Sunday assorted links

Comments

1. Ban seat belts.

1.a. Make airbags optional.
1.b. Abolish roll over standards.
(Although the resulting increase in driver visibility may save more pedestrians, so it may be a net wash)

Now let's look at an illustration of this an opinion opener.

Just recently my self began to strengthen. They
may be running, cycling, or cross-country ski-ing, or
machines that mimic those individuals.

Replace selective service registration with a harvest lottery.

As one comedian said, put a metal spike on the steering wheel. No one will ever tailgate again.

1. The simple change being to expand a taxpayer funded program? Who knew that Prof. Cowen endorses the idea of taxpayer funded programs to help Americans reduce health care costs - 'Luckily, there’s a group that helps people with the travel costs associated with donating. It’s called the National Living Donor Assistance Center (NLDAC), and it’s funded by the federal Department of Health and Human Services (HHS), which administers Medicare.'

Though considering that the well-off don't currently benefit from this program might explain Prof. Cowen's interest - 'The problem is not a lack of funding; NLDAC actually has more funding appropriated to it than it can spend. There are a few reasons for that. One is that only donors in cases where both the donor and recipient earn less than three times the poverty line can receive help. That was why I didn’t get any money. The income cap would be a reasonable limitation if NLDAC were budget constrained, but at the moment, it’s not.'

3. Are you sure they don't agree? Looks like only 7% of the panel disagreed that ranked choice is a better system. The comments all seem to agree that while it's not perfect, it better reflects voter preferences, but that it is potentially too complicated for voters.

Maybe Tyler's joke is at the economists didn't use ranked-choice voting to rate ranked-choice voting. They used some sort of confidence weighted voting instead.

This was a trivial ranking. The question compared two candidates: ranked choice voting and runoff election. With only two candidates, the rankings would be the same using any voting method.

But RCV and Plurality with Runoff are not mutually exclusive. The former is a method of voting. The latter is a method of choosing a winner.

Perhaps Tyler was reacting to the total absence of any of Trump's forgotten men.

Those who live where cutting costs is the number one priority to create wealth.

As in Maine, RCV is clearly a plot by leftists to steal elections from them.

Which is ironic given 1992 would have been a win for them if RCV had been used for president. Clinton was the least loser, HW, the second biggest loser, with Perot the 3rd biggest loser. Voting rounds had eventually eliminated Perot, would Clinton have exceeded 50%?

Approval voting is far superior because the system focuses on concensus, not focusing on creating losers. Those in favor of RCV do so because they want to vote for an obvious loser while making sure one candidate loses. Ie, in 2000, voting against the Democrat but making sure GW Bush loses. In 2016, voting against Clinton, but voting to ensure Trump loses.

In the GOP primaries, two-thirds disapproved of Trump, but there was no agreemennt on which of the candidates 80% approved of was the first on the list to beat Trump. For many, Trump was the obvious loser, so a vote for Trump signalled disapproval of the winnowed field as candidates they approved of dropped out. Imagine the results if primary voters cast votes for as many of the 12-15 down to 8 or so candidates on the ballot in 2016. Any doubt that Bush wouldn't have gotten approval from half the voters, even though he wasnt their first choice? A Trump vote was a blow up the system protest vote for many, a universal none of the above.

Think of voting for the single diner for your town. Chinese? Mexican? German? Italian? Family aka midwest meat and potatoes. If the town is shrinking and diners closing, its approval voting that likely picks the family diner which mixes up the meat and potatoes with sauces on the table. Note, thanksgiving and christmas is basically meat and potatoes except for those who go Chinese.

You're correct that Approval voting would have been more efficient in the 17 candidate GOP primary. But Approval voting is less efficient than Plurality with Runoff when there are only three candidates.

You can't make any inferences about the 1992 election. No one has the data to do so.

I suspect if you put it to a democratic vote you could claim agreement, as we do in political elections....

That said, Hart's and Hall's statements are perhaps informative. One might ask on what basis "the 'wrong' two" might have risen to the top?

It's been used in Australia for over a century and in Ireland for decades. It's not too complicated for voters given some run-up public service campaigns. The problem is that elections administration in this country is done with the left hand and innovations commonly make problems worse. Our state and federal legislatures need to get over their chronic case of the stupids.

That was my reaction too, that might be greater agreement than economists can find on any issue outside of comparative advantage and marginal cost.

Some of the anti statements suggest that a little research would be informative: how often does it make a difference, when is it better and when is it worse than plurality voting, runoffs, etc.? If it works in Australia and Ireland, those are some useful data points. OTOH, I was very unimpressed with the reasoning of the anti statement by the guy who said our system's been working fine for two centuries so don't change it.

In general, I think the question is being asked of the wrong people. Economists can recite Arrow's Impossibility Theorem by heart but the best answers to the question require some empirical research, and economists are going to have much less familiarity with the evidence than political scientists.

You can also use discrete optimization to boost kidney donation success rates: https://youtu.be/2IbJf4oXOxU?t=627

And I'd imagine Amazon-like drone networks will only streamline the process further.

5. I have been reading from The Biological Brain, by Alan Jasanoff. It isn't so much a big idea book as an anti big bad idea book. The big bad idea is that "brains" can be "simplified" and thought of in terms of the technology of the day. Be that telephone networks, digital circuits, or more elaborate computation engines. Brains produce answers, yes, but in a wet and squishy biological way. A very complicated biological way, which should give pause to anyone who thinks their behaviors can easily be simulated. Maybe simulated someday. But not easily, and not soon.

It’s “The Biological Mind”. I skimmed it and wasn’t impressed, for reasons somewhat similar to your insightful comment: namely that it’s an “anti” book . I’m always leery of books blurbed by hyper-partisan types like George Lakoff who is hardly a neurophilosopher. I’m more — way more — excited about Robin Hanson’s book “The Elephant in the Brain.”

“The Biological Mind” - I blame RNA.

Lol. Btw, have you looked at Hanson’s book?

Based on reviews, I consider the core ideas pretty conventional and not wrong. Things I've read by António Damásio, Steven Pinker, William Calvin, and Frans de Waal point the same way. We can could also go back to Oliver Sachs. But thid is not to say political extrapolations all point the same way. :-)

People on different points of the political compass are comfortable with different parts for a general human nature.

Dan Ariely, Daniel Gilbert, Lee Dugatkin.

Simulating how a biological brain produces behaviors is tricky. But simulating behaviors... Well, not many people work on solar panel production lines these days.

We do not have to simulate a "biological brain" to produce realistic utility distributions. We can start with simple approximations, for instance putting candidates and voters on a left-to-right political axis, and assigning utilities based on voters' distance to each candidate. We can make this more realistic by doing multiple dimensions with Gaussian, bimodal, and other types of distributions. It turns out that the more "realistic" you make utility distributions, the more they just look random. More impressively, vast changes to utility distribution makes very little difference in outcomes, and the results are so robust as to leave a huge amount of room for error. See the Bayesian regret chart I linked to.

Simulating behaviors is also incredibly straightforward. For instance, in Plurality Voting, an honest voter who prefers the Green will vote for the Green. The tactical version of that voter will vote for the preferred major party candidate, namely the Democrat. Bayesian regret doesn't actually use (or need to use) party labels per se. It just treats certain candidates as the "frontrunners", which has the same effect as imaging them to be the "Democrat and Republican" (to use American politics). The results remain quite consistent regardless of whether we have 100% honest voters, or 100% strategic voters, or any incremental mixture in between.

I would recommend you look over the procedure in detail if you want to craft anything approaching a robust criticism.

seriously dude, if you "start with simple approximations"
you are gonna fubar something

1. The concern is that a wage benefit for a kidney donor will be used as a subterfuge for paying the donor for her kidney. "One simple change" might refer to paying the donor for her kidney. Why not a market in body parts. After all, we have a market in death (it's called "life" insurance, but it's actually death insurance). Sure, there are restrictions on recovery of death insurance (the beneficiary must have an "insurable interest") to prevent the macabre (death insurance salesmen with a booth at the hospital), but a market in body parts would save lives while death insurance rewards death.

4. Guess who the tiger represents. Readers didn't quite get my prior comment on Brooks's Sidney Award to Brian Phillips for his essay on tigers. I will repeat the quote from the essay: "The arrival of a tiger, it’s true, is often preceded by moments of rising tension, because a tiger’s presence changes the jungle around it, and those changes are easier to detect. Birdcalls darken. Small deer call softly to each other. Herds do not run but drift into shapes that suggest some emerging group consciousness of an escape route.” Does the American herd include an emerging group consciousness of an escape route?

Young economics graduate student: missed the opportunity to report people to KKK, GPU, NKVD, or STASI ?
Here is a new chance for you: sign the AEA lynch petition.
Anonymous cowards welcome.

#2/#3 -- How could economists have time to agree on ANYTHING while they're so busy oppressing the marginalized and marginalizing the oppressed?

#2 -- semester after semester pretty young women manipulate male teachers with phony sexy-friendliness. If anything were to "happen," it's the student who'd have power over the teacher, not the teacher over the student.

It is a plea from the constantly aggrieved to force the administration to give them the power to destroy the careers of anyone who disagrees with them. Bake the effing cake or your life is over.

+1

These movements are a transparent power grab.

But somehow they are working. Did we learn nothing from the French Revolution?

"Did we learn nothing from the French Revolution?"

Keep the guillotines well oiled and ready for duty. Their time is almost nigh.

Just like that
poor man in
that book by Nabokov

*inhales*

Actual failed academic...or incel loser undergraduate ? Or even more likely, bitter divorced dude with a hatred of women?

*exhales*

#2. a)specifically, what is the difference between a challenging environment and a hostile one? b)it's obvious to me that doing "Noble Prize" level work and doing outstanding mentoring are not necessarily linked. How is this discrepancy addressed systematically in large academic departments (and if it's not, why not)? Should ever prof be a mentor? c) Does Callisto keep track of the number of times and number of people each complainant files against? That is relevant, isn't it? d) Some people can maintain adequate GPA`s but fail to have the competence & suitability expected of a holder of a specific advanced degree, especially from the more prestigious schools. How does the program filter/ direct those to careers paths with a better fit? Couldn't any such a process be considered "hostile"? e) Academics often treat those peers they disagree with poorly; how is this behavior compatible with the environment they create in their own groups? It is reasonable to expect we can change this (human nature)? f) is diversity like porn, we know it when we see it? The usual claim is that its rewards are greater than its costs, but how sure are we of that? (It's certainly not true in all circumstances, is it?) Isn't diversity a social goal and not infrequently in conflict with other academic goals? (Making this social engineering, with only a weak linkage to 'economics'). g) seems to me that the anonymous signers fail to understand that academics are expected to have some courage of their convictions. There's something rotten going on if so many felt the need for anonymity. Whether it's in the student's (lack of) character or in their institutions (or both) is the question.h)One size does not fit all, not all (American) Econ depts offering graduate degrees have the same administrative resources. i) Social skills can be taught (and possibly learned), and so can management/leadership skills. Seems to me that PhDs should all have been through a management/leadership skills course prior to getting their degree. (horse to water, I know). j) shouldn't all of a University's departments be held to the same diversity standards? (including Athletics and Feminist studies)

"b)it's obvious to me that doing "Noble Prize" level work and doing outstanding mentoring are not necessarily linked. How is this discrepancy addressed systematically in large academic departments (and if it's not, why not)? Should ever prof be a mentor?"

If you cannot advise students how are you supposed to be able to communicate to the profession your Nobel prize-winning work? The capacity to teach and the capacity to advise are linked to the capacity to communicate in academia and without the capacity for communication, it is almost impossible to make scientific contributions.

Communicating with someone who has extensive background in the subject matter and is a top level intellect is very different than “communicating” with a relative neophyte.

When you are doing nobel prize work you are doing something new. When you are doing something new it means that everybody else is a neophyte in what you are doing.

By the way, if you do not know how to explain your research to a 12 year old it means that you do not really understand it yourself.

Really? Explain put/call parity in terms a 12 year old can understand.

I would disagree. First of all, economists don't do Nobel prize work; they get a related prize (a memorial prize). That was just to help keep things a bit more focused in terms of terminology. I suspect your statement was more in the all these prized fall under a name that does include Nobel in the prize name. (see https://en.wikipedia.org/wiki/List_of_prizes_known_as_the_Nobel_of_a_field#Economics)

That said much of the work done by those awarded the memorial prize is not so much new, as very insightful extension of existing theory (new applications) or interesting additional to the larger body of work rather than completely new. As such the nuances will often escape the new learner and require a broad understanding of the underlying context of the philosophy/science.

However, that type of thinking and skills that allow such research and argument regarding the theory/argument is completely different from mentoring a younger mind. A good mentor can take that young, less mature, less informed and less disciplined mind and help it become the next generations prize winner. A poor mentor will likely inhibit such growth in most under his/her guidance. Both could be highly capable researchers and even hold such esteemed awards.

Summarized: apples and tatpoles being compared.

1. Kudos to them for actually conducting an evaluation to see if the change will be beneficial before implementing it. And here we have an article complaining about the fact that they aren't just implementing it nationally without checking the impact first.

A bit of history on 5 from wiki:

The Worm Runner's Digest (W.R.D.) was created in 1959 by biologist James V. McConnell[1] after his experiments with memory transfer in planarian worms generated a torrent of mail enquiries.

Worms dunnit!

say it loud
say it proud
the roads around here
are pretty well plowed

These economists aren't voting methods experts. For instance, one says:

"
We have too many voting impossibility theorems to state flatly that one method is better."

This is false old school thinking. Talking about various impossibility theorems and properties is like comparing race cars based on aerodynamics or horsepower. You instead want to put them on a racetrack and get a statistically significant sample of race times.

The voting methods analog of this is a technique called Bayesian regret, which allows us to measure the average voter satisfaction with election outcomes.

http://scorevoting.net/BayRegsFig.html

Those econ student names sound like the output of a random name generator app.

If you want to remove someone for any reason, accuse them of sexual misbehavior, or making you feel uncomfortable. If you want to be excused for anything, claim you were sexually molested as a child. Not to say some claims don't have merit. I robbed a bank because I was sexually molested by aliens when I was 2 years old.

Oops, I might have been inappropriate with that last comment. I didn't mean the kind of aliens that the "President" proposes to keep at bay with a wall while tearing innocent children away from their "families". I meant UFO space-aliens. (I hope that wasn't too inappropriate).

It is impossible to tell how hateful your space-alien comment is as we don't know where they fit in the patronage power structure.

They might be white male aliens or they might be brown LGBTs. But the protocol seems to be take whichever route leads to the greatest offence, and then change the subject later if it isn't warranted.

"Aliens" meant "foreigners" hundreds of years before it meant "extraterrestrials."

"Tearing innocent children away..."

Their parents are guilty of crimes. No one accused of a crime and detained gets to keep their children with them.

A gem from #2: "1. Listen to us.

We are the experts on how your colleagues treat their advisees and research assistants."

Another one: "Moreover, it is well documented that these abuses of power disproportionately harm women, minorities, and queer individuals."

Well documented. I am sure Hitler and Goebbels too documented their propaganda very well.

Here's a response to the various clueless economists who decided to opine on Ranked Choice Voting with virtually zero expertise.

https://medium.com/election-science/economists-disagree-on-instant-runoff-voting-b0bdde6a1b

This "expert" is as ignorant as the people he criticizes.

Ranked Choice Voting is a method of collecting data on individual preferences. It is NOT a method of choosing a victor. RCV data can be used to implement any of a number of collective choice methods including but not limited to: majority, plurality, plurality with runoff, Haare, Coombs, Approval, Disapproval, Condorcet.

The question was poorly worded and most of the respondents didn't fully understand either the question or the issues, so the results are mostly nonsense.

If by RCV the survey means "instant runoff," then this method is less costly than having a separate election. But respondents who said that there is a tradeoff of voter confusion may also be correct, although that belief is probably speculative. Ranking three candidates is a very simple exercise. Ranking more than three candidates is more complex, but largely because of a lack of information to distinguish between lower ranked candidates, not because of technical difficulty. While RCV might pose some technical difficulty, all of the selection methods implicitly assume that voter preferences are accurately collected.

So there is a conflation here between two entirely different issues:

1. How to collect data on voter preferences.
2. How to select a winner from among multiple candidates.

I'm confident though that these respondents viscerally understand these issues even if they can't or don't articulate them. From this we derive one and only one valuable piece of information from the survey: the majority of economists favor collecting the data on a single election day rather than on multiple days. And this result is rather obvious because it is less costly.

3. Yeah, they really do agree. Sixty percent agree that ranked choice voting is preferable to a runoff election. Of those who disagree or who are unsure, I think additional clarity would raise that 60% even higher.

Let's be clear: ranked choice voting (RCV) and plurality with runoff (PWR) are terms associated with two different things. RCV is a method of VOTING, I.e. how you express your preferences. PWR is a method of collective choice, I.e. choosing a winner among two or more alternatives. The two are not mutually exclusive. RCV can be used to implement PWR.

RCV has been called "instant runoff" because it involves going to the polls only once. PWR is ranking your first two candidates while RCV assigns ranks to all candidates. In a three candidate race, they are identical.

RCV can allow for more robust collective choice methods such as Haare, Coombs, Condorcet, Approval, Disapproval, Method of Marks. These have different efficiency properties depending on how many candidates there are.

The poll is therefore ill defined, and thus the respondents are likely giving ambiguous answers.

1. A simple system for cowards, liars, and the hypersensitive to destroy the careers and reputations of people they don't like.

There should be policies and reporting procedures, but they should NEVER be anonymous and the accused should always be entitled to the presumption of innocence, notice, and opportunity to respond.

Harassment is severe or pervasive by DEFINITION. Under such circumstances, it is imperative to have a named witness. The only time a named witness is unnecessary is when the harassment was so public that several people report the exact same infraction; a witness' identity then becomes irrelevant. But when there are reports of multiple actions in private, each incident relies crucially on a single person's testimony. The accused can seldom get notice and opportunity to respond without knowing those person's identities.

This open letter is further evidence of the continuing assault on due process.

Sorry, that should be 2, not 1.

2 so why exactly after 20 years constructing
a huge expensive title 9 poo-racracy
do the economic students demand to report
behavior to a NON title 9 person?

Bueno puento
our model she sez if people sortadonnawanna report to the title 9
people that means there mustabeproblemma with title 9 people?

so what exactly is the problem with the title 9 people?

Because Trump's Ed Dept just enacted rules that guarantee due process rights for the accused. They are no longer operating under Obama era rules where the person is guilty the moment they are accused.

Now they want to do an end run around Title IX.

this large voting bloc always wondered if letting sociologists investigate
sex crimes was a good idea because they don't actually have that sorta training.

3. Actually, most of them do agree it's an improvement. Now please explain why economists qua economists should have an opinion on this matter. Barry Eichengreen was the one among them who begged off. (There doesn't appear to have been a 'agree / disagree as a citizen'; 'no opinion as a professional' option).

5. (My first attempt to post this got dropped so I'm re-posting). Tyler's link reminds me of this article about why ocotopi and squid are so intelligent. The article talks about RNA editing and the possible evolutionary tradeoffs but doesn't mention epigenetics.
https://www.nytimes.com/2017/04/06/science/octopus-squid-intelligence-rna-editing.html

In the past couple of decades biology in general and genetics and neuroscience in particular have been perhaps the most intriguing fields of science. We already knew that we don't know how brains really work, but these discoveries continue to surprise.

Comments for this post are closed