How should we measure poverty?

Most countries in the world measure their poverty using an absolute threshold, or in other words, a fixed standard of what households should be able to count on in order to meet their basic needs. A few countries, however, have chosen to measure their poverty using a relative threshold, that is, a cutoff point in relation to the overall distribution of income or consumption in a country.

The chart above shows the differences between relative and absolute poverty headcount ratios for countries that have measured both. You can select other countries from the drop down list, but for example, you can see that Romania switched from measuring poverty in absolute terms to measuring poverty in relative terms in 2006.  Absolute poverty headcount ratios steadily declined from 35.9% in 2000 to 13.8% in 2006. However, by relative measures, the national poverty headcount ratio in 2006 was 24.8%.  This does not mean that poverty bumped up in 2006. These two numbers are simply not comparable, but what exactly do they both mean?

The national poverty indicator
The commonly understood concept of poverty is the condition households or individuals experience when they are unable to access the set of goods and services needed to ensure a decent life in the society to which they belong. When measuring economic poverty, a decision has to be made in regard to a specific monetary threshold or poverty line, which is the value used to distinguish between the poor and the non-poor. To learn more about how countries choose and estimate poverty lines, click here.

If you have ever worked with national poverty data, you have probably used the national poverty headcount ratio indicator from the World Bank’s World Development Indicators (WDI) database.  The WDI compiles the poverty headcount ratio measured at the national poverty line, a threshold which is individually determined by each country’s government. The government then estimates the national poverty rate using this line (sometimes with technical assistance from the World Bank), and publishes it as its official poverty estimate. The WDI accepts these estimates, as long as they can be backed by sound estimation methodologies. Up until recently, the WDI only published absolute poverty rates. The latest update now also includes relative poverty rates for some countries.

Absolute vs relative poverty
Most countries in the world measure their poverty using an absolute threshold, or in other words, a fixed standard of what households should be able to count on in order to meet their basic needs. A few countries, however, have chosen to measure their poverty using a relative threshold, that is, a cutoff point in relation to the overall distribution of income or consumption in a country. To learn more about how countries choose and estimate poverty lines, clickhere.
Setting poverty lines in relative terms is especially prevalent in European countries. As many of the developing countries in Europe are middle-income countries, it is often difficult to define a common set of goods and services (essential for setting an absolute poverty line) that can be easily perceived by a broad range of the population in those countries. Thus, following the EUROSTAT practice and Europe 2020 Strategy, most European countries definepersons at risk of poverty as those living in a household with an equivalized disposable income below the risk-of-poverty threshold, which is set at 60% of the national median equivalized disposable income (after social transfers). Many of these relative poverty estimates are based on per adult-equivalent income, rather than per capita income (see more details here).
One challenge of the relative poverty measure is understanding how it behaves. For example, relative poverty numbers in a country may not decline continuously, and can be persistent, because the poverty line is based on a threshold of 60% of the national median equivalized household income. It is also important look at how relative poverty measures behave at the time of crisis. As the chart below suggests, relative poverty rates in Romania dropped during the “Great Recession,” as the disposable income of the population shrank. While this may seem counterintuitive, it is perfectly acceptable given the distributional nature of this indicator.

Changes in the WDI Series
As of April 2014, the WDI has switched from publishing absolute poverty rates to publishing relative poverty rates for several European countries. If you have previously downloaded this indicator for any of the countries listed in the chart below, your saved series may not be comparable with what is currently available in the WDI.

As an institution committed to poverty reduction, it is important for the World Bank Group to monitor countries’ commitment and ability to measure poverty indicators, irrespective of the definition. Given that only a small number of countries use relative poverty measures, the WDI publishes them under the same series as absolute poverty numbers for other countries. The following countries have relative poverty numbers included in these series: Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia. The footnotes in the metadata distinguish them from absolute poverty numbers, but it is important to be aware of this difference.

Furthermore, it is important to recognize that there are now two separate series available in the Poverty and Equity Database for the poverty headcount ratio at national poverty line (% of population) indicator: the SI.POV.NAHC and SI.POV.NAHC.NC.  The SI.POV.NAHC is a series reflecting the latest official numbers that are comparable within a country. While in the past this series only reported absolute values, it now includes relative values for those countries which have changed their monitoring systems. On the other hand, the SI.POV.NAHC.NC is a new series which has been created to include all historical relative and absolute values, including those not comparable overtime within a country due to methodological changes.

To illustrate the difference, let us go back to the case of Romania in the year 2006, when the government switched to officially reporting relative poverty but continued to measure absolute poverty. The SI.POV.NAHC series reports only the official relative values for 2006 and thereafter, whereas the SI.POV.NAHC.NC series reports the previously published absolute values prior to 2006 as well. The SI.POV.NAHC.NC series provides a fuller picture of the number of countries that have an official monetary poverty monitoring system. This is valuable information, as it better reflects countries’ efforts to collect of household surveys as well as publish national official poverty data.

Moreover, for all of these recently updated countries, the required microdata to replicate this indicator is also available upon accreditation and in accordance with specific terms of use (to learn more about EU-SILC survey data and request access to this data visit Eurostat SILC and the World Bank Microdata Library). This level of transparency from countries is certainly commendable and represents a good example of great data documentation, transparency and availability.

Indicator and code used in this post:

• Poverty headcount ratio at national poverty line (% of population) SI.POV.NAHC
• Poverty headcount ratio at national poverty line (% of population), including noncomparable values SI.POV.NAHC.NC

This post first appeared on The World Bank Blog

Author: Juan Feng, a Chinese national, is an Economist at the World Bank in Washington DC. Minh Cong Nguyen is an Economist in the Poverty Global practice at the World Bank. 

Image: Residents are seen in the Villa 31 slum in the heart of Buenos Aires, February 9, 2014. REUTERS/Enrique Marcarian.