Did the gains reach households?

Where did the benefit stop?

“The question is not who was helped along the way — it is where the benefit stopped.”

Thom Barrett, Where the Benefit Stopped, section 1.

When public support for enterprise is defended as a benefit to everyone, the citizen is entitled to ask for the result. A larger economy, higher profits or a rising stock price does not answer that question. Show what reached households. Show what they could afford after the bills were paid.

The record here gives that question weight. Family health premiums rose faster than the cited hourly wage measure over 2005–2025. Nearly half of renters faced high housing costs in 2024. The top 1% held a larger share of wealth in 2026 than in 1989. These are different tests over different periods. Together they give citizens reasons to demand proof of broad benefit; they do not identify which policy caused each result.

Some gains did reach households. Real median income rose over the longer periods checked here. That fact belongs in the record. The question is whether a particular promise was delivered, to whom, and at what cost. “The economy grew” is the start of that inquiry.

The public-benefit standard proposed here: when public resources support a private activity, name the promised household benefit, measure delivery and assign responsibility for correcting a shortfall.

Use the citizen request below or examine the evidence by topic.

The question / the measurement / the limit

A result that changes with the period

Health premiums and wages

KFF reports different relative growth over different windows. Select a period; both bars always use the same dates and a zero baseline.

A recent period in which wages grew faster qualifies an “always outruns wages” claim. It does not establish that coverage is affordable for every worker.

KFF 2025 survey, summary and Figure 1.15. Employer-sponsored coverage; nominal percentage changes. KFF’s wage comparator is not automatically the same as PF-01’s production/nonsupervisory wage series.

Distribution, not a national average

Renter housing costs, 2024

The middle segment is approximately 23%, calculated from rounded shares. These are renter households, not all households. Harvard Joint Center for Housing Studies, 2026 report release, using 2024 observations.

Who owns the homes? Start with the right denominator.

Census reports a 37.9% homeownership rate for households whose householder was under 35 in the fourth quarter of 2025. This measures households, not all people under 35. The 90% margin of error is ±0.8 percentage points.

Only November and December were surveyed: a funding lapse prevented October collection. Census cautions against comparisons with other quarters. This snapshot does not establish the chapbook’s claimed ten-point fall since 1980. Census Table 6 and survey notes.

GAO studied Cincinnati, Dallas, Jacksonville, Nashville, Phoenix and Seattle. It counted investors with at least 5,000 single-family homes nationwide and holdings in at least five metro areas. The selected sample cannot support a national estimate. GAO, March 24, 2026.

Ownership concentration deserves scrutiny, but concentration alone does not establish its effect on prices or access. GAO’s 2024 review found evidence of possible price and rent increases and neighborhood stabilization; effects on homeownership opportunities and tenants remained unclear. GAO’s research review.

Make a local housing request that can be answered.

Ask the city or county housing office: “What evidence do you hold on investor ownership, rents and first-home access in this area? Please identify the dates, geographic boundary, owner-size threshold and housing stock counted. Which housing decisions fall within your office’s authority, and what result will you measure?”

Keep the response with dated local ownership records, rent notices or purchase figures. Separate a recorded owner from its verified parent company. Track the answer in the Action Workbench; a national percentage cannot answer a neighborhood question by itself.

Student debt: the total is only the first question.

The New York Fed reported $1.651 trillion in student debt in the second quarter of 2026. That was $7 billion less than the preceding quarter and $13 billion more than a year earlier. Its estimate comes from a national sample of anonymized credit reports. August 11, 2026 report release and table.

This total is not the average borrower’s balance. It does not tell us how long borrowers take to repay, or whether their interest grows faster than their wages. Those claims need a defined group of borrowers, matched dates, income and interest records, and a record of actual repayments. Different debt datasets can also have different coverage.

Compare the monthly bill, total cost and end date.

Federal Student Aid’s calculator estimates eligible plans, monthly payments, total principal and interest, and repayment end dates. Eligibility depends on loan type and disbursement date. Estimates are not final terms; the servicer sets those after processing. A lower monthly payment can mean a longer term and more interest. Federal Student Aid’s calculator guide.

Keep your loan type, disbursement date, principal, interest, payment history and current plan together. Ask the servicer: “Please reconcile my current balance with my payment history. Show what went to principal and interest, identify any other balance changes, and explain my current payment and projected end date with the assumptions used.”

For the public-policy question, ask your representative what a proposal changes in the monthly burden, total cost and access to education, and for which borrowers. Record the response in the Action Workbench. A change in the national balance alone does not answer those questions.

Who received compensation growth?

Two pay series, one period

EPI reports real compensation growth of 1,094% for CEOs and 26% for its typical-worker measure between 1978 and 2024. Starting both at 100 makes the ending levels 1,194 and 126.

The worker series increased; it did not keep pace with the CEO series. This is a distribution comparison, not evidence that workers received no gains.

EPI’s published figures reproduce all five PF01 pay claims: about 281 to 1 in 2024 using realized CEO pay, 212.6 to 1 using granted pay, 21 to 1 in 1965, and 60 to 1 in 1989, alongside the growth figures above. EPI labels its 2024 CEO estimate a projection. Its ratios average firm-specific comparisons; the historical worker-growth series and the industry worker comparator used in the ratios are distinct. Read the definitions and verification limits.

Spending is not the same as benefit received.

CMS reports national health expenditure of $5.3 trillion in 2024: $15,474 per person and 18.0% of GDP. These are spending measures, not each person’s bill or a measure of health outcomes.

CMS National Health Expenditure fact sheet. To test benefit received, add coverage, access, financial burden and health outcomes for a defined population.

What would establish the broader claim?

A complete test needs matched-period evidence on household income, essential costs, benefits and access, including households that improved and households that did not. It also needs a separate causal design before assigning a result to a particular law or institutional practice.

Legal authority and household outcomes are separate questions. Explaining what a legal rule permits does not measure whether a household gained. Measuring hardship does not, on its own, determine the rule’s legal status.

A price comparison with its method visible

Infant care and college tuition

PF-01’s $15,636 infant-care figure is confirmed in Child Care Aware of America’s equal-state average. Its other methods yield different prices:

The first row reproduces PF-01’s 30.8% comparison. All three exceed the college tuition figure. Their spread shows why the averaging method belongs beside the number.

Child Care Aware of America: prices and three methods. These are alternative estimates, not three years or three bills. The report’s $13,184 overall average combines ages and settings; it is not the infant-center figure. Its methods note says the “2025” collection covers December 2025 through March 2026.

College Board reports $11,950 in published tuition and required fees for full-time in-state undergraduates at public four-year institutions in 2025–26. This excludes living expenses and does not subtract aid. College Board: 2025 pricing highlights.

This compares the scale of two prices. It does not compare identical services, the same households, or what a family pays after assistance. It is not a measure of price growth.

Inspect the inputs, arithmetic and limits

Household figures you can trace

This reading table incorporates the source checks and two numerical corrections. Each row links to its definitions and limits. Three housing questions remain: the exact historical home-price ratio, the severe-burden percentage for low-income renters, and long-run rent versus income growth. The 83% total-burden rate is now confirmed and included below. One has a specific discrepancy: Harvard’s indexed 2026 report text gives a 4.7 home-price-to-income ratio for 2025, while PF01 says 5.0. Direct report-page inspection is still pending, so this row is excluded from the checked table.

Low-income renters: keep the denominator visible

In 2024, 83% of renter households earning under $30,000 spent more than 30% of income on housing. This is the rate for a low-income group, not all renters. The housing-cost measure includes utilities.

The severe-burden figure needs reconciliation. Harvard’s June summary says 66% of this group spent more than half their income; the original workpaper gives 67%, as does the publicly available America’s Rental Housing 2026 PDF, printed page 25. The unrounded tables and edition dates are still needed to explain the difference. Neither value is presented here as settled. Severe burdens are part of total burdens, so the percentages must not be added.

Read the counting rules before comparing households

Figure 21 of the rental report adjusts household income for inflation with the Consumer Price Index. It counts households with zero or negative income as severely burdened, and those owing no rent as unburdened. These are stated measurement rules. A calculation that drops those households may produce a different rate.

Read Figure 21 and its notes, printed page 26. The available file name includes “embargoed”; no identity with a later edition is assumed.

Ask for both reports’ unrounded tables: “Please identify the data year, income cutoff, household exclusions and edition behind each severe-burden percentage.” Save the answer with the table. A rounded percentage alone cannot explain the difference.

Harvard’s June 17, 2026 summary, item 8 and Figure 7 notes · Readable check record

Which source claims have been checked?

Six housing rows, the family-premium row, national health spending, and student debt now match the original publishers. The rental-stock threshold is inflation-adjusted. The health-premium figure covers employer-sponsored insurance, not every household medical expense. National health spending includes public and private payers; its per-person amount is not a family’s out-of-pocket bill. Student debt is the balance still owed, not an annual expense. The exact home-price historical comparison and remaining long-run calculations are still open.

Read the row-by-row verification and limitsHarvard housing releaseKFF employer surveyCMS health spendingNew York Fed student debt

Food costs: prices and the household burden

The grocery-price claim checks out: the BLS food-at-home price index rose 33.0% from its 2019 annual average to July 2026. That measures prices for a basket of groceries. A family’s spending also depends on what it buys and how much.

The income comparison checks out too. In 2024, food spending represented 33.0% of before-tax income for the lowest-income fifth, compared with 6.4% for the highest-income fifth. The same necessity takes a much larger share of the smaller income.

These are different measures. One tracks grocery prices over time; the other compares food spending across income groups in one year. Use each for the question it answers.

BLS grocery-price series · USDA food and income comparison · Readable verification record

Which income is being counted?

The Congressional Budget Office reports these shares for the top 1% of households:

“Before” already includes Social Security, Medicare and other social insurance. “After” adds means-tested benefits and subtracts federal taxes. Some benefits are services, not cash a family can spend.

CBO also finds that average real income rose in every fifth of the distribution from 1979 to 2022. Growth was greatest at the top. That supports a finding of unequal gains, not a claim that no one else gained.

CBO report and definitions, January 2026

Why another study gives different numbers

Auten and Splinter’s May 2025 workbook puts the top 1% share in 2022 at 15.5% before tax and 10.3% after tax. Both figures match the original paper’s table.

Their national-income measure excludes capital gains. Their estimates also allocate public spending and deficits. These choices help explain why their figures differ from other studies. A smaller number does not establish a universal lower bound.

Authors’ methods, pages 7–9 · Published data, table T-A1, row 67

What WID counts

The World Inequality Database reports 20.73% in 2022 for the top 1% of adults under its pre-tax national-income measure. It divides income equally within couples, then ranks adults. This is different from ranking households. Pension payments are already included in this income measure.

The download repeats 20.73% in 2023 and 2024. The files checked do not establish whether those years were independently estimated or carried forward. Repeated values are not proof that inequality stayed flat.

WID United States data · Definitions

Do not average these studies or present their spread as a margin of error. Each estimate needs its own definition and source.

Citizen test: When an official promises broad gains, ask which households gained, what costs they faced, and whether the measure includes benefits they cannot spend. Request the same measure before and after the policy.

Open the readable verification record

Labor’s share: two measures, two definitions

Each latest reading is the lowest among the 318 quarters checked from 1947 through 2026 Q2. The first measure is an index: 93.446 does not mean workers received 93.446% of income. The second is a calculated percentage, using employee compensation and gross domestic income.

These measures have different scopes. Neither tells us that every worker’s real pay fell, and the remainder is not automatically corporate profit. Read them alongside actual pay and household costs.

BLS index · BEA compensation · BEA domestic income · Calculations and full-period checks

Health premiums: use the right comparison

KFF’s current table lists average family premiums of $10,962 in 2005 and $26,993 in 2025. That is a 146.2% increase: divide the later amount by the earlier amount, subtract one, then multiply by 100. PF01’s original table below says 148.1%; that number has not been reproduced from this source.

The period matters. Over 2020–2025, KFF reports family-premium growth of 26%, wage growth of 28.6%, and inflation of 23.5%. Premiums did not outpace wages over that interval. For 2005–2025, the downloaded BLS series reproduce PF01’s wage increase of 94.5% and inflation of 64.9%, using calendar-year averages. The 2025 CPI average contains eleven months: October data were unavailable during the funding lapse.

For a citizen’s own comparison, use the same dates and distinguish the total insurance premium from the part paid out of the paycheck. The 2025 survey covers employers with at least ten workers; earlier survey coverage differs.

KFF tables, Figure 1.12 · KFF explanation · BLS wage series · BLS price series · Readable calculation record

Profits grew. Where did the benefit go?

Adjusted after-tax corporate profits divided by GDP reached 12.07% in the second quarter of 2026. The average quarterly ratio for 1947–1999 was 6.13%. Checking all 318 quarters from 1947 through this observation confirms that 12.07% is the highest reading in these downloaded series.

This establishes the scale of profits relative to economic output. To test the rising-tide claim, it still needs to be read alongside household income, essential costs, access, and distribution. A high profit ratio alone does not show who benefited or what caused the result.

The numerator includes BEA’s inventory and capital-consumption adjustments. Both inputs use current dollars at annual rates. This ratio is not a profit margin calculated from company sales.

BEA profit series · BEA GDP series · Calculation, dates, and record test

Income gains: name the household and the period

These calculations reproduce PF01’s figures. They show gains, with very different results depending on the period and population. Household income can include more than one earner. Weekly earnings describe full-time wage and salary workers aged 16 and older, before deductions; they exclude self-employed workers.

The weekly comparisons use the second quarter in each year. Neither series follows the same people over decades. Changes in household size or who is working can affect the median.

Census household income · BLS weekly earnings · Inputs and calculations

Tuition: keep the college categories consistent

Public four-year tuition and required fees rose about 244% after inflation between 1980–81 and 2023–24. NCES lists $2,872 and $9,872 in constant 2023–24 dollars.

For all private four-year institutions, the comparable amounts are $12,925 and $35,911: an increase of about 178%. PF01’s original 205% figure can be reproduced by using the later nonprofit-private amount, $39,408, against the earlier all-private amount. That mixes categories. Use 178% for this table’s all-private comparison.

These are published tuition and fees, not what students pay after grants and other aid. They exclude room and board. The table’s institutional coverage also changes over time, so it does not follow an identical set of colleges.

NCES table and definitions · Readable calculation record

Who holds the wealth?

Both columns total 100%, and all eight values match the Federal Reserve’s published series. Net worth means assets minus debts. It is different from income earned during a year.

The top 1% held a larger share at the later date. That does not mean every other group lost wealth in dollars: a smaller share of a growing total can still be worth more. These groups also need not contain the same households at both dates.

The estimates combine the Federal Reserve’s aggregate accounts with its household finance survey. They are not a new census of households every quarter.

Top 1% series · Next 9% series · Next 40% series · Bottom 50% series · Verification record

Complete numerical source inventory

The rest of PF-01 remains inspectable.

The tables below preserve the supplied workpaper’s numerical entries, dates and source labels. They are attributed source claims, not additional independently verified findings of this HTML. The checked household table above incorporates the tuition, premium, food, and other completed checks. The original figures below remain unchanged for comparison, including the two figures corrected above. Three housing comparisons and the status of WID’s later-year estimates still require completion of their underlying checks. The published CBO, Auten–Splinter and WID income-share values are checked above. The wealth-share table is checked above. The real-income growth calculations are checked above. The pay claims match EPI’s published tables with the projection and method limits explained above. Both labor-share rows are checked above. The profit ratio, historical average, and record test are checked above.

The workpaper’s “uncontested” classifications are not a guarantee against revisions, definitional differences or errors. Its productivity/pay and top-income disputes remain unresolved here.

Download the numerical source register

Ask for the proof of public benefit

Choose one public subsidy, contract, tax provision or other support measure. Save the document that promises a public benefit. Address your request to the office that administers the measure, or to the legislative office asking citizens to support it. Name the measure and the promise; do not accuse an individual on the strength of a national chart.

Record the date sent, the office, the promised reply date and the records received. Compare the reply with the original promise. Separate a missing record from a record showing a failed result. If the office redirects you, preserve that response and follow the named route.

Use the citizen workbench to prepare and retain your action record. A useful outcome is a checkable answer, a corrected claim or a dated commitment to measure delivery. This is a request for explanation; formal records requests must follow the relevant office’s procedure.

Keep the test reproducible.

For any comparison, record the claim, the household group, the dates, the measure, whether prices are adjusted for inflation, the source and any contrary result. Ask what evidence would change the conclusion.

No information is collected or sent. This HTML and its embedded charts work offline; external source links require a connection.

Why does public purpose become optional?

The Law’s Ideological Shield connects the Constitution’s public purposes, enforceable claims, capital and the citizen’s demand for material security. Read the source essay, legal distinctions and Federalist Society discussion.

What did the laws deliver?

Read the legislative appendix and checked findings: public promises, household results, rights and duration. The supplied source is preserved; full law-by-law verification remains open.

Return to the book

Return directly to the place in the online book that opened this tool:

• The household invoice

These links return you to the same place in the current online book.