Inequality and wage differentials

Defining inequality

Economic inequality refers to the differences that exist between identifiable groups of people within a larger population, or class, or that exists within a particular group. Economic inequality can be assessed in terms of narrow economic measures, such as income and wealth and material standards of living, or broader ones, such as quality of life, and economic welfare.

The study of inequality begins with the identification of groups and then selecting the criteria for assessing inequality. For example, we can take pay as a potential source of inequality, and identify age as a broad demographic group to consider. We can then create separate ‘age’ ranges, such as under 18, 19 to 24, 25 to 39, and so on, and then compare differences in pay between these groups – say between 19-24 year olds and 25 to 39 year olds.

We can also identify differences within a group – such as differences in pay that exist within the 19-24 year olds group. In fact, we can keep zooming in, to get even smaller and smaller sub-groups to make comparisons.

As well as age, characteristics identified where inequalities may exist include gender, social class, ethnic background, and education level.

This list is not exhaustive, especially when looking at sub-categories within a group. For example, it is common to consider differences related to race and ethnic background, but ethnicity has a wide variety of sub-sets.

Difficulties in defining groups

Identifying discreet groups for comparative purposes presents some difficulty. While there are ‘objective’ ways of identifying groups, such as using age or country of origin, it becomes far more complex when elements of subjectivity are included, such as when individuals ‘self-identify’ as being in a particular category, or where groups are defined too broadly. For example, the now commonly used BAME (Black, Asian and Minority Ethnic Groups) acronym is an example which illustrates how groups can be identified in ways which make comparisons difficult. For example, it is possible that inequality within a particular subset – say Asian – is statistically greater than inequality between Asian, and other ethnic groups. Indeed, a recent report in the UK questioned the validity of creating overly broad groups such as BAME for comparative purposes.

Opportunity vs outcome

Economic inequality can be classified in terms of inequality of opportunity – also called ‘upstream’ inequality – and inequality of outcome – also called ‘downstream’ inequality.


Inequality of opportunity refers to differences in terms of access to the systems of reward that exist within a particular economy. For example, if pay and wages are the primary reward system, then inequality of opportunity relates to differential access to the jobs that provide the highest reward. This could be as a result of barriers to entry into higher paid work for a specific group, or as a result of barriers within firms to promotion to higher paid jobs.


Inequality of outcome – or downstream inequality – refers to differences in the rewards or benefits going to different groups, and includes inequality of income and wealth.

Clearly, the two dimensions of inequality are related – inequality upstream makes inequality downstream more likely.

In the following analysis we focus on inequality of income - rather than wealth. This is primarily because the measurement of income is more straightforward than the measurement of wealth. Wealth is often hidden, and its value frequently changes, in comparison with income which is more measurable and has a more stable value.

Inequality of income

There are two main sources of income – earned and unearned.

Earned income

Earned income refers to earnings gained by supplying labour, and include basic wages, commissions, and royalties from creative work.

Unearned income

Unearned income is income other than that from normal employment, and includes earnings from renting property, interest from personal saving, dividend payments distributed by firms, and income from other securities, such as bonds.

Consumption or income?

Some economists4 suggest that differences in consumption levels are a better indicator of ‘persistent’ inequality. One reason for this is that consumption can approximate to the idea of ‘permanent’ income. Permanent income is the average income resulting from past, present and future incomes. Over time, consumption can be stabilised by using previous income which is saved, or future income which is borrowed against. Hence, any sudden or random changes in income – such as when experiencing temporary unemployment – can be smoothed with current consumption less effected.

Measuring inequality

Equivalised disposable income is frequently used when measuring inequality of income1. Equivalised means that household income is adjusted to take into account the number of people in a household.

The Gini co-efficient

The Gini coefficient is a widely used indicator of income inequality, and is a value that lies between 0 and 1 - the higher the number, the greater the inequality, so that complete (or perfect) inequality has a value of 1, with a perfectly equal distribution of income having a value of 0. [Read more on the Gini co-efficient].

However, reliance on just one statistic may miss important changes in inequality, especially within a given group. For example, it is possible that, over a period of time, the Gini coefficient remains largely stable while the income gap between people within a particular group (say females) widens (or even narrows). Indeed, even when a particular decile (say the top 10%) is looked at, inequality might have remained stable, but zooming in on the top 1% could show increased inequality.

Hence, having a range of measures could help uncover more information about inequality.

Other measures

There are several other indices that attempt to measure inequality, including:

The Robin Hood index

The Robin Hood index (also known as the Hoover index, and Schutz index) shows the proportion of all income that would have to be redistributed to achieve a state of perfect equality. This means that it estimates the amount of income that must be transferred from those on higher incomes to the lower income earners to achieve perfect equality. The higher the value, the more redistribution is required.

The Lorenz curve

Graphically, inequality can be illustrated through Lorenz curves. The position of a Lorenz curve relative to the 45 degree line - which represents a perfectly equal distribution of income - indicates the degree of inequality existing. In the following graph, Lorenz curve 2 indicates greater inequality than Lorenz curve 1.

effect of eduction on wages

[Read more on the Lorenz curve].

The Palma ratio

The Palma ratio is the ratio of the national income shares going to the top 10 per cent of households compared with the bottom 40 per cent. It is based on the observation that changes in income differences at the extremes (the top and bottom) gives a better clue of changes in the overall level of inequality, given that the ‘middle’ groups are more stable over time.

The S80/S20 ratio

Similarly, the S80/S20 ratio compares the ratio of the average income of the richest 20 per cent of the population to the average income of the poorest 20 per cent.

UK inequality between 2000 and 2019

Indices of inequality

For the UK, all measures used suggest a broadly stable level of inequality over the last twenty years. The impact of COVID-19 on inequality is yet to work its way into the statistics.

It has been noted that the inequality growth of the last forty years is mostly attributable to growing gaps within groups rather than between them2.

US inequality between 1991 and 2018

US inequality 1991 to 2018

For the US, the Gini coefficient shows an increase over the last 30 years. Like other countries, the impact of COVID-19 on inequality is yet to work its way into the statistics.

Consequences of inequality - costs

Inequality brings with it some significant costs to an economy, including:

Lower growth - as those at the lower end of the distribution of income lack purchasing power, which translates into lower aggregate demand and national income.

Opportunity cost – those undertaking low skilled jobs or remaining unemployed or underemployed could contribute more and add more value to the economy. Hence, an economy may experience a considerable opportunity cost by denying access to reward systems to certain groups or sub-groups.

Loss of efficiency - it is likely that inequality of opportunity has a negative effect on efficiency – again, in large part because of low skill levels, and the inability to use technology.

Externalities - it is also likely that there is a direct rise in externalities associated with inequality, such as those resulting from social deprivation, poor health, and a propensity to engage in criminal activities, and political and social unrest.

Impact on mobility - mobility of labour and social mobility, are also likely to be adversely affected. Immobility is likely to be both a cause and consequence of inequality.


Incentives allow markets to work

While inequality has costs, it can also be argued that market systems rely on inequality to function. Market based economies allocate scarce resources through the operation of the price mechanism. An essential element in the allocation of resources to meet the wants and needs of consumers is the role of incentives. Incentives also play a role in how labour is employed and rewarded.

Example - Netflix

The demand for labour is derived from the demand for goods and services. As tastes and preferences change - such as a preference for the convenience of watching films at home - demand for programming also changes, which then affects the demand for workers in this sector.

New entrepreneurs are likely to enter the market - as Netflix did some 20 year’s ago – as a mail order company hiring out DVDs. Allowing customers to hire DVDs online enabled the company to increase its efficiency compared with market leaders, Blockbusters, which eventually filed for bankruptcy in 2010. Netflix introduced its streaming service in 2007, and today (2021) controls some 15% of global internet bandwidth3.

In search of increased profits, or simply survival, other providers responded by introducing their own subscription channels, Netflix began to produce its own ‘original’ output. This worked its way back into the relevant labour markets, with increased demand for producers, screenwriters, designers, actors and so on. In short, the greater the demand for Netflix services, the greater the value of designers, actors, and screenwriters to Netflix. This drives up pay, which provides an incentive for existing workers to develop their skills even further, and provide an incentive to those still in education to train and acquire the skills in current demand.

The entry of Amazon, and then Apple, into TV streaming saw the rise of the super-writer, which can be traced back to Netflix’s announcement in 2017 that it was poaching Shonda Rhimes (responsible for Grey’s Anatomy and Scandal) from ABC in a deal worth $150 million12.

Demand for screenwriters following an increase in demand for streaming services

wages of screenwriters

Given that the supply of elite screenwriters is limited and very inelastic, there is a large impact on wages as a result of increased demand.

Also, as tastes and preferences change, some workers suffer from falling demand. The outcome of this is that wages fall relative to others - also increasing inequality.

Hence, wage inequality provides incentives to the suppliers of labour just as they do in terms of the suppliers of products, indicating that some inequality in the rewards to labour is required for the efficient use of labour. This is known as the equity-efficiency trade-off – if all inequalities were removed there would, it is argued, be a loss in efficiency.

Wage differentials which result from the operation of free markets provide a clear insight into the larger issue of inequality of income. However, as is the case with product markets, labour markets can fail to achieve an efficient allocation of scarce resources.

Causes of inequality

The Netflix example provides a modern illustration of how product and labour markets interact to provide incentives to producers and workers through differential rewards.

When looking at a particular group or comparing groups it is likely that there are many cause of inequality within and between these groups. The workings of the labour market are likely to provide, perhaps, the most important clues about the causes of inequality.

Education and the wage premium

The gains from education are often expressed in terms of a ‘wage premium’ – education provides knowledge and skills, and the more highly educated the greater the skills. This can be illustrated with a standard demand and supply of labour diagram.

effect of eduction on wages

The wage premium is the increase in wages (from W to W1) as a result of workers gaining knowledge and skills through effective education and training. Technological change has meant that those with relevant skills in using technology can gain from greater employability, higher current wages, and promotional prospects. This has led to an increase in the value of human capital for certain high skilled jobs which has widened the gap between skilled and unskilled workers.

Skills and market failures

Education and training are regarded as merit goods, and as such are unlikely to be fully provided by firms. This means employers are under no obligation to educate or train above the level required to satisfy the immediate needs of the firm. This means that, while firms may train for the development of specific skills relevant to the firm or industry – there is no incentive to educate or train in terms of ‘generic’ skills, such as communication skills, numeracy, literacy and general IT skills.

The problem of ‘poaching’ provides an insight into the lack of incentives to train. It is not commercially rational to provide training in skills which then make an employee more likely to leave a firm in search of higher pay. These workers will improve the competitiveness of rivals. Hence, the poacher can wait for the new employee to be trained (at a cost to the employing firm) and then poach them with a pay rise (justifiable because they have saved on training costs) - labour market poaching is an example of the 'free-rider' problem.

Hence, free labour markets might fail to educate and train, so that continuing education is restricted and workers may not acquire new skills.

Dual labour markets

Dual labour market theory suggests that there is a growing gap between those working in the official labour market – when employment is more secure and generally better paid – compared with those working in the unofficial labour market doing ‘non-standard’ jobs.

The rise of the ‘gig’ economy has led to a growing number of workers whose income and benefits lag behind those in the official-formal labour market. In the UK alone, nearly 5m people work in the gig economy, providing services to digital platforms, including Uber, Just East and Deliveroo, but often working on zero-hours contracts5.

Rise of the ‘elite’ worker

There are certain industries where workers can derive very high earnings, including sports and, as we have seen, arts and entertainment. Sports companies and entertainment organisations, as well as global businesses, seek out the most gifted and talented from around the globe. 

Research has shown that around 20% of top earners in the most developed countries work in the financial sector, with those being rewarded with a mix of pay and stock options. The rise of remuneration through the use of stock options can be seen as an attempt by large corporations to solve the ‘principal-agent’ problem (also called the ‘agency problem’). The problem arises from the separation of ownership and control, with the managers of corporations able to introduce their own objectives – such as achieving an increased salary – while the goal of profit maximisation is sacrificed. Stock options ‘tie-in’ managers and align them more closely to the objective of shareholders6.

Elite pay can also be understood through the lens of the Lemons Problem – where pay in large corporations is set well above the going market rate to ‘screen out’ applications from the less qualified (‘poor quality’ applicants).

Barriers to entry

Barriers to entry into the labour market may explain some of the differences in pay. Barriers to entry include discrimination against a particular group of employees, by age, race, gender or other human characteristics. Despite the fact that most advanced economies have legislation in place to deter and penalise discrimination against particular groups, it is likely that some groups are denied full and free access to higher paid, and elite, employment. The same is likely to be true for promotion at work, with the so-called ‘glass ceiling’ preventing women from access to higher paid work.

Participation rates

Barriers into the labour market may lead some not to participate in the labour market, and thus do not get access to higher incomes.

The impact of monopsonies

Individuals employed in markets dominated by single buyers of labour - monopsonists - are likely to experience lower pay relative to more competitive labour markets.

Protection of labour

Trade union membership can have an impact on pay, and hence on the distribution of income between ‘unionised’ and ‘non-unionised’ workers. The difference is called the union wage premium, or the union mark-up. Declining membership of labour unions has reduced the bargaining power of organised labour. In the UK, the union mark-up has declined from around 10% to 5% over the last 30 years11.

Globalisation (see below) has also contributed to this, given that multi-nationals often locate where union influence is minimal.


Those who are unemployed will, at least during their period of unemployment, receive a relative low personal income. Of course, long term unemployment can have a significant impact on ‘permanent’ income. The rate of unemployment may vary between different groups, and add to inequality.

Not only does unemployment lead to a fall in family incomes, it leads to the shrinking of a worker's skills during the time they are unemployed. Those experiencing long term unemployment are likely to suffer most from a loss of skills, which makes it harder to acquire work, and especially work which uses the newest technology.

Other factors

Globalisation and free trade

Globalisation and free trade may also have contributed to rising inequality. Many developing countries have specialised in low-value technology manufacturing which, it may be argued, has resulted in lower skilled jobs being exported from developed to developing countries. This suppresses wages of the unskilled in developed economies, again increasing the inequality facing unskilled workers.

In short, globalisation has pushed up the remuneration of skilled or highly skilled labour, and for those that own capital7.

Taxes and benefits

Changes in taxation and benefits can also widen or narrow the income gap. Increases in the ‘progressiveness’ of the tax system will reduce disposable income for higher income earners relative to those on low incomes, or receiving benefits. Regressive taxes, such as sales taxes and VAT will impact the lower paid. Hence, changes in the degree of progression will have a redistributive effect between different groups.

One reason for the rise in inequality in the UK during the 1980s was the effect of the reduction in progressiveness of the income tax system resulting from the removal of  the highest tax bands.

Policies to reduce income inequality

Policies to reduce inequality can either be applied ‘downstream’ by redistribution to reduce the effects of inequality, or ‘upstream’ at the source of the inequality to prevent it happening – such as be opening up more opportunities to earn income. Policies can also be used to regulate how specific markets work8.

Policies focusing on upstream inequality

Education and skills

Improvements in education and skills in order to raise the level of achievement, and enable young people to leave education with a set of key skills will enable them to work in higher value-added, higher paid employment. This could also involve more effective career’s advice9.

Reducing discrimination

Reducing discrimination against particular groups through a combination of legislation, incentives and information, will encourage more social mobility and increase access to the upper-end of the reward system. Schemes to reduce workplace discrimination, and encourage diversity, such as training to reduce unconscious bias against certain groups, will not only encourage movement up the promotional ladder, but will encourage those from disadvantaged backgrounds to enter the labour market.

Increasing labour market participation.

As will all such policy interventions, there may be considerable time-lags between the implementation of any policy, and possible effects on downstream inequality, especially initiatives involving educational provision.

Policies focusing on downstream inequality

Redistribution through taxation

Increasing the progressiveness of the tax system to ensure that the marginal tax rate is rising will redistribute income from the higher paid to the lower paid. However, the danger is that a high marginal tax rate will have disincentive effect on work and effort. [See the Laffer curve].

Another criticism is that high marginal tax rates may deter inward investment, which may have a negative knock-on effect on efficiency and productivity.

Increased welfare payments

Increasing the level of welfare payments may reduce income inequality. Welfare benefits include direct cash benefits, tax credits, and 'benefits in kind' such as assistance with housing costs. These vary considerable across developed economies. Many countries distinguish between 'family' benefits, which involves payments to those with children to support, and 'social assistance' which can be provided to everyone. Perhaps the most significant criticism of increasing welfare benefits is that it can create moral hazard, and trap individuals into long term welfare dependency.

Minimum wage

Setting a minimum wage, or raising its level, will impact on the low paid, while having no direct effect on the higher paid. The main argument against an increase in the minimum wage is that it can create a ripple effect, with wages increasing to maintain differentials. If this happens then any reduction in inequality may only be temporary. [See minimum wage.]

Policies to reduce unemployment

Given that unemployment may be a significant cause of inequality, policies to reduce unemployment should also reduce inequality. Policies might include a mix of fiscal, monetary and supply-side policy. [Read more on unemployment.]

Regulation of the gig economy

Regulation of the gig economy, to increase worker protection and extend employment rights, has been considered and implemented in several major economies, including the US, the UK and the EU10.

However, as with other types of intervention, governments are wary of over-regulating the gig economy as this might stifle innovation, and drive entrepreneurs to move to countries with looser regulatory regimes.


While incentives play an important role in the efficient operation of market forces, making some income inequality functional, there is clearly a 'tipping point' beyond which it becomes dysfunctional in terms of inefficiency, opportunity cost, and low economic growth. What is clear is that the reward to labour has declined relative to that of capital.

There is also evidence that inequality within a particular group is just as great as between groups. Looking at the labour market provides important clues to the origins of inequality, but perhaps even more important is to go back a step, and look at education, and the education system. While policies can reduce downstream inequality, improving equality of opportunity provides the most effective way of reducing inequality in the long run.

Laffer curve and policy

How useful is the Laffer curve?

Laffer and policy
Supply-side policy

How effective is supply-side policy?

Supply-side policy
Monetary policy

How effective is monetary policy?

Monetary policy

Sources and endnotes

1. BIS Working Papers No 654 World changes in inequality: an overview of facts, causes, consequences and policies by François Bourguignon Monetary and Economic Department August 2017)

2. An Anatomy of Economic Inequality in the UK - Report of the National Equality Panel 2010

3. Netflix facts

4. Trapeznikova, I. Measuring income inequality. IZA World of Labor 2019: 462 doi: 10.15185/izawol.462

5 Euronews report -

6. Keeley, Brian (2015), “Why is income inequality rising?”, in Income Inequality: The Gap between Rich and Poor, OECD Publishing, Paris. DOI:

7. BIS Working Papers No 654 - World changes in inequality: an overview of facts, causes, consequences and policies by François Bourguignon Monetary and Economic Department August 2017

8. BIS Working Papers No 654  - World changes in inequality: an overview of facts, causes, consequences and policies by François Bourguignon Monetary and Economic Department August 2017

9. Good work - The Taylor Review of Modern Working Practices, July 2017,

10. Fortune -

11. Trade Union Report 2015 -

12. As reported in Fast Company -