Who Built the American Welfare State?

POSC 459 — Social Welfare Politics and Policy
Week 5 · Monday, September 21

From the async week: Howard ch. 3, "Twice in a Lifetime" · Pierson (1993)
Today: Howard ch. 4, "Ogres, Onions, and Layers," first half

Today

  1. The policy brief: your program, due tonight (~10 min)
  2. The Scaffold, and the path without it (~15 min)
  3. The async week in ten minutes: Howard ch. 3 and Pierson
  4. Howard ch. 4: the gas pedal and the brake (~25 min)

Pick a program. Due tonight, 11:59 p.m.

OASDI (Social Security) · SSI · TANF · Unemployment Insurance · Medicaid · Medicare · SNAP · the EITC

One paragraph on Canvas, about 150 words:

  • Which program.
  • Why you chose it. A personal connection, a question you already have, plain curiosity. "It looked easiest" is an honest answer and I would rather you tell me.
  • One thing you already think is true about it, which you will test against the sources.

Credit / no credit. You can switch until Friday, October 2, by emailing me.

Where the brief goes

When What Weight
Tonight, 9/21 Program selection credit / no credit
Fri 10/9 Source and claims memo: one page, two columns 5%
Wed 10/28 Policy brief, 3–4 pages 15%
Wed 12/2 Term paper, built on the brief's reform debate —

The brief has four required elements: stated purpose · who it serves (eligibility is not enrollment) · major design features, each tied to a consequence · a position on one reform debate, defended against the strongest counterargument you can construct.

The fourth carries 30 percent of the brief grade. A both-sides summary earns at most half of it.

Weeks 5 through 7 walk through most of these eight programs. Once you know which one is yours, those weeks are your research.

The Policy Brief Scaffold

A coaching module in PapyrusAI, one for each of the eight programs. It knows your program and what you were assigned to read, and it asks you questions about both. It diagnoses; it does not write.

It will not:

  • Write anything you can paste. No paragraph, no sentence, no example of the format.
  • Tell you what the reform debate about your program is, or which side to take.
  • Give you a dollar figure, an income threshold, or an enrollment number. It sends you to the primary source.
  • Summarize a reading. It points you at the document.

Opens tonight. Ungraded: the module, the log, and whether you open it at all.

Getting in, and what to do first

  1. PapyrusAI in the left-hand navigation of our Canvas course.
  2. If you have never opened it: set up the account and enter the course code exactly, FALL2026-POSC459. A mistyped code gives no error. It puts you somewhere that is not this class.
  3. Open Policy Brief Scaffold, the one for your program.

Before your first session, read something. The module opens by asking what you have read. If the answer is nothing, it names two documents for your program and stops.

The minimum is the administering agency's own description of your program plus one source document. Both are on the Policy Brief — Program Sources page, in Start Here and again in the Week 5 module. The module cannot show you the documents; get them from that page.

Three sessions, 25 to 40 minutes each

Session When What it pushes on
1 After you pick, late September Say in your own words what the program is designed to do, who it serves, how it is built. Three confusions: eligibility vs. enrollment · financing vs. benefits · authorization vs. appropriation
2 Before the memo, due 10/9 Every claim in your supported column: which document, which section, what it says. A source you found through AI is an assumption until you have read it
3 On a complete draft, mid-to-late October Four failures: a claim with no citation · a design feature never connected to a consequence · a figure that appears in no source you named · element 4 written as a summary

The log. Download the conversation from its three-dot menu and attach the file to your Canvas submission. That file is your AI disclosure for work done inside the module. Nothing downloads on its own, and the tool's data may not be kept indefinitely.

I can see your conversations from the instructor side, archived ones included.

The path without PapyrusAI

You can complete this entire course without PapyrusAI, and taking that path costs you nothing in points or standing.

The memo due 10/9 asks for one conversation about your two columns before you submit. Three ways to have it, worth exactly the same:

The conversation What you attach
A Scaffold session in PapyrusAI The conversation you downloaded
The CSUF Writing Center, in person or virtual The confirmation email
Me: office hours (Mon 3:00–4:00 and 5:30–6:30, Thu 12:00–1:00), by appointment, or email me your draft columns and I will write back Our exchange, forwarded
  • It is a completeness check. It is never scored; what gets graded is the memo.
  • You owe me no reason for your choice, and you can change it at any point. Anything you have already produced still counts.
  • For the brief itself, the same pushback from a person: send me an outline the week before rather than the night before.
  • If none of the three works for you, tell me and we will find a fourth.

The async week: two readings, one question

How did we get the welfare state we have, and why does it stay?

Howard, chapter 3, answers the first half. He counts when the major programs were enacted and who held power when they were.

Pierson (1993) answers the second half. Political scientists had long treated policy as the thing to be explained. Pierson reads four books and argues that policies are causes too. Schattschneider's line, which he opens with: "new policies create a new politics."

Discussion Paper 1 asked you to put one of them against Growing Up Poor in America. Those came in last night.

Howard ch. 3: twice in a lifetime, or all the time?

The standard story: two big bangs, the New Deal and the Great Society, and "the most obvious thread linking these two periods is Democratic control of government" (p. 55).

Howard's recount: take all six tools from Week 3 and about fifty major programs appear in nearly every decade, starting in 1913 with the mortgage interest deduction.

"New social programs do not appear once in a lifetime, or even twice. They happen all the time." (p. 60)

Who held power, 1900–2000 Social insurance, grants Tax breaks, regulation, loans, insurance Total
Unified Democratic (36 yrs.) 15 13 28
Democratic president, divided govt. (12 yrs.) 1 4 5
Republican president, divided govt. (28 yrs.) 5 8 13
Unified Republican (24 yrs.) 0 3 3

Howard, table 3.2.

Divided government produced eighteen major programs, among them disability insurance, Section 8, the EITC, ERISA, and the ADA. Party control decides less about whether a program gets built than about which tool builds it.

Pierson: when effect becomes cause

"Policy feedback" is a label covering several different arguments. Pierson sorts them by mechanism and by who is affected:

Government elites Interest groups Mass publics
Resource and incentive effects Administrative capacities "Spoils" · organizing niches · financing · access "Lock-in" effects
Interpretive effects Policy learning Policy learning · visibility and traceability Visibility and traceability

Pierson 1993, figure 1, p. 626.

"If interest groups shape policies, policies also shape interest groups." (p. 598)

His example: health benefits for the elderly were thin, which gave the AARP a niche. It sold health insurance before Medicare and "Medigap" policies after, a reason to join that had nothing to do with lobbying.

He bets the right-hand column will matter most (p. 597). Pierson tells us why programs stay. Howard's next chapter asks who keeps building them.

Who signed it?

Democrat or Republican in the White House when each became law. Hands up.

Program Year
Disability insurance 1956
Supplemental Security Income 1972
Section 8 housing vouchers 1974
The Earned Income Tax Credit 1975
The Americans with Disabilities Act 1990
Medicare Part D, prescription drugs 2003

The gas pedal and the brake

"If the American welfare state were a machine, then Democrats would be the gas pedal and Republicans would be the brake." (p. 73)

Chapter 3 left that picture mostly standing. Chapter 4 opens on why it should have been truer than ever after the Great Society:

  • In 1960 all twenty-two senators from the former Confederate states were Democrats. By the 1990s about one-third were.
  • On Poole and Rosenthal's ideology scores, where higher is more conservative, the average House Republican went from 0.267 in 1965–66 to 0.429 in 1995–96. Gerald Ford scored 0.271. Newt Gingrich scored 0.471.

"The ideological divide separating the two major parties was larger in the 1990s than at any point since before the New Deal." (p. 74)

Add divided government for most of those years. The prediction is gridlock. New programs kept arriving anyway.

Three programs that should not exist

ERISA ADA Child Tax Credit
Year enacted 1974 1990 1997
Policy tool(s) Social regulation, insurance, tax expenditure Social regulation Tax expenditure
Target group Retirees Disabled Families with children
President Republican Republican Democratic
Congress Democratic Democratic Republican

Howard, table 4.1, "Policy Innovation under Divided Government."

He chose them for four reasons: nontraditional tools, real size, little scholarly attention, and three different versions of divided government.

What he expected to find: "classic logrolling behavior." Republicans hold their noses and let a Democratic program through in exchange for something they want.

What to watch for in each case: who introduced it, who stalled it, and what the Republicans in the story thought the program was for.

ERISA: seven thousand workers at Studebaker

1963. Studebaker goes bankrupt and terminates its pension plan. Seven thousand workers lose hope of a full pension, at a time when Social Security checks were far smaller than they are now.

"It was what political scientists call a 'focusing event' (and what everyone else calls a disaster)." (p. 77)

A congressional study of 1950 to 1970 found that the vast majority of people who worked at firms offering pensions "never received a penny in benefits." Many never qualified because of "extraordinary vesting requirements." Some who did qualify lost out to badly managed funds.

The Employee Retirement Income Security Act of 1974 used three tools in one statute:

  • Regulation of private pension plans, enforced by Labor and Treasury.
  • Insurance: the Pension Benefit Guaranty Corporation, a new federal agency that pays when a plan fails.
  • A tax expenditure: the Individual Retirement Account.

"Based on the sheer number of lives touched, the passage of ERISA is arguably the third 'big bang' of the American welfare state." (p. 77)

ERISA: who pushed, who stalled

Pushed. Jacob Javits, Republican of New York, was the "driving force" from 1967 to 1974: his staff drafted the bill, he held field hearings in major cities, and in 1972 his office received twenty thousand letters in two weeks. His partner from 1971 was the committee chair, Harrison Williams, Democrat of New Jersey.

Stalled. The Chamber of Commerce and the National Association of Manufacturers. Wilbur Mills at Ways and Means and Russell Long at Finance, both Democrats. Lyndon Johnson, who refused in 1964 and withheld support in 1967. The Nixon administration, which wanted reform "as toothless as possible."

What broke it loose. Watergate: a weakened White House, and the IRA added to placate it. Congressional reform that cut down committee chairs. Gerald Ford signed it on Labor Day 1974.

"It would be fair to say that ERISA was a bipartisan effort led by a Republican legislator." (p. 80)

Why would a Republican build this?

Congress raised Social Security benefits by 20 percent in 1972 and indexed them to prices. Republicans watched the payroll tax climb with alarm. Javits on what ERISA was for:

"In order to keep the Social Security system viable, these taxes have been increased to the point where employers and workers are now paying the government more than 12 percent of the wage package for Social Security taxes. If there were no ERISA, no insured private pension plans, we would have to pay about 20 percent of our income to the government to provide an adequate income for retired persons. If we allow the government to support the system of retirement benefits to such a degree, the state will become too large a factor in our lives." (p. 89)

A new social program, built to keep an older one from growing.

Put Pierson next to it. Social Security had created its own politics: beneficiaries, election-year increases, a tax nobody could cut. Repeal was off the table. What was left was to build around it.

Is a program built to contain another program an expansion of the welfare state, or a limit on it?

Due and next

  • Tonight, 11:59 p.m.: policy brief program selection, on Canvas. The Scaffold opens once selections are in.
  • Wednesday: finish Howard ch. 4 (pp. 81–91). Howard expected logrolling and found something else; read asking what evidence would tell the two apart. Then Berkowitz and DeWitt, "Social Security" and Oberlander, "Medicare" (Oxford Handbook chs. 15 and 17, in the Week 5 module).
  • Also for Wednesday, one short document, by last name. A–L: the Heritage brief (Greszler, 8 pages). M–Z: the CBPP fact sheet (10 pages). Both are in the Week 5 module. We use them in class: what does your document need the origin story of Social Security to be?
  • Office hours today: 3:00–4:00 and 5:30–6:30.
  • Discussion Paper 2 is due Friday, October 2.