Friday, September 30, 2011

More population pyramids

From the UN. I was especially interested in the continent-wide/regional aggregates, which illustrate what is summarized on page here (PDF): between now and 2050, the developed world's total dependency ratio will increase from 0.5 to 0.7 while the developing world's total dependency ratio will decrease from 0.6 to 0.5. So, for the purposes of my retirement savings, investing in "emerging markets" and investing in US corporations that do business in those countries are the obvious (but risky) ways to attempt to "escape demographics".

From another angle, if US equities really trended up for decades because so many boomers saved and are now trending down because the boomers are starting to retire, then we could see a global version of this cycle: stocks around the world booming in 2050 as billions of people in Asia and Latin America, who will have much higher living standards than their parents have today, save for retirement. However, I have no idea what extent savers in Asia and Latin America will globally diversify their portfolios.

Decades after that, there would be another bust, provided you trust the UN's projections out that far, and don't expect peak oil, nuclear annihilation, climate change, the singularity, or whatever to economically swamp the effects of demography. Addendum: The demographics behind this hypothetical post-2050 bust will be "locked in" before 2050. The lock-in process starts next decade, as Phillip Longman points out:

Indeed, the U.N. projects that by 2025, the population of children under 5, already in steep decline in most developed countries, will be falling globally -- and that's even after assuming a substantial rebound in birth rates in the developing world.
Actually, annual world births already peaked in the late 1990s. The under-5 population is still growing mostly because of declining child mortality.
A map of books of speculative fiction.

Thursday, September 29, 2011

Bad patents: an aggregate and an anecdote.

Wednesday, September 28, 2011


Is the author of this article is arguing for more laid-back parenting, or is he sneakily arguing for parents to be pickier about their teens' friends?
In the game, you try to drive across town in as little time as possible... You save time—and score more points—if you drive through before the light turns red. But if you try to drive through the red and don't beat it, you lose even more time than you would have if you had stopped for it... When teens drive the course alone, in what Steinberg calls the emotionally "cool" situation of an empty room, they take risks at about the same rates that adults do. Add stakes that the teen cares about, however, and the situation changes. In this case Steinberg added friends: When he brought a teen's friends into the room to watch, the teen would take twice as many risks, trying to gun it through lights he'd stopped for before. The adults, meanwhile, drove no differently with a friend watching.


Would the world be a better or worse place if more people paid close attention (consciously, not just subconsciously) to the difference between real and fake smiles?

FDR's best economic policy...

...was going off the gold standard. Many of his other economic policies were counterproductive.

In more normal times, great economies make great presidents, not the other way around. But in our time, Obama, thinking monetary policy ran out of steam in 2008, has not even tried recess appointments to the two vacancies on the seven-member Federal Reserve Board of Governors. If Obama is punished for the economy, he mostly deserves it. Obama managed some fiscal stimulus, but so did Hoover:

Hoover's budget strategy over his term of office was not to balance the budget. The budget ran a small deficit of -.06% of GDP in 1931, followed by a much larger deficits of 4.0% of GDP in 1932 and 4.5% of GDP in fiscal year 1933... During the Great Depression, Franklin Roosevelt faced unemployment rates of 25% and continued the Hoover policy of budget deficits, running deficits no larger than 5.9% of GDP and more usually in the range of 3-4% of GDP through the 1930s. During the Great Recession, the U.S. economy experienced unemployment of nearly 10%, and has responded with fiscal stimulus on the order of 10% of GDP.
Fiscal stimulus fails if money stays tight.

Monday, September 26, 2011


I love this one and I expect to see more like in the future. However, the bigger problem is irreproducability in general, not fraud. There are structural causes for this problem, such as the statistical significance filter.

Mathematics is blessed from this point of view. My results can be reproduced every time someone reads them and understands the proofs. Only for a handful of famous problems are false proofs a major nuisance; even for these, the errors are found quickly.

What are nuclear bombs good for?

Why did Japan surrender? Very interesting---the scholarly claim reported is that Japan surrendered only because of the Soviet Union's entry into the war deprecated their strategy to try "to convince the Soviet Union, still neutral in the Asian theater, to mediate a settlement with the Americans. Stalin, they calculated, might negotiate more favorable terms in exchange for territory in Asia." Moreover:
The bomb - horrific as it was - was not as special as Americans have always imagined... The photos of charred Tokyo and charred Hiroshima are indistinguishable. In fact, more than 60 of Japan’s cities had been substantially destroyed by the time of the Hiroshima attack.

However, the journalist speculates a bit too carelessly at the end:

Hasegawa’s scholarship disturbs this simple logic. If the atomic bomb alone could not compel the Japanese to submit, then perhaps the nuclear deterrent is not as strong as it seems. In fact, Wilson argues, history suggests that leveling population centers, by whatever method, does not force surrender... If killing large numbers of civilians does not have a military impact, then what, Wilson asks, is the purpose of keeping nuclear weapons? We know they are dangerous. If they turn out not to be strategically effective, then nuclear weapons are not trump cards, but time bombs beneath our feet.
Sometime between 1945 and 1961, the stakes increased from the destruction of a few major cities to the destruction of all major cities. Mad or not, MAD worked: the Soviets withdrew their missiles from Cuba and Kennedy secretly agreed to withdraw US missiles from Turkey. Moreover, in a counterfactual 1945 in which Japan didn't surrender and we kept dropping nukes, the eventual result would have been a sparsely populated, poisoned Japan. At that point, surrender would be irrelevant.

Demographics and retirement

You can't escape demographics. Except, that post ignores the distinction between global and national demographics. Globally diversified retirement savings might buy developed nations more time. For example, compare projected US and World population pyramids.

Wednesday, September 21, 2011

Electoral votes by congressional district?

Which party would benefit on net if this idea was implemented in most states?

The hypothetical is that most states switch to a new scheme of awarding electoral votes for the president where 2 EVs go to the statewide winner & 1 EV is awarded to the winner of each congressional district in that state. It seems to me that this would help the GOP on average. Increasing voter turn-out is easier in urban districts than in rural ones, and on net this helps the Democrats in contests for state-wide majorities/pluralities. This new EV scheme would make state-wide vote totals much less important.

Nate Silver suggests that public disgust with gerrymandering influencing presidential elections leads to the abolishment of the electoral college in favor of an direct national vote for the president (e.g., national plurality, national majority with possible run-off election, or some other scheme that treats votes from different states equally). Unlikely! It only takes 13 states to prevent a constitutional amendment, but there are 33 states with below-average adult population. (I computed 33 from here.)

Also, states don't have to gerrymander forever. Several states already have independent redistricting commissions. EV-by-district could push more states towards this solution. Or maybe some big states would just keep gerrymandering, leading to a constitutional amendment that somehow restores the de facto status quo of awarding most EVs to the winners of state-wide pluralities. (The de jure status quo is that states can award choose their electors however and whenever they want, and electors vote however they want on the Monday after the second Wednesday in December.) Both of these seem more likely than the less populous states giving up their disproportionate power in selecting the president.

Let us also consider the robustness of election results. For example, Minnesota is typical in requiring a recount of a state-wide vote if the margin is less than half a percent. 2 out of the last 47 presidential elections had a national popular vote margin less than half a percent (1880, 1960). (1 out of 47 had a margin less than a quarter of a percent (1880); 6 out of 47 had a margin of less than one percent (1880, 1884, 1888, 1960, 1968, 2000).) If those had been direct elections, would expensive national recounts have been required? You might invoke the law of large numbers to argue for a lower threshold than 0.5% at the national level, but the law of large numbers is about random, uncorrelated errors. What recounts are really about is the fear of election fraud. Think about how the incentives of election officials in deep red Idaho and deep blue Massachusetts would change if presidential elections became direct.

In actuality, 2 of 56 presidential elections were disputed, but, thanks to the electoral college, the disputes were limited to a few states (FL in 2000; FL, LA, SC, and OR in 1876). If EVs were awarded by congressional district, disputed presidential elections would still be rare, and within this small set of disputed elections, most disputes would not require any state-wide recounts.

P.S. I dismiss purely moral arguments for "one man, one vote" because government should be a mere means to ends like life, liberty, and the pursuit of happiness. Having my vote and your vote weighted equally either promotes these ends or does not promote them.

Tuesday, September 20, 2011

The case for QE3

A concise explanation with lots graphs (PDF slide show; hat tip: MR).

Wednesday, September 14, 2011

Health advice for us at desks

No special furniture required. Just sitting is bad, but so is just standing. Therefore:
Sit to do computer work. Sit using a height-adjustable, downward titling keyboard tray for the best work posture, then every 20 [or 30] minutes stand for 2 minutes AND MOVE... just walking around is sufficient.
So, maybe pace back and forth in the office and write on the whiteboard. The trick is the frequency. I don't want a beep every 30 minutes, but perhaps a discreet blinking light in the corner of the screen. I'll have to experiment with this.

My biggest concern is loss of "flow." You know the movie trope, a montage of a hacker typing for hours? I do more scribbling on paper and whiteboards (and forego a soundtrack), but it's basically the same thing. I really enjoy having a big block of time to work on Just One Thing. (The great danger is the temptation to "work" for a long time on reading my favorite economics blogs or perfecting a blog post of my own.)

Unfortunately, it is not clear to me how important "flows" of an hour or longer are to my productivity, nor even how I could perform a self-experiment to find out. (Is there applicable large-N psychology research?) My subjective experience, which I should not naively trust, is that some problems are intricate and that I need a big block of time just to get all the moving pieces properly configured in my working memory. On the other hand, many flows have turned out to be inefficient ex post because of a wrong turn made early in the flow. One the third hand, it is not clear that breaks help detect wrong turns sooner. For an extreme example, it seems easy to dabble briefly in something every day and not notice a lurking error for many, many days.

Tuesday, September 13, 2011

"Imagine a host talking to a panel of four pundits about someone about to roll a 6-sided die..." Scroll down and read the rest.

Monday, September 12, 2011

T/C and T/(C+T)

...are better notions of a person's average tax rate than T/I or T/(I+T). Here T is tax, C is consumption, and I is income. Scott Sumner estimates that Warren Buffett has a 90% tax burden in the sense of T/(C+T). In the sense of T/C, it would be about 1000%. Sumner makes the case for C instead of I succinctly and eloquently at the end of the post:
  1. Buffett’s consumption is the resources he takes out of the economy for his own personal enjoyment.
  2. Taxes paid are what he contributes to the common good.

Sunday, September 11, 2011

Faith is revealed belief...

... and such locutions reveals my assimilation of economic jargon.

Saturday, September 10, 2011

Inputs to teacher productivity

Some strong negative results:
As with most previous research, we found no relationship between a teacher’s earning a master’s degree, certification, or years of experience and the teacher’s classroom performance as measured by student test scores. Though we found that some pedagogy course work was related to teaching effectiveness, the magnitude of the effect was mild: even very detailed information about the teacher’s preparation in college told us very little about how effective that teacher would be in the classroom.
If you want to teach, say, geometry to high-school students, it surely helps to know geometry. At my university, you can take a course with this description:
Selected topics from the foundations of Euclidean and non-Euclidean geometries. Includes the study of spherical and hyperbolic geometries, as well as transformational geometry, with techniques from linear algebra. Intended primarily for students seeking secondary certification.
After that course, your understanding of geometry will go much deeper than it did in high school. However much or however little your future geometry students will benefit from your deeper understanding, my point is that this is an undergraduate course. From another undergraduate course description:
Selected topics from secondary school mathematics. Content, materials, and contemporary issues specific to teaching of mathematics at the secondary school level.
That course has a corequisite; from its description:
This course will focus on field-based supervision of elementary and secondary education pre-service teachers.
After taking such courses, why would I be surprised to learn that you won't help your students by getting a master's degree? There are diminishing returns to education. For an example from higher education, graduate students who just got their bachelor's degrees are assigned to teach lower-level undergraduate courses. As a math PhD student, I taught a trigonometry class and was a TA for calculus classes and for a linear-algebra-plus-differential-equations class. There were no graduate course prerequisites for these teaching assignments. My graduate courses were intended to make me a better scholar, not a better educator.

More surprising is the failure to detect a student test score benefit from more experienced teachers. There are diminishing returns to experience, but is 5 years of experience really no better than 0 years? If you follow the link, they note that most other studies indicate that "the benefit of that experience appears to plateau after the third to fifth year," which is more plausible than no effect at all.

As for the null result for certification, one plausible explanation is that teachers without certification were hired or kept on because they had otherwise better resumes or job performance than their certified competition. However, we must again ask for evidence.

Tuesday, September 06, 2011

Saturday, September 03, 2011

Time, heat, and origins

Ten things everyone should know about time. I really like the list overall, but #9 would better titled "aging can be reversed, in principle" (as opposed to what will happen in practice, which is very speculative).

I feel compelled to give a more substantive clarification of #8:

8. Complexity comes and goes. Other than creationists, most people have no trouble appreciating the difference between “orderly” (low entropy) and “complex.” Entropy increases, but complexity is ephemeral; it increases and decreases in complex ways, unsurprisingly enough. Part of the “job” of complex structures is to increase entropy, e.g. in the origin of life. But we’re far from having a complete understanding of this crucial phenomenon. (Talks by Mike Russell, Richard Lenski, Raissa D’Souza.)
Yes, there are complex very-low-entropy configurations (human bodies) and there are simple very-low-entropy configurations (perfect crystals). However, complexity and entropy are not completely orthogonal. The very-high-entropy configurations of the gas molecules in my apartment is better described as random than complex. What complexity it has, such as macroscopic airflow patterns ornamented with turbulence, is due to the configuration not having as a high an entropy as it could. If my apartment was a closed system, without electricity coming in to power the fans and the air conditioner, and without food coming in to keep the residents breathing and moving, then all these patterns would decay, leaving nothing but random microscopic motion. On a grander scale, the heat death of the universe really would be the death of us all.

Evolution is consistent with the second law of thermodynamics because living organisms are not closed systems; they use energy to pump out their entropy into their surroundings. The author, in #9, makes exactly this point using the example of refrigerators, which take in energy and use it to pump entropy from the food inside them into the air outside them. (If evolution wasn't consistent with the second law, then, given the quality of the evidence for evolution, the better inference might be that the second law is wrong, not that the theory of evolution is wrong.)

In the context of the origin of life, "complexity comes" is actually a very bold scientific hypothesis. The second law is consistent with the earth transitioning from lifeless to full of life, but so far all estimates of the probability of such a transition are highly speculative. Even if we knew the probability, I don't think it would settle the question of our ultimate origin. If life is probable, a theist would say that God made the universe conducive to life. If life is improbable, an athesist would say that say that we're in one of the lucky universes.

Friday, September 02, 2011

Painter envy?

Paul Halmos, 1967: Mathematicians are like painters.

Paul Graham, 2003: Hackers are like painters.

Hackers and mathematicians both wish the world understood them better. If I want someone to understand me better, I'll refer them to Halmos. If I want someone to give me money, I'll refer them to anything but Halmos.