Posts
Comments
Why would there be a difference in muscle loss between losing weight with semaglutide and losing weight by "regular" dieting? Both methods involve taking in fewer calories than you burn. Why would there be a difference?
Also, if income is lowered in Tulsa, housing prices must drop, because you have fewer dollars chasing the same amount of housing.
Perhaps your argument is that, with lots of new housing built in NYC, prices drop in NYC and Tulsa, but consumers of housing are not necessarily better off because (for instance) they might be "forced" to move from Tulsa to NYC, making them less happy than they would have been otherwise, despite the lower housing prices in both cities.
But that's a completely different argument.
I may have erected a straw man here. But to that argument I would respond by noting that any kind of change makes some people worse off and some people better off, but market-based change almost always (in theory) results in more positives than negatives. The drop in price of word processors means that Adam may lose his job at the typewriter factory, but economic theory says that in the absence of anything unusual, the change is a benefit to the world as a whole.
This is especially true when changes are an increase in the abundance of a beneficial consumer good, like housing space.
Why didn't Adam's bosses move him to NYC before the new construction? Because, I assume, the bosses knew Adam and his colleagues wouldn't move because rent is so expensive. Or, which amounts to the same thing, they knew they couldn't attract enough talent in NYC because of the high housing prices.
This implies Adam and his colleagues DO have a choice. It's just that the new, lower housing prices in NYC provide enough incentive to make the move that Adam chooses to make the move, although perhaps reluctantly.
It seems that your argument, therefore, depends on housing prices being lowered in NYC. Otherwise, why wait for new housing before making the move?
I am confused by your number 3. Edited for what I think you mean, and using NYC (high cost) and Tulsa (low cost) to make your example more vivid:
"Building more in [NYC] causes people to move [to NYC because now it's cheaper to live there,] which causes jobs to move [to NYC], which causes more housing to be unused in [Tulsa, which causes housing costs to drop in Tulsa]."
If my understanding is correct, then housing prices drop in NYC, and housing prices drop in Tulsa. Therefore, housing prices drop on a national level. But you say,
"...which means housing costs don't decrease significantly on a national level ..."
But clearly they do.
Unless you're saying that the average American's housing cost doesn't drop. That's possible; it's just Simpson's Paradox. Housing is cheaper in NYC than before, and housing is cheaper in Tulsa than before. But housing is more expensive in NYC than Tulsa. Because you now have a higher population ratio NYC:Tulsa, the average housing cost might not drop.
But I would argue that kind of "housing costs don't decrease" is misleading. If the price of both types of housing drop, and that leads people to choose more expensive housing because it's now more affordable, that's not the cost increasing. That's people spending more money for a higher quality/quantity of product.
For instance, suppose the quantity of all types of housing increases 50%, and the price per square foot drops 33.3%. Now, everyone gets 50% more square feet for the same price as before. The total expenditure on housing is the same, but housing is definitely cheaper. Any apartment size that was $1500 before is $1000 now. Any house size that was $3000 before is $2000 now.
Substitute "in New York instead of Tulsa" for "50% more square feet" and it's obvious the cost has dropped, which makes the quantity demanded increase.
The fact that people spent 100x as much on face masks in 2020 doesn't mean that the cost of face masks increased 100x. People just bought 100x as many of them because the benefits increased by 100x. It's the total amount spent on face masks that increased 100x.
It seems to me that your argument conflates two different senses of "cost." One, cost per unit of housing; second, total amount spent on housing. If the first decreases, the second can increase, decrease, or stay the same. But it's the first that's the important one for the argument. Because the issue is, does building more housing change the cost per unit.
Having said all this, it's possible that I've misunderstood your argument.
How long did it take to feel the difference?
FWIW, my friend who lives in downtown Ottawa sent me this link, written by a neighbor he knows personally. (It's an account of him meeting some of the truckers parked on his street, who are nice people and considerate.) My friend went down to meet them too, and confirms this account.
https://maybury.ca/the-reformed-physicist/2022/02/03/a-night-with-the-untouchables/?fbclid=IwAR0se_AVoi1p4Ae7l3KQSsU3oxoCNNYfNi3SWaaay-2Qvkiqig35oNqElTk
I live a few miles from downtown and so haven't seen what's going on personally.
Attackers aren't given infinite attempts, and even if they are, God doesn't give them infinite time. So what you really want is to minimize the probability that the attacker guesses your password before giving up.
Suppose the attacking bot can make 200,000 attempts. By the first scheme, the probability the attacker guesses the password is .95 (plus an infinitesimal). By the first scheme but with a three-character password on a high roll, the probability is 1.00 (with 50 different characters, there are only 125K three-character words, so success in the remaining 199,999 attempts is certain).
By this measure, both passwords are weak, but the second is weaker than the first.
My Blackberry locks attackers out after 10 tries. So I would choose n=10 rather than n=200000. By that measure, the first scheme is roughly p=.950000, and the second is roughly p=.950072.
Fair enough. The question is then, does a vaccinated person's immune system take care of the virus so fast that the viral load remains "extremely low" enough to result in a negative test?
That seems counterintuitive given that Elizabeth says vaccinated people are more likely to be symptomatic, but I suppose it's possible that the immune system would trigger covid-19 symptoms even while maintaining a low viral load.
Is the vaccinated person's lower viral load enough to trigger a positive test, especially for those with symptoms?
If it is, shouldn't we be thinking of "reinfections" as those cases of serious disease, rather than simply positive tests?
One thing I've never found the answer to: is a positive test evidence of disease? It seems to me that a vaccinated person inhales the virus just as readily as an unvaccinated person, but the vaccinated person's immune system fights it off before symptoms (or before serious symptoms) appear.
In that case, wouldn't it be normal and expected for vaccinated people to sometimes test positive, in the sense of "there exist copies of the coronavirus in the upper respiratory system"?
Why do you say research before 2013 is of lower quality?
With a seven-day incubation period, does that mean it's 0 protection until about day 4, then near-perfect protection after that? (As per jimrandomh's comment of 4/17.)
Very well stated. I would be interested in a link to something that describes that principle, the outcome of the prediction process.
Correction to above: the quote from p. 206 refers to high schools, not colleges.
For colleges, I found a page here that lists 25th and 75th ACT percentiles. Some pairs of schools have no overlap at all; for instance, Ohio State's middle interval is (27, 31), while Vanderbilt is (32, 35). The average for college enrolees, per this study, was 20.1, with an SD of 4.33. So Vanderbilt's 25th percentile is almost +3 SD.
For GPA ... the 25th percentile for Vanderbilt is 3.75. The mean in this study was 2.72, with an SD of 0.65. So the 25th percentile for GPA was only around +1.6 SD.
For ACE at Vanderbilt, the 75th percentile is 0.92 SD higher than the 25th. If the same was true for GPA, the 75th percentile would have to be 4.34, which is clearly impossible, since the upper limit is 4.00.
So that supports the idea that for a given school, ACE has a narrower range than GPA.
I realized I forgot to provide evidence from the paper that the range of ACT within colleges is smaller than the range of GDP.
From p.207 of the paper:
"Thus, ACT scores are related to college graduation, in part, because students with higher scores are more likely to attend the kinds of colleges where students are more likely to graduate..."
(I think they obviously have this backwards, for the most part. Seems to me more likely that the higher graduation rates of those "kinds of colleges" are the ones that choose students with the higher ACT scores.)
From p. 206:
"Many schools do not have students with very high ACT scores, and a number of schools do not have students with very low ACT scores [which explains why some colleges do not have students from the full ACT range, even though they do have students from the full GPA range]."
In other words: students DO sort themselves into schools based on ACT score more than they do by GPA.
Here's an argument for why the study's conclusions are unsupported.
-----
Suppose that there are lots of things that go into predicting what makes a student successful. There's ACT score, and GPA, and leadership, and race, and socioeconomic status, and countless other things.
Now, suppose colleges have tried to figure out the weightings for each of those factors, and shared their results with each other. They all compute "success scores" for each student.
Harvard takes the top 1000 applicants by score. MIT takes the next 1000. Princeton takes the third 1000. And so on.
So, what happens when you run a regression to predict success from ACT/GPA/etc, while controlling for school?
Well, if the formula is correct, nothing is significant!
Consider Princeton. All its success scores are, say, between +2.04 (Z-score) and +2.02, because it takes a specific thin slice of the population. That means that all the students are roughly equal. So if you find a student with a higher ACT score, he's probably got a lower GPA. Because, if he was that high in both, he'd be higher than +2.04 overall and wind up at Harvard instead of Princeton.
In other words, NOTHING correlates to success, controlling for school, if colleges are good enough at predicting who will succeed.
Sure, there's a small amount of slack, between +2.02 and +2.04, but it's nowhere near enough to produce statistically significant evidence that any factor is important. Almost 100% of the variance is between schools, not within schools.
So that leaves noise. Any coefficients you find that are non-zero are probably just random artifacts.
Or ... they are systematic errors in how schools evaluate students.
In this particular study, they found that controlling for school, GPA was important to success but ACT score was not.
Well, all that means is that colleges are not weighting GPA highly enough. It does NOT mean that GPA is more important than ACT score, or any other factor -- only that GPA is more important *after you account for the college's choice in whom to admit*. It could be that the colleges are giving GPA/ACT a 1:15 ratio, and it should be only 1:10 instead. In other words, ACT could still be hugely more important than GPA, but the schools are making it a little TOO huge.
Even if everything in the study is correct, I would argue they misunderstood what they were measuring, and what the results mean. They only mean colleges are underestimating GPA relative to ACT, not that GPA is more important than ACT.
-----
Here's an analogy:
A store will only let you in if you have exactly $1000 worth of large bills in your wallet. An academic study measures how much stuff you get based on all the money in your wallet, including small bills. Since everyone has exactly $1000 in large bills, the regression can't deal with those, and it finds that 100% of the differences in success come from small bills.
That doesn't mean that large bills don't matter! It means that large bills don't matter given that you got admission to the store. Large bills DO matter, because otherwise you wouldn't have gotten in!
Similarly, this study's results don't mean that ACT doesn't matter. They mean that ACT doesn't matter given that you got admission to the college. If college admission criteria include ACT, then ACT does matter, because otherwise you wouldn't have gotten in!
Do any of the cited effects of higher air pollution depend on the subject recognizing the higher levels of pollutants, by sight or smell? Or is it invisible except for the effects?
FWIW, I remember reading about the Chevy Orlando, sold in Canada (but not the US) until maybe 2015. I recall it was said to be the cheapest new vehicle that could seat seven.
It seems cruel to me to ask someone to sit middle seat in front! Maybe not a small child, though.
The logic would be correct if, when Yovanni lied, he would always say it was Xavier Williams. In that case, there would be (roughly) 1/100 "Yovanni lies and says it was Xavier" for every 1/1,000,000 "Yovanni tells the truth and says it was Xavier."
But if Yovanni lies randomly, and you have no prior that he would lie and say Xavier any more than he would lie and say anyone else, you have 1/100 * 1/1,000,000 "Yovanni lies and also Yovanni says it was Xavier" for every 99/100 * 1/1,000,000 "Yovanni tells the truth and says it was Xavier," which is 99% truth.
I'm saying that if previously expensive goods become very cheap due to automation, the total for all goods will be valued higher in "real dollars". For that one good, the total dollar value could indeed be lower, even after overall inflation (such as, for instance, if the price drops by a factor of 20, but only 10 times as many items are produced).
But for the economy as a whole, the value in "real dollars" will always at least stay the same after productivity improvements that lower some prices relative to the status quo. That's because even though that one good may be lower in value even after adjusting for deflation caused by the lower price, the other goods in the economy will make up the difference and more by being higher in value after adjusting for deflation.
But even if workers move to less productive industries, productivity must still go up, adjusted for inflation.
Suppose 5 workers lose their jobs because it takes 5 fewer workers than before to make 10 widgets. The country is now making the same as before, but with 5 fewer workers. So productivity is higher than before, if the 5 workers remain unemployed. (Same output, less labor).
If the 5 workers get jobs elsewhere, even if they are almost completely unproductive and make only 1 grommet combined, the country is still more productive than before -- more output (1 extra grommet), same labor.
If productivity is output/labor, it must always be true, mathematically, that even if the (now) surplus labor is even minimally productive, average productivity rises.
For the case where the workers stay put making widgets and it's just that more widgets get made, that's just a special case where the surplus labor stays in the same industry, and the "proof" is the same as before.
I'm not an economist, but here's an answer based on my understanding.
Suppose the market produces 10 widgets with 10 hours of labor. Those widgets are worth $1 each. Now, an innovation comes along that allows twice as many units to be produced with the same amount of labor.
The company can now produce 10 units with only 5 hours of labor. It then reallocates the five hours, either by assigning the workers to other products, dismissing them to find other work, or whatever.
Clearly, the economy is no less productive at this point. When the now surplus labor moves to another use, the economy is strictly more productive, producing more than before for the same amount of labor.
As you point out, if production is twice as efficient, the price will drop, and quantity demanded will increase at the new, lower, price. So, most likely, the company will wind up keeping the same amount of labor, but producing 20 units instead of 10.
It's true that all things being equal, the value of the 20 units will not be twice the value of the 10 units from before, since the price will drop substantially. So it's true that twice the productivity will not measure twice the monetary value.
However: the lower price produces deflation. That product deflates in price by around 50 percent, but in the overall US (say) economy, it amounts to a very small amount of deflation. Seeing that deflation, the Fed realizes it should increase the money supply (print more money) to keep the overall price level the same (or to keep it at its target 2% inflation, or whatever).
What happens, then, is: the price of widgets is lower, both before and after inflation, but the price of everything else is slightly higher to compensate.
If you add up the economy using all the old quantities but the new prices, they have to stay exactly the same, because inflation is zero. But: adding in the additional 10 widgets means that GDP (after inflation) has increased by their value, which means GDP is higher, with inflation at zero, and the same amount of labor.
------
In summary:
In terms of goods produced, the country is obviously more productive after the innovation, because it has 10 more widgets with the same amount of labor. The monetary value might not show that -- it could indeed go down if the price of widgets falls enough. However, if you choose to measure in dollars instead of widgets, you have to adjust for inflation to keep the dollars constant. If you do that, you can prove mathematically that the overall value of everything produced must be higher. That's because the more the price of widgets drops, the more deflation you have, and the two cancel out, leaving only the value of the extra widgets.
Tried a few variations of the simulation, and found that if you seed a population with high superspreaders, you can indeed get to herd immunity in the 20-30% infection range.
Wrote it up for my blog at the link below. Let me know if I've screwed anything up.
http://blog.philbirnbaum.com/2020/05/herd-immunity-comes-faster-when-some.html
Oops! There was a problem with my simulation, where the random numbers were repeating. I fixed it, and the results changed. It now took about 45% infected before R dropped below 1. That's for a geometric (exponential) distribution of spreaders. For a uniform distribution, it should take 75% with R0=4.
It's hard to figure out what geometric distribution gives the equivalent initial R0=4 by trial and error, but maybe I'll calculate it by expectation just to get the 45% firmed up better.
When I tried a more spread-out distribution, I didn't get that much below 45% for anything plausible. I actually squared the relative weightings (so if A had 4x as many chances to spread as B, he now has 16x), and I don't think it dropped below 40%. Too lazy do walk over to double-check my notes as I write this.
I wrote a simulation where some people are bigger spreaders than others (I used a geometric distribution), and I indeed found cases where 20 percent infected was enough to send R dropping to zero with no other interventions.
I had never thought that a number as low as 25 percent (say) would be enough, but your logic convinced me it was plausible, and the simulation confirmed it.
Zvi, what specific distribution were you thinking? I can find one where the top 20 percent makes up 80 percent of interactions, but if you have one in mind, I'll try that one in the sim and see what happens.
Regarding where the new cases are coming from:
The City of Ottawa, where I live, published a daily digest of "detailed" data up to April 29. They broke down all cases to date by where the exposure came from.
As of April 21, there were 179 cases of exposure with "no travel and no exposure to a known case." As of April 28, there were 183.
So, in that seven-day period, there were FOUR new cases with no travel/contact, out of 354 new cases total.
In other words, a maximum 4 people were infected from surfaces, grocery stores, bike paths, making deliveries, etc. As far as I can tell, that might even include working in a grocery store. That's less than 2 percent of new cases, or four cases out of a population close to one million.
That seems like great news, but nobody has commented on it.
Source: Table 4 of the digest. April 22 pdf is here: https://www.ottawapublichealth.ca/en/reports-research-and-statistics/resources/Documents/covid-19/22-Apr/Web-PDF-COVID-2019-epi-update_20200422wOutbreakEB-MA9MA.pdf
April 29 pdf is here: https://www.ottawapublichealth.ca/en/reports-research-and-statistics/resources/Documents/covid-19/29-Apr/Web-PDF-COVID-2019-epi-update_20200429-46y8.pdf
Would "salting" with zinc or copper help too?
All these points make sense. But aren't they also (with the exception of the one about members of the same household) subject to the logic that they reduce roughly proportionally to reduced contacts? For instance, even in the unlikely case my contacts' contacts are not reducing, I am still reducing contacts with my contacts' contacts by reducing contacts with my contacts.
I am puzzled at how mild interventions don't show a much bigger decrease.
In Ottawa, where I live, we have social distancing and have shutdown non-essential places of business. If you work in a closed business, you have probably reduced your person-to-person contacts from (guessing) 300 per week to maybe 50 -- those 50 being people at grocery stores, etc. Moreover, the intensity of those instances of contact has dropped. You may have played poker with 10 people, mutually touching cards and chips and sitting together for hours. Now, you stand within 2m of a store cashier for a couple of minutes.
Just this 83% drop -- which I think is conservative -- should push R down from 3 (without intervention) to 0.5.
Add in hygiene improvements and more aggressive quarantining of those with symptoms, and R should drop even farther below 0.5.
If my numbers and logic are reasonable, the reason we haven't seen a lot of dropoff yet must be because of legacy cases (from before intervention) still coming in and obscuring the current trajectory. (We've only had serious social distancing for 12 days or so.)
Unless there's there something wrong with my calculations or logic. Are my estimates of contact frequency (300, 50) badly off?
What am I missing?
In comment 2, Emmett said, "You're not accounting for the costs of investigating the prediction market itself."
Which is true, but I think the point is not that you should bet. The point is that if you have strong beliefs on the subject, you should bet. Obviously, if you think Obama is going to do better than the polls suggest, you have strong opinions and have already investigated the issue.
The combination of "I have a strong rational belief in X and believe the betting market is wrong" and "I am nonetheless choosing not to bet" indicates that the belief is not rational at all.
To those people who say that they don't bet on principle ... I say, set up a sting operation where you offer them a no-brainer, like even money that Obama will get more than 10% of the vote. When they jump at that one, they will never again have a decent excuse for not betting.
Not to draw attention away from your main argument, but how does 1101 map onto {0, 2, 3}? It's probably obvious, but I don't see it.