Skip to main content
U-M Poverty Solutions Logo U-M Poverty Solutions Logo

Poverty Solutions at the University of Michigan

Joan and Sanford Weill hall
Suite 5100
735 South State Street
Ann Arbor, MI 48109-3091


Poverty Facts

Though poverty may seem a straightforward concept, how we define and measure it is a matter of great debate. And this debate is important, because how we measure poverty is central to how we define, understand, and improve well-being in our society.

Learn more about current poverty metrics and how we measure poverty in the U.S. below.


Lived in poverty
Children were poor
Income for 4-person family at
poverty level
Detroit poverty rate vs.
national average
Median household income
Received food assistance

SOURCES: U.S. Bureau of the Census (2019 1-year estimates) and USDA (2019)

How Do We Measure Poverty?

Though poverty may seem a straightforward concept, how we define and measure it is a matter of great debate. And this debate is important, because how we measure poverty is central to how we define, understand, and improve well-being in our society. We are constantly confronted with articles and analyses describing the growth or decline in the poverty rate, a certain program’s impact on poverty, how we can best eliminate poverty. But what does it all mean? How do we actually define what it is to be poor in the United States? And why do we do it at all?

We largely use poverty measures for two central reasons: to target resources to those in need, and to measure our progress in improving well-being in society.

The most common way we measure poverty in the United States is through the Official Poverty Measure (OPM), a measure of income poverty developed in the 1960s. The OPM sets a certain income threshold, which varies by family size and is pegged to the rate of inflation, below which a family is considered poor. In 2020, the OPM threshold for a family of four is $26,200, and it’s estimated that in 2019, 10.5 percent of Americans were living beneath that OPM threshold.[1] The OPM is the most commonly referenced poverty measure, and is used in determining the eligibility criteria for a whole range of government programs.

Read our brief on what the U.S. spends on anti-poverty programs

Though the OPM is the most ubiquitous poverty measure in the United States, scholars and policymakers have long pointed to the weaknesses of the OPM as an accurate measure of well-being. The OPM thresholds we use today were largely developed by Mollie Orshansky, an economist working in the Social Security Administration in the early 1960s, and were operationalized by the Johnson administration in its efforts to measure progress in its War on Poverty.[2] Leaning on survey data which showed that American households in the mid 1950s spent roughly a third of their budget on food, the adopted threshold was set at three times a subsistence food budget as defined by the Department of Agriculture, and adjusted for family size. Other than adjustments for inflation, this threshold is basically the same we use today.

While there may be some utility in using a consistent measure over such a long period of time, the world has also changed considerably since the mid 20th century. For example, housing eats up far more of household budgets than it did in the 1960s, and food much less, meaning the current threshold fails to properly account for an adequate basic-needs budget in today’s world.[3] In addition, the OPM is not adjusted for regional variation in cost-of- living, a significant flaw when the cost of housing varies considerably by geography. The OPM also fails to properly account for other essential household expenses, such as transportation and childcare.[4] Finally, the measure does not account for many in-kind benefits and tax credits meant to improve overall well-being in low-income households, meaning the official poverty measure is unresponsive to many policies meant to combat poverty.[5]

In 2011, the Census Bureau began publishing the Supplemental Poverty Measure (SPM) to address the shortfalls of the official measure. Rather than basing the threshold off of historical food budgets, the SPM threshold uses expenditure data on food, clothing, shelter, and utilities.[6] The SPM also accounts for regional variation in housing costs, and, perhaps most importantly, includes in-kind benefits and tax credits in overall household resources. This is important because while the OPM shows a relatively stable poverty rate over time,[7] historical analyses using the SPM show that poverty has dropped considerably since the late 1960s, largely as a result of government programs.[8]

While the SPM was developed in response to critiques of the OPM, many scholars believe that any attempt to define a poverty threshold based on a “basic necessities” basket of goods is an inherently flawed approach in a wealthy democracy.[9] Rather, these scholars believe that a relative poverty measure is more appropriate, defining the poverty threshold at 50 percent of a country’s median income, for example. A relative measure of poverty moves away from measuring the share of individuals facing absolute deprivation, and seeks to measure the share of individuals who are unable to fully participate in society. Rather than trying to adjust and readjust an absolute threshold as costs, demands, and lifestyles of a particular society change, a relative threshold constantly adjusts to these changes, measuring who is being left behind. Many scholars believe that while an absolute threshold may be beneficial in a developing economy, a relative measure is far more useful in a wealthy democracy, allowing for better policy responses and more accurate cross-national comparisons of poverty. If we place the poverty threshold at 50 percent of a country’s median, the poverty rate in the U.S. rises to roughly 18 percent, far above the rate seen in other developed nations.

Still other scholars suggest that the use of any threshold – absolute or relative – may fail to accurately assess the state of well-being in the United States. The OPM, SPM, or even a relative threshold are all indirect measures of economic well-being, estimating the amount of resources a household requires to meet their needs. An alternative way to measure well-being, however, is by simply asking households directly about their ability to meet basic needs: do they have enough to eat, can they pay their bills, are they able to see a doctor when they need to?[10] There are a number of surveys that seek to understand the state of material hardship in U.S. households – most notably the Survey of Income and Program Participation (SIPP) – and when this data is combined with the OPM and SPM, it offers a more holistic view of the state of well-being in America. For example, research finds that hardship often exists in households with incomes well above the poverty line, while some households in poverty – perhaps due to the aid of government programs – don’t demonstrate hardship.[11] Relying on measures of material hardship, Poverty Solutions postdoctoral scholar Richard Rodems found that roughly 45 percent of U.S. households move into and out of either material hardship or income poverty over a two-year period, presenting a picture of widespread economic precarity in the U.S.[12]

Read the journal article on avoiding material hardship.





[4] Rodems





[9] Brady, Rethinking the Sociological Measurement of Poverty.



[12] Rodems

What should we pay attention to in determining the state of well-being in American households?

As it turns out, we need to pay attention to a little bit of everything. All of these measures have some advantages over the others, and all of them combined give us a more accurate picture of the state of well-being in America than any one measure on its own. As noted above, direct measures of material hardship add context to official poverty measures, and can also serve as a check on alternative poverty estimates. For example, some scholars have suggested that poverty, when measured by household consumption, has fallen dramatically over the past two decades, and is far lower than estimates put out by the Census Bureau.[13] However, research from Poverty Solutions has found that while measures of material hardship tend to track closely with the OPM and SPM, they don’t track with these measures of consumption poverty, casting doubt on the reliability of these new measures in determining the well-being of U.S. households.

In addition, while the SPM surely has benefits over the OPM, forthcoming research from Poverty Solutions has found that the OPM may actually do a better job of capturing the truly disadvantaged.[14] This research found that those who were classified as poor by the OPM, but not classified as poor by SPM – because government programs moved them out of poverty as measured by SPM – were far more disadvantaged than those whose pre-tax income put them above the OPM threshold, but who were defined as poor by the SPM. In other words, while the SPM was developed to help get a more accurate picture of the truly disadvantaged in our society, and is supposed to be an improvement upon the potentially antiquated OPM measurement developed in the early 1960s, it turns out the OPM still yields valuable information about who is most in need of support.



History and progress on poverty

The earliest date for which we have poverty estimates from the Census Bureau is 1959. That year, the poverty rate was over 20 percent. Through the 1960s and the Johnson administration’s War on Poverty, the poverty rate declined significantly, falling to 11.2 percent by 1969.[15] Since then, the rate has remained relatively stable, peaking at around 15 percent during economic recessions, dropping down around 11 percent at economic peaks. In fact, the estimated rate of 10.511.1 percent in 2019 is the lowest rate on record[PC1] ,[16] though recent analyses suggest we could see a sharp increase in poverty due to the economic effects of the COVID-19 pandemic.[17]   

Read our paper comparing trends in poverty and material hardship over the past two decades

Though a historical look at OPM poverty may lead one to think that progress in reducing poverty is halting and inconsistent, due to its design the OPM doesn’t capture the impact of many policies that have been implemented over the past five decades to alleviate poverty, such as the expansion of SNAP benefits (food stamps) and the Earned Income Tax Credit (EITC).[18] The SPM accounts for in-kind benefits and tax credits in determining well-being, and therefore might better capture the impact of these government programs. Indeed, some scholars have found that, as measured by the SPM anchored in 2012 living standards, poverty has fallen by 40 percent since the late 1960s, largely as a result of government programs meant to alleviate material hardship.[19] This presents a far different picture from what we get from the OPM, which suggests we’ve made little progress in our fight against poverty.


[16] The original estimated 2019 poverty rate was 10.5 percent, but the Census Bureau adjusted the estimate upwards due to a high rate of nonresponse biasing the results.



How does the U.S. compare with other rich nations?

Compared to other wealthy nations, poverty rates in the United States are quite high. Many European countries, and the Organization for Economic Cooperation and Development (OECD), use a relative poverty measure, typically setting the poverty threshold at 50 percent of the country’s median income.[27] These measures, therefore, are more of a gauge on inequality than deprivation, setting an income threshold that allows citizens to fully participate in society, rather than live at some minimal level of subsistence.[28] 

Using this relative measure, the poverty rate in the United States is roughly 18 percent, the fourth highest among 40 countries measured by the OECD, and ahead of only Romania, Costa Rica, and South Africa.[29] 

By comparison several countries in northern, western, and eastern Europe have relative poverty rates below 10 percent.[30]  More international comparisons can be found at the LIS Cross-National Data Center, Eurostat, and the OECD.

[27] Rodems

[28] Rodems




Measuring hardship with ALICE

In recent years, much attention has been paid to the United Way’s ALICE report, which sets a minimum basic needs threshold far above the poverty threshold, based on estimated household expenses for different family sizes.[31] For example, the ALICE threshold for a family of four (two adults and two young children) is roughly $61,000, including expenses like childcare, housing, transportation, and health care. ALICE stands for Asset Limited, Income Constrained, Employed and is designed to capture those households whose incomes put them above the income poverty threshold, but who are still struggling to afford basic household necessities. The ALICE rate combines these households with those under the poverty line.


The connection between poverty and race

Since we first began recording poverty data in 1959, poverty rates have been far higher for Black and Hispanic Americans than those for whites.[20] In 1959, the poverty rate for Black Americans was a shocking 55.1 percent. In the 1960s, when poverty dropped steeply, much of this reduction was due to the fall in the Black poverty rate, though by 1969 still roughly a third of Black Americans remained poor.[21] Beginning in the early 1970s, the Census Bureau began recording poverty rates for Hispanic Americans as a separate ethnic group, enabling comparisons between the poverty rates for Black Americans, Hispanic Americans, and non-Hispanic whites. The disparities were vast. In 1976, the poverty rate was 31 percent for Black Americans, 25 percent for Hispanic Americans, and 8 percent for non-Hispanic Whites.[22]

 Since then, Black and Hispanic poverty as measured by the OPM has declined significantly, but is still far above the poverty rate for white Americans. The poverty rate for Black and Hispanic Americans took a significant dip in the booming economy of the late ‘90s, falling from roughly 30 percent for both groups in 1995, to roughly 22 percent in 2000.[23] In 2018, 21 percent of Black Americans and 18 percent of Hispanic Americans were living in poverty, while the poverty rate for non-Hispanic whites was the same as it was in 1976, 8 percent.

While the poverty rate for non-Hispanic whites has been relatively stable (it only rose above 10 percent in three years since 1976, and never rose above 11 percent), the poverty rate for Black and Hispanic Americans is highly contingent on the state of the economy, in part due to their overrepresentation in low-wage industries that suffer losses during recessions, and only see wage growth when the economy is strong.[24] As noted above, this is why the poverty rate for Black and Hispanic Americans fell substantially in the economic expansion of the late 1990s, and shot back up during, and in the years following, the Great Recession.

These disparities persist when we look at rates of material hardship. In an analysis of hardship measures from the Survey of Income and Program Participation (SIPP) from the mid 1990s to 2011, Poverty Solutions postdoctoral scholar Richard Rodems found that nearly 40 percent of Black households and 36 percent of Hispanic households experience some form of material hardship, versus just 20 percent of white households.[25] The share of households experiencing “hidden” hardship – households that are not income poor but still experience a form of material hardship – are also higher amongst Black and Hispanic households (25 percent and 24 percent respectively) than amongst white households (16 percent).[26]




[23] Author analysis of historic Census tables




Map of the U.S. showing the most disadvantage communities by color coding counties

Understanding Deep Disadvantage

This index represents a holistic look at disadvantage, using health indicators (life expectancy, low infant birth weight), poverty metrics (rates of poverty and deep poverty), and social mobility data (Opportunity Insights Mobility Metrics).

Read more