Bohemian Swarm · Episode 3 · Interactive companion

The Slow BoilHow your paycheck lost its power

Tom and Jess don't do vibes. They do receipts. Drop your number in the pot. Watch 88 years of official data rewrite what that paycheck actually buys — and which signatures turned up the heat.

Your $ in 1998 money
$43,870
Missing if pay tracked output
$54,243
Productivity since 1948
+318.8%
Typical pay since 1948
+154.9%

Start here · Your number

This is not a vibe. It is arithmetic on your life.

The book opens on a four-job family making $90,000 and still using SNAP. Use their number, or yours. The charts below will talk back in your dollars.

You entered around 2007, into the China Shock and then the crash. Banks recovered in years. Households waited fifteen.

1998 buying power

$43,870

$90,000 today buys what $43,870 bought in 1998.

If pay had tracked productivity

$144,243

$54,243 a year is sitting on the other side of the 1973 split — 60% more.

Years of your income to buy a new home

4.6×

1950: 2.2 years of a typical income. Grandpa was not richer. The house was cheaper.

Paste it under the video. That is how the algorithm meets the data.

The frog in the boiling pot never leaps. The heat rises so gradually that its nervous system registers only a pleasant warming bath.

The Slow Boil of the American Frog, Introduction

Interactive · Purchasing power (CPI)

The dollar time-machine

Inflation is the silent tax on a paycheck. Pick any amount and any two years since 1913. This uses official CPI-U annual averages — what a dollar actually buys.

Convert money across time

$inbuys the same as

$90,000 in 2026$43,870 in 1998

Flip side: to match 1998's $90,000 lifestyle, you would need $184,638 in 2026. 2026 CPI is a partial-year estimate.

CPI-U annual averages, BLS 1913–2025 (via Minneapolis Fed); 2026 uses the Fed's partial-year estimate. Purchasing-power conversion, not investment return.

Interactive · Slice by decade

A year of work, then and now

Drag through the decades. For each stop: typical income, the median price of a new home, and the only number that matters — how many years of income the house costs.

1950: a home = 2.2 years of income

Income $3,319 · Home $7,354

Policy era: Post-war promise: FLSA floor + GI Bill + ~1-in-3 union jobs

1950: median family income (Census); median home value (Census). Income series switches from family (1950–80) to household (1990→). Existing-home basis puts 2026 near 5.1× (Harvard JCHS).

Hours of work to buy it

  • A year of public college tuition

    1964: ~97 hours · 2026: ~369 hours3.8× more work

  • A gallon of gas, on minimum wage

    1950: ~22 min · 2026: ~34 min

  • The median new home

    1950: ~2.2 years of family income · 2026: ~4.9 years (new homes). Harvard JCHS puts all homes at ~5.1×.

Interactive · 123 years of budgets

Where the paycheck went

BLS has surveyed American household budgets since 1901. Tap an era. Food’s share collapsed. Housing climbed. A new giant appeared: insurance, pensions, and everything else required to stay afloat.

Food 29.7%
Housing 27.2%
Clothes 11.5%
Transit 13.4%
  • Food29.7% 12.8 vs 1901
  • Housing27.2% 3.9 vs 1901
  • Clothing11.5% 2.5 vs 1901
  • Transportation13.4%
  • Healthcare5.2% 0.0 vs 1901
  • Entertainment4.4% 2.8 vs 1901
  • Other (insurance, pensions, everything else)8.6% 4.8 vs 1901

Share of average household spending, 1950. BLS Report 991, Table 14 — the car enters the budget.

Interactive · The spine of the story

The Great Decoupling — and the policy choices pegged to it

For 25 years after WWII, productivity and worker pay rose in lockstep. Around 1973 the lines split — and never rejoined. Scrub the year. Tap a marker. That divergence is the spine of the book, and it was not an accident.

Productivity vs typical worker pay, indexed 1948 = 100 (EPI)

Productivity Typical pay If pay had tracked output

1973 · Productivity 196 · Pay 191 · gap 5 pts

1973 · Money & taxes

The Great Decoupling

Productivity keeps climbing; typical pay flatlines. The lines never rejoin. 'That divergence is the spine of this book, and it was not an accident.'

EPI, “The Productivity–Pay Gap” through Q1 2026: productivity +318.8% vs pay +154.9% since 1948. Lines interpolated between published anchors (1948, 1973, 1979, 2026). Pay = production & nonsupervisory workers (~80% of the workforce).

76°

The water temperature

1973: 76°F

A pleasant warming bath. The frog relaxes.

A metaphor gauge, not a statistic. 1948 = bath. 2026 = check the frog.

Interactive · Nominal vs real

The minimum wage: same number, shrinking money

The federal minimum wage has been $7.25 since 2009 — the longest freeze since 1938. Toggle the number on the paycheck against what it actually buys. The peak was not recent. It was 1968.

1968's $1.60 ≈ $14.80 today. Today's $7.25 has lost ~33% of its buying power just since 2009. Had the floor tracked productivity since 1968 it would be roughly $26/hour.

Union membership

~1 in 3 workers in the mid-1950s → 1 in 10 today. Post-PATCO union decline costs workers ~$180B/yr in today's dollars (EPI 2023).

Part 1 · The jobs number nobody wanted

July 2026: the month the water got visible

Part 1 opens on one number: payrolls fell 23,000 in July when forecasters expected +83,000 — and revisions quietly erased another 103,000 jobs from May and June. Labor-force participation hit 61.4%, its lowest since 2021.

Monthly job growth — first reported vs revised

May

+129k first print

+63k actual

June

+57k first print

+20k actual

July

+83k expected

-23k actual

BLS Employment Situation, July 2026 (Aug 7, 2026); consensus via CNBC/Dow Jones. 12-month average job growth after revisions: ~34,000/month.

Watch Part 1Part 2 drops this week

The escape plan · from the book

If policy turned the heat up, policy can turn it down

The book’s final chapters lay out a three-tier blueprint — every item already proven somewhere in the world or already used by a past U.S. administration. The blueprint is not missing. The political will is.

  • Raise wage floors: 8 states and 52 cities are already above $15 (San Francisco: $18.67). California’s $20 fast-food minimum raised earnings ~11% with no widespread job loss (UC Berkeley IRLE).
  • “Just cause” firing laws: NYC’s fast-food law cut terminations 22% and arbitrary discipline 31% in year one.
  • Paid family and medical leave (California model — costs under 1% of payroll) and fair-chance hiring.
  • Zoning reform: Minneapolis ended single-family-only zoning in 2018 — construction rose, rent growth moderated.

The water is still heating. But we know the temperature now. And we know how to turn down the flame.

The Slow Boil of the American Frog, Conclusion

Sources & data notes

  1. BLS Employment Situation, July 2026 — −23k payrolls; −103k revisions; 4.1% unemployment; 61.4% participation
  2. CPI-U annual averages 1913–2026 (BLS, via Minneapolis Fed) — powers the dollar time-machine and all real-dollar lines
  3. FRED MSPUS — median sales price, new houses (Census/HUD), 1963–2026 · 1950/1960: Census Historical Housing
  4. FRED MEHOINUSA646N — median household income · Harvard JCHS price-to-income ~5×
  5. DOL — federal minimum wage history 1938–2009
  6. BLS Report 991 — 100 Years of U.S. Consumer Spending · BLS Consumer Expenditures 2024
  7. EPI — The Productivity–Pay Gap (productivity +318.8% vs pay +154.9%, 1948–Q1 2026)
  8. Union density: BLS Union Members 2025 (10.0%) · 1983: 20.1% · 1945/1954 peaks: CRS RL32553
  9. What work bought: NCES (1963-64 tuition $243) · College Board 2025 ($11,950) · FRED AHETPI · AAA gas Aug 2026
  10. AI & robots: Morgan Stanley $5T/2050 · Tesla 1M robots/yr claim · Stanford Digital Economy Lab on young developers
  11. Policy history, quotes, counterfactuals and the blueprint: The Slow Boil of the American Frog, Jessica M. McDonald (2026), and sources cited therein.

Honesty notes: the decoupling chart interpolates between EPI's published anchor years; the water temperature is a metaphor gauge; spending-survey methods differ across eras; 2026 CPI is a partial-year estimate; MSPUS covers new homes (all-homes basis runs ~5.1×); the 1950 car price lacks an official source, so it is left out of the math. Forecasts (Morgan Stanley, Tesla) are projections and claims, not facts.