Before these figures can be meaningfully compared, they need to be converted into common units. International dollars (int.-$) are a hypothetical currency that is used for this.

The idea is simple: one international dollar should buy the same quantity and quality of goods and services, no matter where or when it is spent. To achieve this, international dollars adjust for two things. First, they account for inflation within each country, so that values from different years can be compared (showing “constant” prices). Second, they account for differences in living costs across countries. This second adjustment uses purchasing power parity (PPP) rates, which reflect how much local currency is needed to buy what one US dollar would buy in the United States.

The United States is the benchmark, so that one 2021 int.-$ is defined as the value of goods and services that one US dollar would buy in the US in 2021. One 2011 int.-$ is defined in the same way, but for prices in 2011.

You can read more in our article, What are international dollars?

One key issue is that high-income countries typically measure people's incomes, while lower-income countries more often measure consumption expenditure — what households spend on goods and services. Pooling both types of survey is unavoidable if we want a global picture of inequality, but it means that somewhat different things are being measured depending on the country or year.

The two concepts are closely related: the income of a household equals its consumption plus savings.

At the bottom end of the income distribution, people’s consumption may be somewhat higher than their income. While zero consumption is not a feasible value — people must consume something to survive — a zero income is a feasible value. A common example is retired people drawing down their savings: they may have a very low, or even zero, income, but still have a high level of consumption.

At the top end of the distribution, consumption is typically lower than income. The gap rises with income, with households generally saving a higher share of their income the richer they are.

For both reasons, the distribution of consumption is generally more equal than the distribution of income. This means that inequality estimates tend to be somewhat lower when based on consumption surveys.

There are other comparability issues too — differences in survey design, coverage, and methodology. The PIP Methodology Handbook provides a good summary of the comparability and data quality issues affecting this data and how it tries to address them.

To help readers see where comparisons may be less reliable, the World Bank groups data points within each country into "spells" — periods where the underlying surveys are considered more comparable. Where available, you can reveal these breaks in our charts using the "breaks in data" option.

For its poverty and inequality data, the World Bank relies on household surveys that are conducted nationally. In order to produce global or regional estimates, the survey data from different countries is “lined up” and aggregated. For each year, the World Bank finds the most recent survey for each country and projects the data forward (or backward) to the year being estimated. This is necessary, particularly since surveys are less frequently available in poorer countries and for earlier decades.

These projections are generally based on the assumption that incomes or expenditure grow in line with the growth rates observed in national accounts data. You can read more about the interpolation methods used by the World Bank in Chapter 5 of the Poverty and Inequality Platform Methodology Handbook.