The directors and staff of the Cato Institute, and indeed friends of liberty throughout the world, are saddened by the passing of David Koch.

David’s accomplishments as a businessman and philanthropist are estimable. But we most remember and admire him for his efforts to advance liberty in the United States and around the world, and his steadfast dedication to libertarian principles.

David is a director emeritus of Cato, having joined our board of directors in 1986 and serving for nearly 30 years. Over this time, the Institute, the Cato community, and our work benefited greatly from his service, insight, generosity, and example. We will not forget the many ways in which he contributed to our mission.

Because the quality of our civil discourse has deteriorated in recent years, David’s dedication to our principles often earned him unfair criticism and excoriation in the public arena. But like us, he was motivated by a firm belief that liberty is the means to human flourishing, through which every individual is able to live a prosperous, meaningful life in a country and world at peace. He also believed strongly that the rights granted to us by nature and protected by the Constitution cannot be denied to any American for any reason. That he bore such unwarranted public criticism with dignity, and that it did not deter him from his work in advancing freedom, merits great respect from all of us.

We have no doubt that David’s partnership will be missed by so many of the causes to which he dedicated himself, not least of which is the cause of liberty. But of course, he will be missed most of all by his family, to whom we extend our heartfelt sympathy and to whom we turn our thoughts at this difficult time.

Since 2007, Mexico has seen a massive surge of gang warfare and violence unlike anything it has witnessed since the 1980s. Intuitively, this fact should lead more Mexicans to want to flee to the United States. Yet that hasn’t happened. In fact, illegal immigration from Mexico has all but disappeared. In 2019, Central Americans have now far outnumbered their more populous neighbor.

Figure 1 illustrates the relationship between the homicide rate in Mexico and the rate at which Border Patrol agents apprehended Mexicans. The number of apprehensions per agent is the best available measure of total illegal crossings because it controls for the level of enforcement. More agents will lead to more apprehensions without more crossings.

As violence dwindled in the late 1990s and early 2000s—from its peak of 21 homicides per 100,000 residents in 1986 to 8 in 2007—Mexican crossings fell precipitously as well, dropping from a 1986 peak of 510 apprehensions per agent to 60 in 2007. Suddenly, in 2008, the homicide rate jumped and remained at least twice as high as 2007, currently standing at three times the rate that year and the highest rate of the entire period. Yet Mexican illegal crossings continued to fall and have not returned.

The most important change related to Mexican immigration to the United States was the large increases in temporary work visas issued under the H‑2A agricultural and H‑2B nonagricultural programs. Figure 2 shows the number of H‑2 and H‑2B workers admitted from Mexico for each year from 1979 to 2019. In the late 1990s, the number of guest workers admitted legally started to rise, and in the post-2009 recession period, they have skyrocketed. Now that workers have a way to cross legally, they don’t need to cross illegally,

Most of the uptick in violence in Mexico is from gangs. Guest workers allow normal Mexican workers to obtain enough resources to relocate their families or otherwise insulate themselves from the gang wars. It also makes more violent communities more tolerable if adequate food and shelter are not also an issue, leading fewer to seek to enter illegally. Overall, this economic-focused strategy has clearly succeeded, and as I point out in my recent policy analysis, Congress should look to replicate it for Central America.

Maritime Administrator Mark Buzby has a problem. As head of the Maritime Administration he is charged with crewing and operating the Ready Reserve Force (RRF), a government-owned fleet used for the rapid deployment of U.S. military forces. Speaking at a Navy League-sponsored breakfast earlier this week, however, Buzby expressed worry there aren’t enough mariners to operate these ships. The RRF, while used in a military role, relies upon civilian mariners to operate them in wartime scenarios. And those mariners are in short supply.


In fact, a 2017 government report found that for a sustained sealift campaign the United States faces a deficit of approximately 1,800 mariners for those needed to crew the RRF and maintain commercial fleet operations—and that’s in a best-case scenario. The obvious remedy according to Buzby: increase the number of U.S.-flag ships to provide more employment opportunities.

“We believe we’re around 1,800 mariners short. So how do you make that up? That’s the question I get asked every single time. We need more places for people to work in peacetime. We need more…we need a larger U.S.-flag fleet by probably about 45 ships.”

Notably, Buzby is an ardent supporter of the Jones Act, the 1920 law which restricts the domestic waterborne transport of goods to vessels that are U.S.-built, U.S.-flagged, at least 75 percent U.S.-crewed, and at least 75 percent U.S.-owned. Indeed, at the same event he listed defending the law remains as among his top priorities. Yet the Jones Act’s U.S.-build requirement is a direct impediment to realizing the goal of more U.S.-flag ships.


That’s because commercial ships built in U.S. shipyards are expensive—frightfully so. A May 2019 Congressional Research Service report found that a U.S.-built tanker is roughly quadruple the price of a foreign-built vessel, while a U.S.-built container ship may be quintuple the price of one constructed abroad. For perspective, the same report said that the cost premium attached to U.S.-built ships shortly after the Jones Act’s passage was 20 percent.


This rise in price has correlated with a pronounced decline in the number of Jones Act-compliant ships. Fewer ships means fewer mariners to crew the RRF fleet.

It stands to reason that if Americans had access to cheaper ships that there would be more of them. But don’t take my word for it—U.S. shipyards themselves admit that high prices are a deterrent to the use of the ships they build.


In 2007 the Metal Trades Department of the AFL-CIO filed suit against the U.S. Coast Guard over its ruling allowing the use of foreign-built equipment modules in the construction of ships deemed to be U.S.-built. Unsurprisingly, U.S. shipyards sided with the Coast Guard. Preventing the use of foreign-built components, Aker Philadelphia Shipyard and General Dynamics-NASSCO argued, would make U.S.-built ships more costly and less attractive to purchase. That would mean both less shipbuilding and fewer vessels in the Jones Act fleet.


As Aker (now known as the Philly Shipyard) stated:

[p]reventing shipbuilders from using more efficient methods in constructing vessels will increase the vessel owners’ capital cost. This in turn will increase the rates that the vessel owners must charge, decreasing their competitiveness and further reducing their share of the domestic transportation market. The lower market share will lead to a reduction in the size and number of vessels needed to fulfill the demand for domestic shipping.

If more expensive ships means fewer ships, the reverse logically holds true—cheaper ships means more of them. And the cheapest solution of all would be to allow Americans to transport goods using ships built in other countries, just as they can for all other forms of transportation. That’s not just a good way to expand the U.S.-flag fleet, it’s what free people should be allowed to do.

In the last week, the Fifth and the Eighth Circuits, sitting en banc, have each issued major, fractured decisions on the subject of qualified immunity — the judge-made defense to civil rights claims under Section 1983, which shields state actors from liability for their misconduct, even when they break the law. In Cole v. Hunter, decided yesterday, the Fifth Circuit, in an 11–7 decision, affirmed the denial of summary judgment for two defendant police officers, who shot a teenage boy and then lied about what happened. The lawsuit brought by the victim and his family will therefore be able to go to trial, making this one of the rare instances where a civil rights plaintiff is able to overcome qualified immuniity. But in Kelsay v. Ernst, decided last week, the Eighth Circuit held, 8–4, that a police officer was entitled to qualified immunity, after he had grabbed a small woman in a bear hug and slammed her to ground — because she walked away from him. Although the courts here reached different outcomes, both cases amply illustrate the legal, practical, and moral infirmities with qualified immunity. The fractured decisions and many separate opinions in both cases also make clear that the doctrine is on increasingly shaky footing with both the judiciary and the general public.

I. Cole v. Hunter: A rare but narrow victory for a victim of egregious police misconduct

This Fifth Circuit case arose out of an incident in Garland, Texas in October 2010, when police were looking for Ryan Cole, then a 17-year-old boy, who had reportedly been walking around the neighborhood with a handgun. Ryan was seen by some officers and ordered to stop, but Ryan pointed the gun at his own head and walked away toward a wooded area. When Ryan reemerged, a group of officers observed him for about five seconds and did not announce themselves or give any warning. Then, while Ryan was facing away from them with the gun still pointed at his head, one or more of the officers fired at him, striking him several times and causing him to involuntary discharge his own gun into his skull and brain. When two of the officers were questioned after the shooting, they falsely claimed that Ryan had turned to face them and pointed his gun at them before they fired — an assertion belied by ample forensic and physical evidence (specifically, the location of Ryan’s bullet wounds, and the location of shell casings and Ryan’s blood).

Ryan and his family brought a civil rights suit against these officers, claiming that they used excessive force and fabricated evidence in violation of Ryan’s Fourth and Fourteenth Amendment rights. The district court denied qualified immunity to the defendants at summary judgment, a panel of the Fifth Circuit affirmed, and the Fifth Circuit then agreed to rehear the case en banc. The court then held, 11–7, that a reasonable jury could have found that Ryan posed no threat to the officers, and that it was clearly established that “shooting a mentally disturbed teenager, who was pointing a gun the entire time at his own head and facing away from the officer, in an open outdoor area, and who was unaware of the officer’s presence because no warning was given prior to the officer opening fire, was unlawful.” Although the defendants presented a starkly different view of the facts, the majority correctly recognized — as courts often fail to do in qualified immunity cases! — that it lacked jurisdiction to consider the officers’ competing factual narrative in this appeal. Rather, the disputed facts would have to be resolved by a jury.

Seven judges dissented, however. The principal dissent, by Judge Edith Jones, accused the majority of defining “clearly established law” at too high a level of generality, arguing that even if the facts clearly established that Ryan posed no threat (and thus that shooting him violated the Fourth Amendment) the relevant question for purposes of qualified immunity was “whether every reasonable officer in this factual context would have known he could not use deadly force” — and then arguing that no prior cases involved this precise factual context. This framing in the abstract is dutiful to the Supreme Court’s qualilfied immunity jurisprudence, but in application, it’s clear that the degree of specificity employed by the dissent would be practically impossible to overcome. To wit, the dissent went on to say:

[T[he importance of grounding the inquiry in a specific factual context cannot be overstated. In this case, if Officer Hunter had stood a hundred feet away from Cole, or Cole had not been turning toward the officers, or Cole had put the handgun in his pocket and wasn’t touching it, the analysis of qualified immunity could be quite different.

On the one hand, the dissent is correct that the Supreme Court has insisted that immunity analysis be “particularized” to the facts of individual cases. On the other, the example of factual distinctions used by the dissent here plainly illustrate that there will never be a prior case involving all of the potentially relevant facts — and even the Supreme Court has purported to say that a case exactly on point is unnecessary. Thus, while the dissent’s analysis is not a wholly unreasonable application of existing precedent, it demonstrates how the “clearly established law” standard is inherently amorphous, and incapable of consistent, predictable application.

Most notably, Judge Don Willett filed a separate dissent in this case. Although he would have held that “the Supreme Court’s unflinching, increasingly emphatic application of ‘clearly established law’ compel[led] dismissal,” the bulk of his opinion is devoted to explaining that “[t]he entrenched, judge-invented qualified immunity regime ought not be immune from thoughtful reappraisal.” Judge Willett’s dissent therefore echoes his recent criticisms of qualified immunity in Zadeh v. Robinson, although he did stress that the Supreme Court has “several ‘mend it, don’t end it’ options,” which would substantially revise the doctrine without eliminating it entirely. Also, he again discussed the Cato-organized cross-ideological amicus brief from Doe v. Woodard, noting that “perhaps the most ideologically diverse amici ever assembled” were urging the Supreme Court to reconsider qualified immunity.

II. Kelsay v. Ernst: Greenlighting egregious and unnecessary police violence against the supposed victim of a crime

Melanie Kelsay, her three children, and an adult friend of hers were swimming at a public pool in Wymore, Nebraska. She and her friend were engaged in what she called “horseplay,” but some onlookers thought he might be assaulting her and called the police. The police arrested her friend and put him a patrol vehicle, even though she repeatedly told them he hadn’t assaulted her; they then decided to arrest her, the alleged victim of this non-crime, because she was “getting in the way of the patrol vehicle door.” While talking with Deputy Matt Ernst, Kelsay saw that her daughter had gotten into an argument with a bystander, and tried to go check on her. Ernst grabbed her arm and told her to “get back here,” but released her. Kelsay then said she needed to go check on her daughter, and again began walking toward her. At that point, without giving any further instructions, Ernst ran up behind her, grabbed her, and slammed her to the ground in a “blind body slam” maneuver, knocking her unconscious and breaking her collarbone.

Kelsay then brought a Section 1983 suit against Ernst, and the district court denied qualified immunity, but a panel of the Eighth Circuit reversed, 2–1. The Eighth Circuit then agreed to rehear the case en banc, and affirmed the panel’s grant of qualified immunity, in an 8–4 decision. The majority, of course, relied on the idea that there were no prior cases involving the “particular circumstances” of this case; i.e., no prior cases specifically held that “a deputy was forbidden to use a takedown maneuver to arrest a suspect who ignored the deputy’s instruction to ‘get back here’ and continued to walk away from the officer.” No case exactly on point, qualified immunity, Q.E.D. The principal dissent by Chief Judge Lavenski Smith correctly noted that that the Supreme Court has never required “a case directly at point,” and that here, an ample body of case law would have “put a reasonable officer on notice that the use of force against a non-threatening misdemeanant who was not fleeing, resisting arrest, or ignoring other commands violates that individual’s right to be free from excessive force.”

To make matters worse, the majority refused even to decide whether Ernst’s conduct did, in fact, violate Kelsay’s Fourth Amendment rights. So any officer could engage in exactly the same misconduct tomorrow, and it still would not be “clearly established” that the conduct was unlawful. Judge Steven Grasz wrote a separate dissent taking issue with this particular aspect of the court’s decision. Though acknowledging that courts have discretion under Pearson v. Callahan to grant immunity without deciding the merits, he argued that the exercise of such discretion was “inappropriate in this case as it perpetuates the very state of affairs used to defeat Ms. Kelsay’s attempt to assert her constitutional rights.” Judge Grasz clearly grasps the circular nature of qualified immunity, noting that the judiciary’s persistent refusal to decide constitutional questions under Section 1983 “imposes a judicially created exception to a federal statute that effectively prevents claimants from vindicating their constitutional rights.”

III. The rising tide of opposition to qualified immunity

Two circuit courts recently decided to take qualified immunity cases en banc, even though neither case involved any new legal questions or suggested reversals of circuit precedent; that fact itself is a testament to how important this issue has become, and how much rising pressure there is to modify or abolish this doctrine. Cole and Kelsay hardly stand alone as examples of recent or ongoing high-profile qualified immunity cases. Just last month, the Eleventh Circuit issued a truly appalling decision in Corbitt v. Vickers, granting immunity to a deputy sheriff who shot a ten-year-old child lying on the ground, while repeatedly attempting to shoot a pet dog that wasn’t posing any threat. The Supreme Court unfortunately denied cert in Doe v. Woodard at the end of its last term, but there’s another outstanding cert petition explicitly calling for the Court to reconsider qualified immunity (Baxter v. Bracey), which will be considered at the long conference on October 1st. And there will probably be additional cert petitions challenging one or more of the decisions in Corbitt, Kelsay, and Cole.

In other words, this issue isn’t going away. Day by day, more and more lower court judges add their voices to the growing chorus calling upon the Supreme Court to reconsider this noxious doctrine, and agitation by public policy groups across the ideological spectrum continues to grow. Some of the Democratic presidential candidates have even started to call for the abolition of qualified immunity. The day of reckoning may well be coming soon.

Between the print version of Rebecca Traister’s August 5 New York magazine profile of Elizabeth Warren, and the version now online, there can be spotted an amusing correction. Print version:

Online version: 

Let’s hope editors in the nation’s leading financial center continue to keep in mind that lending money to someone doesn’t necessarily make you a predator. 

Yesterday, President Trump said the following about how he was “taking on” China in relation to its trade policy:

But one thing I have to do is economically take on China because China has been ripping us off for many years. President Clinton, President Bush, and President Obama, and others should have done this long before me. My life would be much easier — although I enjoy doing it — but my life would be much easier if I just said, “Let China continue to rip off the United States.” All right? It would be much easier, but I can’t do that.

We are winning against China. They’ve lost two and a half million jobs in a very short period of time. They want to make a deal. It’s got to be a deal that’s good for the United States, where they want to make a deal — probably, we will make a deal.

But if I didn’t do that — and I’m not doing this — somebody said it’s Trump’s trade war. This isn’t my trade war. This is a trade war that should have taken place a long time ago by a lot of other Presidents.

Over the last five or six years, China has made $500 billion. $500 billion. Ripped it out of the United States. And not only that — if you take a look, intellectual property theft. Add that to it. And add a lot of other things to it. So somebody —

Q (Inaudible.)

THE PRESIDENT: Excuse me. Somebody had to do it. I am the chosen one. Somebody had to do it. So I’m taking on China. I’m taking on China on trade. And you know what? We’re winning. Because we’re the piggybank. We’re the one that all these countries — including the European Union — wants to rob and takes advantage of. European Union — $200 billion. China — more than $500 billion. Sorry.

Q So it sounds like a recession is worth it —

THE PRESIDENT: I was put here —

Q — is that what you’re saying?

THE PRESIDENT: I was put here by people — I was put here by people to do a great job. And that’s what I’m doing. And nobody has done a job like I’ve done.

Now, would China rather wait for a little more than a year and try and get Sleepy Joe Biden to negotiate with, instead of President Trump? Maybe. But I don’t think so. You know why? They’re losing too many jobs too fast. They had the worst year in 27 years, but I think it was actually 52 or 54 years. It’s the worst year they’ve had in a half a century. And that’s because of me. And I’m not proud of that. But you know what? They want to negotiate.

And Sleepy Joe doesn’t have a clue. Sleepy Joe said, “Oh, China is wonderful.” Well, China is wonderful for China. But I’m wonderful for the U.S.A.

The transcript does not quite do this justice. It’s worth watching the video.

Putting aside the inaccurate description of the U.S. trade balance with China (they are not “ripping us off,” we are simply trading with them), is President Trump “taking on” China? It depends what you mean by that. If you mean, is President Trump taking actions that could induce China to reduce its protectionism, to protect intellectual property better, and to let foreign companies invest in China without transferring technology to their Chinese partners, then I’m not sure he is. The Trump administration has imposed tariffs on Chinese imports in a way that has led to China retaliating with tariffs of its own. But there hasn’t been much indication so far that a deal to address China’s bad practicies is in sight. We may just end up with higher tariffs that stay in place as long as Trump is president.

Instead of public bluster and an arbitrary use of tariffs, the better approach would be to negotiate in the normal way, which often results in trade liberalization. We make some demands and offer some concessions, and the other side does the same. Why didn’t that happen under previous administrations? It’s hard to say for sure, but it’s possible that one reason it didn’t happen under President Bush is that we were so distracted by the mess in the Middle East that China was overlooked. President Obama “pivoted to Asia,” and as part of that negotiated the Trans Pacific Partnership, which could have put pressure on China by excluding it from the TPP’s trade liberalization and thus leaving it at a disadvantage. But Trump withdrew from the TPP.

So here we are. Right now, the Trump administration looks like it favors tariffs over negotiated trade liberalization. The main Democratic contenders for the 2020 election have not said enough to get a sense of how they would approach trade policy in relation to China, although some have expressed skepticism about at least some of Trump’s tariffs. It would be nice if someone would eventually “take on” China, in the sense of adopting an approach to trade policy that encourages China to liberalize.

As I wrote a couple of weeks ago, back-to-school also means the release of lots of education polling, and Tuesday brought us the yearly Education Next poll, which is one of my favorites. (Of course, I love all you crazy polls!) The Education Next folks do a lot of interesting experimentation with their polls, especially when it comes to funding, and they keep nice longitudinal data. I don’t love every part of the survey—looking at you, Common Core question—but overall I think it is well done and highly informative.

As usual, you should read the whole thing, and I’ll just hit some highlights.

More, more, more!

If there is one repeated theme to the poll, it’s that people generally want more of whatever is being discussed: more spending, higher teacher pay, more school choice, more accountability. People also tend to like their public schools and state colleges, and seem to want more higher ed.

Not especially well informed

While the members of the public express opinions on many education issues, they aren’t always especially knowledgeable, as you would expect of people with regular jobs and lives and not a ton of time to delve deeply into public policy issues, including education. They tend to greatly underestimate how much we spend on public schooling (by about 47 percent), public school teacher salaries (by about 30 percent), and they know either very little or the wrong things about charter schools (no, they cannot charge tuition or hold religious services).

School Choice

As you can see below, support for all sorts of choice—charters, vouchers, scholarship tax credits—rose this year, and scholarship tax credits remain the most popular option. This is probably because a lot of people like school choice and even more like tax credits. Charters have also rebounded after taking a sizeable dive between 2016 and 2017.

Common Core

As we’ve seen before, a significant majority of the public likes the idea of common standards across states that would “be used to hold public schools accountable for their performance.” How much this question reflects support for common standards, accountability, or both is unclear. What is clear is if you attach the name of specific curriculum standards to it—the Common Core—support drops, though it has been inching up as the battle over the Core, and the Core itself, fades in the public memory.

Free College

“Free” college is a bad idea for numerous reasons we have discussed many times—and apparently a lot of economists agree—but there is no question that the idea, at least in the abstract, is popular. 69 percent of the public favors it for two-year schools, and 60 percent for four-year institutions. Troubling, but who doesn’t like a free lunch?

Again, lot’s more to see here, so check it all out.