Hi friends,
The Weekend Reccs are back after a week off, with big news, as promised.
After a fourth of a year over email, I’ve moved the newsletter to Substack. Thanks for being with me on this journey and I hope you’ll appreciate some of the new functionality this will enable.
Train-ing the eye: A beautiful collection from, and short profile of, a photographer of the MTA.
The “Dirt” on horse-racing: This is a really interesting writeup about what goes in to calibrating the track horses race on top. The author fails to pause on the irony of perhaps the most interesting part of the story though: despite all this science the track superintendent is superstitious. It is a funny thing to consider the marginal impact of these improvements. If you think your snack choice really changes the fate of one of these beautiful horses, does getting better technology to manage that risk give you more leeway on the snacks? A slightly better sensor means flamin’ hot cheetos are ok even though crunchy is your go-to? I’d suspect the answer is no, the snacks have to always be the same because a superstition is a superstition. But if that’s what has causal weight here, are you going through all this dirt testing for show without believing in it?
Top flight: Relatedly, this is a great video on the issues that arose from golf ball technology getting too good. There’s an interesting set of questions here around the role of technologies and governance in sports that both of these cases bring up. The horse one is easy: we want safer horse racing, and the competitiveness of the race is not really changed as all the horses are slower or faster but the relative thrill is still there. The golfing one is more weird: we want players to be as good as possible, but want to also put them into difficult enough situations that this their good playing is impressive and interesting to watch. If everyone hit a hole-in-one on every hole we’d get quickly bored. We similarly want to make the inter-player struggle more competitive and dynamic to keep it interesting, but not too dynamic, as we want those with the best showing on most days to win (but not every day, as that would be too deterministic). At some level we’re maximizing for thrill while trying to hold a certain level of legitimacy high enough.
A return you can’t refuse: It feels a little like cheating to send another newsletter as a link on my newsletter, but this breakdown [$] of mafia bonds by Matt Levine was just too good not to share. For those that are pay-walled out, the short of it is that front businesses associated with an Italian mafia family had their (sometime illicit) cashflows packaged into a bond which was then sold to investors. Levine has a great tongue-in-cheek riff about the future of such activities, quantitative traders and all. Great commentary aside, I think we also should think about this through a lens of the financialization of everything as a broader trend. A lot of people decry this trend for bringing about various, usually vague, social ills, but that’s to give a generally good process a bad rap. Financialization ultimately is about taking uncertainty and shifting it around, which reduces risk to any single entity. This can be an issue when the tools are misunderstood (see the XIV crash in 2018) or the uncertainty itself is misunderstood. But these are bugs, not features. The same could be said about capitalism. At the end of the day, it is a social technology for distributed information generation and consolidation. These information flows are hampered by quarterly statements causing myopia and by Amazon’s massive market power. As such it makes little sense to point to these as ills of capitalism, but rather as ills against capitalism. They are failures of our current instantiation of it. Bringing it back to the mafia bonds, we might find it ludicrous or socially objectionable that dons can secure for themselves a less volatile income stream by offsetting some of their risks through selling bonds. But that’s not really the fault of the process of lessening income volatility. It is an issue we have with their income streams in the first place. (s/o Benedict Brady for introducing me to Levine’s newsletter)
Where in the World is Carmen Sandiego: As a part of my job I’m learning some spatial analytics. It turns out we don’t know a lot about how to refer to places in the world. Latitude and longitude are not subject to some universal rule as to where exactly they mean. They largely depend on the model of the globe you are using, and, when doing spatial analytics, the representation used to make that globe 2-dimensional. In that same spirit the height that we believe things to be is contentious. Currently a process of “height modernization” is underway to determine how tall things are. The wrinkle — what direction is down?
Out of this world: It turns out we have backup disks that can last 14,000,000,000 years. This is a cool project, but also brings about a really strange phenomenon I’ve never considered explicitly before: all data only exists as physical manifestations. For example, when I think of this, it is the case that it exists on some server somewhere in this form, but it also exists in my thoughts. But my thoughts, ultimately, are a bio-chemical copy of this original version on the server, and in theory could recreate a (bad fidelity) copy if my thoughts were machine-readable. So it is the same when I chop down a tree or drink a cup of coffee, I make a physical impression on the readable data of Earth to archeologists of the future that through complex processes (perhaps involving carbon dating) would be able to record the action as having had occurred.
Critic’s corner: Came across some sloppiness from the Economist. Recessions may substantially reduce happiness, but this is most likely a function of expectations, rather than of money itself. All the difficult psychology of happiness aside, it seems common sense that if I had $70,000 a year and now I have, due to a sudden shock, dropped to $40,000 a year I will be more unhappy than if I always had $40,000. This could be due to liabilities I accrued (rent, whatever) or just that I had developed more expensive tastes and habits of daily life. That I am forced to confront these challenges suddenly probably hurts worse than the marginal effect of losing them by themselves.
Your Weekly Recommendation: I just finished Evicted by Matthew Desmond on audiobook. It was incredibly good. The only critique I can offer is that the analysis at the end left something to be desired. Desmond did not defend many of the points that I, as a college-educated liberal, believe to be true on their face but still want proven to me with data.
I previously wrote about the drip of negative impacts from being barred from homeownership due to a poor credit history. Evicted captures many of these in the struggles of low-income renters. Perhaps the one I found most interesting is below. I’ll be the first to admit I don’t know anything about housing policy or sociology, so maybe this is old hat to most folks. But this paragraph, and its implications, blew my mind:
A single eviction could destabilize multiple city blocks, not only the block from which a family was evicted but also the block to which it begrudgingly relocated. In this way, displacement contributed directly to what [Jane] Jacobs called ‘perpetual slums,’ churning environments with high rates of turnover and even higher rates of resentment and disinvestment. ‘The key link in a perpetual slum is that too many people move out of it too fast — and in the meantime dream of getting out.’ With Doreen’s eviction, Thirty-Second Street lost a steadying presence — someone who loved and invested in the neighborhood, who contributed to making the block safer — but Wright Street didn’t gain one.
Graphs of the week:
[G Elliott Morris of the Economist] A case for trusting the state-wide polls I posted: hyperpolarization decreases volatility, or, as Nathaniel Rakich put it: everything is partisan and correlated and boring.
[WSJ] Graphs like this one make a clear point about the disparate racial outcomes before and during the pandemic. These differences are normatively bad and should be discussed. However, we cannot read too heavily into the right side as far as causes go, as averages obscure heterogeneity. While we can say for sure that Black and Hispanic workers are doing worse in the market, this does not tell us whether that is due to White workers edging them out of roles they previously held or due to differences in the racial composition of sectors and their pace of each sector’s recovery. A great way of getting at this question would be to look at race and ethnicity within sectors. That is: what is the jobless rate among previous leisure and hospitality workers, by race/ethnicity?
As ever, feel free to share this with whoever you may find it worth a read. The move to Substack should hopefully make that easier than ever!
Your friend,
Harrison
you make me a smarter human