Since this section discusses various campaign proposals, I’ll reiterate:
I could not be happier with my decision not to cover the election outside of the particular areas that I already cover. I have zero intention of telling anyone who to vote for. That’s for you to decide.
All right, that’s out of the way. On with the fun. And it actually is fun, if you keep your head on straight. Or at least it’s fun for me. If you feel differently, no blame for skipping the section.
Last time the headliner was Kamala Harris and her no good, very bad tax proposals, especially her plan to tax unrealized capital gains.
This time we get to start with the no good, very bad proposals of Donald Trump.
The details of the announcement speech at the link are pure gold. Love it.
The economists, he said, told him he would get ‘a whole new workforce.’
Yes, that would happen, and now it’s time for Solve For the Equilibrium. What would you do, if you learned that ‘overtime pay’ meaning anything for hours above forty in a week was now tax free? How would you restructure your working hours? Your reported working hours? How many vacations you took versus how often you worked more than forty hours? The ratio of regular to overtime pay? Whether you were on salary versus hourly? What it would mean to be paid to be ‘on call,’ shall we say?
I used this question as a test of GPT-4o1. Its answer was disappointing, missing many of the more obvious exploitations, like alternating 80 hour work weeks with a full week off combined with double or more pay for overtime. Or shifting people out of salary entirely onto hourly pay.
I often work more than 40 hours a week for real, so I’d definitely be restructuring my compensation scheme. And let’s face it, the ‘for real’ part is optional.
This of course is never going to happen. If it did, it would presumably include various rules and caps to prevent the worst abuses. But even the good version would be highly distortionary, and highly anti-life. You are telling people to intentionally shift into a regime where they work more than 40 hours a week as often as possible, the opposite of what we as a society think is good. This is not what peak performance looks like, even working fully as intended.
Less fun Trump proposals are things like bringing back the SALT deduction (what, why, I am so confused on this one?) and a 10% cap on interest on credit cards. Which would effectively be a ban on giving unsecured credit cards with substantial limits to anyone at substantial risk of not paying it back or require other draconian fees and changes to compensate, and lord help us if actual interest rates ever approached 10%. Larry Summers notesthat this is a dramatic price cut on the order of 70% for many customers, as opposed to other proposed price controls that are far less dramatic and thus less destructive, so it would have far more dramatic effects faster. If payday loans are included they’re de facto banned, if not then people will substitute those far worse loans for their no longer available credit cards.
Then there’s ‘I’m going to bring down auto insurance costs by 50%’ where I could try to imagine how he plans to do that but what would even be the point.
Also there is his plan to ‘make auto loan interest tax deductible’ which is another fun one. Already car companies often make most of their money on financing. The catch is the standard deduction, which you have to give up in order to claim this. If the car loan is the only big item you’ve got, it won’t help you. What you need is some other large deduction, which will usually be a home loan. So this is essentially a gift to homeowners – once you’re deducting your mortgage interest, now you can also deduct your car loan interest. It makes no economic sense, but Elon Musk will love it, and it’s not that much stupider than the mortgage deduction. Of course, what we should actually do is end or phase out the mortgage deduction (as a compromise you could keep existing loans eligible but exclude new ones, since people planned on this), but I’m a realist.
Also there’s Trump’s other proposed huge giveaway and trainwreck, which is a quiet intention to ‘privatize’ Fannie Mae and Freddie Mac. I put privatize in air quotes because if you think for one second we would ever allow these two to fail then I have some MBS to sell you. Or buy from you. I’m not sure which. Quite obviously we are backing these two full on ride or die, so this would mean socialized losses with privatized gains and another great financial crisis waiting to happen.
As Arnold Kling suggests, we could and likely should instead greatly narrow the range of mortgages the government backs, and let the private sector handle the rest at market prices. When we back these mortgages, the subsidy is captured by existing homeowners and raises prices, so what are we even doing? Alas, I doubt we will seriously consider that change.
Another note on the unrealized capital gains issue is what happens to IP that pays out over time. For example, Taylor Swift suddenly owns a catalog worth billions, that could gain hundreds of millions in value when interest rates shift. Are you going to force her to pay tax on all that? How is she going to do that without selling the catalog? You want to force her to do that? Or do you want her to find a way to intentionally sabotage the value of the catalog?
We have some good news on the grocery price control front, as Harris has made clear that her plan would not involve global price controls on groceries and widespread food shortages. Instead, it will be modeled on state-level price gouging laws, so that in an emergency we can be sure that food joins the list of things that quickly becomes unavailable at any price, and no one has the incentive to stock up on or help supply badly needed goods during a crisis.
Tariffs are terrible, but not as bad as I previously thought, if there is no retaliation?
Justin Wolfers: Here’s a rule of thumb that Goldman draws from the literature:
Roughly 15% of a tariff is borne by exporters from the other country.
Another 15% results in compressed margins for American importers.
70% of the burden is borne by consumers paying higher prices.
The first 15% is indeed then ‘free money’ and the second 15% is basically fine. So if you were to use the tariff to reduce other taxes, and the other country didn’t retaliate, you’d come out ahead. You get deadweight loss from reduced volume due to the 70%, but you face similar issues at least as much with almost every other tax.
We use an advanced model of the global economy to consider a set of scenarios consistent with the proposal to impose a minimum 60% tariff against Chinese imports and blanket minimum 10% tariff against all other US imports. The model’s structure, which includes imperfect competition in increasing-returns industries, is documented in Balistreri, Böhringer, and Rutherford (2024). The basis for the tariff rates is a proposal from former President Donald Trump (see Wolff 2024). We consider these scenarios with and without symmetric retaliation by our trade partners.
Our central finding is that a global trade war between the United States and the rest of the world at these tariff rates would cost the US economy over $910 billion at a global efficiency loss of $360 billion. Thus, on net, US trade partners gain $550 billion. Canada is the only other country that loses from a US go-it-alone trade war because of its exceptionally close trade relationship with the United States.
…
When everyone retaliates against the United States, the closest scenario here to a US-led go-it-alone global trade war, China actually gains $38.2 billion.
Noah Smith does remind us that no, imports do not reduce GDP. Accounting identities are not real life, and people (including Trump and his top economic advisor) are confusing the accounting identity for a real effect. Yes, some imports can reduce GDP, in particular imports of consumer goods that would have otherwise been bought and produced internally. But it is complicated, and many imports, especially of intermediate goods, are net positive for GDP.
The context of his comment was a hearing where people quite insanely proposed to ban lap infants on flights, which the FAA has to fight back against every few years by pointing out that flying is far safer than other transportation.
So such a ban would actively make us less safe by forcing people to drive.
A comment points out Jeremy is playing loose here: 4% is who listed this as the primary reason for being out of the labor force. A lot more did have difficulty.
Jeremy: Also, the prime-age employment rate is near all-time highs — some men aren’t in the LF, this is true, but women are employed at by far the highest rate ever. This suggests that the number of jobs isn’t the problem, but something (or things) are making men drop out (see above).
And the prime age employment rate is highest for native-born workers
Yes, a lot of those jobs are terrible. But that has always been true.
Kalshi will pay 4.05% on both cash and open positions, which will adjust with Fed rates. That’s a huge deal. The biggest barrier to long term prediction markets is the cost of capital, which is now dramatically lower.
Election prediction market update: As I write this, Polymarket continues to be the place to go for the deep markets, and they have Trump at 55% to win despite very little news. So we’ve finally broken out of the period where the market odds were strangely 50/50 for a long time, likely for psychological reasons driving traders. The change is also reflected in the popular vote market, with Trump up to 31% there, about 8% above his lows. Nate Silver’s predictions have narrowed, he has Harris at 51% to win, down from a high of 58%.
The move seems rather large given the polls and lack of other events. My interpretation is that the market is both modestly biased in favor of Trump for structural reasons (including that it’s a crypto market and Trump loves crypto) and that the market is taking a no-news-is-good-for-Trump approach.
I haven’t heard anyone think of it that way, but it makes sense to me. Consider the debate. Clearly the debate was good for Harris, including versus expectations. But also the debate was expected to be good for Harris, so before the debate the polls were underestimating Harris in that way. One could similarly say that Harris generally has more opportunity to improve and less chance of imploding or having health issues over the last two months, so her chances go down a little if Nothing Ever Happens.
As many have pointed out, there is little difference between 44% Harris at Polymarket, and 51% Harris at Silver Bulletin. Even if one of them wins decisively, it won’t mean that one of them is right and the other wrong. To conclude that you have to look at the details more carefully.
We’ve gone over this before but it bears repeating, and I like the way this got presented this time around. How bad are our marginal tax rates for those seeking to climb into the middle class, once you net out all forms of public assistance, taxes and expenses?
Brad Wilcox: Truly astonishing indictment of our welfare policies fr @AtlantaFed. A single mother in DC can make no gains, financially, as her earnings rise from $11,000 to $65,000 because benefits like food stamps & Medicaid phase in/out as her income rises. Terrible for work/marriage.
Andrew Jobst: Talked to someone who lost their job in the GFC (highly educated, driven, professional credentials). Wanted to start her own business. Commented about how demoralizing it was to hustle all day to earn another dollar, only for her unemployment benefit to drop by a dollar.
Benefits are not ‘as good as cash’ so the problem probably is not quite as bad as ‘100% effective marginal tax rates from $10,000 in income up to $65,000’ but it could be remarkably close, especially in places with high additional state taxes. Can you imagine what would happen if you took a world like this, and you stopped counting tips as taxable income, as proposed by both candidates?
Effectively, you’d have a ~100% tax rate on non-tip income, but 0% on tips (and Trump would add overtime). Until you could ‘escape’ well above the $65k threshold, basically everyone would be all but obligated to fight for only jobs where they could get paid in these tax-free ways, with other jobs being essentially unpaid except to get you to the $10k threshold.
Given these facts, what is remarkable is how little distortion we see. Why isn’t there vastly more underground economic activity? Why don’t more people stop trying to earn money, or shift between trying to earn the minimum and then waiting to try until they’re ready to earn the maximum, or structuring over time?
My presumption is that this is because the in-kind benefits and conditional benefits are worth a lot less than these charts value them at. Cash is still king. So while the effective rate is still quite high, we don’t actually see 100% marginal tax rates.
If you want more income, Tyler Cowen suggests perhaps you could work more hours? A new estimate says 20% of variance in lifetime earnings is in hours worked, although that seems if anything low, especially given as Tyler points out that working more improves your productivity and human capital.
Tyler Cowen: In the researchers’ model, 90% of the variation in earnings due to hard work comes from a simple desire to work harder. Note again this is an average, so it does not necessarily describe the conditions faced by, say, Elon Musk or Mark Zuckerberg.
In my experience, vastly more than 20% of my variance in income comes from the number of hours worked and how hard I was working generally. One could draw a distinction between hours worked versus working hard during those hours. I’d guess the bigger factor is how hard I work when I’m working, but the times I’ve succeeded and gotten big payoffs, it wouldn’t have happened at all if I hadn’t consistently worked hard for a lot of hours. The time I wasn’t able to deliver that effort, at Jane Street, it was exactly that failure (and what caused that failure) that largely led to things not working out.
Working hard also applies to influencers. In this job market paper from Kazimier Smith, he finds that the primary driver of success is lots of posting. Sponsored posts grow reach the same as regular posts, which is nice work if you can get it, although this results likely depends on influencers selecting good fits and not overdoing it, and on correlation, where if you are getting sponsorships it is a sign you would otherwise be growing.
The abstract also introduced the question of focus and audience capture. Influencers and other content creators have to worry that if they don’t give the people what they want, they’ll lose out, and I’ve found that writing on certain topics, especially gaming, creates permanent loss of readers. I’d love to see the proper version of that paper too.
Since we’ve now had some major storms, it’s time for another round of reminding everyone that laws against ‘price gouging’ are a lot of why it we so quickly run out of gas and other supplies in emergency situations. Why would you stock extra in case of emergency, if you only can sell for normal prices? Why would you bring in extra during an emergency, if you can only sell for normal prices?
Dr. Insensitive Jerk: Our relatives in the Florida evacuation zone just told us I-75 is a parking lot, and no gasoline is available.
Do you know why no gasoline is available? Because of price-gouging laws.
Pointing this out provokes a predictable emotional response from adult children. “He should give me gas cheaply! He should store an infinite amount of gasoline so he can fill up all the hoarders, and still have gas left for me, and he should do it for the same price as last week!”
Now when Floridians need gasoline desperately, they can’t buy it at any price, because other Floridians said, “It’s cheap, so I might as well fill the tank.”
People outside Florida with tanker trucks full of gasoline might have considered helping, but instead they said, “I won’t risk it. If I charge enough to make it worth my while, I will be arrested and vilified in the press.”
But at least the Floridians won’t have to lie awake in their flooded houses worrying that somebody made a profit from rescuing them.
Alas, the Bloomberg editorial board will keep on writing correct takes like ‘Price Controls Are a Bipartisan Delusion’ (the post actually downplays the consequences in a few cases, if anything) and we will go on doing it.
Maxwell Tabarrok: High prices during emergencies aren’t gouging – they’re bounties for desperately needed goods. Like a sheriff offering a big reward to catch a dangerous criminal, these prices incentivize the entire economy to rush supplies where they’re most needed.
With two major hurricanes in the last couple of weeks, “price gouging” is in the news. In addition to it’s violent name, there are good intuitive reasons to dislike price gouging.
But imagine if you were the sheriff of Ashville, NC, and it was your job to get more gasoline and bring it into town.
You might offer a bounty of $10 a gallon, dead or alive.
That’s a lot more than the usual everyday bounty, but this is an emergency.
Prices aren’t just a transfer between buyer and seller.
They’re also also a signal and incentive to the whole world economy to get more high-priced goods to the high-paying area; they’re a bounty.
The last thing you’d want if you were the sheriff is a cap on the bounty price you’re allowed to set.
High prices on essential goods during an emergency are WANTED posters, sent out across the entire world economy imploring everyone to pitch in and catch the culprit.
The difficulty that many people may have in paying these higher prices is a serious tragedy, and one that can be alleviated through prompt government response e.g by sending relief funds and shipping in supplies. But setting prices lower doesn’t mean everyone can access scarce and expensive essential goods. In an emergency, there simply aren’t enough of them to go around.
Setting low prices might mean the few gallons of gas, bottles of water, or flights that are available are allocated to people who get to them first, or who can wait in line the longest, but it’s not clear that these allocations are more egalitarian.
These allocations leave the central problem unsolved: A criminal is on the loose and a hurricane has made it difficult to get these goods to where they’re needed.
When there’s an emergency and a criminal is on the loose, we want the sheriff to set the bounty high, and catch ‘em quick. High prices during other emergencies work the same way. Let the price-system sheriff do his work!
Scott Sumner points out that customers very much prefer ridesharing services that price gouge and have flexible pricing to taxis that have fixed prices, and very much appreciate being able to get a car on demand at all times. He makes the case that liking price gouging and liking the availability of rides during high demand are two sides of the same coin. The problem is (in addition to ‘there are lots of other differences so we have only weak evidence this is the preference’), people reliably treat those two sides very differently, and this is a common pattern – they’ll love the results, but not the method that gets those results, and pointing out the contradiction often won’t help you.
Chinese VC fundraising and VC-backed company formation has fallen off a cliff, after China decided they were going to do everything they could to make that happen.
Bill Gurley: Many in Washington are preoccupied with China. If this article is accurate, the #1 thing we could do to improve US competitiveness, would be to open the door much more broadly & quickly to skilled immigration. Give these amazing entrepreneurs a home on US soil.
It’s important to note these are private VC funds and VC-backed companies only. This is not the picture of all new enterprise in China. There are plenty of new companies.
According to FT, venture capital has died because the Chinese government intentionally killed it. They made clear that you will be closely monitored, your money is not your own and cannot be transferred offshore, your company is not your own, the authorities could actively go after the most successful founders like Jack Ma, that you are to reflect ‘Chinese values’ or else. Venture capital salaries are capped.
What is left of venture is often suing companies to get their money back, so the government doesn’t accuse them of not trying to get the money back on behalf of the government. New founders are required to put their house and car on the line.
The advocates of Venture Capital and the related startup ecosystem present it as the lifeblood of economic dynamism, innovation and technological progress. If they are correct about that, then this is a fatal blow.
Often we hear talk about ‘beating China,’ along with warnings of how we will ‘lose to China’ if we do some particular thing that might interfere with venture capital or the tech sector. Yet here we have China doing something ten or a hundred or a thousand times worse than any such proposals. Yet I don’t expect less worrying about China?
One perspective listing what 2% compounding annual economic growth feels like once you get to your 40s. It is remarkably similar to my experience – I look around and realize that the stuff I use and value most is vastly better and cheaper, life in many ways vastly better, things I used to spend lots of time on now at one’s fingertips for free or almost free.
We argue that workers must take costly actions (“conflict”) to have nominal wages catch up with inflation, meaning there are welfare costs even if real wages do not fall as inflation rises.
We study a menu-cost style model, where workers choose whether to engage in conflict with employers to secure a wage increase.
…
We conduct a survey showing that workers are willing to sacrifice 1.75% of their wages to avoid conflict. Calibrating the model to the survey data, the aggregate costs of inflation incorporating conflict more than double the costs of inflation via falling real wages alone.
Matt Bruenig: Also worth considering the degree to which “conflict costs” constitute another of the frictions that prevent job-switching (people don’t like upsetting their boss/colleagues), which again points towards collective bargaining as important and a limitation of anti-monopsony.
I got a job once that I left after 6 weeks because I got an unexpected offer that paid about $20k more per year and boy did I have to hear what a piece of shit I was from the person who hired me in the first job. It’s as if they had never even read the textbook.
Matt Yglesias: This resonates with me as I ask myself why I re-upped my Bloomberg column contract at the same nominal salary without even attempting to negotiate for a higher fee.
Except I have seen unions, and whatever else you think of unions they do not exactly minimize such conflicts, instead frequently leading to deadweight losses including strikes. And I have no doubt that inflation substantially increases the average costs of such conflicts.
The reason a worker would pay to avoid conflict with the boss is partly it is unpleasant, partly The Fear, and partly because it can result in anything from turning the work situation miserable up through a full ‘you’re fired,’ or in the union case a strike. At minimum, it risks burning a bunch of goodwill.
Also Matt should realize that when you take a new job after six weeks and quit, you have imposed rather substantial costs on your old employer. During those six weeks, you were probably a highly unproductive employee. They spent a lot of time hiring you, training you, getting you up to speed, and then you burned all that effort and left them in another lurch.
Of course they are going to be mad, although the bigger the gap in offers the less mad they should be. We’ve decided that the employee doesn’t strictly owe the employer anything here, it’s a risk the employer has to take, but at minimum they owe them the right to be pissed off – you screwed them, whether or not it was right to do that.
Another way to look at this is that the decline in real wages is a cost, which then often means other costs get imposed, including deadweight losses like switching jobs or threatening to do so, in order to fix it, but that as is often the case those new costs are a substantial portion of the original loss.
There are also the actual real losses. This is especially acute in situations that involve wages being sticky downwards, or someone is otherwise ‘above market’ or above their negotiating leverage. For example, when I joined [company], I was given a generous monthly salary. I stayed for years, but that number was never adjusted for inflation, because it was high and I needed my negotiating points for other things – I didn’t want to burn them on a COLA or anything.
Often salary negotiations happen at times of high worker leverage, when they have another offer or are being hired or had just proven their value or what not. Having to then renegotiate that periodically is at minimum a lot of stress.
As one commenter noted, sufficiently high inflation can actually be better here. If there’s 2% inflation a year, then you’re tempted to sit back and accept it. If it’s 7%, then you have a fairly straightforward argument you need an adjustment.
But this policy was sufficiently known and reliable that it resulted in absolutely massive tax evasion, as in 95% of people earning under $2,000 a year flat out not bothering to file. Needless to say, at that point you might as well set the tax for such people to $0 and tell them they don’t need to file.
John Horton: If you listen, insurance companies are giving you solid, data-driven advice about stuff not to do or buy—don’t own a pit bull, don’t have a trampoline, don’t under-water cave dive, don’t own a “cyber” truck…
what’s kind of nuts is that when instead of just quoting you a higher price, they explicitly just will not cover it. To me, that suggests they think adverse selection is a problem. It’s not *justthat pit-bulls are natural toddler-eaters, they think you’re a reckless idiot and a higher price just increases the average idiocy of the customers, with predictable results
Gwern: Or they don’t have enough data.
The problem is, insurance companies only need correlates. So none of that is good advice about stuff you should do – unless you are planning to starting to transition to a woman because of lower insurance rates for women on many things…?
Robert Parham: Upon inspection, it seems like a externality issue. The cybertruck is so tough that any accident with it leaves the truck unscathed while totalling the other car. The Insurance company is liable for the totalled car, hence the decision.
Insurance is indeed pretty great for things like internalizing that your cybertruck would be very bad for any other car that got into an accident with it. The problem is that when you price out trampoline insurance, a lot of this is that people who tend to buy trampolines are reckless, so you don’t know how much you should avoid owning one.
I even wonder if ‘arbitrary’ price differentials would be good. If you charge less for insurance on houses that are painted orange than those painted green, and someone still wants to insure their green house, well, do they sound like responsible people?
As the tech job market continues to struggle, I’m seeing more threads like this asking if it’s time to reevaluate career and college plans based around being a software engineer. My answer continues to be no. Learning how to code and build things is still a high expectancy path.
Work from home allows workers to be paid for the 10 hours they actually work, without having to semi-waste the other 30. What is often valuable is the ability to suddenly work 60-80 hours a week when it matters, or that one meeting or day when you’re badly needed, and it’s fine to work 10 hours (or essentially 0 hours) most other weeks, and the payment is so you’re on standby.
Detty: The most surreal aspect of the WFH vs. in-office debate is how it’s widely acknowledged that hundreds of millions of people do very little all day every day and yet the economy continues to just churn & those who don’t have the magic piece of paper work very hard for very little.
Seth Largo: Lots of corporations and institutions are so wealthy that it makes sense to pay someone a full time salary for 10 hours of work per week, because those 10 hours really do help keep the machine running, and no one’s gonna do it for 10 hours of pay.
Lindy Manager: Also managers need people available who can activate for bursts when needed who have all the context and information to create or present something of sufficient quality on short notice for a client or executive.
Seth Largo: Don Draper knew this.
ib: Yep. A lot of corporate salaries are effectively retainers.
Always Adblock: Yes. And to keep their institutional knowledge. And to keep them away from competitors.
Had this section in reserve for a post that likely will never come together on its own, so figured this was a good time for it.
Paper concludes minimum wage increases drive increased homelessness due to disemployment effects and rental price increases, and dismisses migration as a potential cause. I mean, yes, obviously, on the main result.
A better question is, what does the minimum wage do to rental costs? The minimum wage does successfully cause some work to become higher paid. Most such workers will not be homeowners. It is entirely plausible that landlords could capture a large portion of these gains via higher rents for low-quality housing, perhaps all of it. In which case, what was the point?
Restaurants in Milan used to be forced to be distant from each other, then they stopped requiring that, resulting in agglomeration that caused diverging amenities in different neighborhoods, and increased product differentiation. Tyler Cowen notes ‘I am myself repeatedly surprised how much the mere location of a restaurant can predict its quality.’
I would think of this less as returns to agglomeration and more as it being costly to force restaurants to locate in uneconomical locations, and to effectively undersupply some areas, leading to lack of competition and variety there, while oversupplying others. By creating product differentiation in location, this reduces their incentive to otherwise differentiate or seek higher quality.
The U.S. college wage premium doubles over the life cycle, from 27 percent at age 25 to 60 percent at age 55. Using a panel survey of workers followed through age 60, I show that growth in the college wage premium is primarily explained by occupational sorting. Shortly after graduating, workers with college degrees shift into professional, nonroutine occupations with much greater returns to tenure.
Nearly 90 percent of life cycle wage growth occurs within rather than between jobs. To understand these patterns, I develop a model of human capital investment where workers differ in learning ability and jobs vary in complexity. Faster learners complete more education and sort into complex jobs with greater returns to investment. College acts as a gateway to professional occupations, which offer more opportunity for wage growth through on-the-job learning.
Tyler Cowen suggests this causes problems for the signaling model of education. I disagree, and see this result as overdetermined.
Path dependence. Those who go to college then enter professions and careers that allow for such wage growth, from a combination of skills development and social and reputational accumulation. Thus, whatever mix of signaling, correlation and education is causing these other paths, the paths are opened by college, and this has a predictable effect over time.
In particular: Gatekeeping. I don’t buy that future employers will no longer care if you went to college. Many high paying jobs will be difficult or impossible to get without a degree, and the degree helps justify paying someone more, since pay is largely about affirming social status. Gatekeeping thus keeps such people increasingly down over time as results compound, and also discourages investment. Why develop human capital that no one will pay for?
Correlational. If you go to college, this is a revealed preference for longer time horizons and longer term investment, including the capacity and capability to do it. It makes sense that such folks would continue to invest in human capital growth over time relative to others.
In particular: Signaling. Alas, those more willing to invest more time and resources in signaling likely get better compensated over time. Also college plausibly teaches you how to signal.
Catching up. If you take a job rather than go to college, you are going to start out with several years of practical experience, which gives you a temporary advantage that fades over time. College students first entering the workforce are famously out of touch and useless, lacking practical skills, and are coming from a sheltered academic world with unproductive norms. Over time, you get over it.
This paper estimates differences in human capital as country-of-origin specific labor productivity terms, in firm production functions, making it immune to wage discrimination concerns. After accounting for wage and experience, estimated human capital varies by a factor of around 3 between the 90th and 10th percentile. When I investigate which country-of-origin characteristics correlate most closely with human capital, cultural values are the only robust predictor. This relationship persists among children of migrants. Consistent with a plausible cultural mechanism, individuals whose origin place a high value on autonomy hold a comparative advantage in positions characterized by a low degree of routinization.
I don’t understand why we want to be shouting this from the rooftops. These types of correlations are the kind that very much do not imply causation, the whole thing is doubtless confounded to hell and back and depends on a bunch of free variables. Autonomy is one of those values that maps reasonably closely with ‘The West’ and so does the level of human capital.
The core claim is that if your culture values autonomy, then you are better suited to a less routine production activity and hold comparative advantage there. Which is a case where I am confused why we needed a study or mathematical model. How could that have been false? Less routine is not the same as more autonomous but the correlation is going to be very high. People with cultural value X hold comparative advantage in activities that embody X, paper at conference?
War Discourse and the Cross Section of Expected Stock Returns finds that the paper’s model of what war tail risks should be worth does not match the market’s past evaluation of what war tail risks should be worth, and decides it is the market that is wrong. I am highly open the market mispricing things like this, especially in response to media salience, but I’m even more open to the academics being wrong.
Paper claims that we are gaining 0.5% per year in terms of how much welfare we get from across a variety of categories from increased product specialization and variety. Households increasingly spend funds on specialized products that exactly fit their preferences, with the increased variety driving the divergence in consumption.
This is also evidence we are richer. Increased product variety requires people able to consume enough, and pay enough extra for quirky preferences, to justify greater product variety. This represents a real welfare gain. However, instead of making people feel less constrained and wealthier, it puts strain on budgets and competes with and potentially puts additional strain on raising families rather than making it cheaper to raise one.
I very much appreciate the product variety, but increasingly I think we need to consider three different measures of wealth:
The welfare value of the experience of the items in a typical consumption basket.
The combined welfare value including goods that remain unpriced.
The difficulty in purchasing the typical consumption basket, and what affordances that leaves for life goals especially retirement, marriage and children.
Or: The Iron Law of Wages proposes that real wages tend toward the minimum to sustain the life of the worker. So we can measure four things.
The minimum real wages required to sustain the life of the worker.
The welfare value of that minimum consumption basket.
The surplus available after that to the typical worker and what that buys them.
What else is available that is not priced.
When we either effectively mandate additional consumption, such as purchasing additional safety, health care, residence size, education or other product features, or our culture effectively demands such purchases, or the cheaper alternatives stop being available, what happens?
We do increase the welfare value of the minimum basket. We also raise the cost of that basket, which reduces everyone’s surplus.
What happens when things that people value, like community and friendship and the ability to raise children without being terrified of outside intervention, and opportunities to find a good life partner, are degraded?
Life gets worse without it showing up in the productivity statistics or in real wages.
The current crisis and confusion could be thought of as:
The value of the minimum consumption basket is going up a lot.
The cost of the minimum consumption basket is going up less than that.
Real wages are going up, but less than the cost of the basket, so the surplus available after purchasing the basket is also declining.
Key other goods and options are taken away, like those mentioned above.
Economists say ‘workers are better off,’ and in many ways they are.
People say ‘but I have little surplus and do not see how to meet my life goals and I have no hope and my life experience is getting worse.’
Paper explores the impact of the 2010 dissolution of personal income tax reciprocity between Minnesota and Wisconsin. This looks like it on average raised effective taxes on work across state lines by about 8% of remaining net income. This resulted in a decline in quantity of cross-border commuters between 3% and 5%, with the largest impact on low and young earners. My hunch is that the impact size is so low primarily because of inertia, switching costs and lack of understanding of the costs. Whereas jobs that don’t pay as well, and those of the young, are less sticky. It would be shocking if an 8% tax had this small an effect at equilibrium.
Paper estimates that the CARD Act, which limits credit card interest charges and fees, saved consumers $11.9 billion per year, lowering borrowing costs by 1.6% overall and by 5.3% for those with FICO below 660. What is odd is they also find no corresponding decrease in available credit, despite this making offering credit less profitable. There is no free lunch. A potential story is that credit cards adjusted their other costs and benefits, or the counterfactual here is not well established and there would have been growth in credit otherwise, or the good version is that the whole enterprise is so profitable and useful that the banks ate the reduced profits.
There’s also the strange graph below, which requires explaining. Patrick McKenzie points out that the part of the FICO curve where offering credit cards is unprofitable is still a good place to do business, because those in the unprofitable range are unlikely to stay there so long and their business will remain somewhat sticky as they move.
Has real median income gone up under Biden? This clart implies that it perhaps hasn’t, even if weird timing is involved, and that this explains a lot. Yes, pay has increased since 2019, and increased since 2022, but the question people often effectively ask is since the end of 2020.
‘Total compensation’ is cool but what people look at is the actual money.
The first season opened with the murder of Juliette’s lover, George (Ferdinand Kingsley), who collected forbidden historical artifacts, which silo sheriff Holston Becker (David Oyelowo) investigated at Juliette’s request. When he chose to go outside, he named Juliette as his successor, and she took on George’s case as well as the murder of silo mayor Ruth Jahns (Geraldine James). Many twists ensued, including the existence of a secret group dedicated to remembering the past whose members were being systemically killed. Juliette also began to suspect that the desolate landscape seen through the silo’s camera system was a lie and there was actually a lush green landscape outside.
In the season finale, Juliette made a deal with Holland: She would choose to go outside in exchange for the truth about what happened to George and the continued safety of her friends in Mechanical. The final twist: Juliette survived her outside excursion and realized that the dystopian hellscape was the reality, and the lush green Eden was the lie. And she learned that their silo was one of many, with a ruined city visible in the background.
The official S2 trailer picks up there but doesn’t provide many additional details. We see Juliette in her protective suit walking across the desolate terrain toward the other silos, human skulls and bones crunching under her feet. When Juliette’s oxygen runs out, she finds shelter and survives, and we later see her trying to enter a silo—whether it’s her original home or another one is unclear. Meanwhile, Holland gives an impassioned speech to his silo residents, declaring her a hero for sacrificing herself. But rumors swirl that she is alive, and rebellion is clearly brewing, with Juliette becoming a symbol for the movement.
The second season of Silo debuts on Apple TV+ on November 15, 2024. Ferguson has said that there are plans for third and fourth seasons to wrap up the story, which will hopefully be filmed at the same time.
Studies suggested it could treat cancer, but the clinical trials were a bust.
Pamela Goodwin has received hundreds of emails from patients asking if they should take a cheap, readily available drug, metformin, to treat their cancer.
It’s a fair question: Metformin, commonly used to treat diabetes, has been investigated for treating a range of cancer types in thousands of studies on laboratory cells, animals, and people. But Goodwin, an epidemiologist and medical oncologist treating breast cancer at the University of Toronto’s Mount Sinai Hospital, advises against it. No gold-standard trials have proved that metformin helps treat breast cancer—and her recent research suggests it doesn’t.
Metformin’s development was inspired by centuries of use of French lilac, or goat’s rue (Galega officinalis), for diabetes-like symptoms. In 1918, researchers discovered that a compound from the herb lowers blood sugar. Metformin, a chemical relative of that compound, has been a top type 2 diabetes treatment in the United States since it was approved in 1994. It’s cheap—less than a dollar per dose—and readily available, with few side effects. Today, more than 150 million people worldwide take the stuff.
The French lilac, Galega officinalis, has been used medicinally since medieval times, including for symptoms associated with diabetes. Investigations of the plant’s chemical galegine led to the development of metformin, a related molecule synthesized in the lab. Credit: Wikimedia Commons
Metformin has a variety of effects, such as improving immune function and the body’s responses to insulin, which in turn regulates blood sugar. It can also slow growth of cancer cells in the lab. Many of these benefits seem to stem from metformin’s action in the cell’s powerhouses, the mitochondria, where it slows the production of energy and limits the generation of damaging chemicals called free radicals.
Researchers have considered metformin for treating a plethora of conditions, from glaucoma to polycystic ovary syndrome to pimples. “It really has a reputation of being a potential wonder drug,” says Michael Pollak, an oncologist and researcher at McGill University in Montreal. “There’s still a lot of work to be done on metformin.” (Pollak consults for biotechnology companies interested in metformin analogs as medicines.)
But the latest research has convinced Pollak and some others that treatment of cancers should be taken off the list.
More studies, but no proof
One of the first hints linking metformin to anticancer effects came in a short note in the British Medical Journal in 2005. Researchers analyzed medical records of almost 12,000 people from the Tayside region of Scotland who were newly diagnosed with diabetes between 1993 and 2001. Of those, more than 900 went on to develop cancer. Interestingly, those who’d taken metformin at some point during the study period were 23 percent less likely to have received a later cancer diagnosis.
This finding fueled further research on people with diabetes taking metformin and the risk for breast cancer, liver cancer, ovarian and endometrial cancer, and other types. The authors of a 2013 analysis, covering more than 1 million patients in 41 observational studies like the original one, concluded that metformin “might be associated with a significant reduction in the risk of cancer.” But such associations are not proof.
Researchers went on to explore the link in studies with cells in dishes and in lab animals, finding that metformin slowed growth of blood, breast, endometrial, lung, liver, stomach, and thyroid cancer cells. It also seemed to make cancer cells extra sensitive to chemotherapy drugs. In one mouse study, scientists grafted human breast, prostate, or lung cancer cells into the animals and treated them with either standard chemotherapy drugs, metformin, or a combination of both. The combination worked best, preventing tumor growth and prolonging relapse.
These findings made sense to researchers. Metformin treats metabolic problems in diabetes, and cancer has also been linked to metabolic issues such as obesity. Even before the 2005 British Medical Journal study, Goodwin had noticed that breast cancer patients with high insulin did worse than those with normal insulin levels.
That logic, plus the promising data, led scientists to conduct a number of randomized controlled trials—the gold-standard experiment in medicine. Researchers would enroll people with cancer and split them into two groups. One group would get standard cancer therapy plus metformin; the other group would get standard therapy plus a placebo, a pill containing no medication.
And metformin flopped, big time. While a number of studies are ongoing, trials for two types of cancer recently reported no benefit overall from metformin. In June 2024, at the American Society of Clinical Oncology meeting in Chicago, researchers reported a Canadian trial with 407 men with low-risk prostate cancer. The enrollees had been diagnosed within six months before starting the trial and had decided to monitor their cancer without starting immediate treatment. Half took metformin and half took a placebo. After biopsies at 18 and 36 months to test whether their disease had progressed, there was no difference between the two groups.
A larger British and Swiss trial including nearly 1,900 patients with newly diagnosed or relapsed prostate cancer that had spread to other body parts was reported at the European Society for Medical Oncology Congress in Barcelona, Spain, in September. This trial also found that metformin plus standard treatment, compared to standard treatment alone, did not improve overall prostate cancer survival in the study population.
A multinational study of breast cancer helmed by Goodwin also led to disappointment. The researchers enrolled more than 3,600 patients between 2010 and 2013; these patients had been diagnosed about a year before enrollment and had already undergone chemotherapy and surgery. In addition to standard cancer treatment, half received metformin and half received a placebo.
By 2016, it was clear that metformin wasn’t doing anything to enhance survival for about 1,100 participants with a particular cancer subtype. When the study wrapped in 2020, the researchers analyzed the rest of the patients, counting how many were alive and free of breast or any other form of cancer. Metformin made no difference in those results, or to survival overall, the team reported in 2022.
Fatal flaws in the research
In retrospect, researchers think they know why earlier studies oversold metformin’s potential. Many of the studies that examined medical records had a crucial flaw, says Samy Suissa, a pharmacoepidemiologist at McGill.
Here’s what happens: Researchers sift through old medical records to see if someone ever took metformin. Then they compare cancer rates among people who took the drug at any point to those who never took it. But you have to be alive to take metformin. Anyone who died, of cancer or other causes, before having a chance at a metformin prescription is left out of the calculations. This skews the results; it’s called the “immortal time bias.” It makes any drug, metformin or otherwise, look like it helps patients to survive because it can only be taken by people who are alive, says Suissa.
Plus, scientists are more likely to publish studies that show metformin is promising than ones where it makes no difference, skewing the scientific literature.
As for those studies of cells in dishes and of lab animals, many experiments used much higher doses of metformin than are used in people. Too much metformin risks a buildup of lactate, a byproduct of low oxygen metabolism that acidifies the blood and can be fatal.
Researchers still suspect metformin might treat specific subgroups of cancer. For example, the authors of the prostate cancer trial presented in Barcelona suggested that metformin might help patients whose cancer has spread to other tissues or multiple sites in their bones. And Goodwin saw a hint in her trial that it might help women whose cancers contain a certain version of a cell-growth gene called ERBB2. But it would require another trial, focused on women with that particular cancer, to prove it.
And there are now better treatments for those patients than there were more than a decade ago when Goodwin started her study, reducing the opportunity to test metformin. Goodwin doesn’t currently have the funding to follow up on this theory.
It may also be that the clinical trials recruited patients with cancers that were too far along. “I always thought we were asking too much of metformin,” says Victoria Bae-Jump, a gynecological oncologist at the University of North Carolina Lineberger Comprehensive Cancer Center in Chapel Hill. “Maybe it just needs to be earlier in the pathway of growth.” Bae-Jump is now testing metformin in women who have early-stage endometrial cancer or a precursor to it.
Others are investigating metformin for people who have precancerous lesions in their mouths. “The idea would be to keep them from progressing, or reverse the tissues to be more normal,” says Frank Ondrey, a head and neck cancer surgeon at the Masonic Cancer Center of the University of Minnesota in Minneapolis. In a small, uncontrolled study of 23 people, metformin halved lesion size in four of them. Ondrey is involved in two ongoing studies, one a randomized, controlled trial, to further test metformin in people with precancerous lesions; these should yield results within a few years.
Subdued expectations
Metformin is also being tested for other conditions such as dementia and a genetic disorder called fragile X syndrome. And perhaps the ultimate potential use for metformin is to slow aging itself. “I think it’s much easier to treat aging and prevent cancer than to treat cancer,” says Nir Barzilai, a geroscientist at Albert Einstein College of Medicine in New York and president of the nonprofit Academy for Health & Lifespan Research. Through its enhancement of insulin action and metabolism plus its minimization of free radical production, metformin influences all the key hallmarks of aging, such as problems with DNA, mitochondria and stem cells, says Barzilai.
He and colleagues are gathering funds for a randomized, controlled trial of metformin in 3,000 people age 65 through 79 who are showing signs of age-related disease already. The trial will test whether fewer people taking metformin die over six years. Barzilai, who is 68, says he is confident in metformin’s anti-aging ability and already takes the drug himself.
Others, mindful of what happened with cancer, are more circumspect. Pollak says that many of the studies in other areas of medicine are too small to prove metformin works, and Suissa notes that some of the studies finding benefits in populations taking metformin, including for longevity, have the same problems the oh-so-promising early cancer research did.
In short, Suissa says, “Don’t believe everything you hear.”
On Thursday, AMD announced its new MI325X AI accelerator chip, which is set to roll out to data center customers in the fourth quarter of this year. At an event hosted in San Francisco, the company claimed the new chip offers “industry-leading” performance compared to Nvidia’s current H200 GPUs, which are widely used in data centers to power AI applications such as ChatGPT.
With its new chip, AMD hopes to narrow the performance gap with Nvidia in the AI processor market. The Santa Clara-based company also revealed plans for its next-generation MI350 chip, which is positioned as a head-to-head competitor of Nvidia’s new Blackwell system, with an expected shipping date in the second half of 2025.
In an interview with the Financial Times, AMD CEO Lisa Su expressed her ambition for AMD to become the “end-to-end” AI leader over the next decade. “This is the beginning, not the end of the AI race,” she told the publication.
According to AMD’s website, the announced MI325X accelerator contains 153 billion transistors and is built on the CDNA3 GPU architecture using TSMC’s 5 nm and 6 nm FinFET lithography processes. The chip includes 19,456 stream processors and 1,216 matrix cores spread across 304 compute units. With a peak engine clock of 2100 MHz, the MI325X delivers up to 2.61 PFLOPs of peak eight-bit precision (FP8) performance. For half-precision (FP16) operations, it reaches 1.3 PFLOPs.
Two years ago, Tesla’s Optimus prototype was an underwhelming mess of exposed wires that could only operate in a carefully controlled stage presentation. Last night, Tesla’s “We, Robot” event featured much more advanced Optimus prototypes that could walk around without tethers and interact directly with partygoers.
It was an impressive demonstration of the advancement of a technology Tesla’s Elon Musk said he thinks “will be the biggest product ever of any kind” (way to set reasonable expectations, there). But the live demos have also set off a firestorm of discussion over just how autonomous these Optimus robots currently are.
A robot in every garage
Before the human/robot party could get started, Musk introduced the humanoid Optimus robots as a logical extension of some of the technology that Tesla uses in its cars, from batteries and motors to software. “It’s just a robot with arms and legs instead of a robot with wheels,” Musk said breezily, easily underselling the huge differences between human-like movements and a car’s much more limited input options.
After confirming that the company “started off with someone in a robot suit”—a reference to a somewhat laughable 2021 Tesla presentation—Musk said that “rapid progress” has been made in the Optimus program in recent years. Extrapolating that progress to the “long term” future, Musk said, would lead to a point where you could purchase “your own personal R2-D2, C-3PO” for $20,000 to $30,000 (though he did allow that it could “take us a minute to get to the long term”).
And what will you get for that $30,000 when the “long term” finally comes to pass? Musk grandiosely promised that Optimus will be able to do “anything you want,” including babysitting kids, walking dogs, getting groceries, serving drinks, or “just be[ing] your friend.” Given those promised capabilities, it’s perhaps no wonder that Musk confidently predicted that “every one of the 8 billion people of Earth” will want at least one Optimus, leading to an “age of abundance” where the labor costs for most services “declines dramatically.”
Dan Riccio, one of Apple’s most prominent executives for more than two decades, will retire from the company this month, according to a report in Bloomberg that cites people with knowledge of the move.
Reportedly, Riccio has said he has been planning his retirement for the past five years, and his last day will be Friday, October 11.
Riccio began working at Apple in 1998, and by 2012, he had become the chief of hardware engineering. In that role, he oversaw several major hardware developments for Apple, including AirPods, the evolution of the modern iPhone, the iPad Pro, and more.
He held the title of senior vice president of hardware engineering during that time, then moved into a new role within the company in January of 2021. The public at first only knew that he was working on a “new project” at that time, but before long it became clear the project in question was what became the Vision Pro, Apple’s augmented-reality headset that launched this February.
The group that produced the Vision Pro is called the Vision Products Group within the company; that’s the 2,000-engineer-strong group Riccio has overseen since 2021. He was also involved in developing Project Titan, Apple’s smart car initiative that was eventually abandoned.
Google called the DOJ extending search remedies to AI “radical,” an “overreach.”
The US Department of Justice finally proposed sweeping remedies to destroy Google’s search monopoly late yesterday, and, predictably, Google is not loving any of it.
On top of predictable asks—like potentially requiring Google to share search data with rivals, restricting distribution agreements with browsers like Firefox and device makers like Apple, and breaking off Chrome or Android—the DOJ proposed remedies to keep Google from blocking competition in “the evolving search industry.” And those extra steps threaten Google’s stake in the nascent AI search world.
This is only the first step in the remedies stage of litigation, but Google is already showing resistance to both expected and unexpected remedies that the DOJ proposed. In a blog from Google’s vice president of regulatory affairs, Lee-Anne Mulholland, the company accused the DOJ of “overreach,” suggesting that proposed remedies are “radical” and “go far beyond the specific legal issues in this case.”
From here, discovery will proceed as the DOJ makes a case to broaden the scope of proposed remedies and Google raises its defense to keep remedies as narrowly tailored as possible. After that phase concludes, the DOJ will propose its final judgement on remedies in November, which must be fully revised by March 2025 for the court to then order remedies.
Even then, however, the trial is unlikely to conclude, as Google plans to appeal. In August, Mozilla’s spokesperson told Ars that the trial could drag on for years before any remedies are put in place.
In the meantime, Google plans to continue focusing on building out its search empire, Google’s president of global affairs, Kent Walker, said in August. This presumably includes innovations in AI search that the DOJ fears may further entrench Google’s dominant position.
Scrutiny of Google’s every move in the AI industry will likely only be heightened in that period. As Google has already begun seeking exclusive AI deals with companies like Apple, it risks appearing to engage in the same kinds of anti-competitive behavior in AI markets as the court has already condemned. And giving that impression could not only impact remedies ordered by the court, but also potentially weaken Google’s chances of winning on appeal, Lee Hepner, an antitrust attorney monitoring the trial for the American Economic Liberties Project, told Ars.
Ending Google’s monopoly starts with default deals
In the DOJ’s proposed remedy framework, the DOJ says that there’s still so much more to consider before landing on final remedies that it reserves “the right to add or remove potential proposed remedies.”
Through discovery, DOJ said that it plans to continue engaging experts and stakeholders “to learn not just about the relevant markets themselves but also about adjacent markets as well as remedies from other jurisdictions that could affect or inform the optimal remedies in this action.
“To be effective, these remedies… must include some degree of flexibility because market developments are not always easy to predict and the mechanisms and incentives for circumvention are endless,” the DOJ said.
Ultimately, the DOJ said that any remedies sought should be “mutually reinforcing” and work to “unfetter” Google’s current monopoly in general search services and general text advertising markets. That effort would include removing barriers to competition—like distribution and revenue-sharing agreements—as well as denying Google monopoly profits and preventing Google from monopolizing “related markets in the future,” the DOJ said.
Any effort to undo Google’s monopoly starts with ending Google’s control over “the most popular distribution channels,” the DOJ said. At one point during the trial, for example, a witness accidentally blurted out that Apple gets a 36 percent cut from its Safari deal with Google. Lucrative default deals like that leave rivals with “little-to-no incentive to compete for users,” the DOJ said.
“Fully remedying these harms requires not only ending Google’s control of distribution today, but also ensuring Google cannot control the distribution of tomorrow,” the DOJ warned.
To dislodge this key peg propping up Google’s search monopoly, some options include ending Google’s default deals altogether, which would “limit or prohibit default agreements, preinstallation agreements, and other revenue-sharing arrangements related to search and search-related products, potentially with or without the use of a choice screen.”
A breakup could be necessary
Behavior and structural remedies may also be needed, the DOJ proposed, to “prevent Google from using products such as Chrome, Play, and Android to advantage Google search and Google search-related products and features—including emerging search access points and features, such as artificial intelligence—over rivals or new entrants.” That could mean spinning off the Chrome browser or restricting Google from preinstalling its search engine as the default in Chrome or on Android devices.
In her blog, Mulholland conceded that “this case is about a set of search distribution contracts” but claimed that “overbroad restrictions on distribution contracts” would create friction for Google users and “reduce revenue for companies like Mozilla” as well as Android smart phone makers.
Asked to comment on supposedly feared revenue losses, a Mozilla spokesperson told Ars, “[We are] closely monitoring the legal process and considering its potential impact on Mozilla and how we can positively influence the next steps. Mozilla has always championed competition and choice online, particularly in search. Firefox continues to offer a range of search options, and we remain committed to serving our users’ preferences while fostering a competitive market.”
Mulholland also warned that “splitting off” Chrome or Android from Google’s search business “would break them” and potentially “raise the cost of devices,” because “few companies would have the ability or incentive to keep them open source, or to invest in them at the same level we do.”
“We’ve invested billions of dollars in Chrome and Android,” Mulholland wrote. “Chrome is a secure, fast, and free browser and its open-source code provides the backbone for numerous competing browsers. Android is a secure, innovative, and free open-source operating system that has enabled vast choice in the smartphone market, helping to keep the cost of phones low for billions of people.”
Google has long argued that its investment in open source Chrome and Android projects benefits developers whose businesses and customers would be harmed if those efforts lost critical funding.
“Features like Chrome’s Safe Browsing, Android’s security features, and Play Protect benefit from information and signals from a range of Google products and our threat-detection expertise,” Mulholland wrote. “Severing Chrome and Android would jeopardize security and make patching security bugs harder.”
Hepner told Ars that Android could potentially thrive if broken off from Google, suggesting that through discovery, it will become clearer what would happen if either Google product was severed from the company.
“I think others would agree that Android is a company that is capable [being] a standalone entity,” Hepner said. “It could be independently monetized through relationships with device manufacturers, web browsers, alternative Play Stores that are not under Google’s umbrella. And that if that were the case, what you would see is that Android and the operating system marketplace begins to evolve to meet the needs and demands of innovative products that are not being created just by Google. And you’ll see that dictating the evolution of the marketplace and fundamentally the flow of information across our society.”
Mulholland also claimed that sharing search data with rivals risked exposing users to privacy and security risks, but the DOJ vowed to be “mindful of potential user privacy concerns in the context of data sharing” while distinguishing “genuine privacy concerns” from “pretextual arguments” potentially misleading the court regarding alleged risks.
One possible way around privacy concerns, the DOJ suggested, would be prohibiting Google from collecting the kind of sensitive data that cannot be shared with rivals.
Finally, to stop Google from charging supra-competitive prices for ads, the DOJ is “evaluating remedies” like licensing or syndicating Google’s ad feed “independent of its search results.” Further, the DOJ may require more transparency, forcing Google to provide detailed “search query reports” featuring currently obscured “information related to its search text ads auction and ad monetization.”
Stakeholders were divided on whether the DOJ’s initial framework is appropriate.
Matt Schruers, the CEO of a trade association called the Computer & Communications Industry Association (which represents Big Tech companies like Google), criticized the DOJ’s “hodgepodge of structural and behavioral remedies” as going “far beyond” what’s needed to address harms.
“Any remedy should be narrowly tailored to address specific conduct, which in this case was a set of search distribution contracts,” Schruers said. “Instead, the proposed DOJ remedies would reshape numerous industries and products, which would harm consumers and innovation in these dynamic markets.”
But a senior vice president of public affairs for Google search rival DuckDuckGo, Kamyl Bazbaz, praised the DOJ’s framework as being “anchored to the court’s ruling” and appropriately broad.
“This proposal smartly takes aim at breaking Google’s illegal hold on the general search market now and ushers in a new era of enduring competition moving forward,” Bazbaz said. “The framework understands that no single remedy can undo Google’s illegal monopoly, it will require a range of behavioral and structural remedies to free the market.”
Bazbaz expects that “Google is going to use every resource at its disposal to discredit this proposal,” suggesting that “should be taken as a sign this framework can create real competition.”
AI deals could weaken Google’s appeal, expert says
Google appears particularly disturbed by the DOJ’s insistence that remedies must be forward-looking and prevent Google from leveraging its existing monopoly power “to feed artificial intelligence features.”
As Google sees it, the DOJ’s attempt to attack Google’s AI business “comes at a time when competition in how people find information is blooming, with all sorts of new entrants emerging and new technologies like AI transforming the industry.”
But the DOJ has warned that Google’s search monopoly potentially feeding AI features “is an emerging barrier to competition and risks further entrenching Google’s dominance.”
The DOJ has apparently been weighing some of the biggest complaints about Google’s AI training when mulling remedies. That includes listening to frustrated site owners who can’t afford to block Google from scraping data for AI training because the same exact crawler indexes their content in Google search results. Those site owners have “little choice” but to allow AI training or else sacrifice traffic from Google search, The Seattle Times reported.
Remedy options may come with consequences
Remedies in the search trial might change that. In their proposal, the DOJ said it’s considering remedies that would “prohibit Google from using contracts or other practices to undermine rivals’ access to web content and level the playing field by requiring Google to allow websites crawled for Google search to opt out of training or appearing in any Google-owned artificial-intelligence product or feature on Google search,” such as Google’s controversial AI summaries.
Hepner told Ars that “it’s not surprising at all” that remedies cover both search and AI because “at the core of Google’s monopoly power is its enormous scale and access to data.”
“The Justice Department is clearly thinking creatively,” Hepner said, noting that “the ability for content creators to opt out of having their material and work product used to train Google’s AI systems is an interesting approach to depriving Google of its immense scale.”
The DOJ is also eyeing controls on Google’s use of scale to power AI advertising technologies like Performance Max to end Google’s supracompetitive pricing on text ads for good.
It’s critical to think about the future, the DOJ argued in its framework, because “Google’s anticompetitive conduct resulted in interlocking and pernicious harms that present unprecedented complexities in a highly evolving set of markets”—not just in the markets where Google holds monopoly powers.
Google disagrees with this alleged “government overreach.”
“Hampering Google’s AI tools risks holding back American innovation at a critical moment,” Mulholland warned, claiming that AI is still new and “competition globally is fierce.”
“There are enormous risks to the government putting its thumb on the scale of this vital industry—skewing investment, distorting incentives, hobbling emerging business models—all at precisely the moment that we need to encourage investment, new business models, and American technological leadership,” Mulholland wrote.
Hepner told Ars that he thinks that the DOJ’s proposed remedies framework actually “meets the moment and matches the imperative to deprive Google of its monopoly hold on the search market, on search advertising, and potentially on future related markets.”
To ensure compliance with any remedies pursued, the DOJ also recommended “protections against circumvention and retaliation, including through novel paths to preserving dominance in the monopolized markets.”
That means Google might be required to “finance and report to a Court-appointed technical committee” charged with monitoring any Google missteps. The company may also have to agree to retain more records for longer—including chat messages that the company has been heavily criticized for deleting. And through this compliance monitoring, Google may also be prohibited from owning a large stake in any rivals.
If Google were ever found willfully non-compliant, the DOJ is considering a “range of provisions,” including risking more extreme structural or behavioral remedies or enduring extensions of compliance periods.
As the remedies stage continues through the spring, followed by Google’s prompt appeal, Hepner suggested that the DOJ could fight to start imposing remedies before the appeal concludes. Likely Google would just as strongly fight for any remedies to be delayed.
While the trial drags on, Hepner noted that Google already appears to be trying to strike another default deal with Apple that appears pretty similar to the controversial distribution deals at the heart of the search monopoly trial. In March, Apple started mulling using Google’s Gemini to exclusively power new AI features for the iPhone.
“This is basically the exact same anticompetitive behavior that they were found liable for,” Hepner told Ars, suggesting this could “weaken” Apple’s defense both against the DOJ’s broad framework of proposed remedies and during the appeal.
“If Google is actually engaging in the same anti-competitive conduct and artificial intelligence markets that they were found liable for in the search market, the court’s not going to look kindly on that relative to an appeal,” Hepner said.
Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience.
Back in 2019, Google made waves by claiming it had achieved what has been called “quantum supremacy”—the ability of a quantum computer to perform operations that would take a wildly impractical amount of time to simulate on standard computing hardware. That claim proved to be controversial, in that the operations were little more than a benchmark that involved getting the quantum computer to behave like a quantum computer; separately, improved ideas about how to perform the simulation on a supercomputer cut the time required down significantly.
But Google is back with a new exploration of the benchmark, described in a paper published in Nature on Wednesday. It uses the benchmark to identify what it calls a phase transition in the performance of its quantum processor and uses it to identify conditions where the processor can operate with low noise. Taking advantage of that, they again show that, even giving classical hardware every potential advantage, it would take a supercomputer a dozen years to simulate things.
Cross entropy benchmarking
The benchmark in question involves the performance of what are called quantum random circuits, which involves performing a set of operations on qubits and letting the state of the system evolve over time, so that the output depends heavily on the stochastic nature of measurement outcomes in quantum mechanics. Each qubit will have a probability of producing one of two results, but unless that probability is one, there’s no way of knowing which of the results you’ll actually get. As a result, the output of the operations will be a string of truly random bits.
If enough qubits are involved in the operations, then it becomes increasingly difficult to simulate the performance of a quantum random circuit on classical hardware. That difficulty is what Google originally used to claim quantum supremacy.
The big challenge with running quantum random circuits on today’s hardware is the inevitability of errors. And there’s a specific approach, called cross-entropy benchmarking, that relates the performance of quantum random circuits to the overall fidelity of the hardware (meaning its ability to perform error-free operations).
Google Principal Scientist Sergio Boixo likened performing quantum random circuits to a race between trying to build the circuit and errors that would destroy it. “In essence, this is a competition between quantum correlations spreading because you’re entangling, and random circuits entangle as fast as possible,” he told Ars. “We use two qubit gates that entangle as fast as possible. So it’s a competition between correlations or entanglement growing as fast as you want. On the other hand, noise is doing the opposite. Noise is killing correlations, it’s killing the growth of correlations. So these are the two tendencies.”
The focus of the paper is using the cross-entropy benchmark to explore the errors that occur on the company’s latest generation of Sycamore chip and use that to identify the transition point between situations where errors dominate, and what the paper terms a “low noise regime,” where the probability of errors are minimized—where entanglement wins the race. The researchers likened this to a phase transition between two states.
Low noise performance
The researchers used a number of methods to identify the location of this phase transition, including numerical estimates of the system’s behavior and experiments using the Sycamore processor. Boixo explained that the transition point is related to the errors per cycle, with each cycle involving performing an operation on all of the qubits involved. So, the total number of qubits being used influences the location of the transition, since more qubits means more operations to perform. But so does the overall error rate on the processor.
If you want to operate in the low noise regime, then you have to limit the number of qubits involved (which has the side effect of making things easier to simulate on classical hardware). The only way to add more qubits is to lower the error rate. While the Sycamore processor itself had a well-understood minimal error rate, Google could artificially increase that error rate and then gradually lower it to explore Sycamore’s behavior at the transition point.
The low noise regime wasn’t error free; each operation still has the potential for error, and qubits will sometimes lose their state even when sitting around doing nothing. But this error rate could be estimated using the cross-entropy benchmark to explore the system’s overall fidelity. That wasn’t the case beyond the transition point, where errors occurred quickly enough that they would interrupt the entanglement process.
When this occurs, the result is often two separate, smaller entangled systems, each of which were subject to the Sycamore chip’s base error rates. The researchers simulated this by creating two distinct clusters of entangled qubits that could be entangled with each other by a single operation, allowing them to turn entanglement on and off at will. They showed that this behavior allowed a classical computer to spoof the overall behavior by breaking the computation up into two manageable chunks.
Ultimately, they used their characterization of the phase transition to identify the maximum number of qubits they could keep in the low noise regime given the Sycamore processor’s base error rate and then performed a million random circuits on them. While this is relatively easy to do on quantum hardware, even assuming that we could build a supercomputer without bandwidth constraints, simulating it would take roughly 10,000 years on an existing supercomputer (the Frontier system). Allowing all of the system’s storage to operate as secondary memory cut the estimate down to 12 years.
What does this tell us?
Boixo emphasized that the value of the work isn’t really based on the value of performing random quantum circuits. Truly random bit strings might be useful in some contexts, but he emphasized that the real benefit here is a better understanding of the noise level that can be tolerated in quantum algorithms more generally. Since this benchmark is designed to make it as easy as possible to outperform classical computations, you would need the best standard computers here to have any hope of beating them to the answer for more complicated problems.
“Before you can do any other application, you need to win on this benchmark,” Boixo said. “If you are not winning on this benchmark, then you’re not winning on any other benchmark. This is the easiest thing for a noisy quantum computer compared to a supercomputer.”
Knowing how to identify this phase transition, he suggested, will also be helpful for anyone trying to run useful computations on today’s processors. “As we define the phase, it opens the possibility for finding applications in that phase on noisy quantum computers, where they will outperform classical computers,” Boixo said.
Implicit in this argument is an indication of why Google has focused on iterating on a single processor design even as many of its competitors have been pushing to increase qubit counts rapidly. If this benchmark indicates that you can’t get all of Sycamore’s qubits involved in the simplest low-noise regime calculation, then it’s not clear whether there’s a lot of value in increasing the qubit count. And the only way to change that is to lower the base error rate of the processor, so that’s where the company’s focus has been.
All of that, however, assumes that you hope to run useful calculations on today’s noisy hardware qubits. The alternative is to use error-corrected logical qubits, which will require major increases in qubit count. But Google has been seeing similar limitations due to Sycamore’s base error rate in tests that used it to host an error-corrected logical qubit, something we hope to return to in future coverage.
John is Ars Technica’s science editor. He has a Bachelor of Arts in Biochemistry from Columbia University, and a Ph.D. in Molecular and Cell Biology from the University of California, Berkeley. When physically separated from his keyboard, he tends to seek out a bicycle, or a scenic location for communing with his hiking boots.
Brazil’s Supreme Court is allowing Elon Musk’s X to resume operations, apparently ending a months-long battle after the social network paid over $5 million in fines and reluctantly agreed to suspend accounts accused of spreading disinformation.
The court yesterday issued a press release announcing the reinstatement, saying that X has complied with all the orders it previously defied. Brazil Supreme Court Judge Alexandre de Moraes ordered that the suspension be ended and that telecom agency Anatel take steps to allow the platform’s return.
The dispute began in April, when X refused to suspend certain accounts belonging to supporters of former President Jair Bolsonaro. X, formerly Twitter, was banned in Brazil for over a month. Internet providers, including Musk’s Starlink service, were ordered to block the social network.
In late August, X claimed the orders violate Brazil’s own laws and said it would defy them even if it meant being shut down. “Unlike other social media and technology platforms, we will not comply in secret with illegal orders. To our users in Brazil and around the world, X remains committed to protecting your freedom of speech,” the company said at the time.
X now accepts “boundaries of the law”
X also said that de Moraes targeted the platform “simply because we would not comply with his illegal orders to censor his political opponents.” Now that it has suspended the accounts, X said it is still fighting for free speech “within the boundaries of the law.”
“X is proud to return to Brazil,” the company’s Global Government Affairs account said yesterday. “Giving tens of millions of Brazilians access to our indispensable platform was paramount throughout this entire process. We will continue to defend freedom of speech, within the boundaries of the law, everywhere we operate.”
In a rare event, Apple has rolled out substantial updates to the web-based iCloud interface meant to allow users to access Apple services like Mail and Photos when they’re away from a Mac, iPad, or iPhone.
The flagship addition is dark mode; it “will automatically match your device settings with a Light Mode or Dark Mode color scheme,” Apple explains as part of the update.
There is also now a way to customize the background for the iCloud web interface—specifically, you can choose between several colors.
A few apps received features that have been available on iOS and macOS for a while. For example, the Notes web app now supports pinned notes, and iCloud Drive supports shared views.
If you think all that seems like it’s pretty basic and late to the game, you’re not wrong.
The iCloud web interface has long seemed like an afterthought for Apple, and it has always been far behind Apple’s native software platforms in terms of features. How far behind? Well, consider this: dark mode was previously added to iOS way back in iOS 13.
Apple’s narrative to investors has long said that its services like iCloud are key to making up for slowed hardware sales in the mature smartphone market. To that end, the company has made this web interface available and has brought some of its services like Music and TV+ to other platforms like Windows and Android.
However, there seem to be limits to that. As noted, iCloud for web has historically been a subpar experience, and other key services like Messages have not been made available on other platforms at all, possibly to avoid losing the social lock-in advantage of Messages for iOS. (Messages is notably absent in the web app.)
Still, it’s nice to see any movement at all here. While iCloud.com gets infrequent and small updates, it remains actively supported at a basic level.
Injunction in Epic case gives rival app stores three years to catch up to Google.
Google Play gift cards in a shop in New York on July 5th, 2024.
A federal judge yesterday ordered Google to open up the Google Play Store and its collection of apps to third-party app stores as part of a US-wide injunction stemming from Epic Games’ antitrust victory over the company. The injunction is scheduled to take effect on November 1, though Google will have up to eight months to implement certain provisions.
For three years, Google will have to let third-party Android app stores access the Google Play Store’s catalog of apps “so that they may offer the Play Store apps to users,” said the injunction issued by US District Judge James Donato of the Northern District of California.
App developers will have some control over which app stores their software is distributed on. “Google will provide developers with a mechanism for opting out of inclusion in catalog access for any particular third-party Android app store,” the injunction said.
Google will be required to allow distribution of third-party Android app stores through the Google Play Store, making it easier for users to install different app stores without sideloading. Donato further prohibited Google from requiring the use of its own billing system for apps distributed on the Google Play Store, including for in-app purchases.
Some provisions relate to deals with phone makers and carriers that may offer devices with preinstalled app stores. “For a period of three years ending on November 1, 2027, Google may not condition a payment, revenue share, or access to any Google product or service, on an agreement with an original equipment manufacturer (OEM) or carrier to preinstall the Google Play Store on any specific location on an Android device,” the injunction said. A similar condition applies to any “agreement with an OEM or carrier not to preinstall an Android app distribution platform or store other than the Google Play Store.”
Judge gives competitors three years to catch up
In an order explaining the injunction, Donato said he limited the requirements to three years “because the provisions are designed to level the playing field for the entry and growth of rivals, without burdening Google excessively. As competition comes into play and the network effects that Google Play unfairly enjoys are abated, Google should not be unduly constrained as a competitor.”
At trial, the jury ruled in Epic’s favor on its Sherman Act claims of monopolization, unlawful restraint of trade, and tying. Donato explained that a remedy in antitrust cases “is not limited simply to prohibiting conduct found to be anticompetitive. Rather, the Court has discretion to fashion a remedy directed to the effect of the anticompetitive conduct.”
Epic was “illegally and unfairly foreclosed from using its own in-app billing services while distributing its Fortnite app through the Google Play Store because of Google’s anticompetitive practices,” and “illegally and unfairly foreclosed from competing in the market for Android in-app billing services for digital goods and services transactions,” Donato wrote.
Donato added that the “harms are ongoing and cannot be made right simply by Google writing Epic a large check.” The injunction is in the public interest because it will help restore “free and unfettered competition,” he wrote. Google is also “enjoined from sharing Play Store revenues with current or potential Android app store rivals, and from imposing contractual terms that condition benefits on promises intended to guarantee Play Store exclusivity.”
Donato’s order said that Google on several occasions “fired a blunderbuss of comments and complaints that are underdeveloped and consequently unhelpful in deciding the issues.” He also rejected some of Epic’s proposals because they would have “threatened a degree of judicial oversight that would amount to micromanagement of Google’s business. It is not for the Court to decide the day-to-day business issues of Android app distribution and in-app billing.”
Google plans appeal
Epic Games CEO Tim Sweeney wrote that the injunction “means all app developers, store makers, carriers, and manufacturers have 3 years to build a vibrant and competitive Android ecosystem with such critical mass that Google can’t stop it.”
Google issued a response saying it will appeal the underlying verdict and “will ask the courts to pause Epic’s requested changes, pending that appeal.”
The court-ordered “changes would put consumers’ privacy and security at risk, make it harder for developers to promote their apps, and reduce competition on devices,” Google VP of Regulatory Affairs Lee-Anne Mulholland wrote. “Ultimately, while these changes presumably satisfy Epic, they will cause a range of unintended consequences that will harm American consumers, developers and device makers.”
Mulholland also said the injunction will “undercut Android’s ability to compete with Apple’s iOS.”
“These Epic-requested changes stem from a decision that is completely contrary to another court’s rejection of similar claims Epic made against Apple—even though, unlike iOS, Android is an open platform that has always allowed for choice and flexibility like multiple app stores and sideloading,” she wrote.
Judge dismisses Google arguments
Donato’s order allows Google to impose security restrictions on third-party apps, but he said that Google must show that any restrictions are necessary.
“As Google has suggested, there are potential security and technical risks involved in making third-party apps available, including rival app stores,” Donato wrote. “The Court is in no position to anticipate what those might be, or how to solve them. Consequently, Google will have room to engage in its normal security and safety processes. To the extent Google imposes requirements along these lines on rival app stores, it will… bear the burden when challenged of establishing that the requirements were strictly necessary to achieve safety and security for users and developers.”
The injunction, Donato wrote, “must not only prohibit the specific anticompetitive conduct that Google engaged in, but also undo the consequence of Google’s ill-gotten gains.” But the requirements, such as the one forcing Google to let third-party app stores access the Google Play Store catalog, have some limits:
The injunction must bridge the moat. Even so, the catalog access provision is narrowly tailored to remediate the unfairly enhanced network effects Google reaped without unfairly penalizing its success as a first mover. To that end, if a rival app store does not have a relationship with a developer and so cannot fulfill a download request by a user, the rival will direct the download request to the Google Play Store. In that case, the Google Play Store will fulfill the download request and keep the associated revenue, if any, and the download will be made pursuant to the Google Play Store’s policies. All that the catalog access does is level the playing field for a discrete period of time so that rival app stores have a fighting chance of getting off the ground despite network effects and the disadvantage of offering a “catalog of app/games” that is too “limited” to attract users and developers in a two-sided market.
Donato is giving Google eight months to implement the technology needed to allow distribution of third-party app stores through Google Play, and eight months to give third-party stores access to the Google Play Store catalog of apps.
“Google will have up to eight months from the date of this order to implement the technology and procedures necessary to comply with this provision, and the three-year time period will start once the technology and procedures are fully functional,” he wrote. A technical committee will oversee the process, “with the Court serving as the final word when necessary.”
Donato’s 17-page order did not address every one of Google’s arguments, because the judge decided some of them were too weak to warrant a response. “As noted, Google’s modus operandi in this case has been to deluge the Court in an ocean of comments, many of which were cursory and undeveloped. The Court declines to take up Google’s objections that were not fully developed in their presentation to the Court,” he wrote.
Jon is a Senior IT Reporter for Ars Technica. He covers the telecom industry, Federal Communications Commission rulemakings, broadband consumer affairs, court cases, and government regulation of the tech industry.
Prosecutors now have a “blueprint” to seize privileged communications, X warned.
Last year, special counsel Jack Smith asked X (formerly Twitter) to hand over Donald Trump’s direct messages from his presidency without telling Trump. Refusing to comply, X spent the past year arguing that the gag order was an unconstitutional prior restraint on X’s speech and an “end-run” around a record law shielding privileged presidential communications.
Under its so-called free speech absolutist owner Elon Musk, X took this fight all the way to the Supreme Court, only for the nation’s highest court to decline to review X’s appeal on Monday.
It’s unclear exactly why SCOTUS rejected X’s appeal, but in a court filing opposing SCOTUS review, Smith told the court that X’s “contentions lack merit and warrant no further review.” And SCOTUS seemingly agreed.
The government had argued that its nondisclosure order was narrowly tailored to serve a compelling interest in stopping Trump from either deleting his DMs or intimidating witnesses engaged in his DMs while he was in office.
At that time, Smith was publicly probing the interference with a peaceful transfer of power after the 2020 presidential election, and courts had agreed that “there were ‘reasonable grounds to believe’ that disclosing the warrant” to Trump “‘would seriously jeopardize the ongoing investigation’ by giving him ‘an opportunity to destroy evidence, change patterns of behavior, [or] notify confederates,” Smith’s court filing said.
Under the Stored Communications Act (SCA), the government can request data and apply for a nondisclosure order gagging any communications provider from tipping off an account holder about search warrants for limited periods deemed appropriate by a court, Smith noted. X was only prohibited from alerting Trump to the search warrant for 180 days, Smith said, and only restricted from discussing the existence of the warrant.
As the government sees it, this reliance on the SCA “does not give unbounded, standardless discretion to government officials or otherwise create a risk of ‘freewheeling censorship,'” like X claims. But the government warned that affirming X’s appeal “would mean that no SCA warrant could be enforced without disclosure to a potential privilege holder, regardless of the dangers to the integrity of the investigation.”
Court finds X alternative to gag order “unpalatable”
X tried to wave a red flag in its SCOTUS petition, warning the court that this was “the first time in American history” that a court “ordered disclosure of presidential communications without notice to the President and without any adjudication of executive privilege.”
The social media company argued that it receives “tens of thousands” of government data requests annually—including “thousands” with nondisclosure orders—and pushes back on any request for privileged information that does not allow users to assert their privileges. Allowing the lower court rulings to stand, X warned SCOTUS, could create a path for government to illegally seize information not just protected by executive privilege, but also by attorney-client, doctor-patient, or journalist-source privileges.
X’s “policy is to notify users about law enforcement requests ‘prior to disclosure of account information’ unless legally ‘prohibited from doing so,'” X argued.
X suggested that rather than seize Trump’s DMs without giving him a chance to assert his executive privilege, the government should have designated a representative capable of weighing and asserting whether some of the data requested was privileged. That’s how the Presidential Records Act (PRA) works, X noted, suggesting that Smith’s team was improperly trying to avoid PRA compliance by invoking SCA instead.
But the US government didn’t have to prove that the less-restrictive alternative X submitted would have compromised its investigation, X said, because the court categorically rejected X’s submission as “unworkable” and “unpalatable.”
According to the court, designating a representative placed a strain on the government to deduce if the representative could be trusted not to disclose the search warrant. But X pointed out that the government had no explanation for why a PRA-designated representative, Steven Engel—a former assistant attorney general for the Office of Legal Counsel who “publicly testified about resisting the former President’s conduct”—”could not be trusted to follow a court order forbidding him from further disclosure.”
“Going forward, the government will never have to prove it could avoid seriously jeopardizing its investigation by disclosing a warrant to only a trusted representative—a common alternative to nondisclosure orders,” X argued.
In a brief supporting X, attorneys for the nonprofit digital rights group the Electronic Frontier Foundation (EFF) wrote that the court was “unduly dismissive of the arguments” X raised and “failed to apply exacting scrutiny, relieving the government of its burden to actually demonstrate, with evidence, that these alternatives would be ineffective.”
Further, X argued that none of the government’s arguments for nondisclosure made sense. Not only was Smith’s investigation announced publicly—allowing Trump ample time to delete his DMs already—but also “there was no risk of destruction of the requested records because Twitter had preserved them.” On top of that, during the court battle, the government eventually admitted that one rationale for the nondisclosure order—that Trump posed a supposed “flight risk” if the search warrant was known—”was implausible because the former President already had announced his re-election run.”
X unsuccessfully pushed SCOTUS to take on the Trump case as an “ideal” and rare opportunity to publicly decide when nondisclosure orders cross the line when seeking to seize potentially privileged information on social media.
In its petition for SCOTUS review, X pointed out that every social media or communications platform is bombarded with government data requests that only the platforms can challenge. That leaves it up to platforms to figure out when data requests are problematic, which they frequently are, as “the government often agrees to modify or vacate them in informal negotiations,” X argued.
But when the government refuses to negotiate, as in the Trump case, platforms have to decide if litigation is worth it, risking sanctions if the court finds the platform in contempt, just as X was sanctioned $350,000 in the Trump case. If a less restrictive alternative was determined appropriate by the courts, such as appointing a trusted representative, platforms would never have had to guess when data requests threaten to expose their users’ privileged information, X argued.
According to X, another case like this won’t come around for decades, where court filings wouldn’t have to be redacted and a ruling wouldn’t have to happen behind closed doors.
But the government seemingly persuaded the Supreme Court to decline to review the case, partly by arguing that X’s challenge to its nondisclosure order was moot. Responding to X’s objections, the government had eventually agreed to modify the nondisclosure order to disclose the warrant to Trump, so long as the name of the case agent assigned to the investigation was redacted. So X’s appeal is really over nothing, the government suggested.
Additionally, the government argued that “this case would not be an appropriate vehicle” for SCOTUS’ review of the question X raised because “no executive privilege issue actually existed in this case.”
“If review of the underlying legal issues were ever warranted, the Court should await a live case in which the issues are concretely presented,” Smith’s court filing said.
X is likely deflated by SCOTUS’ call declining to review X’s appeal. In its petition, X claimed that the court system risked providing “a blueprint for prosecutors who wish to obtain potentially privileged materials” and “this end-run will not be limited to federal prosecutors,” X warned. State prosecutors will likely also be emboldened to do the same now that the precedent has been set, X predicted.
In their brief supporting X, EFF lawyers noted that the government already has “far too much authority to shield its activities from public scrutiny.” By failing to prevent nondisclosure orders from restraining speech, the court system risks making it harder to “meaningfully test these gag orders in court,” EFF warned.
“Even a meritless gag order that is ultimately voided by a court causes great harm while it is in effect,” EFF’s lawyers said, while disclosure “ensures that individuals whose information is searched have an opportunity to defend their privacy from unwarranted and unlawful government intrusions.”
Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience.
