That is the topic of my latest Bloomberg column, here is one bit:

The onset of a new year brings plenty of predictions, and so I will hazard one: Many of the biggest events of 2018 will be bound together by a common theme, namely the collision of the virtual internet with the real “flesh and blood” world. This integration is likely to steer our daily lives, our economy, and maybe even politics to an unprecedented degree.

For instance, the coming year will see a major expansion of the “internet of things”…


But whatever your prediction for the future, this integration of real and virtual worlds will either make or break bitcoin and other crypto-assets.


So far the process-oriented and Twitter-oriented foreign policies have coexisted, however uneasily. I see 2018 as the year where these two foreign policies converge in some manner. Either Trump’s tweets end up driving actual foreign policy and its concrete, “boots on the ground” realization, or the real-world policy prevails and the tweets become far less relevant.

There is much more at the link, including a discussion of cyberwar,  China and facial surveillance technologies, and the French attempt to ban smartphones at schools.

Yes, yes, I know patents are not the right measure, that is what we’ve got:

I exploit historical natural experiments to study how establishing a new college affects local invention. Throughout the nineteenth to the mid-twentieth century, many new colleges were established in the U.S. I use data on the site selection decisions for a subset of these colleges to identify “losing finalist” locations that were strongly considered to become the site of a new college but were ultimately not chosen for reasons that are as good as random assignment. The losing finalists are similar to the winning college counties along observable dimensions. Using the losing finalists as counterfactuals, I find that the establishment of a new college caused 32% more patents per year in college counties relative to the losing finalists. To determine the channels by which colleges increase patenting, I use a novel dataset of college yearbooks and individual-level census data to learn who the additional patents in college counties come from. A college’s alumni account for about 10% of the additional patents, while faculty account for less than 1%. Knowledge spillovers to individuals unaffiliated with the college also account for less than 1% of the additional patents. Migration is the most important channel by which colleges affect local invention, as controlling for county population accounts for 20-40% of the increase in patenting in college counties relative to the losing finalists. The presence of geographic spillovers suggests that colleges do cause an overall net increase in patenting, although I find no evidence that colleges are better at promoting invention than other policies that lead to similar increases in population.

That is from new research by Michael J. Andrews, via the excellent Kevin Lewis.

Or do we misallocate talent when it comes to innovation?  Here is a not so famous but very interesting paper by Murat Alp Celik:

The misallocation of talent between routine production versus innovation activities has a fi rst-order impact on the welfare and growth prospects of an economy. Surname level empirical analysis employing micro-data on patents and inventors in the U.S. between 1975-2008 combined with census data from 1930 reveals new stylized facts: (i) people with “richer” surnames have a higher probability of becoming an inventor, however (ii) people with more “educated” surnames become more proli fic inventors. Motivated by this discrepancy, a heterogeneous agents model with production and innovation sectors is developed, where individuals can become inventors even if they are of mediocre talent by excessive spending on credentialing. This is individually rational but socially inefficient. The model is calibrated to match the new stylized facts and data moments from the U.S. economy, and is then used to measure the magnitude of the misallocation of talent in innovation. A thought experiment in which the credentialing spending channel is shut down reveals that the aggregate growth rate of the economy can be increased by 10% of its value through a reduction of the misallocation. Socially optimal progressive bequest taxes that alleviate the misallocation are calculated, which serve to increase the growth rate of the economy to 2.05% while increasing social welfare by 6.20% in consumption equivalent terms.

I am not so persuaded by the idea of buying your way into innovative circles with credentials, or the analysis of the inheritance tax, but nonetheless this should stimulate thought.

So says Keith A. Meyers, job candidate from University of Arizona.  I found this to be a startling result, taken from his secondary paper:

During the Cold War the United States detonated hundreds of atomic weapons at the Nevada Test Site. Many of these nuclear tests were conducted above ground and released tremendous amounts of radioactive pollution into the environment. This paper combines a novel dataset measuring annual county level fallout patterns for the continental U.S. with vital statistics records. I find that fallout from nuclear testing led to persistent and substantial increases in overall mortality for large portions of the country. The cumulative number of excess deaths attributable to these tests is comparable to the bombings of Hiroshima and Nagasaki.

Basically he combines mortality estimates with measures of Iodine-131 concentrations in locally produced milk, “to provide a more precise estimate of human exposure to fallout than previous studies.” The most significant effects are in the Great Plains and Central Northwest of America, and “Back-of-the-envelope estimates suggest that fallout from nuclear testing contributed between 340,000 to 460,000 excess deaths from 1951 to 1973.”

His primary job market paper is on damage to agriculture from nuclear testing.

forthcoming study in the Journal of Economic Perspectives doesn’t use any of those terms and explicitly says it must not be read as an “indictment” of tenure. But it suggests that research quality and quantity decline in the decade after tenure, at least in economics.

The authors of the paper — Jonathan Brogaard, an assistant professor of finance at the University of Washington at Seattle; Joseph Engelberg, professor of finance and accounting the University of California, San Diego; and Edward Van Wesep, associate professor of finance at the University of Colorado at Boulder — started with a question: “Do academics respond to receiving tenure by being more likely to attempt ground-breaking ‘homerun’ research and in this way ‘swinging for the fences?’”

After all, they wrote, “the incentives provided by the threat of termination are perhaps the starkest incentives faced by most employees, and tenure removes those incentives.” (The question is sure to annoy academic freedom watchdogs. In the authors’ defense, they do cite the benefits of tenure, including job stability’s potential to encourage risk taking.)

Looking for answers, Brogaard, Engelberg and Van Wesep collected a list of academics who worked and were tenured in economics or finance departments at 50 top-ranked institutions at any time between 1996 and 2014. The final sample included 980 professors, all of whom were tenured by 2004.

Here is the link.

Andy was great, here is the text and audio, here is the introductory summary:

Before writing a single word of his new book Artemis, Andy Weir worked out the economics of a lunar colony. Without the economics, how could the story hew to the hard sci-fi style Weir cornered the market on with The Martian? And, more importantly, how else can Tyler find out much a Cantonese meal would run him on the moon?

In addition to these important questions of lunar economics, Andy and Tyler talk about the technophobic trend in science fiction, private space efforts, seasteading, cryptocurrencies, the value of a human life, the outdated Outer Space Treaty, stories based on rebellion vs. cooperation, Heinlein, Asimov, Weir’s favorite episode of Star Trek, and the formula for finding someone else when stranded on a lonely planet.

My favorite part was this, which Andy answered with no hesitation:

COWEN: What if there were two immortal people, let’s say it’s the two of us, placed on opposite sides of the Earth, an Earth-like planet, and we can wander freely with no constraints but just foot speed. How long does it take us to find each other?

WEIR: Can we collude in advance in any way?

COWEN: No, we cannot.


COWEN: But we know we’re trying to find each other.

WEIR: We know we’re trying to find each other. Well, we should both — but can we have a — are we both rational actors and we —

COWEN: We’re as rational as you and I are; take that as you wish.

WEIR: So, no?



WEIR: I think the best thing to do would be for both of us to pick an arbitrary great circle to walk, around the planet, and leave markings along the way denoting what direction you’re walking. So I would arbitrarily pick a direction to go and I would just go that direction with the intention of circumnavigating the entire globe, and I would walk at maybe half what is a comfortable speed for me. And you would do the same thing. Now, somewhere, our two — in fact, in two points — our great circles will intersect.

COWEN: Right.

WEIR: And when one of us reaches the other one’s, then they start following the markers at full speed, and then you get the guy. Right?

COWEN: And what’s your best guess as to how long that would take?

WEIR: Well, if you pick two points, I’m guessing one of us would have to walk probably about a quarter of the way around the planet before we found the other one’s great circle. And then you’d have to walk again. So in terms of circumnavigation times, it would take you 2x to get all the way around the planet, because my initial plan was you’d walk half-speed. So I’m guessing it would be a quarter of that, so one-half x to get to your great circle, and then a quarter x to find you along your great circle, on average, I’m guessing. So one-half plus a quarter, so .75x. So three-quarters of the time that it would take to circumnavigate the planet.

COWEN: OK, great answer.

WEIR: That’s my guess.

Do read/listen to the whole thing

That is the topic of my latest Bloomberg column, here is one excerpt:

Virtual reality technology can create vivid multiprojected environments, designed to feel real in some ways. In essence, with virtual reality we will be able to manage our empathetic and emotional reactions in a manner currently beyond us. The technology may make our medical treatments seem less painful by providing distractions. It could help alleviate post-traumatic stress disorder, by allowing sufferers relive the bad experience in a way that helps them get over it. Athletes and test-takers might use simulations to get over “choking” and other performance problems. There are plenty of other uses we probably haven’t much thought of — I was struck by a recent report of a virtual reality “death simulation machine,” to help prepare people for their passing.

In this future, we will be able to steer and manage our emotional reactions to a greater degree. Do you think you don’t care enough about starving babies around the globe? There probably will be a virtual reality program to fix that, at least temporarily. You will be able to enter their world and experience their suffering in a manner that will seem almost real, perhaps in preparation for writing a check to your favorite charity.

One key question is which emotions we will decide to have more of. It would be nice to think we will use virtual reality to make ourselves more caring and more empathetic, but I’m not convinced. Just as gossip magazines and celebrity-based reality TV have long been popular, we might use virtual reality to vicariously sample the lifestyles of the rich and famous. That could make us more callous rather than more caring, or at least less involved in the suffering of others, as competing experiences will seem so much more exciting.

There is much more at the link, interesting throughout, Smith’s TMS lurks throughout.

Understanding the purpose of sex is a fundamental unresolved problem in evolutionary biology. The difficulty is not that there are too few theories of sex, the difficulty is that there are too many and none stand out. To distinguish between theories, we ask: Why are there no triparental species with offspring composed of the genetic material of three individuals? A successful theory should confer an advantage to biparental sex over asexual reproduction without conferring an even greater advantage to triparental sex. Of two leading theories (red queen and mutational), we show that only one is successful in this sense.

That is a new Economic Journal paper by Motty PerryPhilip J. Reny, and Arthur J. Robson.  Of course the core question is a classic example of thinking at the margin.  The core conclusion is that mutations continue to rise with the number of sex-participating partners, but in simple Red Queen models the limiting features of the genotypes is the same whether there are two, three, or more partners.  The argument on pp.2739-2741 is not readily blog-summarizable, and I do not grasp it fully, but at the moment I have the following intuition.  If a parasite attack comes, the species needs only move away from the targeted genome to continue reproducing, due to some all-or-nothing assumptions about the nature of the attack.  This differs from the mutational game, where there is always some marginal (expected value) gain from moving yet further away from the initial nature of the species.  Playing a game against an identified opponent brings a better-specified and more stable and less varying response strategy than playing a game against an as-yet-unidentified opponent.  That isn’t how the authors put things, but…

Since we don’t observe much three-party reproduction (hardly any in fact), that suggests the Red Queen model is more likely to apply.

For the pointer I thank TEKL.

The great Ken Regan on AlphaZero

by on December 18, 2017 at 6:20 pm in Games, Science, Web/Tech | Permalink

A must-read for anyone who has been following this issue, Ken considers how close to God AlphaZero actually came:

We must pause to reflect on how clarifying it is that this single heuristic suffices to master complex games—games that also represent a concrete face of asymptotic complexity insofar as their size n-by-ngeneralizations are polynomial-space hard…

It may be that we can heuristically solve some NP-type problems better by infusing an adversary—to make a PSPACE-type problem that hits back—and running AlphaZero.

That sort of thing.  And don’t neglect the comments.

Policy analysts at the Centers for Disease Control and Prevention in Atlanta were told of the list of forbidden words at a meeting Thursday with senior CDC officials who oversee the budget, according to an analyst who took part in the 90-minute briefing. The forbidden words are “vulnerable,” “entitlement,” “diversity,” “transgender,” “fetus,” “evidence-based” and “science-based.”

That’s the WaPo piece everyone is abuzz about.  A few observations:

1. This story may well be true, but I’d like more than “…according to an analyst who took part in the 90-minute briefing.”  Here is another account of what exactly is known.  Wasn’t “not publishing the article until it is better sourced” the evidence-based thing to do?

2. I don’t have a great fondness for the terms “evidence-based” or “science-based.”  When they are used on MR, it is often as a form of third-person reference or with a slight mock or ironic touch.  When I see the words used by others, my immediate reaction is to think someone is deploying it selectively, without complete self-awareness, or as a bullying tactic, in lieu of an actual argument, or as a way of denying how much their own argument depends on values rather than science.  I wouldn’t ban the words for anyone working for me, but seeing them often prompts my editor’s red pen, so to speak.  The most er…evidence-based people I know don’t use the term so much, least of all with reference to themselves.

3. In any case, the suggested replacement phrase — “CDC bases its recommendations on science in consideration with community standards and wishes” — I do not find offensive or anti-science, and I can imagine a plausible case that it is an actual improvement.  Science is (ought to be) value-free, yet CDC and more broadly federal policy should embody values too.  It should not think of itself as “the handmaiden of science.”

4. There is a fine line between “censorship” and “a bureaucratic organization which can be badly damaged by individual freelancing deciding to adopt uniform terminologies.”  I don’t doubt both might be going on here, but I’d like to see the extant Twitter takes show a little more subtlety on the broader point.  Don’t forget that the executive branch of government reports to the…executive, it is not a freestanding committee for debate, however much it might sometimes like to imagine otherwise.

5. The word “diversity” usually isn’t specific enough, or is channeling unstated preconceptions about how diversity should be interpreted.  We should improve our use of this word.  I have similar feelings about “vulnerable.”

6. People react to changes rather than levels.

7. “Fetus” — look, it is fine to disagree with the “pro-life view” (I’m not even sure what is the most neutral way of labeling it).  But is banning the use of the word “fetus” in institutional documents censorship?  What if an employee, during the Obama years, in an official CDC release had referred to a “fetus” as a “child”?  Would that have been changed back to fetus?  I am inclined to say yes.  Is it censorship in only one direction, or are both decisions censorship?  Or is this better seen as a disagreement over matters of fact?  A disagreement over values?  I am genuinely unsure, and I am unsure what a majority of the American public would think.  But I would say this is sooner worth a ponder than a rant.

8. If nothing else, Sam Altman can show up in China, post “here is my vulnerable entitlement diversity transgender fetus, who is evidence-based and science-based” on his Weibo account, and then go order some Chairman Mao’s braised pork belly.

9. What are the forbidden words in other parts of the federal government, whether de jure or de facto?  Will anyone be showing us a list?  Or is that list censored too?

“Where are they?”, cried out Enrico Fermi in anguish.  We have wondered ever since.  In spite of some subsequent refinements, I still find the Fermi paradox a…paradox.  Where are they?

Now, Oumuamua comes along…

And furthermore:

The object’s trajectory is so strange and its speeds are so blistering that it probably did not originate from within our solar system. Its discoverers concluded that the object is a rare interstellar traveler from beyond our solar system, the first object of its kind observed by humans.

So what do the academics say?

“The possibility that this object is, in fact, an artificial object — that it is a spaceship, essentially — is a remote possibility,” Andrew Siemion, director of the Berkeley Search for Extraterrestrial Intelligence Research Center, told The Washington Post on Monday.

Given the Fermi paradox, shouldn’t we assume a fairly high probability this is in fact some form of alien contact or display?  It’s like when you are expecting a package from UPS and then finally the doorbell rings…

So I’m excited, even though I don’t see much of a chance of a visit.  p = 0.3?  I need to crack open those old Arthur C. Clarke novels.

François Derrien, Ambrus Kecskes, and Phuong-Anh Nguyen have a new paper on this topic, and basically the answer is yes, because of labor supply effects:

We argue that a younger labor force produces more innovation. Using the native born labor force projected based on local historical births, we find that a younger age structure causes a significant increase in innovation. We use three levels of analysis in succession – commuting zones, firms, and inventors – to examine or eliminate various effects such as firm and inventor life cycles. We also find that innovation activities reflect the innovative characteristics of younger labor forces. Our results indicate that demographics increase innovation through the labor supply channel rather than through a financing supply or consumer demand channel.

Here is the SSRN link.

That is a new paper by Aghion, Akcigit, Hyytinen, and Toivanen, here is one brief excerpt:

In particular we see that IQ is by far the main characteristic for the probability of becoming an inventor in terms of the share of variation it explains, followed by parental education.  These two groups of variables account for 66% and 16% of the overall variation captured by our model.  In contrast, IQ plays a relatively speaking much more minor role for becoming a medical doctor or a lawyer.  Parental education is the main explanatory variable for the probability of becoming a medical doctor or a lawyer (40% and 53%), with base controls and parental income also playing clearly more important roles than for inventors.

The paper offers many other points of value, and you will note the data are from Finland.

Where?, I hear you asking.  No, that is the title of a new book by Karl Sigmund and the subtitle is The Vienna Circle and the Epic Quest for the Foundations of Science.  I enjoyed this book very much, though I don’t recommend it as a balanced introduction to its chosen topic.  I liked it best for its whims and interstices:

1. The mathematician Richard von Mises (brother of economist Ludwig) was a patron of Rilke, and he established a foundation for the sole purpose of supporting Robert Musil.

2. Carl Menger was planning on writing a philosophical treatise, and one which would have had a “Vienna Circle” anti-metaphysical slant.

3. Arguably Karl Popper learned the most from a polymathic cabinetmaker he was apprenticed to in his youth.

4. Friedrich Wieser had supported Mussolini, but a young Oskar Morgenstern, in his diary, complained that Wieser was too liberal.

5. Morgenstern later became a confirmed liberal, and he also remarked a few times that game theory was for the social sciences completing the research program of Kurt Gödel.

6. Karl Popper complained that Wittgenstein threatened him, in a lecture, with a poker.  It is not obvious this was the case.

7. I came away from my read wanting to sample more Ernst Mach, more Moritz Schlick, and thinking Otto Neurath was perhaps badly underrated.

Note that most of the book is more serious than this, and less concerned with economists, much more with math and science and some psychoanalysis and positivism too.

There was the ever-present worry that aircraft would make war even more horrific.  Some called for the international control of aviation to prevent its misuse.  A few even advocated the complete destruction of all aircraft on the grounds that even civilian machines could be adapted for war.

…At the opposite end of the spectrum were the enthusiasts who expected that soon everyone would be able to fly their own personal aircraft…As early as 1928, Popular Mechanics predicted a car that could be turned into a helicopter, but most commentators thought the autogyro was a better bet — although it did need a short horizontal run before take-off…As late as 1971, Isaac Asimov was still expecting that VTOL [vertical take-off and landing system] machines would eventually take the place of automobiles.

That is from Peter J. Bowler, A History of the Future: Prophets of Progress from H.G.Wells to Isaac Asimov.

One thing I learned from this book is that “money crank” Frederick Soddy was an early prophet of nuclear power, before many others understood the potential.  I am reminded of how “socialist crank” [oceans of lemonade with ships pulled by dolphins] Charles Fourier once prophesied that all of Europe would be tied together by railways.