How Do We Measure Poverty?
A recent Angus Reid study, reported widely in the news, showed that many Canadians are experiencing financial stress. A large proportion report needing to borrow money to buy groceries, eschewing dental care, and are experiencing hardship in ten other money-related scenarios presented to them. While sadly no one is shocked that a substantial fraction of Canadians is struggling financially, what is surprising is that the study’s estimate of that fraction is considerably higher than the official governmental numbers. The study found that about 16% of the population is “struggling”, while a further 11% is “on the edge” (i.e., at risk of struggling). This gives a total of 27% who are in some sort of financial jeopardy. Whereas, according to the official estimate, 4.8 million Canadians (about 14%) live beneath the line of indigence.
The difference between Angus Reid’s and the government’s estimates comes about because of the differences in philosophies and methodologies embraced by those doing the calculations. The Angus Reid Poverty Index is innovative and meaningful, but arises from a distinctly different set of assumptions and definitions of poverty than does the government’s official low-income cut-off line. The numbers are concerning and in need of policy attention, regardless of how they were manifested. But there is some educational utility in taking a moment to review the various philosophies underlying how poverty is conceptualized and measured.
Defining poverty and low income is important from a policy standpoint. Many governments use the designation to decide which communities receive special attention and resources. As a result, operationalizing any unspecific sense of deprivation can be a process fraught with unintentional bias.
About 21% of children in the USA live beneath the poverty line, while 10.2% of Danish children are thus beset. But do those statistics mean that Denmark is twice as wealthy as the USA? Is a poor person in Denmark as similarly deprived as a poor person in the USA? The confounding factor here is that different countries and organizations define the low-income cut-off line differently, making it difficult to compare across jurisdictions.
Some see poverty as strictly an economic state. Others see it as a condition of political vulnerability. Still others see it predominantly as a measure of social class. All of these interpretations are simultaneously correct and insufficient. The challenge is that we all have a sense in our minds of what a financially challenged existence looks like, but struggle to put that vision into a set of measurable indicators. Several competing measures have thus emerged, each telling only part of the story.
It is important to distinguish between “relative” and “absolute” poverty. A measure of relative poverty defines poverty as being below a threshold computed from within the population of interest. Someone who is “relatively impoverished” has significantly less wealth than other members of society. For example, a common practice is to define someone as poor if his or her income is less than 60% of the population’s median income. This concept is sometimes called “economic distance.” The advantage to this approach is that it ensures that poverty is understood in the context of its specific community.
The great disadvantage is that, regardless of actual levels of wealth, a relative measure necessitates that at least some members of the community are defined as impoverished. Imagine a nation of millionaires. It is a mathematical necessity that some of those millionaires will have an income less than 60% of the median. We would identify those people as impoverished… even though they are millionaires.
“Absolute” poverty, on the other hand, is a level of poverty defined in terms of the minimal requirements necessary to afford a basic standard of living. The inability to achieve that minimal standard is considered deprivation and therefore poverty. To compute absolute poverty, we define a set of necessities that we place in our grocery basket. We can therefore call this a “market basket measure” (MBM). In Canada, the official MBM includes the costs of food, clothing, footwear, transportation, shelter and other expenses for a reference family of two adults and two children, adjusted for geographical region.
According to Canada’s MBM, a Torontonian from a 4-person family earning less than $40,595 would be low-income in 2015.
On the other hand, one of the “relative poverty” thresholds commonly used in this country is the Low-Income Measure (LIM), which is the median income of the population, adjusted for household size. The Statistics Canada website offers a good explanation for how the LIM is computed. Based on 2015 numbers, a Canadian in a four-person household is considered low-income if he or she makes less than $44,266 after taxes. As can be seen, the LIM and MBM give different threshold estimates for the same person.
Statistics Canada also uses something called a Low-Income Cut-off (LICO), which is an income threshold below which a family will likely devote a larger share of its income to the necessities of food, shelter, and clothing than an average family would. For various family sizes and community sizes, data are collected about how much income families typically retain after paying taxes, and about how much of that income they spend on food, shelter, clothing, and the other necessities of life. The income value associated with 20 percentage points above the average expenditure is defined as the LICO for that particular family and community size. Yes, it can be confusing. The LICO really hasn’t been updated since 1992 except to adjust for inflation.
The U.S. Census Bureau determines poverty status by comparing pre-tax cash income against a threshold that is set at three times the cost of a minimum food diet. This is called the “Official Poverty Measure”, or OPM, and was first devised in 1963, with the cost of food is updated each year to reflect inflation. The OPM varies by family size, with the definition of family rigidly prescribed. In 2017, the OPM poverty threshold for a family of four was $30,750 (about 13% of Americans).
The Supplemental Poverty Measure (SPM) was introduced in 2010 in the USA to account for the criticisms of the OPM. Rather than focusing just on food expenditures, the SPM attempts to capture what Americans expend on all basic necessities: food, clothing, shelter, and utilities. It also accounts for taxes paid and government benefits received. According to the SPM, about 14% of Americans live beneath the poverty line. But the OPM is still the official measure.
Perhaps the most commonly applied poverty threshold, used since 1990, is the so-called global poverty line (GPL), which is assessed by the World Bank. It is based solely on the cost of living, and is assessed by computing the global prices of key commodities. Up until 2008, the GPL was conveniently defined as 1 U.S. dollar per day. Anyone making less than this amount was said to be unable to purchase the necessities of life and was thus impoverished. In 2015, due to global changes in commodity prices, that line was redefined as $1.90 per day.
From 1997 until fairly recently, the United Nations relied upon an indicator called the Human Poverty Index (HPI). It is a mathematical combination of scores that describe a population’s longevity, literacy, and standard of living. And the formula is different for poor countries versus wealthy countries. For poor countries, less longevity is expected, while the standard of living is based on very basic conditions, such as whether treated water is available and whether children tend to be underweight. For wealthy countries, the formula includes a measure of social exclusion, typically estimated from the nation’s unemployment rate.
There are ongoing efforts to produce more comprehensive and thorough poverty estimators. Among the more impactful of recent innovations is the Multidimensional Poverty Index (MPI), which was developed in 2010 by the Oxford Poverty & Human Development Initiative, and which has replaced the HPI in most serious international comparisons of poverty. The MPI is intended to measure acute poverty in terms of both the proportion of experiencing multiple deprivations and the intensity of such deprivations. It does so by measuring individuals’ experiences with health, education, and standard of living (via traditional development indicators, like the availability of clean drinking water, and whether the home has a proper floor), and summing those experiences across the measured population.
The MPI is usually only computed for low income countries. In 2016, computations were completed for 102 countries, which accounts for about 75% of the total human population. From those data, 1.6 billion people were found to be multidimensionally poor (compared with 1.3 billion who live beneath the Global Poverty Line). According to the MPI, the poorest country in the world is Niger, whose 2012 MPI estimate is 0.605. In second place is Ethiopia, whose 2011 data results in an estimate of 0.564.
The MPI method is a bit more inclusive than other methods, since it accounts for impoverishing factors beyond the cost of goods.
Clearly, there is a multitude of methods for measuring financial distress, each rife with its own strengths, biases and values. Do we only care about whether a poor person survives, or do we also care whether he thrives? Are food and shelter the only baseline requirements, or should we include access to services –health, legal, financial– or even access to information and electronic connectivity? The challenge when discussing low-income and deprivation is to select the most appropriate measure for the story being told, and to avoid comparing computations that were derived from dissimilar assumptions, values, and intents.
(Portions of the above were adapted from chapter 6 of my 2017 textbook, Determinants Of Health)