Every analytical framework makes choices that shape its conclusions. This page
states those choices plainly, so the scores in this treatise can be used with an accurate sense
of what they are — and are not.
What the framework is
Each term is scored against six metrics (real median household income, income share of the
bottom 50%, wealth share of the top 10%, homeownership by age cohort, labor share of national
income, and intergenerational mobility indicators), each rated −2 to +2. The six scores
sum to a raw total. An external adjustment of −2 to +2 is then applied by hand for terms
containing a documented structural event too large for six broad metrics to capture on their
own. Raw total plus adjustment produces a net score on a −14 to +14 scale, placed on a
seven-band classification from “Great For Wealthy” to “Great For Common
Man.”
Separately, each term's Policy Impact document identifies roughly 15 significant policies,
screens each for whether it meaningfully affected inequality, and gives a sourced narrative
explanation for every policy that passes that screen.
What the framework is not
It is not a peer-reviewed economic model. No formal statistical inference,
regression, or counterfactual estimation was performed. Every score is a qualitative judgment,
informed by cited sources, about the direction and rough magnitude of a term's effect on
inequality.
It does not establish causation. A president does not fully control the
economy. Federal Reserve monetary policy, congressional negotiation (see the
Congressional Control page), global trade and
technology trends (see International Context),
and pure business-cycle timing all shape outcomes attributed here to a presidential term. Where
a term's outcome is substantially driven by forces outside the administration's control (the
2008 crisis's origins, the 2020 pandemic), the documents say so, but the term is still scored
as a unit because that is how the treatise is organized — not because the administration
caused everything within it.
The external adjustment is the most subjective element. It exists because
six numeric metrics cannot fully capture events like the 2008 financial crisis or a
once-in-a-generation tax law. But precisely because it is applied by judgment rather than
formula, it is the step most open to reasonable disagreement. Every external adjustment in
this treatise is explained in prose in its Inequality Trends document; a skeptical reader
should read that explanation before accepting the number.
Sourcing quality is not uniform across terms. Recent terms (2001 onward)
draw on a deep base of contemporaneous CBO, Census, and academic analysis. Earlier terms,
especially Carter's, rely more heavily on retrospective historical assessment, simply because
less real-time distributional data existed in 1977 than exists today. This is disclosed
wherever it materially affects a document (see, for example, the Racial
Wealth Gap page's note on pre-1989 data).
The 15-policy list per term is a selection, not an exhaustive inventory.
Thousands of regulatory and legislative actions occur in any four-year term. The list for each
term reflects a judgment about which actions were most significant, and a different analyst
compiling an independent list might include different items.
The final term is incomplete. The second Trump term (2025–2026) is
scored on roughly 18 months of an ongoing four-year term. Every document for that term is
explicitly marked interim and should be revised once the term concludes in January 2029.
Why keep a scoring system at all, given these limits
The alternative to a structured, disclosed framework is not more objectivity — it's
undisclosed judgment. Any narrative account of “how a presidency affected inequality”
involves the same kinds of choices this treatise makes explicitly: which metrics matter, which
policies mattered most, how to weigh a policy's stated intent against its documented outcome
(see the recurring examples of this across the treatise: Opportunity Zones, the 2016 overtime
rule, the 2022 student loan forgiveness plan). Making the scoring mechanism visible, and
grounding every individual judgment in a cited source, is what lets a reader agree or disagree
with a specific claim rather than simply trusting or distrusting the whole account.
How to use this treatise responsibly
Treat the net score and classification band as an organizing summary of the prose analysis,
not as a replacement for reading it.
Where a term nets close to zero (Clinton 1, Biden) or where a term contains two large,
opposing actions (Trump 1's TCJA and CARES Act; Biden's Child Tax Credit expansion and its
expiration), read the underlying explanation rather than the single number — the number
alone can understate how consequential a term actually was in both directions.
Cross-check specific factual claims against the cited sources in each document, or against
the master bibliography, rather than treating this
treatise as a primary source in its own right.