The World Bank Handbook on Poverty and Inequality (Haughton & Khandker, 2009, p. 1) suggests it is important to measure poverty in order to:
- keep poor people on the agenda;
- identify poor people and thus be able to target appropriate interventions;
- monitor and evaluate projects and policy interventions geared to poor people;
- evaluate the effectiveness of institutions whose goal is to help poor people.
The Handbook on Poverty and Inequality identifies three steps that need to be taken in measuring poverty (Haughton & Khandker, 2009, p. 10):
- ‘Defining an indicator of welfare.
- Establishing a minimum acceptable standard of that indicator to separate the poor from the non-poor (the poverty line).
- Generating a summary statistic to aggregate the information from the distribution of this welfare indicator relative to the poverty line’.
Where data is available, there is a need to adjust for: i) differences in needs between households; ii) intra-household inequalities; and iii) differences in prices across regions and at different times (Coudouel et al., 2002).
Most rich countries measure poverty using income, as it is comparatively easy to measure (much of it comes from wages and salaries), while consumption expenditure is more complex and hard to quantify (Haughton & Khandker, 2009). It also important to account for wealth from assets which can be measured using Inland Revenue taxation data (Piketty, 2014).
Most poor countries use expenditure to measure poverty. Expenditure is easier to track than income, which comes largely comes from self-employment and/or irregular and informal sources (Haughton & Khandker, 2009). Some analysts argue that expenditure is a better indicator of poverty because it shows more accurately whether a person has enough to meet current basic needs (Coudouel et al., 2002). Access and availability are needed alongside income, for the consumption of goods (Coudouel et al., 2002).
The construction of a poverty line is arguably the most difficult step in the practical measurement of poverty (Haughton & Khandker, 2009). Poverty lines tend to be defined using three methods: the cost of basic needs (estimated cost of acquiring enough food for adequate nutrition plus the cost of other essentials such as clothing and shelter), which is generally the preferred approach; food energy intake (expenditure or income per capita against food consumption, in calories per person per day, to determine the expenditure or income level at which a household acquires enough food); and subjective evaluations (asking people what minimum income level is needed just to make ends meet; Haughton & Khandker, 2009). The poverty line ‘may be thought of as the minimum expenditure required by an individual to fulfil his or her basic food and nonfood needs’ (Haughton & Khandker, 2009, p. 41).
National poverty lines can be used to measure trends in countries but cannot be used for comparisons across countries. The choice of poverty line can differ from country to country, in different parts of a country, or based on family composition (Haughton & Khandker, 2009). It often depends on the intended use (Haughton & Khandker, 2009). Absolute poverty lines can be used at the national level, as well as globally. Some countries use an absolute poverty line, adjusted for inflation, which remains fixed over time to enable comparison with past levels and judgement of the effect of antipoverty policies over time. Absolute poverty lines may change to reflect changes in consumption patterns (roughly every ten years or so) if this remains comparable to past absolute poverty lines. A national level example is the US’s absolute poverty line, which in 2012 was around USD 16 a day for an individual. Most countries revise their poverty lines from time to time to reflect the evolution of what is considered poverty in that country beyond the adjustments made to the absolute poverty line. These revised poverty lines are used to measure relative poverty in relation to others in society rather than absolute poverty (Haughton & Khandker, 2009).
Global poverty measures
Differences in the estimates of global poverty occur as a result of the following (Anand et al., 2010):
- the use of different poverty lines;
- the use of different purchasing power parity (PPP) exchange rates to convert incomes in local currencies into a common international currency, which has been particularly problematic;
- different approaches to estimating within country distributions of income; and
- different calculations of mean incomes within countries.
These differences arise out of disagreements over the best way to measure poverty (Anand et al., 2010). The cost of enough food and other essentials is different in different countries, which is reflected in different absolute poverty lines. These need to be converted in order to compare poverty internationally. A common currency is established using PPP exchange rates. One PPP dollar should buy the same basket of goods in Kenya, India or the US, for instance.
To legitimately compare poverty rates across countries a global absolute poverty line, which adjusts for purchasing power parity, is needed to reflect differences in the cost of buying goods. The global extreme poverty line was adjusted by the World Bank in 2015 from USD 1.25 (PPP) a day (used since 2008) to USD 1.90 (PPP) as a resulted of recalculations using new PPP data. See the box below for more information on the global extreme poverty line and PPP.
The USD 1.90 (PPP) international poverty line
The World Bank’s new international absolute poverty line of USD 1.90 dollars a day was estimated based on the recalculated 2011 purchasing power parity (PPP) estimates of the average of the national poverty lines of 15 very poor countries. The increase to 1.90 dollars is an adjustment for inflation so that the real standard of living it represents remains the same. PPP is used to convert to a common currency the amount of money needed in each country to buy the same amount of goods and services in the domestic market. See a World Bank chart showing differences in regional poverty headcounts as a result of the recalculated poverty line.
Some concerns over the use of PPP relate to the consistency of PPP rates over time as a result of methodological concerns over price surveys; the differences in actual consumption patterns of the poor in different countries; and the contrasting results produced by the use different data sources for PPP adjusted GDP data, which include Maddison, Penn World Table and the World Bank (Ortiz and Cummins, 2011).
For example, the 2014 Asian Development Bank report on key indicators for Asia and the Pacific highlighted problems with using the conventional (at the time) USD 1.25 a day poverty line for all regions in the world (ADB, 2014). Here this poverty line was inadequate as it underestimated the costs required to maintain a minimum living standard in Asia (ADB, 2014).
The World Bank’s PovcalNet is the source of, and allows users to replicate, the Bank’s official global, regional and internationally comparable country level poverty estimates. The World Bank also provides poverty data through its:
Table 1 below shows the percentage of people living in extreme poverty (at USD 1.25) in different regions. There are some proposals for the use a much higher global poverty line of USD 10.00, USD 12.50 or even USD 15.00 a day in some cases, as it is claimed that this will better recognise poverty across the world. It is argued that standard absolute poverty measures probably underestimate poverty rates in rich countries, where people face higher welfare costs of social inclusion and relative deprivation (Ravallion & Chen, 2013).
Relative poverty lines are set as percentage of a country’s median income. In many European countries, including the UK, the poverty line is set at 60 per cent of the median income. Luxembourg has one of the highest poverty lines at USD 43.00 a day – see figure 2 below (Ravallion & Chen, 2013, p. 259). As the median income in countries increases, the poverty line tends to increase too as ‘the minimum resources needed to participate fully in society probably rise over time’ (Haughton & Khandker, 2009, p. 43). Relative poverty lines are thus also measures of inequality. A USD 1.90 a day poverty line has little relevance in rich countries where few people would fall below this line. Those considered poor in a rich country would often be considered well-off in poor countries (Haughton & Khandker, 2009). The use of relative poverty lines make it hard to compare poverty across countries or over time as they do not represent the same welfare levels. They are also are limited as targets. Some experts argue that standard measures of relative poverty probably underestimate the extent of poverty in poor countries, given that ‘these measures attach little value to social inclusion needs at low mean income’ (Ravallion & Chen, 2013, p. 258).
Poverty lines across the world
Source: Ravallion & Chen, 2013, p. 259
The proposed truly global poverty measure was designed by Ravallion and Chen (2013) as a response to the neglect of relative poverty in existing international measures. It proposes combining absolute and relative poverty measures in an attempt to capture the costs of social inclusion in a consistent way across countries. A person is counted as poor if they are either absolutely poor or relatively poor against a national relative poverty line (the cost of social inclusion on top of absolute needs being met). The measure is ‘weakly relative’ as it is not set as a constant proportion to the mean (Ravallion & Chen, 2013).
The Foster-Greer-Thorbecke (FGT) class of poverty measures is one of the best known. It includes the headcount index (P0) which measures the proportion of the population that is poor. This is the most popular because it is easy to understand and measure. But it does not indicate how poor the poor are (Haughton & Khandker, 2009).
The poverty gap index (P1) measures the extent to which individuals fall below the poverty line (the poverty gaps) as a proportion of the poverty line. The sum of these poverty gaps gives the minimum cost of eliminating poverty. The measure does not give any special weight to the poorest among the poor (Haughton & Khandker, 2009).
The squared poverty gap index (or the poverty severity index, P2) averages the squares of the poverty gaps relative to the poverty line. Extreme poverty is given greater weight than less poverty. It allows for variation in the weight placed on the income (or expenditure) level of the poorest members in society (Haughton & Khandker, 2009). It is possible to disaggregate the FGT poverty measures for population subgroups (Haughton & Khandker, 2009).
The Sen-Shorrocks-Thon index combines measures of the proportion of poor people, the depth of their poverty, and the distribution of welfare among the poor. These measures allow a breakdown of poverty into three components to see if: there more poor; the poor are poorer; there is higher inequality among the poor (Haughton & Khandker, 2009). They are a different way of measuring the same thing.
Wellbeing, poverty and inequality can also be defined in terms of assets. Three suggestions for asset-based measures of poverty look at the aspect of poverty related to a household’s wealth, their real and financial asset holdings, as well as their access to the credit market. They include: income-net worth measures, asset-poverty, and financial vulnerability (Brandolini et al., 2009). Those who are asset poor are not always the same as those who are income poor. The importance of income and wealth differs depending on the situation in country. For example, wealth (indicated by assets) is less important in countries with secure employment and benefits, where regular income flows ensure living standards are maintained. Personal wealth can provide a cushion against shocks and uncertainties (Brandolini et al., 2009). Sources of data include the Luxembourg Wealth Study (LWS) database; the Eurosystem Household Finance and Consumption Survey and the Demographic and Health Survey (DHS)’s wealth index, and household surveys, although they are not used routinely.
High-frequency poverty measures allowing more real-time information on poor people’s well-being are not well developed, as evidenced by the difficulties in tracking poverty as the global economic crisis unfolded (Poverty Analysis Discussion Group, 2012). Some World Bank pilot projects using mobile phones have occurred in Peru, South Sudan and Tanzania. The South Sudan and Tanzania projects are to be rolled out across Africa as part of a programme called ‘Listening to Africa’ (Croke et al., 2012).
Health, nutrition and education poverty can be measured by looking at the nutritional status of children, incidences of specific diseases, life expectancy, and level of literacy (Coudouel et al., 2002).
Various in-country data on poverty can be gathered from several agencies, such as statistical offices or government poverty analysis units (Coudouel et al., 2002, p. 62). However, in some context responsible agencies may not exist, or there may be challenges with capacity and quality of data collection and analysis.
- Anand, S., Segal, P., & Stiglitz, J. E. (Eds.). (2010). Debates on the measurement of global poverty (Initiative for Policy Dialogue Series). Oxford University Press.
- Brandolini, A., Magri, S., & Smeeding, T. M. (2009). Asset-related measures of poverty and economic stress.
- Coudouel, A., Hentschel, J. S., & Wodon, Q. T. (2002). Poverty measurement and analysis. In The PRSP Sourcebook. Washington, DC: World Bank.
- Croke, K., Dabalen, A., Demombybes, G., Giugale, M., & Hoogeveen, J. (2012). Collecting high frequency panel data in Africa using mobile phone interviews (Policy Research Working Paper 6097). Washington, DC: World Bank.
- Haughton, J., & Khandker, S. R. (2009). Handbook on poverty and inequality. Washington, DC: World Bank.
- Ortiz, I., & Cummins, M. (2011). Global inequality: Beyond the bottom billion – A rapid review of income distribution in 141 countries. UNICEF.
- Ravallion, M., & Chen, S. (2013). A proposal for truly global poverty measures. Global Policy, 4(3), 258-265.