Links – 10/19/2017

  • I recently came across this interesting blog post, in which the author describes his experience dealing with state insurance regulators in the US. He points out that regulators are often ignorant of even the most basic statistical techniques used in the industry. This is a good post to get a sense of what regulation often looks like in practice.
  • There has been a lot of discussion recently about surveys based on online non-probability samples. Pew Research Center recently published a study that compared a bunch of surveys of that sort with a survey based on a probability sample and a variety of benchmarks. It turns out that online non-probability samples are very heterogeneous, but that some of them do pretty good. This is also relevant to my post about antisemitic violence in Europe, where I relied on a non-probability online sample, so I added a note at the end.
  • Pseudoerasmus recently published a very detailed post, in which he tries to explain why Japan’s textile had become the world’s dominant exporter by the 1930’s while Indian’s textile industry stagnated, even though they had similar levels of productivity at the beginning of the 20th century. He argues that the main cause of this divergence is that workers had much more bargaining power in India than in Japan, so that they were able to successfully demand high wages relative to effort, whereas Japanese capitalists were able to squeeze more labor out of their workers relative to their wages and, in particular, to make them operate more machines. Thus, the difference was not that India’s textile industry lacked the technological innovations that Japan’s textile industry relied on to sustain its growth during that period, but that Indian capitalists were prevented from increasing the share of capital in production by workers, who frequently struck to maintain high employment and low effort, even if that meant lower wages. He also examines how culture and the state explains why workers had more bargaining power in India than in Japan. This post is really economic history at its best. Not only does it shed light on a very important historical phenomenon (the divergence between India and Japan during this period), but it also shows how factors that are often ignored by economic models (namely labor relations and the cultural/political context which affects them) play a very important role.
  • Andrew Sullivan recently published a very good essay on the rise of tribalism in New York Magazine. I don’t agree with everything he says, but if this were not the case, it would probably mean that he didn’t do a good job. I want to discuss this topic at some point, though I don’t know when.
  • I recently came across this critical post about Cordelia Fine’s latest book, which I already mentioned several times here. I think it’s very good and encourage you to read it. If you liked the post, you can also follow the author on Twitter, where he is pretty active.

2 thoughts

  1. Did you see this Steve Sailer post?

    relevant excerpt:

    “Any false reports of bias incidents are seized on by those who want to create the impression that no hate crime reporting is legitimate,” said Ibrahim Hooper, the spokesman for the Council on American-Islamic Relations. “We have seen an unprecedented spike in the number of hate- and bias-related incidents targeting American Muslims and others across the board.”

    Perhaps. Or perhaps not.

    Let’s think about the relevant question: What percentage of nationally-publicized hate incidents turn out to be hate hoaxes?

    Of the non-local stories that make, say, the New York Times, the Washington Post, CNN, CBS, NBC, ABC, NPR, or The Guardian:

    – What percentage turn out to be confirmed true?

    – What percentage go down in flames as hate hoaxes?

    – What percentage turn out to be legal (e.g., the two Babson College dudes who were the subject of multiple NYT stories for driving around the Wellesley Campus cheering for Donald Trump, which turned out, remarkably enough, to still be legal in Wellesley, MA)?

    – And what percentage get left in credibility limbo?

    Keep in mind that a lot of Hate Hoaxes go down the memory hole. For example, in 2012 the New York Times made a really big deal out of a police blotter item from 3,000 miles away:

    Iraqi Immigrants in California Town Fear a Hate Crime in a Woman’s Killing

    That proved to be a complete Hate Hoax: the Iraqi murderer tried to blame white people for murdering his wife. But even I barely remember this incident.

    Another impression of mine is that statistics of hate incidents are padded out with a lot of Dog-Bites-Man events that are so boring that they don’t make the national news.

    Among the bottom rungs of society, a lot of bad things are done, and a certain percentage of the bad things really do fit the preconceptions of the elites. Every so often during his career, for example, Jeff Sessions comes down like a ton of bricks on a murderer of some member of a Protected Category. Why? Because Sessions is strongly opposed to murder.

    A lot of statistical padding comes from intra-minority fights (a Mexican gang beats up a black or vice-versa) that readers of the NYT assume must have been committed by whites.

    Other statistical padding can stem from e.g., at a softball game, one drunk lesbian punches her girlfriend for flirting with someone else and calls her the D-word.

    I think padding can also come from inclusion of guys with white prison gang tattoos committing crimes.

    Most of these kind of padding cases don’t make the NYT national news pages.

    So, of the Hate Incidents that do get national coverage, what percentage ever get confirmed?

    This is a really important question and it’s time somebody did this study.”


    Anyway, I don’t presuppose to know how much discretionary time you actually have available, but he’s right, that’s a study somebody should do.

    Maybe that’s up your alley?

    fwiw, I’ve really enjoyed your work, the National Review piece especially, was great

    1. Sorry your comment only shows up now, but for some reason it had ended up in my spam folder. I only noted it now, after a reader emailed me to tell me his comments were not published. I would also be interested in a study of that sort, but it would take a lot of time and I don’t have any right now.

Comments are closed.