Tariff chaos, AI distrust, and China's airspace embarrassment
Nonrival
Expert analysis from think tanks and academia, on public policy, economics, and technology.
Human experts. AI summaries.
The Supreme Court Clipped Trump's Tariff Powers, But American Businesses Are Still Paying a Heavy Price
equitable_growth
- The Supreme Court struck down Trump's broadest tariff authority in early 2026, forcing $175 billion in refunds and causing average tariff rates to fall — but the administration quickly pivoted to alternative legal authorities to keep tariffs high.
- The relief wasn't evenly distributed: retailers saw faster tariff relief than manufacturers, while metals and wood products industries remain heavily burdened because their tariffs were imposed under separate laws the Court didn't touch.
- Trade with China keeps declining as its average tariff rate stays near 23%, with the slack picked up by Southeast Asian countries like Vietnam and Thailand — some legitimately, some suspected of routing Chinese goods to dodge tariffs.
Self-Driving Cars Are Dramatically Safer Than Human Drivers — So Why Don't We Trust Them?
noema
- Waymo's driverless cars have 94% fewer serious injury crashes than human drivers on the same roads, but a minor incident during a San Francisco blackout — where cars safely stopped and waited — generated far more public alarm than the ~115 Americans who die in human-caused crashes every single day.
- The gap between being safe and feeling safe is a well-documented problem in medicine, where hospitals learned that objective outcomes and patient experience must be managed separately — and the autonomous vehicle industry hasn't learned that lesson yet.
- The fix isn't just better PR: it requires standardized, transparent reporting (current federal rules count crashes but not miles driven, making comparisons meaningless), trusted spokespeople to interpret incidents, and patience for familiarity to close the fear gap, as it eventually did with seatbelts and airbags.
Markets Actually Like Independent Regulators — and the Supreme Court Just Took That Away
promarket
- The Supreme Court's Trump v. Slaughter ruling lets presidents fire the heads of independent regulatory agencies at will, ending 91 years of removal protections that insulated agencies like the FTC from direct presidential control.
- New research by the authors found that when Delaware's requirement for politically balanced courts was briefly struck down in 2017, stocks of Delaware-incorporated companies fell measurably — and rebounded when the requirement was restored — suggesting investors genuinely value stable, predictable regulatory and judicial structures.
- The authors argue the same logic applies to federal agencies: removing job protections for commissioners eliminates a key source of policy predictability, which research consistently links to higher investment and economic activity.
In the News
A Civilian Plane Just Crashed Into Beijing's Tallest Skyscraper — and It May Be China's Mathias Rust Moment
hoover
- A 66-year-old Chinese pilot deliberately crashed a small civilian aircraft into Beijing's CITIC Tower in June 2026, exposing a stunning gap in China's supposedly impenetrable air defense system, which reserves over 80% of its airspace for military control.
- The incident raises hard questions about whether years of purges within the PLA's aerospace and air-defense leadership have hollowed out military readiness, and whether China's militarized airspace created bureaucratic silos that prevented civilian and military authorities from coordinating a response.
- The author draws a parallel to Mathias Rust's 1987 landing near Red Square — an event that humiliated the Soviet military and gave Gorbachev cover to purge resistant officers — arguing that small incidents can shatter authoritarian regimes' most politically essential myths: that they see everything and control everything.
Analysis
Young people use AI more than anyone — and they're the most likely to punish politicians for using it
lse
- A new survey from Scotland finds that 38% of voters would be less likely to support a party caught using AI in its campaign, with that figure jumping to 53% among 16-24 year-olds — the age group most likely to use AI themselves.
- Voters draw a clear distinction between back-office AI use (data analysis, drafting) and front-facing uses like writing campaign materials or generating images and videos, which they find far less acceptable.
- Transparency measures like labeling AI-generated content can soften the backlash for some voters, but roughly one in five say AI use in campaigns is never acceptable under any circumstances.
Also Worth a Look
- ChatGPT Thinks the Whole World Shares Western Moral Values — New Research Shows It Doesn't (the_conversation)
- The NFT Art Market's Collapse Is a Warning for Wall Street's Blockchain Ambitions (lse_business_review)
- The Postal Service Is Being Pushed to Decide Who Gets a Mail Ballot — and It's Not Ready For That Role (brookings)
- Climate models are predicting an El Niño so strong it would shatter every record in 150 years of data (yale_climate)
- When T. rex fossils sell for $50 million, science loses something money can't replace (the_conversation)
- Europe Is Buying Its Most Expensive Medicines as 27 Competing Countries — and Paying the Price (voxeu)
- State Abortion Bans Are Linked to Rising Suicidal Thoughts Among Teen Girls (econofact)
Don't miss what's next. Subscribe to Nonrival: