Why might certain sociologists prefer to measure inequality based on wealth instead of income?

Wealth inequality is in the news, with our friends at IPPR pointing to just how ‘unevenly divided’ wealth is in the UK. The problem is huge: the top 10% of households are 875 times wealthier than those at the bottom.

But why is there so much focus on wealth inequality – and what’s the difference between that and income inequality?

Personal wealth means a stock of valuable possessions: anything from cash under your mattress, through shares and bonds, to the value of your house or your car. Income, on the other hand, is a flow of money you receive, such as wages for employment.

Here in the UK, we’ve heard talk that inequality hasn’t increased since before the financial crisis. Claims like that refer to income inequality. In July 2017, the Institute for Fiscal Studies (IFS) caused a stir with a report showing that the gap between the highest wages and the lowest has not changed much since 2008.

Statistics on income inequality risk misinterpretation. Although it has fallen by some measures, that doesn’t mean that those at the bottom are doing any better. In fact, as the Resolution Foundation revealed in a report for the Social Mobility Commission today, people on low pay are increasingly finding themselves stuck there, unable to ‘escape’ to better employment. The results in the IFS report are due mostly to salaries in the financial and insurance services sector, which are among the highest, falling dramatically after the crash; it also showed that those same salaries have been climbing faster over the past couple of years.

Wealth inequality is much more severe than income inequality. A tiny fraction of the population owns most of the UK’s pile of riches. In our recent work, we found that, between 2006-8 and 2012-14, the richest fifth of households gained almost 200 times as much in absolute wealth terms compared to the poorest fifth.

So, irrespective of the story on incomes, Britain is becoming much more unequal. Once we consider the consequences of wealth inequality, there’s much more cause for concern.

In the first place, wealth is itself a source of income. Holding stocks and shares on financial markets guarantees a source of income in the forms of dividends and capital gains; holding bonds or savings generates interest. The effect goes further: wealth allows people to purchase better healthcare and education, and assets like a house or a car themselves enable people to save time and take on better jobs (this article over at Quartz has a great summary of this point). Income can be stored as wealth, but wealth begets income.

This means that wealth is stockpiled by the rich and inequality gets worse over time, as Thomas Piketty’s groundbreaking book Capital in the 21st Century outlined with painstaking historical clarity. Since the return on capital (wealth) is higher than the rate of economic growth in general, wealth comes to dominate wages as the determinant of how prosperity is shared.

As the authors at IPPR point out, these facts have a necessary intergenerational bite to them. ‘Every generation since the ‘baby boomers’ now has less wealth than the generation before them had at the same age.’ With policies like exemptions for inheritance tax and the crushing weight of the housing market bearing down on young people, wealth inequality doesn’t only reinforce itself within the same cohort – it can multiply to appalling levels from one generation to the next.

Conservative politicians tearing their hair out over attracting younger voters would do well to take a long, hard look at wealth inequality statistics. Income inequality threatens to deteriorate. But the real news is in wealth and the patent unfairness associated with it.

8.What is the difference between income and wealth? Why might certain sociologistsprefer to measure inequality based on wealth instead of income?9.What is “structural mobility” and how does this concept describe the decline of

Get answer to your question and much more

manufacturing job in the US since early 1970s?

Get answer to your question and much more

10. According to Max Weber, what do being a teacher, having a cool car, and being amember of a prestigious association have in common?

Get answer to your question and much more

In this paper, we used the data of the Survey of the Household Finance and Living Conditions (SHFLC), which is the linked data of the survey data and administrative data. The SHFLC has been collected by the Statistical Office in South Korea since 2010. The Statistical Office of South Korea began to collect information about household income, debt, wealth, and welfare to know the dynamics of finance after the global financial crisis in 2008–2009. The sample size was 10,000 cases until 2011, and after that, it increased to 20,000 households since 2012. The SHFLC has been replaced by 20% of the total sample every year. Thus the SHFLC will be utterly new after 5 years.

In this paper, we used SHFLC 2017 with the sample size 18,497 in 2017. The Statistical Office in South Korea began to link the SHFLC with the administrative data to overcome shortcomings of the survey data on income. Linking survey data with registered administrative data provides a new research possibility on inequality research with more accurate information on personal and household income. Administrative registered data in the SHFLC included the various data source to measure household income: the annual income and tax from the National Tax Office, the transfer income and pension from the Social Insurance Office, the health insurance and health service information from the National Health Insurance, educational allowance from the Ministry of Education, and child allowance from the Ministry of Health and Welfare.

Table 1 reports different measures of inequality of income and wealth in the survey data and the linked data in 2017. As we expected, the linked data displays a higher maximum income than the survey data, revealing that the survey data underreport the income of the top income. It also shows that the survey data do not accurately identify the income of the low-income groups since the median shifts from 3698 to 3966. As a result, Gini coefficients measured by the linked data, 0.4416, is higher by 4.25% than that of the survey data, 0.4236.2 The linked data that uses income from the administrative record shows the fact that the survey data collected by the Statistical Office in South Korea tend to underestimate the level of the top income and the degree of inequality of the household income. The concentration of income on the top 10% and top 20% of income groups also increases from 0.2883 in the survey data to 0.3165 in the linked data. The share by the top 10% and the top 20% also show a similar pattern of the rising concentration of income. However, p90/p10, which indicates the income gap between the income of the lowest 10% and the highest 10% income group, shows a slightly different pattern, by reducing the ratio by 1.85%. On the contrary, the gap between the poor 40% and the richest 10% widen when the linked data is used. The Palma ratio, the ratio of the richest 10% of the households’ share of income divided by the poorest 40%, shows an increase from 2.4542 to 2.6421.

Table 1 Measurement of income inequality and wealth inequality, 2017

Wealth inequality is much higher than income inequality across the whole country (Sierminska, Smeeding, and Allegrezza, 2013; Jäntti, Sierminska, and Kerm, 2013). On average, the popular perception of the rising inequality has been derived by not only income inequality but also wealth inequality. Wealthy persons consume luxury goods and cars and live expensive housings in gated communities. Also, the increasing housing price contributes to the perception of wealth inequality as well as wealth inequality itself. In South Korea, the rising housing price in recent years has contributed to the fear of the youth of the middle class and the working class, as housing ownership was very low among young adults (see Fig. 1). The Korean youth has shown the lowering marriage rate and birth rate of the married couple mainly due to the housing shortage and housing bubbles over the decades as well as the high rate of unemployment (Kim, 2017).

Fig. 1

Distribution of earnings and housing ownership, 2017. Notes: Earnings ratio refers to the ratio of earnings relative to average earnings, and the ratio of housing ownership indicates the proportion of housing owners among each age group

In this paper, income is measured by annual household income containing the market income and public transfer income. The market income includes earnings from work, profits from business, property income such as interests or rents, and the private transfer. Pre-tax income will be used in the following to focus on the impacts of contributors to before-tax income inequality. Wealth is measured by the price of a variety of possession, including properties such as houses, lands, buildings, and cars. Linking the survey data with the administrative data provides more accurate data on income and wealth than any other data set in South Korea.

The joint distribution of income and wealth shows highly polarized distribution, characterized by the two poles around income poor-wealth poor and income rich-wealth rich. Figure 2 displays a topographic picture of the joint distribution of income and wealth in 2017. We can see that the lowest 5% income and wealth group displays one pole with 3.56 times higher proportion than the average. The highest 5% of income and wealth shows 7.19 times higher proportion than the average. The lowest 10% and the highest 10% in both income and wealth distribution show the second-highest density. However, in the middle of income and wealth distribution in Fig. 2, there is more frequent mobility than any other parts of the joint distribution of income and wealth. It shows that there is a low level of association between income and wealth in the middle of income group and in the middle wealth group. High income and high wealth clusters are formed in the poles.

Fig. 2

Joint distribution of income and wealth in 2017

Table 2 reports the summary of the description of major independent variables. Those are mostly the characteristics of the head of households except family size. Thus, sex in the SHFLC refers to the sex of the household head. The majority of the household head is male (75.51%), relative to females (24.49%) in South Korea. The measurement of education is done by the calculation of years of education rather than categories based on the level of education. Family size refers to the number of family members. In the subsequent analysis, family size is categorized from one to six since there is a non-linear relationship between family size and income. Occupation includes non-working people as well as the working population since it deals with the entire household. Employment status also includes the category of non-working people. Soldiers in the occupational category are excluded in the following analysis because the market does not determine their incomes.

Table 2 Descriptive summary of major variables (N = 18,497)

To identify the extent to which some factors contribute to inequality of income and wealth, we use the regression-based inequality decomposition developed by Gary Fields (2003).

Assuming that the income is

$$ \ln\ y= X\beta +\varepsilon $$

(1)

where y is an n × 1 vector of incomes; X is an n × (K + 1) matrix of individual and household characteristics (age, education, employment status, occupation, household size, etc.) including the constant; β is a (K + 1) × 1 vector of coefficients, and ε is an n × 1 vector of residuals. A sample of observations {yi, xi, i = 1, 2,...n} can be used to estimate the model.

The linear model (1) can be rewritten as:

$$ \ln\ y={\beta}_0+{\beta}_1{X}_1+{\beta}_2{X}_2\dots +{\beta}_{\mathrm{K}}{X}_{\mathrm{K}}+\varepsilon $$

(2)

$$ ={\beta}_0+{Z}_1+{Z}_2\dots +{Z}_{\mathrm{K}}+\varepsilon $$

(3)

where each Zk is a “composite” variable, equal to the product of a regression coefficient and its variable (Zk = βKXK), with k = 0, 1,...K and X0 = 1.

For inequality decomposition calculations, the value of β0 is irrelevant as it is constant for every observation.

The OLS estimate of (3) can be used for inequality decomposition:

$$ \mathrm{lny}={\mathrm{b}}_0+{\hat{b}}_1{X}_1+{\hat{b}}_2{X}_2\dots \kern0.5em +\cdots +{\hat{b}}_k{X}_k+\hat{\in} $$

(4)

Marginal contribution of the kth factor ($ {\hat{Z}}_k $ = $ \hat{b_k}{X}_k $) to the variance of logarithm of income is as follows:

$$ {S}_k=\hat{b_k}\mathit{\operatorname{cov}}\left( Xk,\ln y\right)/{\sigma}^2\left(\kern-0.5em \ln y\right) $$

(5)

$$ {\sum}_{i=1}^k{S}_k={\sum}_{i=1}^k{\hat{b}}_k\mathit{\operatorname{cov}}\left( Xk,\ln \kern0.5em y\right)/{\sigma}^2\left(\kern-0.5em \ln \kern0.5em y\right)={\sigma}^2\left(\kern-0.5em \ln \hat{y}\right)/{\sigma}^2\left(\kern-0.5em \ln y\right)={R}^2 $$

(6)

The less $ \hat{\in} $ in Eq. (4), the larger the explanatory power (R2).