[link back to the original links post: here]

On the article about privateers, local naval expert Bean writes:

It’s time for the standard disclaimer any time Proceedings comes up: Proceedings is intended as a forum for discussion of matters of interest to naval officers, and it is not peer reviewed. Often very not peer reviewed. Like in this case. Please don’t judge the USNI on the basis of this stuff. They do a lot of good work.

And yes, it is that stupid. First, privateering is probably illegal today. The US didn’t sign the 1856 Paris declaration outlawing it, but the ban is almost certainly considered customary international law today, and thus binding on the US, too. (International law is very weird.)

Second, it makes no sense. It was something that people did in an era when the ability of the state to do things was sharply constrained, and it was never all that profitable. These days, the government is a lot more effective, and if it wants to hunt Chinese commerce (never mind the issues about who owns the cargo, which is rather different in the days of worldwide communications and the shipping container) it will make auxiliary commerce raiders of its own. There’s definitely no need to have a DDG sit outside a Brazilian port waiting. Take any reasonable civilian ship (big yacht, fishing boat, tug, whatever) and fit it with a couple of 40mm guns and a boarding party. Have it do the waiting instead.

And our other defense expert, John Schilling, writes:

Modern naval weapons are too good at sinking ships, whereas privateering requires capturing ships intact to be profitable. For a trivial, and in this context uncontroversial, investment, China can equip their merchant fleet with defensive weapons that will sink any privateer, unless the privateer sinks them first.

- Privateers, being incapable of surviving a fight with real warships (especially modern ones), need to be able to hide from and if necessary outrun enemy warships. That’s a lot harder to manage in a world of radio, radar, maritime patrol aircraft, and satellites. Harder still if you insist on taking prizes, which will be Lojacked beyond your ability to clear at sea. Even in a hot war with the United State, China will probably be able to spare e.g. an H-6K for a day to sink the privateer that just sank one of China’s freighters, and that’s all it will take.

- The rest of the world regards privateering as flat-out illegal, so virtually all of the ports of the world will be closed to the privateers and their prizes. Operating in the South China Sea directly from Hawaii, without any intermediate bases (what’s left of Guam will have its hands full), is going to be logistically challenging to say the least. And the value of that prize ship you just took is greatly diminished if it can only be used in the US coastal trade, its cargo sold only on the US domestic market never to be reexported.

This is a stupid idea that keeps coming back every year or two because somebody read too many Napoleonic sea-adventure stories and thinks they’re the only one who read those stories so their clever “obscure” idea is something the rest of us haven’t heard and rejected a dozen times already.

There’s a problem in medicine where people think doctors are trustworthy experts. While this is often true, there are about a million doctors, and some tiny fraction of them are insane. The reasonable doctors mostly keep their mouths shut, but sometimes an insane doctor will endorse some sort of terrible alternative medicine, and then people will get excited: “A doctor endorsed it! It must be real!” The fact is, you can find doctors saying pretty much any bizarre thing - I hear some of them even have Substacks.

My thought when reading that article was “this sounds crazy…but wait! It’s written by a colonel and published by the US Naval Institute! That sounds just wacky enough to make a good link!”

Now I am concerned that colonels work the same way as doctors. I wonder what else is like this.

On the Kossin paper showing worsening hurricanes, Trevor Klee writes:

I did a college research paper on the changes in hurricanes over time with global warming. What I found was that the data sets are really bad. They’re mostly based on tidal gauge measures of storm surges, but tidal gauges malfunction in massive waves. The paper Scott linked is based on satellite imagery, which seems like it should be better, but I wonder how they backtested their “Dvorak” technique to estimate storm intensity from satellite imagery. If they had to test it on the old data sets (i.e. the tidal gauges), I’d worry about the reliability of the Dvorak technique as well.

And Hari Seldon adds:

Worth noting as well that NOAA has been adjusting the data from previous pre-satellite era hurricane seasons, often by as much as 20%; the adjustments might be entirely called for, but if so that indicates that the historical data being adjusted can’t be trusted to within 20%, making long-term calibration of historical data to satellite data even more difficult. 8% per decade is a very small signal to try to extract from this noise.

Walter Sobchak links this article by Roger Pielke, who argues that hurricanes have been getting slightly stronger if you start in 1970, but slightly weaker if you start in 1945. But I don’t know how good the older data are, plus he switched from Atlantic to Atlantic+Pacific which feels suspicious for data hacking. He also claims that “ the Intergovernmental Panel on Climate Change has concluded that it is premature to assert the detection of trends in hurricanes resulting from human-caused climate change”, but the link just goes to IPCC homepage and I can’t see what exactly they say. The article also contains the fascinating tidbit that “historically, about 60% of all economic damage from disasters worldwide comes from landfalling hurricanes in the United States”. [EDIT: Commenters note this is false; the original source specifies “weather-related”]

Regarding the post on Harvard’s budget, Pete writes:

I wish I could find it again, but I have a facebook friend with insider-type knowledge of how these public budgets are created, and they basically said that since money is fungible, universities can and do make the budgets look like anything they want at that level of abstraction. And, they claim, this is why the bigger budgets are always sort of milquetoast, across the board rises. As opposed to them being actually milquetoast across the board rises. My facebook friend didn’t say what they thought the money was actually being allocated to, and since I can’t cite my source this is more like a trailhead or invitation for someone to look more deeply.

When I asked for clarification, Warren explained some accounting tricks, for example this one used by software companies:

The first thing to keep in mind is that an income statement (like Harvard’s) isn’t a measure of assets or inventory, but a measure of the activity in some particular period (e.g. 2020).

On that statement, there will be some lines for revenue (value of cars sold, tuition) and also some lines for expenses. Some of those expenses are direct inputs into the production process (e.g. the cost of car parts). Some of those costs are for labor (salaries, benefits). Some of those costs are for other things, like rent expenses or IT.

Generally, you’re not allowed to move expenses between sections. If a company could pretend salaries were actually food expenses, or car parts costs were actually rent, financial statements would be useless!

However, there are some weird cases where this is allowed. Many of these cases are because of a fundamental principle of accounting: the matching principle. We want to associate revenue with the costs of that revenue whenever possible. So if the company buys a building, it doesn’t want to suddenly have the whole price of the purchase sit on one month’s income statement. After all, the building will be used for many years.

So instead, what happens is that the company takes the whole cost of the building and spreads it out over some period of time based on the expected useful life of that asset. For computers, it can be as short as three years, and for actual buildings, it can be as long as thirty years.

Instead of having one big expense for “buildings” or “machines”, you instead have a small expense each month in “depreciation”. This makes it so that your financial statements are more representative of the actual profit and loss for each period, but also makes them less informative.

The story of capitalized software is that some accountants made a convincing-enough argument (to auditors) that if a company is making big software projects, it should be allowed to treat some kinds of salaries in the same way as old-school physical assets. Hence, instead of having a big expense one year in “salary”, there is a smaller expense for several years in “depreciation”.

Long story short, you generally cannot obfuscate what the expense categories mean, except in certain special cases. Most of the fudging comes from having expense categories cover enough kinds of spend that it all kind of blurs together.

shambibble on Sanders and class-first leftism:

Any sold-out-on-wokeness hypothesis for why the Sanders campaign failed has to grapple with several issues:

- Why New Bernie did so much better in Nevada in 2020 (with a much more diverse base). The article claims that caucuses “place a premium on ground-level organizing, where Sanders excelled.” But this is not a distinction since Nevada was also a caucus in 2016 and the failures in 2020 South Carolina are being alleged at that same level, not that Bernie ran bad television ads or failed to show up.

- Why Classic Bernie did almost as badly in South Carolina in 2016. He got 26% of the vote there in 2016 and 20% in 2020. This is not a very dramatic difference considering the first contest was one-on-one against Hillary and it can be parsimoniously explained by an “anti-Hillary but not especially leftist” cohort.

- The specifics of Sanders’ supposed sellout. What issues, specifically, does Bernie stand accused of shifting on? As near as I can tell, the only one on offer is that Bernie seems to have agreed with most Democrats that there was some sort of corruption involving Trump and Russia. Setting aside for the moment the various mottes and baileys imposingly shadowed here, I simply do not see a case made that any significant mass of (Democratic primary) voters were specifically motivated by an absence of this. And while Bernie was not a contrarian on this issue, neither did he especially lean into it. You can point to statements he made while being interviewed on news shows, but certainly it came up much less in his stump speeches than class issues.

As someone who followed his primary campaigns closely in 2016 and 2020 and supported him in both, I can’t buy into this thesis. My view is much simpler: Bernie never had a majority of the primary electorate. In 2016 his numbers were puffed up by non-ideological anti-Hillary votes and in 2020 Democratic candidates simply coordinated to stop Bernie in exactly the way Republican candidates failed to do with Trump. The article skewers a straw man by contrapositing its arguments from “if the Democratic party was so desperate to rally around Biden….” They were not, any more than Republicans were eager to rally around Ted Cruz! The difference is they sucked up and did it when the circumstances left it as the only option. This is not corruption, A endorsing B to put them over C is basic coalition politics as they have been practiced in every Presidential nominating process since the introduction of the party system.

And while I know people here are loath not to take others at their word, I would also like to suggest a distinction between class-first leftists and “class-first leftists.” Class-first leftists spend most of their time talking about class, are usually found in obscure academic journals and activist groups, and often have wonkish opinions about monetary policy and the labor theory of value. “Class-first leftists” spend most of their time talking about identity politics, can be found on Twitter, Fox News, or erstwhile pro-Trump outfits like American Affairs, and have often have wonkish opinions about how Republicans are correct to say that idpol is bad and Russiagate is fake, prefaced with “as a class first leftist,”

Nobody commented on this, but I noticed it going back over the post: it’s not entirely correct to describe the Osmunden study as “failing to replicate” the finding that conservatives have stronger threat responses than liberals. The study finds that, using a kind of objective measure of threat response called “electrodermal activity”, American, but not Danish, conservatives have stronger threat responses than their respective liberals (but what’s there to feel threatened by in Denmark anyway?) Using subjectively-reported threat response, both American and Danish conservatives do. The study does find this field is complicated and hard to measure, but so is every field, and overall these results seem at least somewhat positive. But there are apparently other replications that really do fail to find the effect, and it’s probably fair to consider it suspicious.

Also, tkling, on the study showing no relationship between politics and personality:

On item #21–I checked the methods section, and the scope of the paper is much narrower than the abstract makes it out to be. The study’s survey of “political attitudes” only includes social and religious issues (such as abortion and evolution). There’s nothing about economic policy, foreign policy, immigration, the environment, or any of the other things people do politics about.

Neal Zupancic on police reform:

[from original post] “I’m interpreting this to mean that there are effective ways to reform the police, but that the atmosphere created by media saturation and protests produces ineffective counterproductive reform instead.”

The authors seem to suggest it is mostly the investigations themselves causing the increase in crime, rather than any particular policy changes. The mechanism they propose is that police officers greatly reduce their quantity of policing when under federal investigation after a “viral” incident, but there is little indication that this comes about as the result of any particular policy reform - the suggestion is that police are either reducing public contact in an effort to avoid having their own actions scrutinized, or are trying to make a point (in the case of deliberate strikes and slowdowns/sickouts). There’s also a section (page 27) where the authors talk about the possible impact of increased paperwork, and estimate it might account for about 20% of the reduction in police activity in one city. I’m not sure if we’re calling this “reform” but even if we do it’s a small proposed effect.

Alex Passos, on the Montreal police strike:

Local police departments in Brazil go on strike all the time (I remember 3 different ones while I was in school in Salvador in the early aughts) and no big terrible things happen.

And Big Worker:

Seems misleading to bring up the Montreal “Night of Terror” without the context that it involved the police going on strike when the city was already tearing itself apart over Quebec separatism and general late 1960s craziness. The way it reads currently it sounds like the police went on strike and it prompted a riot, rather than the police going on strike because they were tired of dealing with riots.