How US wealth concentrated at the very top between 1978 and 2026: the laws, court rulings and market mechanics involved, in causal order, with each link graded by how strong the evidence for it is.
Federal Reserve Distributional Financial Accounts (DFA), 1989:Q3–2026:Q2. The DFA series begins in 1989; the 1978 starting point used in this brief comes from the academic series on the next slide.
| Group | 1989 | 2026 | Change |
|---|---|---|---|
| Top 0.1% | 8.6% | 15.0% | +6.4 pts |
| Rest of top 1% | 14.2% | 17.5% | +3.3 |
| Next 9% (90–99) | 38.0% | 36.4% | −1.6 |
| Middle 40% (50–90) | 35.7% | 28.8% | −6.9 pts |
| Bottom 50% | 3.5% | 2.3% | −1.2 |
Fed DFA. In real dollars every group grew: top 0.1% $4.7T → $27.9T (6.0×), middle 40% $19.4T → $53.4T (2.7×), bottom 50% $1.9T → $4.3T (2.2×), CPI-U to June 2026. Household count rose ~40% over the period.
The two leading academic reconstructions disagree on the size of the rise and agree on its timing: the top 0.1% wealth share bottomed out around 1978 and has risen roughly 2.6× since.
Measured Magnitude contested
Tap a domino to jump to it. Slide 14 weighs the leading alternative explanation (technology and "superstar" firms).
Capital gains, dividends, corporate profits and estates came to be taxed at lower rates than wages. Lower tax drag lets capital compound faster, and capital is concentrated at the top.
| Year · Act | Change | Signer · Senate vote | Evidence |
|---|---|---|---|
| 1978 Revenue Act | Top capital-gains rate cut to 28% | Carter (D) · 86-4 | No study isolates its wealth effect. Measured |
| 1981 ERTA | Top income-tax rate 70% → 50% | Reagan (R) · 67-8 | Income lever; affects wealth through saving capacity. Measured |
| 1986 Tax Reform Act | Top rate 50% → 28%; capital gains raised 20% → 28% | Reagan (R) · 74-23 | Part of the post-1986 jump in reported top income is income shifting onto personal returns (Auten & Splinter). Contested |
| 1993 OBRA | Top rate raised to 39.6% | Clinton (D) · 51-50, no Republican votes | Runs against the trend. Measured |
| 1997 Taxpayer Relief Act | Capital gains 28% → 20% | Clinton (D) · 92-8 | Top 0.1% share 9.6% (1993) → 11.2% (2000). Correlation |
| 2003 JGTRRA | Dividends and capital gains both 15% | G.W. Bush (R) · 51-50 | Dividend cut changed corporate investment and employee pay by zero; payouts rose (Yagan, AER 2015). Causal study |
| 2013 ATRA | Made most 2001–03 cuts permanent | Obama (D) · 89-8 | Measured |
| 2017 TCJA | Corporate rate 35% → 21%; 20% pass-through deduction | Trump (R) · 51-48 | 81% of corporate-cut gains to the top 10%; none to workers below their firm's 90th pay percentile; investment +~11% at firms with bigger cuts (Kennedy, Dobridge, Landefeld & Mortenson 2026). Causal study |
| 2025 OBBBA | TCJA made permanent; ~$930B Medicaid and ~$285B SNAP reductions | Trump (R) · 51-50 | CBO: bottom tenth −$1,600/yr (−3.9% of income); top tenth +$13,600/yr. Measured |
Today's top federal income-tax rates: 37% on wages; 20% on long-term capital gains and qualified dividends (23.8% with the net investment income tax). Roll calls from congress.gov / govtrack. Estate-tax changes are under Domino 8.
Unions raise the wage share and compress pay gaps. When union coverage shrinks, a larger share of firm income stays with owners.
Unions measurably compressed income inequality across the 20th century, and their decline accounts for a meaningful part of the rise (Farber, Herbst, Kuziemko & Naidu, QJE 2021). Causal study
Counter: density was already falling before 1981, and PATCO's weight as a signal to private employers rests on historians' judgement, not a measured effect. Contested
If mergers are judged only on whether they raise consumer prices, firms can grow larger and more dominant. Higher market power raises profit margins, which accrue to shareholders.
1982 Merger Guidelines, Reagan Justice Department (Assistant AG William Baxter), adopting the consumer-welfare standard associated with Robert Bork. Agency guidance, no vote.
Average markups of US public firms: 21% above cost (1980) → 61% (2016) (De Loecker, Eeckhout & Unger, QJE 2020). Measured
Whether the rise came from lax enforcement or from more productive firms winning is unresolved. Autor, Dorn, Katz, Patterson & Van Reenen (QJE 2020) attribute rising concentration mainly to technology-driven "superstar" firms. Contested
Three rule changes widened the channels through which corporate income reaches shareholders and stock-paid executives, and raised pay in finance.
| Year · Rule | What it did | Finding | Counter-finding |
|---|---|---|---|
| 1982 SEC Rule 10b-18 | Safe harbor for companies repurchasing their own stock (SEC Chair John Shad; no vote) | Buybacks absorb most of corporate profits and inflate stock-based executive pay (Lazonick). | Net of new share issuance, payouts are ~41–50% of income, not 90%+ (Fried & Wang). Contested |
| 1993 IRC §162(m) | $1M cap on deductible executive pay, except "performance-based" pay | Equity rose from 37% to 55% of top-five executive pay, 1993–2003. Measured | Little effect on total pay level (Rose & Wolfram). Contested |
| 1999 Gramm-Leach-Bliley | Repealed the Glass-Steagall separation of commercial and investment banking. Senate 90-8; Clinton signed | 30–50% of the finance-sector pay premium is rents linked to deregulation (Philippon & Reshef, QJE 2012). Causal study | Earlier branch-banking deregulation reduced inequality by raising low-end wages (Beck, Levine & Levkov 2010). |
When a larger slice of national income goes to profits instead of pay, the value of the companies that earn those profits rises. Dominos 1–4 converge here.
| Source of the rise in US equity value, 1989–2017 | Share |
|---|---|
| Reallocation of income from labor to shareholders | 43% |
| Real economic growth | 25% |
| Lower risk premium demanded by investors | 24% |
| Lower interest rates | 8% |
Greenwald, Lettau & Ludvigson, "How the Wealth Was Won: Factor Shares as Market Fundamentals," Journal of Political Economy (NBER w25769). Peer-reviewed decomposition
It is the largest single measured component of stock-market wealth growth across the period, larger than economic growth itself. It is also where the policy dominos (1–4) and the market dominos (6–7) meet: rules that shift bargaining power and payout channels show up as a lower wage share, and a lower wage share shows up as higher equity value.
The study measures the size of the shift, not its cause; it does not apportion the 43% among unions, market power, technology or globalization.
Each group's return depends on what it owns. The top holds equity; the middle holds a house. When equity outperforms housing, wealth shares diverge without any further policy change.
| Group (2026:Q2) | Share of all directly held stocks & funds | Stocks + private business, % of own assets | Real estate, % of own assets |
|---|---|---|---|
| Top 0.1% | 25.0% (14.6% in 1989) | 74% | 7% |
| Rest of top 1% | 25.9% | 63% | 14% |
| Next 9% | 37.2% | 41% | 21% |
| Middle 40% | 11.4% | 16% | 37% |
| Bottom 50% | 0.6% | 5% | 47% |
Fed DFA levels. Excludes stock held inside 401(k)/pension accounts, which the DFA reports separately.
Near-zero rates and large-scale asset purchases raise the prices of stocks, bonds and homes. Whether that widens or narrows wealth gaps depends on which asset rises most.
Real wealth, June-2026 dollars, Fed DFA deflated with CPI-U. FOMC decisions are made by appointed officials; no congressional vote.
Two provisions decide how much of a fortune survives a generation: how unrealized gains are treated at death, and how much of an estate is exempt from estate tax.
At death, an asset's cost basis resets to its current value, so gains accumulated over the owner's lifetime are never taxed as income. Retained by every Congress.
Average unrealized gains: ~$4.67M per top-1% household vs ~$27K for the median household (Becker Friedman Institute). Measured
Repeal is scored at only ~$105B over ten years (JCT), because holders would change behavior. Borrowing against appreciated assets instead of selling them ("buy, borrow, die") depends on this provision.
| Year | Exemption | Act |
|---|---|---|
| 2001 | $675K, phasing up | EGTRRA (Bush) |
| 2013 | $5M, indexed | ATRA (Obama) |
| 2018 | ~$11.2M | TCJA (Trump) |
| 2026 | $15M ($30M per couple), indexed | OBBBA (Trump) |
No study cited here measures the effect of these provisions on the top wealth share directly; the link is mechanical (less tax at transfer = more capital retained) rather than estimated.
Court rulings removed limits on independent political spending. Concentrated wealth can then fund the campaigns that write the next round of rules.
Gilens & Page (2014), 1,779 policy questions: economic elites and business groups had substantial independent influence on outcomes; average citizens' preferences had near-zero independent effect.
Counter: the preferences of the affluent and the middle overlap on most issues, so their effects cannot be cleanly separated (Bashir 2015); policy tracks rich and poor about equally in other data (Branham, Soroka & Wlezien 2017). The Gilens–Page data end before 2002, so they predate Citizens United.
In the wealth data: the top 0.1% gained +3.1 share points 1989–2008 and +1.8 points 2010–2019; no step change appears after 2010. Measured Policy effect contested
A second school argues that policy is secondary: digital technology and globalization let the most productive firms capture whole markets, and their founders and shareholders captured the gains.
| Act | Signer | Signer net worth (est.) | Senate | Vote pattern |
|---|---|---|---|---|
| 1978 Revenue Act | Carter (D) | — | 86-4 | Bipartisan |
| 1981 ERTA | Reagan (R) | ~$4M (1981) | 67-8 | Bipartisan |
| 1986 Tax Reform Act | Reagan (R) | 74-23 | Bipartisan | |
| 1993 OBRA | Clinton (D) | ~$1.3M | 51-50 | Democrats only |
| 1997 Taxpayer Relief Act | Clinton (D) | 92-8 | Bipartisan | |
| 1999 Gramm-Leach-Bliley | Clinton (D) | 90-8 | Bipartisan | |
| 2003 JGTRRA | G.W. Bush (R) | ~$20M | 51-50 | Party-line |
| 2013 ATRA | Obama (D) | 89-8 | Bipartisan | |
| 2017 TCJA | Trump (R) | ~$3.5B (Forbes) | 51-48 | Party-line |
| 2025 OBBBA | Trump (R) | ~$5.1B (Forbes) | 51-50 | Party-line |
Of the seven capital-favorable acts through 2013 (excluding the 1993 rate increase), six passed the Senate with large bipartisan majorities. 2003, 2017 and 2025 passed on party lines. Net worths are contemporaneous press estimates; blanks are left blank rather than estimated.
| Domino | Strongest evidence | Grade |
|---|---|---|
| Outcome: top 0.1% share up ~2.6× since 1978 | Fed DFA; Saez–Zucman; Smith–Zidar–Zwick agree on direction | Measured |
| 1 · Tax on owning below tax on earning | Yagan 2015; Kennedy et al. 2026; CBO on 2025 act | Causal study |
| 2 · Bargaining power declines | Farber et al. QJE 2021; BLS density | Causal study |
| 3 · Antitrust narrows | Markups 21% → 61%; cause disputed | Contested |
| 4 · Payout channels widen | Philippon–Reshef (finance); buybacks disputed | Contested |
| 5 · Labor share → stock value | Greenwald–Lettau–Ludvigson: 43% | Peer-reviewed |
| 6 · Stocks outrun houses | Kuhn–Schularick–Steins JPE 2020; Fed DFA | Causal study |
| 7 · Monetary policy | Net effect small, sign unclear | Correlation |
| 8 · Untaxed inheritance | BFI unrealized gains; statutory exemptions | Measured (effect mechanical) |
| 9 · Political spending | Spending measured; policy effect disputed | Contested |
··· The best-supported links are the middle of the chain: income moving from wages to profits, and profits compounding in the hands of those who own stock.
··· Stale Chips · because innovation doesn't just happen