Chapter 28 — Social Policy: Healthcare, Welfare, and the Safety Net

"We had a choice as a country. We could keep the system we had — half-baked, expensive, leaving 50 million people uninsured and the rest one job loss away from disaster. Or we could try something new. We tried something new. It is not what either side wanted. It is not what either side hates. It is what American politics produces: a compromise that nobody salutes, that has saved a lot of lives anyway."

— Theda Skocpol, paraphrasing the political bargain that produced the Affordable Care Act

Why this chapter matters

Social policy is where Americans most directly meet their government. Most Americans will never appear before a federal judge, sign a treaty, or testify before Congress. But almost every American will at some point receive a Social Security check, fill a prescription paid by Medicare, swipe a SNAP card at a grocery store, claim the Earned Income Tax Credit, send a child to a Pell-Grant-supported college, walk into an emergency room covered by Medicaid, or rely on unemployment insurance after a layoff. The American social safety net touches more lives than any other branch of federal activity, including the military.

It is also the most contested terrain in American politics. The structure of the safety net — who is covered, how generously, with what work conditions, financed how — encodes Americans' deepest disagreements about the role of government, the meaning of personal responsibility, the boundaries of solidarity, and whether and how the state should redistribute. Republicans and Democrats agree on far more than the cable-news framing suggests (both parties defend Social Security and Medicare; both parties accept the EITC; both parties have come to accept the Affordable Care Act's coverage gains). They disagree, sometimes sharply, on direction: whether to expand or constrain, to standardize federally or to let states experiment, to fund through general taxation or through earmarked payroll taxes, to deliver through public agencies or through regulated private markets.

This chapter walks the safety net's structure, the major program-by-program battles, and the genuine values disagreements that drive ongoing reform. Education policy, which was once treated as a piece of the broader social-policy frame, gets its own chapter (Ch. 29) because of its size and distinctness. Healthcare gets the most space here, because healthcare is by far the largest single category of social spending, the most politically contested, and the area where comparative evidence is most complicated.

By the end, you should be able to:

1. Map the architecture of the American safety net — universal-by-age, means-tested, work-based, categorical — and explain how it differs from the more uniform welfare states of comparable democracies.
2. Explain the financing and politics of Social Security and Medicare, including the trust-fund accounting and the projected solvency dates.
3. Describe the Affordable Care Act, including its three-legged stool, the NFIB v. Sebelius (2012) outcome, the post-2017 mandate change, and the empirical record on coverage and outcomes.
4. Steel-man the cases for and against single-payer "Medicare for All," and explain why the comparative-democracies evidence is more complicated than partisans on either side allow.
5. Trace the post-1965 trajectory of the safety net — Great Society, the 1980s retrenchment, the 1996 welfare reform, the 2010 ACA, the 2021 ARPA expansion — and explain what each shift represented.
6. Distinguish empirical claims about program effects from normative disagreements about goals.
7. Apply the framework to a current policy question in your district.

28.1 The architecture of the American safety net

The American safety net is not one program. It is more than eighty federal programs, plus thousands of state and local programs that braid together in ways that make even specialists dizzy. The Congressional Research Service (CRS) routinely publishes program inventories that exceed five hundred pages. The Government Accountability Office (GAO) periodically reports on program duplication and gaps. A typical low-income working family with children may simultaneously receive: an Earned Income Tax Credit refund (federal tax code), a Child Tax Credit (federal tax code), SNAP benefits (USDA, state-administered), Medicaid coverage (HHS, state-administered, with rules that vary state by state), childcare subsidies (HHS Child Care and Development Block Grant, state-administered), Section 8 housing voucher (HUD, locally administered), Pell Grant for an older sibling in college (Education Department), and free or reduced-price school meals (USDA, school-district-administered). Each program has its own application, its own eligibility rules, its own benefit-cliff structure, and its own caseworker. None of them talks to the others without effort.

This fragmentation is not accidental. It is the product of more than a century of policy decisions, each made for reasons that seemed sufficient at the time. The federal-state structure of Medicaid reflects 1965 compromises with conservative governors. The work-requirement structure of TANF reflects 1996 compromises between a Democratic president and a Republican Congress. The employer-tied structure of most working-age health coverage reflects World War II wage controls and a 1954 IRS ruling that excluded employer-paid premiums from taxable income. Each layer was added without removing the previous one. The result is a system that political scientists call layered — historically accreted rather than rationally designed.

For analytical clarity, the safety net's eighty-plus programs sort into four broad categories:

Universal-by-age (or "social insurance"). These programs provide benefits to people who meet a non-means-tested categorical criterion — typically age or work-history — without regard to current income. The two giants are Social Security (retirement, disability, and survivors benefits, financed by a payroll tax of 12.4% split between employer and employee) and Medicare (health coverage for adults 65 and older and certain disabled adults, financed by a payroll tax of 2.9% split, plus general revenue and premiums). Together these two programs spend roughly $2.4 trillion a year and reach more than one in five Americans. Their politics are different from means-tested programs because the recipients are seen — by themselves and by other voters — as having earned the benefits through prior payroll-tax contributions. Social Security is the third rail of American politics for this reason: cuts to it touch a politically large and unusually mobilized constituency.

Means-tested. These programs deliver benefits based on current income (and often on household composition, work status, citizenship, and state of residence). The means-tested category includes Medicaid ($800+ billion combined federal and state in 2024, covering roughly 80 million people at peak), **SNAP** ($110+ billion, around 41 million recipients), TANF (the cash welfare successor to AFDC, now around $30 billion combined federal-state with about 1.4 million adult recipients), **housing assistance** (Section 8, public housing, and homeless programs, around $60 billion combined), childcare subsidies (around $11 billion in CCDBG plus Head Start), and the refundable portions of the **Earned Income Tax Credit** ($60+ billion) and Child Tax Credit (around $100 billion when at full strength). Means-tested programs reach more than 100 million Americans at some point in any given year, but their politics are weaker than universal programs because their recipients are politically smaller, less mobilized, and (in the case of immigrant-eligible programs) sometimes politically toxic.

Work-based. These programs are conditioned on employment history and are typically state-administered with federal standards. Unemployment insurance (around $30 billion in normal times, ballooning to over $700 billion during the COVID-19 expansion) is the largest. Workers' compensation (state-administered, $60+ billion) is universal across states but with widely varying generosity.

Categorical. These programs target a defined population for non-income reasons. Veterans benefits (around $300 billion combined VA medical, disability, education, and pension), **child welfare** (Title IV-E foster care and adoption assistance, around $9 billion federal), and the disability programs SSDI (Social Security Disability Insurance) and SSI (Supplemental Security Income) are the largest. SSDI and SSI together cover about 13 million Americans and spend around $200 billion.

Three points about this architecture matter for everything that follows.

First, the United States spends a comparable share of GDP on social benefits as most peer democracies once tax expenditures (the EITC, the CTC, the employer-health-insurance exclusion, the mortgage-interest deduction) are counted alongside direct spending. The OECD's Social Expenditure Database finds the United States in the middle of the pack on total social spending; it stands out for delivering a large share of that spending through the tax code and through employer-sponsored arrangements rather than through direct public programs. American social spending is large; it is also less visible to its beneficiaries.

Second, the United States is unusual among comparable democracies in tying healthcare coverage for working-age adults primarily to employment. Roughly 155 million Americans get health insurance through an employer; the next largest source is Medicare (~67 million), then Medicaid (~80 million but with significant churn), then individual-market plans on and off the exchanges (~24 million), then military and veterans coverage (~9 million), then the uninsured (~26 million in 2024, down from 47 million pre-ACA). No other advanced democracy has anything like this employer-anchored structure. Whether that is a strength (private-sector innovation, choice) or a weakness (job-lock, churn at job loss, administrative complexity) is a long-running debate.

Third, the safety net is administered by tens of thousands of separate institutions. Medicaid is administered by 56 separate state and territorial agencies. SNAP is administered by 50 state agencies and (in some states) by counties. Section 8 vouchers are administered by more than 3,000 local public housing authorities. The result is dramatic variation in program experience by state and even by county. The same family, with the same income and circumstances, will have a meaningfully different safety net depending on whether they live in California or Mississippi, Massachusetts or Texas. American social policy is American in the federalist sense: powerfully shaped by where you live.

28.2 Social Security — the bedrock and the looming reform

History and structure

The Social Security Act of 1935 was Franklin Roosevelt's signature New Deal achievement. It created a payroll-tax-funded retirement program for workers in covered industries, a federal-state unemployment-insurance system, and a federal-state Aid to Dependent Children program (the predecessor of AFDC and TANF). The retirement program was deliberately designed to look like an insurance program — workers paid in, accumulated credits, drew benefits in retirement — even though it operated from the start on a pay-as-you-go basis. Roosevelt himself reportedly explained the design choice this way: "We put those payroll contributions there so as to give the contributors a legal, moral, and political right to collect their pensions and their unemployment benefits. With those taxes in there, no damn politician can ever scrap my social security program."

Subsequent decades expanded coverage and benefits. The 1939 amendments added survivors and dependents benefits. The 1950 amendments brought farm and domestic workers under coverage. The 1954 and 1956 amendments added disability coverage. The 1972 amendments indexed benefits to inflation (preventing the political theater of ad-hoc cost-of-living increases) and created Supplemental Security Income for the aged, blind, and disabled poor. Coverage expanded to virtually all employed Americans and most self-employed.

The 1983 reform — negotiated by a bipartisan commission chaired by Alan Greenspan and signed by Ronald Reagan — was the last major Social Security overhaul. It raised payroll taxes, gradually raised the full-retirement age from 65 to 67, taxed a portion of benefits for higher-income retirees, and brought new federal employees under coverage. The 1983 changes were projected to keep the program solvent for 75 years. They have, in fact, kept it solvent for 40 years and counting.

The benefit formula, in plain English, works like this: the Social Security Administration takes a worker's highest 35 years of inflation-adjusted earnings (up to the annual taxable maximum, $168,600 in 2024), averages them, and applies a progressive replacement formula. Lower earners get a higher percentage of their pre-retirement earnings replaced; higher earners get a lower percentage. Workers can claim reduced benefits as early as 62 or delayed-credit benefits as late as 70. The full-retirement age is currently 67 for those born in 1960 or later.

The trust fund and the projected reserve depletion

Social Security's payroll-tax revenue has, for most of the program's history, exceeded benefit payments. The surplus has accumulated in the Old-Age and Survivors Insurance (OASI) Trust Fund and the Disability Insurance (DI) Trust Fund, which are credited with special-issue Treasury securities. Critics call the trust fund an accounting fiction (the surplus was used to fund other federal spending; the IOUs are claims against future taxpayers); defenders call it a real legal claim (the securities have to be honored when redeemed, just like any other Treasury security; the question of where the federal government finds the cash is a separate macroeconomic problem).

What is not contested is the demographic fact: the ratio of workers to retirees has fallen from about 5-to-1 in 1960 to about 2.7-to-1 today, and is projected to fall to about 2.2-to-1 by 2040 as the Baby Boom generation completes retirement. As a consequence, payroll-tax revenue stopped covering full benefit payments in 2010 (combined OASDI). The trust fund is being drawn down. The most recent Social Security Trustees' report and CBO projections converge on a key date in the early 2030s — around 2034 for OASI alone, somewhat later for combined OASDI — when the trust fund is projected to be depleted. After depletion, payroll-tax revenue will cover roughly 77 to 80 percent of scheduled benefits unless Congress acts.

This is not the same as "bankruptcy" in the colloquial sense. The program does not vanish at depletion; it pays benefits at the level current taxes cover. But that level is meaningfully below what current law promises. A retiree drawing $2,000 a month would, absent Congressional action, see benefits cut to roughly $1,560 a month. That is politically intolerable, which is why the assumption among most analysts is that Congress will act before the depletion date — but Congress has not yet meaningfully begun.

The reform debate

The arithmetic of Social Security reform is straightforward: revenue has to come up, benefits have to come down, or some combination. The politics is hard because every option creates losers and the losers are politically organized.

Raise the payroll-tax cap. The current cap ($168,600 in 2024) means earnings above that level are exempt from Social Security tax. About 6 percent of workers earn more than the cap. Eliminating the cap entirely (taxing all earnings at the full 12.4% rate) would close roughly 70 percent of the long-run shortfall, according to Social Security Administration actuaries. Critics of this approach argue that it severs the contributory link between taxes paid and benefits received (high earners would pay more without proportional benefit increases), turning Social Security into a more redistributive program than its founders intended; supporters argue that high earners have captured a disproportionate share of national income growth since the cap was set, and that asking them to contribute more is fair.

Raise the retirement age. Gradually moving the full-retirement age to 70 would close about 25 to 30 percent of the long-run shortfall. Critics argue that life-expectancy gains have been concentrated among higher-income Americans, so raising the retirement age is effectively a benefit cut for lower-income workers (who cannot reasonably continue physical work into their late sixties); supporters argue that 65 was set when life expectancy at 65 was much shorter, and that aligning retirement age with longevity is principled.

Adjust the benefit formula. "Progressive price indexing" — keeping current-law benefits for low earners, but slowing benefit growth for middle and high earners — would close roughly half the shortfall. Critics call this a stealth benefit cut that erodes Social Security's universal character; supporters argue it preserves the program's redistributive function while making it sustainable.

Means-test benefits. Limiting Social Security benefits for high-income retirees (essentially clawing benefits back through the tax system or denying them outright above a threshold) is rhetorically appealing but actuarially small (high earners are a small share of total benefits, and the political cost of breaking the contributory bargain is high).

Partial privatization. George W. Bush proposed in 2005 allowing workers to direct a portion of payroll-tax contributions into individual investment accounts. The proposal collapsed politically; markets crashed in 2008; the idea has not seriously returned. Defenders argue that long-run equity returns would generate higher retirement income and create individual ownership; critics argue that transition costs are enormous (the current generation of retirees is paid out of current workers' contributions, which would have to be replaced if those contributions were diverted) and that markets generate sequence-of-returns risk that ill-suits retirement security.

The reform debate has been frozen for two decades. Neither party wants to be the side that proposes benefit cuts; neither party wants to be the side that proposes large tax increases. Most analysts expect the eventual deal — when it comes — to combine modest tax increases (probably some cap increase), modest benefit-formula adjustments at the high end, and possibly a small retirement-age increase. The longer Congress waits, the larger the necessary changes, because the trust-fund depletion does not pause for legislative inaction.

28.3 Medicare — the second pillar

Medicare was enacted in 1965 alongside Medicaid as the central healthcare component of the Great Society. It covers virtually all Americans 65 and older, plus certain disabled adults under 65 and patients with end-stage renal disease. It is structured in four parts:

Part A — Hospital Insurance. Covers inpatient hospital stays, skilled-nursing-facility care, hospice, and some home-health services. Financed primarily by a 2.9% payroll tax (split 1.45/1.45 between employer and employee, plus an additional 0.9% on high earners under the ACA). Most beneficiaries pay no Part A premium because they paid the payroll tax during working years.

Part B — Supplementary Medical Insurance. Covers physician services, outpatient hospital care, preventive services, and durable medical equipment. Financed by a combination of beneficiary premiums (~25% of costs) and general federal revenue (~75%). The standard Part B premium in 2024 was $174.70 per month, with higher amounts for high-income beneficiaries.

Part C — Medicare Advantage. Created in 1997 (then called Medicare+Choice; renamed in 2003), Medicare Advantage allows beneficiaries to enroll in private managed-care plans that receive a per-enrollee capitated payment from the federal government in lieu of fee-for-service Medicare. Medicare Advantage has grown rapidly — from about 13 percent of beneficiaries in 2004 to over 50 percent by 2024 — and has become the dominant Medicare delivery model. Conservatives generally favor Advantage as introducing competition and cost discipline; progressives generally argue that Advantage plans are systematically overpaid (MedPAC and CBO have documented that risk-adjustment scoring leads to payments above traditional-Medicare costs for the same patient), use prior-authorization more aggressively to deny care, and that the apparent Advantage cost savings reflect favorable risk selection rather than genuine efficiency.

Part D — Prescription Drug Benefit. Created by the Medicare Modernization Act of 2003 (signed by George W. Bush, passed with bipartisan support but along largely party-line floor votes), Part D is delivered through private prescription-drug plans subsidized by Medicare. Until the 2022 Inflation Reduction Act, Medicare was statutorily prohibited from negotiating drug prices directly. The IRA changed that for a limited set of drugs, capped Medicare beneficiaries' out-of-pocket drug spending at $2,000 per year (effective 2025), and capped insulin at $35 per month for Medicare beneficiaries. The IRA's negotiation provisions are limited (10 drugs in the first round, expanding over time); they represent a meaningful but not transformational shift in Medicare drug policy.

Medicare's solvency challenge

The Medicare Hospital Insurance trust fund (Part A) faces solvency pressures similar to but more imminent than Social Security's. The most recent Medicare Trustees' report projects HI trust-fund depletion in the early 2030s, slightly earlier than OASI. At depletion, payroll-tax revenue would cover roughly 90 percent of HI expenditures.

Medicare's broader fiscal challenge is more severe than Social Security's because healthcare costs grow faster than wages and have grown faster than overall economic activity for most of the program's history. The Congressional Budget Office's long-term outlook projects that Medicare and Medicaid combined will rise from about 5.4 percent of GDP today to roughly 8 percent by 2050. The implications for the federal budget — given current revenue levels and other spending commitments — are profound.

Reform debates parallel Social Security's: raise revenue (payroll-tax increases, premium increases for higher-income beneficiaries, reductions in physician and hospital payments), reduce benefits (raise eligibility age, increase cost-sharing), or change the program's structure (premium-support / "voucher" models proposed by various Republican congressional plans, single-payer expansion proposed by progressive Democrats). None has commanded the political coalition needed to pass.

28.4 Medicaid and CHIP — federalism's social-policy flagship

Medicaid was the smaller of the 1965 twins (Medicare was originally projected to be the larger; Medicaid has since grown to be larger by enrollment, though Medicare remains larger by spending). It is the joint federal-state program covering low-income Americans, children, pregnant women, parents of dependent children, the elderly poor, and certain disabled people. CHIP — the Children's Health Insurance Program, enacted in 1997 — covers children in families with incomes too high for Medicaid but who lack other coverage.

The financing structure is a federal matching grant. The federal government pays a share of state Medicaid spending — the Federal Medical Assistance Percentage (FMAP) — that varies by state from a floor of 50 percent (in higher-income states) to a ceiling near 78 percent (in lower-income states). The ACA Medicaid expansion population (adults up to 138 percent of the federal poverty line) is matched at 90 percent federal, a much more generous rate. Federal matching is open-ended: there is no cap on federal contributions; spend more, get more.

Medicaid is administered by states under federal rules. States have substantial flexibility — to set provider payment rates (which are typically far below Medicare's, leading to provider-acceptance problems), to add optional eligibility categories or services, to apply for federal "waivers" to test alternative program designs (work requirements, premium contributions, restricted networks). The result is that Medicaid varies dramatically by state. New York's Medicaid is far more generous than Mississippi's; California's is far more administratively complex than Wyoming's.

The biggest single Medicaid policy debate of the past fifteen years has been the ACA expansion. The ACA as enacted in 2010 required all states to expand Medicaid to all adults up to 138 percent of the federal poverty line; in NFIB v. Sebelius (2012), Chief Justice Roberts held that the federal government could not coerce states into expansion by threatening their existing federal Medicaid funding. The expansion was effectively made optional. Since then, expansion has rolled out unevenly. As of 2025, 41 states plus the District of Columbia have adopted Medicaid expansion. The non-expansion states — clustered in the South — have created a coverage gap: adults too poor to qualify for ACA exchange subsidies (which start at 100 percent of FPL) but too "rich" for state Medicaid (which in non-expansion states often cuts off well below FPL for adults without dependent children). Roughly two million Americans fall into this gap.

The empirical record on Medicaid expansion is now well-established and broadly converged: expansion states see substantial declines in uninsurance rates, increases in healthcare utilization (particularly preventive care), reductions in medical debt, and — in the strongest causal studies — measurable mortality improvements among the newly covered. The Oregon Health Insurance Experiment (a randomized lottery created when Oregon could fund expansion for some but not all eligible adults) found significant improvements in self-reported health, depression, and financial security, though no statistically significant improvement in measured physical-health markers in the short follow-up. Subsequent observational studies of broader expansion (Sommers, Baicker, and Epstein's work in particular) have found mortality reductions of roughly 4 to 6 percent among newly eligible adults.

Conservative critics of expansion raise several concerns. First, Medicaid expansion grew federal spending substantially (the program has cost more than the 2010 CBO projections) and accelerated the trend of federal-state Medicaid dependence. Second, expansion crowded out private coverage in some cases (people who would otherwise have purchased exchange plans enrolled in Medicaid). Third, the 90-percent federal match is unsustainably generous; if states had to fund expansion at the standard FMAP, fewer would have done it. Fourth, Medicaid's low provider payments mean newly eligible enrollees face access problems (many physicians don't accept Medicaid; emergency-department use remains high). These critiques are real and not motivated reasoning; the question is whether they outweigh the documented benefits, and reasonable analysts disagree.

28.5 The Affordable Care Act — fifteen years on

The Affordable Care Act of 2010 was the most significant healthcare law since 1965. Its passage and implementation also became a long-running case study in the working of the American constitutional system: legislative compromise, judicial scrutiny, executive implementation, and political backlash. It is the running legislative example throughout this textbook (see Ch. 9 on bills, Ch. 14 on the Roberts Court, and case studies in this chapter).

The pre-ACA problem

In the early 2000s, the American health-insurance system had several well-documented problems. Roughly 47 million Americans were uninsured at any given moment, many more cycled in and out. Pre-existing-condition exclusions meant that people with chronic illnesses — diabetes, cancer, pregnancy in many states — could be denied coverage in the individual market or charged unaffordable premiums. The individual market was a small, dysfunctional pool with adverse selection problems (the healthy bought in only when sick, raising premiums and driving out healthier customers). Employer-based coverage was both subsidized through the tax code (the largest single tax expenditure in the federal budget, worth ~$300 billion a year) and a source of "job lock" (workers stayed in jobs they'd otherwise leave to keep coverage). Medical bankruptcy was the leading cause of personal bankruptcy filings.

A range of reform proposals had been advanced — Bill Clinton's failed 1993-94 plan, John McCain's 2008 proposal to replace the employer exclusion with individual tax credits, various single-payer proposals from progressive Democrats. The political center of gravity that produced the ACA came from a different source: the Heritage Foundation's 1989 plan, modified and signed into law in Massachusetts by Republican Governor Mitt Romney in 2006. The Romney/Heritage approach combined three policy levers: regulate insurance markets to forbid pre-existing-condition discrimination ("guaranteed issue"); require everyone to buy insurance to prevent the resulting adverse-selection death spiral ("individual mandate"); and subsidize purchase for those who couldn't afford it. Barack Obama took this design — with origins in Republican policy thought — and made it the basis of his 2008 healthcare proposal and the 2010 law.

The three-legged stool

The ACA's structure rests on three legs:

1. Insurance-market reform. Insurers cannot deny coverage or charge higher premiums based on pre-existing conditions. They cannot impose lifetime or annual benefit caps. They must cover ten "essential health benefits." They must spend 80-85 percent of premium revenue on medical care (the medical-loss ratio). Children can stay on parents' plans until age 26. These rules apply in both the employer-group market and the new individual exchanges.

2. Individual mandate. Originally, the law required all adults to maintain minimum-essential coverage or pay a tax penalty. The mandate's purpose was to prevent adverse selection: without it, people would buy coverage only when they got sick (since they could no longer be denied), driving premiums up and the market into a death spiral. The Trump tax law of 2017 reduced the mandate penalty to zero, effectively eliminating it; subsequent litigation challenged the entire ACA on the theory that the un-funded mandate could no longer be saved as a tax (which was the basis of the 2012 Supreme Court ruling); the Court rejected that challenge in California v. Texas (2021) on standing grounds.

3. Subsidies. Premium tax credits scale with income and the cost of a benchmark plan. Cost-sharing reductions further reduce out-of-pocket costs for lower-income enrollees. The 2021 American Rescue Plan and 2022 Inflation Reduction Act enhanced these subsidies, capping premium contributions at 8.5 percent of income and extending eligibility above 400 percent of FPL. The enhanced subsidies are scheduled to expire at the end of 2025; their fate is one of the most consequential health-policy questions in the 2026 Congress.

NFIB v. Sebelius (2012)

The constitutional challenge to the ACA came in National Federation of Independent Business v. Sebelius. The challenge had two main strands. First, the individual mandate exceeded Congress's Article I powers: it was not regulating interstate commerce (the activity being regulated — not buying insurance — was the absence of activity), and it was not a tax (Congress had specifically called it a penalty). Second, the Medicaid expansion was unconstitutionally coercive: Congress was using its spending power to force states to expand Medicaid by threatening their existing Medicaid funding.

Chief Justice Roberts wrote the controlling opinion. On the mandate, Roberts agreed that Congress's commerce-clause power did not reach the inactivity of failing to buy insurance, but he found that the mandate could be sustained as an exercise of Congress's tax power: it functioned like a tax, the IRS administered it, and its purpose was to raise revenue. The mandate survived. On the Medicaid expansion, Roberts (joined by a different majority) held that conditioning all Medicaid funding on expansion was unconstitutionally coercive — "a gun to the head" of the states — and that the expansion had to be made optional. The Medicaid expansion structure changed, and the post-Sebelius uneven state adoption pattern followed.

The case is constitutionally important well beyond healthcare. It is the Court's most significant Spending Clause decision since the New Deal era, and the most sustained articulation of limits on conditional federal grants to states. It is also the case where Roberts most visibly engaged in a politically attentive judicial maneuver — saving the law's central provision while reining in its federalism reach — that defined his approach to politically charged cases for a decade.

The empirical record

After fifteen years of ACA experience, the data are clearer than the rhetoric.

Coverage gains: roughly 20 million previously uninsured Americans gained coverage through some combination of Medicaid expansion, exchange enrollment, and dependent-coverage extension. The uninsured rate fell from about 16 percent of the population in 2010 to about 8 percent by 2024.

Cost trajectory: mixed. The ACA was sold partly on the promise that it would "bend the cost curve." Healthcare spending growth slowed in the years immediately after passage (2010-2013), but most analysts attribute that slowdown to the lingering effects of the 2008-09 recession rather than to ACA reforms. Cost growth resumed in the late 2010s and continues to outpace overall economic growth. The ACA did not transform the cost dynamics of American healthcare; it expanded coverage within an existing high-cost system.

Health outcomes: the strongest evidence for outcome improvements comes from comparisons of expansion vs. non-expansion states for the Medicaid-eligible population. These studies (Miller, Johnson, and Wherry; Sommers; Baicker et al.) find roughly 4 to 6 percent reductions in mortality among newly eligible adults. Causal estimates for the broader exchange population are weaker but generally positive on financial security and self-reported health. Maternal mortality and infant mortality show some expansion-state advantages but with confounds.

Political position: by the mid-2020s, both parties have effectively accepted the ACA's coverage architecture. Republican efforts to "repeal and replace" failed in 2017, with the famous late-night Senate vote where John McCain joined Susan Collins and Lisa Murkowski to defeat the "skinny repeal." Subsequent Republican policy has focused on regulating exchange plans more flexibly (allowing short-term plans, association health plans), opposing the IRA's enhanced subsidies as too generous, and supporting state-level work requirements for Medicaid. Democratic policy has focused on expanding subsidies, closing the Medicaid coverage gap in non-expansion states (through federal alternatives), and pursuing further drug-pricing reforms. The ACA's core — guaranteed issue, exchange structure, Medicaid expansion as an option, subsidized individual-market coverage — is now politically settled. The fights are at the margin.

28.6 Single-payer and the Medicare-for-All debate

The American left, particularly its progressive wing led by Senator Bernie Sanders and the Congressional Progressive Caucus, has continued to advocate for single-payer healthcare — typically branded "Medicare for All" — as the next-generation reform after the ACA. The debate reaches values questions that go beyond technical policy.

The case for

Universal coverage. A single public program, financed through taxation, would cover every resident automatically. The 26 million Americans currently uninsured (often because of citizenship issues, gaps in eligibility, churn between programs, or affordability of premiums even with subsidies) would be covered.

Lower administrative costs. Multi-payer systems require billing infrastructure for thousands of payers, denial-and-appeal processes, marketing and competition expenses. Studies — Woolhandler and Himmelstein's work being the most-cited — estimate American administrative spending at 25-30 percent of healthcare expenditures, well above single-payer systems (Canada at ~13 percent, U.K. at ~6 percent). A single-payer reform could capture some of this difference.

Lower drug prices. A single national purchaser could negotiate prices in the way comparable democracies do. American drug prices are roughly 2-4 times comparable-democracy prices for branded drugs, with most of that gap attributable to negotiation differences.

Comparative-democracies evidence. On most population-health metrics — life expectancy, infant mortality, maternal mortality, preventable mortality from causes amenable to medical care — the United States performs worse than peer democracies despite spending substantially more. The single-payer or near-single-payer systems of the U.K., Canada, Sweden, Australia, and others perform better on these metrics.

The case against

Cost shifts to taxes. "Free at point of service" is not free; it is paid through taxation. The Medicare-for-All proposals advanced by Sanders and others envision payroll-tax increases of substantial magnitude (estimates of necessary tax increases vary by analyst from $20 trillion to $40 trillion over ten years). The political viability of those tax increases is unclear; politically attractive financing options have shrunk as the math has been examined more closely.

End of private insurance. The Sanders Medicare-for-All bill would prohibit private duplicate coverage. Roughly 155 million Americans currently get insurance through their employer; 30+ million more buy individual-market coverage; many are satisfied with what they have. Polling consistently shows public support for "Medicare for All" dropping by 20-30 percentage points when respondents are asked about eliminating private insurance. The political coalition for total replacement has been hard to build.

Transition disruption. Replacing the largest sector of the American economy in a four-year transition (the Sanders proposal) would disrupt patient relationships, provider networks, employment in the insurance sector, and state Medicaid agencies. Whether the transition could be managed without dangerous discontinuities of care is contested.

Comparative systems aren't uniformly single-payer. This is the point most often elided in the American debate. Germany has a multipayer "social health insurance" system with regulated nonprofit sickness funds. Switzerland has mandatory private insurance with strong regulation and subsidies (essentially the ACA's design, but enforced and funded more thoroughly). The Netherlands is similar to Switzerland. France has a mixed system. Singapore has individual medical-savings accounts paired with catastrophic insurance. Canada and the U.K. are single-payer; Sweden's system is regional but largely public. The lesson of the international comparison is that universal coverage with good outcomes is achievable through multiple structures, not that single-payer is uniquely superior.

Worse outcomes on some metrics in single-payer systems. Single-payer systems typically have longer waiting times for elective procedures (Canada in particular has long waits for non-emergent specialist care), more rationing of high-cost interventions, and less innovation in pharmaceuticals and medical devices (the United States produces a disproportionate share of new drugs and devices, partly because high American prices reward innovation). These costs are real even if the population-health verdict remains generally favorable to peer-democracy systems.

Multipayer reform alternatives

A meaningful share of the policy-analyst community — including some progressive analysts (David Blumenthal, Jacob Hacker on his early "Medicare for some" proposals) and most center-right analysts (Avik Roy at the Foundation for Research on Equal Opportunity, Yuval Levin, Lanhee Chen) — has converged on multipayer-reform proposals: extending the ACA exchanges' subsidy structure, creating a public option, automating Medicaid enrollment for those already income-eligible, simplifying the eligibility-determination machinery, and reforming drug pricing within the existing structure. These proposals would close most of the remaining coverage gap without single-payer's transition costs. Whether they would capture single-payer's administrative savings is contested.

28.7 Welfare and the working-age safety net

From AFDC to TANF

Aid to Families with Dependent Children (AFDC), enacted in 1935 and expanded over decades, was the principal cash welfare program for low-income mothers and children for sixty years. By the early 1990s, AFDC had become deeply unpopular politically — in part because of well-documented case-by-case dysfunction (long-term dependency for some recipients, perverse incentives in the benefit-cliff structure), in part because of racialized public perceptions that political scientists have documented were not fully grounded in fact (the typical AFDC recipient was a short-term claimant, often a working-age woman moving in and out of low-wage work; the long-term-dependency caricature applied to a minority of cases).

The 1996 Personal Responsibility and Work Opportunity Reconciliation Act, signed by Bill Clinton and championed by a Republican Congress led by Newt Gingrich, replaced AFDC with the Temporary Assistance for Needy Families (TANF) block grant. The structure changed in several fundamental ways. First, the funding became a fixed block grant of $16.5 billion to states, without inflation adjustment (so the real value has eroded by roughly 50 percent over thirty years). Second, the entitlement structure ended: states were no longer required to provide cash assistance to all eligible families. Third, work requirements became central: most adult recipients had to engage in 30-hour-per-week work or work-related activities to receive benefits. Fourth, lifetime time limits (60 months in most states) capped the cumulative time a family could receive assistance.

The empirical record after thirty years is genuinely mixed.

What worked: work participation rates among single mothers rose dramatically in the late 1990s and early 2000s, reaching levels comparable to those of married mothers. Earnings for low-income families rose. Child poverty fell from about 1996 through about 2000. The combination of work requirements, the rapidly expanding EITC, a tight labor market, and federal childcare and health-insurance support produced what supporters call the "make-work-pay" success story.

What didn't work: the cash-assistance safety net atrophied. TANF caseloads fell from 12.6 million recipients in 1996 to 1.7 million by 2024 — a much steeper drop than the underlying poverty rate justified. The share of poor families receiving TANF cash assistance fell from about 68 percent in 1996 to about 21 percent by 2024 (the so-called "TANF-to-poverty ratio"). When the labor market weakened — in the 2001 recession and especially in the 2007-09 Great Recession — TANF did not respond as a counter-cyclical safety net the way AFDC had. Deep poverty (income below 50 percent of the poverty line) among single-mother families rose from the early 2000s through the 2010s. States diverted TANF funding to other purposes (in some states, less than 25 percent of TANF block-grant funds went to direct cash assistance; the rest funded foster care, marriage-promotion programs, and general state budget needs). The cash safety net became a much smaller piece of the overall safety net.

The honest verdict is that 1996 reform succeeded in some of what its supporters intended (raising work, raising incomes for those who could work) and failed in some of what its critics warned (eroding cash support for those who could not work, weakening counter-cyclical response). Both sides retain evidence.

SNAP — the working horse

The Supplemental Nutrition Assistance Program — formerly food stamps — is the largest non-health means-tested program in the federal budget, at roughly $110 billion in fiscal year 2024 with about 41 million recipients. Unlike TANF, SNAP is an entitlement: anyone who meets the income, asset, and (for non-elderly non-disabled adults) work-effort requirements can receive benefits.

The empirical record on SNAP is unusually strong and broadly bipartisan. SNAP is highly effective at reducing food insecurity. It is substantially counter-cyclical: enrollment rises in recessions, falls in recoveries. Hilary Hoynes and Diane Whitmore Schanzenbach's work documenting long-term effects of childhood SNAP receipt is particularly important: childhood SNAP exposure correlates with significant adult-life improvements in metabolic health, educational attainment, and earnings — gains that exceed program costs by typical economic-evaluation standards.

SNAP's politics have nonetheless been periodically contested. Republican farm-bill positions in 2014 and 2018 sought to tighten work requirements for "able-bodied adults without dependents" (ABAWDs), to limit categorical eligibility (states' ability to extend SNAP automatically to TANF recipients), and to restrict the benefit's use to certain food categories. Democratic positions have generally resisted tightening and supported benefit-level increases. The 2018 farm bill was a compromise that preserved most of the program structure. The political consensus around SNAP is broader than around TANF.

The EITC and the CTC

The Earned Income Tax Credit, enacted in 1975 and dramatically expanded under Clinton in 1993, is the largest income-support program by spending. It is a refundable tax credit for low- and moderate-income working families, with the refundable amount scaling with earnings up to a phase-in threshold and then phasing out at higher incomes. It is administered through the tax code (the IRS sends checks; there is no caseworker) and reaches roughly 25 million households.

The EITC is unusual in American social policy in commanding genuine bipartisan support. Conservatives (Ronald Reagan called the 1986 EITC expansion "the best anti-poverty, the best pro-family, the best job creation measure to come out of Congress") support the EITC because it is conditioned on work; progressives support it because it raises low incomes substantially. The 2024 maximum EITC for a family with three or more children was $7,830, paid as a tax refund.

The EITC has well-documented effects: it increases labor-force participation among single mothers (the 1993 expansion's labor-supply effects have been estimated by Hilary Hoynes and others as among the largest in any social-policy intervention), reduces poverty (Census's Supplemental Poverty Measure shows the EITC reducing the post-tax-and-transfer poverty rate by about 3 percentage points), and shows multi-generational effects (children in EITC-receiving households have higher educational attainment and earnings as adults).

The Child Tax Credit, enacted in 1997 and expanded several times since, is structurally similar but conditioned on having dependent children rather than on earned income. Its 2024 standard form provides up to $2,000 per child, with $1,600 of that refundable.

The 2021 American Rescue Plan Act made a temporary but consequential change. For one year, it expanded the CTC to $3,600 per child under 6 and $3,000 per child 6-17, made it fully refundable (so families with no earned income could receive the full benefit), and paid it monthly rather than annually. Researchers tracked the effects in close to real time.

The empirical record on the 2021 expansion (covered in detail in Case Study 2) shows that child poverty fell from 9.7 percent in 2020 to 5.2 percent in 2021 — the largest single-year reduction in measured American history. Food insecurity in households with children fell substantially. Hardship measures across multiple dimensions improved. When the expansion expired at the end of 2021, child poverty rebounded sharply (to 12.4 percent in 2022).

The political fate of the expansion has been contested. Senator Joe Manchin, the pivotal Senate Democratic vote, opposed permanent extension on grounds that included full refundability without a work requirement (potential effect on labor supply), the program's cost (several hundred billion dollars over ten years), and the absence of work requirements. Senator Mitt Romney and a small group of Senate Republicans proposed an alternative — the Family Security Act — that combined a similar child-allowance structure with a partial work-test and consolidation of other family-support programs. The Romney plan did not advance in 2021-22 but is the basis of subsequent conservative family-policy thinking. The 2025 tax-policy negotiations and ongoing 2026 family-policy debates will likely return to these alternatives.

28.8 The post-1965 trajectory

The American welfare state has had four major inflection points in the past sixty years.

1965 — Great Society creation. Lyndon Johnson and the 89th Congress created Medicare, Medicaid, the Older Americans Act, Head Start, the Higher Education Act (Pell Grants in 1972), Title I of the Elementary and Secondary Education Act, and the food-stamp program (made permanent in 1964, expanded through the 1970s). The Great Society was the largest expansion of the American social state since the New Deal.

Late 1970s through 1980s — retrenchment and restructuring. Stagflation and the conservative political turn produced the Reagan-era welfare reductions: cuts to AFDC and food stamps in the 1981 Omnibus Budget Reconciliation Act, the slow erosion of the real value of cash welfare benefits, the growth of the EITC (which was the bipartisan compensating mechanism for working-poor families). The shift was a recalibration toward work-based and tax-code-delivered support and away from cash assistance.

1996 — welfare reform. The TANF block grant, work requirements, and time limits restructured the cash-welfare system as discussed above.

2010 — the ACA. The largest healthcare expansion since 1965, achieving roughly 20 million additional insured through a regulated-private-market structure with means-tested subsidies.

2021 — the ARPA expansion (mostly temporary). The American Rescue Plan Act dramatically expanded the safety net — through the enhanced CTC, expanded EITC for childless adults, enhanced ACA subsidies, an expanded child-care subsidy structure, expanded SNAP, and stimulus payments. Most of these expansions were temporary; the enhanced ACA subsidies were extended through 2025; the CTC expansion was not extended.

The pattern across these inflection points: expansions that pass through Democratic Congresses with brief unified-government windows, retrenchments that follow when control shifts; programs that survive once enacted (Medicare and Medicaid have proven politically durable), programs that don't (the 1971 comprehensive child-care bill that Nixon vetoed; the Clinton 1993-94 healthcare effort that failed; Build Back Better's permanent CTC expansion that failed). The American welfare state grows in punctuated bursts, not steady trajectories.

28.9 Comparative outcomes

The cross-national health-outcome data are well-documented and worth stating cleanly.

The United States spends roughly 17-18 percent of GDP on healthcare — about $13,000 per capita as of 2023, far above any peer democracy (the next-highest is around $7,500 per capita in Switzerland and Germany; the U.K. is around $5,500; Japan around $4,800). On most population-health metrics, American outcomes are worse:

• Life expectancy at birth — about 76.4 years in the United States, compared to the OECD average of about 81 years.
• Infant mortality — 5.4 per 1,000 live births in the U.S., compared to OECD average of 3.9.
• Maternal mortality — 17.4 per 100,000 live births in the U.S., compared to OECD average of 7.0.
• Preventable mortality (deaths avoidable through medical care or public health) — substantially higher in the U.S. than in peer democracies.

The reasons are complex and disputed. Some of the gap reflects health-system performance; some reflects upstream factors that the health system doesn't fully control: higher rates of obesity and diabetes, higher firearm-related and motor-vehicle mortality, a more unequal income distribution (with lower-income Americans driving most of the comparative-mortality gap), drug-overdose deaths driven by fentanyl, racial-equity gaps in care that persist for reasons beyond pure access. Pure system-design fixes will not close the entire gap. But the system-design contribution is real: prices for the same procedures and drugs are 2-4 times higher in the U.S. than elsewhere; administrative complexity and care fragmentation generate measurable costs without commensurate benefits.

The honest steel-man on each side:

The universal-system case: peer democracies achieve better population health for less money. The American outlier status is not explained by demographics alone; the system contributes. Universal coverage has independent value beyond cost — financial security, freedom from job lock, the simple ethical proposition that no one in a wealthy democracy should die of a treatable disease for inability to pay. Multipayer or single-payer alternatives could capture much of these gains.

The market-mechanisms case: the American system, for all its flaws, generates a disproportionate share of pharmaceutical and device innovation that benefits the world; high prices fund that innovation. American medicine is the best in the world for the most complex cases (cancer survival rates for many cancers are higher than in single-payer systems). Universal-system reforms that crush prices may capture short-run savings at long-run innovation costs. The right reform is targeted: address the specific market failures (drug-pricing leverage, provider-pricing transparency, surprise billing, anti-competitive consolidation) without destroying what works.

Both arguments capture something real. Neither is conclusively settled by the comparative data alone.

28.10 Disability, housing, and family policy

Disability policy

Two federal programs anchor working-age disability support. Social Security Disability Insurance (SSDI) covers workers with sufficient work history who become unable to work because of medical disability; benefits derive from the worker's earnings record. Supplemental Security Income (SSI) covers low-income disabled adults and children regardless of work history, financed from general revenue. Together these two programs cover about 13 million Americans and spend around $200 billion. SSDI's eligibility process is famously complex and slow; appeals can take years. SSI's asset limits ($2,000 for an individual, unchanged since 1989) effectively prevent recipients from saving.

Beyond cash programs, the Americans with Disabilities Act of 1990 (Ch. 6) establishes accommodations rights and accessibility standards. Medicaid is the largest funder of long-term services and supports, covering about $200 billion of the roughly $700 billion spent annually on long-term care (the rest is split between private insurance, Medicare's limited post-acute coverage, out-of-pocket spending, and unpaid family caregiving). The aging-baby-boomer generation will sharply increase long-term-care demand over the next two decades; how to finance it is an unresolved policy challenge.

Housing policy

Federal housing assistance reaches roughly 5 million households through HUD programs — primarily Section 8 Housing Choice Vouchers (tenant-based rental subsidies, ~2.3 million vouchers), public housing (declining stock, ~900,000 units), Section 8 project-based rental assistance, Low-Income Housing Tax Credit (LIHTC) developments, and homeless-services programs. Federal housing assistance is not an entitlement; only about one in four eligible households receives any assistance, with multi-year waiting lists in most metros.

The most significant housing-policy debate of the 2020s is the relationship between zoning and housing affordability. The "YIMBY" (yes-in-my-back-yard) movement, drawing across both parties, argues that local zoning restrictions — single-family-only zones, minimum-lot-size requirements, parking minimums, design-review processes — are the central driver of housing-cost growth in expensive metros. State-level reforms in California, Oregon, Montana, Massachusetts, and elsewhere have moved (with mixed political coalitions) toward upzoning and reducing local discretion.

The diagnoses span the spectrum:

• Progressive: the housing crisis reflects underinvestment in public housing and subsidized housing; the answer is direct provision and expanded vouchers.
• Conservative: the crisis reflects excessive regulation; the answer is deregulating zoning to let markets build.
• Center-aligned YIMBY: both. Deregulate zoning and expand subsidies; the supply problem and the affordability problem are both real and require both responses.

The bipartisan acknowledgment of the housing crisis — and the cross-party YIMBY coalition (from Senator Ron Wyden's tax-credit reforms to Senator Tim Scott's Opportunity Zones to Massachusetts Governor Maura Healey's MBTA Communities Act enforcement) — is one of the more interesting recent developments in social policy.

Childcare and family policy

Federal childcare support flows mainly through the Child Care and Development Block Grant (CCDBG, ~$11 billion), **Head Start** (~$11 billion), and the dependent-care tax credits. These programs reach a small minority of eligible children; the typical low-income working family faces childcare costs that consume 20-30 percent of pre-tax income.

The United States is unusual among advanced democracies in not having national paid family leave. The Family and Medical Leave Act of 1993 provides up to 12 weeks of unpaid job-protected leave but only for employees of larger firms. State-level paid-leave programs in California (2004), New Jersey (2009), Rhode Island (2014), New York (2018), Washington, Massachusetts, and others have created a state-by-state patchwork.

Family policy has become an unusually cross-cutting issue. Conservatives have advanced family-supportive economic policies they would not have proposed a decade ago: Senator Mitt Romney's Family Security Act, the JD Vance / Compact / American Compass "common-good capitalism" framing, Senator Marco Rubio's child-tax-credit expansion proposals, the Family Hub policy work coming out of the Niskanen Center and elsewhere. Progressives have continued to advance universal child care, expanded CTC, and a paid-leave national framework. The areas of overlap — child allowances, paid leave, and substantial family-support spending — are larger than the polarized rhetoric suggests, though differences over work conditions, benefit structure, and program delivery remain real.

A note on data churn and what to trust

A practical word about the numbers in this chapter. Safety-net statistics are produced by multiple agencies — the Centers for Medicare and Medicaid Services (CMS), the Social Security Administration, the Census Bureau (Current Population Survey, American Community Survey, and the Supplemental Poverty Measure), the Department of Agriculture (SNAP data), HUD (housing data), the Tax Policy Center and the IRS (tax-credit data), and the Congressional Budget Office for projections. They use slightly different definitions, reporting periods, and methodologies. Estimates of "the uninsured rate" can vary by 1 to 2 percentage points depending on whether one uses the Census Bureau's CPS, the ACS, or HHS administrative data. Estimates of trust-fund depletion dates shift annually with each Trustees' report.

When you are evaluating a contested claim — "X program is in crisis," "Y reform reduced poverty by Z percent" — the discipline is to find the primary source, note the methodology, and verify the time period. Most heated political claims about safety-net programs survive translation into numbers; some shrink dramatically when the actual data are examined. The exercises in this chapter ask you to practice that translation.

The other thing to bear in mind: the safety net is not static. The 2026 Congress will face decisions on enhanced ACA subsidy extension, the Trump tax law's expiring individual provisions (which include the Child Tax Credit's structure), Medicare drug-negotiation expansion under the IRA, and likely a renewed push on Social Security solvency as the depletion date approaches. Whatever you read in this chapter about the current state of programs may have moved by the time you read it. The frameworks for analyzing tradeoffs — universal-vs-means-tested, public-vs-regulated-private, conditional-vs-unconditional — should remain useful regardless of which way the legislative weather has turned.

28.11 Bringing it together

The American social safety net is large, fragmented, layered, and contested. Its largest components — Social Security, Medicare, Medicaid, the ACA, SNAP, the EITC, and the Child Tax Credit — together reach more than half of all Americans at some point in their lives. They work, on the data, far better than their political reputations suggest. They also fall short of comparable democracies' achievements on outcome metrics that matter: coverage gaps, mortality differentials, child poverty levels that swing dramatically with policy choices.

The genuine values disagreement at the heart of social-policy debates is not whether the safety net should exist (both parties accept its core programs) but how generously, with what conditions, financed how, and through what institutional structure. Progressives generally favor expansion (closing remaining coverage gaps, fully refundable credits, public options or single-payer), public delivery, and unconditional supports. Conservatives generally favor work conditions, family-formation supports, market mechanisms, state-level variation, and fiscal discipline. The bipartisan zone — the EITC, SNAP, the ACA's now-accepted core, the bipartisan housing-reform openings, the new family-policy convergence — is larger than at any point in the past thirty years, even as polarization on margin issues remains intense.

The next chapter takes up education policy, which shares the social-policy frame but operates through a fundamentally different federal-state-local structure. The chapter after that turns to environmental and energy policy, where the politics have shifted dramatically over the past fifteen years.

Connections to other chapters. The ACA is the running legislative example throughout this book — see Ch. 9 (the bill's congressional path), Ch. 11 (HHS implementation), and Ch. 14 (NFIB v. Sebelius). Federalism in Medicaid is covered in Ch. 4. The 2024 election's Medicare and Medicaid politics appear in Ch. 17. Interest groups in healthcare (PhRMA, the AMA, AHIP) are discussed in Ch. 19. Economic-policy effects on the safety net (especially counter-cyclical adequacy) are in Ch. 27.

Your district. The Democracy Audit project for this chapter asks you to map the safety-net programs in your congressional district: Medicaid expansion status (and the coverage-gap population if you live in a non-expansion state), ACA exchange enrollment, SNAP participation, EITC claims, TANF caseload, public-housing and voucher inventory, and Medicare beneficiaries. Cross-reference with your representative's voting record on safety-net legislation. The exercises file walks you through the data sources.