As a 62 year old man, I think I’m entitled to rant once in a while. But instead of “you kids get off my lawn!”, this is my rant: In principle, economics can help advise most any decisions, like when to wake up, or whether to own a second car. But there are fixed costs to doing explicit econ analysis, and also persuasion costs when you try to influence the decisions of some audience. Thus econ analysis seems most valuable for the biggest decisions whose the audience respects economists for those decisions. Or perhaps many similar but smaller decisions which can all be analyzed at once in the same framework. As we economists are most known for our work evaluating institutions, and as our institutional choices are some of the biggest ones we have, this all suggests our biggest wins come there.
Correct me if I am wrong but you seem to propose that the liability for harms stemming from infecting somebody should fall on the infector, without apportioning some or all of the liability on the infectee. If so, this would produce inefficient outcomes.
Rationally we value a social control system/policy by the efficiency of the results it produces, compared to alternatives. Usually it is possible to identify a set of practices that are efficient, and then additional control mechanisms, such as laws and punishments, are created to encourage such practices. For driving it is efficient to select one side of the road and have either left or right-hand driving, and punish driving on whichever side goes against the grain, which means that in Britain you will be punished for driving on the side that is legal in America. Finding this efficient practice is simple and obvious.
It is less obvious how to create an efficient way of dealing with harms from infection. Assuming low transaction costs, set property rights, and the ability to trade, an efficient outcome can be created by bargaining, per Coase's theorem. Simply apportioning all blame for an infection to the infector skips the bargaining part, implicitly assuming that everyone has the right not to be infected. But this is not how societies have been dealing with infections in the past. Intentionally exposing yourself to syphilis by visiting brothels did not constitute a basis to seek redress from the brothel. In many specific situations laws would assume a duty of self-preservation and see being infected as the result of negligence on the part of the infectee, rather than a tort committed by the infector. This is reasonable, since in many situations it is easier to for a potential infectee than for a potential infector to modify behavior to avoid harms from infection.
Any attempt at improving our society's way of dealing with infection should involve a way of identifying efficient practices, including efficient apportionment of duties, before additional methods (law, insurance premiums, etc.) are used to make sure those practices are followed.
How would you try to find such efficient practices?
A lot of Zvi's COVID posts talk about where our current institutions fall short of basic competency (with him frequently repeating "FDA Delenda Est") yet he has a consistent undertone of "none of these reforms are inside the Overton Window, and trying to persuade people to make better institutions is pointless". I mean, prediction markets are a significantly more powerful forecasting method, which continues to be illegal. I would be shocked to hear that anyone in public office actually started to overhaul institutions, or even just stopped shooting ourselves in the foot. See COVID test regulations, or taking away approval for J&J. The one time that people actually listened to cries for reform, everyone involved was surprised. I wish we lived in the world where people pointing out inadequacies in our current institutions led to actual change, but it really doesn't seem like we do. Not that theorizing over potential better institutions is pointless, as sometimes people are able to actually implement them (which is one good thing about the crypto economy, in that it wasn't constrained by things like anti-prediction market regulations), but it just doesn't seem like there's any chance of actually reforming the FDA, or CDC, or other government institutions that desperately need reform.
I didn't really anticipate the scale of the wave of fear and panic that accompanied the pandemic. Attempts to control the virus mostly proved futile - resulting in minor delays at best. The pandemic was like a dress rehearsal - it was a mild pathogen and we got off very lightly. However, our collective response to the situation was terrible. Probably, attempts to improve outcomes next time should focus quite a bit on that aspect. Attention-seeking fearmongering should have consequences. We need better institutions to tell us who are credible experts on matters of public interest - and so on.
Apparently, 'omicron parties' are now a thing. Now that we have some milder variants, some people who are not inclined towards vaccination are finding some alternative approaches.
Variolation everywhere: intensively sequence every unexpected infection so we know exactly what's going around. Everybody gets a personalized regular cocktail of variolated exposure to everything going around, subject to personal risk factors and everything we've learned about optimal exposure dose
Color-coded geographical regions: anything goes in the red zones. Yellow has moderate controls. Red zones are strictly border controlled, monitored and quarantined. Everyone gets a housing voucher for the duration based on their current housing costs and free storage unit for their junk. Choose your zone depending on your risk tolerance. Only yellow zone actually bears much inconvenience -- green zone can act like new zealand: tight border controls and almost no disease or restrictions internally. Bonus: people usually do better when they shake up their surroundings anyway
Structure colonies: (goal: break through all the crappy nutrition/diet/medicine studies by getting actual high qualit data). Like a really intensive medical study. Voluntary participants, mostly people who really value externally imposed structure/meaning and don't like to provide for themselves. All your needs are provided for, but you eat/exercise/work exactly according to the dictates of the ongoing experiment, with intensive monitoring of everything you do. City-sized, with everyone participating.
Pandemic archipelago: devolve regulation of medicine and the right to travel across state boundaries to the states. Fed role is to track policies, compliance and results and maintain a public leaderboard of state performance.
Bounty medicine: large bounty offered for effective vaccines/treatments. 15% of awarded bounties get divided up among study participants. All other bio-ethics rules suspended
“And if we have spit samples from two people infected with covid, we can compare the DNA in their viruses to see if they match. “
Can anyone provide a rough estimate of the probability that two random COVID sufferers in the same area have viruses with indistinguishable DNA, if neither infected the other?
For law to work for assault, theft, or car accidentsWhich is less and less the case these days. The spike in murder is now getting admitted, but the rise in car accidents despite fewer man-hours driving is really nuts.https://www.unz.com/isteve/...https://www.slowboring.com/...
Regarding the system you mention I think it's really important to keep in mind that, while these measures look neutral, they have substantial distributional effects. It costs some people much more utility to avoid going out (perhaps their jobs) when they fear they may be sick while others can spend a day working on their computer from home.In theory this isn't a problem. Simply use tax dollars to pay those whose probability of having infected others is sufficiently low rather than fining (or some balance of the two) those who do infect others.
However, I expect part of the reluctance with such schemes is because we intuit that this ideal is unlikely to ever really be realized. Moralizing, resentment and our ingrained expectation that we shouldn't have to give someone some of our money merely for not doing a bad thing. Politics tends to solve problems particularly poorly and once you've given a government the authority to intervene in some way they rarely give it up and the net result is the government gets more and more nobs which they turn based on what people want to signal about not what's fair.
In other words, I fear that in the long run the cost from inserting more and more kinds of regulation into the overton window of acceptability will dominate any short term benefits.
On the issue of data I've had the opposite journey. When I was young, it seemed like the data was king. The data was the thing to be explained and it was the job of theory to bend to accommodate it. I loved the fact that data would show you this crazy thing about human nature was true despite it's absurd, counterintuitive nature.
Over time I've come to the exact opposite conclusion. When we try to figure out almost problem in economics that's relevant to live policy questions the data is so messy and allows so many potential confounders that it offers little direct guidance at all. Take a question like the effects of the minimum wage or the effects of rent control. You can't do an RCT and it's easy for forking paths, bad choice of model or an uncontrolled for confounder to give wrong results even when studying a physical system. But physical systems never try and manage their reputation and shift how they react based on what people believe about them.
So, on issues like the minimum wage, I've mostly come to the conclusion that the data only convince me that the system's noisy and complex but the theory gives me compelling evidence that, for sufficiently large minimum wage hikes there must be a loss of jobs or benefits or something and suggests what that curve must look like at a sufficiently large step size and that at least gives me good reason to match the mean of my probability distribution to that curve. Sure, my probability distribution has high variance (if I was a professional economist maybe I'd be able to do somewhat better) but it's way more useful than what I could get just pouring over a bunch of studies all making different assumptions.
Ideally, we apply theory to the questions we actually care about but then build a sufficiently extensive theoretical picture that you can then test little pieces of it in very simple hard to screw up experiments. You can't ignore the data but it's only the theory that can hope to deal with the complexities of the real world or guide you to the experiments which are simple enough to raise or lower your confidence in the theoretical framework.
You do not get ranting rights until after you are retired for a few years, and no one listens to you anymore. And you get double ranting rights if no one listened to you, even before you retired.
Oops, you're right. Thanks for pointing that out.
Existing liability law richly considers all the parties one could blame and how best to blame them all. Pandemic liability law can do similarly.
Correct me if I am wrong but you seem to propose that the liability for harms stemming from infecting somebody should fall on the infector, without apportioning some or all of the liability on the infectee. If so, this would produce inefficient outcomes.
Rationally we value a social control system/policy by the efficiency of the results it produces, compared to alternatives. Usually it is possible to identify a set of practices that are efficient, and then additional control mechanisms, such as laws and punishments, are created to encourage such practices. For driving it is efficient to select one side of the road and have either left or right-hand driving, and punish driving on whichever side goes against the grain, which means that in Britain you will be punished for driving on the side that is legal in America. Finding this efficient practice is simple and obvious.
It is less obvious how to create an efficient way of dealing with harms from infection. Assuming low transaction costs, set property rights, and the ability to trade, an efficient outcome can be created by bargaining, per Coase's theorem. Simply apportioning all blame for an infection to the infector skips the bargaining part, implicitly assuming that everyone has the right not to be infected. But this is not how societies have been dealing with infections in the past. Intentionally exposing yourself to syphilis by visiting brothels did not constitute a basis to seek redress from the brothel. In many specific situations laws would assume a duty of self-preservation and see being infected as the result of negligence on the part of the infectee, rather than a tort committed by the infector. This is reasonable, since in many situations it is easier to for a potential infectee than for a potential infector to modify behavior to avoid harms from infection.
Any attempt at improving our society's way of dealing with infection should involve a way of identifying efficient practices, including efficient apportionment of duties, before additional methods (law, insurance premiums, etc.) are used to make sure those practices are followed.
How would you try to find such efficient practices?
But it crosses a social norm and I feel that once that line is breached there will be less resistance.
FYI, the Wikipedia "prediction markets" article credits their invention to Friedrich Hayek (1945) and Ludwig von Mises (1920).
A lot of Zvi's COVID posts talk about where our current institutions fall short of basic competency (with him frequently repeating "FDA Delenda Est") yet he has a consistent undertone of "none of these reforms are inside the Overton Window, and trying to persuade people to make better institutions is pointless". I mean, prediction markets are a significantly more powerful forecasting method, which continues to be illegal. I would be shocked to hear that anyone in public office actually started to overhaul institutions, or even just stopped shooting ourselves in the foot. See COVID test regulations, or taking away approval for J&J. The one time that people actually listened to cries for reform, everyone involved was surprised. I wish we lived in the world where people pointing out inadequacies in our current institutions led to actual change, but it really doesn't seem like we do. Not that theorizing over potential better institutions is pointless, as sometimes people are able to actually implement them (which is one good thing about the crypto economy, in that it wasn't constrained by things like anti-prediction market regulations), but it just doesn't seem like there's any chance of actually reforming the FDA, or CDC, or other government institutions that desperately need reform.
I didn't really anticipate the scale of the wave of fear and panic that accompanied the pandemic. Attempts to control the virus mostly proved futile - resulting in minor delays at best. The pandemic was like a dress rehearsal - it was a mild pathogen and we got off very lightly. However, our collective response to the situation was terrible. Probably, attempts to improve outcomes next time should focus quite a bit on that aspect. Attention-seeking fearmongering should have consequences. We need better institutions to tell us who are credible experts on matters of public interest - and so on.
Apparently, 'omicron parties' are now a thing. Now that we have some milder variants, some people who are not inclined towards vaccination are finding some alternative approaches.
Thinking about institutions is taboo now that we’re officially a non-foundationalist society. Actually, we have dialectic now so thinking is taboo.
Variolation everywhere: intensively sequence every unexpected infection so we know exactly what's going around. Everybody gets a personalized regular cocktail of variolated exposure to everything going around, subject to personal risk factors and everything we've learned about optimal exposure dose
Color-coded geographical regions: anything goes in the red zones. Yellow has moderate controls. Red zones are strictly border controlled, monitored and quarantined. Everyone gets a housing voucher for the duration based on their current housing costs and free storage unit for their junk. Choose your zone depending on your risk tolerance. Only yellow zone actually bears much inconvenience -- green zone can act like new zealand: tight border controls and almost no disease or restrictions internally. Bonus: people usually do better when they shake up their surroundings anyway
Structure colonies: (goal: break through all the crappy nutrition/diet/medicine studies by getting actual high qualit data). Like a really intensive medical study. Voluntary participants, mostly people who really value externally imposed structure/meaning and don't like to provide for themselves. All your needs are provided for, but you eat/exercise/work exactly according to the dictates of the ongoing experiment, with intensive monitoring of everything you do. City-sized, with everyone participating.
Pandemic archipelago: devolve regulation of medicine and the right to travel across state boundaries to the states. Fed role is to track policies, compliance and results and maintain a public leaderboard of state performance.
Bounty medicine: large bounty offered for effective vaccines/treatments. 15% of awarded bounties get divided up among study participants. All other bio-ethics rules suspended
“And if we have spit samples from two people infected with covid, we can compare the DNA in their viruses to see if they match. “
Can anyone provide a rough estimate of the probability that two random COVID sufferers in the same area have viruses with indistinguishable DNA, if neither infected the other?
For law to work for assault, theft, or car accidentsWhich is less and less the case these days. The spike in murder is now getting admitted, but the rise in car accidents despite fewer man-hours driving is really nuts.https://www.unz.com/isteve/...https://www.slowboring.com/...
Allowing particular kinds of lawsuits, such as re someone infecting you, does NOT authorize all sorts of random regulations.
Regarding the system you mention I think it's really important to keep in mind that, while these measures look neutral, they have substantial distributional effects. It costs some people much more utility to avoid going out (perhaps their jobs) when they fear they may be sick while others can spend a day working on their computer from home.In theory this isn't a problem. Simply use tax dollars to pay those whose probability of having infected others is sufficiently low rather than fining (or some balance of the two) those who do infect others.
However, I expect part of the reluctance with such schemes is because we intuit that this ideal is unlikely to ever really be realized. Moralizing, resentment and our ingrained expectation that we shouldn't have to give someone some of our money merely for not doing a bad thing. Politics tends to solve problems particularly poorly and once you've given a government the authority to intervene in some way they rarely give it up and the net result is the government gets more and more nobs which they turn based on what people want to signal about not what's fair.
In other words, I fear that in the long run the cost from inserting more and more kinds of regulation into the overton window of acceptability will dominate any short term benefits.
On the issue of data I've had the opposite journey. When I was young, it seemed like the data was king. The data was the thing to be explained and it was the job of theory to bend to accommodate it. I loved the fact that data would show you this crazy thing about human nature was true despite it's absurd, counterintuitive nature.
Over time I've come to the exact opposite conclusion. When we try to figure out almost problem in economics that's relevant to live policy questions the data is so messy and allows so many potential confounders that it offers little direct guidance at all. Take a question like the effects of the minimum wage or the effects of rent control. You can't do an RCT and it's easy for forking paths, bad choice of model or an uncontrolled for confounder to give wrong results even when studying a physical system. But physical systems never try and manage their reputation and shift how they react based on what people believe about them.
So, on issues like the minimum wage, I've mostly come to the conclusion that the data only convince me that the system's noisy and complex but the theory gives me compelling evidence that, for sufficiently large minimum wage hikes there must be a loss of jobs or benefits or something and suggests what that curve must look like at a sufficiently large step size and that at least gives me good reason to match the mean of my probability distribution to that curve. Sure, my probability distribution has high variance (if I was a professional economist maybe I'd be able to do somewhat better) but it's way more useful than what I could get just pouring over a bunch of studies all making different assumptions.
Ideally, we apply theory to the questions we actually care about but then build a sufficiently extensive theoretical picture that you can then test little pieces of it in very simple hard to screw up experiments. You can't ignore the data but it's only the theory that can hope to deal with the complexities of the real world or guide you to the experiments which are simple enough to raise or lower your confidence in the theoretical framework.
Dear Young Punk:
You do not get ranting rights until after you are retired for a few years, and no one listens to you anymore. And you get double ranting rights if no one listened to you, even before you retired.