Because the AI is a blunt instrument. Ask it to grade the same unchanged article several times and it reliably returns the same letter — but not the same + or −. The half-step isn’t reporting anything real about the article; it is statistical noise, so we don’t publish it. (The underlying reason is that our rubric describes five levels, A through F, and nothing in between, so there is no standard for the AI to apply.)
Human reviewers do use + and −, because a person means something by it. A combined AI-and-reader grade can land on a half-step too, since it is an average of the two.
Usable with care—serviceable on the basics and unreliable past them. A reader learns who Schiff is, what he does and roughly what he believes. A reader does not reliably learn what became of the matters the article raises, and in a biography of a living person that is the failure that matters most.
Note that this Overall sits below Neutrality, which came back at C. That is deliberate and it is the case the rubric was written for: the article is not badly slanted, but it is thin, stale, and sourced through press summary where primary records exist. An accusation reported without its documented outcome does a subject more harm than a hostile adjective, and does it without any wording a reader could object to.
The article is also twelve years behind on the thesis its subject is known for. Krugman’s 2014 verdict on inflation stands as the last word in a section that never reaches 2021–23. That is not partisanship; it is neglect. But its effect on the reader is indistinguishable from partisanship, which is why the grade is what it is.
The article renders verdicts in its own voice on one side only: “This prediction went unrealized” appended to the credit-card forecast, and in the Bitcoin section “This claim was said to be false as later explained by Erik Voorhees”—where “as later explained” treats the rebuttal as established and hands Voorhees the closing taunt with no reply.
Proportion in Responses is the main problem. Schiff’s best-known claim, calling the housing crash, is reduced to “a YouTube fan video” and a hedge that followers “may have had strong, average or poor returns.” The losses case gets a blog’s “down anywhere from 40% to 70%” plus a Wall Street Journal paragraph. Krugman appears four separate times, always last. Structurally one contested school is represented by a Nobel laureate quoted at length and the other only by the subject defending himself—the Austrian position is answered rather than stated.
Two things hold this well above the floor. The article explicitly frames the inflation dispute as definitional—Krugman “defines inflation very differently from Schiff, focusing on CPI increases rather than monetary and asset price inflation”—and concedes that asset price inflation has been significant. And the BLP allegations are followed promptly by outcomes: the defamation ruling, the payment and the withdrawal of the broadcast sit in the same short section as the fines.
The lede is four sentences of corporate affiliations and never states the thing Schiff is actually known for—the 2006–07 housing and credit forecasts, the sound-money framework, the six books, the 2010 Senate run, or the bank’s liquidation. A reader who stops there learns only that he “has criticized US banking and credit practices.”
The investigation section reports that the bank was “suspected of having facilitated money laundering and tax evasion” and that a multi-country enforcement inquiry drove its closure, but never states the documented outcome that no criminal charges were brought against Schiff. The 2024 IRS FOIA suit likewise has no disposition. In a biography of a living person, an accusation left hanging is an omission that changes meaning.
Economic and public policy views runs out of material around 2014 and omits the 2021–23 consumer-price surge, which any complete treatment of his inflation thesis must address in one direction or the other. Media career stops at a 2012 radio slot and misses the podcast and social platforms that have been his primary outlets for over a decade; the book list ends in 2012 with no discussion of content or reception. The net effect is an article whose last twelve years exist almost solely as a regulatory action of unresolved severity.
Every legal and regulatory outcome is sourced through press summary rather than the record. The Puerto Rico suspension and liquidation order, the fine, the Federal Court of Australia judgment, the 2023 settlement terms and the 2024 FOIA complaint all have filed documents that go uncited—as does any charging decision, which matters precisely because none was brought.
For a subject whose business is regulated broker-dealer and advisory work, FINRA BrokerCheck, SEC Form ADV filings and FEC reports for the 2010 Senate campaign are the authoritative records, and none appear; campaign contribution figures rest on the campaign’s own website. Contentious claims sit on the weakest sources available: the assertion that clients lost 40–70% comes from a blog relaying hearsay, and the passage impeaching Schiff’s account of his Bitcoin wallet rests on an opponent’s posts.
The critical framing of his economics is carried almost entirely by Krugman’s newspaper blog posts, cited four times as though adjudicative, with no Austrian-school or academic literature on the monetary-base-versus-CPI question. The bibliography section is degraded to the point of carrying injected spam and bare publisher-page labels—“Book details”—in place of citations.
The article repeatedly evaluates predictions without supplying the terms that make them evaluable. “This prediction went unrealized” is appended with no date at which it was judged, and the 2014–15 rate-hike passage ends in mid-air: the Fed did hike in December 2015, and the article never says so.
It then measures Schiff by his opponent’s metric. Having correctly noted that Krugman focuses on CPI rather than monetary and asset price inflation, it scores the dispute on CPI anyway—“consumer price inflation rates remained very low in the five years that followed.” The disagreement is about which measure counts, so settling it with one side’s measure is a category error rather than a finding.
Straightforward errors compound this. Euro Pacific Bank is called “a full reserve banking operation” in Wikipedia’s own voice rather than as the firm’s claim. The defamation account conflates a 2022 ruling with a 2023 settlement and mixes currencies without saying whether figures are USD or AUD. And “He has criticized US banking and credit practices” is a limp mischaracterisation of an Austrian-school critic of central banking and fiat currency—a description the subject would not recognise.
The newest event in the article is from July 2025; everything else stops far earlier. The Bitcoin section ends at January 2020, the economics section at Krugman’s 2014 remarks and a 2014–15 prediction, and the media and books sections at 2012.
Several described situations no longer obtain. The lede has him co-founding a firm that amalgamated into another in June 2024 and which he had left years before. Euro Pacific Capital is said to be “currently headquartered” in Westport, with branch offices listed, in the same breath as the admission that he sold it. The 2024 FOIA suit is left pending, with no mention of the 2025 follow-on litigation or its 2026 dismissal.
Most consequentially, the 2021–23 consumer-price surge is entirely absent, so a 2014 jibe stands as the last word on a debate that has since moved substantially. A 2009 line that Schiff “foresaw gold at over $5,000 per ounce” is presented as an outstanding forecast although gold set records near that level during 2025–26, leaving the reader with the opposite impression of the record. He is still labelled a “radio personality” on the strength of a 2012 slot.
The lede is plain enough—“stockbroker, financial commentator, and radio personality” needs no gloss—but it drops three corporate names as a bare list before giving the reader any reason to care about them.
The body then piles on terms of art unexplained at first use: “full reserve banking operation,” “monetary base,” “asset price inflation,” “a few basis point rate hike,” “moneybomb,” and most damagingly “quantitative easing,” which appears once and is thereafter reduced to “the QE program” without ever being defined. CPI is never expanded, in the very sentence where the contrast between measures is the whole point.
Elsewhere the prose is genuinely clear and concrete—the Bitcoin wallet story, the Senate campaign percentages, the bank suspension—which shows the harder passages were avoidable rather than forced by the subject. A biography of a media commentator should not require the reader to arrive already knowing what the monetary base is.