The 1963 Grocery Budget That Still Decides Who Counts As Poor

The American poverty line is a grocery bill multiplied by three. In 1963 a Social Security Administration statistician priced the government's cheapest emergency food plan, assumed food ate a third of a family's budget, and tripled the number. That arithmetic, indexed for inflation ever since, set the 2024 threshold for a family of four at $31,812.

Where Did The Poverty Line Actually Come From?

It came from one researcher working largely on her own initiative. Mollie Orshansky (1915-2006), a statistician at the Social Security Administration, built the thresholds in 1963 and 1964, publishing her first version in the July 1963 article "Children of the Poor" and a fuller version in January 1965 in "Counting the Poor: Another Look at the Poverty Profile."

Orshansky was not designing a national standard. She was trying to describe how many American children were growing up without enough, and she needed a defensible number to describe them with. The Social Security Administration's own account of her career notes that she was born in New York to Ukrainian immigrant parents and grew up poor herself — a biographical fact her colleagues later read into the seriousness with which she treated the question of what "enough" meant.

The government caught up with her fast. In May 1965 the Office of Economic Opportunity, the agency running Lyndon Johnson's War on Poverty, adopted her thresholds as a working definition of poverty for statistical, planning, and budget purposes. In August 1969 they became the federal government's official statistical definition, indexed annually to the Consumer Price Index rather than recalculated from scratch. Sixty years later, the recalculation still has not happened.

Why Multiply The Food Budget By Three?

Because in 1955, food was roughly a third of what a household spent. Orshansky took the U.S. Department of Agriculture's economy food plan — the cheapest of four plans, designed for temporary use when money ran short — and multiplied its cost by three, using the ratio observed in the USDA's 1955 Household Food Consumption Survey.

The logic is elegant and entirely dependent on its moment. If you know what the minimum food budget costs, and you know food is one-third of spending, you can infer the whole budget without pricing housing, transportation, medicine, or child care. Orshansky did not have to decide what a family deserved for rent. The multiplier decided for her.

That is also the flaw everyone has argued about since. Food is no longer a third of the household budget for most Americans; housing is a far larger share, and medical care and child care have grown into line items Orshansky's method never priced. The multiplier of three was a measurement of 1955 consumption habits frozen into a definition and carried forward on inflation adjustments alone.

What Is The Difference Between A Poverty Threshold And A Poverty Guideline?

They are two different numbers issued by two different agencies for two different purposes. The Census Bureau issues thresholds, which are statistical — they exist to count how many people are poor. The Department of Health and Human Services issues guidelines, a simplified version used administratively, to determine financial eligibility for certain programs.

The distinction matters more than the vocabulary suggests. The Department of Health and Human Services explains in its own FAQ that thresholds carry no geographic variation — the same figures apply to all fifty states and the District of Columbia — while the guidelines come in three sets: one for the forty-eight contiguous states and D.C., and separate, higher figures for Alaska and Hawaii.

So the number that decides whether a household appears in the national poverty count and the number that decides whether it qualifies for assistance are relatives, not twins. When a program advertises eligibility at "138 percent of the federal poverty level," it is the guidelines doing that work, downstream of Orshansky's arithmetic but not identical to it.

What Does The Official Measure Refuse To Count?

Almost everything that has changed since 1963. The official measure compares a family's gross before-tax cash income against the threshold. Earnings, Social Security, unemployment compensation, pensions, dividends, and child support received all count. Capital gains and losses do not. Neither do noncash benefits — nutrition assistance and housing subsidies among them.

Read that list twice and the strangeness surfaces. The largest antipoverty programs the United States actually runs deliver help that the official poverty measure is structurally incapable of seeing. Food assistance does not raise a family above the line, because the line does not look at food assistance. Neither do refundable tax credits, which arrive as tax policy rather than as income. A policy can reduce material hardship substantially and leave the headline poverty rate untouched.

The Institute for Research on Poverty at the University of Wisconsin-Madison, in its primer on the subject, lists the standard complaints: the measure counts how many people fall below the line but not how far below, ignores taxes and work and medical expenses, makes no adjustment for geographic cost of living, does not track changing living standards, and applies a definition of "family" that fits fewer households than it once did. The United States, unlike several peer countries, uses an absolute measure rather than a relative one — the line does not rise as the country gets richer.

Does The Supplemental Poverty Measure Fix It?

It fixes the accounting, not the politics. The Supplemental Poverty Measure, introduced in 2009 after a 1995 National Academy of Sciences recommendation, sets its thresholds from a five-year moving average of what households actually spend on food, clothing, shelter, utilities, telephone, and internet, drawn from the Consumer Expenditure Survey.

It then does the arithmetic the official measure skips. It adds noncash benefits and tax credits to a family's resources, subtracts income and payroll taxes, child care and other work-related expenses, child support paid to another household, and out-of-pocket medical costs, and adjusts thresholds for regional differences in housing costs. It is, in other words, a measure that can detect whether antipoverty policy worked.

The two measures disagree, and the disagreement is instructive. In the Census Bureau's report Poverty in the United States: 2024, released in September 2025, the official poverty rate was 10.6 percent — 35.9 million people, down 0.4 percentage points from 2023 — while the Supplemental Poverty Measure came in higher, at 12.9 percent, statistically unchanged from the year before. Same country, same year, two and a third percentage points apart.

Why Hasn't Anyone Replaced The Official Line?

Because changing the definition changes the count, and the count is political. The Supplemental Poverty Measure has existed for over fifteen years and remains supplemental. The official measure, governed by federal statistical policy directive, is the one that anchors the historical series stretching back to 1959 — and that continuity is the argument for keeping it.

The case for the old number is stronger than its critics usually allow. A poverty series is useful partly because it is long: a measure redefined every decade cannot tell you whether the 1960s worked. Orshansky's threshold, whatever its origins, has been applied consistently for six decades, which makes it a reliable instrument for detecting change even if it is an unreliable instrument for detecting hardship. Statisticians distinguish between those two jobs; headlines do not.

But consistency has a cost, and the cost is that the official line now measures something no one would design on purpose. It answers the question "how many Americans have less cash income than a tripled 1963 emergency grocery budget, adjusted for inflation?" That is a real question with a real answer. It is simply not the question most people think they are asking when they ask how many Americans are poor.

The honest reading is that the United States has two poverty lines and uses the older one for the headline. Orshansky herself reportedly understood the limits of what she had built — she was describing children, not legislating a standard. The country took a working estimate and made it permanent, which is what countries tend to do with numbers that arrive at a convenient moment. The Social Security Administration's remembrance of her is worth reading as cultural history: a single researcher's back-of-the-envelope method became, and stayed, the way a nation defines its own poor.

For a related lifestyle perspective, read The Tiny Federal Slice That Nearly Broke America's Libraries.