29 February 2012

Manufacturing Stats

I caught up today with last week's EconTalk with Adam Davidson. The topic was American manufacturing, and it was based on some reporting he did at an autoparts factory in South Carolina.

One of the main points of discussion was the skills needed by the machinists who operate multi-axis CNC machine tools, and the extreme precision with which they turn out fuel injectors.

This reminded me of something I wanted to post about years ago, but I don't think I ever got around to it. We've got good statistics for the number of people employed in manufacturing both through the years and across countries. And decent statistics of the dollar value of manufacturing outputs. But do we have good numbers for the tolerance of manufacturing processes across time or borders? Or the number of components being assembled? Ceteris paribus, I'd be impressed by more pieces being assembled at higher precisions.

Obviously computing the averages for these things if difficult, if it's not undefined entirely. But I'd like to see someone take a try at estimating them.

23 February 2012

Computer Science is "How"

Carlos Bueno | Corrupting the Youth

On the one hand I think this a really good piece, the thesis of which I agree with entirely. Computer Science is about how to do things, and understanding it is a very general skill which helps explain the world.

On the other hand, I think this is a bad example:
Alan Kay has a classic talk about different modes of thinking and an experiment he did with young people of different ages. Imagine a robot that can do four things: move forward some distance, move backward, rotate some number of degrees, and make a mark on the floor. The goal is to write a program that tells the robot to draw a circle.

Before I tell you what happened, do take a minute and try it yourself. It's a pretty problem.

Precocious teenagers are stuffed full of facts about circles and curves, πr2 and calculus and all of that. More to the point, they have been trained in school to think abstractly and to leave aside visual or visceral thinking. Given this robot exercise they would spend 45 minutes on the blackboard writing equations and looking up things in books and scratching their heads before giving up.

Then Kay found some 10-year-olds and gave them the same problem. A 10-year-old doesn't know much about equations. She's a visual thinker. She doesn't know much about circles either, but she does know that a circle is defined as all points that are the same distance from the center. After about ten minutes of thinking and doodling she writes the program this way:
  1. Start at center.
  2. Move forward N inches.
  3. Make a mark.
  4. Move backward N inches.
  5. Rotate 1 degree.
  6. GOTO 1
That is a significant result. Ten-year-olds beat the pants off teenagers by using an officially discouraged mode of thinking. Not satisfied with that, crazy old Kay invited in a bunch of five-year-olds. You don't let little kids play with chalk if you want to get any work done, so for them he turned it into a game:

"Ok kids, you are robots. Say Beep beep beep! Good, now cover your eyes, and I want you to walk in a circle."

The kids would cover their eyes and cock their heads for a few seconds, thinking about how to walk in a circle while blind. Then they did it. They walked forward just a little, then turned a little, then walked forward again, and so on until their inner ear told them they'd done a full circuit. Not only did these children solve the problem immediately using visceral thinking, their solution was hundreds of times more efficient. Somehow they were intuiting differential equations.
Actually the 10-year old is very clever. I think that's a great example. The example with the 5-year olds is bad though.

I think one of the very valuable things about learning computer science and learning to program is figuring out how to construct a solution to a problem in an explicit enough way that you can explain it to someone else. In particular, you need to explain it to someone who is pathologically literal-minded — ie the computer.

What the 5-year olds are doing is entirely unlike that. An answer to "how to do get a robot to move in a circle?" can not be "close your eyes and walk around in a circle." Their "visceral," "intuitive" solution is great, but it's the opposite of what we're looking for. It's not the steps required to solve the problem, it's simply a solution to this particular instance of the problem, performed once, that they understand but which can not be expressed to anyone else, particularly someone inhuman.


PS The hack solution for this problem that I like goes as follows:
  1. Tie a string of length N to the robot.
  2. Tie the other end to a post at center of the circle.
  3. Have the robot go forward 2πr inches.
  4. Stop


PPS I could have mentioned this kind of thing in the previous post regarding what political candidates could do to impress me instead of wearing hard-hats while touring industrial plants and pretending they enjoy bowling.

If they can solve this sort of algorithmic problem then I'll respect them a little more. Nothing terribly complex. I'm thinking about questions easier than the "Dutch Flag Problem" which Buenos mentions, and not as explicitly computational. Good examples would be the non-automotive, mathematical questions that are asked on Car Talk's "Puzzler" segment like "The Coney Island Crab Cake Company" or "Stone Temple Farmers." If you can reason out answers to those, or at least take a running start at them, you're more likely to have my vote. Plus getting candidates to answer these questions — live! — would dissuade people from confusing oratorical skill with intelligence.


PPPS From Buenos' article:
If you've ever seen an old war movie you've heard the characters reading numbers over the radio: One niner five alpha bravo ---repeat--- one niner five alpha bravo!. Numbers have to be clear so they say "niner" because "nine" might be mistaken for "five". They use words for letters so no one confuses B with D, P or E. They also repeat important information as an extra safety. It takes longer to communicate this way but getting it right is more important than efficiency. Computers communicate this way too. In fact the theory behind the tradeoff between efficiency and reliability, called Shannon's information theory, is one of the ideas that made modern communications possible. One of the neat things about coding theory is that you can control exactly how efficient or reliable you want something to be.
(1) I'd never thought about the NATO phonetic alphabet as an Information Theory application before, but now it's so obvious.

(2) Isn't nine pronounced "niner" to distinguish it from the German "nein" not "five"?

Job Requirements

Atomic Nerds | LabRat | Show Me Your Moves

So I think it’s obvious I’m not going to get anything even remotely resembling what I want in a presidential candidate this year, but I thought it would be interesting to think about what sorts of skills or traits or displays I’d genuinely want to see in a candidate as opposed to what they all seem to think will appeal to voters. [...]

So what, besides the obvious agreement with my various political positions, would actually appeal to me and turn my head even if I disagreed with some or most of theirs? (Which, let’s face it, is going to happen anyway, given my positions.) What would actually demonstrate some form of respectable competency and, dare I say it, character?

1. The ability to teach a horse or dog at least one thing, without losing their temper. “Sit” or the equine equivalent would be too simple. How to walk nicely on a leash or long line would do well- convincing another critter that you’re worth following seems germane. Key point here is actually accomplishing the goal set out, with the critter’s full cooperation, without losing the temper and blaming the “student” or making excuses.

2. The ability to accurately and fairly describe the beliefs and structure of at least five faiths the candidate does not share.

3. The ability to accurately and fairly describe the logic and structure of at least five political positions the candidate is in direct opposition to.
These last two are essentially passing Caplan's "Ideological Turing Test," which I think is quite a good idea. LabRat goes on to list several more good qualifications, and then calls for further suggestions. Here are some of mine:

(1) Can you put together a moderately complicated piece of Ikea furniture (like a desk or dresser, not a coffee table)? This tells me you can follow directions, you have some spatial reasoning skills, and you can work within the very real limitations imposed by geometry and physics. There's no talking you way out of those. Like being able to train a dog without blaming it, can you put the furniture together without cursing at the stick people in the pamphlet or throwing a hex wrench across the room in disgust? This tells me about your temperament.

(2) On a similar vein, could you offer a credible entry for the types of hands-on projects that high school physics students do, like an egg drop or popsicle-stick bridge? I'm not looking for award winners here, just a basic competency understanding the physical world and some intuition for artifice.

(3) Can you walk into an art gallery or museum and carry on a 15 minute conversation about some of the pieces displayed there? The topic of this conversation should in no way relate to politics; no credit if you look at Guernica and then go off about invading Iraq and Afganistan. I'm looking for some evidence of aesthetic judgement and indication you have thought before about the more refined things in civilization and have developed your own personal taste to a level that you can articulate it.

(4) Can you show competency in a simple strategic game? Something you can learn in a few minutes, like Warlight or Dice Wars. I'm not looking for grandmaster skills, just the ability to put up average results against the game's standard AI. This demonstrates you can understand the arbitrary structure of the game and tailor your response to it (the way you citizens will be tailoring their responses to the rules you create), that you can think several steps ahead, and that you have sufficient Theory of Mind that you can anticipate your enemies' thought processes by mentally placing yourself in their position.

(5) Can you cook a basic meal for four? Nothing fancy. I just want a hot meal, with protein and side dishes, all coming to the table at roughly the same time. This is mostly a test of temperment as well. Everyone must eat. If you've gone through your life for forty or fifty years without learning this very, very basic survival skill and instead relied on parents, spouses and professionals to tend to your needs three times a day every day there is something I don't trust about you.

(6) Can you do any of these things or the ones on LabRat's list, or even simply behave like a decent, civilized, courteous human being, without sleeping the previous night? I think lack of sleep brings out the worst in everybody. Can you keep your base instincts in check even when tired? Do you have enough self-control to refrain from becoming grumpy or crabby or petulant even when your body desperately wants you to?

(7) Given 48 hours notice, can you prepare a 30-minute presentation, at an introductory level, about a topic of my choosing? (Something like "the history of the automobile" or "how the cell phone network works" or "recent trends in menswear.") Of course, you'd have to do this without the aid of a research staff. If you're going to be asked to make very important decisions about things you're not an expert in you better be able to learn about them quickly and synthesize some understanding.

16 February 2012

Why not get a doctorate from Sandia, MITRE or Google instead of Michigan State, CMU or UMass?

The Atlantic | Megan McArdle | Envisioning a Post-Campus America

I can see all sorts of factors that might combine to preserve the status quo, from signaling and status and networking, to the desire of college students for a four-year debt-financed semi-vacation. On the other hand, disruption never looks inevitable until it suddenly is--if you'd told someone in 1955 that GM was going to have its lunch eaten by some Japanese upstart, they would have laughed until the tears came. So it's interesting and maybe even useful to contemplate what the college system would look like if this sort of distance learning [like MITx] becomes the norm.

1. Education will end up being dominated by a few huge incumbents.
2. Online education will kill the liberal arts degree.
3. Professors (course developers) will be selected for teaching instead of research brilliance. 4. 95% of tenure-track professors will lose their jobs.
5. The corollary of #4 is the end of universities as research centers.
6. Young job-seekers will need new ways to signal diligence.
7. The economics of graduate school will change substantially.
8. Civil society will have to substitute for the intense friend networks that are built at college.
9. The role of schooling in upward mobility will change.
10. The young will have a much lower financial burden in their 20s.
11. The tutoring industry will boom.
12. If the credentials become valuable, cheating will be a problem.
This is an interesting and provocative list. McArdle expands on and defends each of these, but I want to discuss just #7:
7. The economics of graduate school will change substantially. I'm not sure what would happen to the master's and professional degrees--would there be a market for intense, focused instruction in small class groups? Medical school yes, law school probably, social work . . . um, as long as the government requires it, I guess.

But the PhD would be radically upended. Right now, graduate students get miserly stipends in exchange for considerably easing the teaching and research loads of their professors. But in an online model, we won't need so many teachers. And the online schools will not necessarily be research centers any more.

The implication is that most students, especially outside of STEM, will have to pay for their PhDs. Which should, at the very least, take care of the oversupply problem.
I don't think wiping out traditional universities would actually matter much in my field (Comp.Sci.) because the university is so much more tangential to the process of getting a PhD than most people think.

A bit of background. A CS PhD at Maryland, which is a fairly typical large research university program, goes like this. You spend two years taking classes. Even during this time you're encouraged not to make coursework your top priority. (UMD got rid of qualifying exams years ago, but many programs still require them at this point.) Then you spend a couple of years doing preliminary work and proposing your dissertation topic. Then you spend three or so more years writing your dissertation. Then you defend it to a small panel of faculty members.

Everything centers around satisfying those five people on your dissertation committee. Everything else you do is explicitly a preliminary building up to that or a sideshow.

It's important that there's no central authority for determining what does and doesn't constitute PhD-level work. There's no CPA or Bar exam or any other standardized assay of your capability. The only thing separating a PhD from a non-PhD is the approval of less than half a dozen people who have already gotten their PhDs.

The sin qua non of having a PhD program is simply other people with PhDs. You don't need the campus and the grades and the offices and the undergrads and the rest. We could transition pretty easily into a world where independent organizations like The Santa Fe Institute or government labs like Lawrence Livermore or even industry groups like Microsoft Research granted doctorates.

You take an extra couple of semesters of courses, maybe wherever you did undergrad or maybe you go get an MS, or you pass a Qual Exam. Then you spend some years as a journeyman for someone who already has a doctorate. Then whenever a small cluster of PhDs think you're worthy, you're "graduated." It's not that different than the way we do things now, except it would decouple your coursework from your dissertation work, which would probably be a lot more efficient.


PS See also "The economics of higher non-profit and for-profit education" at Marginal Revolution.

14 February 2012

Size matters (?)

The Money Illusion | Scott Sumner | Santantango

This post is completely off topic, but something I’ve been thinking a lot about recently. What determines the length of novels and films? Novels are usually around 200 to 500 pages, and films are usually around 2 hours. I understand that the length of films is somewhat constrained by the desire of theaters to run two showings after dinner, but I don’t think that’s a complete explanation. The same is true for “art films.” Even art films based on novels, despite the fact that an ordinary novel would take 5 to 10 hours to show on film, especially at the pace of art films.
We do have films that are dozens or even hundreds of hours long. They're called "TV Shows."
I don’t know if it’s just me, but I’ve generally enjoyed “mega-novels” more than novels of ordinary length. I just finished “1Q84,” which is now my favorite Murakami book. In recent years I also read and greatly enjoyed mega-novels like 2666, The Man Without Qualities, and Lord of the Rings. I still haven’t read many of the longer classics (War and Peace, The Brothers Karamazov, In Search of Lost Time, etc) but am told these are also outstanding novels. Why aren’t there more long novels? The only disappointing long novel that I ever finished was Phillip Pullman’s Golden Compass trilogy, and even that was pretty good until the third volume.
Thousand page books seem great for the same reason hundred year old furniture all seems sturdy: the stuff that wasn't built to last hasn't. We never see the cheap, breakable chairs from 1912, and we tend to forget about the boring, disposable 1000 page novels. (If, indeed, they ever get printed in the first place.)
Maybe I’m attracted to books that allow me to escape into other realities. It took me so long to finish 1Q84 that by the end I felt like I was partly inside Murakami’s imaginary world.
If Murakami can build that world in 944 pages, good for him. That's an impressive feat. But if an author can make me feel the same emotions in 500 pages, or 50, or 5 that's really impressive.

A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away.
  — Antoine de Saint-Exupery

I used to prefer long novels and series. Now I like the opposite. Not because I don't have patience for longer books, but because the opportunity cost is high. If I coudl read one 1000 page book or three 330 page books, ceteris paribus I'll diversify my entertainment and go with multiple, shorter books.
Colin Marshall expresses a similar view, even without having seen the shorter version:
I submit to you that, while some stories are indeed best told in 90- to 120-ish-minutes, most others, by pure logic of probability — are not.
I submit to you that, while some paintings are best expressed on rectangles with between one and three meters diagonal, most others, by pure logic of probability — are not.

Where then are the billboard-sized masterpieces. Or the football field-sized masterpieces?


For some loosely related commentary, check out Julian Sanchez's discussion of the differences between self-contained and "extended universe" fiction, specifically as it related to the recently-announced Watchmen prequels.

13 February 2012

"Balance out of Life"

Vimeo | Wyatt Hodgson | Balance out of Life

Koyaanisqatsi (1982) at 1552% speed


Via Planetary Folklore.

I don't really buy into the environmental/philosophical message of Koyaanisqatsi, but it sure is pretty to look at.

This sped-up version also drives home a point the original made me think about. We've got this class of things called "movies" which are shown at specific times in special places where everyone has purposefully gathered to see them. And we have this other class of things called "paintings and photographs" which we hang on the walls or pretty much every type of space for people to observe whenever they happen by.

Why do these need to be separate things? Now that we have the technical capability to easily display a moving image on a table top or wall, why aren't we doing that more often? Where are all the portraits and landscapes that happen to be in motion? Where are the movies whose sole point is "this is an interesting/pretty/pleasing thing to look at"?

09 February 2012

Even if The Dude bowls with Hank Rearden everyday he'll never be a Morlock

NY Times: Opinion | David Brooks | The Great Divorce

I doubt Murray would agree, but we need a National Service Program. We need a program that would force members of the upper tribe and the lower tribe to live together, if only for a few years. We need a program in which people from both tribes work together to spread out the values, practices and institutions that lead to achievement.

If we could jam the tribes together, we’d have a better elite and a better mass.
(1) I was involved in a scheme like this as a child. I, along with the other students at the magnet program I attended, were bused into the shittiest part of the county.*The same county in which Brooks lives, I might add. The ostensive objective was to do exactly what Brooks describes: let our good habits rub off on the disadvantaged. (The cynical objective was to juke the stats by raising the average test scores of a poor, minority school.)

It didn't work at all. In my estimation, it backfired completely. No one from genpop wanted to be mistaken for one of us nerds, so they were extra careful not to put in effort, turn in assignments, show up, participate, etc. The few bright students in the general population were mocked for wanting to actually learn like all the magnet maggot kids.

(2) All the same objections I raised back in 2008 about compulsory volunteering still hold (e.g. one, two, three, four). Forced labor, even for the "common good" is still slavery. It's ripe for corruption, giving favored groups and causes access to free labor and well-connected "volunteers" opportunities to shirk. The government already controls enough of my economic output so I don't need them controlling my time as well. It's ageist: no one ever suggests we require service from retirees as a condition of SS/Medicare. Do I need to go on?

(3) Institutions matter. If you take average people and put them in situations with dysfunctional institutions and bad incentives you'll get dysfunctional workers. The DMV will still be the DMV when the zero-marginal-product epsilon moron who works there gets done rubbing elbows with some Vickies. The union contract will still be there; the rock-solid, untethered-to-results budget will still be there; the state-enforced, police-backed monopoly will still be there. None of the incentives exist for that DMV worker to be productive, but because she spent the summer raking trails in the park with a go-getter she's suddenly going to turn into an "achiever"?

(4) Let me get this straight. Brooks' plan is to take the most productive members of society, the ones with the habits and outlooks and decisions who make them the most successful, and remove them from the workforce for a couple of years? Does that not strike you as the goddamned opposite of what they ought to be doing? Instead of designing chemical plants and writing software he wants to force these people to babysit ostensible adults who haven't gotten their own shit together and need a culture transplant?

(via Brian Dunbar)

08 February 2012

Peter Randall-Page


TateShots | Peter Randall-Page, sculptor

Bondholder Discipline

I largely agree with Bill Black in this EconTalk discussion about fraud in finance. I particularly appreciate that he doesn't make the issue of financial regulation into a Red Team/Blue Team fight.

However, this part of the exchange in particular doesn't sit well with me.
EconTalk | Russ Roberts | William Black on Financial Fraud

Black: But don't forget, in terms of key stuff about moral hazard: the government wasn't the only entity. There were shareholders at most of the worst places. And there was subordinated debt at many of the most fraudulent places. And of course in economic theory, it was supposed to be subordinated debt that was the perfect form of private-market discipline. You had the right incentives; you had the sophistication; and you should have done something. But there were zero cases of effective private market discipline by either shareholders or sub-debt holders in the Savings and Loan crisis.

Roberts: So, let me just review that, because we've talked about this in passing in many different podcasts, many episodes of this show. The debt holders have a fixed upside. They cannot make more than they are promised. Their downside is being wiped out. So in general, they are going to be the watchdogs of risk taking and the enforcers of prudence on the part of the people who they've lent money to through this debt. And what I have been worrying about for a long time, and it's a worry I've learned from Gary Stern's work, is: Well, that's true, that under a market system the debt holders are supposed to be the disciplinarians of risk taking, the watchdogs, but if the bond holders think that they might get their money back even when the firm goes out of business, which happened with Continental Illinois in 1984, you might start to not worry so much about that and be willing to accept a fixed rate of return even when there is a chance that the investment will amount to nothing, because you might get your money anyway. Correct?

Black: True, but not basic enough. In other words, you are quite right that Continental Illinois did a terrible thing and it bailed out sub-debt holders. We never did that in the Savings and Loan. We always wiped out the sub-debt holders. And of course they are supposed to be wiped out. That's the concept of risk capital. Despite that, there was never effective private market discipline; indeed there was no even effort at private market discipline. It failed, by sub-debt holders during that entire crisis.
I don't know that much about the S&L crisis; it was before my time. But my instinct is that Black is only looking at half the outcome matrix when he says there was "zero cases" of private market discipline from debt holders.

Anti-lock brakes are supposed to help prevent car crashes. If I looked at 100 cars which got in crashes, and saw that they all had anti-lock brakes, could I conclude that anti-lock brakes always fail to prevent crashes? Of course not! I'd have to consider how many other, non-crashed cars also had anti-lock brakes, as well as the crash rate of cars without anti-lock brakes.

Similarly looking at failed banks and noting that none of their subordinated creditors prevented excessive risk-taking does not allow us to conclude that creditors "never" provide discipline. To reach that conclusion we would also have to know how many other banks would have failed in the absence of discipline from their creditors.

There's another factor to consider, which is the discipline provided by not making the loan to the S&L. Suppose one lender provides an S&L with $10million, and nine others turn the S&L down for $10million each because they believe they are being too risky. If the bank later fails, is this evidence that the market provided no discipline? In Black's view, yes, because the lender who was supposed to prevent excess risk did not. In my view, no, because the losses were only one tenth what they would have otherwise been. It's the potential lender who is most optimistic about the risks who will make the loan, so we should look to the other potential lenders for evidence that market actors judged the risk to be too high.

There's a big difference between "provided no discipline" and "did not provide enough discipline." Since risks can only be minimized and never eliminated completely, it's especially important to recognize that distinction.

07 February 2012

Step One: If you're going the wrong way, turn around.

Charlie's Diary | Cat Valente | How Do We Get There?

I think every writer has a genre or subgenre that they admire, but find baffling. [...] Well, for me, one of those genres is post-scarcity SF. To my mind it's one of the most difficult to pull off. Scarcity has been a fact of the human condition for more or less ever, and once you remove it you have to figure out what it means to be human aside from that endless parade of want. Before you start chapter one.
So far we agree. Scarcity will always be with us. I think most "post-scarcity" writers pull this off by just shifting what is scarce. So Iain Banks' characters still compete for prestigious jobs, attention, travel, reputation, control and power, entertainment, prizes and rankings, etc. just not for food, shelter, or clothing.
On top of that, it's damnably hard to fashion a sympathetic protagonist out of someone who has never struggled in the way we struggle in our own lives, to present someone who does not come off as a monster of privilege.
I think Valente has a pretty facile view of psychology. Or at least one that is pathologically focused on material needs. Just because you grow up never wanting for any physical goods doesn't mean you don't have struggles. Daddy never approved of you; you always lived in the shadow of your older, favored siblings; you were expected to follow a path you didn't want. Bottom elephant: everyone – rich, poor, privileged, down-trodden, or in between – has problems.

Side note: I would not want to have dinner with someone who thinks that people who have their material needs met are monstrous.
Yet I've been thinking about it constantly, as even this morning the lead news story on the radio are about tens upon tens of thousands of jobs being vanished as a cost-cutting measure for American Airlines, who surely have not lost ten billion dollars in the last ten years due to cargo carrier and flight attendant salaries.
"Surely"? Really? She's sure about that? Because depending on fuel prices, labor is either the number one or two largest cost to an airline. There's nothing they can do about fuel costs, short of marrying the CEO's daughter into the House of Saud. Is it really that crazy that American would try and reduce the single biggest cost they can?

(And don't forget Amdahl's Law. If labor is 30% of your expenses and you cut it by 10%, your total costs only go down 3%.)

One more thing – if you're concerned about scarcity, you shouldn't want people to carry on doing things that are unproductive. Those "tens of thousands of jobs" aren't "vanishing" because American Airlines is run by mean people, they're going away because they aren't productive. They're destroying utility. You're never going to reduce scarcity by trying to increase the consumption of labor inputs.
But here's the thing--in most (not all, of course) post-scarcity SF, the fact of post-scarcity is a given. The Culture exists. The question of how we got there might be alluded to or skimmed over in an infodump, but I have so often been left feeling like there's us here, and then SCENE MISSING, SCENE MISSING, transeconomic future humans. Like the opening credits of Enterprise--I see all the steps in the space travel evolution chart, but there's a big gap between the space shuttle and Zefram Cochrane. I am a snake charmer--I can't see how we can get so high, in such spangles, how we can fly with such daring.
Okay, fine. Post-scarcity requires even more hand-waving than all the other things we easily take for granted in SF, like faster-than-light travel and radical life extension. People's desires outstripping their ability to fulfill them is a more ironclad law than Special Relativity.
I think it's a slightly less murky path in Europe than it is in the US right now. Our powers that be would rather drink cognac on a pile of our bones than even give us health care. The word "socialism" might as well be "Voldemort": it which must not be named.
Well there's your problem right there!  No wonder you can't imagine a way to reduce scarcity: you're looking in the wrong direction! Turn around!!

I might as well say I'm pessimistic about there ever being a cure for cancer because the homeopaths and herbalists and lysenkoists are still so far away from solving the problem. That would make about as much sense as Valente does.

Why would the solution to scarcity possibly come from the people who deny scarcity exists? How would the people who focus their efforts on shuffling resources from one group to another ever come up with a way to produce magically vast amounts of resources? If you think scarce resources are things you deserve to be "given" by the "powers" then you don't even begin to have the right mindset to imagine alleviating scarcity.

Things must be produced; they do not miraculously appear, to be "given" to the deserving.

Mobility is not and can not be a ratchet

The Atlantic | Megan McArdle | Income Mobility Means Some People Have to Lose Everything

I've said before that I don't care about income inequality per se, and that focusing on it seems more like institutionalized envy than sound policy. I care about the absolute condition of the poor--do they have the basics of a decent life?
One thousand times Amen.
On the other hand, income mobility is a very important issue. Regardless of how far the top is from the bottom, children born in America should have an equal chance to move from the latter to the former. This is especially important given that so many of the highest-paid jobs are also the most pleasant.

Many people apparently agree with me: the issue of income mobility has become more prominent in policy debates over the last few years. And yet I submit that this agreement is entirely theoretical. How many of the people reading this blog would actually tolerate a one-in-five chance that their children would end up poor?

Because that's what income mobility actually means. It doesn't just mean giving a lift to the folks at the bottom--superior health care, better K-12 education. Everyone in the country cannot be above average. For the poor to have a better shot at ending up in the top quintiles, the folks in the top few quintiles have to run the risk of ending up in the lowest.

Who among the parents fighting so hard to get their kids into a good school is going to volunteer to have their kid give up the slot in the upper middle class? People are willing to accept a certain amount of slippage, but only as long as it comes with added job security (government) or special fulfillment (the ministry, the arts)--and even in the latter cases, Mom and Dad will often be strenuously arguing against following your calling. But how many doctors and lawyers would simply glumly accept it if you told them that sorry, junior's going to be an intermittently employed long-haul trucker, and your darling daughter is going to work the supermarket checkout, because all the more lucrative and interesting slots went to smarter and more talented people?
I don't understand McArdle's logic here. Just because I don't want my kid to be a long-haul truck driver (like several people in my family tree, I'll add) that's supposed to mean that I don't want anyone to be a trucker? Or I don't want anyone to be a trucker if their father had a better job? Her position doesn't even begin to compute with me.

Downward mobility isn't something I desire for my kids, but it's something I think ought to happen to some of my cousins, classmates and other friends. (At least until they straighten out their lives and develop the character that supports high income/wealth.) Just because I want my children to do well for themselves doesn't mean I don't have a couple of lazy, arrogant, eloi cousins who I think deserve to be knocked down a couple of pegs from where previous generation stood.
You don't have anyone in your family you think deserves to do worse than their parents? Really? Everyone in your clan behaves in a way that makes them deserve to rise higher than they are? Your reunions must be a lot more pleasant than mine then.
One of the reasons this is so hard is that so many of the problems poor people deal with are created by living near other poor people. Most poor people are not criminals, but most criminals are poor people, because crime actually doesn't pay (very well). Most poor people take out their trash, maintain their homes, and stay off drugs--but the kind of people who don't do those things are disproportionately likely to end up in poverty. Which is to say, in your neighborhood, if you are poor--shooting at each other and hitting bystanders, breeding vermin that migrate into your living space, pilfering your stuff to support their drug habit.

Someone has to live near those people; whatever your expectations for antipoverty policy, it surely does not include the end of drug addiction and slovenly habits. But should it be your kid? Would you want them to have a one-in-five chance of living in those conditions?
Who said the worst part of being poor is being surrounded by poor people? He was right. One more good reason to decouple geography from government services, especially operating schools.
[Middle class parents] pay lip service to mobility, but they work damn hard to make sure that their kids don't get exposed to a peer group that might normalize dropping out and working low-wage, dead end jobs, or going on welfare.
I know many football coaches who work damn hard to make sure their teams aren't the ones who lose. Does this mean they don't believe in football, or in competition? That they don't actually want their to be winners and losers? That they don't want the refs to keep score? That the they don't think the team with the lowest number of points should record an 'L' on their schedule? That their talk about competition is just "lip service"?

Just because you work to avoid an outcome for yourself or your loved ones doesn't mean you think that outcome shouldn't be a possibility.

I want my future children to be tested. Yes, I also want them to pass that test and build successful lives for themselves. And I will work hard to increase the chance of them doing that. That doesn't mean I don't want them to be tested.*In fact I think growing up knowing you have no chance of slipping down is pretty toxic. How many Roman emperors born in the purple turned out not to be headcases? Two? It's not that I want my children to move up the income ladder. It's that I want them to have the habits, make the decisions, and cultivate the behaviors that enables them to do so. And that's impossible unless there is the chance that they will slide down as well as climb up.

Maybe most other people don't feel this way. Maybe they want mobility to be one way, at least for their loved ones. I don't know. But I wouldn't conclude that just because people work to help their children succeed they don't also think it should be possible for their children to fail.

06 February 2012

Firing Aversion: Naming the Puppy

No, this is not literally about puppies. Though Mrs SB7 and I are in the process of looking for one.
EconLog | Bryan Caplan | Naming the Puppy: Firing Aversion and the Labor Market

In fiction (and "reality" television), firing workers almost seems fun. How many times has Mr. Burns gleefully hissed, "Fire than man, Smithers!"? In the real world, though, bosses dislike being the bearer of bad news. They feel guilty when they let someone go. So guilty, in fact, that some hire consultants to help them fire people. To coin a behavioral econ phrase, most employers feel "firing aversion."

How does firing aversion play out in the real world? For starters:

1. Firms often fail to hire workers who would be profitable in the short-run. [...]

2. Signaling matters more. [...]

3. Outsourcing looks better. [...]
(1) I had a professor at ND who said he left industry for academia specifically because it meant he would never have to fire anyone again.

(2a) Wouldn't giving more workers fixed-length contracts ameliorate these problems? You still have to deliver the bad news that the contract won't be renewed, but it seems a lot better because the frame is that you're declining to give the person something, rather than taking away something they see as theirs.

(2b) Speaking of which, "your job" is only "yours" in the sense that "your girlfriend" is "yours." Employment is a relationship between two parties, not the possession of one of them. If more people accepted that we would have a lot less whining about foreigners "taking our jobs."

(2c) I assume there are legal problems standing in the way of (2a) which I do not know about.

(3a) The mail clerk in our department recently "failed to pass her employment probation." I have a cousin who similarly failed a probationary period at a hardware store. Why is this not more common?

(3b) Someone — Ryan Avent? — once said that labor markets in America were odd because the decision to hire someone preceded all discussion and negotiation of salary.  Relative to the current status quo with salary offers following employment offers, and an alternative where salary is agreed before a final offer is made, would determining salary after a probationary period make more or less sense?

(3c) I suppose that question depends on wether you're the hirer or the hiree, and wether the hiree is leaving another job or is unemployed, but I'll leave it as is.

(3d) I hypothesize that the situation Avent (?) describes persists because the results of the salary negotiation will make only a very small difference to the employer's cost of taking on the new hire, and so they have little reason to negotiate compensation with multiple candidates and then decide which one to hire.

Also, the employer's ability to predict the new hire's productivity is very imprecise, so it is very difficult to select the candidate who minimizes compensation costs while maximizing worker production. As a result, you might as be willing to make an employment offer despite the uncertainty in their compensation because that is dwarfed by your uncertainty about their production. Why invest a lot of effort over $2,000 in salary and an extra sick day when the total cost of your new hire is going to be $200,000 annually and you estimate they will produce $250,000 in revenue, give or take $100,000?

Huzzah

I feel like I spend a lot of time around here calling out things I don't like, so let me take a moment and list some things I've been enjoying recently. This seems like a healthy way to put the disappointment of the Patriots' loss behind me and start the week.
  • Downton Abbey.
    You know that feeling you get when you finish a really good book and you're disappointed that there's no more left to enjoy? Mrs SB7 and I tore through both seasons of this in about a week, and now I have that feeling. And I don't typically care for costume dramas very much.
  • "One Cello X 16," by Zoe Keating.
    Great music to work to: lots of energy, but without distracting lyrics.
  • Artisan Bread in Five Minutes a Day
  • As of this morning, having paid down all of our credit card balance. This one feels especially good.
  • Dueling Michael Caines by Rob Brydon and Steve Coogan from The Trip
  • High West Distillery's Double Rye, a blend of peppery 2-year-old and silky 16-year-old. Oh, and their Bourye, a blend of sweet bourbon and fiesty rye. These are some of the most complex spirits I've tasted not made in Scotland, and the theory behind them is so simple. Is there anyone else out there doing something like this? Their barrel-aged Manhattan is aso delicious. It sounds crazy, but try it before judging.
Now: time to do some Science!


PS Apropos of nothing, you know what I'd like to see for a Super Bowl half-time show? I'd like some straight-ahead rock band like the Black Keys to wheel a drumkit out to midfield, bang out two songs, and walk off. No dancers, no light show, no special guest appearances, no theatrics. Go 100% minimalist. Be over-the-top with the refusal to attempt to be over-the-top.

An open request to the UMD Dept. of Transportation

Dear DOTS,

I would greatly appreciate it if you would either (1) fix the wifi service you provide on your busses so that I can get online during my commute, or (2) remove the "This Bus Has WiFi!" signs from the inside of your buses.

I have a slight preference for (1), but really either would satisfy me. To leave me with busted wifi and force me to look at signs bragging about wifi is a bit mean.

Respectfully,
A grad student

03 February 2012

Weeding STEMs

Blunt Object | Is STEM captured by faculty?

There’s this cute thing that non-STEM majors do every once in a while: They imagine that the process of earning, say, a Computing Science degree is essentially similar to the process of earning whatever degree they actually attained. Presumably, they imagine, it begins with a number of survey courses that give a broad overview of the history of the discipline and the various schools of thought and specialization within it, and continues with more specialized courses in those various specializations — and at the same time broadens into possibly-related courses from other departments — and entails the acquisition of knowledge in a more or less organic fashion. [...]

Intro STEM classes aren’t designed to scare people away so much as they’re designed to impart a wide range of utterly fundamental tools in a short period of time — and, yes, to weed out students who lack facility with the basics. This weeding-out is less a matter of elitism than it is of efficiency: If you don’t have a sufficient grasp of calculus, programming, statics, or organic chem to pass the first-year requirements you’ll only find yourself further behind in subsequent courses.
Indeed. I won't speak for all of STEM, but in my CS experience professors don't wash out underperforming first years because they want to get rid of them.*Yes, it's true that they often would rather not be teaching undergrad classes, but reducing the class size isn't helpful. It's the class prep that's the problem in CS, which is con-stant in relation to the number of students. The cost of a marginal student is very low, even in relation to other STEM fields, because most assign-ments are by nature graded automatically. The marginal cost of a CS student is especially low compared to that of a human-ities student, where essay grading seems to be a large component of the workload. They wash out underperforming students because the advanced classes are hard.

There is a fundamental difference between the way we teach STEM and humanities. Upper level STEM classes build explicitly on lower ones in a way that humanities classes do not. That's one of the reasons I could take 400-level electives in history and film and philosophy and keep pace — rather handily, if I do say so — with the people majoring in those disciplines. Professors expectations in upper-level humanities classes may be higher than for the intro classes, but the material is not particularly harder, nor does it require a mastery of the material in the lower-level classes. This is not at all true in 300 and 400 level STEM classes. That's one big reason you do see CS students taking (and excelling) in Arts & Letters classes but you don't often see history majors acing Operating Systems or Machine Learning.


I think there's actually a big value to weeding out undergrad CS students in particular, and that's because too many of them think they're going to get rich without having to put in effort.

Let me explain by way of a digression. I went to high school with a guy who everyone thought of as a musical genius. He won all the talent shows, etc. because he had what amounted to a dead-on Steven Tyler impersonation. Then he went away to a prestigious conservatory to study music and washed out in his first year. Last I saw him he was directing cars in a concert venue's parking lot.

Within the narrow confines of my hometown, the kid was a star. But he, along with many others that sang his praises, didn't realize that you need to put in a lot of effort mastering artistic fundamentals in addition to a having decent singing voice and a large helping of stage presence. There's a lot more to being a musician than being able to hit the high notes on "Walk this Way."

Guys like that show up as a freshman CS majors every year. They're the guy their family or neighbors go to for help with computer stuff, so they're the biggest fish in their pond. And they figure they're going to become a millionaire inventing some app by the time they're 25, because they're the only kid in school who knows HTML. They don't realize there's a world of difference between building a custom template for their Tumblr account and building Tumblr itself. If you let that kid stumble through intro-CS classes without putting them to the test they're eventually going to crash and burn in Algorithms or Data Structures or Compilers or OS. Coddling Mr. I'm-Going-To-Invent-The-Next-Facebook in their freshman and sophomore years doesn't do him any favors.

Mrs SB7's eight graders had visits earlier this week from department representatives at the local high school about the courses they can sign up for next year.  Following the presentation by the technology department all five of her classes had quemments about taking computer programming and getting rich, and none of those people seemed to understand that computer programming is complicated. Granted, these are 14 year olds, but I seee a lot of college freshmen who still have that attitude.

I don't know if other STEM fields have the same situation. I certainly didn't meet many kids who thought they were going to get rich quick by majoring in CHEG.

02 February 2012

Ban everything which is not a Pareto improvement?

Earlier I commented on an exchange between Jacob Grier and Bill Gardner. Since then something else that Gardner said has been rubbing me the wrong way.
Something not unlike research | Bill Gardner | In which I am chastised for arrogance

But what about liberty? I am not arguing that government should ban bone lugers from employment. My point is that I do not see a compelling argument against employers choosing not to hire them. By extension, we should have no public policies protecting access to employment for smokers (or bone lugers). This is, I believe, consistent with Mill's view that
The only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not sufficient warrant.
Honestly, I would be satisfied if power was only exercised to prevent harm to others. It's certainly a step forward from the society we have now, where power is so routinely exercised to protect people from themselves.

My problem is that in basing the entire decision on whether harm is done Mill and Gardner entirely abdicate even attempting to balance costs and benefits. If the only criterion necessary to ban an action is that someone, somewhere, believes themselves to be harmed then we are all at the mercy of the person most willing to claim injury. (And frankly, I don't like that person very much.) We also lose the ability to take actions that are of great benefit to many, and mild nuisance to some, as well as the ability to decide to mitigate bad outcomes rather than banning them entirely.

There is also the problem of determining the frame of reference. On the one hand, we can ban smoking in cigar shops because it will cause some harm to people who (voluntarily) work there or (voluntarily) patronize the shop. To hell with the good some people derive from the shop, because that shop is causing harm to some people!

On the other hand, such a rule will harm the business owner and patrons of the shop. So let's use our power to stop people from enacting bans on cigar shops. To hell with the good that some people will derive from the ban on cigar shops, because those bans are causing harm to some people!

We see this frame of reference problem in trade policy all the time. I've heard from the White House that we need to slap tariffs on imported tires because they are harming American tire makers. But on the other side of the ledger the tariff is harming American tire consumers. (Not to mention foreign tire makers, who I'd like to think have at least some moral standing.)

Bootleggers and Baptists at the gun range

Gilbert Indoor Range: Fees

We will have a wide variety ammunition available. Under our special exception to operate an indoor shooting range, Montgomery County placed the requirement that all ammunition be purchased and consumed at the facility. [...] They did not want guns and ammo together in the car while driving around the County, even though it would have been perfectly legal going to and from a range.
So the firm given a monopoly on indoor shooting in the county also gets to sell all the ammunition. Both the Grundyists on the County Council and the crony capitalists at Gilbert must be so pleased with themselves. What a coincidence that their desires would align so well!

False Consciousness is the hollow trump card of every "reformer."

Liquidity Preference | Jacob Grier | Smoking: More than just a vice

Something Not Unlike Research” is a health care blog I recently came across written by two university professors. One of them, Bill Gardner, wrote this week in defense of employers choosing not to hire smokers. He concluding by noting that “Smoking is a vice that benefits no one.” I took him to task for this last line on Twitter:
“Smoking is a vice that benefits no one.” — @Bill_Gardner Oh please. I like it. It benefits me! Arrogant assumption.
To my surprise, he responded in a new post:
[...] Having conceded that smoking may benefit Jacob’s subjective well-being, can I still say that “smoking is a vice that benefits no one”? I certainly can, if it’s understood that benefit refers to the long-term well-being of smokers, and those who depend upon them, rather than immediate subjective well-being. In that sense of benefit, there is nothing to be said for smoking or binge drinking.
Semantics aside, activists’ unwillingness to consider the benefits of smoking leads to excessively restrictive policies. Let’s take smoking bans for example. Consider two businesses:

Business 1 is a tobacco shop with an attached lounge that offers beer and wine. Customers are allowed to smoke there. It’s a freestanding building with no immediate neighbors, so no one except customers and employees is affected by the smoking. Four people are employed serving drinks in the lounge. A smoking ban passes that forces the business to eliminate drink service. The day the ban takes effect those four employees lose their jobs.

Business 2 is a restaurant that serves Dungeness crab caught in the Pacific Northwest. Commercial fishing has one of the highest fatality rates of any occupation and crabbing in this region is often the highest of all. For comparison, the average annual fatality rate for all occupations is 4 per 100,000 workers. For fishing as a whole the rate is 115/100,000. For Dungeness crab fishermen in the Pacific Northwest the rate is 463/100,000. (Source here.) There are no proposals to forbid restaurants from serving Dungeness crab.

The comparison might seem silly, but why? Dungeness crab is delicious but it’s hardly a staple in the food supply. Fishermen are literally dying to put it on our plates. Though the level of risk associated with secondhand smoke exposure is in dispute, it would be astonishing if the danger of pouring beer in a smoky room was at all comparable to crabbing on a stormy ocean. So again, why the disparity in how we treat these workers? [...]

Yet exceptions to smoking bans are often unreasonably narrow, preventing consenting adults from making free exchanges with each other. This is because policy makers view smoking as inherently without value. The thought process goes something like this:

1) Smoking has no value.

2) Protecting workers has value.

3) Therefore it’s OK to ban smoking everywhere without worrying about smokers’ preferences.
Grier makes an excellent point. Of course Dungeness crab is only the beginning. There is literally no end to the things people can — and do! — claim have no value and are not beneficial in the long run. Sure, motorcycle owners claim their vehicles enhance their "subjective well being" but I know better, and I can tell that in the long run they are too risky to truly benefit them.

Similarly drinking, eating red meat, sky diving, gambling, having premarital sex, taking a pay-day loan, driving long distances, watching pornography, watching television, downhill skiing, majoring in art, working as a salaryman, getting a tattoo, getting high, ultramarathoning, having a debit card with late payment or overdraft fees, using previously owned children's toys or cribs, eating sugar, ...

When you claim you know what benefits people more than they know themselves there is nothing you can't justify.

Murray, Elites, Empathy

EconLog | Arnold Kling | Kling on Charles Murray

So, I downloaded Coming Apart. I am not disappointed. It is well argued. [...]

I can think of one interesting indicator for being in the "bubble." In the nearest public high school, take the ratio of the number of seniors who will attend an Ivy League school to the number of seniors who will enter the military. Where I live, it is about 0.2. In the better parts of Montgomery County, it has to be at least 5. I would say that if it is more than 1, you are in the bubble.
I grew up in "the better parts of Montgomery County." My graduating class was 450 people. I'd say about three dozen people went to Ivy League schools. IIRC three people joined ROTC programs. I don't think that's what Kling has in mind for typical people entering the military, so let's ignore them. Two enlisted, one in the USMC and one in the Israeli army. Again, I don't think the latter is what Kling has in mind for typical military service. That left my graduating class with a ratio over 30:1!
Beyond college, high-IQ people sort themselves into what Murray calls Superzips, which are zip codes that contain a concentration of people with high educational attainment and high income.
It is not a problem if truck drivers cannot empathize with the priorities of Yale professors. It is a problem if Yale professors [and others in the elite] cannot empathize with the priorities of truck drivers.
I disagree. The inability of the elite professors to empathize with the truck driver is only a real problem if the professor has control over how the truck driver lives his life.

We don't care about wether the truck driver empathizes with the professor precisely because he does not get to make decisions for the professor.

If we're truly worried about this disconnect in empathy one solution is to make people more empathetic. Another is to remove the chance for their lack of empathy to cause harm by reducing the power of elites to coerce others.