Assessing *Modern Times*

In 2010, I reread Paul Johnson’s Modern Times. I’ve loved it since I was a teenager, but professional historians typically rolled their eyes in disdain when I mentioned it. After finding one mistake, I asked:

Does anyone know of any other glaring factual errors in Johnson? Please show your work.

Besides this list of errors about Japanese history, almost no one showed any work. Which, given the scarcity of human time, is totally understandable.

Recently, however, I decided to take advantage of the abundance of AI time to belatedly make progress on my original question. Here’s me talking to ChatGPT. Big picture: I struggle to imagine a better conversation with a human being on this topic.

Bryan:

What are the most important factual errors in Paul Johnson’s Modern Times?

ChatGPT:

There are real factual errors in it, but the interesting question is whether they are numerous and serious enough to justify the book’s reputation among some professional historians for unreliability. Bryan Caplan raised essentially this question in 2010, after noticing that historians often dismissed the book as inaccurate without supplying examples.

The best answer I can give is that Johnson makes a fair number of specific mistakes and some much more consequential misleading claims, but I have not found evidence that Modern Times is riddled with elementary factual errors. The more serious problem is his habit of turning a disputed interpretation into an apparently settled fact.

Here are the most important examples I would flag.

1. “The Democrats had been out of office for fifty-three years” before Woodrow Wilson — flatly false

Johnson says this in discussing Wilson’s election in 1912. Grover Cleveland had been president from 1885–89 and 1893–97, only sixteen years before Wilson took office.

This is the cleanest indisputable factual error I know of, and it is the one Caplan independently spotted. It is probably just a careless mistake rather than evidence of a substantive misunderstanding.

2. Johnson’s account of the origins of modern antisemitism is seriously historically misleading

In his Hitler discussion, Johnson describes modern theoretical antisemitism as essentially deriving from Marxism. Contemporary scholarship makes that genealogy extremely difficult to sustain. Racial/pseudoscientific antisemitism emerged from a much wider nineteenth-century mixture of nationalism, racial theory, Christian anti-Judaism and social Darwinism; major antisemitic thinkers such as Wilhelm Marr were certainly not simply products of Marxism.

This is much more important than the Wilson error because it affects Johnson’s causal interpretation.

3. His Spanish Civil War chronology is tendentious enough to cross the line into factual distortion in places

Johnson portrays the Spanish Left as having initiated the process of political violence that culminated in civil war. The problem isn’t that left-wing violence didn’t occur—it certainly did, including the 1934 Asturias revolt—but that serious histories trace substantial political violence to both left and right well before 1936.

Treating strikes or the Left’s violence as the beginning therefore produces a misleading chronology.

4. His treatment of the 1936 Spanish election conflates vote share with legitimacy

Johnson emphasizes that the Popular Front received less than 50% of the popular vote and uses this to cast doubt on its democratic mandate. The vote totals themselves have long been difficult to reconstruct precisely, and recent historical work has raised genuine questions about irregularities.

But winning less than 50% does not itself make a government constitutionally illegitimate under the electoral system then in force.

This is characteristic of Johnson’s more consequential problem: the numerical observation can be true while the narrative inference attached to it is much stronger than the evidence warrants.

5. Some of his claims about Japan appear to be straightforwardly wrong

This is potentially the richest vein of actual errors. In the discussion prompted by Caplan, a reader pointed to a detailed review identifying numerous alleged mistakes in Johnson’s Japan chapter. Unfortunately, the old Amazon review is now difficult to retrieve reliably, so I don’t want to repeat its accusations as established facts without independently checking each one.

This is worth pursuing because it may provide a much stronger test of Johnson’s factual reliability than disagreements over Spain or the Depression.

There is also an important distinction between error and Johnson being Johnson. His portraits of Gandhi, Roosevelt, Kennedy, Keynes, the New Deal, decolonization, Vietnam, etc. are often aggressively revisionist. A hostile reviewer can easily call these “inaccurate” when what he really means is “selective, tendentious, or contrary to the historiographical consensus.” That’s not the same thing.

Conversely, Johnson sometimes does something subtler than making an error: he selects genuine facts that all point in the direction of his thesis and omits facts pointing the other way. That can produce a seriously distorted picture without containing a sentence that can easily be marked “false.”

So my provisional ranking would be:

Clear factual mistakes: Wilson/Democrats; probably a number of details in the Japan chapter.

Substantive historical claims that are demonstrably misleading: genealogy of antisemitism; origins/sequence of Spanish political violence; some claims about the democratic legitimacy of the Spanish Popular Front.

Mostly interpretations rather than errors: Hoover/New Deal, FDR, Keynes, Gandhi, colonialism, Vietnam, 1960s radicalism, etc.

And that leaves me somewhat closer to Caplan’s position than to Johnson’s harshest critics: I haven’t yet found a catalogue of major factual errors remotely commensurate with Modern Times‘ reputation for factual unreliability.

There are errors, but a surprising amount of the criticism concerns Johnson’s interpretations and selection of evidence rather than dates, events, numbers, or quotations that are simply false.

Caplan made precisely this challenge—“show your work”—and the readily accessible responses don’t decisively answer it.

If you want, I can do something considerably more rigorous: go chapter by chapter through Modern Times and compile the 20–30 strongest candidate errors, then check each against modern specialist scholarship and classify them as (A) definitely false, (B) misleading/obsolete, (C) disputed but defensible, or (D) Johnson basically right.

That would give us a much better answer to whether the book actually deserves its reputation.

Bryan:

Do it.

ChatGPT:

I went through the 1991 revised edition looking specifically for claims that can be checked rather than merely arguing with Johnson’s interpretations. My conclusion is more favorable to Johnson than his reputation might suggest, but less favorable than my first answer.

There are several unmistakable errors, including a few surprisingly bad ones. At the same time, many of the passages most likely to anger historians are not factual errors at all.

I would use four categories:

A — Definitely false
B — Materially misleading
C — Seriously disputable interpretation
D — Basically defensible

The distinction matters enormously with Johnson.

Wilson and the Democrats — A: Definitely false

Johnson says Democrats had been out of office for 53 years when Wilson arrived.

Grover Cleveland was Democratic president in 1885–89 and 1893–97. Wilson took office in 1913. This is simply wrong, presumably a slip. Caplan independently caught it.

Einstein and relativism — C/D: Disputable but substantially defensible

Johnson suggests Einsteinian relativity helped inaugurate an age of moral/intellectual relativism.

Johnson actually acknowledges that confusing relativity with relativism was a misunderstanding and that Einstein disliked it. His broader claim is about the cultural reception of relativity, not the content of physics. It’s highly speculative intellectual history, but calling it a factual error is unfair.

Marxism and modern antisemitism — C: Seriously disputable

Modern antisemitism is presented in unusually close connection with socialism/Marxism.

Johnson’s discussion is selective and rhetorically loaded, but he does not simply say “Marx invented antisemitism,” as some summaries of his argument imply. His actual chapter discusses independent völkisch, racial, Christian, nationalist and populist currents.

I would retract the stronger criticism in my previous answer: this is disputable synthesis, not a clean factual error.

Japan “takes Korea” — A: Definitely false

In discussing the Sino-Japanese War of 1894–95, Johnson says Japan “took Korea, Formosa and Port Arthur.”

This is materially wrong. Japan acquired Taiwan/Formosa and the Liaodong Peninsula, but the Treaty of Shimonoseki formally recognized the “full and complete independence and autonomy of Korea.” Korea became a Japanese protectorate in 1905 and was annexed only in 1910.

Japan takes “the Sakhalin islands” — A: Definitely false

After the Russo-Japanese War, Johnson says Japan took “the Sakhalin islands.”

Japan received southern Sakhalin, south of the 50th parallel—not Sakhalin as a whole.

Japanese immigration restricted since 1894 — A: Major error

Johnson says Japanese had been “restricted by treaty from entering the United States as long ago as 1894, the first national group to be so controlled.”

This contains two problems.

The Japanese labor-migration restriction was the Gentlemen’s Agreement of 1907–08, not 1894.

And Japanese could hardly be the first nationality singled out, since the Chinese Exclusion Act was enacted in 1882.

Japanese exclusion in 1924 — D: Basically right

Johnson says the 1924 U.S. immigration law effectively excluded Japanese immigration.

This part Johnson gets right. The 1924 act barred aliens ineligible for naturalization, terminating the Gentlemen’s Agreement and effectively excluding Japanese immigration.

Sun Yat-sen and Lenin — A/B: Chronologically misleading

Johnson says Sun Yat-sen founded the Hsing Chung Hui “at exactly the same time” Lenin was developing his vanguard-elite strategy.

Sun founded the Revive China Society/Hsing Chung Hui in 1894. Lenin’s classic formulation in What Is to Be Done? was written in 1901–02. Seven or eight years is not “exactly the same time.”

The comparison may be illuminating, but Johnson compresses the chronology to make it work.

China’s “three millennia” of imperial government — B: Materially misleading

Johnson says China’s imperial system “had lasted for three millennia” when it collapsed in 1911–12.

Chinese monarchy was much older, but the unified imperial order inaugurated by Qin Shi Huang dates to 221 BCE—about 2,130 years before 1911, not 3,000.

“Three millennia” works only if Johnson loosely folds the pre-imperial Zhou monarchy into the same system.

Port Arthur — B, but largely exonerated

Johnson says Japan’s 1894–95 victory resulted in taking Port Arthur.

Japan did take it militarily and was awarded Liaodong at Shimonoseki, but Russia, France and Germany immediately forced Japan to relinquish the peninsula.

Johnson actually says this in the following sentence, so isolated descriptions of this as an error would be unfair.

“Commercial supremacy” in Manchuria — B/C

Johnson says Japan received “commercial supremacy in Manchuria” after 1905.

Japan gained extremely important rights in southern Manchuria, but “commercial supremacy in Manchuria” overstates its immediate position. Russia remained dominant in northern Manchuria.

Japanese democracy becoming “meaningless” — C

Johnson suggests constitutional government had essentially collapsed by the late 1920s and Japanese elections had become “meaningless.”

Far too categorical. Military intimidation and political assassination were real, but party cabinets, the Diet and contested elections continued to matter substantially into the early 1930s.

This is Johnson converting a genuine trend into an accomplished fact too early.

The Spanish Popular Front — B/C

Johnson uses the Popular Front’s failure to receive 50% of the vote to undermine the significance or legitimacy of its 1936 victory.

The premise is roughly correct: modern estimates put the Popular Front around 47%, narrowly ahead of the right. But the electoral law deliberately awarded large seat bonuses to successful coalitions; neither major bloc needed 50% to win legitimately.

There is a genuine modern controversy over irregularities and seat allocation, so Johnson’s skepticism cannot simply be called false.

Political violence in Republican Spain — C

Johnson’s account tends to make left-wing violence the initiating force in the Republic’s breakdown.

This is selective causation rather than a falsifiable error. The revolutionary left committed enormous violence, including the Asturias insurrection of 1934, but monarchist, Falangist and other rightist violence and military conspiracy were also integral to the spiral.

I would not classify Johnson’s narrative as “factually false” without specifying individual incidents.

Pentagon Papers and the Fourth Amendment — A: Glaring error

Johnson says Pentagon Papers participants could not be jailed in America because the press enjoyed constitutional privileges “under the Fourth Amendment.”

The press clause is of course in the First Amendment.

More importantly, the Pentagon Papers decision prevented prior restraint of the New York Times; it did not establish immunity for Ellsberg from prosecution.

The Supreme Court case concerned publication, not a constitutional right for leakers to escape prosecution. This mistake, including Johnson’s Fourth Amendment wording, was also identified in the comments on Caplan’s original challenge.

Ellsberg couldn’t be prosecuted — A/B

Johnson implies Ellsberg could not be prosecuted because of American press freedom.

False as stated. Ellsberg was prosecuted. His case collapsed because of government misconduct, including the burglary of his psychiatrist’s office—not because the Supreme Court had declared his disclosure constitutionally protected.

The 1979 Egypt-Israel treaty causes the Lebanese Civil War — A: The worst error I found

Johnson says the Egypt–Israel peace treaty of 26 March 1979 “led directly to the Lebanese civil war.”

Chronologically impossible.

Serious civil fighting in Lebanon began in April 1975, nearly four years before the treaty Johnson says caused it.

The PLO “started” the Lebanese Civil War — B/C

Palestinian armed organizations were an extremely important cause and participant, but “started by the PLO” is an excessive monocausal formulation of a civil war involving Lebanese Christian militias, Palestinian organizations, Lebanese Muslim and leftist forces, Syria and later Israel.

It becomes especially misleading when attached to Johnson’s impossible 1979 chronology.

Syria’s “Awali sect” — A, but trivial

Johnson refers to Syria’s rulers as belonging to the “Awali sect.”

Presumably a typo for Alawi. It matters little substantively, but it is another example of loose copy editing in the added material.

Egypt’s separate peace increased Syria’s importance — D

Johnson argues that the 1979 Egyptian-Israeli treaty altered Arab power politics and increased Syria’s relative importance.

Perfectly defensible. Johnson’s mistake is making it a cause of a war already four years old, not noticing that Egypt’s separate peace transformed the subsequent regional environment.

What surprised me most

The Japan/China chapter is clearly the weakest early chapter for elementary factual precision. In just a relatively short stretch Johnson gets Korea’s acquisition wrong, Sakhalin wrong, Japanese immigration chronology wrong, the “first nationality restricted” claim wrong, and forces the Sun–Lenin chronology.

These aren’t disagreements between conservative and progressive historians. They are checkable mistakes.

The old Amazon reviewer mentioned in Caplan’s thread was therefore apparently onto something when singling out the Japan chapter. Caplan’s commenters specifically remembered a list of errors there.

The Pentagon Papers passage is worse than I initially appreciated. Calling the First Amendment the Fourth might be dismissed as a typo, but Johnson then makes the substantive constitutional mistake that goes with it. New York Times Co. v. United States concerned the government’s attempt to prevent newspapers from publishing the documents. That is quite different from saying Ellsberg himself was constitutionally immune from prosecution.

And the Lebanon error is probably the most embarrassing single mistake I’ve found. There isn’t even room for historiographical interpretation: an event in March 1979 cannot have caused a civil war underway since 1975.

On the other hand, my audit also cleared Johnson of several charges I was initially inclined to make.

His Einstein discussion is more sophisticated than its caricature: he explicitly knows relativity ≠ relativism.

His antisemitism section does not actually reduce antisemitism to Marxism.

His Spanish-election argument is tendentious, but the underlying vote-share fact is broadly correct and there really is scholarly disagreement over the extent of electoral irregularities.

So how inaccurate is Modern Times?

If I had to quantify my judgment after this exercise, I’d say something like 70–80% of the things likely to provoke a professional historian are Johnson’s interpretations, selection and rhetoric, while perhaps 20–30% are actual factual or chronological problems.

That’s a judgment from the audit, not a measured statistical error rate.

Among the actual errors, most are small: dates slightly compressed, geographic possessions overstated, terminology mangled. But there is a meaningful tail of serious mistakes.

I’d rank the five worst as:

1. 1979 Egypt–Israel treaty → 1975 Lebanese Civil War.
2. Japanese immigration restricted since 1894 / Japanese supposedly first nationality restricted.
3. Japan “taking Korea” in 1894–95.
4. Pentagon Papers protections attributed to the Fourth Amendment plus the mistaken implication that Ellsberg could not be prosecuted.
5. Japan acquiring Sakhalin rather than southern Sakhalin.

Those substantially change my assessment from “Johnson’s reputation for inaccuracy seems largely undeserved” to “the reputation has a real basis, but is usually stated too crudely.”

The key distinction is that Modern Times is not an unreliable book in the sense that you should assume every striking Johnson fact is wrong. In fact, a great many of the bizarre-looking details I spot-checked turned out to be real.

His problem is different: he writes with exactly the same supreme confidence when he has the facts nailed down, when he is advancing a controversial causal interpretation, and when he has simply gotten something wrong. That makes the errors more dangerous than their raw frequency suggests.

So I would still recommend Modern Times, but as an extraordinarily stimulating argumentative history that should not be used as a sole factual reference.

Caplan was right to demand that Johnson’s critics “show their work.” Once you actually do that, however, you can produce a nontrivial bill of particulars.

If I were assigning the book a factual-reliability grade on this audit, I’d give it roughly a B−: much better than “notoriously inaccurate hack history,” noticeably worse than you would want from a standard scholarly survey.

Bryan:

Shouldn’t he get massive credit relative to mainstream historians for accurately describing mass murder and other widespread human rights violations of Communist, nationalist, anticolonial, and post colonial movements? Many leading mainstream historians like Fairbanks apologized for the kind of thing with little damage to their reputations.

ChatGPT:

Yes. That should materially raise Johnson’s relative standing, and my previous B− grade was too focused on ordinary factual accuracy while underweighting what we might call moral-historical accuracy: correctly identifying enormous crimes that influential contemporaries minimized, rationalized, or contextualized away.

There are two separate questions:

1. How often did Johnson get discrete facts wrong? As we found, more often than one would like.
2. How well did he identify the central realities of twentieth-century regimes and movements? Here Johnson’s record looks considerably better—especially if the comparison class is what respectable Western historians were saying at the time, rather than what historians say after archives opened.

Fairbank is an excellent example

I assume you mean John King Fairbank. His case makes your point particularly well.

Fairbank was not some fringe Maoist. He was arguably the preeminent American historian of modern China and a foundational figure in American China studies.

Yet his contemporary treatment of Communist China repeatedly gave the regime the benefit of the doubt in ways that aged badly.

Jonathan Mirsky’s 1991 reassessment of Western scholarship on Mao’s China is quite revealing on this point. Mirsky quotes Fairbank’s favorable assessment of the revolution and describes the extent to which critics of the prevailing picture encountered hostility within China studies.

Fairbank was hardly alone. Mirsky describes a broader intellectual environment in which journalists, academics and visitors frequently accepted extraordinarily favorable pictures of Communist China. He also notes that reports of the Great Leap famine existed contemporaneously, even though some influential observers discounted or denied them.

That is important because subsequent scholarship moved much closer to Johnson’s view of the empirical facts about Maoist rule.

By Fairbank’s later years, his own assessment had darkened considerably. Mirsky quotes Fairbank describing Communist China’s transition to party dictatorship as bringing modernized indoctrination, surveillance and intimidation into Chinese villages and families.

So there was a remarkable historiographical change.

Where Johnson deserves particularly substantial credit

It isn’t merely Communism.

One of the most distinctive things about Modern Times is Johnson’s refusal to grant an automatic moral presumption to movements because they fall into categories that were fashionable among Western intellectuals: anticolonial, revolutionary, socialist, Third World nationalist, racial-liberation, etc.

Consequently, he is unusually attentive to things like Communist terror, forced collectivization and famine; the brutality of postcolonial dictatorships; ethnic expulsions and massacres committed by newly independent states; corruption and kleptocracy among nationalist leaders; political violence by anticolonial movements; and the enormous human cost of ideological social engineering.

That approach sometimes leads Johnson in the other direction—he can be excessively charitable toward colonial governments or right-wing regimes.

But his basic methodological instinct is valuable:

Judge governments and movements by what they actually did to human beings rather than by the moral attractiveness of their stated objectives.

That sounds banal today. It wasn’t consistently applied in twentieth-century historiography.

But there’s an important qualification about Fairbank

I wouldn’t characterize Fairbank simply as an apologist throughout his career. His views changed as the evidence changed.

There is also a legitimate epistemic defense available to Fairbank and others. China was extraordinarily closed.

Fairbank himself defended Edgar Snow against the charge that Snow should have uncovered the Great Leap famine. Their exchange over Snow and Mao illustrates both the difficulty of observing Mao’s China and the continuing dispute over how much contemporary observers could reasonably have known.

So we should distinguish:

“Fairbank knew Mao had killed tens of millions of people and concealed it” — false.

from

“Fairbank adopted interpretive assumptions unusually favorable to the Communist revolution, discounted hostile evidence too readily, and consequently failed for a long time to appreciate the scale and nature of the catastrophe” — much more defensible.

And that distinction actually strengthens the interesting case for Johnson.

Johnson’s advantage was partly methodological

Johnson was willing to take seriously information that respectable specialists often discounted precisely because it came from refugees, defectors, émigrés, missionaries, anti-Communists, colonial officials, opposition movements, etc.

Those sources obviously can exaggerate. A careful historian shouldn’t believe every atrocity story.

But there’s an asymmetric-error problem here.

Imagine in 1962 you have:

• Chinese official sources implying things are broadly fine;
• Communist sympathizers reporting enormous social progress;
• refugees describing starvation, coercion and mass death;
• fragmentary demographic evidence suggesting catastrophe.

A historian worried principally about repeating Cold War propaganda may systematically discount the last two.

Johnson’s instinct was almost the reverse: authoritarian governments have extremely strong incentives to conceal their own crimes, so hostile testimony deserves serious consideration even when precise verification is impossible.

In several extremely important cases, history vindicated that instinct.

Mirsky’s retrospective is especially striking here. He notes both the existence of contemporary reports about China’s famine and the tendency of some academics and prominent visitors to discount evidence that contradicted the favorable picture of Maoist modernization. His account is worth reading in full.

And you’re right about reputational asymmetry

This is probably the strongest form of your argument.

Suppose Historian A writes in 1975:

Mao’s China is a murderous tyranny responsible for millions upon millions of deaths.

He gets the broad picture right but says 25 million when the best eventual estimate is 30 million, gets three provincial details wrong, and mistakenly dates an event to 1959 rather than 1960.

Historian B writes a beautifully sourced monograph emphasizing popular participation, modernization, emancipation, peasant agency and Western misunderstanding of China while fundamentally underestimating the regime’s coercion and enormous mortality.

There is something strange about retrospectively regarding A as the unreliable historian because his book contains more checkable errors.

The magnitude of errors matters.

Getting “southern Sakhalin” versus “Sakhalin” wrong is a factual error.

Getting the character of Maoism wrong is also an error—and vastly more historically consequential.

That doesn’t excuse Johnson’s mistakes. But an evaluation that counts the former and treats the latter merely as an innocent “interpretive difference” can badly distort comparative accuracy.

I would therefore revise my assessment

Dates, names, constitutional details and geography — B−

Johnson has a real tendency toward small factual mistakes and overconfident compression.

Careful qualification of uncertain claims — C+/B−

He frequently states debatable propositions with more confidence than the evidence warrants.

Avoiding ideological selectivity — A−

One of his greatest strengths. Johnson is unusually willing to apply ugly empirical facts against movements with morally attractive stated goals.

Recognizing Communist mass repression — A

This is one of the areas in which his large-scale judgment aged especially well.

Recognizing brutality of anticolonial/postcolonial regimes — A−

Again, a major strength relative to a good deal of contemporary writing.

Separating noble rhetoric from actual outcomes — A

Arguably the organizing strength of Modern Times.

Evenhandedness toward right-wing/Western/colonial abuses — B−/C+

Johnson’s sympathies sometimes produce a different standard of contextualization.

Overall big-picture twentieth-century judgment — A−/B+

So I’d now say calling Modern Times “unreliable” without qualification is itself misleading.

Johnson is surprisingly careless about numerous small and medium facts while being remarkably prescient about several of the largest facts that mattered. And some historians who were considerably more meticulous at the micro level made much larger macro-level errors about revolutionary regimes.

That’s actually a much more interesting verdict on Modern Times than either “Johnson was a right-wing hack” or “Johnson was vindicated about everything.”

His greatest strength and greatest weakness stem from the same trait: extraordinary willingness to make strong judgments. It produced some howlers when he was wrong, but it also meant that he was willing to call a mass-murderous regime mass-murderous at times when greater scholarly caution could shade into credulity.

Bryan:

He was also harsh on fascism and Nazism, while correctly noting that Franco and Pinochet were orders of magnitude less awful.

ChatGPT:

Yes. That is an important correction to any characterization of Johnson as simply applying a harsher standard to the revolutionary left. His taxonomy of authoritarian regimes is one of the stronger aspects of Modern Times.

Johnson is unequivocally hostile to Hitler, Nazism, Mussolini’s Fascism, racial ideology, aggressive nationalism, and the Axis dictatorships. What distinguishes him from a lot of twentieth-century political writing is that he doesn’t infer from this that every right-wing dictatorship belongs in essentially the same moral category as Nazi Germany.

That distinction has held up quite well empirically.

Consider rough magnitudes:

• Nazi Germany: approximately 6 million Jews murdered in the Holocaust, plus millions of Soviet POWs, Poles, Roma, disabled people and other civilians deliberately murdered; on top of that, Hitler initiated a European war whose death toll ran into tens of millions.
• Stalin’s USSR: millions dead through executions, Gulag mortality, deportations and collectivization/famine, with the exact attribution and totals heavily dependent on definitions.
• Mao’s China: tens of millions of excess deaths during the Great Leap Forward alone, plus large-scale killing and persecution during other campaigns.
• Francoist Spain: extremely bloody, particularly during and immediately after the Civil War. Estimates for executions and extrajudicial killings by Franco’s side are generally in the 100,000-plus range, with tens of thousands more executed after the war depending on definitions and periodization.
• Pinochet’s Chile: Chile’s official truth-commission processes ultimately recognized roughly 3,000 people killed or disappeared and tens of thousands imprisoned and/or tortured.

So saying that Pinochet was vastly less murderous than Hitler, Stalin or Mao is not apologetics for Pinochet. It’s an elementary observation about scale.

Even Franco—whose regime was responsible for vastly more killing than Pinochet’s—doesn’t approach Hitler/Stalin/Mao levels.

And Johnson deserves some credit for being willing to maintain both propositions simultaneously:

Fascist/Nazi totalitarianism was monstrous.

and

Therefore it does not follow that every anti-Communist authoritarian government was equivalently monstrous.

That sounds obvious when expressed that way, but historiographical and political categories have often obscured it.

There is another dimension on which the distinction matters: what the regimes were trying to accomplish.

Pinochet’s dictatorship committed murder, disappearance, torture and political repression. But it did not undertake anything analogous to the Holocaust, Stalinist collectivization, the Great Terror, or Mao’s repeated attempts at revolutionary transformation of society.

Franco’s early regime was substantially more ambitious and brutal, but it too evolved into a comparatively conventional authoritarian state rather than an endlessly mobilizing totalitarian one.

This relates to the classic distinction between authoritarianism and totalitarianism. It isn’t merely a question of body counts. Nazi Germany and Stalinist Russia sought enormous ideological transformations of society and attempted to penetrate institutions and private life to degrees that many traditional dictatorships did not.

Where this strengthens the case for Johnson

It makes the “Johnson was just a Cold Warrior counting Communist atrocities” interpretation considerably less convincing.

His underlying framework is closer to:

How much coercive power does an ideological state acquire, what does it attempt to do with that power, and what happens to actual people as a result?

That leads him to condemn Hitler extremely strongly, condemn Stalin and Mao extremely strongly, condemn many postcolonial dictators, and still distinguish Franco or Pinochet from all of them.

You can disagree with individual judgments while thinking that this is the correct kind of comparison.

In fact, I’d amend one line from my previous report card. I gave Johnson B−/C+ for “evenhandedness toward right-wing/Western/colonial abuses.” That’s probably too low. Something like B/B+ is fairer.

Johnson has strong sympathies, but he certainly isn’t unwilling to recognize atrocities merely because their perpetrators are anti-Communist or right-wing.

The more defensible criticism is subtler: he sometimes gives conservative authoritarian regimes more benefit of contextualization than revolutionary regimes. That’s a genuine source of bias. But it’s very different from saying he whitewashes fascism.

And the distinction you’re pointing to is important for evaluating his historical accuracy.

If Johnson says Hitler was extraordinarily evil, Pinochet was evil but vastly less murderous, and Mao’s revolutionary regime produced human suffering on an altogether greater scale than Pinochet’s dictatorship, subsequent quantitative historical research broadly vindicates all three propositions.

A historian shouldn’t lose points for refusing to flatten morally objectionable regimes into a single category.