Persuasion

Persuasion

By Yascha Mounk
Download on the App Store

Persuasion episodes

  • A New State Capitalism
    A worker producing semiconductor products for export to Europe and the United States at a semiconductor manufacturer in Binzhou, China, on April 1, 2024. (Photo by Costfoto/NurPhoto via Getty Images.)

    This article is brought to you by American Purpose, the magazine and community founded by Francis Fukuyama in 2020, which is proudly part of the Persuasion family.

    Not so long ago, the distinction between America’s market-based economic model and its main alternatives—or rivals—looked reasonably clear. Before 1989, there was the sclerotic shortage economy generated by Soviet-style economic planning. More recently, there was Europe’s “cuddly” capitalism, and China’s state-run version of it—all standing in sharp contrast to America’s “cutthroat” variety.

    Reasonable people could dispute the relative merits of European and American versions of capitalism, but most would agree that the Chinese model, lacking democratic accountability, suffered from serious flaws—notwithstanding China’s impressive growth record. Those included picking economic winners and losers, the exercise of state and party control over “strategic” firms, the fuzzy line between the national interest and the balance sheets of leading corporations, and a lack of inhibition about deploying nominally private assets in pursuit of government policy.

    Those flaws were real, which is why it is so disconcerting to see the rise—largely under the radar—of an American version of state capitalism. Yes, there is the gaudy, scandalous self-dealing of White House envoys such as Steve Witkoff and Jared Kushner, obscuring the difference between private interests and official government policy. There is Trump’s exercise of arbitrary power in setting tariff rates, or extorting the private sector as a precondition for clearing regulatory and antitrust obstacles set by the federal government.

    America’s drift toward state capitalism, however, is deeper and more lasting than just Trump’s idiosyncrasies and corruption. The return of industrial policy into “strategic” sectors was a hallmark of the Biden presidency. And while the amount of norm breaking during Trump’s second term is an aberration by any standard, it would be a mistake to dismiss all of the precedents it has set as flukes to be rolled back by a future Democratic administration.

    Consider Trump’s oftentimes confusing policies on semiconductors, mangling long-term strategic consideration in pursuit of short-term revenue objectives. In April 2025, the administration banned the export of Nvidia’s China-bound H20 processor on national-security grounds. By August, it had reversed itself—on the condition that Nvidia and AMD hand Washington 15 percent of their China revenues in exchange for the licenses.

    American Purpose at Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal democratic values wherever they are threatened, please consider becoming a paying subscriber today!

    By December, the more powerful H200 processor—roughly six times the H20—was cleared for “approved” Chinese customers, this time with the government taking a 25 percent cut. The same export controls Washington spent a decade building were undone—for an arbitrary fee. A bipartisan bloc in Congress, alarmed that the H200 would hand Chinese AI labs a meaningful lift, scrambled to block the sale—in the end, the deal was blocked by Beijing after finding a domestic alternative.

    The effective export tax is ineffective in preventing the chips from falling into the hands of our adversaries—and it may well end up being unconstitutional. Unless courts rule against it forcefully, future presidents will enjoy a new tool of discretionary power over the private sector.

    The Intel case brings America even closer to explicitly state capitalist territory. In August 2025, the government took a 9.9 percent stake in the chipmaker—433.3 million shares at $20.47, an $8.9 billion investment—by converting unpaid CHIPS Act grants and Defense Department “Secure Enclave” money into equity. Commerce Secretary Howard Lutnick’s logic was simple: the government should get an equity stake for its money.

    For now, the holding is passive, with no board seat. The deal has also been wildly profitable on paper: after Intel’s shares surged, the stake was worth roughly $36 billion by the spring—one of the most lucrative investments in the history of the federal government. Washington also pocketed a five-year warrant for another 5 percent of the company, exercisable if Intel ever ceases to control a majority of its foundry—a hedge designed to keep the chipmaker’s manufacturing arm American. Setting aside the cast of characters populating this administration, one could easily imagine a Democratic president deploying similar tools in pursuit of more or less well-defined national security goals.

    The Department of Defense has also become the largest shareholder in MP Materials, operator of America’s only rare-earth mine, with a roughly 15 percent position wrapped in price floors and offtake guarantees. The Department of Energy took stakes in the lithium developer Lithium Americas. The Pentagon also took 10 percent of Trilogy Metals, tied to an Alaskan mining project the Biden administration had blocked. A wave of pre-revenue rare-earth startups followed. By April 2026, the running total had reached $20.9 billion across fifteen deals—the largest federal push into strategic industry since the Second World War, according to the Council on Foreign Relations—with the administration reportedly eyeing dozens more.

    Again, there is (or could be) a national security logic to these moves. China’s threat to choke off rare earths is real, and the war in Iran exposed how fragile the supply chains have become—for everything from magnets to interceptor drones. A government that lets its adversaries monopolize the inputs of modern defense is failing at its job. And there is an old American tradition—from the transcontinental railroad to Operation Warp Speed—of the state midwifing industries the market alone would underprovide.

    Yet what risks emerging, especially in the light of America’s rule-of-law crisis, is a system in which access to the American market, export licenses, and capital itself, will run through political favor—especially beyond a certain scale.

    Conversely, one must worry about the outsized power of the enormous private sector institutions that undergird the ongoing technological transformation—and with it the future of our democracy. Nvidia is the largest company in the world, valued at more than $5 trillion, followed by Alphabet (Google), Apple, and Microsoft.

    SpaceX, freshly public at a valuation north of $2 trillion, runs the satellite network on which the Pentagon increasingly depends. Can the U.S. government, and American voters, tolerate a situation in which our basic technological infrastructure and some critical functions of the federal government depend on, and generate profit for, a narrow segment of gargantuan private actors?

    There is no clean exit from the problem of concentrated economic power in frontier industries that are laying the groundwork for the economy of tomorrow. For free market conservatives, this may seem like no problem at all—at least in theory. In a reality of flawed political institutions, however, concentrated economic power is bound to translate into exactly the kind of corruption and kleptocracy that we see flourishing under the second Trump administration. It should also give one pause that SpaceX’s float on Nasdaq was preceded by an agreement to modify the exchange’s underwriting and index rules in SpaceX’s favor—a highly unusual step, which speaks to how far removed we might be from a world of unadulterated free markets.

    The political left, meanwhile, will want to treat compute as a utility, as the railroads and the power grid were before it. Under both extremes, the substantive outcome risks being the same—once-innovative industries transformed into lazy, rent-generating incumbents, whether under private or government ownership.

    Managing the ongoing transition successfully will require transparent, durable rules, as well as impartial anti-trust enforcement. What America does not need are nationalizations, heavy-handed statism, or a series of opaque, deal-by-deal shakedowns negotiated by whoever holds the pen. But laying down rules of the game for the economy of tomorrow, as neutral and economically efficient as possible, is no mean feat under America’s broken politics.

    At the same time, continuing on the default trajectory leads nowhere good. A U.S. government that owns a tenth of a critically important company it regulates is not a neutral umpire. A White House that is able to grant or withhold 25 percent export waivers wields a lever over Silicon Valley that no federal statute ever conferred. The temptation to use such leverage for ends that have nothing to do with national security—a flattering headline, a campaign contribution, a rival punished—is precisely the temptation the rule of law exists to constrain.

    Forget Trump and Trumpism: America’s biggest challenge in the next generation is to avoid the entrenchment of a system of state capitalism with U.S. characteristics rather than Chinese.

    Dalibor Rohac is a senior fellow at the American Enterprise Institute, an advisor at GLOBSEC, and a contributing editor at American Purpose.

    Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.

    And, to receive pieces like this in your inbox and support our work, subscribe below:

    Subscribe now

    0 min
  • The Supreme Court Isn’t Trump’s Lackey
    The U.S. Supreme Court on June 29. (Rui/Xinhua via Getty Images.)

    President Trump has a lot of obsessions. If you had to name the ones he cares about the most, the list would probably include imposing tariffs on imports, getting the Federal Reserve to lower interest rates (even if it means ousting board members who disagree), abolishing mail-in voting, ending birthright citizenship for children of illegal immigrants, and—last but not least—fighting personal lawsuits brought against him since his first administration.

    And yet on every one of these fixations, the Supreme Court in its just-completed term dealt Trump a defeat. Having already ruled in February against his attempt to levy tariffs by invoking a statute on international economic emergency powers, the Court has now struck down his effort to end birthright citizenship by executive order; blocked his attempt to fire Lisa Cook as a Fed governor; permitted the counting of mailed-in ballots received after Election Day; and denied his appeal of a lower court ruling upholding the $5 million in damages for sexual abuse awarded to E. Jean Carroll in 2023.

    Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal values wherever they are threatened, please become a paying subscriber today!

    True, the Court also overturned a New Deal-era precedent to allow Trump to reshape the Federal Trade Commission and other independent regulatory agencies by firing officials at will. This was a major victory for Trump that enhanced his power and control over the executive branch. But it was also a victory for a conservative legal theory—the “unitary executive”—that had been in development long before his presidency. The presidential power it undoubtedly magnified will be available to a Democratic president the next time there is one.

    All of the above refutes the notion that the Court’s conservative Republican majority is in the tank for Trump. As such, it represents a reassuring development for the rule of law and American constitutional democracy. The Court is not a rubber stamp for Trump. The majorities that ruled against him, in various key cases, included one or more of his own appointees.

    Progressive critics need to acknowledge this reality even as they understandably lament the many times the Court has reached results Democrats abhor, whether in ending race-conscious districting; allowing Trump to abrogate Temporary Protected Status for Haitian nationals and others; or permitting states to enact laws barring transgender athletes from girls’ sports. There is a difference between facilitating a long-standing, albeit controversial, conservative agenda—as each of those rulings did—and going out of your way to enable the president’s legally far-fetched or authoritarian pet projects. A Court that consistently did the latter would merit the charges of corruption progressive Democrats so often hurl at it; this one seems appropriately concerned with preserving a measure of independence and legitimacy.

    This hasn’t stopped commentators and politicians from proposing radical solutions to the Court’s supposed authoritarian drift in recent weeks. A mindset of “desperate times call for desperate measures” has taken hold in the Democratic camp, with Maine Senate candidate Graham Platner among those calling for the Court to be stacked with new justices. (He has also called for at least two of the sitting justices to be impeached.)

    There are many problems with such schemes, not least of which is that they are contingent upon all sorts of hypothetical events—starting with a Democrat-controlled House and Senate undoing the filibuster. Court packing would set a dangerous precedent that both parties would likely make use of in years to come, as well as inject ugly partisanship into the nation’s top court.

    Subscribe now

    Thankfully, the situation is nowhere near as dire as that. And the Court seems aware of the need to protect its political capital—Chief Justice John Roberts in particular. In every case in which the Court has resisted Trump, Roberts has been in the majority, joined by the Court’s three Democratic-appointed liberals, plus a fifth or sixth justice drawn from among Trump appointees Neil Gorsuch, Brett Kavanaugh, or Amy Coney Barrett.

    A Supreme Court justice can be political in three ways: inevitably, improperly, and institutionally. It’s inevitable that a justice might be influenced, consciously or not, by his or her priors about policy and ideology; presidents select them for that, in fact. In many cases, law and precedent are so open to interpretation that a certain exercise of individual judgment can’t be avoided.

    At the same time, it’s improper for a judge consciously to calculate the personal or partisan pros and cons of a particular decision, and to rule based on that.

    What’s occasionally necessary, and indeed admirable, is to take into account the potential impact of a case on the overall constitutional system and the judiciary’s interests within it. This kind of politically-minded judging can be thought of as judicial statesmanship.

    In the Court’s just-completed term, Roberts was political in this latter, institutional, sense. He chose his battles with Trump carefully, selecting those he was likeliest to win not only on the law, but among an ideologically diverse coalition on the Court—and in the court of public opinion. Justices Clarence Thomas and Samuel Alito, by contrast, repeatedly dissented in support of Trump’s positions, even where they were least defensible and most destabilizing, as in the tariff case.

    Roberts and the other right-leaning justices who occasionally voted with him advanced long-standing conservative objectives without sacrificing the Court’s independence. If the system didn’t crack under the pressure of this year’s high-stakes cases, there’s reason to hope it never will.

    Charles Lane, a contributing writer at Persuasion, is a nonresident senior fellow at the American Enterprise Institute and a columnist for The Free Press.

    Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.

    And, to receive pieces like this in your inbox and support our work, subscribe below:

    Subscribe now

    0 min
  • Reading Is Dead. And We Killed It.
    (Photo by Andreas Arnold/picture alliance via Getty Images.)

    Earlier this year, publishing house Hachette pulped the upcoming horror novel Shy Girl by Mia Ballard, following allegations that Ballard relied on artificial intelligence to write the book. Meanwhile, half of novelists in the UK fear they will be completely replaced by AI. As artificial intelligence continues to replace creative activities previously considered uniquely human, there’s a fear that fiction will grow ever more distant from the human experience, with predictable plots and simplistic dialogue and characters. Literature will start to function as synthetic junk food for the brain.

    Unfortunately, we didn’t need AI to do this. Literature has been functionally artificial for a number of years now, since long before ChatGPT came on the scene. It wasn’t computers that did this, it was us—publishers, agents, writers, and readers. Over the last 20 years, modern novels have become shorter, with simpler language and shorter sentences. At the same time, people are reading less. In 2023, the number of Americans aged 15 and older who read for more than 20 minutes a day was 15 percent, compared to 22 percent in 2003. In some ways, shorter novels can be a blessing (I know it’s supposed to be Charlotte Brontë’s best novel, but I spent most of Villette wishing the main character would just get on with it), but it’s worrying that authors are now writing for diminished attention spans.

    Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal values wherever they are threatened, please become a paying subscriber today!

    A major culprit dumbing down literature is the rise of influencer culture—including influencers cosplaying as readers. BookTok is one of the largest communities on TikTok. Posters share which books they’re reading (or, more accurately, buying), and review their favorites, sometimes coordinating their books with a particular aesthetic. In 2025, there were over 35 million videos with the hashtag #BookTok, with a total of 370 billion views. In the United States alone, around $59 million of print sales could be linked to BookTok in 2024. Users tend to be young and spend an average of 1.5 hours a day on TikTok.

    This content is necessarily short-form, which means it has to be easily searchable. Featured books are broken down by trope, with popular themes including things like “enemies to lovers,” “romantasy,” and “grumpy x sunshine” (the latter being a trope describing a relationship between a miserable person and a happy one). Such labels, which also function as hashtags, make it easy to discover a book based on your exact mood or preference—without the risk of any surprises or complexity. In this way, BookTok echoes online fan fiction spaces, where readers can select their favorite characters from any popular book and settle in to read a story about those two characters having a spicy relationship—again, without much in the way of surprise or literary merit.

    By reducing books to their key tropes, BookTok actually makes it harder to find new books outside the select few favored by the algorithm. Author Stephanie Danler, when reflecting on her experience of BookTok, said: “It seemed impossible to discover different fiction. It was the same 20 books over and over.” Categorizing books in this way shows the limitations of labels—it’s hard to imagine Tolstoy reducing Anna Karenina to “grumpy x sunshine.”

    That hasn’t stopped publishers and agents from trying to anticipate the next BookTok trend and responding to market demand. Many, for example, are choosing to prioritize romantasy fiction. Publishing houses are in a bind here—this stuff sells (a lot), but when the fiction bestseller list is dominated by easy reads driven by short-form internet creators who slap digestible labels on them, something has gone wrong with the industry.

    This is not to say that all genre books are worthless. Donna Tartt’s 1992 classic The Secret History, which is certainly not shallow, is often used as an example of “dark academia,” a trend which pairs books with an aesthetic of gothic clothes and posing moodily amongst old buildings. But even then, the trouble with trends is that they quickly pass. Reading “dark academia” books, no matter their literary merit, may one day seem as passé as wearing skinny jeans or a baker boy hat.

    Subscribe now

    A core feature of BookTok is that it’s female. The largest group of BookTokers are women aged 25-34, followed by women aged 18-24. This makes sense—women read more fiction than men, a trend that holds despite both sexes reading less than a few decades ago. BookTok users also tend to be predominantly Millennial and Gen Z, so it gives a useful insight into how online culture is shaping girlhood and young womanhood. Young women today are caught between the pressure to live performatively online and the pressure to prove that they’re more than just their appearance. It’s understandable that a reaction to these conflicting pressures is to perform the fact that they’re reading.

    Works by writers such as Sally Rooney and Jia Tolentino are labelled “Hot Girl Books.” One is made more attractive, the internet claims, merely by reading or even just holding works by these authors. “Reading While Hot” reduces books to a fashion accessory—the highbrow equivalent of Labubus.

    This focus on appearance is reinforced by changing styles in book covers. Those that are instantly recognizable on video—especially if they label which tropes are held within—are the most popular. Many videos glamorize choosing a book purely on how pleasing its cover is, as if a book’s primary job is to match your outfit or interior design.

    What you read drives what you think, and BookTok not only limits exposure to new fiction and difficult themes; it also makes literary criticism shorter, snappier, and shallower, as analyses that would once have been essays or long YouTube videos are compressed into a few minutes or seconds. What’s more, cancel culture remains alive and well on BookTok, perhaps because of its similarities to teen blog sites such as Tumblr. In June 2025, Ali Hazelwood, a romance author popular on BookTok, was bullied off Instagram after saying she preferred a different love interest for the protagonist of the teen series The Hunger Games. Meanwhile, blogs and videos helpfully point out why common romance tropes are “harmful.” And of course, only the very, very brave would admit on BookTok that they still read JK Rowling.

    Reading trashy books is one of life’s great joys. But presenting them as great literature just because everyone else is reading them closes readers off to more difficult but rewarding works that genuinely challenge them. On the surface, BookTok seems to encourage reading, but in the long term it reflects the broader crisis of the humanities. No longer are great works of literature a vehicle for self-development or understanding the inner lives of others. Books are now just another way to curate a performative self online, even as we read—and understand—less and less.

    Leonora Barclay is Head of Podcasts at Persuasion.

    Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.

    And, to receive pieces like this in your inbox and support our work, subscribe below:

    Subscribe now

    0 min
  • Trump’s Threat to Scientific Research
    (Photo by Scott Olson/Getty Images.)

    This article is brought to you by American Purpose, the magazine and community founded by Francis Fukuyama in 2020, which is proudly part of the Persuasion family.

    In 2025, the Trump administration scrambled scientific research by turning off the tap of funds. Among those to be shut down were clinical trials for promising treatments for cancer and dementia. In the National Institutes of Health alone, the administration terminated nearly $4 billion across 2,300 grants.

    The Trump team reluctantly began restoring funding and then, in March, a federal appeals court ruled that the administration’s action was arbitrary, capricious, and likely to produce “irreparable harm.” The administration has responded by counter-attacking with a proposed regulation, published on May 29, that would take decisions about research grants out of the hands of scientific experts and give ultimate control to political appointees.

    The scientific community is in an uproar, with funding having stopped, then started, and now threatened. One survey found that, among researchers receiving part of the $40 billion in funds distributed every year by the National Institutes of Health, more than half found their pipeline disrupted. More than four out of five younger researchers are worried that the cuts could destroy their careers.

    The administration’s new proposal threatens even more damage in the future.. In a field where success only comes through the long term, the future would suddenly be unpredictable.

    It is not enough to say that just some projects would be affected. All would be. And not just now. But forever—even as the tastes and values of new political appointees change.

    Subscribe now

    The proposal runs to 108 pages and covers every agency making grants to researchers. Wading through it all requires superhuman patience or an unlimited budget for legal billing. Still, let’s compare the current process with the key points of what the Office of Management and Budget (OMB) proposes.

    Currently, the grant process across most of government works in a similar way. The agency posts a notice that it’s going to make grants on a particular problem and gives a deadline for applications. Agency staff members screen the proposals when they come in to ensure that they fit the basic requirements of the program. The agency staff then sends proposals out for peer review, and the reviewers score each proposal. The agency staff then sorts the applications by score and generally makes grants in order of the ranking until the money runs out. Successful grantees work on the projects, make regular progress reports, and submit to financial auditing. They continue work until the end of the grant period and, with luck, science blooms.

    The new OMB proposal, on the other hand, goes down a very different road—paved by a series of reports from the right-wing Heritage Foundation, which contend that the federal government had developed a “leftwing funding apparatus,” which “only employs progressives.” The OMB initiative seeks to defund “a woke policy agenda” that runs against the Trump administration’s plans.

    The most important parts of the Trump administration’s new proposal are:

    • Peer scientists would still do reviews and rankings, but political appointees would make the final decisions. They could overrule the experts on which projects have the most merit. The decisions of Health and Human Services Secretary Robert F. Kennedy, Jr., and the rise of his Make America Healthy Again movement show how some research get canceled and new, sometimes offbeat projects take their place. Instead of building systematically, scientific research would careen from administration to administration, in a way that would inevitably harm America’s place in the research world.

    • Once a project starts, political appointees could shut it down at any point, including ‘‘‘for convenience’ whenever it determines that termination is in the government’s interest.” Any new administration could come in and shut down any—and all—existing grants by using the “for convenience” clause. Important work requires sustained effort, and no researcher can keep a lab going if the spigot is turned on and off. Moreover, much medical research requires double-blind studies over several years. Researchers worry whether they can ethically enroll patients in a study if their trial can be quickly shut off.

    • Grantees couldn’t use federal funds to attend conferences, unless the funding agency explicitly approves the trip. Scientists shake out the bugs in new projects and share their findings with others at conferences. They often come back with ideas for new projects and improvements for old ones. This would end with the new rule. Cross-fertilization of work would dry up.

    • Grantees can’t use grant funds to pay publication costs in professional journals. A little-known development in the recent past has been the charge that journals make for publishing articles. Articles are often available online (eventually) or through university libraries, but in fast-moving fields researchers need “open access,” or immediate free distribution to readers. For the Journal of the American Medical Association, the cost ranges from $4,000-$6,000. The leading journal in public policy, Journal of Policy Analysis and Management, charges $3,960. Grants now often pay these charges, but if that ends, an important avenue for quick access to important new work would shut down. Few researchers can pay these fees out of their own pockets.

    • All grant recipients will be required to take part in the E-Verify program, which employers use to track whether workers are eligible to work in the country, and ensure that everyone paid with the funds does as well. This could make it much harder to hire the advanced researchers from other countries that are essential to many labs.

    Draining the Gold from Science

    The proposed rule promises “Gold Standard Science,” without defining how it would get there. Researchers themselves worry that the rule will be anything but. The editors of the New England Journal of Medicine bluntly wrote, “We know that what counts as gold-standard science can be determined only by peer review conducted by experts in the field—the current system.”

    Moreover, the editors said, “When science becomes politicized, everyone loses.” The American College of Surgeons posted that the proposed rule “threatens to devalue surgeons’ scientific expertise,” while the American Physical Society, collaborating with more than two dozen other professional associations, said that “the proposed changes would fundamentally alter the practice of science in the United States.”

    The proposed rule is open for public comment until July 13. OMB has already collected tens of thousands of comments, and the wheel on the comment counter is rapidly spinning upward.

    Subscribe now

    Boosting the President’s Powers

    Stretching beyond the worries of scientists, the proposed rule is part of a broader Trump administration strategy emanating from OMB. It flows as well to efforts to rein in control of federal grants to state and local governments and turn more federal civil servants into at-will employees.

    The basic principles are the same: a foundational belief that the radical left had taken over the federal government and was spending public money to advance woke causes; a desire to shrink the federal footprint, both financial and administrative, across the government; and tactics to pull control of programs and money out of the hands of career experts and give it to political appointees—who can more directly steer who does what and who benefits.

    So OMB’s proposed rule on research is important in its own right. But it’s also part of the Trump plan to magnify the president’s powers. Article II, Section 3 of the Constitution says, “he shall take care that the laws be faithfully executed.” Trump takes “he” personally: he views “faithfully executed” in terms of his own policy goals, not what the law says. In fact, that clause imposes a duty on the president to implement the laws that Congress has enacted, not create his own law independently through Executive Orders.

    This draft rule ought to be opposed by every scientist and responsible citizen.

    Donald F. Kettl is author of The Right-Wing Idea Factory (2026). He is also Professor Emeritus and Former Dean of the University of Maryland School of Public Policy.

    Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.

    And, to receive pieces like this in your inbox and support our work, subscribe below:

    Subscribe now

    0 min
  • Charles Fain Lehman on Why Cities Got Safer

    Charles Fain Lehman is a fellow at the Manhattan Institute and a senior editor at City Journal, where he covers crime, policing, and urban policy.

    In this week’s conversation, Yascha Mounk and Charles Fain Lehman discuss why Baltimore failed to follow the crime declines that transformed other American cities, why strategic policing works, and how focused deterrence breaks cycles of retaliatory violence.

    Watch the conversation below—the full version is behind the paywall on this page!

    This transcript has been condensed and lightly edited for clarity.

    Yascha Mounk: There is a broad question in the United States about why there was a lot of crime in the eighties and nineties, then a significant drop in crime in many cities across the United States in the following decades, and then possibly a slight uptick since the pandemic—that is somewhat contested and depends on the kinds of crime you look at. You have been trying to think about this by looking at a particular city that I know somewhat well: Baltimore, because I teach there some of the time. Baltimore to me was always a big intellectual puzzle because it had not gone through some of those same transformations. In the 2000s and 2010s, when Washington, D.C. became a lot safer, Philadelphia became a good bit safer, and New York became much, much safer than it used to be, Baltimore kind of did not. Perhaps we can start with that—why is it that even as all of these other cities on the eastern seaboard were really making inroads in reducing murder rates and becoming quite a bit safer, Baltimore at the time did not seem to be following the same positive trends?

    Charles Fain Lehman: The first thing I want to emphasize is just the extent to which that is unusual—that Baltimore really struggled with its experience with homicide well after other cities had dealt with homicide as an issue or had substantially reduced the homicide level in their cities. If you look at young Black men in the city of Baltimore compared to other major cities, they experienced far less significant a decline in their risk of death by homicide over the period we are talking about, in the 1990s and 2000s. It is a dramatic reduction. As a result, when you get to even 2013 or 2014, their risk is substantially elevated over the rest of the nation. That is before 2015, when the death of Freddie Gray resulted in citywide protests, some rioting, and a durable crime spike. There have been a number of papers arguing persuasively that whatever little control the city had in the immediate prior period was totally lost yet again for the next several years.

    There are different factors that are conceivably going on there. There has long been a struggle over the management, funding, and staffing levels of the BPD—the Baltimore Police Department’s trust in the community and the civilian leadership’s trust of the Baltimore Police Department. That relationship has long been problematic, and I think Freddie Gray was a microcosm of broader community concerns. I once met an MPD police officer in DC who would commute an hour and a half each way to work for MPD because he did not want to work for Baltimore PD. That is sort of the struggle that they had.

    Mounk: Perhaps one way of thinking about this is through the comparative lens—this is a classic question of comparative politics as well as many other social science disciplines. Why does something happen in one set of places and not in another? Standard methodological advice holds that you can only explain a difference in outcome by a difference in independent variables, by a difference in causal factors. So perhaps let us make it comparative: what was happening at that time in New York City, in Washington, D.C., and so on that was reducing the crime rate?

    Lehman: To a first approximation, the answer is deliberate, strategic policing. New York City is the prime example—the deployment of strategic policing under Bill Bratton, where they actually, in the 1990s, for the first time asked: where is crime really happening? What are the hotspots? How can we deploy NYPD resources, and how can we do that strategically to enhance safety in any given area? Think about Bryant Park or Times Square, where they really surged resources and cleaned up. The same thing happened in other jurisdictions—in LA later on, in Washington, D.C.—sort of across the board. There were efforts to do this in Baltimore in the nineties and two thousands as well.

    Mounk: Before we get back to Baltimore, tell us a little bit more about what strategic policing looks like. There is obviously a huge debate about what led to this decline in crime in places like New York and LA. Explain the different kinds of explanations that have been given, and why you think strategic policing in particular is the thing really responsible—according to you, at least—for a significant chunk of that decline.

    Lehman: It has to be a significant chunk, insofar as you see a large increase in homicide and other major crimes in the 1960s and 1970s, peaking in the eighties, starting to decline in the nineties, with a precipitous decline through the 2000s and early 2010s. For listeners who may not know: a large portion of that is demographic and structural, by which I mean the baby boomers aging into and out of crime. If you do age adjustment, about half of that effect disappears. People commit more crimes when they are younger; the baby boomers, who were in their peak crime-committing years during that hump, then aged out of it. That accounts for about half of it.

    Mounk: To be clear, you are saying not that there is something particular about baby boomers that made them more prone to violence, but simply that there was a particularly large number of young men—a demographic that commits a lot of crimes—and so you would just expect there to be more murders when the number of people who are twenty years old is higher.

    Lehman: Exactly. If we are talking systematically about the increase and decrease in violence, that is a big part of the story. Then there are other explanations worth addressing. I think there are a number of them—things like lead and crime, where I just do not think the evidence is there. Then you have to think about specific cities where we see very large decreases in violence. New York again is the paradigmatic example; its crime decline is twice as deep and twice as long as the rest of the nation. Boston is another good example, where there was a precipitous decline, particularly in youth violence in the 1990s.

    When you look at what those departments did—and often those departments were talking to each other and communicating about their tactics, often run by the same people, very small circles—their theory was that through the seventies and eighties, policing had been highly reactive. You hear about a crime, you go, you arrest somebody, maybe, maybe you do not. They said: we have these resources and tools, let us use them proactively. Let us identify the people who are frequent flyers—the individuals we know are driving violence disproportionately—and the places where violence is happening disproportionately. Let us go there and focus on those people. If we do that, maybe we will see some substantial improvement. Indeed, many jurisdictions implemented this kind of strategy and saw the substantial improvement you would expect.

    Mounk: Why is that so successful? What about the model causes this decrease in crime? It is not immediately obvious why police officers saying, this is where crimes happen, we will hang out there is going to reduce crime that much. If a lot of murders are connected to the drug trade and to gangs, you might think: if a lot of these shootings happen in place A, but we are still in competition with a rival gang and still want to avenge some other crime, we will just carry out a murder somewhere else. There would be all kinds of reasons, if you were telling me about this as an idea before the empirical evidence existed, to think: how can that possibly work? Why does it seem to be working?

    Lehman: Your specific question is about displacement—the so-called balloon effect. Why is it that if you do something here, people do not just move over there? A lot of the answer comes down to the fact that murder is objectively very rare. Overwhelmingly, you are unlikely to be murdered, I am unlikely to be murdered. It is extremely unlikely, because murder—gun violence in particular—requires a brew of different things to go wrong in order for it to occur. You need, as I like to say, young men with guns, a culture of honor or culture of beef, and low levels of formal and informal social control. That mix does not occur naturally in most places or at most times, which is why violence is usually concentrated in pockets even within cities—and even within those places, often in a specific set of people.

    Violence is not an efficiently produced phenomenon; it is not a market phenomenon. If you do not have the right environment for it, if one person gets pulled out of the social network, it can collapse really easily. In a modern, highly surveilled, enforced society, violence is not the norm—it is the exception. If you take one of the Jenga blocks out, all of a sudden violence becomes much less likely.

    Mounk: What is it about this form of policing that is so successful at taking one of those Jenga blocks out? What does it actually look like? Cops know that a particular part of a particular neighborhood is where gang warfare is especially endemic or where there is a particularly high level of crime—they go and they hang out there. Is it to make people feel surveilled? Is it to build relationships? Is it to warn people? What does that look like concretely?

    Lehman: “Strategic policing” is really just an umbrella term for police departments thinking proactively—rather than reactively—about how to allocate their resources, identifying places that are problems, and remediating them. One version of this is hotspot patrol policing, for which we have a substantial body of evidence. Basically, you make a heat map of crime, identify where it is happening, and put a cop there. We have a lot of evidence that when that happens, crime levels go down and do not displace. That hotspot is a hotspot because it is an area uniquely conducive to crime. Police are a visible deterrent—in many senses, deterrence is the primary way policing works. If you put a cop in a place, that officer is going to deter criminal behavior there and remove the opportunity for crime to be committed.

    That is not the only approach. There is also place-based policing. We have a lot of evidence that the built environment matters a great deal for crime—that environments can be more or less criminogenic. One thing you can do is surge police resources to help clean up, beautify, and facilitate the surveillance of a neighborhood, a park, or an area that has been taken over by crime. That can durably reduce crime. Then there is a third approach—focused deterrence—which is about deterring the spread of violence as a kind of infectious pathogen within a social network.

    Mounk: Tell us about some of these in a little bit more detail. You are not a kind of airy-fairy policy analyst, not one of the progressive voices in this debate who thinks the only cause of crime is deprivation and people being too hard on criminals. So it is perhaps somewhat surprising to hear you say that beautifying a public park or changing the built environment really has this big an impact on where crime takes place. In one register, that sounds like something the German artists I grew up around might say—if only our cities were more beautiful, there would be less crime. Yet here is a hard-nosed, right-of-center—if that is how you describe yourself—policy analyst saying that going out and beautifying a park is really going to bring down crime. Why should we believe that? What is it about these architectural interventions, these interventions in public space, that actually prove effective in influencing crime?

    Lehman: The first answer is simply that we have a fair amount of high-quality evidence showing that cleaning up vacant lots, forcing landlords to clean up problem properties, and rehabilitating houses all causally reduce crime. The question then becomes: why? Some people argue it affects the psychology of criminals, which I do not think is true. What I do think is that it renders the space more governable. This goes back to Jane Jacobs and The Death and Life of Great American Cities—the idea that what deters antisocial behavior in public is first and foremost the informal surveillance that we all engage in, what Jacobs calls “eyes on the street.” The reason many spaces are not crime-ridden is that there is a high ratio of law-abiding to non-law-abiding people in that space.

    We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

    Subscribe now

    Set Up Podcast

    If you have any questions or issues setting up the full podcast feed on a third-party app, please email [email protected]

    When you clean up a park, restore community space, clear lines of sight, or clear out Penn Station, you allow—or encourage—people who had otherwise been absent from that space to recolonize it. The natural governance that comes from informal social actors is restored. It is about facilitating that informal enforcement, that informal surveillance and coercion that is made possible by having a space that feels like people are welcome in it, can govern it, and feel that it is theirs.

    Mounk: It is not about the psychology, then, but about the ability to exercise social control over the most dangerous parts of a city. How does that relate to this third idea you raised—focused policing?

    Lehman: One of the things you get out of the criminological literature, as I said earlier, is that violence is really concentrated. A lot of things have to go wrong. A corollary of that is that violence is highly concentrated in social networks—most violence in America today is driven by people who know each other, who are often in conflict with one another, who have beef with one another, who are all dating the same people, who are all from the same blocks, and so on.

    As a result, violence becomes contagious within these groups. You and I pick a fight, I shoot you, your friend comes back and shoots me, my friend comes back and shoots him—you get this cycle of escalation. I mentioned Boston earlier; part of how they drove down violence there is a strategy called focused deterrence or ceasefire. The idea is simple: you identify the people in the network who are connected to each other and commit crimes. You go to them and say, we are no longer tolerating violence. If you commit a shooting, we are going to come after you with everything we can. When a shooting happens, you follow up and say, remember what we told you—there are going to be consequences for escalating. You establish a deterrent threat. Implementation fidelity is challenging; it is hard to get this right. But when you do get it right, there are often dramatic reductions in violence, because you have established for that population: we know who you are, we are watching you, there will be consequences if you escalate, so do not escalate. If you stop the escalation—much like stopping a cancer from growing—you prevent many more murders than just the one interaction would otherwise generate.

    Mounk: The idea is that one shooting can lead to a whole retaliatory cycle, but because you intervene fast enough, it does not. If you succeed at that, perhaps for six months there are no shootings, and there is no particular reason for another shooting to start.

    In the background, I was wondering about the account of what actually drives these shootings in the first place, and how that matters. One way of thinking about it is as something very strategic: you have two different drug gangs trying to expand territory, engaging in a kind of urban warfare where the logic can be modeled with something like a rational choice framework—at a certain level, a very rational set of processes, and you might think you need one kind of intervention to contain it.

    On the other hand, you might have a model that is far less rational: people at the age when men are most violent, carrying deadly weapons around for all kinds of reasons, who just happen to run into someone who slept with a girlfriend three months ago outside a bar when both are drunk, and that is where it kicks off. Or someone has the impression that someone else is disrespecting them, undermining their honor, and feels they must defend that honor in order not to become vulnerable to further attacks—and in that moment, they engage in conflict.

    If you think a lot of the causes of violence are near-random in that way—where the structure of a situation determines the likelihood of conflict, but each particular conflict is not itself a rational choice—that might lead to a very different model of what it takes to rein in these crimes. Is it clear which of these two types of things really drives the most damaging violence? How do criminologists think about this?

    Lehman: The short answer is it is the second one. The way I would push back on your framing is that there is an inherent rationality to honor culture. It is not just impulse—there is a logical form to it.

    Mounk: I was starting to get at that, which is that if you allow yourself to be disrespected, you become a target for further disrespect down the line.

    Lehman: Exactly. Being aggressive in a way that is antisocial can garner you benefits. The guy who is picking fights on the train, playing his music too loud, getting in somebody’s face, smoking on the train—he is doing it to establish his dominance in the public space. So it is not just reactive, not just: I am going to engage in this beef because otherwise I lose face and there is risk to me. It is also: if I provoke this beef, I will gain face, I will gain relative social standing, and I will be better off.

    It is also the case—and there are a bunch of great books on this topic, Jill Leovy’s Ghettoside, and A Code of the Street by Elijah Anderson—that when you are in an environment of relatively low formal social control, where there is relatively low state legitimacy or projection of state power, honor culture is a very natural thing to emerge and to persist over time. When you are in that kind of environment, particularly if you are a young man, you are stuck in the game—you are going to fight back and forth, you are going to be obliged to engage in this sort of dispute of honor, in order to protect yourself but also in order to gain status. I think that is overwhelmingly more important as an explanation for why violence happens in America today than the story about rationally acting drug gangs. People tend to overstate the degree of violence associated with the drug trade. There is a particular political economy of the crack crisis in the 1980s, but that does not always generalize to what every illicit marketplace looks like. Today, a much smaller share of violence is explained by that versus by the honor dynamic.

    Mounk: If you are saying it is about honor culture in that way, another way of reframing it is: how do you discourage honor culture, or differently put, how do you reduce the consequences within an honor culture of either having high status or having low status? The strange thing about honor culture is that the examples I can think of where it really governs social interactions include some schoolyards. A lot of what you were talking about applies to the school bully who you just know is crazy enough that they will always escalate—so if they ask for your lunchbox, you are just going to give it to them, because the stakes if you do not are too high. It applies to parts of inner-city America, and there is a lot of interesting sociological and anthropological research on this. It applies to parts of Brittany in France. It also applies to 17th and 18th century European aristocrats. When you read about all of these people dueling—including some of the founders of the United States—that is because they were embroiled in an honor culture where allowing someone to disrespect you had downstream consequences for your life and your reputation, and so you needed to engage in extremely risky acts to protect it.

    This is not somehow characteristic of particularly poor neighborhoods mostly populated by a particular ethnic group in the United States today. Very different groups at very different ends of the socioeconomic scale, including some of the most privileged people in 17th and 18th century Europe, have been embroiled in these kinds of honor cultures. One question is: how do you change those cultures such that, as the 18th century gives way to the 19th, the response to an insult is no longer meet me tonight at midnight with your second but rather the bourgeois idea that sticks and stones may break my bones, but your insult will never hurt me—and so you simply smile and say, you are really just degrading yourself, and move on? How do these changes happen?

    Lehman: Barry Latzer, the emeritus professor and CUNY criminologist, in his book The Rise and Fall of Violent Crime in America, makes a related argument about why you see honor culture disproportionately in inner-city Black communities in America. His answer is that it is simply southern honor culture—the American South has a long tradition of honor culture, and it is being transplanted northward as people move through the Great Migration. That gets to the point: it is not sui generis to race or specific to race or poverty status. It is about where the culture comes from.

    Mounk: As a good Hobbesian, I would have thought it is about the absence of an external authority.

    Lehman: I am getting there. What happens in the South is a little more complicated than simply saying “weak state formation”—they do ultimately have weaker states, but it is more nuanced than that. I think that is true also in the other environments you are talking about: honor culture is ultimately the result of the state having insufficient power to be the ultimate arbiter, to resolve all conflicts, or to claim a monopoly on the resolution of conflicts. The reason you do not have an honor culture is because the state undermines any rival claim to settle disputes. The way you get away from honor culture is through legitimate dispute resolution through the courts. How do you get to that bourgeois attitude? Because you have recourse to the legal infrastructure, which is itself undergirded by the state’s threat of violence.

    Mounk: Imagining being someone who is growing up embroiled in honor culture and who is aware enough to recognize that it is dangerous—that it can lead to an early death through a duel or a shooting—and who is trying to get out of it: the ability to secure your basic rights, belongings, and safety through authorities is going to govern a lot of what your incentives are. Perhaps here we are in a kind of strategic logic. If the schoolyard bully comes to take your lunchbox and you feel you can tell a teacher who will reliably punish the bully—and doing so will not lead to your classmates turning on you as a snitch—then why fight? But if you think the teachers will not care, or that even if they do, your classmates will ostracize you for snitching and your social life will be miserable, then you face a different calculation. It is not just about the lunchbox today. Once you acquire the reputation of being easily bullied, perhaps you should fight—because that is the only way to protect yourself going forward. That ability to seek recourse against injustice, and to avoid becoming vulnerable by stepping out of honor culture, governs the strategic incentives. That is true for a trivial example like the lunchbox, but it also seems to be true for far more life-and-death situations.

    Lehman: The bully analogies are useful because the bully gets something out of it too—it redounds to his benefit to engage in bullying behavior. In some sense, the reason we want states and law enforcement and everything that goes with them is that the predatory conduct of the bully is not something society should have to tolerate. To aggressively abuse the metaphor: we want one very big bully whose job it is to suppress all the others. We want the teacher. Having to live around predators is extraordinarily socially toxic. If you spend your whole life worrying about being victimized, you are much worse off. We know that kids who are exposed to shootings do worse in school, that kids regularly exposed to violence have worse outcomes as adults. There is disinvestment, harm to property values—all of these objective measures that correspond to the intuition that if I have to walk through my community every day worrying that one to five percent of the young men there are going to victimize me at random, I will be much worse off.

    Mounk: Let us bring that back to the areas in the United States where there is the most crime. Is this partially about trust between residents of those neighborhoods and the police? If your best friend or your brother is shot and you more or less know who did it, one response is to go to the police and make sure that person goes to jail. Whether the norms of that neighborhood allow you to do that without yourself becoming a target for further violence and retaliation is going to strongly influence whether, faced with tragedy, you go to the police—or decide that is not a route available to you, and instead get in a car with a couple of friends and go looking for the person who did it.

    Lehman: When we are talking about the modern era of high violence—the great crime wave of the sixties, seventies, and eighties—there are a couple of different factors going on. One is that, to return to this idea of governance, many parts of America’s big cities become ungoverned or ungovernable. There is a systematic decay of both the small things and the large things. Think about New York in the 1980s—or cities that never recovered, like Detroit, or Gary, Indiana, or large parts of Washington, D.C.—that become just totally lost. There is a fixed stock of police resources, and if crime becomes sufficiently bad, the police may be inadequate to addressing the issue in the stable equilibrium. You have to do something to shift the equilibrium, because in the stable equilibrium the police are simply outmatched.

    Subscribe now

    In that period, that dynamic is exacerbated by—to put it oversimplistically—our collective belief that policing does not do anything. This is a real position held by many people of influence in American society in the sixties, seventies, and eighties: that policing has no effect on crime, that the criminal justice system cannot work until you address the root causes of crime. This goes back to the Kerner Commission and the Katzenbach Commission in the 1960s under the Johnson administration: until you address the root causes—poverty, deprivation, racism—you will not have any impact. So not only has crime gotten bad enough that the corresponding police response is relatively ineffective in equilibrium, but we have also decided that police cannot do anything, and so police gravitate toward the least useful things to do with their time—and are often quite corrupt on top of that. It is not a good equilibrium on either side of the equation.

    Mounk: There is a huge debate on this in Europe and in Britain as well. One of the things that marked out Tony Blair’s third way was a famous soundbite from the run-up to his 1997 victory: tough on crime, tough on the causes of crime. He was essentially saying we do not need to choose between those—but you can only understand the impact of that sentence because the conventional wisdom in the Labour Party for the previous twenty years had been that it is just about the causes of crime, not about reining crime in directly.

    To people who might be tempted by that position: why is it unhelpful to think that we simply need to eliminate the deeper causes of crime? Certainly, when I look at relatively affluent countries—Switzerland, Singapore, Japan—there are much lower levels of crime, perhaps because they have wealthier states and less social deprivation. What role do the root causes of crime play in bringing about the relatively high levels of crime in the United States—even in places where things have improved a great deal? New York is much safer than it was, many cities in America are much safer than they were, and yet they remain a lot less safe than Zurich, Singapore, or Tokyo.

    Lehman: Two things. One is that there are structural factors that explain that variance and that matter—though I tend to point to more direct ones. We have a lot more guns in the United States than peer countries. I am actually a squish on guns and the Second Amendment relative to others on the right, but it is a fact of American life, and it explains a lot of the variance in violence specifically—it is just much easier to harm other people in the United States than in peer countries. That does not explain all of it, but it explains some. There are structural factors that, plugged into a model, will explain some of the variance.

    That said, it is important to observe that rising affluence is not necessarily correlated with receding crime. There is a great line from James Q. Wilson, the political scientist, who observed that in the 1960s we dramatically overhauled the criminal justice system, made it much less punitive and much more rehabilitative, launched the Great Society, and made huge investments in the inner city—in welfare, in transfers—through the first half of the decade. Then crime exploded. Similarly, in the 1990s, inequality was rising, there were real problems with poverty particularly in major cities, the poverty rate went up under Giuliani in New York City—and yet crime dramatically declined.

    You can talk about the big structural variables, but they are relatively hard to move. Part of my argument is that crime is policy-accessible. We actually know today, in a way we did not in the seventies, what tools work to meaningfully reduce crime. We know we can use policy to bring crime down. Last point: crime is itself a cause, not just an effect. Crime is a cause of deprivation. If you live in a crime-ridden neighborhood, you are systematically worse off, systematically disadvantaged. Policy interested in remediating disadvantage should say: we need to free children from the risk—and the fear—of being shot and murdered. That is a disadvantage they should not have to suffer. Policy can change crime in a way that it is much harder to change those other core variables. We should understand crime as an independent cause of the so-called root causes, not just an effect of them.

    Mounk: To spell this out a little: it sounds like you are making two different kinds of responses. The first is that looking at the root causes does not give you much explanatory power. You look at times when the supposed root causes improve—the United States in 1970 had much more generous welfare programs and was much less overtly racist than twenty years previously—and yet crime is higher in 1970 than in 1950. Then you look at the period when crime really declines, and it is not obvious that those factors had gotten much better. So the root causes are just not that explanatory.

    The second point, which is slightly different but important, is that sometimes the root causes are not the things that are actionable. I have drawn on this James Q. Wilson example in talking about a very different topic—populism—in some of my work. Perhaps the root causes of populism are globalization, or whatever; I am not sure I fully buy that argument, but even if it were true, we are not going to undo globalization in order to save ourselves from populism. It is just not clear how that might work. But you might learn how to run a different kind of political campaign, or adopt certain kinds of policies—moderate political parties might learn how to speak to people in ways they find less off-putting. Those are interventions that are doable in a way that changing the entire global structure is not.

    Whatever exact role the root causes play, what we do have is a lot of empirical evidence that specific interventions—focused policing, changing the architectural environment of cities, warning people who are at particular risk of committing a serious crime that there will be consequences—can actually reduce crime.

    To bring it back to Baltimore: this is going on in New York, in LA, in Boston, in Washington, D.C., and these cities transform quite radically. There are still some corners of New York that are quite violent, but whenever friends come to visit and ask if they should be worried, I say you would really have to set out to find a dangerous place to end up there. That is very different from what New York felt like in 1990. Washington, D.C. is an even more extreme case—it used to be that the area around Dupont Circle was relatively safe, but only on particular blocks, and straying a couple of blocks in the wrong direction was genuinely dangerous.

    Lehman: There were crack markets behind the White House.

    Mounk: Right. Washington, D.C. today has also radically transformed. Yet throughout all of this, Baltimore did not follow. Is it that people in Baltimore just did not learn these lessons? What happened there?

    Lehman: It is a really great question, one to which I only have part of the answer. There were several parts. One is that, as I alluded to earlier, Baltimore had real problems making the Baltimore Police Department an effective tool for doing some of these things, just at an institutional and cultural level. BPD has had real scandals, real struggles, and a lot of adversarialism between BPD and the civilian leadership of the city. As a result, they never quite built the institutional infrastructure needed. Conversely, when they tried to do these things, they could not build the executive support they needed. We talked about focused deterrence; I have written about their most recent focused deterrence implementation, which I argue works. Everyone I talked to in the city said: we have done this before, we did it in 1999, we did it in 2014, and both times the implementation fell apart. The reason was they did not have executive support, they did not have people at the mayoral level who said, we believe violence is a problem and we are going to put our shoulders into it throughout city government. So they did not get the resources they needed, and there was a great deal of skepticism. That is the challenge of strategic policing and more strategic approaches generally: it takes deliberate action, institutional know-how, and institutional commitment. Without those, you can get stuck in a quagmire. All of that is downstream of the more deeply entrenched problems, which is simply that many neighborhoods in Baltimore have deeply entrenched gang cultures. When I say gang cultures, I mean these tight social networks of young men with guns who all know each other and are all primed to shoot at each other, because the city government has not been committed to solving the problem.

    Mounk: This is a story of the city not quite getting its act together because of structural features that do not allow it to effectively deploy the tactics that have worked elsewhere, even when it tries. If we had recorded this conversation two or three years ago, that is roughly where it would have stopped. Until quite recently, Baltimore was still the most dangerous major city on the eastern seaboard. As Washington D.C., New York, and Boston had improved very significantly, Baltimore had improved a little from the heights of crime in the nineties, but was just notably less safe than those other major cities. It turns out, though, that over the last few years, there have been the beginnings of quite a remarkable success story. What is the evidence that Baltimore is finally turning it around?

    Lehman: The most straightforward answer is that you can draw a straight line down in terms of the homicide rate. In 2022, when things started to turn, there were about 333 murders in Baltimore. Last year, three years later, there were 133 murders, a more than 60 percent decline. In count terms, that is the fewest murders Baltimore has seen since 1965, a stunning and dramatic decline. If you talk to people on the street, and there is some polling, less than I would like, and if you go and talk to people who live in Baltimore’s most violent communities, they will tell you things feel safer, that they can go out and walk around. The change is felt particularly in those places. It is not every category of crime, not that all measures of antisocial behavior have gotten better across the board. It is that the city has said we are no longer going to tolerate murder, and murder has precipitously declined, and shootings have declined in kind.

    Mounk: We talked about the structural features that made it hard for Baltimore to implement some of these things: the tension between the Baltimore Police Department and its leadership, the high levels of corruption relative to other police departments on the eastern seaboard, and so on. One thing I have heard anecdotally, though I do not know whether there is reliable evidence for it, is that a lot of Baltimore Police Department officers do not live in the city of Baltimore but in Baltimore County, and so do not feel the same sense of ownership that perhaps the average NYPD officer might, being more likely to live in the city. Presumably none of that changed overnight. So what did change? What allowed Baltimore to finally implement some of the successful tactics from other places?

    Lehman: There are two answers. One is that they got a new prosecutor who made a real difference. A prosecutor is a kind of gating function: a good prosecutor on its own will not make a difference, but a bad prosecutor can cause real problems when you are trying to fix this situation. So they addressed that, and then they built out an implementation of focused deterrence that actually has executive buy-in, that is a genuine all-of-government effort, and that, for the first time, makes a real effort to get the implementation details right rather than just attempting it and not really doing it. Everyone tends to pick one or the other of those two factors, and I say no, it is probably both, they are probably reinforcing each other, and the bigger picture is that they are simply trying to run the criminal justice machine properly. When you do that, things get better.

    Mounk: Tell me about the role of prosecutors here. How was the previous prosecutor in Baltimore failing, and how has the new prosecutor made a difference? More broadly, you have argued that the movement to install progressive prosecutors in major American cities has backfired and led to significant rises in crime. What is the evidence for that?

    Lehman: Baltimore’s state’s attorney is a man named Ivan Bates, elected in 2022 and taking office in 2023. He replaced Marilyn Mosby, who left under a cloud of scandal, having been under multiple federal investigations and ultimately convicted federally of various fraud charges.

    Subscribe now

    Mosby came in in 2015 as a self-described progressive prosecutor. The idea behind the progressive prosecutor movement is that prosecutors in the United States have prosecutorial discretion: they can decide whether or not to prosecute offenses. Progressive prosecutors argue you can use that power to unilaterally limit the reach of the criminal justice system, or simply stop prosecuting whole classes of crimes.

    What do we know about progressive prosecution? It is a little thorny, because it is very easy to call yourself a progressive prosecutor without implementing any particular policy mix. There are a couple of studies that look at the effect of electing a self-identified progressive prosecutor on crime in a given jurisdiction. Depending on which study you prefer, the findings show an increase in petty or disorder offending, or an increase in property offending. Neither study finds an increase in violence. My response to this is that many progressive prosecutors run on things that do not make a big difference. If you promise to stop prosecuting jaywalking, for example, the response is that American prosecutors do not prosecute jaywalking anyway, so you are not making a policy change of any meaningful significance.

    But then you have to look under the hood at what they are actually doing, and there are decisions that can be really impactful. Refusing to prosecute felons in possession, for instance: if you are a felon, you cannot carry a firearm, and some big-city prosecutors will refuse to prosecute that charge even though it is a way to get a criminal with a gun off the street. Or categorical non-prosecution of drug-related offenses, or not using gang enhancements. In Mosby’s case, I think she was fairly aggressive, particularly after 2020, stopping prosecution of a whole variety of so-called low-level offenses. That, concurrent with de-policing in 2020 and everything else that drove the increase in violence at the time, led to a similar increase in violence in Baltimore as many other jurisdictions saw.

    When you talk about the role of a prosecutor, you need the prosecutor as a backstop, the person who establishes the credibility of the threat of arrest. When you do not have a prosecutor likely to send people to prison, the credibility of that threat declines significantly and it becomes harder for police to do their job.

    Mounk: In terms of the impact of these things, let me steel-man the case on the other side. There is a widespread conviction, often based on somewhat shaky empirical evidence, that a huge share of the U.S. prison population is there because of nonviolent drug offenses. This is partially rooted in Michelle Alexander’s book The New Jim Crow. If you believe that, it seems very reasonable for prosecutors to say we should not be locking all of these people up because they are drug addicts who sadly use drugs, and some cop encounters them with a few grams of marijuana or a more serious substance in their pocket, and we are locking them away for a really long time. That seems genuinely bad.

    More broadly, the argument would be that things like gang enhancements are racially discriminatory: that they are simply a way of discriminating against Black men who grew up in poor neighborhoods with a lot of gang presence, where it is inevitable that some of their associates or acquaintances are gang members, meaning the same crime gets punished far more harshly than the same crime committed by a 17-year-old white teenager in an affluent suburb making a dumb mistake and getting off with a warning. Why should one person go to prison for far longer than the other? That is the steel-man. Why does that story miss the mark?

    Lehman: The answer is that it simply does not line up with the empirical facts. I once did the work of trying to run down some of the claims in Michelle Alexander’s book. I may get this wrong, but I believe she claims that the majority of the increase in the prison population since the 1980s is attributable to drug offenses. If you go into the footnotes and figure out where she got that number, she clearly misread the relevant table, and the actual figure is much smaller. There is a lot of that kind of thing in the literature, where you poke around and the empirics are fairly shoddy.

    The bigger picture is: when you look at who is really getting prosecuted for the kinds of offenses we are talking about, who was the actual object of prosecution in the war on drugs, my argument has been that the war on drugs was strategically misguided. It did not accomplish its primary goal of aggressively suppressing the drug trade. But it gets a bad rap for a bunch of other reasons that are not well supported. One of those is the assumption that we were locking up low-level nonviolent drug offenders for very long periods. There is a study I find useful that looks at the composition of the prison population in 1997, examines drug offenders, and finds that something like 2 percent of state offenders and 6 percent of federal offenders can be characterized as nonviolent, low-level drug offenders. Everyone else has a violent prior, is in a gang, was involved at the trafficking level, or was some sort of major offender.

    That gets to the deeper issue: the reason for prosecutorial discretion is that you need to be able to make decisions about who is and is not worth your time, and that is a person-level decision as much as an offense-level decision. Those critiques often look at what people get locked up for, what they get arrested for, but they cannot observe who is actually getting arrested and locked up. There are lots of arrests for drug possession every year, and almost nobody goes to prison for drug possession. Why? Because drug possession arrests are mostly a pretext to bring in someone you need to bring in, since it is easier to establish probable cause for drug possession than for the thing you actually need to lock them up for. That is a common dynamic.

    Mounk: Presumably, for some portion of that 2 percent of state prisoners and 6 percent of federal prisoners, the nonviolent conviction may also be the result of a plea deal: you likely have evidence of a violent crime, you are not entirely sure you will win the prosecution, or there is a lot of pressure on the system to avoid jury trials because they are so much more lengthy and costly, and so what you agree to in the plea deal is a nonviolent drug offense, even though the underlying reason for the prosecution is something quite different.

    Lehman: That is also often the dynamic, and it gets to the deeper point that in many cases what looks strange on paper is happening because of circumstances outside of what the paper is measuring. The deeper conversation is about how much discretion the criminal justice system should have. Often what progressive prosecutors are doing is saying this whole category of offense, this whole tactic or strategy, is going to be off limits, when a previous prosecutor would have said, I am going to use this as leverage toward my overall goal of incapacitating people who really need to be incapacitated. The progressive prosecutor worries a great deal about false positives, but as a result incorporates a lot of false negatives: this person should not be locked up because I do not like the charge we would use, I do not feel we can get them on the bigger thing, when a prior prosecutor would say, I can get them on this, and they need to go away for five to ten, so I am going to get them on this.

    In the rest of this conversation, Yascha and Charles discuss how to balance public safety whilst minimizing the risk of false convictions, and whether crime in major cities will continue to decline. This part of the conversation is reserved for paying subscribers…

    Read more

    1 hr 1 min
  • America’s AI Policy Is Truly Chaotic
    Vice Chairman Senator Mark Warner delivers an opening statement during a Senate Intelligence Committee hearing on worldwide threats in the Hart Senate Office Building on March 18, 2026 in Washington, D.C. (Photo by Win McNamee/Getty Images.)

    This year has been a crazy time with regard to Washington’s treatment of artificial intelligence, and the pace has picked up in the last couple of weeks. This train of events began in February when Defense Secretary Pete Hegseth declared the leading AI firm Anthropic, a “supply chain risk,” meaning that neither the Pentagon nor its contractors could use Anthropic models without risking legal liability.

    Then, in April, Anthropic announced that its latest Mythos model was so powerful that it would be released to only a handful of organizations. Mythos, it was claimed, had extraordinary abilities to break into computer systems, and these early organizations were asked to use it to test and secure their systems. In June, Senator Mark Warner told the Senate Banking Committee, “thank God it was Anthropic. When the head of the NSA and Cyber Command came and said, ‘This tool broke into almost all of our classified systems, not in weeks, but in hours,’ … we are not going to solve this problem if we rely on a less ethical CEO operating on the basis of plain voluntary testing alone.”

    Experts later added caveats to this description of Mythos 5’s capabilities, but it is clear that the politics of AI regulation has shifted dramatically in the past month. The Trump administration came into office last year being advised by tech bros like David Sacks and Marc Andreessen, who opposed any form of AI regulation. The latter in fact created an organization called Leading the Future that put significant money behind lobbying against AI regulation. The Trump White House nonetheless intervened against Anthropic, but only for what looked like typically bad political reasons. The company, according to Hegseth, was too “woke”; otherwise, the administration’s default position was to oppose all regulation.

    We’re expanding our events offerings! Please check out our events page to join Book Club, Ask the Author, and Intellectual Bootcamp—and to watch recordings of recent events.

    Persuasion events

    This has now changed. The national security community has woken up to the fact that the latest AI models pose a clear and present danger to the security of the government’s data. This is not a speculative outcome years in the future, but one that is here right now. The Sacks-Andreessen faction can’t really argue against this, and so the balance of opinion on the need for regulation—at least, the pre-approval of new AI models—has shifted in favor of intervention. The latest twist is the Commerce Department giving Anthropic the green light to give its latest Mythos 5 models to a group of “trusted partners” so that they could ferret out vulnerabilities in the security of their systems. Similarly, OpenAI announced last week that its newest model would be restricted to a few government-approved organizations. So the White House has now joined the larger consensus in the tech community that believes that, however useful AI may be, it needs regulation.

    The problem is that the Trump administration’s AI policy is typically chaotic. It is not clear who has final authority to issue new rules for AI, and on what basis. It appears that the turn towards regulation is driven by serious people in the national security community raising serious concerns, and not some clownish attempt by political actors like Pete Hegseth to punish woke enemies. But how are such decisions being made, and how will they be made going forward?

    The problem up to now was that pro-regulation advocates could not define clearly the sorts of harms regulation was meant to protect against. Now there is a clear harm (cybersecurity), and so a more thoughtful institutional design effort can begin. Such an effort would need to answer the following questions:

    • Do we need a specialized regulator for AI, or is this a function that can be added on to existing bodies like the NSA or the Defense or Commerce Departments?

    • Regardless of which government entity does the regulation, how do we get sufficient capacity into the bureaucracy so that it can make well-informed decisions? U.S. government capacity in this area is woefully deficient. The typical approach in the past has been to outsource state capacity to private actors, but under present circumstances this could easily lead to regulatory capture by big, self-interested players.

    • How does a regulator do surveillance and enforcement of any rules that it creates? The AI industry is huge and growing bigger by the day; moreover, foreign countries like China also have significant capabilities very close to those of the United States. If we deem a certain AI capability to be dangerous, how will we know that it is being developed, and how will we enforce rules limiting it? As Senator Warner’s remarks indicate, we are currently dependent on the good intentions of the CEOs running today’s large companies.

    • To what extent should we delegate discretionary authority to the new regulator? In the past, statutes have spelled out the specific rules that the regulator was meant to enforce. But the AI field is evolving so quickly that any effort to write such specifications into hard law will be almost immediately overtaken by technological change.

    These are just some of the real-world institutional design questions we need to answer if we are to adequately regulate artificial intelligence. My suspicion is that we will indeed need a specialized regulator, because AI is a sector very different from other parts of the economy. When trucking became an important means of transporting goods early in the 20th century, Congress decided to give regulatory authority to the Interstate Commerce Commission. But this was a mistake: the ICC was designed to regulate railroads, and the economics of trucking are very different from rail transport. This is why the advent of air transport led to the creation of specialized agencies like the FAA and CAB.

    Subscribe now

    It is time to stop talking about whether we need an AI regulator; the dangers are here and now and we need to move quickly. The discussion needs to shift to a much deeper and more specific analysis of how to regulate, understanding that the dangers posed by AI are likely to change over time. Our main competitor, China, is moving in this direction as we speak: the country is not run by a bunch of libertarians who want to let the technology rip, come what may. That may have been our position in the past, but it can’t define our policy today.

    Francis Fukuyama is the Olivier Nomellini Senior Fellow at Stanford University. His latest book is Liberalism and Its Discontents. He is also the author of the “Frankly Fukuyama” column, carried forward from American Purpose, at Persuasion.

    Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.

    And, to receive pieces like this in your inbox and support our work, subscribe below:

    Subscribe now

    0 min
  • Episode 190: Antonia Senior on the Cambridge Five

    Bookstack at Persuasion is delivered to you every other Sunday at 6am EST. Grab a coffee, settle onto the couch, and see your weekend out with a deep dive into the best new book releases…

    This week, Richard talks to Antonia Senior, author of Stalin’s Apostles: The Cambridge Five and the Making of the Soviet Empire (Public Affairs).

    Host: Richard Aldous

    Producer: Bhaasita Athani

    Follow Persuasion on X, LinkedIn, Instagram, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.

    And, to receive pieces like this in your inbox and support our work, subscribe below. Bookstack listeners can take advantage of a 20% offer for paid Persuasion membership, with access to special columns and bonus Good Fight with Yascha Mounk podcast content.

    20% off for Bookstack listeners!

    33 min
  • The Good Fight Club: Why the Center Left Is Losing, the Squad vs. the Median Voter, and How Patriotism Wins Elections

    Will you be in Washington, D.C. on Wednesday July 15? I will be interviewing Francis Fukuyama about how liberalism should respond to the postliberal threat. Find out more and get your free ticket here! —Yascha

    In this week’s episode of The Good Fight Club, Matthew Yglesias, Claire Ainsley, and Yascha Mounk examine why center-left parties are losing groun…

    Read more

    48 min
  • Pakistan Has Never Looked So Important

    Will you be in Washington, D.C. on Wednesday July 15? I will be interviewing Francis Fukuyama about how liberalism should respond to the postliberal threat. Find out more and get your free ticket here! —Yascha

    Pakistan’s prime minister Shehbaz Sharif with JD Vance and Qatar’s prime minister in Switzerland on June 21, 2026. (Photo by Fabrice Coffrini via Getty Images.)

    Pakistan’s reality from the very beginning, so it’s said, has always been about the three A’s: Allah, Army, America. The formal end of Donald Trump’s hot war against Iran appears to have added a fourth: Assertiveness.

    Pakistan brokered both the April 8 U.S.-Iran ceasefire and the longer-term halt to hostilities under a document called the Islamabad Memorandum of Understanding. It subsequently mediated follow-up talks in Switzerland on June 21.

    Key moments from those negotiations, preserved in multiple photos and videos across social media, show Pakistan bathed in a flattering glow. The U.S. delegation led by Vice President JD Vance warmly embraced the Pakistanis and repeatedly thanked them. In reference to the chief of the defense forces, Vance joked that there are “two very, very important people in my life, an Indian and a Pakistani. The Indian is my wife, and the Pakistani is Field Marshal Munir.” The Iranians, led by parliament speaker Mohammad Bagher Ghalibaf, were similarly cordial to the Pakistani leaders, though much less jokey. And Pakistan’s prime minister, Shehbaz Sharif, boasted about his “most cordial telephone conversation” with the ruler of Qatar.

    Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal values wherever they are threatened, please become a paying subscriber today!

    As well he might. Pakistan’s success as indispensable intermediary between the United States and Iran—two countries locked in nearly half a century of animosity—has won it great praise and much international goodwill. Pakistan’s founder, Muhammad Ali Jinnah, set out the bold aspiration for his country to be “the pivot of the world.” Nearly 80 years after its birth, Pakistan may finally be on the cusp of achieving that vision.

    It has played dealmaker before, of course, notably as the backchannel for the startling moment in 1972 when Richard Nixon became the first sitting U.S. president to set foot in the People’s Republic of China. It also had a key role in the 1988 Geneva Accords, which ended the Soviet occupation of Afghanistan.

    But this time feels different—Pakistan appears more self-assured than ever. This might seem strange for a country beset by deep-seated political and economic problems. It is, after all, on yet another International Monetary Fund loan program, its 25th since the 1950s. And there is general unease over its metamorphosis into a democracy with profoundly “Pakistani characteristics”—which, as one Pakistani analyst put it, means combining elections with “military hegemony and weak institutions.”

    But Pakistan’s view of itself and the way it deals with the world does seem to have changed. In his candid book Magnificent Delusion, Husain Haqqani, former ambassador to the United States, explains how it used to be: Jinnah asked America for an eye-popping $2 billion in September 1947, barely one month after the Pakistani state came into being (he got a mere $10 million the first year). In the early days, the most frequent visitor from Karachi to Washington was the Pakistani finance minister seeking aid.

    Now, the main PR pitch is Pakistan’s key location at the crossroads of the Middle East and South Asia, balancing a close relationship with China alongside a renewed relationship with Trump’s America. Though Pakistan is not in the Middle East, it has long had key relationships with regional players. It is a prominent member of the two-month-old group of Muslim countries known as R-4, which also includes Saudi Arabia, Egypt and Turkey, all of whom have either strong armies or deep pockets. ​At the very moment American and Iranian officials arrived in Switzerland for those Pakistan-brokered talks, foreign ministers of the R-4 were engaged in their fourth meeting since February.

    Pakistan is also the Muslim world’s only nuclear-armed state, making it well-placed to provide security guarantees; in fact, it formalized one with Saudi Arabia in the days after Israel’s September 2025 airstrike on the leadership of Hamas. Yoel Guzansky, formerly of Israel’s National Security Council, once mused about Pakistan “openly granting [a] ‘nuclear umbrella’” to Saudi Arabia. With the possibility of Turkey signing on to the Pakistan-Saudi defense alliance—five months ago, the Turkish foreign minister said that “discussions and talks are underway” on exactly this matter—a new alignment may be arriving. In a sign of growing alarm, former Israeli prime minister Naftali Bennett warned that “Turkey is the new Iran,” and claimed it is “trying to flip Saudi Arabia against us and establish a hostile Sunni axis with nuclear Pakistan.”

    Subscribe now

    Despite Israel’s alarm, the Trump administration has reasons for allowing Pakistan’s influence to grow: it does not want India to become the dominant regional power. On a recent visit to New Delhi, U.S. Deputy Secretary of State Christopher Landau declared: “We are not going to make the same mistakes with India that we made with China 20 years ago (by allowing it to) develop all these markets” and then start “beating” the United States in “a lot of commercial things.”

    Not only has Trump overturned decades of patient work by previous administrations who wanted to build a relationship with India, but he has ​also made clear his preference for Pakistan’s powerful Field Marshal Asim Munir. Pakistan has responded with shovel-loads of flattery, hailing Trump as the “savior of South Asia” and nominating him for a Nobel Peace Prize.

    Another reason for the closeness is that, nearly five years after the Taliban’s reconquest of Afghanistan, the United States and Pakistan are finally on the same page about the Taliban. It’s no secret that Islamabad’s relationship with Kabul is in tatters. Each accuses the other of harboring fighters who carry out deadly attacks on its territory. The breakdown of Pakistan’s long alliance with the Taliban could prove the ultimate asset in building a successful new relationship with the United States. As former diplomat Ted Craig has noted, Pakistan’s “disappointment” with Taliban 2.0. has caused it to stop tolerating the group, and the Pakistan military has also reduced “its costly support to Islamist militancy” in general. This has eliminated the main point of friction between the Pakistan security establishment and the United States.

    It’s not clear what will happen with the U.S.-Iran interim deal. It doesn’t resolve the thorniest issues: nuclear inspections, uranium enrichment limits, Iran’s ballistic-missile program, stewardship of the Strait of Hormuz, and how Iran will spend the billions of dollars soon to be released from sanctions. Meanwhile, Israeli operations in Lebanon could torpedo it all, Trump faces criticism at home, and there is a low rumbling of discontent in Tehran for even talking with the “Great Satan.”

    But whatever comes next, one party has done well out of the deal: Pakistan.

    Rashmee Roshan Lall is a journalist and academic. Her Substack is This Week, Those Books.

    Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.

    And, to receive pieces like this in your inbox and support our work, subscribe below:

    Subscribe now

    0 min
  • This Week At Persuasion
    My dog, Maude, enjoying the sunshine before the UK heatwave hit.Events
    • Our weekly Intellectual Bootcamp for paying subscribers is on Fridays at 12pm noon Eastern at this link. This week, Francis Fukuyama will be joining us to discuss The End of History. Find out more about Intellectual Bootcamp on our events page!

    Read more

    0 min

About Persuasion

From the publisher's feed

The community for those who believe that a free society is worth fighting for.