Portable panopticon

Surveillance on four wheels / Doomberg and Graeber on overemployment / Sam Bankman-Fried / Not all GDP revisions are created equal (12 minute read)

Today we cover the powerful surveillance capabilities of the modern automobile, jobs that do not seem necessary, an update on rogue du jour SBF and take a quick look at GDP revisions. Enjoy!

Surveillance on four wheels

Mozilla, the non-profit organisation which makes one of the leading web browsers, came up with the programming language Rust, and is primarily funded by Google, has released a report which shows that, at least in the US, modern cars and privacy are not natural bedfellows.

From the Mozilla report:

Every car brand we looked at collects more personal data than necessary and uses that information for a reason other than to operate your vehicle and manage their relationship with you. (1)

The best were Renault and Dacia, and the worst was Tesla, which achieved a privacy-abusing mark in every category.

A 2023 rewrite of Thelma & Louise would have the ladies in custody before you’ve had a chance to make a dent in your popcorn. (ibid.)

"No officer; now you mention it, I'm not familiar with the latest advances in telemetry"

Cars are now computers on wheels — which means they "have an unmatched power to watch, listen and collect information about what you do and where you go," per the report. That information is then shared with or sold to data brokers, law enforcement and others. (2)

Out of 25 car brands studied by the Mozilla Foundation: 56% will share data with law enforcement in response to an informal request. 84% share or sell personal data. (ibid.)

"Kitt, what do you mean you've just reported me(!)"

To pick a 'privacy policy' at random, from Nissan:

The company can share "sensitive personal information, including driver's license number, national or state identification number, citizenship status, immigration status, race, national origin, religious or philosophical beliefs, sexual orientation, sexual activity, precise geolocation, health diagnosis data, and genetic information." (ibid.)

Moving from Turing-complete on wheels to Turing-complete in your pocket: the bods at growth.design have a case study of the 'privacy policy' of the new-ish Meta Threads app, Zuck's attempt to disintermediate Elon Musk Twitter. In this case, as with so many others, 'privacy policy' is an oxymoron in the same league as 'friendly fire' or 'deafening silence'. (3)

Doomberg and Graeber on overemployment

Unemployment in the US is around 3.8%, it is around 4.3% in the UK, and 6.4% throughout the EU as a whole, with the latter being a record low, according in Eurostat, despite being the highest of the three. (4)

This is unhelpful to central bankers who are interested in bringing inflation down because people without jobs consume fewer goods and services and by definition, they are not going to contribute to the wage-price spiral ('second order effects') that e.g. Huw Pill at the Bank of England has been commenting on in recent months.

Which begs the question, what is the optimal level of unemployment? During the strongly recessionary period of the 1930's John Maynard Keynes famously used the the idea of paying workers to dig holes and then fill them back up as a way to stimulate economic activity. He used this example to suggest that even paying people for unproductive work could be better than allowing high levels of unemployment to persist, since those employed workers would spend their wages, thus stimulating demand and driving further economic activity.

Keynes lived before the advent of 'the laptop class' / 'email jobs' phenomenon, although it does seem that some commentators are suggesting that during a period of what looks and feels like stagflation, some holes are being metaphorically dug.

From Doomberg:

[...] Christopher Joye, Chief Investment Officer and Portfolio Manager of Coolabah Capital, makes an underappreciated observation about the inflationary effects of the modern corporate organizational chart. Over the past few decades, an explosion of “mandatory” jobs that contribute little to the creation of tangible products or services—in other words, to productivity—have invaded large companies like a cancer. Joye points to the big banks, which have seen a massive increase in “compliance departments, risk departments, legal departments, government affairs departments, ESG departments, [and] human resources departments” leading to the inevitable result that “we are producing similar products with many more people”. (5)

The late anthropologist and activist David Graeber published a book with the title Bullshit Jobs: A Theory where he he delved into this phenomena further. Graeber defines a "bullshit job" as a form of paid employment that is so pointless, unnecessary, or pernicious that even the employee cannot justify its existence, although they may feel obliged to pretend otherwise. These are distinct from "shit jobs" which might be hard and underpaid but are often necessary to society e.g. sanitation work. He taxonomises some of the ways in which jobs might not be necessary: jobs that exist only to make someone else look or feel important; jobs that exist to fix a problem that should not exist in the first place; jobs that exist only to allow an organisation to claim it is doing something it is not actually doing.

Perhaps we should not be surprised about the proliferation of meaningless jobs, in part because of the cultural belief that work is of value in and of itself, and in part due to structures of economic decision-making that promote the creation of unnecessary roles. If the value of some tradable asset—a painting, bond, car, carrot—is what someone is willing to pay when they buy it from you, the 'value' of a job can be seen to be what someone will pay you to do it, no matter how pointless it is. Holders of equity should be a break on this kind of activity in theory, but in the private sector the principal-agent problem (that managers manage for managers and not optimally from the perspective of the owner) is well known and in the public sector the government is, to a greater or lesser degree, remote from its main source of income, which in the developed West is the taxation of citizens and resident businesses.

In another of his books, Utopia of Rules: On Technology, Stupidity, and the Secret Joys of Bureaucracy, Graeber touches on a deep paradox within Western developed countries; most people claim to dislike bureaucracy while doing nothing to oppose it. Bureaucracy pervades every aspect of our lives, and represents a form of power that is impersonal and increasingly rooted in technology, something which many technologists prefer to gloss over. The job of the bureaucrat is to decrease entropy, not increase output. This is an optimisation problem like any other; it should be possible to get the level of bureaucracy just right, but there is an asymmetry because it is much easier to build on the edifice that already exists than hew chunks off, which is why so many conglomerates seem to be involved in semi-permanent restructuring and why Schumpeter tells us that recessions are not all bad.

Once spawned, none of these positions can really ever be eliminated, and it necessarily follows that the creation of the same increment of GDP now requires a far higher net wage expenditure. A secondary effect of so many unproductive roles is the dwindling number of qualified candidates to fill the productive ones. These observations form the basis of Joye’s view that we should expect inflation to be a tad stickier than most are modelling. (ibid.)

We would extend Joye’s thinking by pointing out another troubling trend in Western labour markets: that the definition of “doing work” has been blurred with “attending meetings.” As a general rule, meetings are a giant waste of time, and yet nothing makes an employee occupying an unproductive job feel more productive than being invited to one. A packed calendar is the sign of an important person, after all, and what better way to kill time until you can go home and attend your evening meetings with the team in Asia? (ibid.)

In our view, there are only two reasons to have a meeting: to make a decision or to inform others of a recent one. Neither embodiment should take more than 15 minutes [...] These uncontroversial statements are intuitive and universal, making the collective and continued participation in meetings all the more puzzling. (ibid.)

Sam Bankman-Fried

Even as a lobster with recourse to only around 10^5 neurons, it seemed clear to me at the time of FTX's collapse that Sam Fried My Bank Man was a wrong 'un. (6)

Thanks to his trial, we now know that with a good degree of certainty.

There are two things that the US legal system gets right in a way that slackens the jaws of Europeans; very large fines for big business that finds itself on the naughty step and getting former accessories to criminal activity to rat others out -- $104 million for helping to explain how UBS advised its clients to avoid tax by putting diamonds in toothpaste tubes, for example. (7)

And so it wasn't that much of a surprise that Caroline Ellison, SBF's ex-girlfriend and the person nominally in charge of Alameda Research, a related entity to FTX, became a star witness.

“He directed me to commit crimes,” Ellison told the jury. “He was the one who directed us to take customer money and repay our loans.” She previously pleaded guilty to fraud and agreed to co-operate with prosecutors. (8)

“Alameda took several billion dollars ... from FTX customers” (ibid.)

When it came to his political donations, Ellison testified that Bankman-Fried “said he thought it was very effective ... that he could get very high returns in terms of influence”, adding that the defendant had contributed $10 million to Joe Biden’s presidential campaign. (ibid.)

A nice case of guilt-free schadenfreude if there ever was one; if you wanted more, Molly White, a sensible person writing about crypto, tells us about the run up to his trial:

Sam Bankman-Fried, a vegan, has been subsisting on a diet of bread and water, with the occasional smear of peanut butter. His prescribed medications for major depressive disorder and ADHD are running low, and prison officials have been hesitant to replenish his supply. He's not allowed to have a laptop in his cell, and in the periods he's allowed to meet with his attorneys to review documents for his defence, Internet access is slow and spotty, and the battery on the laptop runs out within a few hours. Hard drives of evidence sent from his attorneys for him to review in the jail's offline computer room have been slow to arrive. (9)

Bankman-Fried's attorneys first asked a judge to intervene to force prison officials to be more attentive to his dietary and medical needs, and are now seeking to have him released and returned to house arrest where he would once again have 24/7 access to his legal team, a computer, and a high-speed Internet connection in order to prepare a defence for the complex and document-intensive trial looming in October [...] From December until mid August, Bankman-Fried lived in relative comfort under house arrest at his parents' Palo Alto home. At first, his Internet access wasn't curtailed, and he blogged, tweeted, and played video games in his downtime. Within weeks, however, Bankman-Fried had begun to push the boundaries of his conditions of release, and as a result of two separate incidents, found himself repeatedly in front of a judge who starkly warned him about the possibility of remand as his release conditions were tightened. It wasn't until the third incident, when the judge determined that there was probable cause that Bankman-Fried had attempted to tamper with witnesses at least twice, that he was sent back to jail. His lack of computer access is, largely, a problem of his own making. (ibid.)

Relatedly, Michael Lewis, one of the most recognisable American authors and certainly one of its best non-fiction writers, has been getting a bit of heat for not being critical enough of SBF in his new book Going Infinite: The Rise and Fall of a New Tycoon which would perhaps be better titled Going Infinite: The Greatest Financial Fraudster in History. Here is an unusually direct Noah Smith with what seems like a plausible theory as to why this could be the case:

I don’t have enough information to say Lewis is wrong here, but I expect when all is said and done, there will be solid evidence that SBF knowingly stole from his customers. In fact, I kind of doubt the standard story that SBF gave the money to his hedge fund and gambled it all away; I suspect he just embezzled most of it for himself, and still has most of it stashed in various places. (10)

So my bet is that Michael Lewis is just being way way too nice to SBF here. But why? Lewis didn’t hold back when criticizing the bond traders at Salomon Bros. or the creators of mortgage-backed CDOs; why would he lob such a softball at a guy whose finance empire was infinitely more scammy than those others? (ibid.)

Maybe Lewis did get too close to his subject, or maybe the magic spell of cynical “effective altruism” that SBF and his cronies used to get on so many people’s good side. But I suspect that the deeper thing that’s going on here is that SBF, like so many crypto barons, presented himself as an anti-financier — a good guy taking finance back from the cynical Wall Street banksters who hustled everyone in the 80s and 90s and blew up the world in the 00s. In other words, SBF set himself up to be the antithesis of everything that Michael Lewis spent his early career bashing, when in fact he was just a far scammier version of the same. (ibid.)

Not all GDP revisions are created equal

Because GDP is difficult to calculate the latest data are estimates which are then revised for accuracy as they age; a fresh GDP read-out is provisional and subject to change. John Burn-Murdoch, the FT's dataviz guru, has a good piece on GDP revisions and how accurate GDP estimates are in general.

Ireland's margins of error suggest that the ghost of 'Student' left Dublin's Guinness Brewery a long time ago ...

On average, five years after an estimate of quarterly Irish GDP growth is first published, the latest revision of that figure is two full percentage points off the original value. The equivalent for the UK is almost 10 times smaller at 0.25 percentage points, making the ONS’s initial estimates among the most accurate in the developed world, narrowly ahead of the US at 0.26 and well ahead of the likes of Japan (0.46) and Norway (0.56). (11)

... but then Ireland's beggar-thy-neighbour tax policies make it quite hard to measure actual economic activity there, so the Irish Central Statistics office does have a hard job on its hands. (12)

Amusingly the US revises its GDP down and the UK revises its GDP up. It seems a bit strange that they don't include an extra term in their models to offset this by including some sort of bias based on the error in previous estimates, but maybe they leave that to consumers of their data instead.

Out of 24 developed countries that consistently report quarterly GDP revisions to the OECD, the UK’s initial estimates are the most pessimistic. Britain’s quarterly growth figures typically end up 0.15 percentage points higher than first thought. The Germans go up by 0.07 on average, the French by 0.04, while the Americans, ever optimistic, typically end up revising their estimates down by 0.11 percentage points. (13)

That’s all for this time folks. If the mood strikes you, please leave a comment, and please do share this with anyone else who may find the contents to be of interest.