I am not sure of the relative color gradient of the swan, as it's been a bit since I read Taleb, but I think this was a separate metaphor. The point I was making was that "make every positive E(v) move" is actually a really bad strategy for survival as Taleb has pointed out. E(v) is simply too flat a metric to make a good rule of thumb l…
I am not sure of the relative color gradient of the swan, as it's been a bit since I read Taleb, but I think this was a separate metaphor. The point I was making was that "make every positive E(v) move" is actually a really bad strategy for survival as Taleb has pointed out. E(v) is simply too flat a metric to make a good rule of thumb like that, because the nature of the downside risk matters a great deal. E(v) also doesn't capture the effects of multiple actions; as you say, the too much and too fast aspect is a problem, but no E(v) calculus is going to get you to that, certainly not one done based on individual actions. The amount of risk is not independent across the actions.
We do agree in the general case that you are correct about Taleb and E(V). Picking up pennies in front of a steam roller is a bad idea.
And Grey Swan I’m not even sure is Taleb’s idea. But I think it applies re culture here.
I’m just saying that for each incremental immigrant allowed in, the Black Swan reasoning does not apply, and BC’s E(V) reasoning does. (Unless your beliefs are “even a single life is too much…”, in which case the logic applies, but that is no Black Swan - that is simply the reality of human nature).
Unless you take it to unlimited.
Because at less than unlimited, at any “reasonable” level, the risk is in fact independent. [Of course, reasonable people can argue about what a “reasonable” level is…]
Taleb’s Black Swan point - as illustrated wonderfully by the steam roller and turkey analogies - does not require going to the extremes of unlimited.
I disagree that the risks are independent. Letting in one immigrant family to a city is different from 100,000 immigrant families to the same city. Where exactly it goes from “the family adopts the culture of the new country” to “the families make their part of the new country just like the old country” is debatable, but clearly it becomes more likely the larger the number in a networked effect. So while any given immigrant taken alone might be positive E(v) that doesn’t mean that 100,000 of that exact same person would be. At a certain level of change per year the culture of the host country changes instead of the culture of the immigrants, and the high immigration argument relies on immigrants assimilating.
At any rate, E(v) per individual doesn’t capture that.
In fact we are in complete agreement that depositing 100,000 immigrant families in a single city is problematic.
For the cultural reasons you and I are each describing.
Because depositing 100K families in any city - save perhaps NYC or L.A. - is not “reasonable”.
As opposed to 100K such families scattered throughout the state of CA, which would indeed be reasonable. While 10M such families in a single year across the state would *not* be reasonable.
But the problem we are agreeing on here is not one of “Black Swan” reasons, which was the other of my two initial points.
To try to explain by analogy, “Black Swan” logic says it’s stupid to attempt to pick up even 1 or 2 or 3 pennies in front of a steamroller, precisely because the E(V) of *each attempt* for any sane person is negative. It is Russian roulette.
But for immigration we have already agreed that at the beginning and for a long while, the E(V) is indeed positive. Even if no two of us might agree on exactly how much of such legal immigration each year is safely positive.
It is only if taken too far that there are possible/likely problems.
I think we are talking past each other. I am not making a Black Swan argument, but Taleb's other point that even positive E(v) risks are bad if the downside is "you go broke". He talked about it a lot with regards to his barbell strategy of investing, and how many investors do really well for a few years with a strategy but then the downside happens and they lose everything and have to stop being investors. It is a bad strategy not because the E(v) of any individual trade is negative, but because the upsides gained over time can be completely wiped out by a bad outcome. It isn't a Black Swan because you know the downside of e.g. highly leveraged investing is there; you can see the steam roller coming. You can get away with it if it is infrequent and a low percentage of your wealth so that the downside doesn't wipe you out, but it is hard to know where to draw that line. The line is definitely before "Take every positive E(v) trade", however.
The immigration question isn't 1:1 with this, but is similar. The difference is that there are network effects in the downside, where more immigrants over a given period of time are increasingly risky. As such, any given decision to hand out an H1b doesn't capture the risk, because it doesn't consider the total number. The difficulty there is that the network effects are extremely hard to estimate. 100k families in NYC is higher E(v) if they scatter and don't interact much, much lower E(v) if they form an enclave becoming a city within a city. 10k families will do the same, or even 1000. Where that line is is almost impossible to identify.
What's worse, you can't do anything to mitigate that risk. One can try and put quotas on people from various countries, hoping they won't coalesce into a tiny version of their home country, but that does mean that the ratios are going to be wildly different. It will look like blatant racism, and one might even have to limit by geographical region instead of country.
One can't control where people live once inside the country, so one can't make sure those 100k families are scattered all over CA. Plus, over time as people are added they will tend to merge into local groups. So that's out.
So yes, it is only if it is taken too far that there are problems, but no one can pin down what "too far" is until you start crap like the UK or Sweden are dealing with. Then the question becomes "can we fix this?"
I am not sure of the relative color gradient of the swan, as it's been a bit since I read Taleb, but I think this was a separate metaphor. The point I was making was that "make every positive E(v) move" is actually a really bad strategy for survival as Taleb has pointed out. E(v) is simply too flat a metric to make a good rule of thumb like that, because the nature of the downside risk matters a great deal. E(v) also doesn't capture the effects of multiple actions; as you say, the too much and too fast aspect is a problem, but no E(v) calculus is going to get you to that, certainly not one done based on individual actions. The amount of risk is not independent across the actions.
We do agree in the general case that you are correct about Taleb and E(V). Picking up pennies in front of a steam roller is a bad idea.
And Grey Swan I’m not even sure is Taleb’s idea. But I think it applies re culture here.
I’m just saying that for each incremental immigrant allowed in, the Black Swan reasoning does not apply, and BC’s E(V) reasoning does. (Unless your beliefs are “even a single life is too much…”, in which case the logic applies, but that is no Black Swan - that is simply the reality of human nature).
Unless you take it to unlimited.
Because at less than unlimited, at any “reasonable” level, the risk is in fact independent. [Of course, reasonable people can argue about what a “reasonable” level is…]
Taleb’s Black Swan point - as illustrated wonderfully by the steam roller and turkey analogies - does not require going to the extremes of unlimited.
I disagree that the risks are independent. Letting in one immigrant family to a city is different from 100,000 immigrant families to the same city. Where exactly it goes from “the family adopts the culture of the new country” to “the families make their part of the new country just like the old country” is debatable, but clearly it becomes more likely the larger the number in a networked effect. So while any given immigrant taken alone might be positive E(v) that doesn’t mean that 100,000 of that exact same person would be. At a certain level of change per year the culture of the host country changes instead of the culture of the immigrants, and the high immigration argument relies on immigrants assimilating.
At any rate, E(v) per individual doesn’t capture that.
In fact we are in complete agreement that depositing 100,000 immigrant families in a single city is problematic.
For the cultural reasons you and I are each describing.
Because depositing 100K families in any city - save perhaps NYC or L.A. - is not “reasonable”.
As opposed to 100K such families scattered throughout the state of CA, which would indeed be reasonable. While 10M such families in a single year across the state would *not* be reasonable.
But the problem we are agreeing on here is not one of “Black Swan” reasons, which was the other of my two initial points.
To try to explain by analogy, “Black Swan” logic says it’s stupid to attempt to pick up even 1 or 2 or 3 pennies in front of a steamroller, precisely because the E(V) of *each attempt* for any sane person is negative. It is Russian roulette.
But for immigration we have already agreed that at the beginning and for a long while, the E(V) is indeed positive. Even if no two of us might agree on exactly how much of such legal immigration each year is safely positive.
It is only if taken too far that there are possible/likely problems.
I think we are talking past each other. I am not making a Black Swan argument, but Taleb's other point that even positive E(v) risks are bad if the downside is "you go broke". He talked about it a lot with regards to his barbell strategy of investing, and how many investors do really well for a few years with a strategy but then the downside happens and they lose everything and have to stop being investors. It is a bad strategy not because the E(v) of any individual trade is negative, but because the upsides gained over time can be completely wiped out by a bad outcome. It isn't a Black Swan because you know the downside of e.g. highly leveraged investing is there; you can see the steam roller coming. You can get away with it if it is infrequent and a low percentage of your wealth so that the downside doesn't wipe you out, but it is hard to know where to draw that line. The line is definitely before "Take every positive E(v) trade", however.
The immigration question isn't 1:1 with this, but is similar. The difference is that there are network effects in the downside, where more immigrants over a given period of time are increasingly risky. As such, any given decision to hand out an H1b doesn't capture the risk, because it doesn't consider the total number. The difficulty there is that the network effects are extremely hard to estimate. 100k families in NYC is higher E(v) if they scatter and don't interact much, much lower E(v) if they form an enclave becoming a city within a city. 10k families will do the same, or even 1000. Where that line is is almost impossible to identify.
What's worse, you can't do anything to mitigate that risk. One can try and put quotas on people from various countries, hoping they won't coalesce into a tiny version of their home country, but that does mean that the ratios are going to be wildly different. It will look like blatant racism, and one might even have to limit by geographical region instead of country.
One can't control where people live once inside the country, so one can't make sure those 100k families are scattered all over CA. Plus, over time as people are added they will tend to merge into local groups. So that's out.
So yes, it is only if it is taken too far that there are problems, but no one can pin down what "too far" is until you start crap like the UK or Sweden are dealing with. Then the question becomes "can we fix this?"