Tuesday, March 22, 2011

Doesn't Alton Brown have a hand puppet for this?

Andrew Gelman feels he may have the solution to the mystery of why Nathan Myhrvold (billionaire, physicist and former Microsoft CTO) became so fixated on solar cells acerbating global warming that he convinced the authors of Superfreakonomics to include an almost immediately discredited section on the subject.
Aha! Now, I'm just guessing here, but my conjecture is that after studying this albedo effect in the kitchen, Myhrvold was primed to see it everywhere. Of course, maybe it went the other way: he was thinking about solar panels first and then applied his ideas to the kitchen. But, given that the experts seem to think the albedo effect is a red herring (so to speak) regarding solar panels, I wouldn't be surprised if Myhrvold just started talking about reflectivity because it was on his mind from the cooking project. My own research ideas often leak from one project to another, so I wouldn't be surprised if this happens to others too.
Gelman was referring to Myhrvold's writings on modernist cuisine (or what the slightly less trendy call molecular gastronomy) and specifically to this passage, "As browning reactions begin, the darkening surface rapidly soaks up more and more of the heat rays. The increase in temperature accelerates dramatically."

This may explain why Myhrvold had albedo on his mind, but the comments to Gelman's post suggest another mystery: does the change in color actually have a dramatic effect the rate of browning or is the rate primarily driven by other changes such as water boiling away from the surface of the food*?

Is it possible that Myhrvold is, at heart, basically a freakonomist? Someone who, though brilliant and accomplished, is so eager to find examples of important principles that he sees them where they don't apply?

The following clip has nothing to do with anything in this post, but it does feature an exploding turkey which is really cool.




* From Wikipedia:
High temperature, intermediate moisture levels, and alkaline conditions all promote the Maillard reaction. In cooking, low moisture levels are necessary mainly because water boils into steam at 212 °F (100 °C), whereas the Maillard reaction happens noticeably around 310 °F (154 °C): significant browning of food does not occur until all surface water is vaporized.

Time to call a lie a lie

One of the most effective rhetorical tools in the education reform movement is the "we're just in this for the children" chant. The implication, of course, is that the people who disagree with the movement's proposals must not be putting the children first. It is an obviously unfair suggestion but it done a spectacular job quelling potential criticism on the left.

Of course, the vast majority of people on both sides of the debate are there because of a concern for kids. I disagree strongly with Jonathan Chait and Ray Fisman (just to name two) but I have no doubt that both men are motivated by a desire to see young people get a better education. (For the record, that's a courtesy that many of those in the movement, such as Chait, have been reluctant to extend to the other side.)

This concern does not, of course, preclude self-interest. As functional adults we expect people to act out of a mixture of motives. When policemen lobby for more cops on the street or our dentists advise us to schedule more appointments, we know that their advice to us is also in their self-interest but, barring evidence to the contrary, we believe that they are genuinely concerned about us as well.

These two facts, that everybody has mixed interests and that their advice should still be given the benefit of the doubt, need be kept in mind during all debates. Acknowledging these facts goes a long way toward keeping things civil and, more importantly, honest. That's why, in the context of recent events, the behavior of Michelle Rhee has been so difficult to forgive.

Rhee has always played an aggressive game and has gone out of her way to portray her opponents in a negative light, but with the formation of her lobbying group StudentsFirst, Rhee has crossed the line into claiming that only she and her allies have pure motives.

Consider this quote from an interview conducted by the painfully credulous Guy Raz on Weekend All Things Considered:
"Over the last 30 years, the education policy has been driven in this country by lots of special interest groups, including the teachers union," she says. "I think that one of the missing pieces is that there is no organized national interest group that has the heft that the unions and the other groups do who are advocating on behalf of children."
The trouble with proclaiming your own purity is that someone will remember those proclamations when you have to make compromises. Rhee's recent role as an adviser/advocate for various conservative Republican governors has made some of these compromises unavoidable.

The recent debate over Florida's education bill provided an ideal example:
Among the amendments proposed and rejected as poison pills:
Requiring superintendents to offer a written explanation for denying a teacher's contract renewal, if test scores and evaluations make the teacher eligible for the renewal.
Let's take a minute and unpack this. First let's keep in mind that Rhee's philosophy is based on the assumptions that you can largely fix the current problems in education by putting better teachers in the classroom and you can accurately identify those teachers through test scores and evaluations. The teachers being denied these letters are, by definition, the same teachers Rhee says we need to keep in the classroom in order to save our school system.

Unfortunately, the main effect of denying that letter will be to force many of these teachers out of the profession forever. As I explained it before, the problem is asymmetry of information. It is incredibly difficult and disruptive to make staffing changes during the school year. This makes administrators very skittish about hiring a teacher who has been fired elsewhere. The administrators would, however, probably take a chance if they knew that the teacher got good evaluations and produced high test scores but was fired for something like budgetary reasons. In other words, that letter might have determined whether or not the effective teacher remained employable.

The effect here is two-fold: effective teachers who find themselves caught in this trap will have a great deal of trouble finding another job and may have to leave the profession; other effective teachers will see that competence and accomplishment cannot protect them from arbitrary career-ending decisions and will consider leaving the field as well. Either way, the law Rhee endorses causes us to lose more of the teachers Rhee says we need to keep.

To be blunt as a sock full of sand, from the students' standpoint this is all bad. There is no possible benefit. You simply cannot argue that causing effective teachers to leave the classroom is good for kids. Despite Michelle Rhee's titular claim, rejecting that amendment puts students a poor second.

This doesn't Rhee and the Florida GOP don't care about the quality of teachers (I'm sure they do), but it does mean that, in this case, other things mattered more. Things such as the money to be saved by firing teachers who are likely to max out the merit pay system and the power that comes from running the education department like a political machine.

If Michelle Rhee were concerned solely with the interests of children, she would have been actively lobbying for rules like the one in the amendment, rules that furthered her stated goal of having more teachers in the classroom whom she considered competent. But, of course, Rhee has to balance the interests of children against the interests of those she represents, an alliance that includes, among others, educational entrepreneurs who stand to make a great deal of money from proposed reforms and conservative Republicans who see the current conflict as a way of maintaining political power and moving back to a period when the country was on the right track.

I have no doubt that Michelle Rhee's concern for children is genuine. Rhee is a professional educator and it is exceptionally rare to find someone who has spent a career working in schools who doesn't care about kids. Nor does the fact that she has sometimes put other interests above those of students (including a particularly notorious case involving her own children) indicate a lack of concern -- making compromises is a necessary part of being an adult.

The sin here is in the lie, in claiming purity of motive and suggesting that only she was trustworthy. That was unfair to her opponents, provably false and terribly damaging to the discourse. Michelle Rhee should be ashamed of herself for saying it and Guy Raz and the rest of the press corps should be ashamed of themselves for not holding her accountable.

Monday, March 21, 2011

"How to Erase $70,000 in Debt"

First, get an income of $140,000...

You know, when Steve Martin and South Park did this sort of thing, they meant it to be funny.

Safety of Energy Sources

Here is an interesting discussion of deaths per terra watt-hour of electricity generated. I have always been surprised that coal plants generate more radioactivity in the surrounding community than nuclear plants do.

I won't vouch for the numbers being perfect, nor do I think that these rates include the risk of massive failures (like a nuclear plant blowing up). On the other hand, the otherwise fairly safe Hydro has a disaster with 170,000 deaths included in it.

This comparison does seem to put the relative level of risk in perspective. Certainly, there is no risk free alternative to power generation and we should really be focused on what non-fossil fuel options we might have.

Sunday, March 20, 2011

20 each?

Cory Doctorow's post (via DeLong) on the New York Times' new paywall has me thinking:

New York Times paywall: wishful thinking or just crazy? - Boing Boing: lots of people are going to greet the NYT paywall with eye-rolling and frustration: You stupid piece of technology, what do you mean I've seen 20 stories this month? This is exactly the wrong frame of mind to be in when confronted with a signup page (the correct frame of mind to be in on that page is, Huh, wow, I got tons of value from the Times this month. Of course I'm going to sign up!)

Which means that lots of people will take countermeasures to beat the #nytpaywall. The easiest of these, of course, will be to turn off cookies so that the Times's site has no way to know how many pages you've seen this month

Of course, the NYT might respond by planting secret permacookies, using Flash cookies, browser detection, third-party beacons, or secret ex-Soviet vat-grown remote-sensing psychics. At the very minimum, the FTC will probably be unamused to learn that the Grey Lady is actively exploiting browser vulnerabilities (or, as the federal Computer Fraud and Abuse statute puts it, "exceeding authorized access" on a remote system -- which carries a 20 year prison sentence, incidentally)

I'm running a dual boot operating system (Windows and Linux). Barring Doctorow's recently decanted telepaths, I assume that puts me up to forty a month which is about four times what I expect my monthly demand to be.

A nice post on Education Reform

There is a nice post in the Daily Kos talking about education reform. The whole piece is worth noting but this point seems especially apt:

Those advocating the end of seniority-based retention practices in favor of "performance" based on student test scores have to concede that districts, which must stretch dollars these days like never before, will be tempted to staff their classes in such a way to protect their younger (and, it must be noted, markedly cheaper) staff members.

I will never forget in my third year on the job drawing a Freshman Geography class that felt, on bad days, like a training session for America's Most Wanted. When I half-jokingly teased a counselor about how I managed to draw every wild-eyed boy in the freshman class, she smiled and told me, "But, Steve, we all know how good you are with difficult students."

At the time, I took it for the backhanded compliment that it was. In this brave new world being promoted by the GOP (and an alarming number of Democrats), it would be my ticket to lower pay. Worse yet, it could be my ticket out of the profession.


The worry here is that, in the short term, this approach will save a lot of money. Having a lot of inexpensive and enthusiastic junior teachers will do wonders for budgets (at a time when tax cuts are a priority). While teachers will recognize what is happening, in an environment with unemployment hovering around 10% (and basic things like Health Insurance depending on employment) it is likely that schools will not suffer in the short term. In the long term, the new world of teaching will require much higher pay for equally qualified teachers as we know have to compensate the teachers for the fear and uncertainty in such a system.

Not to mention to concern that class assignments could be used to protect liked but less capable teachers. Do we have a solid plan for preventing this from happening?

Sunday Morning Funnies

Saturday, March 19, 2011

Weekend Gaming -- the heated chess/checkers debate

Andrew Gelman joined in the Chess/Checkers debate and added Go to the mix (which was outside of the scope of the original post but is certainly relevant to the discussion). There have been around two dozen comments so far (and Gelman always has unusually thoughtful discussion threads). Definitely worth a look.

Tenure and the end of mandatory retirement (Canada edition)

Frances Woolley has a nice post about the intersection between tenure and the lack of a mandatory retirement age in Canada. It is a different case than the debate in Canada. The background is that salary scales and tenure agreements (at Canadian Universities) were negotiated when there was a mandatory retirement age of 65. The removal of mandatory retirement was a windfall for professors who were already employed as they work under a salary scale designed for workers who would leave at 65.

In practical terms this can have a fairly important impact on budgets as it adds additional years of salary at the highest levels (often 2.5 x starting salary in a Canadian University). Frances has a nice chart here. The short term implications are stark:

Such a pay structure can be profitable as long as the pay structure is similar to the one shown in the diagram above, where the high costs of paying workers between 45 and 65 are counter-balanced by the low cost of paying workers between 25 and 45. But if the terms of the employment arrangement were changed so that workers stayed on until 75, the firm's pay structure would no longer be profitable: the costs of paying experienced workers more would exceed the gains from underpaying junior workers.


I think that there is an important balance between job security and balancing out employment contracts. In this case, due to regulatory changes, I think it would make a lot more sense if tenure elapsed at the traditional retirement age. In this case we have the reverse of what is happening in the United States for teachers -- the employment contract changed in mid-stream. I think it is consistent to argue, in both cases, that a change of contract terms should not result in a windfall for either party unless the change was by mutual consent.

Friday, March 18, 2011

Not all that broken up about the paywall

[I'm working an a post on the Florida education bill but I'm taking my time. The truth here is so ugly that even the slightest over-reaction would be going too far. In the meantime, here's something light and snarky for your Friday afternoon.]

As you've probably heard, there's a paywall going up around America's most over-rated newspaper (I'd put the Wall Street Journal, the LA Times and maybe a half-dozen other papers above it). The limit for free articles is twenty a month though you can still apparently follow blog links after you've run through those so you should still be able to keep up with most of what you're reading now (almost all of which is probably summarized by bloggers like Thoma and DeLong anyway).

As far as I can tell, the big loss will be those articles that catch your eye while you're browsing the site and most of those tend to read like this piece on the spectacular failure of Mars Needs Moms (a bomb that may leave a nine-figure crater).

The explain-the-fiasco story is one of the annoying perennials of entertainment journalism (the object is to explain why a show tanked without addressing the fact that it stank) and even by the low, low standards of the genre, this article by Brooks Barnes leaves much to be desired, consisting of widely-available facts, conventional wisdom and analysis like this:
It is quite rare for a Disney release to flop as badly as “Mars Needs Moms,” which is based on an illustrated book by Berkeley Breathed, best known for the comic strip “Bloom County.” Part of the problem may have been the story. What child wants to see a movie about his mom being taken away from him? But studio executives also pointed to the style of animation as a culprit.
Do the names Bambi and Dumbo not ring any bells whatsoever? Does Barnes not know that early Disney features (Snow White, Pinocchio, Dumbo, Bambi) expertly played expertly on children's fear of being separated from their parents? Or that this template remains popular to this day (Finding Nemo)? More importantly, does this strike you as an insight you'd pay $15 a month for?

If you're in the mood for more fun at the gray lady's expense, check out this amusing bit of mockery from Wonkette.

Thursday, March 17, 2011

I'm heading out the door but before I go...

Make sure to check out this post by Andrew Gelman. It's relevant to a number of big but underdiscussed issues in the education reform debate.

New Education Bill

Okay, there was a new law passed today in Florida:

School teachers would lose tenure and see future pay raises based on student performance under a politically charged package of education changes the Florida House sent to Gov. Rick Scott Wednesday on a straight party-line vote.


The new bill:

The legislation will establish a statewide teacher evaluation and merit pay system in 2014 and do away with tenure for new teachers hired after July 1 this year. It also chips away at teachers' due process and collective bargaining rights.



Among the amendments proposed and rejected as poison pills:

Requiring superintendents to offer a written explanation for denying a teacher's contract renewal, if test scores and evaluations make the teacher eligible for the renewal.


In the new system contracts need to be renewed annually. I am unclear how not offering an explanation for failure to renew (for teachers that test well) is an unreasonable requirement. After all, if we trust these test-based metrics than it should be perfectly reasonable to explain why a high performing teacher is being let go (e.g. drop in student enrollment at their school). If we do not trust these metrics to give an unbiased picture of how is an effective teacher then why are we tying pay so closely to these metrics?

It just seems to be inconsistent.

It is also unclear where any money for merit raises might come from in a state focused on tax cuts. Overall, I am underwhelmed.

Mark: Any thoughts?

Wednesday, March 16, 2011

Cloning Jaime Escalante -- a thought experiment

One of the fundamental tenets of the reform movement is the belief that we could fix all of our schools problems if, through big bonuses and wholesale firings, we could replace all of the lazy and incompetent teachers with great ones. Counterarguments generally point out that our metrics for identifying good teachers were unreliable and that, even with high-quality metrics, trying to restaff a major industry so that, say, 80% of the new recruits were in what had been the top 10% is simply not practical.

But what if we moved past those arguments altogether? What if we could create perfect duplicates of any teacher we want and place them in a million classrooms? Surely that would do it, but who should we pick?

How about Jaime Escalante, the teacher immortalized in the movie Stand and Deliver? Escalante was beyond question a spectacular teacher and he managed to build one of the country's most successful math programs in a very troubled urban school, Garfield High. By the end of his time at the school, he was teaching huge calculus sections (for HS) and producing better than a 90% pass rate on the AP exam. Only four schools in the country had more students passing the test.

So what would happen if you could clone "the best teacher in America" (as reporter Jay Matthews called him) and have him teach your AP calculus class? We can never be sure but I suspect that it would go something like this:
In 1991, he packed up his bag of tricks and quit Garfield, saying he was fed up with faculty politics and petty jealousies.

He headed to Hiram Johnson High with the intention of testing his methods in a new environment.

But in seven years there, he never had more than about 14 calculus students a year and a 75% pass rate, a record he blamed on administrative turnover and cultural differences.
Jaime Escalante was a great teacher, but to achieve those amazing results at Garfield he had to be in the right place at the right time. He needed a compatible and supportive administrator and, more importantly, a unique and powerful bond with the student body and the community. Compatibility and rapport are difficult to measure and next to impossible to predict but they are often the difference between adequate and astounding results.

Krugman joins the nuclear debate


I'm not sure this is the best time to be having this debate (a once in a millennium disaster tends to interfere with the ability to accurately evaluate risks), but Paul Krugman has a good post on the subject:

As Nordhaus’s RA, I spent the summer of 1973 on this project: my days were spent in the geology library, reading Bureau of Mines circulars on the engineering and costs of alternative energy sources, my nights at the computer center drinking vending machine coffee. (These were still the days of big mainframes and punchcards; you handed a deck of cards to the high priests behind the glass wall, then an hour later you got back a huge stack of hexadecimal garbage because you made an error on one of your cards.)

In short, I was in heaven.

Nordhaus’s paper was wonderful. (Sorry, for technical reasons I can’t put up a full version from my current undisclosed location.) But as it turned out, it was much too optimistic. Not his fault or mine: it was those Bureau of Mines circulars.

What was wrong with those circulars? They were much too optimistic about the costs of alternative energy sources, especially alternatives to oil. Basically, the engineers were understating the difficulties involved. Later Marty Weitzman would formulate a law on this: the cost of alternatives to crude oil is 40% above the current price — whatever the current price is.

And hence my skeptical reaction to the new study about the costs of running an all-renewable economy.

To be fair, we probably have much more solid ideas about the cost of wind and solar power than we did about shale oil and coal liquefaction back in 1973: wind is already a widely used technology, and concentrated solar power — probably the main way we’ll use the sun — is pretty well understood too. But there will be surprises, not all of them positive.

None of this is meant to disparage the work, or the need to use much more renewables than we are using now.

More Glaeserian causality

We all occasionally make too much of anecdotes and jump too quickly from correlation to causality, but with Edward Glaeser, this sort of thing is starting to become a habit.

From the New York Times:
Vast public infrastructure projects, like high-speed rail, helped create Spain’s current fiscal morass and did little to revitalize Japan during its lost decade.
Of course, given the magnitude of the demographic and economic forces acting on Japan, it's difficult to say exactly what effect high-speed rail had.

As for Spain, do we really have reason to believe massive spending on public works helped cause the crisis? Here's Paul Krugman's answer:



On the eve of the crisis, Spain was running a budget surplus; its debts, as you can see in the figure above, were low relative to GDP.

So what happened? Spain is an object lesson in the problems of having monetary union without fiscal and labor market integration. First, there was a huge boom in Spain, largely driven by a housing bubble — and financed by capital outflows from Germany. This boom pulled up Spanish wages. Then the bubble burst, leaving Spanish labor overpriced relative to Germany and France, and precipitating a surge in unemployment. It also led to large Spanish budget deficits, mainly because of collapsing revenue but also due to efforts to limit the rise in unemployment.

Wouldn't Glaeser's argument imply that Spain was spending too much and wouldn't that, in turn, show up in the debt to GDP numbers?