Results for “high frequency”
120 found

The Volume Clock

That is the title of a new paper by David Easley, Marcos M. Lopez de Prado, and Maureen O’Hara:

Abstract:
Over the last two centuries, technological advantages have allowed some traders to be faster than others. We argue that, contrary to popular perception, speed is not the defining characteristic that sets High Frequency Trading (HFT) apart. HFT is the natural evolution of a new trading paradigm that is characterized by strategic decisions made in a volume-clock metric. Even if the speed advantage disappears, HFT will evolve to continue exploiting Low Frequency Trading’s (LFT) structural weaknesses. However, LFT practitioners are not defenseless against HFT players, and we offer options that can help them survive and adapt to this new environment.

The paper has many interesting bits, such as this:

Databases with trillions of observations are now commonplace in financial firms. Machine learning methods, such as Nearest Neighbor or Multivariate Embedding algorithms search for patterns within a library of recorded events. This ability to process and learn from what is known as “big data” only reinforces the advantages of HFT’s “event-time” paradigm, very much like how “Deep Blue” could assign probabilities to Kasparov’s next 20 moves, based on hundreds of thousands of past games (or more recently, why Watson could outplay his Jeopardy opponents).

The upshot is that speed makes HFTs more effective, but slowing them down won’t change their basic behavior: Strategic sequential trading in event time.

One message of the paper is that sequential strategic behavior will occur at any speed.  I liked this sentence:

As we have seen, HFT algos can easily detect when there is a human in the trading room, and take advantage.

And the ending bit is this:

There is a natural balance between HFTs and LFTs. Just as in nature the number of predators is limited by the available prey, the number of HFTs is constrained by the available LFT flows. Rather than seeking “endangered species” status for LFTs (by virtue of legislative action like a Tobin tax or speed limit), it seems more efficient and less intrusive to starve some HFTs by making LFTs smarter. Carrier pigeons or dedicated fiber optic cable notwithstanding, the market still operates to provide liquidity and price discovery – only now it does it very quickly and strategically.

Bad Scrabble strategy, from Alaska

I try not to blog Sarah Palin, but this passage, reproduced on Andrew Sullivan's blog, caught my interest for non-Palin reasons:

"Everybody in the family played Scrabble and took great pride in hoarding Ks and Qs and slapping them down in long, fancy words on triple-letter scores." — Going Rogue, p. 12.

Sullivan's reader objects that there is only one K and one Q but I think permissible to use the plural in this context, referring to general acts of hoarding over time.

My point is that this is bad Scrabble strategy.  The way to do very well is to put down seven-letter words on bonus squares, thereby getting the fifty-point bonus for using all your letters and doubled or tripled at that.  Such a strategy means maximizing one's holdings of S, R, E, T, O, A, and N, essentially, and dumping awkward letters which stand in the way.  "ING" is a powerful combination.  In addition, high frequency letters help you link up with other words running crossways, boosting your score further.

The astute MR reader will recognize here that we are dealing with portfolio theory, albeit where many assets are complements rather than near-perfect substitutes.

K doesn't mesh well with most other letters and so you should try to dump it quickly.  Q is paralyzing unless you have a U to go with it.  If you are happy because you could lay down "quit" on a double word score, for 26 points, I would say you are not a very ambitious Scrabble player, all the more if you hoarded letters and waited turns to do that.  (You have some chance of "aliquot" or "quaeres" or "quinoas," but do you really expect to score "obloquy," "quassia," or "qigongs"?, keeping in mind that if you build upon an already-laid tile you need an eight-letter word with q to score the bonus.)

If this is her game of Scrabble, you can only imagine what her foreign policy would be like.

Correction: If you search inside the book, you will see that she is referring to the Scrabble strategies of her grandparents, not her own Scrabble strategies.  They are the ones who cannot be trusted with U.S. foreign policy and it can also be said that she misses this chance to condemn their weak gaming strategies.

I thank Seth H. for the pointer.

Papers I wish I had written

What is truly scarce inside the human mind?  Hayek (The Sensory Order) and the neuroeconomists have grasped this as a central question of economics.  Here is a new paper:

Common intuition and experimental psychology suggest that the ability
to self-regulate, willpower, is a depletable resource. We investigate
the behavior of an agent who optimally consumes a cake (or paycheck or
workload) over time and who recognizes that restraining his consumption
too much would exhaust his willpower and leave him unable to manage his
consumption. Unlike prior models of self-control, a model with
willpower depletion can explain the increasing consumption sequences
observable in high frequency data (and corresponding laboratory
findings), the apparent links between unrelated self-control behaviors,
and the altered economic behavior following imposition of cognitive
loads. At the same time, willpower depletion provides an alternative
explanation for a taste for commitment, intertemporal preference
reversals, and procrastination. Accounting for willpower depletion thus
provides a more unified theory of time preference. It also provides an
explanation for anomalous intratemporal behaviors such as low
correlations between health-related activities.

My approach to willpower deletion, of course, is to always leave oneself wanting to do a little more of the virtuous task, rather than to overdiscipline.  If you have promised yourself 200 push-ups, stop at 198.  Here is the link.

Police work at the private margin

We study how public sector workers balance their professional motivations with private economic concerns, focusing on police arrests. Arrests made near the end of an officer’s shift typically require overtime work, and officers respond by reducing arrest frequency but increasing arrest quality. Days in which an officer works a second job after their police shift have higher opportunity cost, also reducing late-shift arrests. Combining our estimates in a dynamic model identifies officer preferences over workplace activity and overtime work. Our results indicate that officers’ private costs of arrests have a first-order impact on the quantity and quality of enforcement.

That is from a new NBER working paper by Aaron Chalfin and Felipe M. Gonçalves.

Behavioral Economics and GPT-4: From William Shakespeare to Elena Ferrante

There is a new paper on LLMs by Gabriel Abrams, here is the abstract:

We prompted GPT-4 (a large language model) to play the Dictator game, a classic behavioral economics experiment, as 148 literary fictional characters from the 17th century to the 21st century. 

Of literary interest, this paper analyzed character selfishness by century, the relative frequency of literary character personality traits, and the average valence of these traits. The paper also analyzed character gender differences in selfishness.

From an economics/AI perspective, this paper generates specific and quantifiable Turing tests which the model passed for zero price effect, lack of spitefulness and altruism, and failed for human sensitivity to relative ordinal position and price elasticity (elasticity is significantly lower than humans). Model updates from March to August 2023 had relatively minor impacts on Turing test outcomes.

There is a general and mainly monotonic decrease in selfish behavior over time in literary characters. 50% of the decisions of characters from the 17th century are selfish compared to just 19% of the decisions of characters from the 21st century. Overall, humans exhibited much more selfish behavior than AI characters, with 51% of human decisions being selfish compared to 32% of decisions made by AI characters.

Historical literary characters have a surprisingly strong net positive valence across 2,785 personality traits generated by GPT-4 (3.2X more positive than negative). However, valence varied significantly across centuries. The most positive century, in terms of personality traits, was the 21st — over 10X the ratio of positive to negative traits. The least positive century was the 17th at just 1.8X. “Empathetic,” “fair” and “selfless,” were the most overweight traits in the 20th century. Conversely, “manipulative,” “ambitious” and “ruthless” were the most overweight traits in the 17th century.

Male characters were more selfish than female characters: 35% of male decisions were selfish compared to just 24% for female characters. The skew was highest in the 17th century where selfish decisions for male and female were 62% and 20% respectively.

This analysis offers a specific and quantifiable partial Turing test. In a few ways, the model is remarkably human-like; The key human-like characteristics are the zero price effect, lack of spitefulness and altruism. However, in other ways, GPT-4 reflects unusual or inhuman preferences. The model does not appear to have human sensitivity to relative ordinal position and has significantly lower price elasticity than humans.

Model updates in GPT-4 have made it slightly more sensitive to ordinal value, but not more selfish. The model shows preference consistency across model runs for each character with respect to selfishness.

To which journal might you advise him to send this paper?

Emergent Ventures India, Cohort Five

The following was compiled by Shruti Rajagopalan, who directs Emergent Ventures India.  I will not indent the material:

Ankita Vijayvergiya is a computer Science Engineer and an entrepreneur. She founded BillionCarbon along with her co-founder Nikhil Vijayvergiya, to work on solving two problems that plague India – soil degradation and managing biodegradable waste. At BillionCarbon, they are nutrient mining from biodegradable waste to convert it into liquid bio-fertilizer. Their EV grant is to execute proof of concept with pilots, field trials, and technology validation.

Sujata Saha is an Associate Professor of Economics at Wabash College, Indiana. Her primary research interests are in International Finance and Trade, Open Economy Macroeconomics, and Financial Inclusion. She received her EV grant to study entrepreneurship and economic development in Dharavi, Mumbai, the largest slum in the world.

Aditya Mehta is an Arjuna Award-winning professional snooker player. Through the non-profit organization,  The ACE Snooker Foundation, he aims to teach and promote cue sports in India. He is creating a technology-based digital cue sports coaching solution, specifically aiming to develop a curriculum-based approach for schools and colleges across India.

Aditi Dimri (PhD, Economist) & Saraswati Chandra (Engineer, Entrepreneur) co-founded Cranberry.Fit to develop a virtual menstrual health coach with the aim to break through the traditional silence and apathy regarding painful periods and menstrual health. The EV grant supports the development of the virtual coach to help manage menstrual symptoms with the help of a personalized habits plan.

Vedanth Ramji  is a 15-year-old high school junior from Chennai, passionate about research at the intersection of Math, Computer Science, and Biology. He is currently a student researcher at the Big Data Biology Lab at QUT, Australia, where he develops software tools for Antimicrobial (AMR) research. He received his EV grant to travel to his lab at QUT, to develop deeper insights into AMR research and collaborate with his team on a publication which he is currently co-authoring.

Abhishek Nath is a 43-year-old entrepreneur tackling public restroom infrastructure and sanitation in urban areas head on. He is determined to bring Loocafe – a safe, hygienic, and accessible restroom for everyone – to cities around the world. He seeks to ensure that no city is more than a kilometer away from accessing a safe public toilet, providing youth easy and safe access to hygienic urban sanitation.

Sandhya Gupta is the founder of Aavishkaar, a teacher professional development institute that aims to educate, equip, and enable teachers of K-10 to become excellent science and math educators. Sandhya and Aavishkaar received an EV grant to help create an army of female Math educators helping students enjoy Math while chartering a career pathway for themselves in STEM fields.

Ankur Paliwal is a queer journalist and founder of queerbeat, a collaborative journalism project to cover the historically underserved LGBTQIA+ community in India. Over the last 13 years, Ankur has written narrative journalism stories about science, inequity, and the LGBTQIA+ community. He received an EV grant to build an online community and newsletter alongside queerbeat, to help transform public conversation about LGBTQIA+ persons in India.

Arsalaan Alam is a web developer, machine learning enthusiast, and aspiring rationalist. He is working on improving the conditions of harmonic coexistence between humans and wildlife. He got his Emergent Ventures grant to continue building Aquastreet, which consists of a hardware device that can be attached beneath a boat, after which it takes in audio of fish’s voices and converts the audio into a MEL frequency and then performs machine learning to classify the fish species, which is then displayed on the Aquastreet mobile app.

Soundarya Balasubramani  is a 26-year-old writer, author, and former product manager. She moved to the United States to pursue her master’s at Columbia University in 2017. Immigrants in the US face several barriers, including the decades-long wait times to get a green card for Indians, the lack of a startup visa for entrepreneurs, and the constant political battle that thwarts immigration reform. To reduce the barrier skilled immigrants face, Soundarya is has written a comprehensive book (Unshacked)  and is building an online community where immigrants can congregate, get guidance, and help each other.

Aadesh Nomula  is an engineer focused on cybersecurity. He is working on a single-point cybersecurity device for Indian homes and small-scale factories. His other interest is Philosophy.

Aurojeet Misra is an 18-year-old biology student at IISER Pune. He received his EV grant for his efforts on a radioactive tracing system to detect and locate forest fires. He hopes to test a prototype of this system to better understand its practical feasibility. He is interested in understanding different scientific disciplines like molecular biology, public health, physics, etc., and working on their interface.

Divyam Makar is a 24-year-old entrepreneur and developer working on Omeyo, a platform to connect local pharmacists, which aims to provide a large inventory to users with all the needed items, along with being super low-cost and interactive. They aim to deliver medicine to their users in as little as 20 minutes.

Divas Jyoti Parashar is a 23-year-old climate entrepreneur from Assam. He founded Quintinno Labs, a cleantech company driving the electric vehicle revolution by developing power banks for EVs. These compact and portable devices that fit in your car’s trunk aim to reduce range anxiety and offer emergency relief to EV users in developing countries that lack a charging station network. He is also working on deploying hydro-kinetic turbines in Assam to generate clean energy from flowing water. His recent passion project was a documentary about the impact of the 2021 volcanic eruption on the local population in La Palma Island.

Ray Amjad is prototyping scalable tools for finding and supporting the lost Einsteins and Marie Curies of the world – young people with exceptional math and science ability from under-resourced backgrounds. He received his EV Grant to help him find collaborators. He graduated from Cambridge, where he filmed many educational videos.

Amandeep Singh is a 22-year-old inventor and entrepreneur interested in machine learning and deep learning. He is building ‘Tiktok for India’, a short video-sharing app that allows people to edit and share videos with the world, create communities, and deliver authentic video content. Prior to this, he founded an AI surveillance startup, particularly for CCTV cameras.

Govinda Prasad Dhungana is an assistant professor at Far Western University, Nepal, and a doctoral candidate at Ghent University, Belgium. He is a public health researcher and co-founder of the Ostrom Center and he designs and implements high-impact HIV/Family Planning programs in marginalized communities. His EV grant will be used for piloting the community-based distribution (using Ostrom’s Design Principles and behavior change models) of a new self-injectable contraception (Sayana Press).

Kalash Bhaiya is a 17-year-old high-school student and social entrepreneur. She founded Fun Learning Youth (or FLY), a nonprofit that employs cohort-based mentorship by volunteers in their localities and received her EV grant to help reduce middle-school dropouts within underserved communities.

Kranthi Kumar Kukkala is a serial entrepreneur and technologist from Hyderabad.  He is working on a health care device – HyGlo – a non-invasive anemia diagnosing device. HyGlo is similar to a pulse oximeter, when a person puts their finger in the device probe, it investigates blood inside the finger without taking a blood sample and finds the hemoglobin percentage in the blood. This device can help young girls and women manage anemia (a big problem in India).

Kulbir Lamba is a 35-year-old researcher and practitioner, interested in understanding the startup landscape and received an EV grant for studying the evolution of DeepTech startups in India.

Keshav Sharma  is a 23-year-old entrepreneur working at the intersection of design, technology & marketing. Two years ago, he founded Augrade, a deeptech startup with his college friends. Augrade is an AI+AR platform to streamline the creation, editing, validation & visualization of 3D models at scale.

Srijon Sarkar is a 19-year-old researcher from Kolkata interested in mathematical oncology and applied rationality. He received his EV grant to study cancer systems, particularly Epithelial/Mesenchymal Plasticity through a lens of mathematical models and statistical algorithms, during his gap year. He will start his undergraduate degree (mathematics and biology) with a full scholarship at Emory University starting Fall 2023.

Shubham Vyas s an advocate for open discourse and democratic dialogue in India. With a background in data science and interest in philosophy, he received his EV grant to build his venture “Conversations on India,” into a multi-platform media venture to help shape the Indian political and economic discourse landscape.

Navneet Choudhary is an entrepreneur, and his journey started when he was 21 with a food delivery app for trains and buses across 70 cities in India. He received his EV Grant to develop LAMROD, a mobile application-based platform to manage trucking and cargo fleet operations at one place.

Srinaath Krishnan is a 20-year-old entrepreneur from Chennai. He received his EV grant to work on Zephyr, a start-up making credit scores universal and mobile, to enable immigrants to qualify for financial products using their international credit history.

Venkat Ram is an assistant professor at Indian Institute of Technology (IIT) Jodhpur, researching the development and deployment of human capital. He received his EV grant to study the structure and functioning of labor addas (proverbial marketplaces most daily wage laborers in India find work).

Arvind Subramanian,  is a 25-year-old sailor from Chennai and works as a product manager at Sportstar, the oldest sports magazine in India. He won his EV grant to enable his (and his team’s) participation in the 2022 J80 World Sailing Championship in Rhode Island, USA. He is working towards building and scaling the niche sporting scene in India.

Some past winners received additional grants:

Karthik Nagapuri, a 21-year-old programmer and AI engineer, for general career development.

Akash Kulgod is a 23yo cognitive science graduate from UC Berkeley founded Dognosis, where he is building tech that increases the bandwidth of human-canine communication. His grant will go towards launching a pilot study in Northern Karnataka testing the performance of cyber-canines on multi-cancer screening from breath samples. He writes on his Substack, about effective altruism, talent-search, psychedelics, and sci-fi uplift.

Those unfamiliar with Emergent Ventures can learn more here and here. The EV India announcement is here. More about the winners of EV India second cohort, third cohort, and fourth cohortTo apply for EV India, use the EV application click the “Apply Now” button and select India from the “My Project Will Affect” drop-down menu.

If you are interested in supporting the India tranche of Emergent Ventures, please write to me or to Shruti at [email protected].

Evidence from Italy’s ChatGPT Ban

We analyse the effects of the ban of ChatGPT, a generative pre-trained transformer chatbot, on individual productivity. We first compile data on the hourly coding output of over 8,000 professional GitHub users in Italy and other European countries to analyse the impact of the ban on individual productivity. Combining the high-frequency data with the sudden announcement of the ban in a difference-in-differences framework, we find that the output of Italian developers decreased by around 50% in the first two business days after the ban and recovered after that. Applying a synthetic control approach to daily Google search and Tor usage data shows that the ban led to a significant increase in the use of censorship bypassing tools. Our findings show that users swiftly implement strategies to bypass Internet restrictions but this adaptation activity creates short-term disruptions and hampers productivity.

That is from a recent paper by David Kreitmeir and Paul A. Raschky.  Via Pradyumna Shyama Prasad.

Machine Learning as a Tool for Hypothesis Generation

While hypothesis testing is a highly formalized activity, hypothesis generation remains largely informal. We propose a systematic procedure to generate novel hypotheses about human behavior, which uses the capacity of machine learning algorithms to notice patterns people might not. We illustrate the procedure with a concrete application: judge decisions about who to jail. We begin with a striking fact: The defendant’s face alone matters greatly for the judge’s jailing decision. In fact, an algorithm given only the pixels in the defendant’s mugshot accounts for up to half of the predictable variation. We develop a procedure that allows human subjects to interact with this black-box algorithm to produce hypotheses about what in the face influences judge decisions. The procedure generates hypotheses that are both interpretable and novel: They are not explained by demographics (e.g. race) or existing psychology research; nor are they already known (even if tacitly) to people or even experts. Though these results are specific, our procedure is general. It provides a way to produce novel, interpretable hypotheses from any high-dimensional dataset (e.g. cell phones, satellites, online behavior, news headlines, corporate filings, and high-frequency time series). A central tenet of our paper is that hypothesis generation is in and of itself a valuable activity, and hope this encourages future work in this largely “pre-scientific” stage of science.

Here is the full NBER working paper by Jens Ludwig and Sendhil Mullainathan.

Better predicting food crises

Anticipating food crisis outbreaks is crucial to efficiently allocate emergency relief and reduce human suffering. However, existing predictive models rely on risk measures that are often delayed, outdated, or incomplete. Using the text of 11.2 million news articles focused on food-insecure countries and published between 1980 and 2020, we leverage recent advances in deep learning to extract high-frequency precursors to food crises that are both interpretable and validated by traditional risk indicators. We demonstrate that over the period from July 2009 to July 2020 and across 21 food-insecure countries, news indicators substantially improve the district-level predictions of food insecurity up to 12 months ahead relative to baseline models that do not include text information. These results could have profound implications on how humanitarian aid gets allocated and open previously unexplored avenues for machine learning to improve decision-making in data-scarce environments.

Here is more from Ananth Balashankar, Lakshminarayanan Subramanian, and Samuel P. Fraiberger.

New facts about the game theory of balloons

But it turns out that China’s effort has been underway for more than a decade. According to a declassified intelligence report issued Thursday by the State Department, it involves a “fleet of balloons developed to conduct surveillance operations” that have flown over 40 countries on five continents.

That is from the Washington Post.  And:

Balloon operations obviously make sense for the Chinese. The United States has military bases in Japan and elsewhere from which it can launch daily flights by P-8 and other surveillance planes that fly perilously close to Chinese airspace. China doesn’t have similar options.

The frequency of these American “Sensitive Reconnaissance Operations,” or SROs, has increased sharply from about 250 a year a decade ago to several thousand annually, or three or four a day, a former intelligence official told me. China wants to push back, and collect its own signals; it wants its own version of “freedom of navigation” operations. Balloons are a way to both show the flag and collect intelligence…

Let’s look at another tit-for-tat motivation: China claims in its internal media that the Pentagon has aggressive plans to use high-altitude balloons, in projects such as “Thunder Cloud.”

It turns out the Chinese are right. Thunder Cloud was the name for the U.S. Army’s September 2021 exercise in Norway to test its “Multidomain Operations” warfighting concept, following a similar test in the Pacific in 2018, according to the Pentagon’s Defense News.

Here is my previous post on the game theory of the balloons.  Worth a reread.

Modeling persistent storefront vacancies

Have you ever wondered why there are so many empty storefronts in Manhattan, and why they may stay empty for many months or even years?  Erica Moszkowski and Daniel Stackman are working on this question:

Why do retail vacancies persist for more than a year in some of the world’s highest-rent retail districts? To explain why retail vacancies last so long (16 months on average), we construct and estimate a dynamic, two-sided model of storefront leasing in New York City. The model incorporates key features of the commercial real estate industry: tenant heterogeneity, long lease lengths, high move-in costs, search frictions, and aggregate uncertainty in downstream retail demand. Consistent with the market norm in New York City, we assume that landlords cannot evict tenants unilaterally before lease expiration. However, tenants can exit leases early at a low cost, and often do: nearly 55% of tenants with ten-year leases exit within five years. We estimate the model parameters using high-frequency data on storefront occupancy covering the near-universe of retail storefronts in Manhattan, combined with micro data on commercial leases. Move-in costs and heterogeneous tenant quality give rise to heterogeneity in match surplus, which generates option value for vacant landlords. Both features are necessary to explain longrun vacancy rates and the length of vacancy spells: in a counterfactual exercise, eliminating either move-in costs or tenant heterogeneity results in vacancy rates of close to zero. We then use the estimated model to quantify the impact of a retail vacancy tax on long-run vacancy rates, average rents, and social welfare. Vacancies would have to generate negative externalities of $29.68 per square foot per quarter (about half of average rents) to justify a 1% vacancy tax on assessed property values.

Erica is on the job market from Harvard, Daniel from NYU.  And they have another paper relevant to the same set of questions:

We identify a little-known contracting feature between retail landlord and their bankers that generates vacancies in the downstream market for retail space. Specifically, widespread covenants in commercial mortgage agreements impose rent floors for any new leases landlords may sign with tenants, short-circuiting the price mechanism in times of low demand for retail space.

I am pleased to see people working on the questions that puzzle me.

Labor Unions Reduce Product Quality

A very nice paper in Management Science by Kini, Shen, Shenoy and Subramanian finds that labor unions reduce product quality. Two strengths of the paper. First, the authors have relatively objective measures of product quality from thousands of product recalls mandated by the FDA, the Consumer Product Safety Commission and the National Highway Traffic Safety Administration covering many different industries. Second the authors use 3 different methods. First, they find that unionized firms are more likely to have recalls than non-unionized firms (a simple difference in means subject to many potential cofounds but I still like to see the raw data), second they find that in a panel model with industry and year fixed effects and other controls that firms which are more unionized have a greater frequency of product recalls. Finally they find that firms where the union just barely won the vote are more likely to have subsequent product recalls than firms for which the union just barely lost the vote–a regression discontinuity study.

In this paper, we study the impact of labor unions on product quality failures. We use a product recall as our measure of quality failure because it is an objective metric that is applicable to a broad cross-section of industries. Our analysis employs a union panel setting and close union elections in a regression discontinuity design framework to overcome identification issues. In the panel regressions, we find that firms that are unionized and those that have higher unionization rates experience a greater frequency of quality failures. The results obtain even at a more granular establishment level in a subsample in which we can identify the manufacturing establishment associated with the recalled product. When comparing firms in close elections, we find that firms with close union wins are followed by significantly worse product quality outcomes than those with close union losses. These results are amplified in non–right-to-work states, where unions have a relatively greater influence on the workforce.

The authors put more weight on financial strains caused by unionization as a mechanism whereas my story would be that unionization prevents firms from disciplining shoddy workers and that leads to lower product quality. Note that my theory would also cover teachers unions which the author’s mechanism would not.

Hat tip: Luke Froeb.

Photo Credit: Joe Piette.

Increased politicization and homogeneity in NSF grants

  1. This report uses natural language processing to analyze the abstracts of successful grants from 1990 to 2020 in the seven fields of Biological Sciences, Computer & Information Science & Engineering, Education & Human Resources, Engineering, Geosciences, Mathematical & Physical Sciences, and Social, Behavioral & Economic Sciences.
  2. The frequency of documents containing highly politicized terms has been increasing consistently over the last three decades. As of 2020, 30.4% of all grants had one of the following politicized terms: “equity,” “diversity,” “inclusion,” “gender,” “marginalize,” “underrepresented,” or “disparity.” This is up from 2.9% in 1990. The most politicized field is Education & Human Resources (53.8% in 2020, up from 4.3% in 1990). The least are Mathematical & Physical Sciences (22.6%, up from 0.9%) and Computer & Information Science & Engineering (24.9%, up from 1.5%), although even they are significantly more politicized than any field was in 1990.
  3. At the same time, abstracts in most directorates have been becoming more similar to each other over time. This arguably shows that there is less diversity in the kinds of ideas that are getting funded. This effect is particularly strong in the last few years, but the trend is clear over the last three decades when a technique based on word similarity, rather than the matching of exact terms, is used.

That is from a new CSPI (Richard Hanania’s group) study by Leif Rasmussen.

A few observations on my latest podcast with Amia Srinivasan

I am reluctant to do this, as I have never offered ex post commentary on a Conversations with Tyler before.  It seems unfair to the guest (who may or may not have comparable platforms), and perhaps it is the guest who deserves the last word?  Still, I think I can at least try to clear up a few misunderstandings about the episode, as I see a number of important points at stake here.  So here goes, with some trepidation:

1. The number, frequency, and extremity of reactions to the episode, both on Twitter and in the MR comments section, I think shows that women simply have a much, much tougher time in the public sphere.  There is a much smaller intellectual and emotional space they are allowed to inhabit comfortably and without condemnation or excess judgment.  Had the episode been with a man, and had been comprised of the exact same words, it would not have received nearly the same attention or criticism.  But people don’t like women who argue back.  I realize that is a kind of cliche, but it is largely true.

In this regard, even if you largely disagree with Amia Srinivasan, you should take the strength of the reaction to the episode as a sign she might have a valid point after all.

And to put it bluntly, if said female guest plausibly can be perceived as attractive, the reaction will be all the more disproportionate.

2. Some listeners are teed off about “disabled individuals” vs. “disabled men.”  I’ve committed numerous tongue and memory slips in my time, and they are hardly ever pointed out.  Now you might be upset that she insisted I said “men” (when I didn’t), but in fact my interior monologue at the time was something like this: “We all know this is mostly about men.  But if I just say “men,” she will react to that word and drive the conversation in a different direction.  So I will say “individuals.””  Maybe she gets points for insight?

3. If I challenge a guest directly, it is typically a sign of intellectual respect for said guest or person (just ask Bryan Caplan, though perhaps by this point he has suffered too much?).  And if the guest comes back at me forcefully, I usually (and consistently) take that as a sign of respect.  If I don’t seem frustrated, it is because I am not.

4. If a guest challenges my questions (or indeed anyone’s questions) for having sexist premises, I don’t consider this an illegitimate response.  I may or may not agree, but I don’t think it should upset me (or you).  I think a lot of people’s questions have for instance highly statist or collectivist premises (and should not).  I may or may not be right, but surely that too is a response deserving of consideration, should I decide to raise it.

5. To be fully forthright, if you wish to hear my “negative take” on her responses, I don’t think she was very good at handling empirical evidence in the context of a discussion, and furthermore this is a major shortcoming.  I find this to be common amongst philosophers, if I may be allowed to continue my moment of condescension.  I also had the feeling she is not challenged sufficiently often with said evidence, and that may partly be the fault of Oxford.  This is exactly the point where I feel bad/uncertain offering ex post commentary on the episode, but still leaving off this opinion would not be offering my honest assessment of what happened.

6. I have studied her work carefully, including reading her doctoral dissertation and some undergraduate work, and I then and still now fully believe she will be one of the more important philosophers over the next few decades.  As I mentioned before, super-impressive in terms of combining intellect, depth, breadth, determination, and relevance, plus has the all-important “willing to put oneself out there.”  And if you don’t trust me as talent-spotter, dare I point out that Oxford University has a not too shabby history choosing and developing philosophical talent?  But to return a bit to boasting, I think my relatively strong ability to differentiate emotional response from the talent judgment is in fact one reason to trust my talent judgments.

7. You have to learn to learn from people who bother, annoy, or frustrate you.  If you do, they will not in fact bother, annoy, or frustrate you.  One central point under consideration is her view that even today in the Western or also Nordic countries, the treatment of women (among other groups) could plausibly be much, much better, and with general emancipatory effects for many other groups as well.  You may or may not agree, but is that such a crazy question to ponder and think through?  No.

So I thought it was a good episode.  I would gladly do another one with her someday, and I hope the feeling is mutual.

The NYTimes on the FDA and Rapid Tests

In July of 2020 I wrote in Frequent, Fast, and Cheap is Better than Sensitive:

A number of firms have developed cheap, paper-strip tests for coronavirus that report results at-home in about 15 minutes but they have yet to be approved for use by the FDA because the FDA appears to be demanding that all tests reach accuracy levels similar to the PCR test. This is another deadly FDA mistake.

…The PCR tests can discover virus at significantly lower concentration levels than the cheap tests but that extra sensitivity doesn’t matter much in practice. Why not? First, at the lowest levels that the PCR test can detect, the person tested probably isn’t infectious. The cheap test is better at telling whether you are infectious than whether you are infected but the former is what we need to know to open schools and workplaces.

It’s great that other people including the NYTimes are now understanding the problem. Here is the excellent David Leonhardt in Where are the Tests?

Other experts are also criticizing the Biden administration for its failure to expand rapid testing. Even as President Biden has followed a Covid policy much better aligned with scientific evidence than Donald Trump’s, Biden has not broken through some of the bureaucratic rigidity that has hampered the U.S. virus response.

In the case of rapid tests, the F.D.A. has loosened its rules somewhat over the past year, allowing the sale of some antigen tests (which often cost about $12 each). But drugstores, Amazon and other sellers have now largely run out of them. I tried to buy rapid tests this weekend and couldn’t find any.

The F.D.A.’s process for approving rapid tests is “onerous” and “inappropriate,” Daniel Oran and Dr. Eric Topol of Scripps Research wrote in Stat News.

For the most part, the F.D.A. still uses the same cumbersome process for approving Covid tests that it uses for high-tech medical devices. To survive that process, the rapid tests must demonstrate that they are nearly as sensitive as P.C.R. tests, which they are not.

But rapid tests do not need to be so sensitive to be effective, experts point out. P.C.R. tests often identify small amounts of the Covid virus in people who had been infected weeks earlier and are no longer contagious. Rapid tests can miss these cases while still identifying about 98 percent of cases in which a person is infectious, according to Dr. Michael Mina, a Harvard epidemiologist who has been advocating for more testing

Identifying anywhere close to 98 percent of infectious cases would sharply curb Covid’s spread. An analysis in the journal Science Advances found that test frequency matters more for reducing Covid cases than test sensitivity.

As I said on twitter what makes the FDA’s failure to approve more rapid antigen tests especially galling is that some of the tests being sold cheaply in Europe are American tests just ones not approved in the United States. If it’s good enough for the Germans it’s good enough for me!