Category: Data Source
Is U.S. deposit outflow an overrated problem?
US bank deposits are a third higher than they were at the start of 2020, which makes worries about a banking system liquidity crisis seem a little overwrought (though, to be fair, it is changes in liquidity, not absolute liquidity levels, that matter most to markets)…
So there was a one-time outflow of about $185bn, or about 3 per cent of small banks’ deposits. The next week, however, small bank deposits were stable. With the usual qualifiers — things can always go wrong later, and so on — US banks do not seem to have a deposit outflow problem.
Here is more from the FT.
The Causal Effect of Economic Freedom on Female Employment & Education
While we have decades of evidence that economically free economies grow faster and are more productive than un-free ones, we have less knowledge about the effect of economic freedom on groups that have traditionally been disadvantaged. I study the causal effects of large and sustained jumps in economic freedom on women’s labor force participation and primary school enrollment. I find that these jumps have a positive and significant effect in both cases–economic freedom is good for women’s labor force opportunities and female education.
That is from Robin Grier at TTU.
Personality Differences and Investment Decision-Making
We survey thousands of affluent American investors to examine the relationship between personalities and investment decisions. The Big Five personality traits correlate with investors’ beliefs about the stock market and economy, risk preferences, and social interaction tendencies. Two personality traits, Neuroticism and Openness, stand out in their explanatory power for equity investments. Investors with high Neuroticism and those with low Openness tend to allocate less investment to equities. We examine the underlying mechanisms and find evidence for both standard channels of preferences and beliefs and other nonstandard channels. We show consistent out-of-sample evidence in representative panels of Australian and German households.
That is from a new NBER working paper from Zhengyang Jiang, Cameron Peng, and Hongjun Yan.
Côte d’Ivoire claim of the day
Côte d’Ivoire citizens pay the highest income taxes in the world according to this year’s survey findings by World Population Review.
While both its sales and corporate tax regimes may be considerably lower than those of other countries globally, at 60%, Côte d’Ivoire’s income tax rates are markedly higher compared to developed countries.
Only Finland (56.95%), Japan (55.97%), Denmark (55.90%), and Austria (55%), closely follow Côte d’Ivoire to round up the top five countries with the highest income tax, in a study that surveyed over 150 countries.
How much people pay of course is yet another matter. Here is the link, via Jodi Ettenberg.
Lifespans of the European Elite, 800-1800
I analyze the adult age at death of 115,650 European nobles from 800 to 1800. Longevity began increasing long before 1800 and the Industrial Revolution, with marked increases around 1400 and again around 1650. Declines in violent deaths from battle contributed to some of this increase, but the majority must reflect other changes in individual behavior. There are historic spatial contours to European elite mortality; Northwest Europe achieved greater adult lifespans than the rest of Europe even by 1000 AD.
Here is the paper by Neil Cummins, via Matt Yglesias.
Are social media making us miserable?
Stuart Richie rebuts some of the recent studies:
And here’s the thing: when the authors of the “Facebook arrival” study raised their standards in this way, running a correction for multiple comparisons, all the results they found for well-being were no longer statistically significant. That is, a somewhat more conservative way of looking at the data indicated that every result they found was statistically indistinguishable from a scenario where Facebook had no effect on well-being whatsoever.
Now let’s turn to the second study, which was a randomised controlled trial where 1,637 adults were randomly assigned to shut down their Facebook account for four weeks, or go on using it as normal. Let’s call it the “deactivating Facebook” study. This “famous” study has been described as “the most impressive by far” in this area, and was the only study cited in the Financial Times as an example of the “growing body of research showing that reducing time on social media improves mental health”.
The bottom-line result was that leaving Facebook for a month led to higher well-being, as measured on a questionnaire at the end of the month. But again, looking in a bit more detail raises some important questions.
First, the deactivation happened in the weeks leading up to the 2018 US midterm elections. This was quite deliberate, because the researchers also wanted to look at how Facebook affected people’s political polarisation. But it does mean that the results they found might not apply to deactivating Facebook at other, less fractious times – maybe it’s particularly good to be away from Facebook during an election, when you can avoid hearing other people’s daft political opinions.
Second, just like the other Facebook study, the researchers tested a lot of hypotheses – and again, they used a correction to reduce false-positives. This time, the results weren’t wiped out entirely – but almost. Of the four questionnaire items that showed statistically-significant results before the correction, only one – “how lonely are you?” – remained significant after correction.
It’s debatable whether even this result would survive the researchers corrected for all the other statistical tests they ran. Not only that, but they also ran a second model, controlling for the overall amount of time people used Facebook, and this found even fewer results than the first one.
Third, as well as the well-being questionnaire at the end of the study, the participants got daily text messages asking them how happy they were, among other questions. Oddly, these showed absolutely no effect of being off Facebook – and not even the slightest hint of a trend in that direction.
Here is the entire piece, which is well thought out.
Time Passages
Here’s an interesting idea it wouldn’t have occured to me to ask. What is the length of time described in the average 250 words of narration and how has this changed over time? Most famously James Joyce’s “Ulysses” is a long novel about single day with many pages describing brief experiences in minute detail. In contrast, Olaf Stapledon’s Last and First Men covers 2 billion years in fewer words than Joyce uses to cover a single day.
Using human readers grading 1000 passages, Underwood et al. (2018) finds that the average length of time described in a typical passage has declined substantially since the 1700s, from a day to about an hour so a decline by a factor of 24. Writers have become much more focused on describing individual experiences than events.
In a blog post, Underwood revisits his earlier work but this time comparing human readers with various statistical techniques and also GPT. To use GPT he prompts
Read the following passage of fiction. Then do five things. 1: Briefly summarize the passage. 2: Reason step by step to decide how much time is described in the passage. If the passage doesn’t include any explicit reference to time, you can guess how much time the events described would have taken. Even description can imply the passage of time by describing the earlier history of people or buildings. But characters’ references to the past or future in spoken dialogue should not count as time that passed in the scene. Report the time using units of years, weeks, days, hours, or minutes. Do not say zero or N/A. 3: If you described a range of possible times in step 2 take the midpoint of the range. Then multiply to convert the units into minutes. 4: Report only the number of minutes elapsed, which should match the number in step 3. Do not reply N/A. 5: Given the amount of speculation required in step 2, describe your certainty about the estimate–either high, moderate, or low. The passage follows: <omitting this to save space>
Here’s an example of the analysis:
ABRIDGED PASSAGE FROM Jane Eyre: And signs, for aught we know, may be but the sympathies of Nature with man. When I was a little girl, only six years old, I one night heard Bessie Leaven say to Martha Abbot that she had been dreaming about a little child … Of late I had often recalled this saying and this incident; for during the past week scarcely a night had gone over my couch that had not brought with it a dream of an infant, which I sometimes hushed in my arms, sometimes dandled on my knee, sometimes watched playing with daisies on a lawn, or again, dabbling its hands in running water. It was a wailing child this night, and a laughing one the next: now it nestled close to me, and now it ran from me; but whatever mood the apparition evinced, whatever aspect it wore, it failed not for seven successive nights to meet me the moment I entered the land of slumber.
ASSISTANT REPLIES.
1: The narrator recalls a saying about dreaming of children being a sign of trouble and remembers an incident from her childhood. She then describes her recent dreams of infants over the past week.
2: The passage mentions the narrator’s childhood memory, which happened when she was six years old, and the past week of dreams. The time elapsed in the passage is the week of dreams.
3: 1 week has elapsed. At seven days a week that’s 7 days. 24 hours in a day, so 168 hours. 60 minutes in an hour, so 10080 minutes.
4: 10080 minutes.
5: Moderate confidence, as the childhood memory is not counted as time passing in the scene.
Overall, Underwood finds that GPT-4 is more accurate than previous statistical techniques in that its judgments correlate with those of human readers nearly as much as human readers correlate with themselves.
Addendum: See also my paper with Tyler on how to use GPT models.
Bike riding is falling in Portland
Overall, Portland bicycle traffic in 2022 dropped more than a third compared to 2019, to levels not seen since approximately 2005-2006 (Table 1). This is based on a comparison of people counted at the 184 locations that were counted in both 2022 and 2019. Volunteers recorded 17,579 people biking at those 184 locations in 2022, a 37% drop from the 27,782 counted at the same locations in 2019. This reversion to earlier and lower volumes is also reflected in bicycle commute data, as well as for driving, walking, and using transit to commute. (Tables 5-6) Looking at data from 2013-2019 we see that bicycling remained relatively flat between 2013 and 2016. However, bicycle counts dropped significantly between 2016 and 2019. This drop is also reflected in census commute data.
And it wasn’t all Covid:
While 2022 data is anomalously low, it is also a continuation of a trend of declining bicycle use in Portland. Both annual count data and Census data demonstrates that bicycle use in Portland peaked in the 2013-2015 period and has been declining since.
Here is the report, here is one abbreviated source. Via Glenn Mercer.
Call me contrarian, but I have never been convinced that bicycles have a promising economic future in a truly Pigouvian city. And as a side point, how popular would bicycles be if they were embedded with software, requiring each bicycle to respect the law, stop at red lights, and so on?
Teaching and Learning Economics with the AIs
Tyler and I have a new paper, How to Learn and Teach Economics with Large Language Models, Including GPT:
GPTs, such as ChatGPT and Bing Chat, are capable of answering economics questions, solving specific economic models, creating exams, assisting with research, generating ideas, and enhancing writing, among other tasks. This paper highlights how these innovative tools differ from prior software and necessitate novel methods of interaction. By providing examples, tips, and guidance, we aim to optimize the use of GPTs and LLMs for learning and teaching economics effectively.
Most of the paper is about how to use GPTs effectively but we also make some substantive points that many people are missing:
GPTs are not simply a chatty front end to the internet. Some GPTs like ChatGPT have no ability to search the internet. Others, like Bing Chat, can search the internet and might do so to aid in answering a question, but that is not fundamentally how they work. It is possible to ask a GPT questions that no one has ever asked before. For example, we asked how Fred Flintstone was like Hamlet, and ChatGPT responded (in part):
Fred Flintstone and Hamlet are two vastly different characters from different time periods, cultures, and mediums of storytelling. It is difficult to draw direct comparisons between the two.
However, one possible point of similarity is that both characters face existential dilemmas and struggles with their sense of purpose and identity. Hamlet is plagued by doubts about his ability to avenge his father’s murder, and his own worthiness as a human being. Similarly, Fred Flintstone often grapples with his place in society and his ability to provide for his family and live up to his own expectations.
Not a bad answer for a silly question and one that (as far as we can tell) cannot be found on the internet.
GPTs have “read” or “absorbed” a great amount of text but that text isn’t stored in a database; instead the text was used to weight the billions of parameters in the neural net. It is thus possible to run a GPT on a powerful home computer. It would be very slow, since computing each word requires billions of calculations, but unlike storing the internet on your home computer, it is feasible to run a GPT on a home computer or even (fairly soon) on a mobile device.
GPTs work by predicting the next word in a sequence. If you hear the phrase “the Star-Spangled”, for example, you and a GPT might predict that the word “Banner” is likely to come next. This is what GPTs are doing but it would be a mistake to conclude that GPTs are simply “autocompletes” or even autocompletes on steroids.
Autocompletes are primarily statistical guesses based on previously asked questions. GPTs in contrast have some understanding (recall the as if modifier) of the meaning of words. Thus GPTs understand that Red, Green, and Blue are related concepts that King, Queen, Man and Woman are related in a specific way such that a woman cannot be a King. It also understands that fast and slow are related concepts, such that a car cannot be going fast and slow at the same time but can be fast and red and so forth. Thus GPTs are able to “autocomplete” sentences which have never been written before, as we described earlier.2 More generally, it seems likely that GPTs are building internal models to help them predict the next word in a sentence (e.g. Li et al. 2023).
The paper is a work in progress so comments are welcome.
Which political tweets do best?
Analyzing every tweet of all US senators holding office from 2013 to 2021 (861,104 tweets from 140 senators), we identify a psycholinguistic factor, greed communication [TC: basically accusing other people of greed], that robustly predicts increased approval (favorites) and reach (retweets). These effects persist when tested against diverse established psycholinguistic predictors of political content dissemination on social media and various other psycholinguistic variables. We further find that greed communication in the tweets of Democratic senators is associated with greater approval and retweeting compared to greed communication in the tweets of Republican senators, especially when those tweets also mention political outgroups.
That is from new research by Eric J. Mercadante, Jessica L. Tracy, and Friedrich M. Götz. Via David Lilienfeld.
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.
Income and happiness, revisited
Measures of well-being have often been found to rise with log (income). Kahneman and Deaton [Proc. Natl. Acad. Sci. U.S.A. 107, 16489–93 (2010)] reported an exception; a measure of emotional well-being (happiness) increased but then flattened somewhere between $60,000 and $90,000. In contrast, Killingsworth [Proc. Natl. Acad. Sci. U.S.A. 118, e2016976118 (2021)] observed a linear relation between happiness and log(income) in an experience-sampling study. We discovered in a joint reanalysis of the experience sampling data that the flattening pattern exists but is restricted to the least happy 20% of the population, and that complementary nonlinearities contribute to the overall linear-log relationship between happiness and income. We trace the discrepant results to the authors’ reliance on standard practices and assumptions of data analysis that should be questioned more often, although they are standard in social science.
That is from a recent collaboration by Matthew A. Killingsworth, Daniel Kahneman, and Barbara Mellers. Via the excellent Kevin Lewis. So if you’re rich, don’t be a sad sack! No need for that. And via Daniel Lippman, here is some Bloomberg coverage of the same.
Uh-oh, reverse Flynn effect edition
A reverse Flynn effect was found for composite ability scores with large US adult sample from 2006 to 2018 and 2011 to 2018.
Domain scores of matrix reasoning, letter and number series, verbal reasoning showed evidence of declining scores.
Three-dimensional rotation scores generally increased from 2011 to 2018.
Differences in ability scores were present regardless of age, education, or gender.
The steepest slopes occurred for ages 18–22 and lower levels of education.
That is from a new paper by Elizabeth M. Dworak, William Revelle, and David M. Condon. Alex T. comments.
Earnings Are Greater and Increasing in Occupations That Require Intellectual Tenacity
That is the title of a new paper by Christos Makridis, Louis Hickman, and Benjamin Manning, here is the abstract:
Automation and technology are rapidly disrupting the labor market. We investigated changes in the returns to occupational personality requirements—the ways of thinking, feeling, and behaving that enable success in a given occupation—and the resulting implications for organizational strategy. Using job incumbent ratings from the U.S. Department of Labor’s Occupational Information Network (O*NET), we identify two broad occupational personality requirements, which we label intellectual tenacity and social adjustment. Intellectual tenacity encompasses achievement/effort, persistence, initiative, analytical thinking, innovation, and independence. Social adjustment encompasses emotion regulation, concern for others, social orientation, cooperation, and stress tolerance. Both occupational personality requirements relate similarly to occupational employment growth between 2007 and 2019. However, among over 10 million respondents to the American Community Survey, jobs requiring intellectual tenacity pay higher wages—even controlling for occupational cognitive ability requirements—and the earnings premium grew over this 13-year period. Results are robust to controlling for education, demographics, and industry effects, suggesting that organizations should pay at least as much attention to personality in the hiring and retention process as skills.
Of course that is very much accord with some of the claims Daniel Gross and I make in our book on talent. Via the excellent Kevin Lewis.
What young people say is driving changes in their mental health
I think consistent with Matt's point here — you see young people identifying social media as a source of worsening mental health; but also emphasizing politics and national media.https://t.co/scOXEMNFjJ https://t.co/MGgdNIsz0m pic.twitter.com/hY47EOZFAO
— Arpit Gupta (@arpitrage) March 1, 2023
I would stress this point. Most of the explanations for declining teen mental health invoke contagion at one stage of the argument or another. That in turn means the initial causes can be fairly small, relative to the final outcome. Contagion itself is arguably the most important cause.