The Trajectory of American Poverty

Date: 2026-06-20

Author: Wealth & Means Staff

Source: https://wealthandmeans.com/essay/the-trajectory-of-american-poverty

A research report tracing how the United States has defined, measured, and funded poverty since 1959 — from the Orshansky food-basket formula to the Supplemental Poverty Measure, and from $88 billion in 1965 to over $1 trillion in 2024.

TL;DR

The U.S. has spent over $1 trillion per year on means-tested programs — yet the official poverty rate sits within the same band it occupied in 1973. That paradox is real, but its interpretation depends entirely on which measure you use. The original poverty formula was built on a 1955 food survey and never updated for the modern transfer state. When you count what government actually spends, spending per poor person rose from $2,670 in 1965 to $29,214 in 2024. Whether that represents progress or dependency is the most contested question in American social policy.

Key Takeaways

• When the U.S. Census Bureau first measured the national poverty rate in 1959, it stood at 22.4%. By 1973, it had fallen to a historic low of 11.1%. Since then — across recessions, booms, welfare reform, and a pandemic — it has never permanently left the 10–15% band. The measure has been flat for over fifty years despite trillions in cumulative spending.
• The original poverty threshold, developed by Mollie Orshansky at the Social Security Administration in 1963–64, was derived from one equation: three times the annual cost of the USDA's cheapest emergency food plan. That multiplier came from a 1955 household survey. It has never been structurally revised — only price-adjusted via CPI — and remains the statutory basis for most federal program eligibility today.
• The Official Poverty Measure (OPM) counts only pre-tax cash income. It is completely blind to SNAP, housing vouchers, the Earned Income Tax Credit, Medicaid, the Child Tax Credit, and every other non-cash transfer. Because the modern safety net is built almost entirely from non-cash benefits, any poverty reduction those benefits achieve is statistically invisible in the official headline rate.
• In constant 2024 dollars, total means-tested spending rose from $88.63 billion in 1965 to $1.048 trillion in 2024 — an eleven-fold real increase. Per person in poverty, spending rose from $2,670 to $29,214. For a family of four in poverty, combined government investment now averages $116,856 per year — while the official poverty threshold for that family is $32,130.
• The Supplemental Poverty Measure (SPM), introduced in 2011, incorporates non-cash benefits, subtracts taxes and out-of-pocket expenses, adjusts for regional housing costs, and uses actual contemporary spending data rather than a 1955 food survey. Under anchored SPM estimates, poverty fell from roughly 26% in 1967 to around 16% by 2012. Consumption-based analyses estimate absolute poverty has declined by over 90% since the early 1960s.
• The 1996 welfare reform (TANF) reduced cash dependency rolls by nearly 80% and drove single-parent poverty from 33% to 11% by 2020 — but left the other 90 means-tested programs largely intact. Critics argue the structural incentives for long-term dependency remain. Supporters note that the group hardest to reach — childless, non-disabled adults — receives virtually no federal support and comprises the majority of the unsheltered homeless population.
• Means-tested spending now represents 4.1% of GDP and roughly 16% of total federal expenditures. The EITC alone lifts approximately four million people — including 1.9 million children — out of poverty each year, while simultaneously encouraging work. The safety net lifted 69% of otherwise-poor elderly individuals and 44% of otherwise-poor children out of poverty, but only 8% of poor childless adults — a rate essentially unchanged since 1970.

The systematic measurement of poverty in the United States began in the late 1950s. When the Census Bureau first calculated the official national poverty rate in 1959, it stood at 22.4%, representing roughly 39 million individuals.

Over the course of the 1960s, this rate underwent a dramatic contraction, falling to 19.0% by 1964 and dropping further to 17.3% in 1965, the year the Economic Opportunity Act was implemented. This decline culminated in a historical low of 11.1% in 1973. Since then, the Official Poverty Measure has remained remarkably flat and cyclical — fluctuating within a narrow band between 11.0% and 15.2% for over half a century, tracking macroeconomic cycles but failing to show permanent downward progress.

Historical Trends

An analysis of the decades immediately preceding the federal declaration of the War on Poverty reveals that absolute poverty was already experiencing a steep decline due to robust market-based economic growth and rising real wages. Between 1939 and 1963, absolute poverty fell by approximately 29 percentage points. The poverty rate among Black families plummeted from 87% in 1940 to 47% in 1960 — a 40-point reduction achieved in an era characterized by a minimal federal safety net and the absence of major civil rights legislation. During this pre-1964 period, only 2% to 3% of working-age adults relied on government transfers for more than half of their income, contrasting sharply with post-reform dependency rates of 7% to 15%.

In the twenty-first century, poverty trends have continued to follow the business cycle. The official rate rose to 15.1% in 2010 in the wake of the Great Recession, before declining to a historic low of 10.5% in 2019. The onset of COVID-19 disrupted this trajectory, but unprecedented federal intervention temporarily altered the picture. While the official poverty rate hovered around 11.6% in 2021, the Supplemental Poverty Measure — which incorporates non-cash transfer payments — dropped to a historic low of 7.8% due to expanded pandemic-era aid. As these temporary benefits expired, the SPM rate rose sharply to 12.9% by 2023 and remained flat in 2024, highlighting the persistent dependence of low-income households on state-directed transfers.

How Poverty Was Defined at the Start

When the federal government initiated its modern anti-poverty policies under President Lyndon B. Johnson, poverty was defined through an absolute, material prism of physical survival. The Council of Economic Advisers in its 1964 annual report utilized a simple flat income threshold of $3,000 in 1962 dollars for families of all sizes.

This flat measure was criticized by Mollie Orshansky of the Social Security Administration, who pointed out that a uniform threshold failed to adjust for family size and structure. Orshansky developed a refined series of thresholds between 1963 and 1964, which the Office of Economic Opportunity adopted in May 1965 as its working definition.

Orshansky's methodology defined poverty as a state of disposable cash-income inadequacy. Her approach relied on a formula that linked nutrition to total household spending:

T = 3 × C(food)

In this equation, T represents the absolute cash-income threshold for a given family size, and C(food) represents the annualized cost of the USDA's "economy food plan" — the cheapest of four plans developed by federal dietitians, explicitly described as "designed for temporary or emergency use when funds are low."

The multiplier of three was derived from the USDA's 1955 Household Food Consumption Survey, which indicated that average families of three or more persons spent approximately one-third of their post-tax income on food. For two-person households, Orshansky used a multiplier of 3.7 based on consumption patterns.

The original thresholds differentiated families across 124 distinct categories based on family size, farm or nonfarm residence, the sex of the household head, the number of children, and elderly status. The thresholds focused solely on pre-tax cash income and were entirely blind to non-cash resources, wealth, assets, or geographic variation in cost of living.

The Contemporary Reconceptualization

In the decades following the 1960s, shifts in consumer spending, housing costs, and the development of the social safety net rendered the original food-to-income formulation highly anachronistic. While the Official Poverty Measure remains the statutory standard for determining eligibility for most federal assistance programs, the federal government introduced the Supplemental Poverty Measure in 2011 to provide a more accurate picture of economic deprivation in the twenty-first century.

The contemporary definition of poverty has shifted from a narrow focus on physical survival to a broader assessment of disposable resources relative to a modern baseline of basic needs.

The SPM threshold is based on actual, contemporary consumer spending data from the Consumer Expenditure Survey, targeting the 30th to 36th percentile of expenditures on food, clothing, shelter, utilities, and basic telephone and internet services, averaged over a rolling five-year period — rather than a food budget from 1955.

This conceptual shift has altered the demographic profile of poverty. Under the official measure, child poverty appears high and elderly poverty appears low — elderly individuals receive cash through Social Security while children often receive non-cash benefits through SNAP and Medicaid. Under the SPM, which counts these non-cash transfers but subtracts medical out-of-pocket expenses, child poverty rates are lower, while poverty rates among the elderly and working-age adults are higher.

The Fiscal Record

Evaluating the financial commitment of the federal government requires analyzing expenditures in constant dollars, adjusted for both inflation and population growth. When the War on Poverty was launched in fiscal year 1965, total means-tested welfare spending across federal, state, and local governments was approximately $8.9 billion in current dollars, absorbing 1.2% of GDP.

To compare historical figures with modern expenditures, they must be adjusted to constant 2024 dollars. The CPI-U averaged 31.5 in 1965 and rose to 313.7 in 2024. Adjusted for inflation, the 1965 total anti-poverty investment was equivalent to $88.63 billion. Given a poverty population of approximately 33.2 million individuals, the constant-dollar investment per person in poverty was $2,669.66 per year.

Over the subsequent fifty-five years, the anti-poverty framework expanded through the addition of refundable tax credits, nutrition programs, housing vouchers, and subsidized healthcare. In fiscal year 2024, total federal spending on means-tested programs reached approximately $1.049 trillion.

With 35.9 million individuals living below the official poverty line in 2024, total federal spending amounted to $29,214 per person in poverty. For a family of four in poverty, this combined governmental investment averages $116,856 — even though the official poverty threshold for such a family is only $32,130.

Even when Medicaid is entirely excluded, the constant-dollar investment stands at $13,406 per person in poverty, which translates to $53,623 for a family of four. Distributed across the total U.S. population of 336.8 million, the modern per capita anti-poverty investment represents $3,113.96 per American citizen. This means-tested spending represents approximately 4.1% of GDP and accounts for roughly 16% of total federal expenditures.

The Efficacy Debate

The dramatic increase in federal resources alongside a relatively flat official poverty rate has generated intensive debate. The core of this debate centers on a statistical paradox: despite trillions of dollars in cumulative spending since 1965, the Official Poverty Measure suggests a stubborn 10% to 15% of the population has remained in poverty. This has led to two distinct, competing interpretations.

The transfer-skeptic perspective argues that the modern welfare state has failed to achieve its original objective of helping the poor secure economic independence. When President Johnson launched the War on Poverty, he explicitly stated that the goal was to strike at the causes of poverty — to prevent it, rather than merely subsidizing dependency. From this viewpoint, the steady expansion of the safety net has created perverse behavioral incentives that discourage work and undermine family stability.

This school of thought emphasizes that the rapid expansion of means-tested benefits in the late 1960s and 1970s coincided with a significant decline in married, two-parent families and a sharp rise in non-marital birth rates among low-income populations. Because means-tested benefits are phased out as household income rises, they create marriage penalties and high effective marginal tax rates, which can discourage low-income couples from marrying or increasing their work hours.

Critics highlight the success of the 1996 Personal Responsibility and Work Opportunity Reconciliation Act, which replaced the open-ended Aid to Families with Dependent Children (AFDC) with Temporary Assistance for Needy Families (TANF). By establishing work requirements and time-limited benefits, TANF cut dependency rates by nearly 80%, prompting a substantial increase in labor force participation among single mothers and driving single-parent poverty rates down from 33% in 1996 to 11% by 2020. Because the 1996 reforms left the other 90 means-tested programs largely intact, critics argue that the structural incentives for long-term dependency remain widespread.

The transfer-advocate perspective argues that the apparent lack of progress in official statistics is a misleading artifact of how poverty is measured. Proponents emphasize that the Official Poverty Measure is fundamentally incapable of reflecting the success of the modern safety net because it excludes all non-cash benefits and refundable tax credits. Because programs like SNAP, housing vouchers, the EITC, and the Child Tax Credit are not counted as income under the official measure, any poverty reduction they achieve is invisible in official cash-only statistics.

When evaluated using the Supplemental Poverty Measure or consumption-based metrics that account for the full post-tax, post-transfer safety net, the decline in poverty is substantial. Researchers utilizing anchored historical supplemental measures find that poverty fell from approximately 26% in 1967 to 16% by 2012, and dropped below 10% when temporary pandemic transfers were fully active. Consumption-based analyses estimate that absolute poverty has declined by over 90% since the early 1960s, falling to roughly 6%.

From this perspective, the modern safety net is highly effective at its primary task: alleviating material hardship and insulating vulnerable populations from macroeconomic shocks. The EITC alone lifts approximately four million people, including 1.9 million children, out of poverty in a typical year, while simultaneously encouraging work by subsidizing low wages.

Advocates point out that the primary deficit in the U.S. safety net is not that it creates dependency, but that it remains deeply fragmented and austere for certain groups. In 2017, the safety net lifted 69% of otherwise-poor elderly individuals and 44% of otherwise-poor children out of poverty, but it lifted only 8% of poor childless adults — a rate virtually unchanged since 1970. This group now comprises half of all Americans living in deep poverty and constitutes the vast majority of the unsheltered homeless population.

Synthesis

The evolution of U.S. poverty policy since 1964 reveals a stark contrast between conceptual measurement standards and the scale of financial resource allocation. Under the original War on Poverty framework, the federal approach was defined by absolute physical survival, centered entirely on the cash required to purchase an emergency food basket. Over sixty years, the conceptualization has shifted toward a broader, relative standard of net disposable resources, incorporating non-cash benefits, tax liabilities, and geographic differences in cost of living.

This conceptual expansion has been matched by an unprecedented increase in fiscal resources. In constant 2024 dollars, total targeted federal means-tested spending has risen from $88.63 billion in 1965 to over $1.048 trillion in 2024 — expanding the safety net from a minor cash-assistance program into a comprehensive system of in-kind benefits, direct health services, and tax credits. On a per capita basis, this represents a shift from $456.16 per citizen in 1965 to $3,113.96 in 2024, or from $2,669.66 per poor person to $29,214.00 per poor person.

The long-term persistence of a flat official poverty rate despite this massive increase in spending remains one of the most contentious issues in contemporary social policy. While critics point to this stagnation as evidence of a system that subsidizes government dependence and discourages work, supporters argue that the official rate is a flawed metric that fails to capture a substantial, transfer-driven decline in material deprivation.

Managing these competing outcomes — providing immediate material relief while designing effective pathways to long-term economic independence — remains the central challenge for the future of American anti-poverty policy. Both sides of the debate agree on one point: the official poverty rate, derived from a 1955 food survey and blind to a trillion dollars of annual government transfers, cannot settle the argument. It was never designed to.