“Inequality” appears to be this year’s flavour of the year in the development discourse. The UN High Level Panel report on post 2015 development goals used the word 24 times. One instance, for example was “The 1.2 billion poorest people account for only 1 per cent of world consumption while the billion richest consume 72 per cent.” The debate will be more influential if it focuses on inclusivity instead of inequality. There is no consensus on definitions of “income inequality” or of “equitable.”

Inclusivity, at least, can be defined by access measures and be represented by minimum standards and income floors.

Income inequality is a contentious topic, due in part to different frames of reference and different definitions and measures. Is the frame of reference the distribution of income across the world’s population? (Of total world income, the top 10% has 42 per cent of global income with the poorest 10 per cent having 1%).  Or is it the distribution across countries (where income inequality in the world rose in the 1980s and the mid-1990s, and decreased after 2000)? Or are we asking about the income distribution within countries (great variation around the world)? Countries with very high inequality are clustered in South America and southern Africa. Countries with low inequality are mostly in Europe. Canada and the U.S. have medium income inequality.

Do we measure world income inequality by the difference of the average income of high-income countries and the average income of low-income countries? Or do we calculate overall inequality by the Gini index? A Gini index of 0 represents exact equality (that is, everybody has the same amount of income); a Gini index of 1 represents total inequality (that is, one person has all the income and the rest of the society has none) Or do we calculate income inequality within each country?

Branco Milanovic (Worlds Apart: Measuring International and Global Inequality) characterized three different concepts of the distribution of income across the world’s population very clearly. His distinction:

·         Each country counts the same. Unweighted international inequality. A country’s population is not considered; income refers to GDP per capita—everyone in the country gets the same income. Comparisons are based on a representative individual from each country, each with the average income per capita. In this concept, you can imagine a world of 200 representatives from 200 countries, each holding a placard with the GDP per capita of his/her country. The representatives are then lined up from richest to poorest and the degree of inequality is determined.

·         Countries with more people count more. Population weighted international inequality “world inequality.”  Assume everyone in the country receives the same income; the country is still the unit of measure; but, the number of representatives from the country is based on the population. So, instead of 200 representatives holding placards, countries with higher populations would have more people holding placards than countries with smaller populations.

·         Each person counts the same. Individuals, not countries are the unit of measurement; and the income is their own income. In each country, household surveys rank individuals from richest to poorest. Each country is represented by each individual with their own income on the placard e.g. rich Chinese will be next to rich Americans. Then countries with larger populations count more (they have more individuals in the lineup).

In the unlikely event we could agree on the appropriate frame of reference, there are differences of opinion on the appropriate techniques for measurement. Results will differ if comparisons are based on purchasing power parity compared to official currency exchange rates. Conclusions will vary if the statistics are derived from household survey-based estimates of mean income compared to national estimates of GDP per capita. Yet another complication is whether we measure expenditures and consumption instead of income. Varying household size and age affects welfare comparisons, so there is the additional complication of the need to refine measures to distinguish between per capita figures and the equivalent income of adults.

There is an extensive literature on the difficulties of measuring inequality. There is no consensus that any indicator or index is appropriate to summarize the degree of inequality of a given income distribution. Indeed indices by their nature are arbitrary in the choice of weights for their components. The GINI coefficient is most popular – measuring the average difference between every possible pair of individuals in a group, divided by the average size (income) of the group. Criticism of the GINI index is that it does not emphasize the lower part of the distribution (the poor), but instead places the same weight throughout the distribution. The Multidimensional Poverty Index (MPI) measures the factors beyond income that constitute a poor person’s level of deprivation. It is an index of ten indicators (chosen based on appropriateness to the specific context) in three dimensions: education, health, and standard of living. Each dimension is weighted equally (1/3) and each indicator within each dimension is weighted equally. The Atkinson index calculates an “equity-sensitive average income” or the level of per capita income which, if enjoyed by everybody, would make total welfare exactly equal to the total welfare generated by the actual income distribution.  The Hoover (or “Robinhood”) Index measures the proportion of income needed to be redistributed to achieve equality.

A distribution statistic currently in vogue is the share of the total income of a country in a certain proportion of the population, measuring wealth concentration at the top of the socio-economic scale. For example, if 80% of the national income is concentrated in the top 5% of the population, then the remaining 20% of the income must be shared among the other 95% of the people. A corresponding measure is the proportion of the total income that is earned by the bottom X% of the population. In a similar vein, the income quintile share ratio is calculated as the ratio of total income received by the 20% of the population with the highest income (the top quintile) to that received by the 20% of the population with the lowest income (the bottom quintile).

To complicate matters, there is a controversy concerning the hypothesis that increasing income inequality is a necessary consequence of development – as countries, like the Chinese case move up the economic development ladder, they initially experience more inequality, and eventually progressively less inequality. As a society develops from an agricultural society into an industrialized one, inequality within that society grows. Average earnings of industrial workers increase relative to those of farm workers. As development continues, inequality among industrial workers also increases. As a society moves into the more advanced stage, the government begins to help redistribute the wealth—through the tax and transfer system as well as through funding universal education—which leads to a decline in inequality. This explanation has been criticized as simplistic and for ignoring other factors—such as trade openness, the willingness of government to intervene in the market, and the efficiency of the country’s financial sector—that also explain income inequality.

Some researchers hypothesize that the process of urbanization is the key explanatory factor. Increased urbanization will make the national Gini index and the Gini index between urban and rural areas rise first and then eventually decline. Others argue that the “villain” is market forces – that technical change and increased globalization, increased demand for highly skilled labour. As developed countries import more low-skilled-intensive goods and export more skills-intensive goods, jobs in low-skilled industries are lost in those developed countries. Others point to the increase in institutional forces, like declines in unionization rates, stagnating minimum wage rates, deregulation, and national policies that favour the wealthy as contributing to inequality.

If work is to be done to encourage the G20 to fulfill commitments to “shared and inclusive growth,” it will be important to develop a consensus on measurement. The notion of shared growth could be defined by targets for decreasing the proportion of people below each country’s locally determined national poverty line or “low income cut off line,” below which people are considered poor. Civil society in each G20 country could determine ambitious, but practical targets for decreasing the proportion in their country, encouraging their government to adopt it.

Dialogue will be more productive if there is a consensus on definitions of “inclusive.” Perhaps the best approach is agreement on indicators that measure access to employment and critical services. One place to start is the well-known list of indicators used for the Millennium Development Goals. The burden on the proponents of a serious discussion on “inclusivity” is to arrange for a more or less universally agreed list of indicators for access to employment, food and water, health and educational services, connectivity to the various forms of infrastructure, and minimum standards of social protection. Otherwise, people will be arguing at cross purposes – with little chance of influencing decisions. Agreement on the definition of equity and inequality will be very challenging. Rather than engage in the quixotic quest to reach consensus on the definition of inequality, efforts should be directed at inclusiveness defined by access and minimum standards.

The opinions expressed in this article/multimedia are those of the author(s) and do not necessarily reflect the views of CIGI or its Board of Directors.