re: a trillion prices, I'm running through the data now, and while it's ugly and needs a lot of cleansing, and as a computer person who recently entered the healthcare system, I can confidently assert that EY is right: they don't have any idea what you can do with computers/data. The entire industry seems 25 years behind on that front. I worked in marketing previously, and they had amazing statistical models that could calculate the % of someone buying a branded mug to 3 decimal places, but I got introduced to a state of the art healthcare model for determining patient risk that ... adds up the # of major diagnoses they have. That's the model. Do you have COPD and diabetes? You're a 2.
There is a massive amount of $ in that dataset if you can make it intelligible (and you can, it just takes some work and some big hard drives). I emailed the guy at Dolt whose post inspired the Hanson and EY comments and he knows it - there are already tons of 3rd party companies who work in healthcare just turning messy data into stuff that backward hospitals, orgs, insurers, etc... can actually absorb. I wish I could dedicate a lot more time to working on it.
How hard would it be, absent legal/regulatory/signaling issues, to actually get a very good set of risk factors, you think? What about not absent those?
I'm very excited, trying to get my day to day done so I can work on it.
Re: risk factors, I think that absent those things, bluntly, you could do clairvoyance magic. Especially if you could get big opt in for people's continuous glucose monitors, oura rings, fitbits, 23andme, apple watches, etc..... I'm an anti-surveillance-state civil libertarian, but a voluntary, deeply "invasive" surveillance co-op between you, your doc and helpful data scientists would have insane benefit.
Not absent those things, you are largely stuck with very basic models - unless, of course, you are already at a healthcare organization that has large amounts of health data that you've already jumped through the legal/regulatory/signaling hoops. Hypothetically speaking, of course.
That said, this trillion prices data could give you a backdoor anonymized aggregated peek into this, because at the end of the day, the price data (because in healthcare, prices are different in every case) IS essentially patient treatment data, modulated through the lens of insurer costs (which isn't bad, just adds more info!). Can you take big hospital/healthcare org A, look at their prices (because there's a trillion of them, it will cover a large % of actual individual cases - and it covers at least a somewhat static population) for 1) treatment of high A1c over time, 2) diabetes related stuff over time, 3) dialysis and then 4) common end of life codes - and how much it is likely to cost (with over time trendlines), and even worse, how much it costs at Hospital A vs Hospital B, or how well insurers are reimbursing for it! Does that seem like an interesting statistical model? It does to me. And because of the Price Transparency law, it's all free and open and no one will prosecute you for HIPAA violations for analyzing it. (I hope)
Why didn't you at least get a home test for covid? They're pretty cheap! What was your logic for not wanting to know? Just curious. One disadvantage of the home tests is that people tend not to report the results, so the numbers the agencies report are most likely on the low side. My whole family just went through this ... first the daughter, then the wife, finally me ... all about 3 days apart according to the home tests. We had all gotten both shots and the booster for my wife and me. Chills, tired, slight headache, but the worst thing was the dizziness (brain fog?) ... move too fast and wow ... it's like getting off one of those roundabouts!
a principle in medicine (or so I've been told, I Am Not A Doctor) is never do a test if the results won't change the treatment. He obviously had an illness of a type that necessitated doing certain things (whatever those things are) - and if it was covid instead of the flu, or a some other nasty respiratory thing, that wouldn't change any of those things. That said, it IS a very cheap test, and you might want to know just out of curiosity, or to inform "do I need to my omicron booster now or later?"
Don't test if it doesn't change the therapy is a good rule. It is not only about the cost of the test which in this case is trivial but knowing the positive result can cause a lot of anxiety in certain people. Maybe not in Zvi but he is showing a good practice for others. You could say that most likely he had covid and that should make someone anxious too. Nevertheless, by not testing the doctor is signalling that his illness is not serious as opposed to testing which would signal that his condition must be so serious that a confirmatory test was required.
Yeah, a big part of it is the severity, as you say! I'm sure if for whatever reason he took a turn for the worse, ended up in the hospital, and the doc thought "hey, if he DOES have covid, I might need to give him covid-specific treatments like Paxlovid" and then would definitely test, because it would change the treatment.
Felt like a normal illness and was never tempted to bother. I've had a bunch of other similarly low-level things happen over the past two years, I have 3 kids. In hindsight this one was somewhat more Covid-flavored. But if I had gotten the test, my presumption is only annoying things would have happened.
I find it baffling that your intellectual curiosity, especially given this blog, wouldn't cause you to take a simple test. Feels like it was a fear based, turn-a-blind-eye decision. You might have have had to inform people you'd been around that they were potentially exposed. You might have felt some anxiety around still hugging and kissing your children. You might have had to wait until you tested negative again to see your personal trainer. Obviously to each his own, but I think it's worth exploring this attitude, because if even Zvi Mowshowitz won't take a test when it comes down to it, to prevent some "annoying things" - yikes. :)
What are the costs? You could choose not to take precautions if you believed it didn't matter. You could choose not to worry about it being positive vs. negative. So the only cost is the guilt or burden you'd feel given the delta between what you feel you maybe should or need to do to protect others vs. what you're actually doing. Otherwise the test is free, at home, and takes 10 seconds.
There's very little upside to knowing. You need a positive test for a Paxlovid script, but Paxlovid might be of little benefit to you.
One downside of knowing is that you are forced to either abandon things like medical appointments or lie when they ask things like if you've been in contact with someone with COVID in the past ten days.
If I'm sick with anything, COVID or not, I'll avoid being around people. If I'm no longer sick, then I won't. Post-symptomatic spread is rare, and unless you're a hermit you're almost certainly going to be exposed to COVID many times.
Lol. “Post symptomatic spread is rare” is not true, esp with a still positive antigen test. So you’re putting your head in the sand because you don’t want the guilt of lying.
"The Working Party was unable to assess the likelihood of post-symptomatic transmission (meaning that an index case has no symptoms during the exposure period of their close contacts, but previously had symptoms) because of an absence of evidence."
However, that no one has bothered to measure it is itself signal that the researchers don't think it's worth measuring.
I think it's a logical to assume that if you are still testing positive on an antigen test, even if your symptoms are 100% gone (hard to even know), you are still shedding enough virus to spread a virus. Anecdotally, I've had many friends contract covid from people who "felt sick last week" but never tested, and only saw that person unmasked.
So I am assuming Zvi didn't want to test because his kids' schools have policies that if they're exposed to covid they need to isolate or test themselves, and that would have turned his life upside down for a week or two. So better to just turn a blind eye to it. I get it - I really get that this is hard - but I think we can stand to be a tiny bit more altruistic.
Hello from the UK. As a near two-year-long reader of your lucid, detailed and thorough posts from across the pond, I am interested why you did not check for covid, given the symptoms of your ailment? Tight chest, cough, trouble breathing etc. Yes it may have been a few other respiratory infections or viruses, but given that tests (I assume) are still available at low/minimal cost in the US why did you not take test from an elimination point of view, given these test only take a couple of minutes? Also - any I know this question has often been a misused moral hammer - did you think at any time it may have indeed been covid, and thus transmittable to those less physically well than yourself?
Finally, you say your trainer assumed you had covid, in some ways it reads as though you assumed you did not. I hope your breathing is easier now and that you will be able to build up strength again soon.
I was already staying home and resting. Not putting other people at risk applies to non-Covid illness as well, so I was already careful about that. I'm always confused why people don't treat other illnesses that way.
Re: Dishwashers and their washing ability. Have you seen this youtube video (https://www.youtube.com/watch?v=_rBO8neWw04) on dishwashers, how they work, and why detergent pods are bad? I adopted it's suggestions (the two most important being: include pre-wash soap and make sure your water is running hot before you start the dishwasher) and have found that there was a noticeable increase in dishwasher efficacy and that #4 from the less-wrong post works quite well for me. Not quite _perfectly_, and I still have to be cognizant of loading configuration to make sure that things don't block other things from the sprayer, but I definitely do not need to take any special pre-washing care unless the item has a considerably amount of nooks and crannies.
Oh, and the other thing I found considerably helped (that wasn't included in the video) was that I run the dishwasher every single day, even on the rare occasions when I can't completely fill it. The time/effort efficiency is, in my opinion, more than worth it, and it makes sure that things never sit _too_ long before getting washed. It doesn't take any particular effort as normally my household requires running it once a day anyways, but on the rare occasions that it doesn't, I'll either put in bulkier items that I would normally handwash or else just run it half full.
I'm glad that somebody else considers a side comment about dishwashers to be the most important part of a post. This indicates that normality is upon us.
I mean, I _do_ think that normality is upon us, but my choice of what to comment on was less about "what's important in the post" and more about "what parts do I have something meaningful to contribute to"
I don't think I really have anything to say about most of the post that I think anyone else should bother listening to. But I _have_ recently had experience in getting my dishwasher to perform better.
At first this was nice to see someone change their mind in response to evidence. But then he tweets:
"I would also add it doesn't really slow the growth of rents in the area either. The study found that rents grew slightly less quickly (≈2%) in the surrounding area.
So for example rather than seeing a $500 increase in rent, you might see a $490 increase.
We need Good Cause.
Landlords are not rational players, they are greedy speculators. The housing market doesn’t exist in a vacuum. New York needs more housing, but it must be socially controlled by the people.
Let's pass Good Cause and house the homeless today."
My skim of the study doesn't support an assertion that growth was merely slowed, the only appearance of "2%" is to say that rents fell by 2%.
I'm not sure what the 2% number is about, but it is true that the actual cases the study is based on in NYC involved new construction slowing rent increases rather than actually creating decreases. The study predicts that a sufficiently large increase in the housing stock would reduce rents, but none of the areas actually see that large an increase.
In fact, the study actually finds that rents grew MORE QUICKLY in areas near new construction, with the rent depressing effects of that new construction only apparent after controlling for the selection effect of more new construction happening in areas with quicker rent rises.
Bottom of page 17: "The positive correlation between completed new high-rises and nearby residential rents is driven by the selection bias that new high-rises are located in areas with rising residential rents.According to the residential rents dataset, nominal rent growth rate is 4% for rental buildings within 500 feet of new high-rises, which is a significant one percentage point higher than the rest of NYC from 2003 to 2013."
So you can see why people without sufficient economic literacy made the wrong conclusion naively looking at that correlation. All you can ask is that those people listen to reason when the more informed confront them about their ignorance - and we see that happening here in real time in twitter!
> My guess is that until we fix the completely broken the way we assign research grants and choose which efforts to back, and are able to try things capable of working, throwing more government money at cancer won’t have much effect.
Do you have any quantification on this effect? My (not an expert) model here is that NIH dollars are in general ROI-positive (something like 2x ROI from a quick search), some projects extremely-ROI-positive (the Human Genome Project claiming something crazy like ~175x ROI, implying many projects are not positive ROI), but I don't know how things look at the margin for the next dollar invested, or what the specific ROI is on similar projects in the field of cancer research.
I suppose the overarching question is whether that marginal dollar would have been more productive left untaxed? (i.e. left circulating in the economy at GDP% growth, or whatever the correct benchmark is). What's the ROI threshold here where you get excited about allocating government funds to research like this?
Agree that "moonshot" is probably a misnomer here, it's more likely to simply be investment in basic/applied research. I think this is mostly a result of the inevitable marketing that goes on with any policy initiative. Though I suppose it is possible that these funds would be invested in a "seed stage accelerator" style like DARPA / ARPA-E, which seems to be a model for success for high-risk/high-reward bets.
I don't have a quantification of this. I would agree that NIH dollars are ROI-positive versus no NIH dollars but the same grant-seeking activities, but if this forces tons and tons of organization around grant-seeking activity this could easily end up negative, and it does not seem like NIH is moonshot-friendly in general, and it is crowding out potential alternatives while eating the majority of many scientists time, the vast majority if you count 'doing incremental worthless work to get grants.'
The salient alternative is not clear but it isn't to not fund science, it's to fix the methods, I'd assume. It's not about ROI of the marginal grant's dollars as such, it's about what kind of work gets done/encouraged/rewarded and how much time is spent trying to fund it.
Re: the social tax study, what exactly is a blocked savings account? Isn't that just a regular savings account? My friends and family don't learn about the savings accounts I've opened unless I tell them, and they don't know the balances either.
It's remarkable that 14% of people would opt for one of these strange public accounts when they could open a normal bank account instead.
Re: overproductive employee, this person should not be fed discrete tasks. They should be given ownership over a problem or a product and allowed to define the work. You can give them direction on what metrics to improve at first. If they're stellar at that then let them figure out their own definition of success.
RE dishwashers There are some good videos regarding using dishwashers to get them to wash your dishes. Main video https://youtu.be/_rBO8neWw04 Followup: https://www.youtube.com/watch?v=Ll6-eGDpimU the main trick is to put detergent in both the pre-wash and main wash sections, then run the dishwasher.
Re: not bothering to test; I'm sympathetic, despite (not because of) similarly-aligned priors. My workplace used to send out near-daily notifications of positive cases, sometimes multiples a day. HIPAA only served as a thin veneer of paper shield, in a place rife with gossip and only around 150 employees...it's obvious from day to day who's missing, and why, and for how long. People picked up A Reputation for being out sick a lot; hawks got super-judgey about coworker behaviour outside of work ("how dare you attend this outdoor BBQ without masks and distancing!" was a huge dramafest during summer 2020). And of course, being in SF, people actually followed the quarantine guidelines of staying home for weeks. This was very painful during panic-buying surges, where those of us who did show up to work had to pick up unprecedented levels of slack with no extra Slack given in recompense. Given this scrutiny and pressure, better to never get sick in the first place. Officially, at least.
Then Omicron happened, and after the initual massive Jan-March wave...suspiciously very little. Some weeks there are no emails at all, often just one or two. (It's not the same 150 people getting reinfected, turnover's been extremely high.) But we've all seen the wastewater data here, we all know cases are undercounted by $LARGE_PERCENTAGE. People are absolutely getting sick without officially reporting it to work, and possibly still clocking in anyway. Even if they won't admit it explicitly, revealed preferences are that most of my coworkers now...simply don't bother to test. Not unless they're *really* sick, in a strong pattern-matches-covid way. Because when you're poor enough to consider bagging groceries a sane economic career choice, it really sucks having to miss at least 5 days of work for no pay, or face the potential guilt of lying and killing Grandma.
It's difficult to be the only altruist when everyone around you is declining to do the same. What profit? No profit, only pain - my own immisseration, and work losing one of their critical irreplaceable employees for a week. Nationwide booster uptake is poor, but in SF it's pretty high, even/especially among the most vulnerable...so to whatever extent I might theoretically be marginally threatening a $RISK_FACTOR, this is one of the "safest" places to do so. The cost-benefit just doesn't pencil out. (Yes, I do stay home if actually-sick, no such instances in last 3 years have resembled covid at all though.)
I also notice that this drop-off just happened to coincide with the official cessation of company- and government-provided covid sick pay, much more cleanly than any actual lull in cases/deaths. Funny how that works. Incentives matter. (For a good long while, employees were forced to quarantine even while waiting for test results, symptomatic or not, regardless of exposure level...you know, the good old days of Only PCR Tests Are Legal.)
----
Re: Overproductive employees anxiously fearful of getting fired when not constantly working; I'm one of those. It feels like some sort of gross violation of the Laws of Capital to get paid for standing around, not adding any value. I truly envy my coworkers who can internalize the "get paid by the hour" mindset, or even ruthlessly Defect on it for personal profit at collective expense. (Oversight is quite lax, accountability for not "killing tasks" is...spotty at best. Everybody Knows who abuses this, but no one wants to be a fink for management.)
Though in fairness, this is because I actually have gotten fired/threatened with discipline for not appearing busy enough in the past. Am sadly not clever enough to generate The Symbolic Representation Of The Thing instead of The Thing. Hence, excessive uncompensated productivity.
I have attempted to implement the suggested ideal solutions, of advocating for more tasks, or getting side projects to fill up free time. This routinely gets denied. Management insists it's very important to encourage idleness, that times of no work are relaxation opportunities. They do not understand why this is not reassuring. Continuous ratcheting-down of job responsibilities, in a high-turnover constant-hiring environment: how can it *not* come across as gently nudging costly veterans (who willingly undersell their talents) out the door?
re: a trillion prices, I'm running through the data now, and while it's ugly and needs a lot of cleansing, and as a computer person who recently entered the healthcare system, I can confidently assert that EY is right: they don't have any idea what you can do with computers/data. The entire industry seems 25 years behind on that front. I worked in marketing previously, and they had amazing statistical models that could calculate the % of someone buying a branded mug to 3 decimal places, but I got introduced to a state of the art healthcare model for determining patient risk that ... adds up the # of major diagnoses they have. That's the model. Do you have COPD and diabetes? You're a 2.
There is a massive amount of $ in that dataset if you can make it intelligible (and you can, it just takes some work and some big hard drives). I emailed the guy at Dolt whose post inspired the Hanson and EY comments and he knows it - there are already tons of 3rd party companies who work in healthcare just turning messy data into stuff that backward hospitals, orgs, insurers, etc... can actually absorb. I wish I could dedicate a lot more time to working on it.
That sounds exciting.
How hard would it be, absent legal/regulatory/signaling issues, to actually get a very good set of risk factors, you think? What about not absent those?
I'm very excited, trying to get my day to day done so I can work on it.
Re: risk factors, I think that absent those things, bluntly, you could do clairvoyance magic. Especially if you could get big opt in for people's continuous glucose monitors, oura rings, fitbits, 23andme, apple watches, etc..... I'm an anti-surveillance-state civil libertarian, but a voluntary, deeply "invasive" surveillance co-op between you, your doc and helpful data scientists would have insane benefit.
Not absent those things, you are largely stuck with very basic models - unless, of course, you are already at a healthcare organization that has large amounts of health data that you've already jumped through the legal/regulatory/signaling hoops. Hypothetically speaking, of course.
That said, this trillion prices data could give you a backdoor anonymized aggregated peek into this, because at the end of the day, the price data (because in healthcare, prices are different in every case) IS essentially patient treatment data, modulated through the lens of insurer costs (which isn't bad, just adds more info!). Can you take big hospital/healthcare org A, look at their prices (because there's a trillion of them, it will cover a large % of actual individual cases - and it covers at least a somewhat static population) for 1) treatment of high A1c over time, 2) diabetes related stuff over time, 3) dialysis and then 4) common end of life codes - and how much it is likely to cost (with over time trendlines), and even worse, how much it costs at Hospital A vs Hospital B, or how well insurers are reimbursing for it! Does that seem like an interesting statistical model? It does to me. And because of the Price Transparency law, it's all free and open and no one will prosecute you for HIPAA violations for analyzing it. (I hope)
Why didn't you at least get a home test for covid? They're pretty cheap! What was your logic for not wanting to know? Just curious. One disadvantage of the home tests is that people tend not to report the results, so the numbers the agencies report are most likely on the low side. My whole family just went through this ... first the daughter, then the wife, finally me ... all about 3 days apart according to the home tests. We had all gotten both shots and the booster for my wife and me. Chills, tired, slight headache, but the worst thing was the dizziness (brain fog?) ... move too fast and wow ... it's like getting off one of those roundabouts!
a principle in medicine (or so I've been told, I Am Not A Doctor) is never do a test if the results won't change the treatment. He obviously had an illness of a type that necessitated doing certain things (whatever those things are) - and if it was covid instead of the flu, or a some other nasty respiratory thing, that wouldn't change any of those things. That said, it IS a very cheap test, and you might want to know just out of curiosity, or to inform "do I need to my omicron booster now or later?"
Don't test if it doesn't change the therapy is a good rule. It is not only about the cost of the test which in this case is trivial but knowing the positive result can cause a lot of anxiety in certain people. Maybe not in Zvi but he is showing a good practice for others. You could say that most likely he had covid and that should make someone anxious too. Nevertheless, by not testing the doctor is signalling that his illness is not serious as opposed to testing which would signal that his condition must be so serious that a confirmatory test was required.
Yeah, a big part of it is the severity, as you say! I'm sure if for whatever reason he took a turn for the worse, ended up in the hospital, and the doc thought "hey, if he DOES have covid, I might need to give him covid-specific treatments like Paxlovid" and then would definitely test, because it would change the treatment.
Felt like a normal illness and was never tempted to bother. I've had a bunch of other similarly low-level things happen over the past two years, I have 3 kids. In hindsight this one was somewhat more Covid-flavored. But if I had gotten the test, my presumption is only annoying things would have happened.
I find it baffling that your intellectual curiosity, especially given this blog, wouldn't cause you to take a simple test. Feels like it was a fear based, turn-a-blind-eye decision. You might have have had to inform people you'd been around that they were potentially exposed. You might have felt some anxiety around still hugging and kissing your children. You might have had to wait until you tested negative again to see your personal trainer. Obviously to each his own, but I think it's worth exploring this attitude, because if even Zvi Mowshowitz won't take a test when it comes down to it, to prevent some "annoying things" - yikes. :)
I'm a little confused that you know of this blog and yet don't already know the explanation for why this doesn't pass a cost-benefit test for Zvi.
What are the costs? You could choose not to take precautions if you believed it didn't matter. You could choose not to worry about it being positive vs. negative. So the only cost is the guilt or burden you'd feel given the delta between what you feel you maybe should or need to do to protect others vs. what you're actually doing. Otherwise the test is free, at home, and takes 10 seconds.
There's very little upside to knowing. You need a positive test for a Paxlovid script, but Paxlovid might be of little benefit to you.
One downside of knowing is that you are forced to either abandon things like medical appointments or lie when they ask things like if you've been in contact with someone with COVID in the past ten days.
If I'm sick with anything, COVID or not, I'll avoid being around people. If I'm no longer sick, then I won't. Post-symptomatic spread is rare, and unless you're a hermit you're almost certainly going to be exposed to COVID many times.
Lol. “Post symptomatic spread is rare” is not true, esp with a still positive antigen test. So you’re putting your head in the sand because you don’t want the guilt of lying.
Are there are any studies that support the idea that post symptomatic spread is a problem? I can't find any that even bother to measure it.
So, I can't say that post-symptomatic transmission is rare, I can just say we don't really know. This literature review also can't make any statements on post symptomatic transmission for lack of evidence: https://bmcinfectdis.biomedcentral.com/articles/10.1186/s12879-022-07440-0
"The Working Party was unable to assess the likelihood of post-symptomatic transmission (meaning that an index case has no symptoms during the exposure period of their close contacts, but previously had symptoms) because of an absence of evidence."
However, that no one has bothered to measure it is itself signal that the researchers don't think it's worth measuring.
I think it's a logical to assume that if you are still testing positive on an antigen test, even if your symptoms are 100% gone (hard to even know), you are still shedding enough virus to spread a virus. Anecdotally, I've had many friends contract covid from people who "felt sick last week" but never tested, and only saw that person unmasked.
So I am assuming Zvi didn't want to test because his kids' schools have policies that if they're exposed to covid they need to isolate or test themselves, and that would have turned his life upside down for a week or two. So better to just turn a blind eye to it. I get it - I really get that this is hard - but I think we can stand to be a tiny bit more altruistic.
Hello from the UK. As a near two-year-long reader of your lucid, detailed and thorough posts from across the pond, I am interested why you did not check for covid, given the symptoms of your ailment? Tight chest, cough, trouble breathing etc. Yes it may have been a few other respiratory infections or viruses, but given that tests (I assume) are still available at low/minimal cost in the US why did you not take test from an elimination point of view, given these test only take a couple of minutes? Also - any I know this question has often been a misused moral hammer - did you think at any time it may have indeed been covid, and thus transmittable to those less physically well than yourself?
Finally, you say your trainer assumed you had covid, in some ways it reads as though you assumed you did not. I hope your breathing is easier now and that you will be able to build up strength again soon.
Yeah, I'm fine now.
I was already staying home and resting. Not putting other people at risk applies to non-Covid illness as well, so I was already careful about that. I'm always confused why people don't treat other illnesses that way.
Another excellent post. These make my lunch on Thursdays interesting and informative. Thank you.
Re: Dishwashers and their washing ability. Have you seen this youtube video (https://www.youtube.com/watch?v=_rBO8neWw04) on dishwashers, how they work, and why detergent pods are bad? I adopted it's suggestions (the two most important being: include pre-wash soap and make sure your water is running hot before you start the dishwasher) and have found that there was a noticeable increase in dishwasher efficacy and that #4 from the less-wrong post works quite well for me. Not quite _perfectly_, and I still have to be cognizant of loading configuration to make sure that things don't block other things from the sprayer, but I definitely do not need to take any special pre-washing care unless the item has a considerably amount of nooks and crannies.
Oh, and the other thing I found considerably helped (that wasn't included in the video) was that I run the dishwasher every single day, even on the rare occasions when I can't completely fill it. The time/effort efficiency is, in my opinion, more than worth it, and it makes sure that things never sit _too_ long before getting washed. It doesn't take any particular effort as normally my household requires running it once a day anyways, but on the rare occasions that it doesn't, I'll either put in bulkier items that I would normally handwash or else just run it half full.
I'm glad that somebody else considers a side comment about dishwashers to be the most important part of a post. This indicates that normality is upon us.
I mean, I _do_ think that normality is upon us, but my choice of what to comment on was less about "what's important in the post" and more about "what parts do I have something meaningful to contribute to"
I don't think I really have anything to say about most of the post that I think anyone else should bother listening to. But I _have_ recently had experience in getting my dishwasher to perform better.
State Senator Jabari Brisport update - it took him about 24 hours to look into the issue and reverse his position: https://twitter.com/JabariBrisport/status/1570486831229317120
At first this was nice to see someone change their mind in response to evidence. But then he tweets:
"I would also add it doesn't really slow the growth of rents in the area either. The study found that rents grew slightly less quickly (≈2%) in the surrounding area.
So for example rather than seeing a $500 increase in rent, you might see a $490 increase.
We need Good Cause.
Landlords are not rational players, they are greedy speculators. The housing market doesn’t exist in a vacuum. New York needs more housing, but it must be socially controlled by the people.
Let's pass Good Cause and house the homeless today."
My skim of the study doesn't support an assertion that growth was merely slowed, the only appearance of "2%" is to say that rents fell by 2%.
I'm not sure what the 2% number is about, but it is true that the actual cases the study is based on in NYC involved new construction slowing rent increases rather than actually creating decreases. The study predicts that a sufficiently large increase in the housing stock would reduce rents, but none of the areas actually see that large an increase.
In fact, the study actually finds that rents grew MORE QUICKLY in areas near new construction, with the rent depressing effects of that new construction only apparent after controlling for the selection effect of more new construction happening in areas with quicker rent rises.
Bottom of page 17: "The positive correlation between completed new high-rises and nearby residential rents is driven by the selection bias that new high-rises are located in areas with rising residential rents.According to the residential rents dataset, nominal rent growth rate is 4% for rental buildings within 500 feet of new high-rises, which is a significant one percentage point higher than the rest of NYC from 2003 to 2013."
So you can see why people without sufficient economic literacy made the wrong conclusion naively looking at that correlation. All you can ask is that those people listen to reason when the more informed confront them about their ignorance - and we see that happening here in real time in twitter!
> My guess is that until we fix the completely broken the way we assign research grants and choose which efforts to back, and are able to try things capable of working, throwing more government money at cancer won’t have much effect.
Do you have any quantification on this effect? My (not an expert) model here is that NIH dollars are in general ROI-positive (something like 2x ROI from a quick search), some projects extremely-ROI-positive (the Human Genome Project claiming something crazy like ~175x ROI, implying many projects are not positive ROI), but I don't know how things look at the margin for the next dollar invested, or what the specific ROI is on similar projects in the field of cancer research.
I suppose the overarching question is whether that marginal dollar would have been more productive left untaxed? (i.e. left circulating in the economy at GDP% growth, or whatever the correct benchmark is). What's the ROI threshold here where you get excited about allocating government funds to research like this?
Agree that "moonshot" is probably a misnomer here, it's more likely to simply be investment in basic/applied research. I think this is mostly a result of the inevitable marketing that goes on with any policy initiative. Though I suppose it is possible that these funds would be invested in a "seed stage accelerator" style like DARPA / ARPA-E, which seems to be a model for success for high-risk/high-reward bets.
I don't have a quantification of this. I would agree that NIH dollars are ROI-positive versus no NIH dollars but the same grant-seeking activities, but if this forces tons and tons of organization around grant-seeking activity this could easily end up negative, and it does not seem like NIH is moonshot-friendly in general, and it is crowding out potential alternatives while eating the majority of many scientists time, the vast majority if you count 'doing incremental worthless work to get grants.'
The salient alternative is not clear but it isn't to not fund science, it's to fix the methods, I'd assume. It's not about ROI of the marginal grant's dollars as such, it's about what kind of work gets done/encouraged/rewarded and how much time is spent trying to fund it.
Re: the social tax study, what exactly is a blocked savings account? Isn't that just a regular savings account? My friends and family don't learn about the savings accounts I've opened unless I tell them, and they don't know the balances either.
It's remarkable that 14% of people would opt for one of these strange public accounts when they could open a normal bank account instead.
Re: overproductive employee, this person should not be fed discrete tasks. They should be given ownership over a problem or a product and allowed to define the work. You can give them direction on what metrics to improve at first. If they're stellar at that then let them figure out their own definition of success.
RE dishwashers There are some good videos regarding using dishwashers to get them to wash your dishes. Main video https://youtu.be/_rBO8neWw04 Followup: https://www.youtube.com/watch?v=Ll6-eGDpimU the main trick is to put detergent in both the pre-wash and main wash sections, then run the dishwasher.
Re: not bothering to test; I'm sympathetic, despite (not because of) similarly-aligned priors. My workplace used to send out near-daily notifications of positive cases, sometimes multiples a day. HIPAA only served as a thin veneer of paper shield, in a place rife with gossip and only around 150 employees...it's obvious from day to day who's missing, and why, and for how long. People picked up A Reputation for being out sick a lot; hawks got super-judgey about coworker behaviour outside of work ("how dare you attend this outdoor BBQ without masks and distancing!" was a huge dramafest during summer 2020). And of course, being in SF, people actually followed the quarantine guidelines of staying home for weeks. This was very painful during panic-buying surges, where those of us who did show up to work had to pick up unprecedented levels of slack with no extra Slack given in recompense. Given this scrutiny and pressure, better to never get sick in the first place. Officially, at least.
Then Omicron happened, and after the initual massive Jan-March wave...suspiciously very little. Some weeks there are no emails at all, often just one or two. (It's not the same 150 people getting reinfected, turnover's been extremely high.) But we've all seen the wastewater data here, we all know cases are undercounted by $LARGE_PERCENTAGE. People are absolutely getting sick without officially reporting it to work, and possibly still clocking in anyway. Even if they won't admit it explicitly, revealed preferences are that most of my coworkers now...simply don't bother to test. Not unless they're *really* sick, in a strong pattern-matches-covid way. Because when you're poor enough to consider bagging groceries a sane economic career choice, it really sucks having to miss at least 5 days of work for no pay, or face the potential guilt of lying and killing Grandma.
It's difficult to be the only altruist when everyone around you is declining to do the same. What profit? No profit, only pain - my own immisseration, and work losing one of their critical irreplaceable employees for a week. Nationwide booster uptake is poor, but in SF it's pretty high, even/especially among the most vulnerable...so to whatever extent I might theoretically be marginally threatening a $RISK_FACTOR, this is one of the "safest" places to do so. The cost-benefit just doesn't pencil out. (Yes, I do stay home if actually-sick, no such instances in last 3 years have resembled covid at all though.)
I also notice that this drop-off just happened to coincide with the official cessation of company- and government-provided covid sick pay, much more cleanly than any actual lull in cases/deaths. Funny how that works. Incentives matter. (For a good long while, employees were forced to quarantine even while waiting for test results, symptomatic or not, regardless of exposure level...you know, the good old days of Only PCR Tests Are Legal.)
----
Re: Overproductive employees anxiously fearful of getting fired when not constantly working; I'm one of those. It feels like some sort of gross violation of the Laws of Capital to get paid for standing around, not adding any value. I truly envy my coworkers who can internalize the "get paid by the hour" mindset, or even ruthlessly Defect on it for personal profit at collective expense. (Oversight is quite lax, accountability for not "killing tasks" is...spotty at best. Everybody Knows who abuses this, but no one wants to be a fink for management.)
Though in fairness, this is because I actually have gotten fired/threatened with discipline for not appearing busy enough in the past. Am sadly not clever enough to generate The Symbolic Representation Of The Thing instead of The Thing. Hence, excessive uncompensated productivity.
I have attempted to implement the suggested ideal solutions, of advocating for more tasks, or getting side projects to fill up free time. This routinely gets denied. Management insists it's very important to encourage idleness, that times of no work are relaxation opportunities. They do not understand why this is not reassuring. Continuous ratcheting-down of job responsibilities, in a high-turnover constant-hiring environment: how can it *not* come across as gently nudging costly veterans (who willingly undersell their talents) out the door?
Brilliant, if very sad, observations.
Always my most informative and thought provoking read of the week.