CLR Fact Check: The Wealth Tweet and the Federal Ledger, 1989–2026
The claim: Bottom 50% wealth fell from $317.8B (1989) to $313B (2026) while the top 20 households rose from $88.6B to $4T. A post by Sen. Bernie Sanders, October 2026.
No verdict yet
Reference build for review. The open items at the foot of this page are not yet cleared, so this brief carries no verdict.
Bottom line
The tweet's direction is real, but its "bottom half went nowhere" line is a measurement choice, and the government it wants to empower grew 14x in debt over the same window.
- Holds [C]: Wealth concentrated at the very top. By the Fed's own data the bottom 50% share fell from 3.5% (1989) to 2.5% (2026).
- Does not hold as stated Cond.: "$317.8B → $313B" is the Saez–Zucman Realtime Inequality series in June 2026 dollars. It excludes Social Security and vehicles. The Fed's Distributional Financial Accounts show the bottom half rising from $718B to $4.25T nominal.
- Likely apples to oranges Cond.: the 1989 top-20 figure ($88.6B) appears to be nominal Forbes data set beside an inflation-adjusted bottom-50% figure. Source not yet located.
- The missing ledger [C]: Over 1989–FY2025, federal outlays rose 6.1x, receipts 5.2x, and gross debt 14x ($2.87T → $40.1T). Net interest now tops $1T a year.
The CLR read (systems, not scapegoats): the bottom half is not losing because the top gained. It is losing to cost structures — housing, healthcare, college — locked by the same government a wealth tax would fund. The fix runs bottom-up: ownership, local fiscal discipline, and unlocking supply.
The claim, line by line
Three of the tweet's six numbers check out or are plausible; the bottom-half pair depends entirely on which ruler you use.
| Tweet line | What the record shows | Stamp |
|---|---|---|
| Bottom 50% wealth 1989: $317.8B | Matches the order of magnitude of Realtime Inequality in June 2026 dollars. Fed DFA shows $718B nominal (Q3 1989). | Cond. |
| Bottom 50% wealth 2026: $313B | RTI-consistent. Fed DFA shows $4.25T (Q3 2025). Sanders cited the same source at $85.4B in Dec 2025, so the series swings widely. | Cond. |
| Top 20 wealth 1989: $88.6B | Source not located. Likely nominal Forbes; if so, not comparable with the inflation-adjusted bottom-50% line. | open |
| Top 20 wealth 2026: $4T | Plausible. Sanders' own bill puts Musk alone at $833B and 938 billionaires at $8.2T. | open |
| "0.00001% (20 households)" | 0.00001% of ~132M households ≈ 13. Twenty households ≈ 0.000015%. Minor. | [C] |
| Implied: the bottom half stagnated | True for share (3.5% → 2.5%, Fed). Not true for dollars under the Fed's measure. | [C] share |
Inside the Realtime Inequality Project
Realtime Inequality is a serious academic series, but its definition of wealth leaves out the single largest asset the bottom half holds: Social Security.
Who runs it [C]. Thomas Blanchet, Léonard Mouillet, Emmanuel Saez and Gabriel Zucman. Saez and Zucman are the leading academic advocates of a U.S. wealth tax and advised the 2019–2020 presidential wealth-tax plans open. Their analysis also scored the current Sanders–Khanna bill at $4.4T over ten years. That is a cui bono flag, not a disqualifier: the code and data are public on GitHub.
What it measures [C] (per its methodology page):
| Choice | RTI's setting | Effect on the bottom 50% |
|---|---|---|
| Dollars | Inflation-adjusted, June 2026 dollars | 1989 figure is restated upward; growth looks smaller |
| Assets counted | All financial and nonfinancial assets, incl. funded pensions (401k, IRA) | Neutral |
| Excluded | Vehicles; Social Security; unfunded pensions | Removes the bottom half's two biggest non-home assets |
| Debts | All debts netted, incl. student loans | Debt counted, but the education it bought is not an asset |
| Unit | Adults (equal split) or tax-code households | Not the same "household" the Fed counts |
Why Social Security matters [C]. A 2025 Journal of Finance paper (Catherine, Miller, Sarin) finds Social Security wealth grew from $7.2T (1989) to $40.6T (2019). It makes up nearly half the wealth of the bottom 90%. Counted in, top wealth shares barely moved in three decades.
The irony worth stating plainly. Social Security is a promise from the government, not property the household owns. Congress can cut it, and the trust fund is projected to run short within ten years. So the critic and the defender are both partly right: the bottom half is "richer" than RTI says, but in a form it does not control.
The government's own ledger
The tweet shows two numbers for households and none for the government; over the same window, gross federal debt grew 14x, faster than anything else on the page except the top 20.
| Line (billions, nominal) | FY1989 | Latest | Stamp |
|---|---|---|---|
| Receipts | 991.1 | ~5,200 (FY2025) | [C] |
| Outlays | 1,143.7 | ~7,000 (FY2025, receipts + deficit) | open |
| Deficit | 152.6 | ~1,800 (FY2025) | [C] |
| Net interest | 169.0 | >1,000 (FY2025) | [C] concepts differ |
| Debt held by the public | 2,190.7 | ~32,050 (Jul 2026) | [C] |
| Gross federal debt | 2,867.8 | 39,830 (Aug 5, 2026) | [C] |
The debt figure is about $295,000 per U.S. household per the JEC, more than four times the Fed's average bottom-50% net worth of roughly $64,000 per household. Interest already absorbs about one in seven federal dollars. FY2026 just closed; CBO's year-end figures will replace the FY2025 column when published.
Hostile read: what a critic of this piece would say
CLR's version survives only if it concedes what is real; these are the five attacks a hostile reader will make, and the honest answer to each.
- "Concentration is real by every measure. You're nitpicking." Conceded. The top 20 went from tens of billions to trillions. CLR does not dispute the gap; it disputes the cause and the cure.
- "Debt and wealth are different things. You're mixing ledgers." Fair. The federal ledger is shown as context, not as an offset to household wealth. Its relevance is narrower: the institution asking for new revenue has run a deficit every year since 1989 except FY1998–2001.
- "Counting Social Security is a trick to hide inequality." Partly fair. It is a revocable government promise, not owned property. CLR shows both views side by side rather than picking the flattering one.
- "Wealth taxes work; Norway and Switzerland still have them." True. But the OECD found 12 member countries had net wealth taxes in 1990 and 4 in 2017. Repeals cited efficiency and admin costs, low revenue, and missed redistribution goals. That is evidence, not proof, either way.
- "Attacking Sanders personally is the scapegoat move CLR says it avoids." Correct, and why this piece names no villain. Congress as an institution, under both parties, built the 14x. Per CLR doctrine, the surface stays structural; name-calling would hand a hostile reader the discrediting move.
Bottom-up: raising the floor instead of lowering the ceiling
The bottom half's problem is not that billionaires own too much; it is that the things the bottom half needs to buy got locked and expensive, and the ladders to ownership got pulled up.
CLR's certified five-good finding already shows the pattern: healthcare is 4.57x less affordable than in the base year, college 2.33x (sticker), housing 1.81x, cars 1.47x, while food got about 2.9x more affordable. The dose tracks the Economic Definition of Mizraim: subsidy on the demand side plus a supply lock produces the worst affordability. A wealth tax adds a third hand; it unlocks nothing.
| Lever | What it changes | CLR evidence on hand | First local move |
|---|---|---|---|
| Unlock starter-home supply | Home equity is the bottom half's main asset; more owners raise the floor directly | Housing 1.81x less affordable [C]; Deer Park and La Porte are "two-lock" (zoning + covenant) [C] | Map lot-size and use rules against the Strong Towns Housing Ready City checklist |
| Hold local debt to a ceiling | Every interest dollar is a tax with no service attached | Deer Park interest-to-revenues ~1% → ~11% (FY2020–24); La Porte ~1% [C] | Publish the Finance Decoder for each city and ISD before every bond vote |
| Make the full tax bill legible | People can't resist a cost they can't see | TEA headline per-pupil excludes debt service and construction [C] | Per-household "total local government bill" on the CLR surface |
| Measure with both rulers | Stops either side from picking the flattering dataset | This brief | CLR dashboard: Fed DFA beside RTI, with Social Security shown both ways |
| Build local ownership capital | Wealth that is owned, not promised | TAAL Directory, Remnant community | Remnant curriculum on household balance sheets and employee/community ownership models |
None of these needs Washington. Each is falsifiable in the CLR's own towns within a budget cycle or two.
Open items
Four items must clear before the Herald authors a public piece.
- Locate the source of the $88.6B (1989, top 20) figure and confirm nominal vs real.
- Pull the RTI bottom-50% series directly from the public GitHub repo (Engineer/Code lane) and match the $317.8B / $313B values.
- Replace FY2025 columns with CBO's FY2026 year-end figures when published (FY ended Sep 30, 2026).
- Verify the CPI multiple (~2.6x) against BLS CPI-U before the chart ships publicly.
- Pin one interest series (net interest, OMB/CBO) for both years on the card and in the ledger.
- Re-cite the 2026 debt figure to Treasury Debt to the Penny and the FY1989 figure to Treasury Historical Debt Outstanding.
Sources
- Realtime Inequality: methodology · project home
- FRED: Net worth held by the bottom 50% · data table
- Catherine, Miller, Sarin, Social Security and Trends in Wealth Inequality, J. Finance 2025
- CBO Monthly Budget Review: Summary for FY2025
- OMB-derived receipts and outlays (UCSB Presidency Project)
- Economic Report of the President 2004, Table B-78
- OMB Historical Tables FY2002, Table 8.1
- JEC Monthly Debt Update, Aug 5, 2026
- Sanders–Khanna Make Billionaires Pay Their Fair Share Act
- OECD, The Role and Design of Net Wealth Taxes (2018)