Category: Data Source

Jason Abaluck writes me about masks and the Bangladesh RCT study

This is all him, no double indent though:

“As a regular reader of your blog and one of the PIs of the Bangladesh Mask RCT (now in press at Science), I was surprised to see your claim that, “With more data transparency, it does not seem to be holding up very well”:

  1. The article you linked claims, in agreement with our study, that our intervention led to a roughly 10% reduction in symptomatic seropositivity (going from 12% to 41% of the population masked). Taking this estimate at face value, going from no one masked to everyone masked would imply a considerably larger effect. Additionally:
    1. We see a similar – but more precisely estimated – proportionate reduction in Covid symptoms [95% CI: 7-17%] (pre-registered), corresponding to ~1,500 individuals with Covid symptoms prevented
    2. We see larger proportionate drops in symptomatic seropositivity and Covid in villages where mask-use increased by more (not pre-registered), with the effect size roughly matching our main result

The naïve linear IV estimate would be a 33% reduction in Covid from universal masking. People underwhelmed by the absolute number of cases prevented need to ask, what did you expect if masks are as effective as the observational literature suggests? I see our results as on the low end of these estimates, and this is precisely what we powered the study to detect.

  1. Let’s distinguish between:
    1. The absolute reduction in raw consenting symptomatic seropositives (20 cases prevented)
    2. The absolute reduction in the proportion of consenting symptomatic seropositives (0.08 percentage points, or 105 cases prevented)
    3. The relative reduction in the proportion of consenting symptomatic seropositives (9.5% in cases)

Ben Recht advocates analyzing a) – the difference in means not controlling for population. This is not the specification we pre-registered, as it will have less power due to random fluctuations in population (and indeed, the difference in raw symptomatic seropositives overlooks the fact that the treatment population was larger – there are more people possibly ill!). Fixating on this specification in lieu of our pre-registered one (for which we powered the study) is reverse p-hacking.

RE: b) vs. c), we find a result of almost identical significance in a linear model, suggesting the same proportionate reduction if we divide the coefficient by the base rate. We believe the relative reduction in c) is more externally valid, as it is difficult to write down a structural pandemic model where masks lead to an absolute reduction in Covid regardless of the base rate (and the absolute number in b) is a function of the consent rate in our study).

  1. It is certainly true that survey response bias is a potential concern. We have repeatedly acknowledged this shortcoming of any real-world RCT evaluating masks (that respondents cannot be blinded). The direction of the bias is unclear — individuals might be more attuned to symptoms in the treatment group. We conduct many robustness checks in the paper. We have now obtained funding to replicate the entire study and collect blood spots from symptomatic and non-symptomatic individuals to partially mitigate this bias (we will still need to check for balance in blood consent rates with respect to observables, as we do in the current study).
  1. We do not say that surgical masks work better than cloth masks. What we say is that the evidence in favor of surgical masks is more robust. We find an effect on symptomatic seropositivity regardless of whether we drop or impute missing values for non-consenters, while the effect of cloth masks on symptomatic seropositivity depends on how we do this imputation. We find robust effects on symptoms for both types of masks.

I agree with you that our study identifies only the medium-term impact of our intervention, and there are critically important policy questions about the long-term equilibrium impact of masking, as well as how the costs and benefits scale for people of different ages and vaccination statuses.”

Your political views are not your own

In a unique sample of 394 adoptive and biological families with offspring more than 30 years old, biometric modeling revealed significant evidence for genetic and nongenetic transmission from both parents for the majority of seven political-attitude phenotypes. We found the largest genetic effects for religiousness and social liberalism, whereas the largest influence of parental environment was seen for political orientation and egalitarianism. Together, these findings indicate that genes, environment, and the gene–environment correlation all contribute significantly to sociopolitical attitudes held in adulthood, and the etiology and development of those attitudes may be more important than ever in today’s rapidly changing sociopolitical landscape.

Here is the full piece from Emily A. Willoughby,  Via the excellent Kevin Lewis.

India fact of the day

India’s most recent National Family Health Survey, which is conducted every five years by the Health Ministry, was released Wednesday and showed the total fertility rate (TFR) across India dropping to 2.0 in 2019-2021, compared with 2.2 in 2015-2016. A country with a TFR of 2.1, known as the replacement rate, would maintain a stable population over time; a lower TFR means the population would decrease in the absence of other factors, such as immigration…

In cities across India — as in other countries — women are opting for fewer children: The urban fertility rate is 1.6.

And the bias against baby girls is diminishing.  Here is the full story.  Via Naveen.

The anatomy of gender discrimination

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

Maybe the men, on average, did have greater ambition and thus promotion potential. One reason could be that women, on average, spend more time at home raising children than men. For very demanding executive jobs, even a small difference in time and travel availability could make a big difference in job performance.

And yet even if that’s the case, there could still be a discrimination problem. Even if women and men differ on average, there is a probability distribution for each group, and those distributions usually will overlap. That is, there will be many women who are willing and able to meet any workplace standard thrown at them, and many men with limited ambition.

If you think men and women are different on average, the unfairness can become all the more severe for the potential top performers. In this context, employers will look at the most talented women and, for reasons of stereotyping, dramatically underestimate their potential, including for leadership positions.

Economic reasoning suggests another subtle effect at play. Promotion to the top involves a series of steps along a career ladder, often many steps. If there is a discrimination “tax” at each step, even if only a small one, those taxes can produce a discouraging effect. It resembles the old problem of the medieval river that has too many tolls on it, levied by too many independent principalities. The net effect can be to make the river too costly to traverse, even if each prince is taking only a small amount.

With a citation to Zaua further below!

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.

Is the growing geographic concentration of innovation flattening out?

U.S. invention has become increasingly concentrated around major tech centers since the 1970s, with implications for how much cities across the country share in concomitant local benefits. Is invention becoming a winner-takes-all race? We explore the rising spatial concentration of patents and identify an underlying stability in their distribution. Software patents have exploded to account for about half of patents today, and these patents are highly concentrated in tech centers. Tech centers also account for a growing share of non-software patents, but the reallocation, by contrast, is entirely from the five largest population centers in 1980. Non-software patenting is stable for most cities, with anchor tenants like universities playing important roles, suggesting the growing concentration of invention may be nearing its end. Immigrant inventors and new businesses aided in the spatial transformation.

That is new research by Brad Chattergoon and William R. Kerr.

Legal Systems and Economic Performance in Colonial Shanghai, 1903-1934

Abstract: How important are legal systems to economic performance? To address this question, I focus on a historical period from colonial Shanghai, where quite different legal systems operated in the International Settlement andFrench Concession. In particular, employing novel historical data, I examine 1903–1934 land value discontinuities at the border between these Settlements. Substantial discontinuities were found in the 1900s, with higher land values associated with the International Settlement. However, by the 1930s, this land value advantage of the International Settlement had disappeared. A closer look at the institutions reveals that the French Concession adapted its operation to be more business friendly, under competition from the neighboring International Settlement. This suggests that the French legal system per se was not a barrier to economic growth, but rather it could function well if interpreted and implemented properly. This paper thus adds to evidence that formal legal system is not a key determinant of economic performance.

That is from Mingxi Li, who is on the job market this year from UC Davis.

The gender gap in preferences

This is taken from new work by Ángel Cuevas, Rubén Cuevas, Klaus Desmet, and Ignacio Ortuño-Ortín.  Here is the abstract:

This paper uses information on the frequency of 45,397 Facebook interests to study how the difference in preferences between men and women changes with a country’s degree of gender equality. For preference dimensions that are systematically biased toward the same gender across the globe, differences between men and women are larger in more gender-equal countries. In contrast, for preference dimensions with a gender bias that varies across countries, the opposite holds. This finding takes an important step toward reconciling evolutionary psychology and social role theory as they relate to gender.

Here is a bit more:

Our premise is that innately gender-specific interests should mostly conform to evolutionary psychology theory, whereas other interests should mostly conform to social role theory. We find strong evidence consistent with this premise.

And some detail on the categories:

We say that an interest is gender-related if it displays a systematic bias toward the same gender across the globe. More specifically, if in more than 90% of countries an interest is more prevalent among the same gender, then we refer to it as gender-related. For example, “cosmetics” and “motherhood” are universally more common among women, whereas “motorcycles” and “Lionel Messi” are universally more common among men. Conversely, we say that an interest is non-gender-related if its gender bias varies across countries. More specifically, if an interest is more common among men in at least 30% of countries and more common among women in at least another 30% of countries, then we refer to it as nongender-related. For example, “world heritage site” and “physical fitness” do not display a systematic gender bias across the globe.

And indeed everything works out as one ought to expect.  In the more gender-equal countries, men have “more male” interests, and the women have “more female” interests.  But for the less gender-specific interests, greater equality ends up resulting.  As for magnitude:

the standardized β is 30% when taking 9 dimensions, meaning that a one standard deviation increase in gender equality increases the difference in preferences between men and women by 30% of its standard deviation. The corresponding standardized β when taking 68 dimensions is 19%. Overall, the evidence points to a positive relation between gender equality and the difference in interests between men and women.

Hope you all are interested in this one!

The”hot hand” depends on location

Here is new research by Robert M. Lantis and Erik T. Nesson:

Do basketball players exhibit a hot hand? Results from controlled shooting situations suggest the answer is yes, while results from in-game shooting are mixed. Are the differing results because a hot hand is only present in similar shots or because testing for the hot hand in game situations is difficult? Combining repeated shots in a location and shots across locations, the NBA 3-Point Contests mimics game situations without many of the confounding factors. Using data on the 1986-2019 contests, we find a hot hand, but only within shot locations. Shooting streaks increase a hot hand only if a player makes his previous shot and only within locations. Even making three shots in a row has no effect on making the next shot if a player moves locations. Our results suggest that any hot hand in basketball is only present in extremely similar shooting situations and likely not in the run-of-play.

This YouTube video, of Stephen Curry, is one of the greatest videos of all time.

Regional personality variation across the United States

Here is a fun piece for Bloomberg, I would say you should take Five Factor personality theory somehow seriously, but not too seriously.  Excerpt:

Let’s consider extraversion. The least extroverted states in the country are Maine, Washington and Oregon, which fits my stereotype that a disproportionate number of the residents of those states are seeking some kind of isolation. Wisconsin has the most extroverted population, with Illinois, Iowa and Nebraska coming in next. The Midwest seems to be a friendly and outgoing place. That is to me also no huge surprise, though I would not have picked Wisconsin to be No. 1. The southern states come in at about average, while New Mexico, Nevada, Vermont and Montana do not measure as very extroverted, relative to the rest of the country.


The data on conscientiousness run counter to stereotype. My expectation was that the Midwest would win out here, but the Southeast ranks the highest. Coastal California fares poorly, as do scattered parts of the Midwest and West, again non-obvious or unexpected results. If it restores your faith in stereotypes, the area surrounding New Orleans, perhaps the most licentious city in the South, also rates low in conscientiousness.

Overall, the two strongest correlates of conscientiousness were Republican share of the vote, and share of married individuals in the population.

When it comes to emotional stability, fans of “The Sopranos” or “Seinfeld” will not be surprised: The Northeast, stretching down through Appalachia, ranks the lowest by a noticeable amount. There’s a reason George Costanza and Tony Soprano fit right in.

Mobile internet and political polarization

So far this paper is my favorite of the job market papers I have seen this year, and it is by Nikita Melnikov of Princeton.  Please do read each and every sentence of the abstract carefully, as each and every sentence offers interesting and substantive content:

How has mobile internet affected political polarization in the United States? Using Gallup Daily Poll data covering 1,765,114 individuals in 31,499 ZIP codes between 2008 and 2017, I show that, after gaining access to 3G internet, Democratic voters became more liberal in their political views and increased their support for Democratic congressional candidates and policy priorities, while Republican voters shifted in the opposite direction. This increase in polarization largely did not take place among social media users. Instead, following the arrival of 3G, active internet and social media users from both parties became more pro-Democratic, whereas
less-active users became more pro-Republican. This divergence is partly driven by differences in news consumption between the two groups: after the arrival of 3G, active internet users decreased their consumption of Fox News, increased their consumption of CNN, and increased their political knowledge. Polarization also increased due to a political realignment of voters: wealthy, well-educated people became more liberal; poor, uneducated people—more conservative.

My read of these results (not the author’s to be clear!) is that the mobile internet polarized the Left, but not so much the Right.  What polarized the Right was…the polarization of the Left, and not the mobile internet.

And please do note this sentence: “This increase in polarization largely did not take place among social media users.”  It seems that on-line versions of older school media did a lot of the work.

Here are further papers by Melnikov.

Cognitive sentences to ponder

The main conclusions are that, although children’s intelligence relative to their peers remains associated with social class, the association may have weakened recently, mainly because the average intelligence in the highest-status classes may have moved closer to the mean.

Here is the paper by Lindsay Paterson, British data, and for the pointer I thank the excellent Kevin Lewis.

Daddy’s girl?

That is the title of a new paper by Maddelena Ronchi and Nina Smith:

We study the role of managers’ gender attitudes in shaping gender inequality within the workplace. Using Danish registry data, we exploit the birth of a daughter as opposed to a son as a plausibly exogeneous shock to male managers’ gender attitudes and compare within-firm changes in women’s labor outcomes depending on the manager’s newborn gender. We find that women’s relative earnings and employment increase by 4.4% and 2.9% respectively following the birth of the manager’s first daughter. These effects are driven by an increase in managers’ propensity to replace male workers by hiring women with comparable education, hours worked, and earnings. In line with managers’ ability to substitute men with comparable women, we do not detect any significant effect on firm performance. Finally, we find evidence of rapid behavioral responses which intensify over time, suggesting that both salience and direct exposure to themes of gender equality contribute to our results.

Note that Ronchi is on the job market this year.  Via Jennifer Doleac.  I am not sure of the last part of that last sentence (are other mechanisms possible?  Maybe these fathers simply become better at understanding female talent?), but still this is an interesting result.