Month: August 2022
Will open science matter?
Yesterday I posted this link, about how federally funded science will have to be made open access right away. I’m all for this, as it has some upside and no downside that I can see. Still, at the margin I am not sure it will make a huge difference.
Who says when the research is “ready” to be posted? No matter what you put in the fine print, de facto that is a decision made by the scientists. If scientists wish to delay open access publication, I doubt if this will stop them. “The paper simply isn’t finished yet.” The law cannot in practice dictate otherwise. Of course scientists already put plenty of works on-line and open to the public, and this private calculation will continue, but with only modest changes.
The actual main effect will be to enable scientists to resist commercial attempts to monopolize publication rights in closed access form. Presumably such contracts now will be illegal. But think about the new equilibrium: there will be a final, published, canonical version of the article published in The Journal of Botchagaloup. There will be an open access, not yet published, non-canonical (no proper pp. at the very least) version on the scientist’s home page. And very often there will be an illegal copy of the canonical version on SciHub, the pirate site for scientific papers. Plus the data copies that circulated before the commercial publisher made the authors take them down.
How is that so different from the status quo? Some scientists who didn’t get a crack at the data the first time around won’t have to wait as long to access it. And maybe scientists will make more of an effort for the open access version of their papers to be closer to the canonical versions published in commercial journals. This could prove a modest benefit, though you, as an outside scientist, wishing to cite “p.43” just won’t know how canonical the open access version will be. And presumably for-profit commercial journals will add extra stages to the final production process, if only to keep interest in the product they are selling, relative to the open access versions on-line. So I don’t think it will “do under” for-profit scientific publishing, not to mention that many articles are not federally funded by the U.S. government.
A more radical policy change would have been to require the journals to make their final versions of the papers open to the public, and in the final, canonical versions. That would create greater benefits, but also run the risk of putting those journals out of business. As I understand the new dictate, it does not do this.
The new law also will give scientists leverage against private companies that wish to buy up the research rights and not publicize the results. Probably this is a benefit, though that doesn’t hold a priori, as it does raise the cost of private sector involvement by forcing them to share the information more. And some unscrupulous scientists might try to get a better deal from the companies by releasing a different and inferior version of “the public research results” to the open access community. Still, on net I expect these are benefits.
Addendum: You may recall that Fast Grants had its own version of an open access requirement. I think this worked quite well! But it is interesting why it might have proven effective. I think we credibly signaled that people with open and early good results would be plausible candidates for additional funding, and soon, and indeed some of them were. To the extent that the federal government can signal the same, good incentives will be all the stronger. But this new OSTP order does not coordinate future funding decisions per se, and I don’t see any clauses that the NIH, NSF, and others are bound to revise their funding policies accordingly, to favor researchers who come out with speedy, open results. So the benefits here could be much greater if the entire federal science apparatus could signal its prioritization of speed and openness. We are still quite far from this. Nonetheless, this policy is a marginal improvement and a step in the right direction and its creates some preconditions for matters getting better yet.
Saturday assorted links
The long-awaited upgrade to Ethereum
Which are the best sources to read on what is happening, and its likely benefits and problems? If you are starting from scratch, here is some NYT coverage. The Ethereum Foundation reports:
- Soon, the current Ethereum Mainnet will merge with the Beacon Chain proof-of-stake system.
- This will mark the end of proof-of-work for Ethereum, and the full transition to proof-of-stake.
- This sets the stage for future scaling upgrades including sharding.
- The Merge will reduce Ethereum’s energy consumption by ~99.95%.
What else should I be reading on this? Here are various Reddit threads. Here is coverage from Timothy B. Lee. And what are you all expecting? (Please don’t just rehash the standard crypto debates.)
The Student Loan Giveaway is Much Bigger Than You Think
Wiping out 10k in student debt is not the most expensive part of the Biden student loan program. Most Federal student loans are now eligible for an income based repayment plan, under these plans students pay a small percentage of their “discretionary” income, say 10%, and then after a fixed number of years the debt is wiped off the student’s books. At first glance these plans don’t seem crazy, but as Matt Bruenig points out they create perverse incentives.
Under the Public Service Loan Forgiveness (PSLF) program, law graduates that go on to work in the public sector, which is a lot of them as the public sector employs many lawyers, only have to pay 10 percent of their discretionary income for 10 years in order to have their debt forgiven.
Law schools figured out many years ago that, for a student who is planning to enroll in PSLF upon graduation, prices and debt loads don’t matter. Ten percent of your discretionary income is ten percent of your discretionary income regardless of what the law school charges you and how much debt you nominally have to take on.
Law schools also realized that they could make the deal even sweeter by setting up LRAPs [repayment programs, AT] that give graduates money to cover the the modest repayments required by the PSLF.
The LRAP schemes work as follows:
- The school increases their tuition.
- The student takes out federal loans to cover the tuition increase.
- The school squirrels away the debt-financed tuition increase into an LRAP fund.
- The school disburses money from the LRAP fund to cover PSLF repayments.
Did you get that? Here’s a stylized example. Suppose a student will make 150k per year for 10 years working in the public sector. If they have 200k in debt they pay 15k every year to the government for 10 years and then 50k is “forgiven.” But now the law school comes to the student and says ‘heh, I have a deal which will make both of us better off. We are going to raise the price of law school to 400k but don’t worry not only won’t that cost you a penny more than the 15k a year you are already obligated to pay it will actually cost you much less because we will pay your payments of 15k per year!’ This indeed is a great deal for the student who pays nothing and it’s a great deal for the law school which gets 200k more revenue immediately in return for 150k of payments paid out over the following 10 years. Win-win! Except for the taxpayer of course.
But wait there’s more. Student loans can be used not only to pay tuition and fees but also to pay “living expenses.” Thus, under these plans, students have an incentive to take out as big a loan as allowed in excess of tuition and fees because no matter how large the loan the student’s costs are zero! Lyman Stone has a good tweet thread giving many examples of how to game the system such as “Every student should borrow their maximum loan eligibility and then find some way to invest it illegally. My strategy would be: rent a wildly oversized apartment and sublet.” And here is a tweet thread from Michael Feinberg showing how even wealthy parents may be able to game the system.
Furthermore, the new Biden plan makes the income driven repayment schemes even more generous!
The IDR changes are four-fold:
- Increase the amount of income not subject to IDR from 150 percent of the federal poverty line to 225 percent of the federal poverty line.
- Reduce the interest rate on IDR-enrolled loans to 0 percent.
- For undergraduate debt, reduce the IDR rate from 10 percent of income beyond the threshold in (1) to 5 percent of income beyond the threshold in (1).
- For IDR-enrolled debts with original loan balances below $12,000, reduce the repayment period from 20 years to 10 years.
Essentially what this means is that every school will now have the possibility of using a law school like program to shift costs onto taxpayers. Thus Bruenig concludes:
…going forward, these new rules could quite radically alter the incentives of colleges and students when it comes to college prices, institutional financial aid, how much debt to take on, and how to approach repayment.
Indeed, these programs are likely to be very expensive and the resulting increase in the price of tuition will lead to calls either to end the program or for price controls on education.
What I’ve been reading
Frances Spalding, The Real and the Romantic: English Art Between the Two Wars. Wonderful text, quality images, and the whole subject area remains underrated, so this book was a big plus for me. The history of modernism is not just cubist and abstract art on the continent.
Chris Miller, Chip War: The Fight for the World’s Most Critical Technology. I liked this book and found it useful, though I wished for more on Taiwan and more recent times, and for less on the earlier years. Just my subjective preference.
Alice Bentinck and Matt Clifford, How to be a Founder: How entrepreneurs can identify, fund and launch their best ideas. Do you have it in you to be a founder? If you are asking that question, this book is maybe the best place to start looking for some answers.
Thomas H. Davenport and Steven M. Miller, Working with AI: Real Stories of Human-Machine Collaboration. Actual examples!
There is also Ajay Agrawal, Joshua Gans, and Avi Goldfarb, Power and Prediction: The Disruptive Economics of Artificial Intelligence, which I have not yet read.
Samuel Gregg, The Next American Economy: Nation, State, and Markets in an Uncertain World is a useful corrective to some recent attempts to overrate the import of industrial policy, especially in an American context.
Celia Paul, Letters to Gwen John I found a moving set of (imaginary) letters from one living female painter to another first-rate deceased female painter, both having lived through some similar situations. Excellent color plates too.
Christopher Marquis and Kunyuan Qiao, Mao and Markets: The Communist Roots of Chinese Enterprise. A good look at the essential continuity in Chinese history between the Maoist period and the “capitalist” period. Of course the main thesis no longer seems so crazy as it might have ten years ago.
How does such a “power grab” stay secret?
Prime Minister Anthony Albanese announced on Friday that former High Court Justice Virginia Bell will report on Nov. 25 on the findings of her inquiry into [Scott] Morrison’s secret power grab.
Morrison secretly appointed himself to five ministerial roles between March 2020 and May 2021, usually without the knowledge of the original minister.
Albanese, who replaced Morrison in May elections, cited Solicitor-General Stephen Donaghue’s legal opinion that while the appointments were legal they undermined the principle of responsible government….
Morrison, who is now an opposition lawmaker, maintains that he gave himself the portfolios of health, finance, treasury, resources and home affairs as an emergency measure made necessary by the coronavirus crisis.
But his only known use of the secret powers had nothing to do with the pandemic. He overturned a decision by former Resources Minister Keith Pitt to approve a contentious gas drilling project near the north Sydney coast that would have harmed his conservative coalition’s reelection chances.
Here is the full story. But really people — how is this secrecy possible? No one was tempted to mention this on Twitter? Did it involve no paperwork? Isn’t such a “secret” minister just a windowless monad? Or why not install loyal “lackey type” subordinates instead? Or is this like Trump’s “document declassification,” existing mainly in the mind of the executive and nowhere else? Model this for me please!!
China’s Heat Wave
Axios: The extreme heat and drought that has been roasting a vast swath of southern China for at least 70 straight days has no parallel in modern record-keeping in China, or elsewhere around the world for that matter.
… “I can’t think of anything comparable to China’s heat wave of summer 2022 in its blend of intensity, duration, geographic extent and number of people affected,” meteorologist Bob Henson, a contributor to Yale Climate Connections, told Axios.
Friday assorted links
1. The sheep human contest in France.
2. The culture that is Illinois.
3. How does fear of failure affect performance on Master Chef?
4. “The AI startup erasing call center worker accents: is it fighting bias – or perpetuating it?”
5. Kevin Erdmann on rising rents.
6. Do people underestimate how much they will enjoy strangers? (NYT)
7. “Baidu claims it’s the world’s first hardware-software integrated quantum computer and it’s already available to use.” I have found such past reports unreliable, but…
8. Howard Rosenthal obituary (NYT).
France fact of the day
French power prices for 2023 are trading €900/MWH with winter prices above €1.500/MWH (or an eye-popping 30X the 5-yr average price).
Here is more information. Nuclear is not enough, you also have to keep your facilities up and running and in good repair…
America is now pulling away from Western Europe
That is the topic of my latest Bloomberg column, the energy discussion is obvious so here is the segment on health care systems:
The pandemic revealed years of capital underinvestment in many of European health-care systems. Many Americans used to admire the UK’s National Health Service, but right now the whole system is ailing. There has been a labor and capital shortage, and a collapse of emergency health care services, which may be costing up to 500 excess (non-Covid) deaths a week. Similar problems exist throughout Europe, though they seem to be worst in the UK.
The American hospital and health care system long has been good — too good — at making expensive, long-term investments in care and technology. Often this meant excess prices and not much of an improvement in basic care. But in the pandemic and post-pandemic environment, that feature of the system has kept US health care up and running. All that capital investment turns out to have been pretty useful in a major crisis.
The longstanding charge that the US does not have universal health care now is less relevant. Obamacare is highly imperfect along a variety of dimensions, but US health care coverage has never been higher — the percentage of the uninsured population is now 8%. Keep in mind that many of those uninsured may have decided not to purchase health insurance, instead preferring to spend their money in other ways. That might be a personal mistake, but that is not the same thing as a systemic failure of the entire US health care regime.
America actually has something pretty close to universal coverage, at least as an option. And remember that some of the European systems, most notably in Switzerland, also require significant out-of-pocket expenditures. Other parts of those systems are paid through relatively regressive systems of a value-added tax, so they are not as “free” as they might seem.
Recommended.
Kin-based institutions and economic development
Though many theories have been advanced to account for global differences in economic prosperity, little attention has been paid to the oldest and most fundamental of human institutions: kin-based institutions—the set of social norms governing descent, marriage, clan membership, post-marital residence and family organization. Here, focusing on an anthropologically well established dimension of kinship, we establish a robust and economically significant negative association between the tightness and breadth of kin-based institutions—their kinship intensity—and economic development. To measure kinship intensity and economic development, we deploy both quantified ethnographic observations on kinship and genotypic measures (which proxy endogamous marriage patterns) with data on satellite nighttime luminosity and regional GDP. Our results are robust to controlling for a suite of geographic and cultural variables and hold across countries, within countries at both the regional and ethnolinguistic levels, and within countries in a spatial regression discontinuity analysis. Considering potential mechanisms, we discuss evidence consistent with kinship intensity indirectly impacting economic development via its effects on the division of labor, cultural psychology, institutions, and innovation.
That is a new and very important paper by Duman Bahrami-Rad, Jonathan Beauchamp, Joseph Henrich, and Jonathan Schulz, the two Jonathans being my colleagues at GMU.
Thursday assorted links
1. Why so few economies of scale in construction?
2. Salem Center/CSPI forecasting tournament.
3. Play Magnus Group receives Chess.com offer.
4. A Hollywood insider’s guide as to what to do in Belfast.
5. Megan McArdle on student loan forgiveness.
6. “We have created the first example of an engineering material that can simultaneously sense, think and act upon mechanical stress without requiring additional circuits to process such signals,” Harne said. “The soft polymer material acts like a brain that can receive digital strings of information that are then processed, resulting in new sequences of digital information that can control reactions.” Link here.
On the Internet Nobody Knows You Are a Dog
Bitcoin.com: A set of hackers managed to impersonate Binance chief communications officer (CCO) Patrick Hillmann in a series of video calls with several representatives of cryptocurrency projects. The attackers used what Hillman described as an AI hologram, a deepfake of his image for this objective, and managed to fool some representatives of these projects, making them think Hillmann was helping them get listed on the exchange.
It’s wild how good fake and synthetic video are becoming. Synthesia, for example, uses synthetic AI video instead of actors for training videos and similar–see the example below. Tyler and I could write econ scripts and then use AIs to create video–much cheaper and less time consuming than hiring a video director. Presumably, we could even create synthetic AIs in our own image.
Notice also that Synthesia reads text but could presumably also repeat text that it hears. Thus, the future of Zoom meetings is avatars. Fake faces on fake backgrounds. Choose your race, gender, age, species and so forth. No discrimination possible! Some people won’t be happy.
My Conversation with Cynthia L. Haven
Here is the audio and transcript, here is part of the summary:
…those two interests converged as they led her to interview and write books about three writers and thinkers whom she also came to call mentors: René Girard, Czeslaw Milosz, and Joseph Brodsky.
Cynthia joined Tyler to discuss what she’s gleaned from each of the three, including what traits they have in common, why her biography of Girard had to come from outside academia, Milosz’s reaction to the Berkley Free Speech Movement, Girard’s greatest talent — and flaw — as a thinker, whether Brodsky will fall down the memory hole, why he was so terrible on Ukraine, why Cynthia’s early career was much like The Devil Wears Prada, the failings of Twitter, and more.
And one excerpt:
COWEN: What is your philosophy of what is missing in most other people’s interviews?
HAVEN: I don’t know that it’s a philosophy.
COWEN: You must think you’re adding something, right?
HAVEN: I’m interested in big questions. I think a lot of people aren’t. A lot of interviewers aren’t. It’s not an era for big questions, is it?
COWEN: 2022? I’m not sure.
HAVEN: Really?
COWEN: Maybe the questions are either too big or too small and not enough in between.
HAVEN: That’s an interesting point of view.
COWEN: There’s plenty of ideology in the world and in this country. It doesn’t have to be a good thing, but —
HAVEN: Ideology is different than big questions, I think.
Interesting throughout.
*Of Boys and Men*
…I was shocked to discover that many social policy interventions, including some of the most touted, don’t help boys and men. The one that first caught my eye was a free college program in Kalamazoo, Michigan. According to its evaluation team, “women experienced large gains,” in terms of college completion (increasing by 50%), “while men seem to experience zero benefit.” This is an astonishing finding. Making college free had no impact on men…So not only are many boys and men struggling, they are less likely to be helped by policy interventions.
And:
In the U.S. for example, the 2020 decline in college enrollment was seven times greater for male than for female students.
And:
The bottom line is that Finland’s internationally acclaimed educational performance is entirely explained by the stunning performance of Finnish girls.
That is from the forthcoming Richard V. Reeves book, one of the most important of this year, perhaps the most important.