21 December 2012

TfA & Marginal Analysis

Coyote Blog | Warren Meyer | Teach for America

But through a strange series of articles that seemed to have used the Sandy Hook massacre as an argument for teacher unionization and against charter schools (yeah, I don't get the connection either), I found out that teachers unions hate Teach for America. Which means that I will likely double my contribution next year.
Mrs. SB7 is a teacher, and when she was getting her certification it was accepted doctrine among the education students and faculty that TfA was not only bad but outright insulting. The fact that my wife had no problem with them was treated as heresy.

I can begin to see their point — some college kid who paradrops into a school with a couple weeks preparation is no substitute for an experienced, trained professional.

But the mistake they make is one made by so many on the Left: they weren't analyzing at the margin. The trade-off isn't "inexperienced college kid" vs. "experienced professional." It's "inexperienced college kid" vs. "whatever other warm body was going to get hired to fill up space in the front of the class/no teacher at all."

Of course if you buy into the NEA world view, where every teacher is (far-)above average (literally), this isn't a relevant distinction.

I will not be celebrating Filter Day

Overcoming Bias | Robin Hanson | Today Is Filter Day

So today, let me remind us all of one key somber and neglected fact: the universe looks very dead. Yes, there might be pockets of life hiding in small corners, but for billions of years billions of galaxies full of vast resources have been left almost entirely untouched and unused. While we seem only centuries away making a great visible use of our solar system, and a million years from doing the same to our galaxy, any life out there seems unable, uninterested, or afraid to do the same. What dark fact do they know that we do not?

Yes, it is possible that the extremely difficultly was life’s origin, or some early step, so that, other than here on Earth, all life in the universe is stuck before this early extremely hard step.
Hanson leaves out one entire end of the probability distribution. Perhaps these civilizations are uninterested not because they've learned a dark fact, but because they've learned a glorious fact?

Hanson is skipping right from "they haven't" to "they can't (because of catastrophe)." What if they aren't "stuck," but instead they don't need to alter entire galaxies to suit themselves? Perhaps they have created VR so good they don't need to leave home. Perhaps they've sublimed. Perhaps Bostrom is right, and there are plenty of advanced civilizations, but they're exogenous to our simulation. To be truly fanciful, perhaps we live in the Slow Zone.

There's another, less profound (and more likely) possibility. As Hanson says, the universe does look dead. Strike that. The parts of the universe which we can see look dead. But the observable universe is what, 100 billion light years across? Life arose on Earth 3.5 billion years ago. That means that if a large scale civilization arose at any time between now and the first single-celled organism on Earth we could only know about it if they were located in the 0.0043% of the observable universe closest to us. The rest of the universe could be teeming with Kardashev I+ civs and we couldn't possibly know unless they were older than bacteria.

PS I will note that just in my lifetime, both f_p and n_e in the Drake equation have been radically revised upward. In a couple of decades the estimate for life on other plantes has gone from "no way, there aren't even many other planets out there" to "hey look, exoplanets!" to "hey, lots of exoplanets!" to "yawn, more exoplanets" to "ah! rocky exoplanets" to "rocky exoplanets in the habitable zone" to "rocky planets in the habitable zone with moons."

20 December 2012

Price Dynamism

HBR | Rafi Mohammed | Why Online Retailers' New Pricing Strategy Will Backfire

It's emerged as this season's hot holiday pricing strategy for online retailers such as Amazon and Best Buy: dynamic pricing. That's the strategy, employed by industries such as airlines, in which a seller moves prices up and down quickly, charging similar customers different prices for the same product or service. [...]

When a retailer uses dynamic pricing, it's only a matter of time before those who realize they've paid significantly more than others for the same product start screaming "unfair." This is when the big problem erupts for a retailer ‘ the trust component of its brand is at risk.

Ironically, retailers that don't employ dynamic pricing stand to profit from this situation. Costco, for instance, pledges it will not mark-up its products' costs by more than 15%. In an environment of price uncertainty, this pledge seems fair and engenders trust with customers.
I think this environment benefits Amazon, if they're wise enough to keep dynamic pricing subtle. Well, really I think it benefits anyone with enough data and analytic capability to make small, targeted adjustments, but that requires a lot of sales to generate the data and a lot of smart people to figure out how to make it work, and to me that list consists of... Amazon.

If you drop prices by $50 people are going to feel cheated. But if you wiggle the price of some item by 73 cents how many people are going to feel insulted?

The difficulty is that you need to be pretty sophisticated to figure out that p+0.73 is an optimal price, with both good data and good algorithms, and you need to be big enough that an extra $0.73 per item is significantly profitable. This benefits big, technologically savvy firms with lots of sales across many SKUs, lots of repeat customers, robust technical infrastructure, quantitatively oriented management, etc.

So I woulndn't cast this issue as "firms with dynamic pricing will lose to those with fixed pricing." I think a much better frame is "firms with the sophistication to wield dynamic pricing like a scalpel will beat firms that use it like a cleaver."

PS How many people are actually using tools like CamelCamelCamel to deal with a dynamic pricing environment? Not many, best I can tell. Which leads me to believe, at current levels of dynamism, they don't care much.

PPS This is an interesting moral situation. I have no doubt that Mohammed is making a proper positive description. People do feel cheated when prices are lowered after they buy things. Firms are going to have to deal with that, because that's the world as it exists. But normatively, this is gross. It's nothing but envy.

If you paid $600 for an iPhone, you must have thought it was worth more than $600. You wouldn't have bought it if you didn't think you were benefitting from the transaction. If Apple later sells it for $400 to someone else, so what? No one has taken anything away from you. So what if someone else later got an even better deal? Yes, you want the consumer surplus that they got, but we all want things other people get. That's not victimization, it's covetousness.

Yes, envy exists in the world, and we need to deal with that. But far too much of our public discourse treats it like it's noble. If someone else got something you want that doesn't make you a victim. The reaction shouldn't be "oh no, you poor thing, you've been cheated by a Big Bad Company." It should be "stop being envious, you whiner." This is a lot harder to explain to people, but it ought to be done.

I would hope at least the R.C.Church would be on board with this approach, but instead I hear a bunch of crypto-socialist liberation theology bullshit.

PPPS Okay, that last line is a rather low blow. (Skipper, if you're reading this, I'm sorry.) But screw it; I'm in a mood and I have no patience for subtlety right now.


John D Cook has two posts about The Lindy Effect that are worth reading. (In Cook's words "The longer a technology has been around, the longer it’s likely to stay around.")
Cook provides some mathematical backing to an intuition I had a couple years ago about style: the longer something has been popular, the longer you can expect it to be popular. My father wore Chuck Taylors 60 years ago. I wear Chucks today. Therefore I'm pretty confident that my children will be wearing Chucks 60 years from now. Ditto OCBDs, boat shoes, etc.

(Okay, so neither of us wore Chucks like this, but we do both love Lichtenstein.)

I had no idea this property had an actual name though. The math is convincing. I think the only big assumption is that lifespans are power-law distributed, which seems reasonable and appropriate in many domains.

17 December 2012

More Best Buy

HBR | Maxwell Wessel | Best Buy Can't Match Amazon's Prices, and Shouldn't Try

To survive their disruption, Best Buy should be looking for opportunities to optimize their business model around the jobs that Amazon can't do for customers. Maybe, for example, Best Buy could offer exclusive products for the customer who fears buying a product without seeing it in person. Customers with this job-to-be-done would happily pay a premium to buy a product they saw in person if it weren't available through online retailers for less. For instance, Best Buy could be the perfect home for retail distribution of products launched and funded through Kickstarter, which has seen 2012 as the year of the crowdfunded consumer electronics.
This seems like the opposite of a good idea.

(1) Kickstarter, almost by definition, will be turning out products that have appealed to people who are comfortable online. These are not the people who want to drive to a store to check out products.

(2) Kickstarter projects can not, by their very nature, afford to build tons of inventory to have it sitting in show rooms.

(3) Kickstarter projects, again by nature, are niche goods. If they were mass market they wouldn't need crowd-funding.
It could offer incredible on-site service that required customers to have purchased at a premium at a Best Buy (in the same way customers flock to the Apple's Genius Bar for help when the littlest thing goes wrong with their hardware).
We've just come through one of those annual blogo-storms about "Why can't Walmart be more like CostCo?" and once again the answer is "because they're different companies, and it is extremely difficult to change corporate DNA that much." Why can't Best Buy have incredible on-site service?" is about as sensible as "Why can't Chrysler make a Camry?" Perhaps I'm being fatalist, but that's just not what they do.
It could even focus more of its operations on the last-minute needs customers have for which no delivery network is fast enough, growing its automatic kiosk network.
Amazon already has same day delivery trials underway. Who has the kind of life where they can't wait 24 hours for electronics? Or, how many people would rather spend 45 running an errand than wait for a next-day shipment? (Answer: people with little employment, and little expendable income.)

Even if this kiosk business worked, why would Best Buy be particularly good at it? Why should an investor give $$$ to Best Buy to build out a network of kiosks when they could give those same dollars to Amazon or Red Box or New Egg or some start-up or WaWa or Aramark Vending or literally anyone else to do the same thing?

PS I realize I'm being a little hard on the HBR bloggers lately who have been posting "Best Buy, etc. could do X to survive" ideas. I think Best Buy is pretty well doomed, if not to bankruptcy than to market irrelevance. This makes it easier for me to say "no, that won't work; no, not that either." If I had to go to work for Best Buy as a consultant I couldn't really say "you're screwed, fold up shop now."  I'd be stuck trying to pitch long-shot ideas about Kickstarter and kiosks, like these HBR bloggers.

You have a 60" waist? Have some free stuff. You have a 23" femur? Tough shit.

The Economist | N.B. | Obese flyers: How should airlines treat larger passengers?

Traveling while obese can be stressful and humiliating. Many American airlines ask overweight passengers to buy extra seats or wait for the next flight if they can't be accommodated. Requesting a seat-belt extender can draw stares. And sitting next to someone who doesn't really fit in their seat can be uncomfortable for all concerned. So I was interested to note, via US News & World Report, Air Canada's fascinating policy with regards to obese flyers:
International airlines such as Air Canada address this issue more amicably: Because the airline considers obesity a medical condition, it provides overweight passengers with a free extra seat as long as they present a doctor's note.
This is remarkable, and appears to demonstrate a fascinating difference between American and international attitudes towards obese people
Why not do this for tall people?

We could have endless debates about how much of obesity is your responsibility and how much is luck of the genetic/social draw, but I think we can agree that tall people are not at fault for being tall.

So why is it that everyone should chip in to make sure people with one form of bad luck don't have to suffer the consequences, but people with another type of bad luck when flying are expected to deal with it on their own?

Is there a consistent moral rationale for why an airline ought to give away seats to the obese but not the tall?
Air Canada and other airlines with policies that say obesity is a medical condition are taking a financial hit for every extra seat they give to overweight passengers. That cost is probably being passed on to other passengers.
Perhaps this is only fair. We all subsidise the costs of providing on-board defibrillators for patients who suffer cardiac events on planes. We all subsidise the added medical costs of obese people. Why not airline tickets, too?
Oh, you want to play this game? Okay. Let's begin.

We don't pool the costs of everyone who bought books to read during the flight, so why pool the costs of extra obese passengers? People who buy over-priced meals in the airport concourse aren't subsidized by those who don't. Some people order a car service to get from the airport, some people take cabs, others take buses; they don't all chip in and evenly split the costs of ground transportation. Now (thankfully!) the people who pack light aren't responsible for paying to transport someone else's XL suitcases, duffle bags, golf clubs, baby strollers and skis.

I guarantee I can list more things we don't collectively buy than things we do. "Buy Your Own Stuff" is the baseline, default way things are done in a free society. The burden is on the collectivists to argue that something like extra-wide airline seats shouldn't obey that rule.

The entire social safety net boils down to protecting people from the consequences of bad luck.* What's always bugged me is how selective we are about which forms of bad luck we must Do Something about.

(* Well, also their own bad decisions, since we try so hard not to draw a distinction between the deserving and undeserving poor, but that's another matter. For simplicity and generosity's sake, let's confine ourselves to bad luck.)

You got diabetes and need $20,000 of medical services and you don't have an insurance policy that will cover it? No problem; it's now apparently a moral imperative that everyone else pitch in to pay for that. You had a tree fall on your home and need to spend $20,000 putting a new roof on it, but don't have an insurance policy that will cover it? Tough shit; you're on your own.

I wish someone could give me an a priori theory about which forms of bad luck we are supposed to insulate people against and which we aren't.

11 December 2012

"Banned on Campus"

Jacob Grier | Banned on campus

One of the quiet ways smoking bans have spread across the United States is via bans on college campuses. [...]

Encouraging schools to go smokefree is now official policy of the Department of Health and Human Services, which recently launched an initiative in collaboration with the University of Michigan to encourage campus bans on campus. The bans apply outdoors, [...]
“The CDC and surgeon general say there is no safe level of exposure to second-hand smoke,” said Julien Guttman, a GWU public health graduate student who is part of the advocacy group Colonials for Clean Air. “No matter how much science we have to back up what we are saying, there will always be individuals who see this as a restriction on their freedom.”
(0) I see it as a restriction on freedom, because it is. You might think it's totally justified. You might be right. But it's still an abridgment of freedom. Sometimes that's necessary. When it is, you justify it. You don't pretend it isn't real with a snotty little "oh, those wacky people who want freedom, how silly."

(1) The science is not on your side on this one, buddy. This is a religious gesture, not a health policy. Even the NIH admits that. And when you say things like this, it's kind of a give-away: "Guttman, like Manzo, emphasized the extent to which a smoke-free campus is as much about education and resources for smokers as about policy and enforcement."

Awareness, out-reach, education, etc. are last resorts of Crusaders. Notice that the game has now shifted from abridging freedom in order to protect non-smokers to abridging freedom in order to educate smokers about the risks of smoking.

(2) The "no safe levels" thing gets me every time. I would be drummed out of the room for asserting that any substance or activity had "no safe levels" in a risk analysis class. Safety and risk are not boolean. It's not a matter of "yes, safe"/"no, risky." These things exist on gradients.

PS Grier posted another gem from Guttman recently:
“We’re trying not to use the word 'ban,'” says Julien Guttman, of the GW campus advocacy group Colonials for Clean Air. “We encourage people to talk about a smoke-free campus rather than a ban on smoking.”
I encourage employers to talk about a Julien Guttman-free firm rather than a ban on hiring Julien Guttman.

Best Buy & Shopping

Newsweek Magazine | Megan McArdle | The Future of Shopping: Get set for luxury to win the battle for the brick-and-mortar marketplace.

“Clearly unsatisfactory” is what Best Buy CEO Hubert Joly called the company’s third-quarter earnings. It’s a droll bit of understatement—so French!—when words like “plunge” and “dismal” more accurately capture the numbers that the company reported on Nov. 20. Sales were $12.1 billion in the third quarter of 2011; a year later, that number had shrunk to $10.7 billion, with net income working out to just 3 cents a share. Five years ago, in December 2007, a share of Best Buy was worth more than $50; today the company is trading just under $13. This may be some sort of record—from profit powerhouse to basket case in under five years.
Like I was saying yesterday about diamonds, it's easy to say that the advantage of physical retailers is that they can provide "service," but it's not clear to me that having an actual salesman around providing "service" is actually that much benefit to consumers. It certainly can be, but I wouldn't jump to list that in the "pro" column without a lot of consideration.

This is an actual service:
[Best Buy] is also undertaking extensive renovations on remaining stores to refocus them around personal service—the one thing that Amazon can’t deliver via UPS. “With things like home appliances, people are going to want the things we offer, for example, the delivery to service and install. Or Geek Squad: thousands of people sitting in homes, doing installations, across all the platforms,” says Stephen Gillett, the digital wizard who helped lead a turnaround at Starbucks before joining Best Buy eight months ago.
But too often I hear people say physical stores can provide customers with service without really laying out what that's going to be. I don't know about you, but my experience with the employees at places like Best Buy is that they're usually disorganized, sullen and uninformed. They're not as bad as the employees at Staples, but they're not particularly pleasant to interact with or helpful. Perhaps this is the introverted geek in me speaking, but at a lot of retail locations I would pay extra not to deal with employees.  If I was in charge of a big box retailer I would be really hesitant about staking out a strategy that revolved around those people as linch-pin assets.

Another thing to consider: if "service" really does become a competitive advantage for traditional retail chains, there's not much reason to believe online chains couldn't snatch that away from them. No advantage is permanent, after all. (See: "What Killed Michael Porter's Monitor Group? The One Force That Really Matters".) Didn't Geek Squad start out as an online operation? Is there some reason Amazon couldn't staff up it's own teams of in-home installation & repair experts?

If you're Best Buy you need to invest in a couple of people in every location who are helpful and knowledgable enough to help people make good decisions about buying a TV. If you're New Egg, you only need a few dozen people who can do that in the whole country. They don't need a home stereo expert waiting around in every single store waiting for someone to come in looking for advice and education about BlueRay players. I would not simply assume that online retailers won't be able scale "service" the same way they have scaled up everything else.

One example: my experience with Land's End customer service over the internet and phone has been exemplary. Really first rate stuff. My experience with them in stores has been maybe 3/10 stars. I can get far better service not in person than I can in person, and that's within the same company.

“Home Depot and Lowe’s will survive, because people who go into those stores need a lot of help,” he says. So do many electronics consumers, and Dubé believes that the best way to win them back is to provide expert in-store service.
I'll believe most of the people who go into a hardware store need help. But I'd be surprised if most of the revenue comes from people who need much help. If someone could siphon off all the small contractors and tradesmen from Home Depot & Lowes I think those stores would end up looking a lot different.

I'm getting into speculation now, but I'd think number of builders who would be comfortable ordering most of their tools & supplies from an online retailer isn't going to be shrinking in the next twenty years. And it doesn't just have to be an online catalog-with-delivery. There's a lot of possibility for new business models like "tools-as-a-service.")
But upskilling the workforce means swimming against the trend for retailers, who have been steadily moving to more part-time workers on lower wages and shorter shifts. Sophisticated software now allows managers to project store traffic flows based on time of day, holidays, even the weather. Workers plug in the hours that they’re available, and the computer spits out a schedule that uses them only when they’re most needed, which can shave labor costs by 4 or 5 percent.

In a business with thin margins and high labor costs, this can mean a big improvement in the bottom line. But it’s hell on workers, who complain that they have to make themselves available for 60 hours in order to get 20—while never knowing which 20. [...] This is not the way to attract a high-end labor force that can sell your customers on service. But giving up those savings means that the price gap between brick-and-mortar and online stores grows wider. And the pressure to use shorter shifts will get even more intense next year when the Affordable Care Act begins to require that companies buy health insurance for any employee working more than 30 hours a week or else pay a stiff penalty.
Again, this works in New Egg's favor. They can hire someone to sit in front of a webcam 40 hours a week and answer customers questions from anywhere in the world rather than trying to guess when to staff experts for a couple of hours in a thousand places. (It's also easier to predict how many people you'll need, since you're trying to estimate a single Poisson process using lots of data rather than an independent process for every single retail location using much thinner samples.)

Is Grandma going to be comfortable using a webcam to consult with an expert before getting a new microwave? Maybe not. But her kids may be. And her grandkids almost certainly are.

And doing so is only going to get easier. If big box stores got into this mess in the first place because people got more comfortable doing things online and using smart phones and gadgets, then I'd say they'd be pretty unwise to adopt a solution that counts on people not being comfortable doing things online and using gadgets to shop.

10 December 2012

Diamonds & Shopping

HBR | Dina Wang | Blue Nile's Cheaper Diamonds Need Not Threaten Tiffany. Here's Why.

Blue Nile's primary advantage is that it offers diamond rings online at prices that are as much as 35% lower than traditional retailers. Their basic premise is that loose diamonds are commodities; quantifiable details (e.g., the 4Cs of carat, cut, color, clarity), a few photos and a price should be all a man needs to select a diamond. Blue Nile also provides simple-to-use educational tools and an impressive virtual inventory (they act as a middleman for diamond wholesalers). For men who hope to optimize affordability, speed and anonymity in their ring search, Blue Nile is a clear winner.
I didn't get Mrs. SB7's ring from Blue Nile, but if I didn't have a friend-of-a-friend connection to a diamond wholesaler/goldsmith in San Francisco I probably would have.  Even so, it was an invaluable resource when I was shopping. It gave me the ability to walk into any jewelry shop and not only judge the competitiveness of their prices, but also the degree to which they were bullshitting me.

I wish I could buy every product the way they allow people to. ("Here's a database of several tens of thousands of options, each measured along a dozen dimensions. Go.") It turns a bewildering decisions fraught with emotional peril into a simple optimization problem.

(I didn't actually build a program to solve that optimization problem for me, but I was sorely tempted to try. If they had an API I would have, just for shits-and-giggles, but the idea of trying to scrape the relevant info out of their database... uggghhhh.)
For example, for men who are hiring a diamond seller because they are nervous about finding the perfect ring, a brick & mortar retailer is a safer bet. In a store, they can gauge a particular diamond's sparkle with their own eyes, or they can confer with the jeweler, relying on his artistry.
Yes, you can confer with a jeweler and rely on their artistry. But more likely you'll be conferring with a salesman whose going to use all the same salesmanship shenanigans you would find at a car dealership.

Also, I don't have any psychophysics research to back this up, but I'll lay dollars to donuts that amateurs can't judge diamonds by eye, especially not in the circumstances you face when you're shopping for an engagement ring. Just for starters, there's a reason jewelry shops pour so much effort into lighting design, and it's not to make things easier on the customers.
As one young man I emailed with put it, he wouldn't buy from Blue Nile for the same reason he "wouldn't buy a used car from eBay motors. A car is more than mileage, year, make and model (does it smell? does it ride smoothly?) Likewise, a diamond is more than what is defined by the 4Cs."
But that's exactly wrong. A diamond is much, much simpler than a car. It can be graded, quantified, ranked, summarized. No, it's not just 4Cs, but 12 or 15 variables covers it. It is a commodity. There's a reason when you buy a stone they give you a map of all the microscopic inclusions and there's a reason certain sellers laser engrave their stones with a unique ID: without these things diamonds are indistinguishable from each other. You can tell your car from other similar cars without looking at the VIN; the same is not true of diamonds.
Though my friend recognizes that he was in part managing his own emotions, he didn't want to lose any sleep over his purchase by "optimizing cost over confidence." And that's the point. Blue Nile's online platform couldn't fulfill his job of being completely confident in his ring choice.

Of course, some men buying diamond engagement rings have another "job" they want done: signaling status. For them, a high-end luxury jeweler is better suited to that task.
This seems to be the thesis of this post, such as it is. There will always be some customers who will overpay for the sake of overpaying.

If I was Tiffany's, I would expect some better advice than "When it comes to a luxury good, some consumers see paying more as benefit rather than a cost." Great, thanks. I know that. They're already over-charging by 50% or 100%. I think they've pretty much figured out that some people see like higher priced luxuries.
For Tiffany, this means they should focus on cultivating their image of luxury and the "job" of signaling status, cutting the lower-priced "Return to Tiffany" line. In making Tiffany's little blue boxes accessible to everyone, they damage the brand's exclusivity and its ability to convey status.
Ah. Now here is some actual, actionable advice. Thumbs up.
Meanwhile, middle-tier firms like Zales and Jared should focus on customers who want to feel confident they're buying something their girlfriend will want.
I don't think this is a long-term winner for these firms. Everyone younger than... well, pretty much younger than me... has spent their entire life buying things online — typically sight-unseen. I think the number of people who feel intrinsically more confident buying things from a salesman in a store is monotonically decreasing.
To fulfill the job of optimizing confidence, they could, for example, host "diamond nights" at their stores during which friends can come try on diamond rings together. Once a woman finds one she likes, she could register her favorite with the retailer. The store could then make that knowledge accessible to her friends or to her boyfriend (preferably when asked!).
Problem One: a lot of women don't want to pick out their own rings, seeing it as unromantic.

Problem Two: a lot of men don't want to give the exact same ring that was asked for. Hell, a lot of people generally are reluctant to give someone the exact gift they requested. For all its efficiency, it seems uninspired and slightly uncaring.

My mother happened to ask for the exact Christmas gift I got her days after I had purchased it. Now it's going to look like I was just following orders rather than putting thought and care into gift selection.

Problem Three: what's to stop Blue Nile from setting up a similar service? Or Amazon? Why can't a man learn about his GF's jewelry preferences from her Pinterest account? I think the demographic window where this plan would work for Zales or Jared is shrinking.
On paper this seems relatively straightforward; so why is it, in practice, so hard? No company facing disruption likes to acknowledge that they can't win it all. But sometimes, relinquishing certain customers isn't an act of desperation. Rather, it returns a brand to where it should have been comfortably enthroned all along.
I'm glad we're on the same page re: this being straightforward. Wang is making a good point here.

East German MiE as Rorschach Test

Marginal Revolution | Tyler Cowen | East German markets in everything
Communist East Germany not only put its political prisoners to work making Ikea furniture, it also sold the bodies of its sick citizens to Western pharmaceutical firms for testing their new products, a new television documentary has revealed.
Here is more, hat tip to Kit in Germany.
One of two thoughts has to occur to you first:
  1. OMG the communist state is so evil for doing that! or
  2. OMG corporations are so evil for doing that!
I'm willing to bet that your first reaction is highly correlated with how you feel about a vast majority of political issues.

Poly Sci grad students, care to test that out?

Call It

This is another post left over from the days immediately following the election.

A lot of people have spilled a lot of ink in the last couple of weeks praising Nate Silver's numerical thoroughness. (Myself included.) This makes is all the more annoying to see things like this...
twitter | @asymmetricinfo

South Carolina projected for Romney with 1% of precincts in and Obama at 77% of counted vote. #statistics
and this...

I'm almost positive these kind of anomalies are the results of something like poorly synchronized database queries and not nefarious media plots to throw the election. But they're still really annoying.

More generally, it's frustrating to sit through an evening of TV coverage with anchors hemming and hawing about "ooooooh, too close to call right now, razor's edge, but it's looking good for Smith, but the margin is X, with Y percent of precincts counted, but it's still too early to know, but we're pretty confident, but stay tuned, but we think it's going this way, but..." No one ever bothers to explain what criteria are being used to "call" a state for one candidate or another.

What lead are they looking for given a certain percent of votes cast? What confidence interval are they waiting for? What statistical tests are they doing? Surely after all the Silvergate brouhaha they aren't "calling" states when some editor or producers just scratched his chain and says "yup, that's it! I feel good about 'calling' this state now."

Media: please give me something — anything — about the quantitative method you're using to make these decisions.

If not for me, then for the millions of people out there who are convinced you're doing it completely subjectively in order to discourage Red Team voters from staying in line. Because I don't think it's a conspiracy, but it sure would be nice to have you try to demonstrate that.

Now maybe that's hard to do on live TV. (Although with all the blathering they were doing, it's not like they couldn't squeeze some actually content-filled explanation in there. So what if it's hard? If we believed 1% of the mythology journalists tell about themselves they ought to be willing to try something hard like explaining some statistics to their audience.) It wouldn't be that confusing or difficult to list the number of votes a candidate is in the lead together with a confidence interval for what the "true" lead will end up being. At the very least give me a sidebar somewhere on all the election results websites explaining the method they use to "call" each race.

Politics, Exponentials and Sinusoids

I wrote the following a couple of days after the election, at which point I promptly forgot about it. I figured it might be worth dragging out, even though an entire month has passed.

I'm a libertarian. Neither Obama nor Romney are, by a long shot. So why was I disappointed (not at all surprised, but still disappointed) by Obama's election?

Option 1 is that I was raised conservative, and while I've mostly out grown that, there is still some lingering affinity.

Option 2 is that libertarians and the right are more natural allies, either in principal or just as a function of contemporary American politics. More ink has been spilled on potential left-libertarianism fusion that I care to rehash.

Option 3 is that I'm surrounded by Dems who I disagree with, and so I want to see there guy loose. If I lived in a Team Red area I would be surrounded by annoying Romney suporters and would be happy to see their guy loose.

Option 4 is that while both Obama and Romney are miserable, Obama is worse. This lesser-of-two-evils situation makes sense, but it also seems like a way to rationalize any of the more emotional reasons listed above. When either one of them is so abhorent to me, picking which one is worse is a mug's game. This post is an attempt to figure out how after the fact how I implicitly measured them against each other, so that doing so again in four years -- when I will inevitably be presented with two wretched candidates -- will be a little more rational.

Megan McArdle was right that no matter what the outcome of the election was, my life will go on mostly the same. I spent last night watching Downton Abbey with my wife, doing a bit of data analysis for one of my advisers, cooking a delicious (and cheap) dinner, teaching myself a bit about Sobel filters and trying to apply them a new art project. These are the bread and butter of my life, and they won't change no matter who wins what office.

(Sidenote: for all the people talking about how polarized American politics is, do me a big favor. Just hold your tongues until we see something like the Nika Riots. Billions of people and man hours were poured into selecting our next leader, the selection happened according to plan, and no one got stabbed over it. Not one single stabbing, or beating, or riot, or arson, or assassination. Do you know how rare that is in the history of the world? Maybe I'm setting a low bar, but can we take a second and appreciate that the worst reactions that happen when we get a new president is that some partisans freaked out Facebook or Twitter or cable news or talk radio? If the worst reaction we get from people is them screaming "OMG! We're all doomed! I'm packing my guns and MREs and holing up outside of Saskatoon!" then we're doing okay.)

Waking up today, or Janurary 21st, 2013, or 2016, it doesn't really matter to me who won the election night. I got up, made some coffee and got to work. I thought about when I'm going to the gym this week, whether I should go to a seminar tomorrow afternoon, and whether I could finish reading Swamp Thing before it's due back at the library.

So on that hand, the election was pretty irrelevant to me, and I'm happy about that. But on the other hand, it might very much matter who won last night once 2063 rolls around, because some small changes can compound a lot between now and then. Some policy changes are exponential, and others are sinusoidal.

Let's start with economic growth, since that's where my metaphor is the most literally true. Say a president implements some policies like, I dunno, a massive health care "reform" that doubles down on one of the worst aspects of our current system, namely, the tight coupling between employment and insurance. Say that lowers economic growth by half a percent a year. That doesn't much matter now, or even four years from now. I'm still going to get up in the morning and do the same things I would have otherwise and live the same life. But in fifty years that lost growth will matter a lot. If growth is 3.5% instead of 4.0%, in 50 years we'll be 21% poorer than we otherwise could have been.

This isn't just true of economics though. I think a lot of political processes are exponential like this. A bad Supreme Court ruling now is bad, but the way one ruling compounds on another, stacking bad ruling on bad precedent like a castle on a swamp, means that a single bad decision now can become really terrible in a few decades.

Giving the police (or pseudo-police) the power to search me without any cause before I get on a plane leads to giving them the power to search me before I get on a subway, which gives them the power to search me before I get on a bus, which gives the them power to search me when I walk down a sidewalk.

A lot of political opinion and policy comes and goes. The pendulum swings back and forth on immigration. For every decade in American history where people are streaming out of steerage class and up onto Ellis Island, there's been another where the borders are locked down. A president who screws up immigration policy is a disappointment, and has costs now, but those mistakes can be corrected later when the pendulum swings back. Sadly a lot of things don't look like pendulums to me, at least not on the couple-of-centuries time scale relevant to a modern nation state.

Privacy doesn't seem to ebb and flow. An administration that claims some prerogative to spy on its citizens will be followed by one who claims even larger prerogatives to do so. One that claims the privilege of arresting and holding indefinitely anyone it claims to be dangerous, anywhere in the world, will be followed by one who claims the ability to execute them as well.

The things I don't like about another Obama presidency seem to be mostly in this exponential growth category: poor economic policy, poor court nominations, executive privilege regarding secrets and whistleblowers, spying, executions & use of force abroad, centralization of local matters like education policy, the president-as-venture-capitalist-in-chief, corporatism, protecting consumers from themselves, etc.

The issues I wound't have liked about Romney seem to fall into one of three categories:

(a) Things that are sinusoidal, and destined to swing back against him sooner or later, like immigration.
(b) Things that are exponential but that Obama will also be terrible about accelerating, like the War on Terror.
(c) Things that are exponential* but moving in the opposite direction as him, like gay marriage and marijuana liberalization.

(* Actually, I need to introduce a third category: sigmoidal policies. These behave like exponentials but only for a limited time, because there's a natural ceiling on how much they can change. Gay marriage can only become so legal. In about half my lifetime it's gone from being non-existant to being treated exactly the same as heterosexual marriage in many states. However there's no place else for it go in those states. It's essentially a non-issue there now. The ride is over.)

Had Romney won he would have done plenty of things I wouldn't like, but I don't think would have much affect in 50 or 100 years, while Obama has the ability to do some bad stuff now that could make my twilight or my children's lives a lot worse than they could have been.

Next time I've got to mentally tot up a pro and con list, I'm going to divide every issue up into exponential, sinusoidal and sigmoidal categories.

06 December 2012

I do not think Kevin Drum and I live in the same universe

Mother Jones | Kevin Drum | Quote of the Day: Maybe Raising Tax Rates on the Rich Isn't Such a Bad Idea

Second, there's Coburn's point. If you want to have any chance at all of broadening the base and lowering rates in the future, you can't close loopholes now. You need to leave them there as bargaining chips. Tax reform will be more likely if rates are higher (making them easier to lower) and loopholes are all still intact (giving you plenty of stuff to close in return for lowering rates).
This seems more than a little bonkers.

Couldn't you apply this logic to any policy you want your opponents to cave in on? "Well, if you really want marijuana to be more legal, you should allow drug warriors to crack down on it more now, so you have room to negotiate in the future." "If you're serious about increasing the role of government in health care, you should back free market reforms now, which will preserve your bargaining chips for later." What?
There are two ways to come at this. First: which is simpler and easier, raising rates or closing loopholes?
By "easier" he means "more politically expedient." There's nothing wrong with that. I simply point it out because that is not what "simple" or "easy" mean to me.
I'd say raising rates is easier, and if it's done now it will make it harder to raise them again in the future. This means that if Democrats want to soak the rich again, they'll have to do it via closing loopholes, which is a harder lift.
Ah. Of course. Because in the annals of government, no one has ever managed to raise tax rates twice in a row. Tax rates never, ever ratchet up. It's practically impossible for a politician to double down on a policy. No one has ever followed up "Hey, let's take some money from those guys over there who I don't like" with "We need more money; Let's get it from those same guys again!"

I suppose this makes a certain kind of sense if you're assuming you live in a world in which you could not possibly simplify the tax code some now, and then do so again later. Well, "sense" is too generous, but it's somewhat consistent.
If you're a slave of Grover Norquist and hellbent on never raising revenue in any way at all, none of this matters. But if you're smart enough to pound sand, you know that raising revenue is inevitable eventually.
Obviously anyone who thinks spending is the problem is just plain dumb. Obviously.

(chart via Coyote Blog)

PS I think the key to understanding Drum world may be that he does not see the GOP's goal as having lower tax rates on a broader base than the US currently does. He seems to see the goal as passing a piece of tax reform legislation which legislators can claim lowers rates and broadens the base from whatever the then-current status quo is. Which tells me Drum has spent too long inside the Beltway and needs to go take a multi-year vacation somewhere on the other side of the Appalachians.

We're at A right now. The GOP's goal is to get to B. What Drum is suggesting is that somehow moving from A to C is the way to get there. The goal isn't just to move down to the right at some point. You'd have to be even more cynical than I am about congress to believe that's the only motivation. No, the goal is to get from A to B. And C is not on the way.

30 November 2012

Loopholes aren't

Politico | Lauren French | Tax loopholes alone can't solve fiscal cliff

Raise revenues and reform the Tax Code? Easy — just eliminate all the tax loopholes, right? Good luck with that.

“Eliminating loopholes” sounds a lot better than “raising rates”: The tax rate is what I pay, and a loophole is what the other guy gets.

But the biggest loopholes in the U.S. Tax Code — generally referred to as tax expenditures — aren’t just the tricks of the trade for millionaires with offshore bank accounts. For the vast majority of Americans, they’re just how things work: You don’t pay taxes on your health insurance or Medicare benefits; you contribute tax-free to your 401(k); and your mortgage interest pushes down your tax bill each year.
(1) "Loopholes" is a red-flag word. When I hear it I immediately raise my estimation that the person talking is either confused or deceitful. Most of the time when people say "loopholes" they're really describing an intentional feature of the system. This is what "loophole" actually means:
"A way of escaping a difficulty, especially an omission or ambiguity in the wording of a contract or law that provides a means of evading compliance."
The mortgage interest deduction, as much as a dislike it, it not an omission, and it's not ambiguous. It's very, very explicit. It's no more of a "loophole" than the income tax itself.

(2) Yes, you do pay taxes on 401k contributions. You just pay them later. The timing is irrelevant; they're still taxed. Maybe we should tax retirement savings now, maybe we should tax them later. But it's not a "give-away" and it's not "costing the government money." It's certainly not an "upper class entitlement." Sure, high earners avoid more taxes now. Then they pay more taxes later.

(3) You don't pay taxes on health insurance if your employer buys it for you. This is terrible way to do things. Unfortunately ObamaCare doubled down on the link between employment and insurance instead of fixing this illogical, historical accident.

(4) Paying taxes on Medicare benefits seems a little silly to me. Why do we want the treasury to write people checks with their right hand then snatch back a portion with their left? Am I missing something? Sure, doing so would technically "raise revenue" but only in the most facile, myopic way. Is this just because we're incapable of admitting that we ought to be writing smaller checks in the first place? (5) The mortgage interest deduction is flat out dumb. I have a hard time imagining a worse way to structure that system. Dumb, dumb, dumb. Why are we rewarding this particular form of indebtedness? There are a thousand ways we could do this better.

(5) Capital gains rates are not a "loophole" either. They're there on purpose.* Investment income has already been taxed, back before the money was invested. It's being taxed a second time (or more, when you consider corporate taxes as well.) The fact that you pay 15% on your gains instead of 35% or whatever doesn't mean the treasury is being cheated out of 20% because of some "loophole." There's no reason to think that when the IRS takes a second bite that it needs to be the same size as the first bite.

(* Some of the stuff with sweat equity and the way financial workers are compensated is a little dodgy. There are some loopholes there. But as a general rule, no. Capital gains taxes are just The Rules of the Game, just like income taxes and property taxes and sales taxes.)

PS French actually does mention some of this stuff later in her piece, like the tax benefit of health insurance only coming into play when your boss buys it for you. But I'm too wound up to go through and excerpt from all over her article and respond.

PPS No, I take that back. I can't resist commenting on two things later in the article.
• The earned income tax credit — $58.4 billion

[...] But recently, deficit hawks have balked at the $4.2 billion the [EITC] has paid out to ineligible illegal immigrants as a reason it should be taken off the books.
Here's a bright idea: how about you don't pay it out to illegal immigrants? (Or any non-citizens, for that matter.) Far be it from me to expect the IRS to run a tight ship, but you know, just maybe, maybe... it might be possible to operate this program correctly rather than just scuttling the whole thing?
• Exclusion of gains at death and the gift carryover exclusion — $51.9 billion

These exclusions allow hundreds of billions of dollars to transfer hands tax-free through gifts and inheritances, a huge revenue loss for the government.
You ever read Pillars of the Earth? I feel like I'm living in that medieval morality, tax-wise. "Oh, you want to ford this river? Well starting today you have to pay a tax to the king for doing that. You want to hold a flea market? The bishop demands a cut for granting you the privilege. You're gonna shear a sheep? Can't do that until you pay the baron his fee." It all seems to boil down to: we need money; a thing is happening; let's make anyone doing that thing pay us. No rhyme or reason.

That's how I feel when people talk about gift and estate taxes. As if every time money changes hands the fisc has some natural right to step in between the two people and demand a rake. The current gift-and-estate tax regime is screwed up in a lot of ways. But I can't abide this idea that whenever money changes hands the state has the right to siphon some off. What's worse is thinking that not only do they have the right to, but that whenever they don't do that they're somehow being cheated.

You know grad school was a good decision...

... when you laugh harder at Chomsky gag than anything else all week.

(Although this is a close second. Mrs SB7 laughed even harder at it than I did. That's how I know she's a keeper.)

Someone came along and cleaned tore down the carefully curated webcomics that we had taped to the lab door. This might go up as the seed of a new collection.


Modern American Trade Unionism

Asymmetric Information | Megan McArdle | Unions Organize Walmart Protests; Rest of the Nation Goes Shopping

But the Black Friday bargain hunters apparently simply pushed past the scattered protests in search of cheap flat-screen televisions—and the progressives who seem most on fire about this campaign are not really very likely to be Walmart shoppers. Which could be a metaphor for the whole US labor movement.
Modern American trade unionism: people who don't shop at, work for, or own shares in companies telling those who do how to run their lives.

28 November 2012

Does complaining about inequality harm people by making them feel more unequal?

EconLog | Garett Jones | How do you Sustain the Second Freest Economy in the World?

In an excellent Singapore econ-travelogue, Scott Sumner writes:
My theory is that leftists don't really mind a place where income is unequal, they don't like places where income looks unequal.
If this is a good theory, then leftists should spend much more time praising the pro-consumer side of capitalism. I wish I saw more pieces like Sebastian Mallaby's recent WaPo column "Progressive Walmart. Really."*

(* One very, very minor criticism of Mallaby's piece. His conclusion begins "Companies like Wal-Mart are not run by saints." He is right. But I wish he would have mentioned that no organization is run by saints. Not Walmart, not PIRG, not the AFL-CIO, not the USPS, not your local school board, not the Red Cross, not the the Boy Scouts, not the Church. None of them.)

If Sumner's theory is correct, we should also see leftists embracing studies which show inflation measures overestimate prices paid by low income people. Instead I tend to see the mood affiliation running the opposite direction.

They may in fact prefer societies in which income is unequal but looks equal, but in my experience they wish to talk about society as if it were even more unequal than it is. Doing so gives them more political ammunition for their favored reforms.

Jones links Sumner to Mickey Kaus:
This is close to what pioneering blogger Mickey Kaus has been pushing for since, oh, the invention of the New Democrats. I've never read his 1992 book The End of Equality, but after 13 years of reading Kausfiles I think I've got the main idea: Higher income inequality is inevitable, let's get used to it and let's respond by increasing civic equality, social equality. His idea is to tinker with the state so that people feel equal in civic life even if money inequality is high.
If we wanted to pursue this, it should be people most concerned about inequality who push all the statistics about money not mattering much in elections. Instead we see the opposite. A leftist following Kaus' playbook should be downplaying the effects of Citizens United, not screaming about the "end of democracy."

I think it also agitates for de-coupling housing purchases from education spending, (ie school choice) since doing so would increase social equality without affecting income equality. (At least not for a generation or more, and even then would do it without redistribution.) Again, we see the opposite from most of the Left. Is this because they disagree with Kaus, or because they are more concerned with the minor inequality experienced by unionized school employees than the major inequality experienced by poor students?

Returning to Jones:
Here's a sentence Kaus likes from Reagan's 1992 GOP convention address:
Whether we come from poverty or wealth... we are all equal in the eyes of God. But as Americans that is not enough — we must be equal in the eyes of each other.
This is an essentially meaningless quote, because it can mean completely contradictory things based on how we choose to define "equal."

I suspect this is precisely why it appealed to Reagan. If you read what follows in his speech, it's all inspirational pablum. 'Crime is bad! Education is good!' The only actual policies he calls for are a balanced budget amendment and a line item veto. Otherwise it's just 'Hooray for Freedom! Hooray for Prosperity! Hooray for America!'

It is the cheap cloth, the cheap cotton and rayon fabric, boots, motorcars and so on that are the typical achievements of capitalist production, and not as a rule improvements that would mean much to the rich man. Queen Elisabeth owed silk stockings. The capitalist achievement does not typically consists in providing more silk stockings for queens but in bringing them within the reach of factory girls in return for steadily decreasing amounts of effort.
— Joseph Schumpeter, Capitalism, Socialism and Democracy (1950)

What's great about this country is that America started the tradition where the richest consumers buy essentially the same things as the poorest. You can be watching TV and see Coca-Cola, and you know that the President drinks Coke, Liz Taylor drinks Coke, and just think, you can drink Coke, too. A Coke is a Coke and no amount of money can get you a better Coke than the one the bum on the corner is drinking. All the Cokes are the same and all the Cokes are good. Liz Taylor knows it, the President knows it, the bum knows it, and you know it.
— Andy Warhol
I'd like to see much more emphasis on consumption inequality than on income inequality. Warhol was no reactionary. Neither was Schumpeter, for that matter. They both managed to recognize how markets can help everyone, not just the rich. I wish more people recognized that, not just because it's the truth, but because like Sumner and Kaus say, it would make for a healthier, less envious society.

27 November 2012


The Economist | Fine art: Collectors, artists and lawyers

Alas, plenty of other experts are now too scared of lawsuits to authenticate pictures, says Clare McAndrew, the founder of Arts Economics, a consultancy. Early this year the Andy Warhol Foundation dissolved its authentication board after spending $7m to fight a lawsuit from a disgruntled London collector. In September the Estate of Jean-Michel Basquiat and the Keith Haring Foundation stopped authenticating works by the two late artists. Last year the Roy Lichtenstein Foundation dissolved its authentication committee rather than “jeopardise our health and well-being”, says Jack Cowart, its director. In the past five years insurance policies taken out by art authenticators have more than doubled at Hiscox, an insurer.
Can we all agree this is bad litigation? It's a sad state of affairs if society can't enforce a contract that says "Party A agrees to tell Party B whether they believe artwork X is authentic, and party B agrees not to sue Party A if they don't like the answer."
How will the art market adapt? China offers a clue. Art expertise in China often carries little weight because authenticators are thought to be in cahoots with a dealer or seller, says Shin-Yi Yang, a curator in Beijing. So living artists make more money, since they can personally assure buyers that a picture is not a fake, he says.
Interesting. Can I propose another possible adaptation? Increased demand for digital works which can be signed using public-key cryptosystems.

Digital signatures aren't really my thing. (Hell, crypto isn't really my thing, but I'm trying to remedy that now.) But I think as long as an artist/their estate keeps their private key secret this should work well. It should even allow authentication of limited additions, which I'd guess is a problem with digital works currently. IIRC they're usually packaged with a physical certificate of authenticity from the artist, which is bound to be orders of magnitude easier to forge than either a traditional painting or a digitally-signed file.

the inverse of "pics or it didn't happen"

Run Free 2013 by Ridiculo.us

So on February 2, 2013, we are faking a marathon. We’ll have all the trappings of a real marathon - race banners, race t-shirts, race medals, race sponsors, race bibs, a race program... you get the idea. We just won’t have any actual... racing. We’re calling it Run Free 2013. Because, well, it’s run-free. It’s free of running.
The fact that this exists: I like it.

PS I want to take this opportunity to tell the Internet that I am proud of Mrs SB7 for finishing her second half marathon recently. Way to go, Lady Friend.

Wrinkly Tax Codes

Asymmetric Information | Megan McArdle | More On That Wrinkly Tax Code

Reader Mark writes in to highlight some of those EITC problems I spoke of earlier:
Now, if the GOP would like to cement a generation of voters that would keep them in office and make them competitive in places like OH, NY and PA, end the injustice of the current tax code and make it easy to be a citizen instead of having that sinking feeling that you are both getting hosed and probably a felon at the same time.And guess what, it’s the right thing to do. Too bad the donor base is addicted to tax breaks and crony capitalism instead of the good of the nation.
If given the choice between the current tax code and one where I would pay more dollars but not simultaneously feel like I'm being cheated and being forced to worry about the IRS claiming I'm a felon, I would take the latter.

I don't care if the Red Team or the Blue Team does this. I'll consider backing whoever allows me out of a tax system that makes me feel simultaneously afraid of being a sucker and of being caught breaking laws I didn't know existed.

PS McArdle concludes with:
This is an enormously tough problem with the EITC. I support the idea that our tax code should support low-wage work. But the phase-outs create large income ranges where raising your salary doesn't make you noticeably better off--may even make you worse off.
In a later post she says the following:
This is a basic problem wtih means testing. But it's not necessarily an argument against means testing so much as it is an argument for very long phaseout periods--and especially, for staggering the phaseouts so that they don't all hit in the same income range. This is why the poor can face marginal tax rates that exceed 100%: hit a certain income range, and your EITC, Medicaid, food stamps, and other forms of assistance all start to phase out at once.
We can try to smooth out these cutoffs, stop using step functions, distribute them across a range of incomes rather than stacking them all together, make phase-outs more gradual, make rates differentiable, but at the end of the day there's no way to have both generous benefits and low marginal rates for low income people. There are plenty of ways to make this system a little more reasonable — and we should take them — but when all is said and done that's a circle that can't be squared. Every time you raise benefits for the $0 income person you make the marginal rates of low income people higher as well.

26 November 2012

In which I piggy back on another blogger's advice column

Asymmetric Info | Megan McArdle | Ask the Blogger

Dear Blogger:

Which leaves me with something of a middle choice: masters programs. I'm very interested in pursuing an MPP--it seems most of those policy-related jobs that used to be available to smart BAs now require a law degree or an MPP.
I've heard statements like "I'm really interested in public policy" way too often in the last year.  I'm trying my best to take off my libertarian hat here, and I still can't find any way of translating such statements into anything other than "I'm really interested in figuring out how to tell other people how to live their lives." Seriously, non-wookies, fill me in here. I'm sure most of the people who say "I want to do public policy" do have pure intentions, but how is this different from wanting to order other peoples' lives?
[From McArdle's response.] And second, to pay very close attention to the ranking of your school within your field of graduate study. Yale has the number one law school, and an MBA program with a so-so reputation. People often erroneously assume that a top-ranked school in a field they know well must mean that all the programs are good. But even a school like Yale has some notso-hotso graduate programs.
This is very, very important.

People have pretty good background knowledge of how undergrad programs stack up. Those rankings are only loosely correlated with grad schools rankings. In turn, those are only loosly corrleated with the rankings of departments in particular fields. And then those are only loosely aligned with sub-specialties within fields. Without even considering which lab you work in or who your advisor is, either of which have their own rankings and prestige, we're already pretty much entirely uncoupled from what you read in USN&WR or Princeton Review a few years ago.
Those of us who have been working for longer have noticed that shockingly few of our acquaintances are still doing data entry or driving a fork lift twenty years later. As long as you're actively looking for the next thing, and talking to employers about how you might become the worker they want, you will eventually figure out what you're good at, and find a better job.
I know too many people who are terrified of sitting down with their boss and asking what they need to do to take the next step up the ladder. They're scared enough of that confrontation (or scared that the boss will tell them to work harder/longer than they want to?) that they decide to take huge gambles on multi-year graduate programs or fliers on total career shifts instead.

"Making assistant manager is hard." Okay, sure. "So I'm quitting and going to law school." Right. Because that isn't hard. Or time consuming. Or expensive.

"I don't see there being much a career path for me at IniTech." Alright. "So I'm quitting to take an internship with an event planner in Denver." I think what you meant to say was "I like reading wedding magazines and it would give me an excuse to be on the bottom rung of the ladder for a few more years."

This poor guy must think he's in Snow Crash

Not Always right | Leaving The Country Is Fined By Us

(I work at the Dutch version of the DMV. We get a lot of calls by people trying to get a fine waived. In most cases we can’t, and in some cases we can give the customer some slack.)

Me: “This is [name]. How can I be of service?”

Customer: “Yes, I want to complain about your service, and I want you to remove me out of your system!”

(I’m a bit confused; since we are a government-controlled agency, it is a bit strange for someone to ‘cancel their subscription’ with us.)

Me: “What issue are you having with us?

Customer: “I got fined because I didn’t get my car inspected in time for my MOT!”

(Note: Translated into English, ‘MOT’ is our Periodic Vehicle Inspection. Normally, we send out reminders as a courtesy, but we cannot be held responsible if a customer forgets to get their MOT done. This customer in particular did not have an MOT for about 9 months.)

Me: “Ma’am, I am going to try to repeat what you are saying so I know that I understand correctly. You want us to ‘cancel’ your ‘subscription’ to us, a government-controlled agency to which you, as a person living inside the country itself, are responsible to abide the laws on which owning a car are set?”

Customer: “Yes, I want you to remove me from your system! I wish to go to your competitor!”

Me: “Miss, with all due respect. I first of all do understand the discomfort of getting a fine for these kinds of laws. But don’t you agree that moving to a different country is a bit drastic, just because you do not want to get fined for your car?”

Customer: *continues ranting*

Me: “Miss, again, with all due respect, I do apologize for letting you feel like this. However, we have laws to which you, me and everyone needs to abide to. How unfair the fine may seem, I cannot undo it and I cannot take you out of our system.”

Customer: “Why not?!”

Me: “Because that would be considered a criminal offense. It would also involve you, leaving our country and living abroad. We do not have a competitor. I wish I had other news for you, so I’m truly sorry.”

Customer: “F*** you!”
"We do not have a competitor." Yeah. That's the problem right there.

My fondest wish is that I can show this story to my grandkids in 50 years and they'll think the DMV worker is the crazy one for supporting a state that doesn't contract this kind of thing out to multiple providers.

I call shenanigans on David Brooks again

NYTimes | David Brooks | Why We Love Politics

I hope everybody who shares this anti-political mood will go out to see “Lincoln,” directed by Steven Spielberg and written by Tony Kushner. The movie portrays the nobility of politics in exactly the right way.

It shows that you can do more good in politics than in any other sphere.
(1) You can also do more evil in politics than any other sphere. Seriously, name me ten people who have done as much evil without politics as Stalin, Mao, Hitler, Pol Pot, Hirohito, Caracalla, Napoleon, Temujin, Robespierre, ... Do I have to go on?

(2) Tell that to Norman Borlaug, or Joseph Lister, or John MacAdam, or Henry Bessemer, or Johannes Gutenberg, or Luca Pacioli, or Vint Cerf, or ... Again, do I have to go on?
The challenge of politics lies precisely in the marriage of high vision and low cunning. Spielberg’s “Lincoln” gets this point. The hero has a high moral vision, but he also has the courage to take morally hazardous action in order to make that vision a reality.

To lead his country through a war, to finagle his ideas through Congress, Lincoln feels compelled to ignore court decisions, dole out patronage, play legalistic games, deceive his supporters and accept the fact that every time he addresses one problem he ends up creating others down the road.
"You know what's great about politicians? That they're so good at lying and cheating and arm-twisting and back-stabbing in order to achieve the goals they've personally decided are right." Wow. How noble. Brooks' entire argument boils down to the ends justify the means.

No, actually it's worse than that. He's claiming that dirty means make the ends even more glorious.

Can you imagine the NYTimes opinion page publishing this blather about someone in any other profession? "Oh, I'm ever so thankful that the CEO of ADM lied and cheated in order to get draught resistant GM potatoes in the hands of peasant farmers. Isn't it great that he ignored courts, doled out patronage, played legalistic games and deceived friend and foe alike, all to achieve a good end?"

If you weren't convinced that Brooks had a tremendous hard-on for the State I hope this column has cleared things up for you.

25 November 2012

"High Virtue and Low Cunning"

EconLog | Bryan Caplan | Modelling the Marriage of High Virtue and Low Cunning

Suppose Brooks' claims are factually correct. What economic, psychological, and/or sociological model(s) would explain them? Why precisely would good political ideas need to be shrouded in deception?
Wait a sec. Perhaps this is my ND-going-to-the-big-game beer induced haze talking. But in what universe are we starting from the assumption that political ideas are not shrouded in deception?

At this point I'm incapable of enough metacognition to know if this is the High Life or my own misanthropic anarcho-libertarianism talking, but why wouldn't deception be the baseline? If we could simulate a billion human civilizations I'd expect good ideas and deception to be going hand-in-hand in a vast majority of them. Maybe that's just my priors talking but... But nothing. I'm going to go finish this pint, wipe down the kitchen, and turn in.

PS I'm beginning to realize that I need a list of "just barely smart enough to be dangerously wrong" list, and that Brooks needs to be highly placed on it. But that's a thought for another evening. G'Irish. I'm out.

21 November 2012

Receiver Operating Characteristics

Cafe Hayek | Don Boudreaux | Testing Jonathan Haidt

I happened across this paragraph from the site breastcancer.org – not a political or ideological site:
Biopsy is usually a simple procedure. In the United States, only about 20% of women who have biopsies turn out to have cancer. By contrast, in Sweden, where cost accounting is much stricter and only the most suspicious lesions are biopsied, 80% of biopsies turn out to be cancerous (malignant).
Notice how this account can easily be read as evidence (1) against the alleged superiority of heavy government involvement in financing health care, and (2) for the alleged superiority of heavy government involvement in financing health care.
Sweden and the US sit on different points along the biopsy ROC curve. That is all you can conclude from this extract.

File under: "benefits to studying Machine Learning that have nothing to do with technology."

17 November 2012

"The learning pill"

Daniel Lemire's Blog | The learning pill

Given a choice between a college graduate in Computer Science and someone who has a leading reputation on stackoverflow, which corporation would pass on the college graduate?
Answer: probably the corporation that built StackOverflow.

Draw your own conclusions.

PS I don't mean StackExchange will hire people with a good reputation on their own system. I mean StackOverflow was built by a team of people who valued ability over credentials. You'll notice Jeff Atwood, one of StackExchange's two co-founders, does not list "look at their college transcript" on this list of "how to hire a programmer." Joel Spolsky, the other co-founder, puts more emphasis on college than Atwood, but he's said that's for the pre-selection effect. He's as happy to see McKinsey or the Marines on a resume as Harvard because what he's interested in is outsourcing the effort of screening candidates to other selective organizations. They probably discussed their hiring process half a dozen times on their podcast, which was extremely educational about how good software companies actually operate.

PPS Spolsky has repeatedly written that he's a big, big fan of hiring directly from his internship program. He recruits sophomores more than seniors, IIRC. So while he places import on college, he doesn't seem to place much on having finished college.

12 November 2012

There's bias and then there's bias

Rhymes With Cars & Girls | Sonic Charmer | Nate Silver finds and quantifies evidence of (D)-favoring voter fraud

I wish this was true. I really do. But Charmer said people should call him out more often, so that's exactly what I'm going to do.

As much as I wish I could embrace this thesis, I think Charmer is making a big mistake, and that's conflating bias and variance.
Nate Silver has an interesting post summarizing how the various polling firms did in their state polls. Generally it looks like he finds most of them had consistently overestimated the (R) vote. He throws out the usual cuffed explanation for this (not enough cell phones, younger people have cell phones, younger people are more (D), bla bla).

But he appears to miss the elephant in the room, which is cheating. But surely no serious, scientific, quantitative genius person such as Nate Silver can possibly forget or just un-scientifically ignore the fact that the final vote contains some nonzero amount of cheating.
So far, so good. Cheating accounts for some fraction of the discrepancy. A non-zero fraction, but also a non-one fraction. Even if all the polls are perfectly done, there will be some tiny, tiny correlation between them. Just like it's safe to assume at least one ballot was fraudulent, it's safe to assume at least the estimate of one vote did not cancel out when all the different polls were conducting and aggregated.
Now – to echo a bunch of arguments I made against Silverbating righties – the fact that almost all these polls from all these different polling companies with all different sorts of methodologies find a consistent, systematic “(R) bias” just beggars belief. No quantitative-minded person can just accept that as the result of random chance. Sure, there will be errors and biases but wouldn’t the errors and biases cancel each other out? How likely is it that virtually all polls would come out with an (R) bias? That is an extraordinary claim which requires extraordinary evidence, which Nate Silver does not have.
Here's where things get semantically tricky, because "bias" has a particular meaning which is different from the one we use when we talk about politics. At least it does according to the quantitatively-minded persons who taught me quantitative things.

If a system has bias, it will make errors on a particular prediction regardless of the training set (or in the case of polling, sample set) used. Variance, OTOH, are errors that will change for that particular prediction based on which training set/polling sample is selected. Bias is, roughly, a systematic error, while variance is a result of "noise."

For example, the C4.5 decision tree learning algorithm has a bias towards orthogonal classes. If the true decision surface isn't orthogonal in a particular dimension then it will always make some errors because of that. If the true decision surface is orthogonal but slightly amorphous, those will cause variance errors.

(The left is variance. If you're predicting blue o's will be on the left of x=0.5 and red x's on the right, you'll have gotten some wrong just based on the random noise. The right is bias. If you still thought x=0.5 was the dividing line between classes but it turns the true dividing line is the solid purple diagonal line then you're going to make consistent and correlated errors no matter what random sample of x's and o's you select.)

The errors of every predictive system are a combination of bias and variance. In order to avoid bias you need flexibility, but increased flexibility makes you subject to over-fitting. As a result you'll always have some of each kind of each error.

So it's totally possible, in fact likely, that each poll could have correlated bias. It's implausible they'll have correlated variance, but bias is extremely likely. If there is in fact problem with getting Dems on the phone, etc. and every poll is based on randomly phoning people, then yes, you would expect all those phone polls to make correlated errors.
The more parsimonious and scientific inference is that the “(R) bias” Nate Silver has found is, of course, nothing other than an estimate of the (D) cheating advantage. What else could it be, after all? Yes, it could theoretically be something else – but that would require an explanation, and evidence. Surely the null hypothesis is that the “bias” showing up from these polls is just the result of voter fraud.
Can we assume that both parties cheat to some extent? There is at least one pro-GOP and one pro-Dem ballot that has been cast fraudulently. In any set of elections, some of them will receive more fraudulent R than D votes. If each party were to cheat the same amount we would see these canceling each other out; there would be no correlation between the cheating levels.

Okay. In order for Charmer's hypothesis to be the most parsimonious, we must conclude that the correlation of cheating exceeds the correlation of poll errors.

I'm totally willing to believe one party cheats more than the other. Fine. But I also have good reason to believe that the poll errors are highly correlated. What I don't have is reason to believe ballot fraud is more highly skewed Dem than is polling error. We're right back where we started: some of the polls' overestimation of GOP votes is due to bias in the polls, and some is due to fraud. Still no reason to conclude the fraction is 100% fraud and 0% polling bias.
In other words, we now have scientific, quant-friendly evidence here that the (D)s get something like a ~1.2% advantage from fucking cheating.
This is true only if you assume all the differential is due to cheating. This is the ceiling of that the Dem advantage to cheating is; the true value is lower.

In order to figure that out we would need to know not only that one party cheats more than the other, but the degree to which they do so.
Again, if you have a better explanation why all these polls would come out with a R+1.2 bias on average, you are welcome to advance your argument, along with your evidence. But fair warning, if you mumble some facile BS about ‘cell phones’ or ‘hurricane Sandy’ I’m going to fucking make just as much fun of you as I made of righties who BS’ed stuff about lefties more likely to lie about being likely voters or pollsters ‘using a 2008 turnout model even though (R) enthusiasm is really high’.
Cell phones? Sandy? I don't know what the underlying cause is, if indeed there is a single one. But you don't have to be able to solve that responsibility assignment problem in order to know if there is bias.

If every poll is making the same assumption about the distribution of the electorate, for example, assuming there is no correlation between being willing to take 5 minutes to talk to a pollster and supporting Romeny, and that assumption is wrong then they could all very easily make correlated errors because they would have the same bias. Not the same variance, but the same bias. Since all predictive systems will make both bias and variance-derived errors, if the bias terms are correlated, the error terms will be correlated. (Less strongly, due to canceling variances, but still somewhat correlated.)
I think I’ve earned the right to say this because I have been and continue to be consistently on the side of the quants: if you don’t see Nate Silver’s table as, absent other quantified and supported explanations, prima facie evidence of the size of the (D) cheating advantage, then guess what? You’re not on the side of the quants, and you must hate math.
I think I've also earned the right to say this, because I have also consistently been on the same side of this issue. So... ummm... there. Take that?

PS If every poll's errors were uncorrelated, why would we see this?

If Charmer's story is correct, and polls are wrong only because of cheating, then it shouldn't matter what kind of poll was conducted. They'd all accurately reflect how people intended to vote, people would go out and vote that way, then the results would be skewed by fraud. All the poll types would show the same GOP overestimation/Dem cheating advantage.

This seems to be pretty clear evidence that the polling process itself introduces correlated errors.

PPS Internet polls most accurate? Interesting. Contradicts conventional wisdom. Would like to know more.

I'm editing the above to reflect a point Charmer (aka RWCG, which I'll call him (?) from now on) made in the comments.

Internet polls more closely reflect the observed outcome? (* Where the observations include effects of fraud.) Interesting. I would like to know more about their process.

08 November 2012

Micromorts & Microlives

Schneier on Security | Bruce Schneier | Micromorts

Here's a great concept: a micromort:
Shopping for coffee you would not ask for 0.00025 tons (unless you were naturally irritating), you would ask for 250 grams. In the same way, talking about a 1/125,000 or 0.000008 risk of death associated with a hang-gliding flight is rather awkward. With that in mind. Howard coined the term "microprobability" (μp) to refer to an event with a chance of 1 in 1 million and a 1 in 1 million chance of death he calls a "micromort" (μmt). We can now describe the risk of hang-gliding as 8 micromorts and you would have to drive around 3,000km in a car before accumulating a risk of 8 mt, which helps compare these two remote risks.
There's a related term, microlife, for things that reduce your lifespan. A microlife is 30 minutes off your life expectancy. So smoking two cigarettes has a cost of one microlife.
(1) Yes, this is a neat idea. The analogy to using grams (or pounds) rather than tons makes a lot of sense.

(2) People are tremendously bad at reasoning numerically about fractions. For instance, people tend to see the difference between 1/50,000 and 1/100,000 as the same as that between 1/200,000 and 1/250,000. This may help alleviate that problem since it holds the denominator constant and presents the numerator to people to reason about.

(3) Perhaps we need a unit like the horsepower: some sort of real world baseline for risk to which we can refer. Maybe everything can be compared to lightening strikes, so the chance of dying in a helicopter crash might be 0.25 L-strikes if four times as many deaths are caused by lightening as helicopter crashes. (I'm totally guessing about those values.)

My first thought was deaths-for-falling-down-stairs, but that is not nearly constant over time or across cultures. Plus I can't think of an easy abbreviation like "L-strike".

(4) I think micromorts and microlives may give more ammunition to scaremongers. It's already easy enough to scare people about tiny risks. I sure don't want to make it any easier. We would have to do some surveys to find out, but I suspect 8.6 micromorts sounds scarier than a 1 in 116,000 chance of dying.

(5) You can gather pretty good statistics for the chance of dying from hang gliding accidents, and get a pretty precise measure of the micromorts involved. Determining microlives involves much more sophisticated actuarial calculations, and I'm pretty certain it will end up being much less accurate. I would be very skeptical of someone who could tell me with any precision at all how many microlives eating a sunny side up egg or medium rare steak will cost me.

I'm all for doing things scientifically and quantitatively, but I am wary of pseduo-science and false precision. "All large calculations are wrong," as a wise man once said. I do not want to give technocrats and professors hungry for grants any more reason to go in front of Congressional hearings to talk about how they've calculated that allowing Americans to smoke weed/drink wine/drive Toyotas/use Sidecar for ride-sharing/eat imported haggis/etc is costing them 59.463271 microlives each.

Yeah, this is a bit immaterial. Micromorts/microlives are still a good idea. I'm just pointing out that they may be used for Evil as well as Good.