The easy parallel between the 1920s and our 2020s is economic and technological. The more important connection runs deeper and into the human conditions of fear, identity, and the habit of reaching for a scapegoat when the world changes too fast.
We are well past the midpoint of a decade that started with a pandemic and gave way to an unparalleled AI-driven technological explosion which is akin to the Roaring Twenties, which also began with a pandemic and experienced rapid technological change. But the real parallel between these periods is predominantly psychological at an individual level. Through this lens, it is about what happens to ordinary, frightened people when the world accelerates past them and no one in power seems to notice or care.
The Anxiety Beneath the Celebration
Beneath the jazz and bootleg gin (my great aunts were wine bootleggers who operated near the University of Pennsylvania in Philadelphia…) and the excitement of new household appliances such as electric refrigerators; the 1920s were also a decade of profound cultural dread. WWI had killed 15–22 million people. The Bolshevik Revolution ignited a Red Scare that cast suspicion on immigrants and labor organizers. This also meant that tens of thousands were surveilled and hundreds deported. The 1920 census delivered another jolt when for the first time, more Americans lived in cities than in the countryside. For millions of rural Americans, the cities, alive with Southern and Eastern European immigrants, jazz, and newly enfranchised women felt less like progress than invasion.
The political response was swift. The Immigration Act of 1924, for example, was explicitly designed, in Congress’s own words, to “preserve the ideal of U.S. homogeneity,” reaching back to the 1890 census that preceded what had been an immigration wave of 20 million people into the United States between 1880 and 1920 to set quotas. This message was not subtle as the America that existed before all this demographic change was the realAmerican “melting pot”. Against this backdrop, the KKK, reimagined as a defender of Protestant nativist values against Catholics, Jews, and immigrants, surged to four to five million members. Rebranding is, after all, an American pastime. And in 1925, the Scopes Trial was nominally about science education but was really a referendum on whose America would survive; the modern urban one or the traditional rural one.
Hemingway’s “Lost Generation” was not lost because of the war alone. They were lost because the cultural scaffolding that had given previous generations their bearings had buckled, and nothing coherent had risen to replace it.
The Same Fear, a Century Later
None of this should sound unfamiliar. We are living through our own version, but in updated attire.
The Red Scare’s fear that foreign ideologies would corrupt America from within has been replaced by the geopolitical threat of China and Iran. Real enough on its own terms, it has also been weaponized to justify an unchecked AI buildout with profound implications for labor, civil liberties, and the concentration of economic power — with minimal democratic scrutiny. The argument: if we slow down to ask questions, they will win. It worked in 1919. It is working now.
The census data are not partisan. The Brookings Institution confirms the U.S. will become a majority-minority nation by approximately 2045. This is precisely the kind of data point that historically triggers political backlash — the Immigration Act of 1924 passed within four years of the census showing urban populations eclipsing rural ones. Today’s immigration enforcement apparatus lands in a political environment shaped by the same underlying demographic anxiety. The clothes are different. Masks are worn. The fear is recognizable.
The Difference That Matters Most
The Great Depression was catastrophic — but it produced something extraordinary: a determination to build institutions that would prevent it from happening again. The FDIC (1933), Social Security (1935), Medicare and Medicaid (1960s), and the ACA were not acts of generosity. They were acts of hard-won institutional memory — an acknowledgment that unprotected prosperity is a trap.
Ironically, the U.S. is now actively dismantling these same institutions in the middle of a capital market boom, materially extending the economic tail risks that accompany the country’s next inevitable recession.
The One Big Beautiful Bill Act — signed on the Fourth of July 2025, in a gesture whose symbolism was presumably not accidental — represents the largest single cut to basic needs programs in American history. Medicaid work requirements begin in late 2026. Six-month re-enrollment requirements are designed to generate bureaucratic attrition among eligible people who struggle with paperwork. Automatic ACA re-enrollment ends in 2028. Medicare faces nearly half a trillion dollars in automatic cuts over the next decade. The FDIC itself has had its budget cut 16 percent and staffing reduced 20 percent, while the White House has explored folding its regulatory functions into Treasury.
Meanwhile, AI is quietly reshaping the entry-level job market. Goldman Sachs estimates AI will affect tasks equivalent to 300 million full-time jobs globally. McKinsey calculates current technology could automate roughly 57 percent of U.S. work hours. Young workers aged 22–25 in AI-exposed occupations have already seen roughly a 16 percent employment decline even as overall numbers hold. The AI boom is generating extraordinary wealth — but for those who own the infrastructure. For the worker whose entry-level job is absorbed into a system prompt, that wealth belongs to someone else.
The convergence is not hard to see. By 2027–2028, the short-term fiscal stimulus of current legislation will have faded just as structural safety net cuts begin to bite. A generation entering the workforce will find entry-level jobs thinned by AI and social supports thinned by policy — simultaneously — without the institutional buffers built for exactly this kind of moment.
What People Remain
The Lost Generation is often romanticized — Hemingway’s prose, Fitzgerald’s parties, the tragic glamour of people who knew the world had broken. Less romanticized is the fear animating the other half of that decade: the people who joined the Klan, cheered the Immigration Act, packed the Dayton courthouse to watch a teacher stand trial for teaching science. Those people were not necessarily monsters. They were frightened. They watched a world they recognized dissolve, and they reached — as people do — for simple explanations and familiar enemies.
This is the deepest parallel. Rapid technological change does not just create winners. It creates dislocation, and dislocation creates fear, and fear — absent institutions and leadership capable of absorbing it — gets organized. The 1920s taught us that a society can be simultaneously prosperous and poisoned; that a booming economy and rising nativist rage are not mutually exclusive. They can coexist, and they can feed each other.
We built safety nets after the Depression not only to protect the poor, but to protect democracy itself — because a society in which ordinary people have nothing left to lose is a society in which demagogues thrive. Strip those nets away in an era of accelerating disruption and you are not running an economic experiment. You are running a political one. And the 1930s showed us, in Europe if not yet in America, exactly how that experiment can end.
Mark Twain is often credited with observing that history does not repeat itself, but it often rhymes. The 1920s and 2020s rhyme in ways that are easy to see: the technology, the markets, the euphoria, the approaching cliff. But they also rhyme in ways that are harder to admit — the fear beneath the excitement, the backlash beneath the progress, and the dangerous temptation to dismantle the very lessons the last century paid such a terrible price to learn.
People, it turns out, remain people — even in a world of ubiquitous AI.
Bottom Line
Much has been written about imminent market threats tied to elevated oil prices and scarcity in key industrial components like sulfur and helium. While real, these risks are currently being offset by the sheer scale of the AI buildout and the fiscal largesse of last summer’s One Big Beautiful Bill. The “paper” side of capital markets remains fine; physical constraints have yet to surface, though they are still coming.
A Middle East peace deal appears to be forming with Chinese and Pakistani involvement. While the writer expects this to ultimately be a disaster for the U.S. — with Iran likely retaining control of the Strait of Hormuz, U.S. military bases abandoned, the petrodollar threatened, and U.S. geopolitical influence in the region materially diminished — a controlled and now censored media will likely provide cover for the administration to claim victory as it pivots geopolitically toward the Caribbean and South America. Near-term, domestic equity markets appear to have room to run.
Given all of the above, I will maintain at least 20 percent cash with a healthy allocation to gold as important components of my overall asset allocation going forward.