Keith Rabois defends tariffs as economic statecraft: fentanyl, manufacturing, and the Monroe Doctrine

Keith Rabois frames the tariff debate as economic statecraft rather than trade policy, arguing that market volatility is the wrong lens for evaluating a strategy designed to play out over years, not hours.

Tariffs as foreign policy lever

Rabois's core argument is that tariffs are a tool of coercion, not just economics. He points to Canada as a proof point: the Trudeau government did nothing about fentanyl trafficking through the northern border until Trump threatened tariffs, at which point interdictions improved. He extends that logic to China, where he says CCP-sanctioned production of fentanyl precursors is killing roughly 100,000 Americans annually. His position is that relaxing some China tariffs in exchange for stopping that precursor supply is a straightforward trade — American lives for stock market stability is not a close call.

On inflation, he argues the framing is definitionally wrong. Tariffs are a one-time price shock, not a sustained rise in the price level, and therefore cannot cause inflation by construction. He adds that consumers substitute goods when prices rise, and that countries are already rushing to negotiate better trade terms, meaning the net price impact could be negligible. His rough estimate, sourced from an external calculation he doesn't fully endorse, is that even the highest China tariffs add something like 70 basis points of cost across all goods over five years.

His personal preference is a flat ~10% tariff plus extremely high tariffs on China, and he says Stanley Druckenmiller's position is closer to right than wrong, while trusting Treasury Secretary Bessent to calibrate the final numbers.

“Trump lets tariff reactions play out for a few weeks, sees the trend of capitulations, and is emboldened to keep going. The most important thing is we need to achieve foreign policy objectives — economic tools are better than military tools. Inflation is the rate of growth of prices; tariffs are at most a one-time trade shock, so they can't cause inflation by definition. I don't agree that we will go into a recession.”

Macro thesis and historical analogies

Rabois places tariffs inside a larger package — debt reduction, deregulation, energy costs, and technology investment — and argues that isolating tariffs as a single variable misses the equation. He cites $1.4 trillion in regulatory costs added under Biden as a burden Trump will reverse. Real-time inflation trackers like Truflation, he says, already show inflation at around 1.28%, which he argues makes rate cuts safe now.

He draws the closest historical parallel to Reagan and Thatcher, both of whom absorbed one to two years of economic pain before delivering durable growth. His timeline for judging this strategy is the midterms, not the daily tape. He is dismissive of evaluating policy from intraday market moves, noting the market was up, down, and up again on the day of the conversation alone.

Geopolitics and the Monroe Doctrine

Rabois is explicit that the administration is pursuing a Monroe Doctrine revival — expanding U.S. influence in the Western Hemisphere. He describes the Panama Canal and Greenland as serious agenda items, not rhetorical flourishes. Taiwan, AI dominance, and fentanyl are the three fronts where he sees the China confrontation playing out concurrently.

On Chinese robotics and hardware, he goes further than tariffs. He argues DJI is a greater threat than TikTok and that Chinese humanoid robots, capable of running 17 miles per hour and carrying embedded backdoors, are an order of magnitude more dangerous. A remote-access backdoor was recently discovered in Unitree's robot dog. He wants legislative and executive action quickly.

AI, manufacturing, and venture

Rabois thinks AI will displace lawyers, accountants, and investment bankers faster than it displaces manufacturing workers. He cites Vinod Khosla's 2012 paper predicting AI-driven doctor replacement as directionally correct and nearer-term than most assume. He notes healthcare is roughly 22% of GDP and argues bringing that cost down dwarfs any tariff-driven price increase in consumer goods.

On manufacturing as a venture category, he is direct: venture is roughly 4% of startup capital and is structurally mismatched with manufacturing, which rarely scales a thousand times year over year. He wants alternative capital structures to fill that gap. For his own portfolio, he invests predominantly at Series A and C, with an eight-to-fifteen year horizon, which means current tariff noise is largely irrelevant to his investment decisions.

His advice to founders is to treat the current environment as an opportunity. His working thesis is that market flux is structurally favorable for startups because inertia, the thing that kills new companies, is already broken.

Scorecard for the strategy

Rabois concedes Derek Thompson's critique is fair — a strategy that explains every outcome is unfalsifiable. His proposed metrics are fentanyl death counts and import volumes, U.S. Treasury yields, inflation, and interest rates. He expects visible progress on the macro reshaping by the midterms. Whether the broader global order resets along the lines the administration intends, he says, won't be measurable this year.