Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Suspicious Discontinuities

danluu.com

Post reply on HN
Print view View original

161–170 of 307 posts

Re: Suspicious Discontinuities

#161

Post by peteretep »

iainmerrick wrote: post #159
Dylan16807 wrote: post #129

Earlier quoted context omitted.

The exact same way as the new system, except instead of 63% it's a much smaller number. Edit: Why are people downvoting this? I don't understand what's so confusing about SilasX's argument. Removing cliffs and changing it to a smooth dropoff is good. But that dropoff doesn't have to be £.63 in benefits per £1 earned. That's a pretty big effective marginal tax rate! Why not £.50 or £.30? It should be extremely easy to…

I totally agree, but I just realised -- there’s another cliff edge here! I surveyed opinions on tax brackets a while back, and it seemed like there was a big psychological barrier at the 50% rate. Most high earners would be willing to pay a 50% marginal tax rate (albeit some very grudgingly) but almost nobody was willing to go over 50%. What’s good enough for the rich should do for the poor as well. The benefit drop…

In the UK there's already a higher-than-50% marginal tax rate between £100 and £120k because you lose your personal tax allowance.

Re: Suspicious Discontinuities

#162

Post by Dylan16807 »

peteretep wrote: post #156
Dylan16807 wrote: post #150

Earlier quoted context omitted.

You said you didn't want to "unilaterally extend benefits to a whole bunch of people", not that the cost couldn't increase even the tiniest amount. But even so, smoothing out the curve some might be better overall even if the max amount drops. Also you could extend the curve so that 'benefits' go negative at the high end. Now your budget is fine...

Your plan both massively increases the number of people on benefits and also increases costs by 20% on your 50% proposal, which in the UK means £15b-ish. Reconciling this with "even the tiniest amount" is impossible. > Also you could extend the curve so that 'benefits' go negative at the high end You've just invented tax, congratulations.

> massively increases the number of people on benefits

You have a very different definition of 'massively' than I do.

> You've just invented tax, congratulations.

Positive taxes can balance out negative taxes, ~tada~.

Re: Suspicious Discontinuities

#163

Post by friendlybus »

toomuchtodo wrote: post #58
userbinator wrote: post #56

Earlier quoted context omitted.

who expect the dating and sexual marketplace to be some sort of meritocracy, or equality to be exhibited Why/how would people come to think that way? As the parent comment says, this is fundamentally a biological thing.

I will leave detailed explanations to domain experts. An example I'd hold up would be when Leonardo DiCaprio catches flak, loudly and publicly, for consistently dating women only in their early to mid 20s [1] [2] [3]. Is it biological? Yes. Is it legal? Absolutely. Do people still vocally decry it? Very much so. [1] https://www.insider.com/leonardo-dicaprio-girlfriends-reddit... [2] https://www.reddit.com/r/dataisbea…

When does he catch flak? Leo is one guy, you could list a whole bunch of ceos and famous people's private lives as counter-examples. Bill gates has been married to Melinda forever.

That story is an advertisement for Leonardo, not a fundamental truth about human sex lives.

Re: Suspicious Discontinuities

#164

Post by Dylan16807 »

iainmerrick wrote: post #159
Dylan16807 wrote: post #129

Earlier quoted context omitted.

The exact same way as the new system, except instead of 63% it's a much smaller number. Edit: Why are people downvoting this? I don't understand what's so confusing about SilasX's argument. Removing cliffs and changing it to a smooth dropoff is good. But that dropoff doesn't have to be £.63 in benefits per £1 earned. That's a pretty big effective marginal tax rate! Why not £.50 or £.30? It should be extremely easy to…

I totally agree, but I just realised -- there’s another cliff edge here! I surveyed opinions on tax brackets a while back, and it seemed like there was a big psychological barrier at the 50% rate. Most high earners would be willing to pay a 50% marginal tax rate (albeit some very grudgingly) but almost nobody was willing to go over 50%. What’s good enough for the rich should do for the poor as well. The benefit drop…

What does 'willing to pay' mean, exactly? Surely they're not turning down a raise out of stubbornness. And pay of that caliber is not often hourly.

Re: Suspicious Discontinuities

#165

Post by fyp »

Another real world example I saw recently was in programming competitions.

There's a tradition of displaying a color for your username based on your ELO rating (for example people would brag about being Red or Yellow on topcoder).

So people tend to stop competing forever after hitting the next level just to preserve their color!

Graph distribution of ratings: https://codeforces.com/blog/entry/52470

Re: Suspicious Discontinuities

#166

Post by dragonwriter »

iron0013 wrote: post #84
thrower123 wrote: post #78

Earlier quoted context omitted.

Government paternalism is terrifying. People are going to screw up, and make bad choices, but having choices is better than being treated like a child your entire life.

Speaking of children, that’s who’s gonna starve if you decide to let them suffer the consequences of their parents’ poor choices. I would have personally starved to death dozens of times over if not for food stamps.

Food stamps don't prevent children from suffering the consequences of their parents poor choices; people can and do trade those benefits for other things.

Children also get to suffer for policymakers poor choices of what the stamps should be spent on, and taxpayers get to suffer the cost of every benefit program duplicating income verification functions and other eligibility functions rather than having universal programs where the tax system takes care of the better off not getting unfair share of benefit.

Re: Suspicious Discontinuities

#167

Post by glogla »

tjr225 wrote: post #77
dragonwriter wrote: post #75

Earlier quoted context omitted.

> The perverse incentives around these discontinuities is one of the worst misfeatures of programs intended to help people. This is a key motivation for replacing many of those programs with UBI funded by progressive taxes.

Either that or a more progressive stance on socialist policies that provide basics to everyone. At least with socialist style policies there isn't the possibility of those spending their UBI on junk food or wine our whatever, but I really would prefer either to the current state of affairs.

I disagree strongly with your idea of "poor people would just spend on drugs" but there's a nugget of reason in it - there's a high risk that UBI without rent control would just be subsidizing the landlords, without actually helping the people.

Re: Suspicious Discontinuities

#168

Post by Fnoord »

umvi wrote: post #113
iron0013 wrote: post #84

Earlier quoted context omitted.

Speaking of children, that’s who’s gonna starve if you decide to let them suffer the consequences of their parents’ poor choices. I would have personally starved to death dozens of times over if not for food stamps.

You might have been more hungry, but death by starvation in the USA is nearly unheard of. I personally have never heard of anyone - homeless or otherwise - dying of starvation in the USA. The only people I've heard of dying of starvation were stranded on a mountain or on the ocean or something.

Starvation isn't literal in this context. One can intellectually starve a child. One can have a child be stagnant in their growth, for example by not having them play.

Re: Suspicious Discontinuities

#169

Post by gwd »

stared wrote: post #36

Ad Polish end of high school exams: quite a few years ago I created an interactive widget to explore it https://github.com/stared/delab-matury . There are a few jokes about the Matural Distribution (Matura is the name of this exam) or the Polish Gaussian. It is the Polish language, which is the most discontinuous. However: - There are quite a few scores for the essay which are subjective (e.g. on the style). - It is…

> It is an official procedure that in the case of results on the edge, the person marking should mark an ambiguous score in the favour of the student.

The funny thing about the "almost there" argument is that what it means is the actual line is more fuzzy, but a bit lower. That is, if the "official" cut-of is 30, and teachers are willing to "find an extra point" for people with scores of 28 and 29 but not 27, then the actual cut-off is 27.

I had a prof at uni who for programming projects would say: "The deadline is 6pm on Friday. But there's a 6-hour grace period where you're allowed to be late." Inevitably someone in the class would say, "Doesn't that mean that the actual deadline is at midnight?" To which he would respond, "Yes, of course. But if someone submits a project at 12:01 and it's rejected, when they say 'But I was only 1 minute late!', I can say, 'No, you were 6 hours late.'"

Re: Suspicious Discontinuities

#170

Post by fizixer »

Off-topic: I'm looking at turnout histograms and I'm having the hardest time making sense of it. Why is registered voter count on the y-axis instead of x-axis, since that should be the independent variable?

Then I realized/recalled that is what histograms do. They flip the plot between the two axis. The original, non-histogram, plot is really turnout (y-axis) vs registered-voters. But that would look very different (generally such plots are very noisy, with a running-average that mimics a horizontal line). Then, if you wish to know what's the distribution of registered-voters with respect to turnout, you take a histogram.

edit: on second thought, the plot still doesn't make sense. The y-axis should be # of polling stations, not # of registered-voters.

edit 2: This article [0] has a plot that confirms my suspicion (y-axis is supposed to be number/frequency of polling stations, not number of registered voters).

[0] https://doi.org/10.1111/j.1740-9713.2018.01141.x

Post reply on HN