Income Distribution in the UK: 2009 and 2017

8540 words (34 pages) Business Assignment

13th May 2020 Business Assignment Reference this

Tags: Business AssignmentsFinanceExternal AnalysisBusinessLiterature ReviewDiversityWorkforce

Disclaimer: This work has been submitted by a student. This is not an example of the work produced by our Business Assignment Writing Service. You can view samples of our professional work here.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of BusinessTeacher.org.

ABSTRACT

This paper evaluates income distribution changes in the UK between the year 2010 and 2017. It also aims to examine income inequality between households of different socio-economic and regional characteristics. By utilizing data from Understanding Society, (UKHLS), income distribution is investigated by applying tabular analyses and inequality measurements such as Gini coefficient. An analysis of the determinants of income in the UK is also reported in this study. The results suggest that there haven’t been significant changes in the income distribution which indicate that income inequality has remained fairly stable with a slight decline.

Keywords: income distribution, income inequality, Gini ratio, Lorenz curve, regression analysis

Introduction

Inequality in income has been a central issue among social researchers for many years. Equality is an important value in most societies and people care about it. Presence of inequality shows that particular segments of the society are suffering from persistent unfairness due to lack of opportunity and income mobility. It is important to study income inequality as widening inequality has significant implications for growth and macroeconomic stability.

The main objective of this paper is to evaluate income distribution changes in the UK and to describe how income inequality in the UK has changed during the period 2010 and 2017. This study also analyses selected socio-economic and regional groups using characteristics of the household to see if income distribution has undergone major changes over this period. This paper explores and updates the information on income distribution in the UK. Based on this information, policy makers can plan and design policies to alleviate income inequality in the UK.

The main objectives of this study are to:

  • Analyse changes in income inequality between 2010 and 2017.
  • Compare income distribution pattern between the period 2010 and 2017.
  • Compare variation in mean income of different socio-economic and regional characteristics between 2010 and 2017.
  • Analyze the impacts of socio-economic and regional characteristics of household on household income.

The remaining sections of this study are organized as follows. Section 2 presents a review of empirical literature on income inequality. Section 3 consists of theoretical framework with methodology, data description, description of variables and model specification. Section 4 discusses the tabular analyses of size distribution and the Gini ratio. Section 5 talks about the results and discussion of regression analyses. Section 6 concludes the major findings of the study.

Literature Review

Issue of global income inequality has generated considerable debate in the academic field. Global inequality is higher today than it was 200 years ago (Bourguignon and Morrisson 2002). Evidence from numerous literatures since the mid-1990s indicate, however, that the world income inequality has declined during the end of the twentieth century (Bhalla 2002; Firebaugh and Goesling 2004; Sala-i-Martin 2006; Clark 2011; Milanovic 2013; Ravallion 2014). However, these research studies have several limitations, including the use of outdated income data (Clark, 2011). Findings from Sala-i-Martin (2002) suggests that world income inequality declined between 1978 and1998. The Gini coefficient showed a decline of 4.9 percent, the Theil index declined by 9.9 percent and a decrease of 12.8 percent as measured by the mean logarithmic deviation (MLD). In the study conducted by Clark (2011), he addresses the limitations to existing research by updating previous estimates of world income with improved and more recent data of 151 countries during the 1990-2008 periods. He concludes that world income inequality has stopped growing and has started to decline during the beginning of the twenty-first-century era. While there has been a decline in between-country inequality, there has been a rise in within-country inequality although at a slower pace since 2000.

A considerable amount of literature has been published on income inequality in the UK. These studies have considered various aspects of the changes in income distribution in the UK since the 1990s. Some assessed the causes of changes in the distribution of household income (Jenkins 1996; Brewer & Wren-Lewis 2011) while other researchers (Clark and Leicester 2004; Hills et al. 2014) analysed the effects in personal tax on income inequality. In recent times, OECD (2011) analyses several factors that may be affecting household income inequality among OECD countries and results indicate that most OECD countries have experienced a rise in income inequality, while some countries like the UK has seen a slight decline since 1990. While some studies have analysed income distribution of household income in the UK during 1980-2009, no similar detailed research exists for later periods. So, my research paper will focus on the period from 2009 to 2017.

There are numerous literatures in income distribution which estimates household income distribution by using tabular analyses with decile and quintile distribution (Ahiram 1964; Mookherjee&Shorrocks 1982;Banskota et al. 1987; Livada 1991; Jenkins 1996; Hills 2004). In a study by Banskota et al. (1987), household expenditure surveys of 1975, 1976 and 1977 were used to investigate the size distribution of household income in Jamaica. Households were categorized by a selective socioeconomic and regional group such as occupation, household size, age group and regional disparities (urban-rural areas). The shares of income were divided into 5 equal parts; each of them called a quintile is generated by the selected variables. The results indicated that the pattern of the size distribution of income in Jamaica remained biased towards the top 20 percent of the households andaccounted for over 40 percent of the total income in most variables. The income share of the bottom 10 percent of the households doubled whereas the income share of the top 10 percent of the households decline by about 30 percent suggesting that the income inequality had reduced in the intervening 20 years. Likewise, in another study on income distribution in Jamaica, Ahiram (1964) found that the bottom 20 percent of the households have 2 percent of the income while the top 5 wealthiest group has a share of 30 percent of total income. In other words, the household from the top group on average has 120 times more income than the household in the lowest group.

Most researchers investigating on income inequality have used Gini coefficient as a tool to measure the extent of inequality (Livada 1991; Xie & Zhou, 2014; Banskota et. al., 1984; Clark, 2011; Azam & Bhatt 2018). Some studies have found that income inequality is rising by using the Gini coefficient. Xie & Zhou (2014) have reported that the Gini coefficient in China nearly doubled from 0.30 in 1980 to 0.55 in 2012. This significant rise in inequality is explained partly by various government policies that favour urban residents over rural residents and partly by the rapid economic growth of China. Similarly, in a study conducted by Azam & Bhatt (2018), they found that in 2011, the Gini coefficient for rural India was 0.508, whereas for urban India it was 0.490. In contrast, several of the studies also report that the Gini coefficient was lower in most recent years compared to earlier years. In an investigation of the world income inequality, Clark (2011) argued that the world income inequality during 1990-1999 slightly declined, this has been supported by the evidence that the Gini coefficient showed a minimal decrease of 1.3 percent from 0.697 to 0.688. Likewise, in a study conducted by Banskota et al. (1987) income distribution in Jamaica from the years 1975, 1976 and 1977 were compared to 1958 and they reported that the Gini coefficient in the later years was lower than in 1958, indicating that the income inequality had decreased during the 20 year period. In the same way, Ipek (2019) investigates income inequality of Turkey during the period 2003-2015 and finds that the Gini coefficient declined by 5.1 percent over the 13 year period. The studies presented here give a sense that income inequality is rising in some countries, but it is also declining in other countries. This may be due to various factors like government policies, growth of the economy, globalization and whether the country is developed or is a developing country.

In addition to using the Gini coefficient, some literatures also use Theil entropy measure. Both measures are particularly sensitive to transfers that occur at different points along with the income distribution. While the Gini coefficient is sensitive to transfer that occur in the middle of the income distribution, the Theil index is more sensitive to transfers that occur at the top of the income distribution (Clark 2011). The results from the Theil index showed an increase of 0.1 percent from 0.930 to 0.931 in the period 1990-1999. During the same period, Theil decomposition showed that between-country inequality decreased by 5.3 percent (0.734 to 0.695), whereas the within-country increased by a remarkable 20.4 percent (0.196 to .236).

Several authors have used the Lorenz curve to help better understand income inequality (Livada 1991; Qingtao et al. 2014). A Lorenz curve can give graphical representation of the distribution of income and shows how much inequality exists in a region or in a country. Similarly, this paper also utilises Lorenz curve to show graphically if income inequality has risen or declined.

There is a large volume of published articles describing the role of household characteristics in determining the household income by using regression-based methods (Banskota et. al 1984; Zang, 2012; Xie& Zhou 2014; Lin et al. 2015; Brewer & Wren‐Lewis 2015). Xie& Zhou (2014) applies regression analysis to examine the determinants of income to estimate the impact of selected variables – region, area type, education level, race/ethnicity, and family structure. They find that income inequality in China is more pronounced by regional variation and urban-rural gap and 10% of the total income inequality is explained by the urban-rural gap. Likewise, the role of education is seen to be important as 15% of the total income inequality can be explained by the level of education of the household head. Similarly, education is found to be higher in urban areas than in rural areas. Lin et al (2015) aim to better understand the relationship between age and income inequality in Taiwan by introducing a spatial factor in the model. Other parameters like education level and unemployment are also added in the model. Results indicate that aging contributes negatively to the increase in inequality which means that the distribution of income has become more equal across regions. Likewise, the higher the educational attainment of individuals, the lesser is the income inequality among households. But, a rise unemployment showed positive relation to income inequality. Brewer & Wren‐Lewis (2016) analyses data from 1978 to 2009 in the UK to better understand why income inequality rose significantly from 1978 to 1991 and then remained fairly stable. They use multivariate regression-based methodology to analyse the effect of household characteristics such as age, sex and education on income inequality. They find that between 1978 and 1991, unemployment and other factors such as education and region increased income inequality. However, neither education nor region appeared to have a significant effect on changes in income inequality since 1991.

Methodology

Theoretical Framework

In this paper income distribution changes in the UK from 2010 to 2017 is investigated by applying tabular analyses and inequality measurements such as the Gini coefficient and Lorenz Curves. A regression statistical analysis of sources of income variation in the UK is also reported to provide explanation of income distribution changes.

For tabular analysis and regressions, the variables used are based on educational qualification (Degree holder) of the household, (2 groups), age class of the household (10 groups), geographical location of the household (4 groups), ethnicity of household head (5 groups), sex of household (2 groups), occupation of household (6 groups) and household size (7 groups). In the tabular analysis, the share of income accruing to the five percentage groups were calculated for the sample partitions generated by particular socioeconomic variables. In the regression analysis, the characteristics for partitioning the households were defined as independent dummy variables. Different explanatory variables were used in expressions with the dependent variable expressed as the log of the household income.

The analysis is carried out by two interlinked method of measuring inequality: the Lorenz Curve and the Gini Coefficient. Apart from computing Gini coefficient, analysis needs to be carried out according to income share accruing to different groups of population in deciles and quintiles. These are described below:

Quintile Distribution

A quintile is a statistical value of a data set that represents 20% of a given population, so the first quintile represents the lowest fifth of the data (1-20%); the second quintile represents the second fifth (21%-40%); the third quintile represents the third fifth (41%-60%); the fourth quintile represents the fourth fifth (61%-80%) and the fifth quintile represents the highest fifth (81%-100%). In tabular analysis, the shares of income were accumulated by using quintile distribution for different socioeconomic and regional groups.

Decile Distribution

Decile is a statistical method of dividing a set of ranked data into 10 equal subsections. To analyse the overall income distribution of 2010 and 2017, the shares of income were accumulated by using decile distribution.

The Lorenz Curve

First developed in 1905 A.D. by Max O. Lorenz, A Lorenz curve is a graphical representation of the distribution of income or wealth. It is a widely used technique to represent and analyze the size distribution of income, wealth as well as many other magnitudes. The curve plots the cumulative portion of income units and the cumulative proportion of income received when income units are arranged in progressive order of their income.

The Gini coefficient

Gini coefficient is the most commonly used measure of inequality. It was developed by the Italian statistician and sociologist Corrado Gini in 1912 A.D.The Gini is derived from the Lorenz curve, which sorts the population from poorest to richest, and shows the cumulative proportion of the population on the horizontal axis and the cumulative proportion of income on the vertical axis.
The Gini coefficient = Area Between Lorenz Curve and DiagonalTotal Area Under Diagonal

Data

This paper utilises data from Understanding Society, UK Household Longitudinal Study (UKHLS), which is a longitudinal survey with approximately 40,000 households at Wave 1 (University of Essex, 2018). This survey is conducted in the UK which consists of four countries: England, Scotland, Wales and Northern Ireland. To study the differences in income inequality in two different years (2010 and 2017), this paper uses the first Wave (Wave 1), corresponding to the survey year 2009/10 and the latest wave (Wave 8), corresponding to the survey year 2016/17.

Understanding Society provides high calibre longitudinal information about subjects, for example, family structure, income, health, employment, education and social life to help comprehend the long haul impacts of social and monetary change, just as arrangement intercessions intended to affect upon the general prosperity of the UK population. The individuals who are interviews are adult household members who are of age 16 or older.

Description of Variables

Dependent Variable used in this study:

Y = Annual Net Household Income

Annual net household income is used as the dependent variable in this paper to evaluate the distribution of income and to examine the determinants of income. Several papers have used net income to analyse income distribution (Hills 2004; Biewen & Juhasz 2012; Jenkins 1996). Net monthly household income is the sum of net monthly income from all households. Net monthly income is given as gross income minus taxes (no taxes deducted other than taxes on earnings) (Al Baghai et al 2015). Annual net household income is derived from net monthly household income assuming that the households earn the same net monthly household income across the whole year. Annual net household income is calculated by multiplying the net monthly household income by 12 to obtain the annualised net household income. Negative and zero income have been omitted from both surveys which bring to a total of 50,529 for Wave 1 (2009) and 38,681 for Wave 2 (2017). Logarithm of annual net household income is used in the regression analysis to interpret the coefficient as semi elasticity.

Independent Variables used in this study:

HH = Household Size

The number of household size is divided into 7 categories in this study: one member, two member, three member, four member, 5 member, 6 member and 7 and above member of the household. Number of household size can affect how much a household can earn. Generally, households with large occupants size is expected to earn more income opposed to small household size. Household size will be used as a categorical variable in regression analysis.

Degree = Education Level of Household

Education level of household is categorised into two groups: one with degree and the other with no degree. Having a degree is an important factor in determining how much an individual earns. It is expected that households having a degree will earn more than households having no degree. Education level will be used as a dummy variable in regression analysis.

Eth = Ethnic Group of Household

There are many ethnic groups living in the UK. For simplicity of this study, ethnic groups are divided into 5 categories: White, Asian or Asian British, Black or Black British, Mixed/Multiple and Other. Ethnic group will be used as a categorical variable in regression analysis.

Country = Geographical Location of Household

UK consists of 4 countries so households are divided into 4 groups: England, Scotland, Wales, and Northern Ireland. England is the largest country in the UK. Country will be used as a categorical variable in regression analysis.

Age = Age of the household head
Age of the household is an important factor in explaining the income of the household. As theory
suggests, income increases but at a decreasing rate with age. 10 classes of age is created in the tabular analyses starting from below 24 to 65 and above.

T = Time
Time is used as a means of comparison of income levels between 2007 and 2017. Time will be used as a
dummy variable in regression analysis. Where T=1 if the year is 2017 and T=0 if the year is 2009.

S = Sex
Income levels vary with sex as well. Sex will also be used as a dummy variable in regression analysis. Where, S=1 if the sex is male and S=0 if the sex is female

Job = Occupation of Household

There are various types of occupation. The Registrar General’s Social Class (SC) of current job is used in this study. In this variable, occupation is divided into 6 categories: professional, managerial and technical, skilled non-manual, skilled manual, partly skilled and unskilled. Type of occupation also determines the level of income one earns. Occupation will be used as categorical variable in regression analysis.

Model Specification

This study uses regression analysis of sources of income variation in the UK to provide explanation of income distribution changes. In the regression analysis, the characteristics for partitioning the households are defined as independent dummy variables. Different explanatory variables are used in expressions with thedependent variable expressed as the log of the household income. Various specifications of different combination of independent explanatory variables are created. Time dummy is created to compare the results of two time periods 2009 and 2017 and is retained in all the specifications.

Specification 1

Time and household size dummies are regressed on the log of household income. The results are deviated around four member households. The equation of specification 1 can be written as:

Y = μ0+ μ1T+ μ2Country+ε

Specification 2

Time and location by country dummies are regressed on the log of household income. The results are deviated around households living in England. The equation of specification 2 can be written as:

Y = β0+ β1T+ β2Country+ε

Specification 3

Time and sex dummies are regressed on the log of household income. The results are deviated around female. The equation of specification 3 can be written as:

Y = δ0+ δ1T+ δ2S+ε

Specification 4

Time and urban-rural location dummies are regressed on the log of household income. The results are deviated around rural households. The equation of specification 4 can be written as:

Y = ρ0+ ρ1T+ ρ2Urban+ε

Specification 5

Time and ethnic group dummies are regressed on the log of household income. The results are deviated around White ethnic group. The equation of specification 5 can be written as:

Y = κ0+ κ1T+ κ2Eth+ε

Specification 6

Time and educational qualification dummies are regressed on the log of household income. The results are deviated around no degree households. The equation of specification 6 can be written as:

Y = λ0+ λ1T+ λ2Ed+ε

Specification 7

Time and occupation dummies are regressed on the log of household income. The results are deviated around one member households. The equation of specification 7 can be written as:

Y = α0+ α1T+ α2Job+ε

Specification 8

All the independent variables are regressed on the log of household income. The equation of specification 8 can be written as:

Y = θ0+ θ1T+ θ2HH+θ3Country+θ4S+θ5Urban+θ6Eth+θ7Degree+θ8Job+ε

Descriptive Statistics

Age of Household

Table 1 and Table 2 summarise the results on the distribution of household income in 2009 and 2017 respectively by age of household. Several observations are suggested in Table 1. It is observed that in 2009 the peak mean income occurs at the age class of 45-49 in 2009 and 2017 with mean income of £38,157.The age class of 45-49 has the highest mean income because generally people at this stage of age reach their peak in their career and earn the most. The second largest mean income is earned by 50-54 age class with a slightly less mean income of £37,208. Table 2 shows fairly identical observations to Table 1. From the estimated mean income perspective, it is observed that the peak mean income occurs at age class of 45-49 earns the highest mean income with £48,428. This suggests that there hasn’t been much change between 2009 and 2017 in the age class who earns the highest income. There is a slight change in the age class who earns the second highest mean income with age group of below 24 (£47,546) having roughly the same mean income as the age group of 50-54 (£47,346). This gives us the idea that young adults of age 16-23 are more involved in the labour force than in educational attainment. Individuals between ages of 24 to 44 have fairly similar mean income. After the age of 54, generally, the mean incomes begin to decline. Age class of 65 and above has the highest percentage of individuals in a household and earns the lowest mean income in both 2009 (£21,795) and 2017 (£31,831). This is expected as the generally people retired during this age and are not actively participating in the labour force and hence, earn low income. But one thing to note is that the overall the mean income has risen.

The distributions of income for each age class by percentile group are also presented in Table 1 and Table 2. Clearly, the distribution of income is rightward skewed in both Table 1 and Table 2. The top 20 percent (81-100%) of the households in all the age group account for over 39 percent of the overall income. Also note that the concentration of income at the top 20 percent of the households is greatest amongst the 55-59 age class. The bottom 20 percent of the household on the other hand account for only about 6 percent of the total income. Gini rations are also reported in Table 6 and Table 7 along with an index of inequality. The index of inequality is simply the Gini coefficient of each class normalized by the lowest Gini coefficient. Income inequality is greatest among households with 55-59 years of agethan among households between age of 30 and 39 years.Similarly, as reported in Table 2, the top percent (81-100%) of the households in all age group account for over 36 percent which is slightly less than what is observed in 2009. The concentration of income at the top 20 percent of the households is greatest among 60-64 age class. Income inequality is greatest among households with 60-64 years old individuals than among households with 40-44 years of age which is indicated by the Gini ratio and the index of inequality. Majority of people generally retire after the age of 60 so they are not actively participated in labour force and do not earn income by employment but they may be earning through passive income (such as pension, rent income, interests, dividends and many other) which varies from individuals. When we compare between the two years (2009 and 2017), it seems that the households with greatest income inequality has passed on a step further from age class 55-59 to 60-64 and households with the lowest inequality has moved from age 30-39 to 40-44.

Occupational Status of household

Income distribution by occupational status of household is reported in Table 3 and Table 4. The income variation among households is greater by occupation status of the household than by age group of the household. A majority of the household reported managerial and technical as their occupation in both Table 3 (35 percent) and Table 4 (37 percent). These households earned the second highest mean income of £45,393 in 2009 and £54,023 in 2018. Table 3 and Table 4, show that the highest earners were those who worked in professional occupation earning almost twice than the unskilled household who earn the least mean income.

The size distribution of income by occupational class of the household was rightward skewed. In Table 3, the top 20 percent of the households received over 40 percent of the income while the bottom 20 percent had less than 7 percent. Based on the index of inequality, inequality was greater among households with professional occupation and was higher by 4 percent than households with skilled non-manual occupation. Similarly, in Table 4, the top 20 percent of the households received about 37 percent of the income while the bottom 20 percent had about 8 percent of total income. Similar to 2009, index of inequality shows that inequality was 8 percent higher in households with professional occupation than households with skilled non-manual. Looking at Table 3 and Table 4, overall, inequality was most pronounced in households with professional occupation while it was least pronounce in households with skilled non-manual occupation.

Household Size

Distributions of household incomes by household sizes are reported in Table 5 and Table 6. In Table 3, households with 7 or more members, which constitute the lowest percentage (2.28%) of all households, had an estimated mean income £49064 which was more than three times the mean income of the households with just one member. The maximum number of members in the household is estimated to be 16 where most of the members may be involved in the labour market which explains the larger share of income. Households with two members constituted the largest percentage (34.16%) and had almost the double mean income relative to unimember households. The results from Table 6 is consistent with Table 5, where households with 7 or more members, which constitute the lowest percentage (2.48%) of all households, had an estimated mean income £60338 and was thrice the mean income of one member households. Similarly, households with two members constituted the largest percentage (33.71%) with a mean income of £38101. The standard deviation of one member households is higher relative to its mean income, which means that there was a lot of variation in incomes earned by individuals in one member households. By simply looking at the estimated mean incomes of households with different members from both 2009 and 2017, we can conclude that income increases as size of the household increases.

The distribution of income by percentiles paint a picture that is again rightward skewed in both Table 5 and Table 6. In Table 5, the top percent of the households account for over 38 percent of the total income as opposed to less than 8 percentof total income received by the bottom 20 percent of the households. Gini ratio and index of inequality indicate that income inequality is greatest among unimember households than among four member households. Similarly in Table 6, the top 20 percent of all the households account for over 35 percent with majority of over 37 percent. The bottom 20 percent account for 8 percent of households. Inequality of income is greatest among one member households than among seven and more member households. Income inequality was most pronounced among unimember households in both 2009 and 2017

Country

The United Kingdom consists of four countries: England, Scotland, Wales and Northern Ireland. England accounts for 53 percent of the total area of the UK, Scotland accounts for just under one third (32 percent), Wales accounts for less than a tenth (9 percent) and Northern Island accounts for only 6 percent of the total area of the UK. Geographically England and Wales are located on the southern part of the UK, whereas Northern Island is located on the North-West and Scotland is situated on the North of the UK. Table 12 and Table 13 provide a summary on the distribution of household incomes by countries. About 80 percent of sampled households live in England in both years. In Table 12, the estimated mean incomes of all the countries are about the same with an average of £31,000. England has a slightly higher mean income of £32,758. Similarly, Table 13 shows that the average estimated mean income of all the countries in the UK is £39,000 with England having £42,706.

The distribution of income by countries is also observed to be rightward skewed as seen in both Table 12 and Table 13. The top 20 percent of the households account for over 41 percent of income while the bottom 20 percent of the households account for about 6 percent of income as reported in Table 12. In 2009, the distribution suggests that the household incomes were more equally distributed in Scotland than the other three countries. Overall, the difference in inequality was lowest between countries in the UK in 2009.In Table 13, the top 20 percent of the households accounts for over 38 percent of income while the bottom 20 percent of the households account for about 7 percent of income. In 2017, Wales had a more equally distributed household income among all the countries in the UK. Based on the index of inequality, inequality of income was 12 percent higher in England than it was in Wales. Overall, it shows that the inequality of income has slightly increased between the countries in the UK during the period 2009-2017.

Ethnicity of Household

White ethnic group constitutes about 80 percent of the sampled households, followed by Asian or Asian British (12 percent) and Black or Black British (5 percent). Table 14 shows that Asian or Asian British ethnic group has the highest mean income (£33,836) closely followed by White ethnic group (£32,612) whereas, Black or Black British has the lowest mean income (£27,609). Table 15 reports are consistent with Table 14 where the highest mean income is earned by Asian or Asian British ethnic group (£45,123) and the lowest mean income is earned by Black or Black British (£37,610). The standard deviation of the Mixed/Multiple ethnic group has a relatively higher value (46,511) than the mean income estimate (43,684) which could mean that there is large variation in income earned by those ethnic group households.

Education of Household

Income distribution by education level (with degree and no degree) is reported in Table 14 and Table 15. A majority of households reported that they have no degree in both 2009 (78 percent) and 2017 (73 percent).Table 14 shows that the individuals in households having no degree had a significantly lower mean income (£29,265) than those households having a degree (£43963). Similarly, individuals in households with no degree had a mean income of £38,986 opposed to individuals in households with degree having mean income of £52,790. This suggests that having a degree has a significant impact on earning higher income.

The distribution of income by education level of the household head was rightward skewed. Table 14 provide us information that the top 20 percent of the all households account for over 40 percent of the total income while the bottom 20 percent of all households account for less than 7 percentof the total income. Inequality is more pronounced with households having no degree by about 5 percent relative to households having a degree, shown by the inequality index. Table 15 presents that the top 20 percent of all households account for over 40 of the total income while the bottom 20 percent of all households account for about 7 percent of the total income. These results are consistent with Table 14 which indicates that inequality has been fairly stable between 2009 and 2017. The Gini ratio and index of equality both give us the idea that both households with degree and no degree have almost the same extent of inequality.

Sex

Income distribution of household by sex is reported in Table 7 and Table 8. There are more female (54 percent) than male (46 percent) in the sampled households in both 2009 and 2017. In Table 7, the estimated mean income of males is higher (£33,548) than the females (£31,494) by about 6 percent. Similarly, as reported in Table 8, Males have mean income of (£43,210) whereas, the females have a mean income of (£40,603).

The income distribution by location was also skewed to the right in both years. In Table 7, the top 20 percent of the households received about 42 percent of the income while the bottom 20 percent had about 6 percent. The Gini ratio for male and female was almost the same which shows that there isn’t much difference in inequality between them. Similarly, in Table 8, the top 20 percent of the households received about 40 percent of the income while the bottom 20 percent had about 7 percent of the total income. The index of inequality shows 3 percent higher inequality in female than in male.

Income distribution in the UK: 2009 and 2017

The estimated mean income in 2009 was £32,427, while in 2017 it was £41,795 with an overall 28 percent increase. The estimates in Table 15 provide a unique opportunity to evaluate the change in the income distribution of the UK over a 9-year period. Based on the results in Table 15, the Gini ratio suggests that the income inequality has slightly declined by nearly 7 percent during the 8 years. Overall, it seems there isn’t much difference in the distribution of income between the two years. So, it can be concluded that the income distribution in the UK has remained fairly stable with a slight decline (authors).

Figure 2 represents the Lorenz curve of the year 2009 and 2017. The 45⁰ straight (blue) line represents the line of equality where each unit of the population receives the same income. This corresponds to the case of perfect equality of incomes. The Lorenz curve of the year 2009 and 2017 is shown by the red and green line respectively. The further the Lorenz curves away from the line of equality the higher the inequality. We can observe that both the green and red Lorenz curves are close to each other. The green curve (2017) is slightly close to the line of equality than the red curve (2009) which supports to the conclusion that inequality has slightly decline during the intervening 8 years (2009-2017).

Results and Discussion

Regression results

The data used for this regression analysis reported in this section are from 2009 and 2017 surveys. In specification 1, time and sex dummy are regressed on the log of the household income (Table 20). The time dummy is positive and statistically significant at 1 percent level, indicating that incomes in 2017 were higher than 2009 by about 29 percent. As the survey data are 8 years apart, it explains its higher growth in incomes. Household incomes (in logs) were greater by nearly 7 percent for males compared to females.

In specification 2, household size dummies were included and sex dummy were excluded from this specification but time dummy was retained. The coefficients estimated are relative to four member households. The estimated time dummy coefficient in this specification adjusting for family size becomes positive and significant at 1 percent level, which means that the incomes were higher in 2017 than in 2009 by nearly 29 percent; this was the same compared to specification 1. Relative to four-member households, the small households had significantly lower incomes. One-member households had over a 100% lower income than that of the four-member households. The larger households had incomes that were not significantly different from the income of the four-member households varying from 2 percent (five-member household) to 17 percent (7 and more member households). It seems that as the size of the household increases above four members, income don’t increase by a significant amount and remain about the same level. Larger households may have more dependent individuals like children and elderly people who do not contribute to household income.

Household size dummies were omitted and location dummies were added in specification 3. The estimated time dummy is still statistically significant and positive with adjustment of location dummies. The incomes were higher in 2017 than in 2009 by about 30 percent which is slightly higher than in specification 1 and specification 2. The coefficients estimated are deviated around households living in England. The coefficients of the location dummies show that, relative to households living in England, all other countries had lower household income and are statistically significant at 1 percent level. It appears that households of all other countries in UK (Scotland, Wales and Northern Ireland) had incomes that were not significant different from the income of households living in England.

Educational qualification dummy was introduced while the location dummies were removed from specification 4. The coefficients were relative to households having no degree. The time dummy retained its sign and significance and had the same magnitude of increase from 2009 to 2017 as was the case in specification 3. Having a degree significantly increases the income by around 37 percent than not having a degree and the coefficient was significant at 1 percent level. It seems that having a degree has a significant impact on earning higher income than not having a degree.

The ethnic group dummies were added while the educational qualification dummy was removed from specification 5. The estimated coefficients are deviated around “White” ethnic group. Only the coefficient of “Mixed/Multiple” ethnic group was insignificant. Relative to “White” ethnic group, “Asian or Asian British” ethnic group had higher income by about 7 percent. The rest of the ethnic group (“Black or Black British” and “Other”) had 15 percent lower income than “White” ethnic group.

In specification 6, only time dummy and occupation class dummies are used. The time dummy coefficient adjusting to ethnic group was positive and significant at 1 percent level. The incomes were higher in 2017 than in 2009 by about 20 percent which is lower than other specifications. Relative to unskilled households, all other types of households had higher incomes varying from 13 percent (partly skilled) to a high of nearly 60 percent (professional).

Finally in specification 7, all the dummy variables were included. The estimated time coefficient is still significant at 1 percent level. Thus, incomes in 2017 were higher than 2009 by nearly 20 percent. The estimated coefficients of sex dummy retained their sign and significance but its value decreased sharply. Incomes for males had 3.6 percent higher than incomes for females and were statistically significant at 1 percent level assuming that other variables remained constant. Similarly, the estimated coefficients for the household size dummies reflecting more than four members, increased significantly relative to the estimates in specification 2 while the opposite effect was observed for the magnitudes of coefficients for dummies for households of smaller sizes (less than four-member). The estimated location dummies are all significant except for Scotland. The coefficients for Wales and Northern Ireland are almost identical to coefficients from specification 3. The estimated coefficients for degree dummy was retained its sign and significance but declined significantly relative to specification 4. Incomes for households having degree were 16 percent higher than having no degree, while it was 37 percent higher in specification 4. The estimated coefficients for ethnic group dummies are now all significant at 1 percent level. The sign of the coefficient for “Asian or Asian British” has changed from positive to negative, indicating that “Asian or Asian British” ethnic group had lower incomes than “White” ethnic group. The estimated coefficients for occupational dummies all retained their sign and significance. The magnitudes of all the coefficients have slightly declined relative to specification 6. In short, household size, sex, location, education, ethnicity and occupation were all important in explaining the variation in the logarithm of household income in the UK.

Values for the explained variation of the log of income reported in Table 20 show that the regressions explained at most 31 percent. Although it is not surprising that the coefficients of determination R2 values are low because of the complexity of income determination. The R2 value for the last specification (specification 7) was observed to be the highest (since it contained more variables) at 0.314 which means that about 31 percent of the total variability in income is explained by the utilised independent variables.

Limitations

There are some limitations to this study. I couldn’t use years before 2009 because the particular variables that I needed for my analysis were only available in the Understanding Society (UKHLS) dataset which started from 2009 onwards to 2017. The British Household Panel Survey (BHPS), Understanding Society, UKHLS, which is a panel data panel data, I only assessed the individual years. I didn’t use fixed effects or random effects in the regression analysis as there was a 8 year gap between the years that I used (2009 and 2017)

Conclusion

This study has updated the information on the income distribution in the UK. Utilizing data from Understanding Society, UK longitudinal Study (UKHLS), the size distribution of income was investigated and related to selected variables. The pattern of the size distribution of incomes in the UK remained biased towards the top 20 percent. For sample partitions studied, the top 20 percent of the households accounted for over 40 percent of the total income in most cases. The Gini ratios in 2017 were lower than in 2009 in most of the cases, suggesting that income inequality had declined during the intervening 8 years.

Income distribution of the UK for the year 2009 and 2017 were compared to each other. Based on the Gini ratio, Theil entropy measure and Mean Log Deviation (MLD), it appears that the income inequality has slightly declined by 7 percent during the 8 years. Lorenz curve also showed that income inequality has declined by a small margin between 2009 and 2017. Based on all the results, it can be concluded that the income distribution in the UK has remained fairly stable with a slight decline.

A regression analysis applying several related specifications was utilized to estimate impacts of selected classification variables on household incomes. Household income varied considerably by household size, ethnicity, sex, location, occupation and educational qualification. Incomes in 2017 were higher than 2009, by about 20 percent. The results suggest that incomes of households having obtained a degree were 16 percent higher than that of households having no degree. This indicates the importance of having a degree in determination of income level. “White” ethnic group had the highest income among all other ethnic groups. The results of indicate that incomes of households having no skill were much less than for the other types of households. In short, In short, household size, sex, location, education, ethnicity and occupation were all important in explaining the variation in the logarithm of household income in the UK.

References

  • Al Baghal, T., A. Jaeckle, et al. (2015). Understanding Society -The UK Household
    Longitudinal Study: Innovation Panel, Waves 1-8, User Manual. Colchester,
    ISER University of Essex.
  • Bourguignon, F. & Morrisson, C., 2002. Inequality Among World Citizens: 18201992. American Economic Review, 92(4), pp.727–744.
  • Bhalla, S., 2002. Imagine there’s no country: poverty, inequality, and growth in the era of globalization. book
  • Firebaugh, G. & Goesling, B., 2004. Accounting for the recent decline in global income inequality. American Journal Of Sociology, 110(2), pp.283–312.
  • Sala-i-Martin, X., 2006. The world distribution of income: falling poverty and … convergence, period. Quarterly journal of economics, 121(2), pp.351–398.
  • Clark, 2011. World Income Inequality in the Global Era. New Estimates, 1990–2008. Social Problems, 58(4), pp.565–592.
  • Milanovic, B., 2013. Global Income Inequality in Numbers: in History and Now. Global Policy, 4(2), pp.198–208.
  • Ravallion, M., 2014. Income inequality in the developing world. Science (New York, N.Y.), 344(6186), pp.851–855.
  • Jenkins, S. (1996). Recent trends in the UK income distribution: What happened and why? Oxford Review of Economic Policy, 12(1), 29-46.
  • Brewer, M. & Wren‐Lewis, L., 2016. Accounting for Changes in Income Inequality: Decomposition Analyses for the UK, 1978–2008. Oxford Bulletin of Economics and Statistics, 78(3), pp.289–322.
  • Hills, J., Brewer, M., Jenkins, S., Lister, R., Lupton, R., Machin, S., Mills, C., Modood,T., Rees,T. and Riddell, S. (2010).An Anatomy of Economic Inequality in the UK: Report of the National Equality Panel, Government Equalities Office, London. Report
  • OECD. (2011). Divided We Stand, OECD Publishing, Paris. report
  • Ahiram, E., 1964. Income distribution in Jamaica, 1958. Social and Economic Studies, 13, pp.333–369.
  • Yu Xie & Xiang Zhou, 2014. Income inequality in today’s China. Proceedings of the National Academy of Sciences, 111(19), pp.6928–6933.
  • Banskota, Kamal; Johnson, Stanley R.; and Stampley, Gary L., “Income Distribution in Jamaica” (1987). CARD Staff Reports. 75.
    https://lib.dr.iastate.edu/card_staffreports/75
  • Mookherjee, D. & Shorrocks, A., 1982. A Decomposition Analysis of the Trend in UK Income Inequality. The Economic Journal, 92(368), pp.886–902.
  • Livada, A., 1991. INCOME INEQUALITY IN GREECE: A STATISTICAL AND ECONOMETRIC ANALYSIS *. Oxford Bulletin of Economics and Statistics, 53(1), pp.69–82.
  • Azam, M. & Bhatt, V., 2018. Spatial Income Inequality in India, 1993–2011: A Decomposition Analysis. Social Indicators Research, 138(2), pp.505–522.
  • Ipek, E. 2019. Consumption and Income Inequality in Turkey. Journal of Yasar University. 14(53). 13-20.
  • Yan Qingtao et al., 2014. A New Lorenz Curve Model Based on Income Distribution Research. 2014 International Conference on Management of e-Commerce and e-Government, pp.295–300.
  • Lin, C.-H., Lahiri, A. & Hsu, S., 2015. Population Aging and Regional Income Inequality in Taiwan: A Spatial Dimension. Social Indicators Research, 122(3), pp.757–777.
  • University of Essex, Institute for Social and Economic Research, 2019, Understanding Society: Waves 1-8, 2009-2017 and Harmonised BHPS: Waves 1-18, 1991-2009, [data collection], UK Data Service, 11th Edition, Accessed 13 August 2019. SN: 6614, http://doi.org/10.5255/UKDA-SN-6614-12

Cite This Work

To export a reference to this article please select a referencing stye below:

Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.

Related Services

View all

DMCA / Removal Request

If you are the original writer of this assignment and no longer wish to have your work published on the UKDiss.com website then please: