution.
Politicians Political participation in reducing inequality between all regions of the country Fair opportunities and protection for all members of society, and not just people who voted for them or are being represented Creation of decent work in a society with higher incomes
The Media (i) Public concern (ii) education of the masses; (i) increase the connections between people and the environment in which they can share your opinions Direct access to accurate information
Civil Society Organisation (i) Ensure justice and fairness(ii) engrossed in advocacy and bargaining unusual policies for government, the private sector, and other institutions. (i) Monitor government activities and political processes and embrace authorities liable for performing sensibly and in harmony with the rule. (ii) Make every effort to create valuable developments in people's social, economic, political, and cultural lives. (i)Drive national governments for regulations that encourage equality.(ii) Protect the delivery of basic services, protect, and support human rights.
Non-Governmental Organisations (NGOs) (i) acts as a intermediary between society and the government, assessing the needs of individuals;(ii) Project funding. Involvement in various aspects of community development such as community mobilization, promotion of child's rights law. (i) Human Empowerment Through Capacity Building(ii) Engage in lobbying governmental actors for social, economic, and political change.
Investors (i) Investment in universal access to healthcare and education (ii)Build assets for working families;(iii) Investment in economic infrastructure (i) Look at ways to improve low wage scales and unsafe working.conditions, and labor standards violations Fair labor standards, access to collective bargaining, and quality job creation
Financial institutions Empowerment of all categories of people through financial inclusion. Financial inclusion to assist in solving the problem of inequality (i) Promotion of a more inclusive financial system.(ii) Expand financial literacy programs and activities.(iii) Incentivise providers to deploy ATMs and POS in rural communities.(iv) Create incentives for MFBs to focus on serving rural communities.
Community Leaders and members of households Promote the integration of hazard mitigation within the community. Mobilize and guide others. Aid in the development of grassroots innovations,
Technology managers and academia Fostering innovation and efficiency Improve the relative position of skilled workers Human Development
Table 2.Power and stakeholder mapping matrix of relevant actors influencing inequality and poverty dynamics in NigeriaSource: Compiled by the authors, 2022


This article reviewed different empirical methodologies to analyze poverty dynamics and inequality and unpacks the the relationships between economic inequality and poverty in terms of the movement of people into and out of poverty over time. The empirical findings on economic inequality and poverty dynamics are highly inconclusive, as there is no collectively established measure of inequality and poverty. The common conclusion in most of the articles reviewed established that poverty dynamics is largely ascribed to the unbalanced spreading of resources, with poverty attributed to inequality. Another striking finding in the research is that the lower the elasticity of poverty-growth, the greater the inequality. This study comprehensively examines the measurements and correlation between economic inequality and poverty dynamics in the literature for a better understanding of the empirical applications. The findings of this study will uniquely contribute to the discourse on growth-poverty inequality in the 2030 Sustainable Development Goals 1 and 10.

To conduct an analysis of poverty transition, a panel (longitudinal) dataset is needed that provides information on the living standards of households over time. This type of data would enable researchers to track changes in a household's poverty level in comparison to a predetermined poverty line. Furthermore, by correcting for variances in these characteristics, panel data can make the association between time-variant, observable nation characters, and explanatory factors more accurate. Latent Markov chain analysis is good for analyzing poverty dynamics and ordered logit/probit regression for the study of factors that influence poverty dynamics.


The lack of reliable dynamic data is one of the reasons for the small number of studies on poverty transitions in developing countries. Without panel data, it may be impossible to distinguish between persistent poor people and people who move in and out of poverty. Precisely

I. Based on the indeterminate relationship between inequality and poverty, policies to combat poverty may or may not be the same as policies to combat inequality.

II. Panel data sets with more detailed information are ideal for tracking poverty over time (economic inequality and poverty dynamics. Household surveys are conducted on a regular basis (longitudinal survey); for example, the “Living Standards Measurement Survey (LSMS)” will shed light on 'poverty dynamics', such as movements into, out of or failures to rise beyond the poverty line.

III. Real income, not nominal income, is of interest at any time income is contrasted over place or time. On the other hand, poverty is considered not just by a shortage of income but also by a lack of purchasing power resulting from both income and prices.

IV. To avoid biases or measurement errors while estimating economic inequality and national poverty dynamics, care should be taken to maintain the same survey design at different times when data are collected from the same respondents.

V. Utilizing the Gini index through Shapley's approach allows us to develop each factor's expected marginal influence on inequality.

VI. Based on the weaknesses in the use of the Gini coefficient, it should not be used as the only measure of economic inequality. Other methods can be used.

VII. Cash transfer programs to low-income families and improved taxation policies can be applied to effectively reduce poverty.

7.1. Suggested Topics for Future Studies

More emphasis should be paid to the factors that allow people to take advantage of opportunities to leave poverty and the endowments that provide the most resistance to shocks. Specifically, future studies can look at the following:

1. The links between the non-farm sector and the dynamics of poverty using panel data sets.

2. Poverty dynamics by household sociodemographic characteristics.

3. Poverty dynamics and inequality Nexus: Empirical evidence from developing countries.