// Wealth Concentration · Median & Envelope

Forty years of wealth concentration

FRAME 1985 → 2025
SERIES 5 · marketable + adj.
UNITS % of total HH wealth

The bold line is the median of four marketable-wealth measures; the sand band is their min–max envelope. The dashed line is a fifth methodology (including Social Security entitlements) plotted separately.

[01] Native coverage by dataset
PSZ ← 1913 DFA SZZ ← 1966 KSS ← 1949 CMS 1985 1995 2005 2015 2025
[02] Concentration trajectory · median across the marketable-wealth measures
% of household wealth
20% 25 30 35 40% PSZ · 38.5 DFA · 28.5 CMS 24.0% · 2019 MEDIAN · 32.8% 2016 — last fully-supported year DFA 31.7% · 2025 PSZ 38.5% · 2022 1 1995 · Financialization regime 2 2008–10 · GFC pause 3 2017 · TCJA + pandemic surge 1985 1990 1995 2000 2005 2010 2015 2020 2025 Top-1% share of total household wealth
Median across DFA · PSZ · SZZ · KSS Envelope min–max of the four marketable-wealth measures CMS marketable + Social Security wealth DFA continues post-2016 PSZ continues to 2022
[03] Three structural breaks · annotated
1
1995
Financialization regime begins
The median trajectory steepens here — roughly +0.4 pp/year before 1995, ~+0.6 pp/year after — as equity-linked compensation takes hold.
2
2008–10
GFC: a pause, not a break
DFA registers a temporary dip (29.5% → 27.3%); PSZ sees no movement. Neither methodology shows a structural break. By 2013 every line is rising again.
3
2017→
TCJA & pandemic surge
DFA top-1% share climbs from 28.9% (2017 Q4) to 31.7% (2025 Q3), driven by post-2020 equity revaluation. Largest sustained run in the full series.
[04] What the envelope itself does
0 4 8 12 pp 1985 1995 2005 2015 2025 2016 post: fewer datasets 1.0 pp 10.0 pp Pp-spread between max & min, by year

The methodologies don't just disagree — their disagreement grows. The marketable-wealth spread widens from 1 pp in 1985 to 10 pp by 2016, tracking the same forces (financialization, pass-through expansion, mega-cap tech, QE) that drove the median upward in the first place. The dispute is the rise, viewed at higher resolution.

[05] What's in the median, and what isn't
Why these four datasets, with CMS as a sidecar?

The median and envelope here are computed across four marketable-wealth measures — wealth that can in principle be sold: DFA PSZ SZZ KSS. They differ in returns assumptions, source data, and pass-through treatment, but they're measuring the same conceptual quantity.

CMS is plotted separately because it includes Social Security entitlements. That's not a flaw — it's a defensible alternative definition of wealth that uniquely changes the level (the SS denominator expands faster than top-1% marketable wealth). Including CMS in the median would conflate two questions: how is marketable wealth distributed versus how is total household economic security distributed. The dashed line lets the reader see both at once without averaging them together.

Post-2016, the envelope is no longer well-defined (SZZ and KSS end), so the chart switches to plotting DFA and PSZ individually. The median line stops at 2016, the last year all four are observing.

[06] Sources & series
DFA · Federal Reserve Distributional Financial Accounts, series WFRBST01134 (Share of Net Worth Held by the Top 1%), retrieved from FRED 2026-05-19.
PSZ · Saez, Emmanuel and Gabriel Zucman, Wealth Inequality in the United States Since 1913, QJE 2016, with updates at gabriel-zucman.eu/usdina (2024 vintage).
SZZ · Smith, Matthew, Owen Zidar, and Eric Zwick, Top Wealth in America: New Estimates Under Heterogeneous Returns, QJE 2023; replication files at ericzwick.com/wealth.
KSS · Kuhn, Moritz, Moritz Schularick, and Ulrike Steins, Income and Wealth Inequality in America, 1949–2016, JPE 2020.
CMS · Catherine, Sylvain, Max Miller, and Natasha Sarin, Social Security and Trends in Wealth Inequality, Journal of Finance 80(3), 1497–1531, 2025.
Median = arithmetic median of DFA, PSZ, SZZ, KSS at each year all four observe (1985–2016). Envelope = min–max across the same four. CMS plotted as a separate dashed line because its wealth concept (marketable + Social Security entitlements) is not commensurable. Post-2016 the chart plots DFA and PSZ individually as the envelope is no longer well-defined. Series-internal revisions across publication vintages not reconciled. Top-1% shares are unitless ratios — no real-terms deflation applies. Inflection-point years are annotation conveniences, not algorithmically detected breakpoints.