Economics of inequality and poverty

• Distinguish equality from equity, and income inequality from wealth inequality.
• Draw a Lorenz curve, show a change in the distribution of income on it, and interpret a Gini coefficient.
• Distinguish absolute from relative poverty and explain how international poverty lines, minimum income standards and the Multidimensional Poverty Index measure poverty.
• Explain why poverty is difficult to measure.
• Explain each main cause of inequality and poverty as a chain ending in incomes.
• Evaluate the impact of inequality on economic growth, living standards and social stability.
• Classify taxes as progressive, proportional or regressive, and as direct or indirect, and evaluate the role of taxation in reducing inequality.
• Evaluate each of the six further policies against poverty and inequality: whom it reaches, how quickly it works and what limits it.

Equality means that everyone receives the same amount or is treated in the same way. It is a description that can be measured: two households either have the same income or they do not. Equity means fairness. Whether a distribution is fair is a normative judgement, so two people can look at the same figures and disagree about whether they are equitable.

The two ideas are related but they are not the same. A society can judge some inequality to be fair, for example higher pay for longer hours, longer training or greater risk. It can also judge an equal outcome to be unfair, for example the same pay whatever the effort. In practice, however, a lack of equity is usually discussed in terms of inequality: a government that says it is pursuing equity normally means reducing inequality of income, wealth or opportunity to a level its society regards as fair. Greater equality is therefore the usual working measure of greater equity, but more equality is not automatically more equitable, and the two words should not be used as if they meant the same thing.

Economic inequality is the degree to which income and wealth are unevenly distributed among the people or households of an economy. It has two parts, which are measured separately.

• Income is a flow: the money a household receives over a period of time. It is earned from the factors of production the household supplies (wages for labour, rent for land, interest for capital, profit for entrepreneurship) or received as transfer payments such as pensions. Income is unequally distributed when some households receive a much larger share of total income than their share of the population.
• Wealth is a stock: the value of the assets a household owns at a point in time (property, shares, savings, a business, pension rights) minus its debts. Wealth is unequally distributed when the ownership of assets is concentrated in a small share of households.

Income and wealth are linked in both directions. Wealth produces income, because assets earn rent, interest, dividends and profit. Income that is not spent is saved and becomes wealth. Wealth is normally distributed far more unequally than income, for three reasons: it accumulates over a whole lifetime; it can be inherited, so a gap can pass from one generation to the next; and it earns income that can itself be saved, so an initial gap tends to widen. The two can also diverge for one household: a young professional may have a high income and, because of student debt, almost no wealth, while a retired home owner may have a low income and substantial wealth.

Great Britain shows the size of the difference. In April 2018 to March 2020 the wealthiest 10% of households held 43% of all household wealth, and the bottom 50% of households held 9%. A policy that acts only on income therefore leaves the most concentrated form of inequality largely in place.

Inequality of income is measured by ranking households (or people) from the poorest to the richest and asking what share of total income each part of the population receives. If the poorest fifth of households receive 5% of total income and the richest fifth receive 45%, income is unequally distributed; if every fifth received 20%, it would be equally distributed.

The Lorenz curve shows this information in a diagram. The horizontal axis measures the cumulative percentage of the population, starting with the poorest. The vertical axis measures the cumulative percentage of total income that those people receive. Each point on the curve is a statement of the form 'the poorest 40% of the population receive 15% of total income'. The curve starts at the origin, because 0% of the population receive 0% of income, and ends at the top-right corner, because 100% of the population receive 100% of income.

If every household had the same income, the poorest 20% would receive 20% of income, the poorest 40% would receive 40%, and so on. The curve would be a straight 45-degree line, called the line of equality (or line of perfect equality). In reality the population is ranked from poorest to richest, so the first groups always receive less than a proportionate share. The Lorenz curve therefore lies below the line of equality and becomes steeper as it moves to the right, where each additional group adds more income than the last. The further the curve bows away from the line, the more unequal the distribution.

A change in the distribution of income moves the curve. If the share of income received by lower-income households rises, for example because progressive taxes and transfer payments redistribute income or because the wages of the lowest paid rise faster than other incomes, the cumulative share at each point is higher and the Lorenz curve shifts towards the line of equality. If the share received by the highest-income households rises, the curve shifts away from the line. The same diagram can be drawn for wealth, with the cumulative percentage of wealth on the vertical axis.

The Gini coefficient summarises a Lorenz curve in one number. It is a ratio of two areas: the area between the line of equality and the Lorenz curve, divided by the whole area under the line of equality. If income were shared equally, the curve would lie on the line, the first area would be zero and the coefficient would be 0. If one household received all the income, the curve would run along the bottom and the right-hand side of the diagram, the two areas would be the same and the coefficient would be 1. Actual values lie between these limits. The coefficient is also published on a scale from 0 to 100, as a percentage or as the 'Gini index'. A higher value means a more unequal distribution, and a Lorenz curve that shifts towards the line of equality gives a lower value.

The Gini coefficient has to be understood and interpreted, not calculated, so the area ratio is something to explain in words and not a formula to apply to data.

• A value has meaning only in a comparison: with another country, with another year, or with the same country before and after taxes and benefits. Calling a single figure 'high' without a comparison is not interpretation.
• The value depends on what is measured. Income before taxes and benefits is more unequal than income after them, and wealth is more unequal than income, so the definition has to be stated with the figure.
• Two different distributions can have the same Gini coefficient, because Lorenz curves can cross. One country may have an unusually poor bottom fifth and another an unusually rich top fifth, and the single number does not show which.
• It measures how income is shared, not how much income there is. A low-income country can have a low Gini coefficient while most of its people are poor, and a fall in the coefficient does not show that poverty has fallen.
• The survey data behind it are weakest at the two ends of the distribution: very high incomes are under-reported, and income from informal work or received in kind is hard to record.

Poverty and inequality are different questions. Inequality is about how income and wealth are shared across the whole population. Poverty is about whether the people at the bottom have enough. An economy can have little inequality and widespread poverty, or great inequality and almost no one without the necessities of life.

Absolute poverty is having an income (or level of consumption) too low to pay for the basic needs of life: enough food, clean water, shelter, clothing and basic health care. It is measured against a fixed line set at the cost of those needs, and the line is changed only to allow for changes in prices. Because the line is fixed in real terms, absolute poverty means the same thing in different years and, when incomes are converted at purchasing power parity, in different countries.

Relative poverty is having an income so far below the typical income of one's own society that a person cannot afford the way of life that is normal there. It is measured against a line set as a proportion of typical income, by convention a fixed share such as half or 60% of the national median household income. The line is not fixed in real terms: it rises whenever median income rises.

Three consequences follow from the definitions. First, economic growth that raises all incomes can reduce absolute poverty on its own, but it reduces relative poverty only if the incomes of the poorest grow faster than the median. Second, relative poverty exists in every economy with unequal incomes, including high-income countries where almost no one is below an absolute line: a person can be relatively poor in a rich country while living far better than a person who is absolutely poor elsewhere. Third, a fall in relative poverty is not always good news. In a recession median income may fall faster than the incomes of the poorest, so fewer people are below the relative line although nobody is better off.

Poverty is measured either with a single indicator, which uses one variable (almost always income or consumption) and one line, or with a composite indicator, which combines several variables in one measure. Two single indicators and one composite indicator are widely used.

The international poverty line is a single indicator of absolute poverty set by the World Bank. It is an amount of income (or consumption) per person per day, expressed in US dollars at purchasing power parity so that it buys roughly the same basket of goods in every country. A person living on less is counted as being in extreme poverty. Since June 2025 the line has been $3.00 a day at 2021 purchasing power parity; it replaced a line of $2.15 a day at 2017 purchasing power parity. The World Bank also publishes higher lines for richer groups of countries: $4.20 a day for lower-middle-income countries and $8.30 a day for upper-middle-income countries. Because the line is revised when new price data become available, a count of the poor must always be quoted with the line and the year it refers to.

A minimum income standard is a single indicator set for one country. It is the income that a household of a given type needs in order to reach a standard of living that people in that society regard as the acceptable minimum. A standard of this kind is typically built in three steps. Groups of members of the public agree which goods and services a household needs: not only food and housing but, for example, transport, heating and the means to take part in social life. Researchers then find the cost of that basket. The result is stated as the income that each type of household (a single adult, a couple with two children) needs to afford it.

A minimum income standard differs from an international poverty line in three ways. It is much higher, because it covers participation in society and not only survival. It is defined by public agreement about what is acceptable, not by experts' estimates of physical needs. And it belongs to one country at one time, so it shows how many households in that country fall short of a decent standard but cannot be used to compare countries.

Multidimensional poverty is poverty measured by the deprivations a person faces in several parts of life at the same time. The Multidimensional Poverty Index (MPI) is the composite indicator that measures it. It is published every year by the United Nations Development Programme and the Oxford Poverty and Human Development Initiative for more than 100 developing countries. It does not use income at all. Instead it records, from household surveys, the deprivations that each household actually experiences, using ten indicators grouped into three dimensions. Each dimension carries one third of the total weight.

A person is counted as multidimensionally poor if the indicators in which their household is deprived have weights that add up to one third or more of the total. A household deprived in both health indicators (1/6 + 1/6 = 1/3) is therefore poor on the index, while a household deprived only in electricity and assets (1/18 + 1/18 = 1/9) is not. The index takes account of two things: how many people are poor, and how many deprivations poor people face at once. It can therefore fall either because fewer people are poor or because those who are poor are deprived in fewer ways.

The two kinds of indicator answer different questions. An income line is simple, can be updated often and shows whether people have the purchasing power to meet their needs. It misses people whose income is above the line but who lack schooling, clean water or electricity because those services are not available where they live at any price. The MPI captures these overlapping deprivations and shows a government which of them to act on. Its weaknesses are that it needs detailed household surveys, exists only for the countries and years surveyed, and rests on value judgements about which indicators to include, how to weight them and where to set the one-third cut-off.

Every measure of poverty is an estimate. The main difficulties in measuring poverty are these.

• Any line is partly arbitrary. People just above a poverty line are barely better off than people just below it, yet only the second group is counted, and moving the line changes the count with no change in anyone's living standard.
• A headcount hides depth and duration. It does not show how far below the line people are, or whether they are poor for a few months or for their whole lives.
• The data are costly and incomplete. Household surveys are expensive and infrequent, and they tend to miss the people most likely to be poor: the homeless, the displaced and those in remote or insecure areas.
• Income is hard to record. Earnings in the informal economy, food grown for the household's own use and payments in kind appear in no tax or wage record and can only be estimated.
• Income is measured for the household. This assumes that it is shared equally inside the household, which can hide the poverty of women, children or older people within it.
• Comparisons need price adjustments. Converting incomes to a common currency relies on purchasing power parity estimates, which are revised, and prices also differ between cities and rural areas inside one country.
• Composite measures add difficulties of their own: the choice of indicators and weights is a matter of judgement, and the surveys used come from different years in different countries.

A household's income is the payment it receives for the factors of production it owns, plus any transfer payments, minus taxes. Each cause of inequality and poverty below is a reason why those payments differ between households, or why the gap is not closed afterwards. There are eight causes, and each should be explained as a chain that ends in incomes.

1. Inequality of opportunity. Children born into low-income households tend to have less access to good schools, health care, adequate nutrition, credit and useful contacts. They reach adulthood with fewer skills and qualifications, so they earn less, and they can offer their own children less in turn. Low income is passed from one generation to the next for reasons outside the individual's control, and ability that could have raised output goes unused.
2. Different levels of resource ownership. A household that owns only its labour receives only wages. A household that also owns land, property, shares or a business receives rent, interest, dividends and profit as well. The ownership of these assets is concentrated in a small share of households and is passed on by inheritance, so the income they earn is concentrated too.
3. Different levels of human capital. Workers with more education, training and better health are more productive, so firms' demand for their labour is higher and they are paid more. Workers with few skills compete for jobs that many people can do, and their wages are low. Because acquiring human capital costs money and time that poor households do not have, this cause is closely tied to inequality of opportunity.
4. Discrimination on grounds such as gender, race, ethnicity, religion, age or disability. It lowers the incomes of the affected group through lower pay for the same work, exclusion from better-paid occupations and from promotion, and restricted access to education, credit or the ownership of land. Because it lowers the reward the group receives for a given level of skill, it also weakens the incentive, and the means, to acquire more skills.
5. Unequal status and power. Workers with little bargaining power, such as those in insecure or informal work with no trade union and few alternative employers, have to accept low wages and poor conditions. Groups with high status and political influence can shape tax rules, regulation and public spending in their own favour. The first effect holds down incomes at the bottom and the second protects incomes at the top.
6. Government tax and benefit policies. Taxes and transfer payments change the distribution that the market produces. A system that relies heavily on indirect taxes, taxes high incomes and wealth lightly, or pays low benefits leaves most market inequality in place. A cut in benefits, or a change that makes the tax system less progressive, raises inequality of income after taxes and benefits even if market incomes are unchanged.
7. Globalisation and technological change. In higher-income countries both raise the demand for highly skilled labour and reduce the demand for low-skilled labour. Automation replaces routine tasks, and freer trade and the relocation of production expose low-skilled workers in manufacturing to competition from lower-wage economies, so their wages fall or their jobs disappear. Skilled workers, owners of capital and the most successful firms gain from larger markets. The gap between high and low pay widens. In the lower-income countries that receive the production the same process can raise wages and reduce poverty, so the effect is not the same everywhere.
8. Market-based supply-side policies. Policies designed to strengthen incentives and make labour markets more flexible include lower taxes on high incomes and on company profits, lower unemployment benefits, weaker trade unions and a lower or abolished minimum wage. Each raises the after-tax income of high earners and owners of capital, or reduces the income and protection of low earners. Greater inequality is therefore a recognised equity cost of these policies, even where they succeed in raising output.

The causes reinforce one another. A household with a low income can save little and can spend little on education and health, so its human capital and productivity stay low, and so does its income. Discrimination and a lack of power make it harder to break out of this circle, and a tax and benefit system that does little redistribution leaves it in place.

Whether inequality damages an economy or helps it is contested. Its impact can be judged on three things: economic growth, standards of living and social stability. Each has to be evaluated, which means setting out the arguments in both directions and the condition that decides between them.

Economic growth: how inequality can slow it. High inequality can reduce growth through three channels. First, households on low incomes cannot afford education and health care and cannot borrow to pay for them or to start a business. Part of the population's potential human capital and enterprise is never developed, so productivity and potential output are lower than they could be. Second, lower-income households spend a larger proportion of any extra income than higher-income households do, so the more income is concentrated at the top, the weaker consumption and aggregate demand are. Third, if inequality leads to social and political instability, firms face more uncertainty and invest less.

Economic growth: how inequality can support it. Differences in reward are an incentive to work longer, to train, to take risks and to start firms. If every outcome were equal whatever the effort, that incentive would disappear. Higher-income households also save a larger proportion of their income, and saving can finance investment in capital.

Standards of living. For a given average income, a more unequal distribution means that more people have incomes too low to meet their needs or to take part in the normal life of their society, so both absolute and relative poverty are higher. Low-income households have worse nutrition, housing, health and education and less access to merit goods, which lowers life expectancy and educational attainment for a large part of the population. Income per person then overstates the living standard of the typical household. The argument on the other side is that if inequality comes with faster growth, the real incomes of the poor may still rise, so living standards at the bottom improve even as the gap widens. That holds only where the gains from growth actually reach low-income households, through jobs, higher wages or better public services; where they do not, average income rises and the living standards of the poorest do not.

Social stability. Large and visible gaps in income and wealth reduce people's trust in institutions and in one another, especially where the gaps follow lines of race, region or class and where people see little chance of moving up. Crime, social tension, protest and political instability become more likely, and governments under pressure may change policy abruptly. Instability is a cost in itself, and it feeds back into growth, because uncertainty about security, property rights and future taxes discourages investment. The effect is not automatic. Inequality that is seen as earned, in a society with real social mobility, is widely accepted; it is inequality seen as unfair and permanent that threatens stability.

Taxation reduces inequality in two ways. It can take a larger share from households with high incomes and wealth than from households with little, which narrows the gap in what they have left after tax. It also raises the revenue that pays for transfer payments and public services, which add most to the living standards of lower-income households. How far a tax does the first of these depends on what happens to the proportion of income paid in tax as income rises.

With a progressive tax, a higher-income household pays more tax not only in money but as a proportion of its income. With a proportional tax, it pays more money but the same proportion. With a regressive tax, it may still pay more money, but it pays a smaller proportion of its income than a lower-income household does. The test is always the proportion of income taken, never the amount of tax paid.

A direct tax is a tax on income or wealth, paid to the government by the person or firm that earns the income or owns the wealth. There are three kinds, and each can be used to reduce inequality.

• Personal income tax is charged on wages, salaries and other personal income. It is usually progressive: an allowance is free of tax and successive bands of income are taxed at higher rates, so the proportion of income paid rises with income. It is the main tax used to reduce inequality of income. Its limits are that very high rates may weaken the incentive to work longer, seek promotion or declare income, and that the highest earners are the people most able to move themselves or their income to a country with lower taxes.
• Corporate income tax is charged on the profits of firms. Profits belong to shareholders, and shares are owned mainly by wealthier households, so much of the tax falls on them, and it raises revenue for redistribution. Its limits are that a high rate lowers the return on investment after tax, that multinational firms can report their profits in low-tax countries, and that part of the burden may be passed to workers as lower wages or to consumers as higher prices.
• Taxes on wealth are charged on the ownership or transfer of assets. A wealth tax may be a tax on property, on inheritance, on capital gains or on a household's net wealth each year. Wealth is more concentrated than income and is passed between generations, so these taxes act directly on inequality of wealth and of opportunity. Their limits are that assets such as private businesses and works of art are hard to value, that wealth can be moved abroad or rearranged to avoid the tax, and that a household with valuable assets but a low income may struggle to pay.

An indirect tax is a tax on spending on goods and services, such as a sales tax, a value added tax or an excise duty. The seller pays it to the government and passes at least part of it on to buyers in a higher price. Indirect taxes are usually regressive, even though every buyer pays the same rate on the item. Lower-income households spend almost all of their income, and much of it on necessities, while higher-income households save a substantial part of theirs. A tax on spending therefore takes a larger proportion of a low income than of a high one. Suppose (invented figures) a tax on spending costs a low-income household $2,000 out of an income of $20,000 and a high-income household $6,000 out of an income of $100,000. The second household pays three times as much tax, but only 6% of its income against 10% for the first: the tax is regressive.

A tax system that raises most of its revenue from indirect taxes can therefore leave incomes after tax more unequal than before. Governments can make indirect taxes less regressive by exempting necessities such as basic food, or taxing them at a lower rate, and by taxing luxuries more heavily. The cost is a more complicated system and less revenue, and part of the benefit of a lower rate on food goes to high-income households, who buy food too.

Besides taxation, there are six kinds of policy. They differ in whether they act on a cause of low income or on its result, in how quickly they work, in whom they reach and in what they cost. Each must be evaluated, so each is set out below as a mechanism followed by the conditions under which it works well or badly.

Policies to reduce inequality of opportunity and investment in human capital. The government provides or subsidises schooling, early-years programmes, training, health care, clean water and sanitation, with particular attention to low-income households. Better education and health raise the productivity of children and workers from those households, so the demand for their labour and their future earnings are higher, and low income is less likely to be passed on to the next generation. Because the quantity and quality of labour increase, potential output rises as well. This is the only policy in the list that acts on a cause of inequality, and it does not weaken the incentive to work. Its limits are of time, cost and quality. The benefits arrive only after many years, so the policy does nothing for people who are poor now. It is expensive and competes with other spending. It works only if the services are of good quality and actually reach poor households, who may be unable to give up their children's earnings or to pay for transport and materials. And higher skills raise incomes only if the economy creates jobs that use them.

Transfer payments. A transfer payment is a payment from the government to a household for which no good or service is supplied in return: a pension, a child benefit, an unemployment or disability benefit. It raises the disposable income of the household that receives it directly and at once, so it is the fastest way to reduce poverty, and it can reach people who cannot earn. Because low-income households spend most of any extra income, transfers also support consumption. Effectiveness depends first on design. A payment made to everyone in a group, such as all pensioners, reaches every poor member of the group but spends much of its budget on households that are not poor. A means-tested payment, paid only to households below a stated income, concentrates the money where it is needed, but it is costly to administer, some entitled households do not claim it, and it has to be withdrawn as earnings rise. The faster it is withdrawn, the less a recipient gains from taking a job or working more hours, so the weaker the incentive to work. Effectiveness depends also on the budget: transfers must be financed every year from tax revenue, which limits them most in lower-income countries. Finally, a transfer treats the result of low income and not its cause, unless it is made conditional on something that raises future earnings, such as children attending school.

Targeted spending on goods and services. Instead of cash, the government provides or subsidises particular goods and services for low-income households: free school meals, social housing, health clinics, food or fuel support. Living standards rise directly, and the household's own income is freed for other needs. The advantage over cash is that the support is certain to be used for the intended purpose, which matters most where the aim is a specific outcome such as child nutrition, or where the service would otherwise not be available to buy. The disadvantages are that recipients have less choice than with cash and may value the good at less than it costs to provide; that finding out who qualifies is costly and can discourage people from applying; and that a service used only by the poor may be of low quality.

Universal basic income. A universal basic income is a regular cash payment from the government to every adult (or every resident), whatever their income and whether or not they work. It gives everyone an income floor. Because there is no means test, everyone entitled receives it, it is cheap to administer, and nothing is lost when a person takes a job, so it avoids the withdrawal problem of means-tested benefits. Its main limit is cost. A payment to every adult that is large enough to live on requires very large tax increases or cuts in other spending, while a payment that is affordable is likely to be too low to end poverty on its own. Most of the money goes to households that are not poor, so each unit of spending reduces poverty by less than a targeted transfer would. Its effect on the supply of labour is uncertain: an unconditional income may lead some people to work less, although no part of it is lost by working. If it replaces benefits designed for particular needs, such as disability or housing costs, some of the poorest households could end up worse off.

Policies to reduce discrimination. Governments pass laws that require equal pay for equal work and prohibit discrimination in hiring, promotion, lending, education and the ownership of property, and they take measures to widen access to education and jobs for disadvantaged groups. These policies raise the pay and opportunities of the groups affected, increase the reward they receive for their skills, and so strengthen their incentive to acquire more. Using everyone's abilities fully also raises productivity. The cost to the government budget is low. The limits are that a law works only if it is enforced, and discrimination in hiring and promotion is hard to prove; that attitudes and social norms change slowly; and that legal rights do not reach the informal sector, where many of the poorest people work.

Minimum wages. A minimum wage is a legal lower limit on the hourly wage, which makes it a price floor in the labour market. Set above the equilibrium wage for low-paid work, it raises the earnings of the workers who keep their jobs, reduces poverty among working households and narrows the gap between low and typical pay, at no direct cost to the government budget.

Three conditions decide how far a minimum wage reduces poverty. The first is the effect on employment. In the competitive model of the diagram, firms employ fewer workers at the higher wage, and the fall in employment is larger the more wage elastic the demand for labour is (for example where workers are easily replaced by machines) and the higher the floor is set relative to typical pay. A moderate floor in a market where the demand for labour is inelastic raises incomes with little loss of jobs. The second is coverage. A minimum wage reaches only people in formal employment. It does nothing for the unemployed, the retired, people unable to work or workers in the informal sector, and many of the poorest households have no one in formal work. The third is who the low-paid are. Some minimum-wage workers are second earners or young people in households that are not poor, and for those who are in low-income households, higher earnings can be offset by the withdrawal of means-tested benefits. Firms facing higher wage costs may also raise prices.

• Equality is sameness and can be measured; equity is fairness and is a normative judgement. Reducing inequality is the usual way of pursuing equity, but the two are not the same.
• Income is a flow and wealth is a stock. Wealth is more unequally distributed than income because it accumulates, is inherited and earns income.
• The Lorenz curve plots cumulative percentage of income against cumulative percentage of population. The closer it is to the line of equality, the more equal the distribution, and the lower the Gini coefficient (0 to 1).
• Absolute poverty is measured against a fixed line; relative poverty against a share of typical income. Growth alone can reduce the first but not the second.
• Single indicators (the international poverty line, minimum income standards) use income; the Multidimensional Poverty Index combines ten indicators of health, education and living standards and counts a person as poor if deprived in one third or more of them by weight.
• There are eight causes: unequal opportunity, resource ownership, human capital, discrimination, status and power, tax and benefit policies, globalisation and technological change, and market-based supply-side policies.
• Inequality can harm growth (lost human capital, weaker consumption, instability) or support it (incentives, saving); it lowers living standards for a given average income and can threaten social stability.
• Progressive, proportional and regressive taxes are defined by what happens to the proportion of income paid as income rises. Direct taxes on income, profits and wealth can be progressive; indirect taxes are usually regressive.
• Six further policies: investment in human capital and equal opportunity, transfer payments, targeted spending, universal basic income, anti-discrimination policies and minimum wages. Each has a mechanism, a group it reaches and a limit.