The Social Welfare Domain: Care vs. Fairness

The bipartisan baseline: Americans want more, not less

The starting point for any honest analysis of spending attitudes is the topline trajectory of the most-asked questions. Americans have always told the GSS, by large and growing majorities, that the federal government underspends on most of what it does.

The education line is the clearest. In the early 1970s, half of Americans wanted more education spending. By the mid-2000s the figure was over 70%; the 2024 reading of 76% is the highest in the series. The health line's profile is similar but starts higher — 63% in 1973, peaking at 78% in 2004 (the year of the Medicare prescription-drug expansion debate), settling at 74% in 2024. Even welfare — the program with the loudest partisan rhetoric and the lowest approval — has grown more popular over fifty years: 21% in 1973, 33% in 2024.

This is one of the more counterintuitive findings in this book, and it bears repeating. The common rhetoric of contemporary American politics — "people want less government" — is, on the evidence of the country's longest-running survey instrument on the question, false as a description of what people say. Americans have not become more skeptical of federal spending over the last fifty years. They have become more enthusiastic about it, on every major domain except in periods immediately following sharp tax-and-deficit debates.

What divides the parties is not the appetite. It is the framing.

Education and Social Security: the bipartisan domains

Two of the four major social-welfare programs in the GSS battery enjoy genuine bipartisan supermajority support, in the sense that majorities of both Democratic and Republican voters say the federal government spends too little. Education is one. Social Security is the other.

Two-thirds of Republicans say the government spends too little on education. That is a fact worth holding alongside the contemporary discourse around "school choice" and "the education establishment." Whatever Republicans are angry about when they discuss education in cable-news terms, they are not telling the GSS that they want the federal government to spend less on it. They are telling the GSS that they want it to spend more.

Social Security is the cleanest case in this chapter of an "earned-benefit" social program — paid into during working life, drawn down in retirement. It is bipartisan to the point of being non-political. In 2024, more Republicans than Democrats said the federal government spends too little on it. Whatever the partisan dynamics around social spending are, they are not running through Social Security.

Health: the same domain, with the ACA's footprint

Health is the third large social-welfare domain, and is the one where partisan dynamics are most visible. From 1973 through the late 2000s, the partisan gap on health spending was modest — typically 7 to 15 points, with Democrats more enthusiastic than Republicans but both saying "too little" by majority margins. In 2010, that changed.

The 2010 reading is one of the largest single-wave partisan reversals in the GSS's entire spending battery. Republican support for "too little on health" fell 27 points in two years; Democratic support also fell, by 14 points (83% → 70%), but considerably less. This is a direct instance of the priming effect that political scientists describe in the post-Reagan partisan literature: when a domain becomes politically named (here, "Obamacare"), the partisan signal in the responses to a generic question about that domain changes, even when the underlying preferences may not have. Republican respondents, asked about "improving and protecting the nation's health" in 2008, said too little by 68%. The same partisan coalition asked the same question in 2010 said too little by 41%.

The recovery in subsequent years is partial but real. By 2024 the Republican reading is 64% — still above the 2010 trough, still below the pre-ACA baseline. The Democratic line dipped through the early 2010s (around 68–70% during 2010–2014) before recovering to roughly 80% by 2018 and holding there. The 2024 partisan gap on health spending (15 points) is, by coincidence, back to where it stood in 2008 (16 points) — but the intervening 2010–2014 period roughly doubled it. The institution did not change. The question did not change. The political naming changed.

Welfare: where the framing matters most

The "welfare" item in the GSS spending battery is famous among survey researchers for a particular kind of asymmetry: the partisan difference is not so much in whether people want welfare to grow as in whether they call it a budgetary priority. Even Democrats are not enthusiastic about more welfare spending — only 49% of them said too little in 2024, the highest reading in fifty years and still a minority. The partisan-divide story on welfare lives mostly on the opposition side.

Two stories run through Figure 5.

First: the Democratic line on "too much welfare" has fallen by 32 points over fifty years (46% in 1973, 14% in 2024). This is one of the largest fifty-year shifts within a single partisan coalition on any spending question in the GSS. Democrats in the 1970s, including Democrats in northern industrial states, were nearly as opposed to "more welfare" as Republicans. Democrats today, even allowing for sample-composition changes, are not. The Democratic coalition has moved leftward on welfare in a way it has not moved on (say) Social Security.

Second: the Republican line has held mostly steady. 66% of Republicans said too much welfare in 1973; 52% say so in 2024. There is real movement — a 13-point softening over fifty years — but it is much smaller than the Democratic shift. The widening partisan gap on welfare is not a story of two coalitions diverging. It is a story of one coalition moving while the other holds.

The same dollar, framed in different words

The 2024 partisan gap on welfare spending ("too little") is 34 points. The 2024 partisan gap on the same question framed as "improving the conditions of Black Americans" is 46 points — the largest spending-domain gap the GSS reports.

The shape of Figure 6 is, in some ways, the canonical "sorting" chart: a Democratic line that moved sharply during a specific cultural moment, and a Republican line that did not. The post-2014 inflection on the Democratic line — from 47% to 74% in eight years — is one of the sharpest movements in the entire spending battery. It is roughly contemporaneous with the Ferguson protests, the killing of Michael Brown, the formation of Black Lives Matter, and the long-tail visibility of the issue in Democratic-aligned media. None of those events drove a corresponding Republican movement.

What is notable about Figures 5 and 6 read together is that the dollar amount in question is, in policy terms, the same dollar. "Welfare" and "improving the conditions of Black Americans" are largely overlapping sets of programs — TANF, SNAP, housing assistance, job training, early-childhood programs. The GSS asks about both, and the Democratic line on the second framing is 20 points higher than on the first (D=69% on race, D=49% on welfare). Republican readings show the same direction at smaller magnitudes (R=23% on race, R=15% on welfare). What this difference primarily measures is not how the respondent feels about transferring resources to low-income Americans. It is how the respondent feels about the words the survey uses for that transfer.

The racial breakdown on welfare confirms what the partisan breakdown suggested: this is a question on which respondents' demographic and partisan identities sort together. Black respondents have been consistently more supportive of welfare than White respondents throughout the series. The gap has narrowed somewhat — from 43 points in 1973 to 24 in 2024 — but it has not closed. The partisan story is on top of the racial one.

Childcare: the new frontier

One last piece of the social-welfare battery is worth flagging because it is the rare domain where the GSS shows a real recent rise in both parties.

Childcare, as a survey item, is a relatively recent addition (the GSS first asked it in 2002) and is presented in family-policy framing — not "child welfare," not "income transfers to low-income parents," but childcare assistance as a generic category. Both partisan coalitions have moved upward on it, and the Republican line in particular has moved more on childcare than on any of the welfare-adjacent items in the long-running battery. Whether this is a leading indicator of where the family-policy debate is going (the contemporary push for an expanded child tax credit, for example) or a one-time political moment, the data cannot say. But it is a place where the bipartisan-baseline story of this chapter shows up most cleanly.

What this chapter changes

Three things to take into the rest of Part II.

First, the polarization story on social spending is not a story about the size of government. Americans want more, not less, on every domain in this battery, and the trend over fifty years has been toward more, not less. The story is about which domains the parties have agreed to count as legitimate uses of federal money. Education, Social Security, and (to a lesser extent) health, the parties agree on. Welfare, the parties agree to disagree on. "Improving the conditions of Black Americans," the parties have decided to disagree on more sharply with each passing decade.

Second, the 2010 ACA episode in Figure 4 is a clean illustration of how priming works in survey data. A question that produces broadly bipartisan support in one year can produce a 27-point single-coalition shift two years later, with no real change in the underlying programs, simply because the political naming of the domain has been claimed by one coalition. Most of the partisan-gap widening this chapter documents has this character: not a divergence of underlying preferences, but a divergence in which framings each coalition will accept as legitimate.

Third, the welfare/race overlap is the single sharpest example in this book of demographic + partisan reinforcement. The partisan gap on welfare is amplified by the racial gap on welfare; the racial gap is amplified by the partisan gap. Neither alone fully describes the structure of the divide. The two together do.

The next chapter turns from social-welfare spending to security and order — military, police, border control — where the too little / too much asymmetry runs in the other direction.