It might be worth our giving a little explanation to The Guardian about how tax systems work. We impose taxes upon certain things. Activities, transactions, even at times unsuccessfully upon mere existence as with the poll tax. These taxes are then paid by those who indulge in such activities, perform such transactions, have the temerity to exist. If we then decide to cut the tax rate or level on an activity, type of transaction or mode of existence then it will be those who formerly paid the tax on such who benefit from the tax cut on such. This shouldn’t be all that difficult for people to understand but we do seem to have an entire newspaper devoted to not grasping the point […]
There is that objectionable idea that not taxing something is a giveaway. The root presumption there is that everything belongs to the State and we’re lucky it allows us to keep anything to deploy as we desire and not as those who stay awake in committee do. This is not an assumption that leads to a free country nor populace, nor a liberal society.
But it’s also to miss that logical point, that if income tax is to be reduced then it must be those currently paying income tax who benefit from not doing so in the future under the new rates. […] The low paid cough up hardly anything in income tax. Therefore the low paid gain hardly anything from income tax being reduced. This should be obvious.
Tim Worstall, “Budget Revelation – Those Who Pay Income Tax Benefit From Income Tax Cuts”, Continental Telegraph, 2018-10-30.
March 6, 2021
QotD: Why “the rich” benefit more from tax cuts
February 19, 2021
Freddie DeBoer’s arguments against successful charter schools
Scott Alexander’s extensive review of Freddie DeBoer’s book The Cult of Smart includes this discussion of DeBoer’s belief that American charter schools are fraudulent and only manage their headline-worthy educational outcomes by “cooking the books”:
I think DeBoer would argue he’s not against improving schools. He just thinks all attempts to do it so far have been crooks and liars pillaging the commons, so much so that we need a moratorium on this kind of thing until we can figure out what’s going on. But I’m worried that his arguments against existing school reform are in some cases kind of weak.
DeBoer does make things hard for himself by focusing on two of the most successful charter school experiments. If he’d been a little less honest, he could have passed over these and instead mentioned the many charter schools that fail, or just sort of plod onward doing about as well as public schools do. I think the closest thing to a consensus right now is that most charter schools do about the same as public schools for white/advantaged students, and slightly better than public schools for minority/disadvantaged students. But DeBoer very virtuously thinks it’s important to confront his opponents’ strongest cases, so these are the ones I’ll focus on here.
Success Academy is a chain of New York charter schools with superficially amazing results. They take the worst-off students — “76% of students are less advantaged and 94% are minorities” — and achieve results better than the ritziest schools in the best neighborhoods — it ranked “in the top 1% of New York state schools in math, and in the top 3% for reading” — while spending “as much as $3000 to $4000 less per child per year than their public school counterparts.” Its supporters credit it with showing “what you can accomplish when you are free from the regulations and mindsets that have taken over education, and do things in a different way.”
DeBoer will have none of it. He thinks they’re cooking the books by kicking out lower-performing students in a way public schools can’t do, leaving them with a student body heavily-selected for intelligence. Any remaining advantage is due to “teacher tourism”, where ultra-bright Ivy League grads who want a “taste of the real world” go to teach at private schools for a year or two before going into their permanent career as consultants or something. This would work — many studies show that smarter teachers make students learn more (though this specifically means high-IQ teachers; making teachers get more credentials has no effect). But it doesn’t scale (there are only so many Ivy League grads willing to accept low salaries for a year or two in order to have a fun time teaching children), and it only works in places like New York (Ivy League grads would not go to North Dakota no matter how fun a time they were promised).
I’m not sure I share this perspective. Success Academy isn’t just cooking the books — you would test for that using a randomized trial with intention-to-treat analysis. The one that I found is small-n, short timescale, and a little ambiguous, but I think basically supports the contention that there’s something there beyond selection bias. Teacher tourism might be a factor, but hardly justifies DeBoer’s “charter schools are frauds, shut them down” perspective. Even if Success Academy’s results are 100% because of teacher tourism, they found a way to educate thousands of extremely disadvantaged minority kids to a very high standard at low cost, a way public schools had previously failed to exploit. That’s not “cheating”, it’s something exciting that we should celebrate. If it doesn’t scale, it doesn’t scale, but maybe the same search process that found this particular way can also find other ways? Surely it doesn’t seem like the obvious next step is to ban anyone else from even trying?
And we only have DeBoer’s assumption that all of this is teacher tourism. Success Academy itself claims that they have lots of innovative teaching methods and a different administrative culture. If this explains even 10% of their results, spreading it to other schools would be enough to make the US rocket up the PISA rankings and become an unparalleled educational powerhouse. I’m not claiming to know for sure that this is true, but not even being curious about this seems sort of weird; wanting to ban stuff like Success Academy so nobody can ever study it again doubly so.
DeBoer’s second tough example is New Orleans. Hurricane Katrina destroyed most of their schools, forcing the city to redesign their education system from the ground up. They decided to go a 100% charter school route, and it seemed to be very successful. Unlike Success Academy, this can’t be selection bias (it was every student in the city), and you can’t argue it doesn’t scale (it scaled to an entire city!). But DeBoer writes:
After Hurricane Katrina, the neoliberal powers that be took advantage of a crisis (as they always do) to enforce their agenda. The schools in New Orleans were transformed into a 100% charter system, and reformers were quick to crow about improved test scores, the only metric for success they recognize. Whether these gains stand up to scrutiny is debatable. But even if these results hold, the notion of using New Orleans as a model for other school districts is absurd on its face. When we make policy decisions, we want to isolate variables and compare like with like, to whatever degree possible. The story of New Orleans makes this impossible. Katrina changed everything in the city, where 100,000 of the city’s poorest residents were permanently displaced. The civic architecture of the city was entirely rebuilt. Billions of dollars of public and private money poured in. An army of do-gooders arrived to try to save the city, willing to work for lower wages than they would ordinarily accept. How could these massive overall social changes possibly be replicated elsewhere? And how could we have any faith that adopting the New Orleans schooling system — without the massive civic overhaul — would replicate the supposed advantages?
These are good points, and I would accept them from anyone other than DeBoer, who will go on to say in a few chapters that the solution to our education issues is a Marxist revolution that overthrows capitalism and dispenses with the very concept of economic value. If he’s willing to accept a massive overhaul of everything, that’s failed every time it’s tried, why not accept a much smaller overhaul-of-everything, that’s succeeded at least once? There are plenty of billionaires willing to pour fortunes into reforming various cities — DeBoer will go on to criticize them as deluded do-gooders a few chapters later. If billions of dollars plus a serious commitment to ground-up reform are what we need, let’s just spend billions of dollars and have a serious commitment to ground-up reform! If more hurricanes is what it takes to fix education, I’m willing to do my part by leaving my air conditioner on “high” all the time.
February 4, 2021
QotD: The (as-yet-unfulfilled) promise of “personalized medicine”
(You can still have some personalized medicine. We don’t have to know the causes of depression to treat it. You might be depressed because your grandfather died, but Prozac can still make you feel better. So it’s possible that there’s a simple personalized monocausal way to check who eg responds better to Prozac vs. Lexapro, though the latest evidence isn’t really bullish about this. But this seems different from a true personalized medicine where we determine the root cause of your depression and fix it in a principled way.)
Even if we can’t get much out of this, I think it can be helpful just to ask which factors and sciences are oligocausal vs. massively polycausal. For example, what percent of variability in firm success are economists able to determine? Does most of the variability come from a few big things, like talented CEOs? Or does most of it come from a million tiny unmeasurable causes, like “how often does Lisa in Marketing get her reports in on time”?
Maybe this is really stupid – I’m neither a geneticist or a statistician – but I imagine an alien society where science is centered around polycausal scores. Instead of publishing a paper claiming that lead causes crime, they publish a paper giving the latest polycausal score for predicting crime, and demonstrating that they can make it much more accurate by including lead as a variable. I don’t think you can do this in real life – you would need bigger Big Data than anybody wants to deal with. But like falsifiability and compressability, I think it’s a useful thought experiment to keep in mind when imagining what science should be like.
Scott Alexander, “The Omnigenic Model As Metaphor For Life”, Slate Star Codex, 2018-09-13.
January 29, 2021
QotD: Banishing racism
We can – we must, in fact – learn to judge individuals as individuals, not as members of racial or other ascriptive groups. This has always been the right thing to do; as knowledge about genetic group differences becomes more detailed and widespread, we will need to learn how to focus rigorously on individuals with the same discipline (and the same justified fear of failure) that we now apply to averting our eyes from genetic group differences.
Part of the reason this evolution won’t be easy is that so much of our politics has been distorted by racial grievance-mongering. It’s not only the obvious bad guys like neo-Nazis, Black separatists like Louis Farrakhan, and Bharatiya Janata who are invested in racialist categorization as a lever to power. The political Left has fallen into a lazy habit of screaming “racist!” at anyone who disagrees with them, won’t readily relinquish that rhetorical club, and have a lot invested in the present system of taboo, resentment, “disparate impact” legislation, and racial identity politics; expect them, too, to be part of the problem rather than part of the solution.
Still, the right strategy is clear. Actual knowledge makes both prejudice and repression unsustainable. “Know thyself!” said the oracle, and behavioral genetics will allow – actually, force us – to know ourselves in ways we never have before. That way lies the pain of revelation, but also the path of redemption.
Eric S. Raymond, “A Specter is Haunting Genetics”, Armed and Dangerous, 2010-06-19.
January 27, 2021
QotD: Open-source the data
Eric S. Raymond, “Open-Sourcing the Global Warming Debate”, Armed and Dangerous, 2009-11-23.
January 23, 2021
QotD: “Genetics is interesting as an example of a science that overcame a diseased paradigm”
Genetics is interesting as an example of a science that overcame a diseased paradigm. For years, basically all candidate gene studies were fake. “How come we can’t find genes for anything?” was never as popular as “where’s my flying car?” as a symbol of how science never advances in the way we optimistically feel like it should. But it could have been.
And now it works. What lessons can we draw from this, for domains that still seem disappointing and intractable?
Turn-of-the-millennium behavioral genetics was intractable because it was more polycausal than anyone expected. Everything interesting was an excruciating interaction of a thousand different things. You had to know all those things to predict anything at all, so nobody predicted anything and all apparent predictions were fake.
Modern genetics is healthy and functional because it turns out that although genetics isn’t easy, it is simple. Yes, there are three billion base pairs in the human genome. But each of those base pairs is a nice, clean, discrete unit with one of four values. In a way, saying “everything has three billion possible causes” is a mercy; it’s placing an upper bound on how terrible genetics can be. The “secret” of genetics was that there was no “secret”. You just had to drop the optimistic assumption that there was any shortcut other than measuring all three billion different things, and get busy doing the measuring. The field was maximally perverse, but with enough advances in sequencing and computing, even the maximum possible level of perversity turned out to be within the limits of modern computing.
(This is an oversimplification: if it were really maximally perverse, chaos theory would be involved somehow. Maybe a better claim is that it hits the maximum perversity bound in one specific dimension)
Scott Alexander, “The Omnigenic Model As Metaphor For Life”, Slate Star Codex, 2018-09-13.
December 29, 2020
The Economics of Wine (Orley Ashenfelter, Princeton)
Marginal Revolution University
Published 30 Sep 2020What does an economist know about wine? Given that many wines need years to mature, how can one predict which ones will be great or not?
Princeton’s Orley Ashenfelter explains how he used economic principles and regression analysis to predict wine quality (and score great deals!). His research helped spawn an entire field dedicated to the economics of wine.
This video is based on the following paper:
Predicting the Quality and Prices of Bordeaux Wines By Orley Ashenfelter
https://www.researchgate.net/publicat…More of Orley Ashenfelter’s work: https://irs.princeton.edu/people/orle…
Orley Ashenfelter’s vineyard: https://cedarrosevineyards.com/
Want to see more Economists in the Wild? Check out our series: https://mru.io/economists-wild-67905
December 14, 2020
QotD: Goodhart’s law
Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.
Or as it has been reformulated:
Goodhart’s law is an adage named after economist Charles Goodhart, which has been phrased by Marilyn Strathern as: “When a measure becomes a target, it ceases to be a good measure.” One way in which this can occur is individuals trying to anticipate the effect of a policy and then taking actions which alter its outcome.
Set a target for tonnes of shoes and you get one tonne shoes. Set a target for 100 shoes and you get 100 left feet. Set a target for being on time and people fiddle their definition of time.
It is, by the way, entirely fine to insist that airlines play fair with telling us how long a flight will take. You said it will take 4 hours, then 4 hours should be about the time it takes. Yes, sure, we understand, airports, crowded places. Idiot passengers forget to board, luggage must be taken off. Winds vary, thunderstorms happen, French air traffic controllers actually turn up to work today, their one day in seven. Sure, there’re lots of variables. But if you say it’s about four hours then it should be about four hours. Great.
But to complain that they pad their number a bit is ludicrous. We’re holding their feet to the fire, insisting that an underestimate will lead to financial costs. Thus, obviously, they will overestimate. That’s not really even Goodhart’s Law, that’s just human beings. But then, as we know, those who would plan everything don’t deal well with the existence of people, do they?
Tim Worstall, “Goodhart’s Law Applies To Economies, To Everything – Why Not Scheduled Airline Flight Times?”, Continental Telegraph, 2018-08-27.
December 13, 2020
QotD: The statistical “Rule of Silicone Boobs”
“Sexy” means “likely to get published in the New York Times and/or get the researcher on a TEDx stage”. Actual sexiness research is not “sexy” because it keeps running into inconvenient results like that rich and high-status men in their forties and skinny women in their early twenties tend to find each other very sexy. The only way to make a result like that “sexy” is to blame it on the patriarchy, and most psychologist aren’t that far gone (yet).
[…]
Anything counterintuitive is also sexy, and (according to Rule 2) less likely to be true. So is anything novel that isn’t based on solid existing research. After all, the Times is the newspaper business, not in the truthspaper one.
Finding robust results is very hard, but getting sexy results published is very easy. Thus, sexy results generally lack robustness. I personally find a certain robustness quite sexy, but that attitude seems to have gone out of fashion since the Renaissance.
Jacob Falkovich, “The Scent of Bad Psychology”, Put a Number On It!, 2018-09-07.
November 7, 2020
Misunderstanding what is meant by “mineral reserves”
It seems to happen almost as regularly as Old Faithful, as someone blows a virtual gasket over the reserves of this or that mineral “running out” in x number of years. Tim Worstall explains why this is a silly misunderstanding of what the term “mineral reserves” actually means:
“Aerial view of a small mine near Mt Isa Queensland.” by denisbin is licensed under CC BY-ND 2.0
It’s not exactly unusual to see some environmental type running around screaming because mineral reserves are about to run out. The Club of Rome report, the EU’s “circular economy” ideas, Blueprint for Survival, they’re all based upon the idea that said reserves are going to run out.
They look at the usual listing (USGS, here) and note that at the current rate of usage reserves will run out in 30 to 50 years. Entirely correct they are too. It’s the next step which is such drivelling idiocy. For the claim then becomes that we will run out of those metals, those minerals, when the reserves do. This being idiot bollocks.
For a mineral reserve is, as best colloquial language can put it, the stuff we’ve prepared for use in the next few decades. Like, say, 30 to 50 years. That we’re going to run out of what we’ve got prepared isn’t a problem. For we’ve an entire industry, mining, whose job to to go prepare some more for us to use.
[…] A mineral reserve is something created by the mining company. Created by measuring, testing, test extracting and proving that the mineral can be processed, using current technology, at current prices, and produce a profit. Proving that this is not just dirt but is in fact ore.
Mineral reserves are things we humans make, not things that exist.
October 15, 2020
QotD: What the GDP is failing to show (even though it’s there)
Our problem is that we know technological revolutions produce growth. Yet economic growth is limp at best, meagre perhaps a better description. So, there’s something wrong here. Either our basic understandings about how growth occurs are wrong and we [are] loathe to agree to that. Not because too much is bound up in that understanding but because too much of it makes sense. The other explanation is that we’re counting wrong.
[…]
We know that we’ve not quite got new products and their falling prices in our estimates of inflation quite correctly. They tend to enter the inflation indices after their first major price falls, meaning that inflation is always overstated. Given that the number we really look at is real growth – nominal growth minus inflation – this means we are consistently underestimating real growth.
[…]
The more we dig into this the more convinced I am that our only real economic problem at present is counting. Everything makes sense if we are counting output and inflation incorrectly, under-estimating the first, over- the second. If we are doing that – and we know that we are, only not quite to what extent – then all other economic numbers make sense. We’re in the midst of a large technological change, we’ve full employment by any reasonable measure, wages and productivity should be rising strongly. If we’re mismeasuring as above then those two are rising strongly, we’re just not capturing it. Oh, and if that’s also true then inequality is lower than currently estimated too.
The thing is, the more we study the details of these questions the more it becomes clear that we are mismeasuring, and mismeasuring enough that all of the claimed problems, the low growth, low productivity rises, low wage growth, simply aren’t there in the first place. And if they ain’t then nothing needs to be done about them, does it? Except, perhaps, count properly.
Tim Worstall, “Where’s All The Economic Growth? Goldman Sachs Blames Apple’s iPhone”, Continental Telegraph, 2018-07-03.
October 9, 2020
QotD: How to analyze complex multivariate systems for the popular press
- Choose a complex and chaotic system that is characterized by thousands or millions of variables changing simultaneously (e.g. climate, the US economy)
- Pick one single output variable to summarize the workings of that system (e.g. temperature, GDP)
- Blame (or credit) any changes to your selected output variable on one single pet variable (e.g. capitalism, a President from the other party)
- Pick a news outlet aligned with your political tribe and send them a press release
- Done! You are now a famous scientist. Congratulations.
Warren Meyer, “Modern Guide to Analyzing Complex Multivariate Systems”, Coyote Blog, 2018-06-25.
August 28, 2020
National “cheater density” for popular online games
Richard Currie summarizes the findings of Ruby Fortune’s cheater research (note that there’s no data on China because reasons):
Ever torn your keyboard from the desk and flung it across the room, vowing to find the “scrub cheater” who ended your run of video-gaming success? Uh, yeah, us neither, but a study into the crooked practice might help narrow down the hypothetical search.
The research, carried out by casino games outfit Ruby Fortune, has produced a global heatmap of supposed cheater density.
According to the website, this was done by analysing “search trend and search volume data to reveal where in the world is most likely to cheat while playing online multiplayer video games”. The report looks at the frequency of search engine queries for the most-played video games and measures them against searches for related cheat codes, hacks and bots, to show which country has the highest density of cheaters, and which cheat categories are the most popular in each location.
[…]
There is a massive hole in the data, however, thanks to the Great Firewall of China, which has a terrible reputation for ruining the experience of online games.
If there was any doubt that the Middle Kingdom would otherwise take Brazil’s crown, consider that Dell once advertised a laptop for the market by saying it was especially good for running PUBG plugins to “win more at Chicken Dinner”, a reference to the “Winner winner chicken dinner” message that comes up on a victory screen.
Data from the Battle Royale granddad’s anti-cheat tech provider, BattlEye, has also suggested that at one point 99 per cent of banned cheaters were from China.
August 16, 2020
This is a “hockey stick” graph you can believe
Brian Micklethwait says this graph, unlike the more famous (debunked) “hockey stick”, shows one of the most important moments in human history:
If that graph, or another like it, is not entirely familiar to you, then it damn well should be. It pinpoints the moment when our own species started seriously looking after its own creature comforts. This was, you might say, the moment when most of us stopped being treated no better than farm animals, and we began turning ourselves into each others’ pets.
Patrick Crozier and I will be speaking about this amazing moment in the history of the human animal in our next recorded conversation. That will, if the conversation happens as we hope and the recording works as we hope, find its way to here.
I’m not usually one for podcasts, in the same way that I’m not an audiobook user: I find I’m unable to do other things while listening to the spoken word, and it’s always far faster to read a text than to have it read to you. In this particular case, I might try to make an exception, and give up hope of doing anything else productive while I listen.
August 8, 2020
Andrew Sullivan – “[T]he Kendi test: does the staff reflect the demographics of New York City as a whole?”
In his latest Weekly Dish, Andrew Sullivan looks at an earnest diversity initiative of The Newspaper Guild of New York:
I’m naming this after Ibram X. Kendi because his core contribution to the current debate on race is the notion that “any measure that produces or sustains racial inequity between racial groups” is racist. Intent is irrelevant. I don’t think many sane people believe A.G. Sulzberger or Dean Baquet are closet bigots. But systemic racism, according to Kendi, exists in any institution if there is simply any outcome that isn’t directly reflective of the relevant racial demographics of the surrounding area.
The appeal of this argument is its simplicity. You can tell if a place is enabling systemic racism merely by counting the people of color in it; and you can tell if a place isn’t by the same rubric. The drawback, of course, is that the world isn’t nearly as simple. Take the actual demographics of New York City. On some measures, the NYT is already a mirror of NYC. Its staff is basically 50 – 50 on sex (with women a slight majority of all staff on the business side, and slight minority in editorial). And it’s 15 percent Asian on the business side, 10 percent in editorial, compared with 13.9 percent of NYC’s population.
But its black percentage of staff — 10 percent in business, 9 percent in editorial — needs more than doubling to reflect demographics. Its Hispanic/Latino staff amount to only 8 percent in business and 5 percent in editorial, compared with 29 percent of New York City’s demographics, the worst discrepancy for any group. NYT’s Newsroom Fellowship, bringing in the very next generation, is 80 percent female, 60 percent people of color (including Asians), and, so far as I can tell, one lone white man. And it’s why NYT‘s new hires are 43 percent people of color, a definition that includes Asian-Americans.
But notice how this new goal obviously doesn’t reflect New York City’s demographics in many other ways. It draws overwhelmingly from the college educated, who account for only 37 percent of New Yorkers, leaving more than 60 percent of the city completed unreflected in the staffing. It cannot include the nearly 19 percent of New Yorkers in poverty, because a NYT salary would end that. It would also have to restrict itself to the literate, and, according to Literacy New York, 25 percent of people in Manhattan “lack basic prose literary skills” along with 37 percent in Brooklyn and 41 percent in the Bronx. And obviously, it cannot reflect the 14 percent of New Yorkers who are of retirement age, or the 21 percent who have yet to reach 18. For that matter, I have no idea what the median age of a NYT employee is — but I bet it isn’t the same as all of New York City.
Around 10 percent of staffers would have to be Republicans (and if the paper of record nationally were to reflect the country as a whole, and not just NYC, around 40 percent would have to be). Some 6 percent of the newsroom would also have to be Haredi or Orthodox Jews — a community you rarely hear about in diversity debates, but one horribly hit by a hate crime surge. 48 percent of NYT employees would have to agree that religion is “very important” in their lives; and 33 percent would be Catholic. And the logic of these demographic quotas is that if a group begins to exceed its quota — say Jews, 13 percent — a Jewish journalist would have to retire for any new one to be hired. Taking this proposal seriously, then, really does require explicit use of race in hiring, which is illegal, which is why the News Guild tweet and memo might end up causing some trouble if the policy is enforced.
And all this leaves the category of “white” completely without nuance. We have no idea whether “white” people are Irish or Italian or Russian or Polish or Canadians in origin. Similarly, we do not know if “black” means African immigrants, or native black New Yorkers, or people from the Caribbean. 37 percent of New Yorkers are foreign-born. How does the Guild propose to mirror that? Ditto where staffers live in NYC. How many are from Staten Island, for example, or the Bronx, two places of extremely different ethnic populations? These categories, in other words, are incredibly crude if the goal really is to reflect the actual demographics of New York City. But it isn’t, of course.
My point is that any attempt to make a specific institution entirely representative of the demographics of its location will founder on the sheer complexity of America’s demographic story and the nature of the institution itself. Journalism, for example, is not a profession sought by most people; it’s self-selecting for curious, trouble-making, querulous assholes who enjoy engaging with others and tracking down the truth (at least it used to be). There’s no reason this skillset or attitude will be spread evenly across populations. It seems, for example, that disproportionate numbers of Jews are drawn to it, from a culture of high literacy, intellectualism, and social activism. So why on earth shouldn’t they be over-represented?
And that’s true of other institutions too: are we to police Broadway to make sure that gays constitute only 4 percent of the employees? Or, say, nursing, to ensure that the sex balance is 50-50? Or a construction company for gender parity? Or a bike messenger company’s staff to be reflective of the age demographics of the city? Just take publishing — an industry not far off what the New York Times does. 74 percent of its employees are women. Should there be a hiring freeze until the men catch up?
