Hate speech posted on economics website is traced to leading universities, research finds
Associated Press (archive.ph)
By Christopher Rugaber
2023-07-20 21:41:53GMT
WASHINGTON (AP) — Anonymous comments with racist, sexist and abusive messages that were posted for years on a jobs-related website for economists originated from numerous leading U.S. universities, according to research released Thursday.
Some economists have long condemned the website, Economics Job Market Rumors, for its toxic content. The site, known by its acronym EJMR, is run by an anonymous individual and is not connected to a university or other institution. That fact had fed speculation that those who posted hateful messages on it were mostly online cranks who might not be economists.
Yet the new research indicates that users of the website include individuals at top-tier colleges and universities, including Harvard, Stanford and the University of Chicago, and many others.
“Our analysis reveals that the users who post on EJMR are predominantly economists, including those working in the upper echelons of academia, government, and the private sector,” the paper concluded. It was written by Florian Ederer, a management professor at Boston University, Paul Goldsmith-Pinkham, a finance professor at the Yale School of Management, and Kyle Jensen, an associate dean at Yale.
A spokeswoman for Harvard declined to comment. Stanford and the University of Chicago did not immediately respond to requests for comment.
“It’s not just a few bad apples,” Ederer said in a presentation Thursday at a conference sponsored by the National Bureau of Economic Research in Cambridge, Massachusetts. “It’s very, very widespread. And the toxicity is widespread.”
The revelations have provoked debate on social media among economists about privacy, free speech and online abuse. Some economists, particularly women who have been attacked on the site, say they hope the revelations lead colleges and universities to investigate the postings. Others have expressed concern that the research could lead to a “witch hunt” among those who posted on the site.
Speaking in an interview with The Associated Press, Goldsmith-Pinkham sought to dispel those concerns, saying the group does not plan on “releasing anything identifying” individuals.
Nearly 2,000 people watched a livestream of the paper’s presentation Thursday on YouTube. That was far more than the 100 or so who watched other NBER presentations the same day, suggesting widespread interest in the topic among academic economists.
The bigoted content on the website makes women and nonwhite economists often feel unwelcome in a profession that is already struggling to diversify, Goldsmith-Pinkham said. Black Americans, for example, are more likely to earn a Ph.D. in mathematics or other social sciences than in economics.
“The idea that in an anonymous space, people behave in this way, it reflects pretty poorly on the profession,” Goldmsith-Pinkham said.
The researchers used publicly available data to determine the internet addresses for about two-thirds of the more than 7 million posts that have been made on the site since 2010. They classified about 10% of those posts as “toxic” because of their racist or sexist content. These posts included the use of racial slurs and assertions that women have smaller brains than men.
About 11% of the postings on EJMR, the researchers found, originated from among several hundred universities, including those they classified as the top 25 research universities. On average, 13% of the posts from universities were considered toxic.
“Things were WAY better when women were focused on rearing children and feeding their husbands,” said one post highlighted by the researchers.
“The biggest enemies of America are: Blks,” read another.
The site has drawn criticism since at least 2017, when Alice Wu, an undergraduate student at the University of California, Berkeley, wrote a paper highlighting the sexist nature of many of the postings on the site.
In response to her paper, Olivier Blanchard, a former chief economist at the International Monetary Fund and emeritus economics professor at MIT, called the website a “cesspool.” Blanchard added that the site had “become a breeding ground for personal attacks of an abusive kind.”
Anya Samek, an economics professor who was first attacked on the site in 2009, said the site persists because there has been no way to hold it accountable. She said she hopes the universities that are being identified as sources of some of the posts will take steps to prevent future abuse.
“I would like to see universities take some action to make sure there’s no hate speech online coming from their own offices,” Samek said.
Samek was a target of false accusations on the EJMR website after she was hired in 2009 by the University of Chicago’s top-tier economics department. She had earned her Ph.D. from Purdue University, which she said some posters on EJMR might not have considered prestigious enough.
She was subject to threats on the site in 2022, after she presented at an economics conference.
“It was a truly horrible experience,” she said.
Samek has since started a petition urging the American Economic Association to consider legal action against threatening and defamatory posts.
The economic association has started its own job board to provide an alternative to EJMR, though so far it hasn’t received much use.
---
EJMR thread about the paper:
Here's a copy of the paper:
PDF (archive.org)
Excerpt:
In this paper, we identify the scheme used to assign usernames for each post written by an anonymous user on EJMR. We show how the statistical properties of that algorithm do not anonymize posts, but instead allows the IP address from which each post was made to be determined with high probability.
To recover IP addresses from the observed usernames on EJMR, we employ a multi-step procedure. First, we develop GPU-based software to quickly compute the SHA-1 hashes used for the username allocation algorithm on EJMR. In total, we compute almost 9 quadrillion hashes to fully enumerate all possible IP combinations and to check which of the resulting substrings of hashes match the observed usernames. For each post, this roughly narrows the set of possible IP addresses from 232 to 216. Second, we measure which IP addresses occur particularly often in a narrow time window and use the uniformity property of the SHA-1 hash to test whether these IP addresses appear more often than would likely occur by chance.
Our statistical test is very conservative and minimizes the probability of falsely assigning an IP address to a post because the p-value thresholds we employ are of the order of approximately 10−11. For example, even though there are 7,098,111 posts on EJMR, we assign exactly zero posts to the large set of bogon IP addresses. Despite this very conservative approach our procedure recovers 47,630 distinct IP addresses of EJMR posters and assigns 66.1% of the roughly 7 million total posts to these IP addresses. We then describe aggregated features of posting behavior on EJMR. Based on the geographic content of the IP addresses we identify and the origin of the associated internet service providers, we show that posting on EJMR is pervasive and very common in academia. Over 10% of posts originate from universities including all top-ranked universities in the United States. A substantial number of posts also come from government agencies, companies, and non-profit organizations employing economists. However, the vast majority of EJMR posts comes from residential IP addresses located in the United States and in particular in cities with elite universities.
Like other online platforms posts on EJMR are very concentrated across posters. A mere 5% of IP addresses generate over 50% of posts and 20% generate more than 80% of posts and this concentration is even more pronounced for toxic posts.
---
I clipped out the presentation from the NBER youtube stream:
https://youtu.be/pFCApVPawI4?t=27354
PDF of slides from presentation (archive.org)
Associated Press (archive.ph)
By Christopher Rugaber
2023-07-20 21:41:53GMT
WASHINGTON (AP) — Anonymous comments with racist, sexist and abusive messages that were posted for years on a jobs-related website for economists originated from numerous leading U.S. universities, according to research released Thursday.
Some economists have long condemned the website, Economics Job Market Rumors, for its toxic content. The site, known by its acronym EJMR, is run by an anonymous individual and is not connected to a university or other institution. That fact had fed speculation that those who posted hateful messages on it were mostly online cranks who might not be economists.
Yet the new research indicates that users of the website include individuals at top-tier colleges and universities, including Harvard, Stanford and the University of Chicago, and many others.
“Our analysis reveals that the users who post on EJMR are predominantly economists, including those working in the upper echelons of academia, government, and the private sector,” the paper concluded. It was written by Florian Ederer, a management professor at Boston University, Paul Goldsmith-Pinkham, a finance professor at the Yale School of Management, and Kyle Jensen, an associate dean at Yale.
A spokeswoman for Harvard declined to comment. Stanford and the University of Chicago did not immediately respond to requests for comment.
“It’s not just a few bad apples,” Ederer said in a presentation Thursday at a conference sponsored by the National Bureau of Economic Research in Cambridge, Massachusetts. “It’s very, very widespread. And the toxicity is widespread.”
The revelations have provoked debate on social media among economists about privacy, free speech and online abuse. Some economists, particularly women who have been attacked on the site, say they hope the revelations lead colleges and universities to investigate the postings. Others have expressed concern that the research could lead to a “witch hunt” among those who posted on the site.
Speaking in an interview with The Associated Press, Goldsmith-Pinkham sought to dispel those concerns, saying the group does not plan on “releasing anything identifying” individuals.
Nearly 2,000 people watched a livestream of the paper’s presentation Thursday on YouTube. That was far more than the 100 or so who watched other NBER presentations the same day, suggesting widespread interest in the topic among academic economists.
The bigoted content on the website makes women and nonwhite economists often feel unwelcome in a profession that is already struggling to diversify, Goldsmith-Pinkham said. Black Americans, for example, are more likely to earn a Ph.D. in mathematics or other social sciences than in economics.
“The idea that in an anonymous space, people behave in this way, it reflects pretty poorly on the profession,” Goldmsith-Pinkham said.
The researchers used publicly available data to determine the internet addresses for about two-thirds of the more than 7 million posts that have been made on the site since 2010. They classified about 10% of those posts as “toxic” because of their racist or sexist content. These posts included the use of racial slurs and assertions that women have smaller brains than men.
About 11% of the postings on EJMR, the researchers found, originated from among several hundred universities, including those they classified as the top 25 research universities. On average, 13% of the posts from universities were considered toxic.
“Things were WAY better when women were focused on rearing children and feeding their husbands,” said one post highlighted by the researchers.
“The biggest enemies of America are: Blks,” read another.
The site has drawn criticism since at least 2017, when Alice Wu, an undergraduate student at the University of California, Berkeley, wrote a paper highlighting the sexist nature of many of the postings on the site.
In response to her paper, Olivier Blanchard, a former chief economist at the International Monetary Fund and emeritus economics professor at MIT, called the website a “cesspool.” Blanchard added that the site had “become a breeding ground for personal attacks of an abusive kind.”
Anya Samek, an economics professor who was first attacked on the site in 2009, said the site persists because there has been no way to hold it accountable. She said she hopes the universities that are being identified as sources of some of the posts will take steps to prevent future abuse.
“I would like to see universities take some action to make sure there’s no hate speech online coming from their own offices,” Samek said.
Samek was a target of false accusations on the EJMR website after she was hired in 2009 by the University of Chicago’s top-tier economics department. She had earned her Ph.D. from Purdue University, which she said some posters on EJMR might not have considered prestigious enough.
She was subject to threats on the site in 2022, after she presented at an economics conference.
“It was a truly horrible experience,” she said.
Samek has since started a petition urging the American Economic Association to consider legal action against threatening and defamatory posts.
The economic association has started its own job board to provide an alternative to EJMR, though so far it hasn’t received much use.
---
EJMR thread about the paper:
Here's a copy of the paper:
PDF (archive.org)
Excerpt:
In this paper, we identify the scheme used to assign usernames for each post written by an anonymous user on EJMR. We show how the statistical properties of that algorithm do not anonymize posts, but instead allows the IP address from which each post was made to be determined with high probability.
To recover IP addresses from the observed usernames on EJMR, we employ a multi-step procedure. First, we develop GPU-based software to quickly compute the SHA-1 hashes used for the username allocation algorithm on EJMR. In total, we compute almost 9 quadrillion hashes to fully enumerate all possible IP combinations and to check which of the resulting substrings of hashes match the observed usernames. For each post, this roughly narrows the set of possible IP addresses from 232 to 216. Second, we measure which IP addresses occur particularly often in a narrow time window and use the uniformity property of the SHA-1 hash to test whether these IP addresses appear more often than would likely occur by chance.
Our statistical test is very conservative and minimizes the probability of falsely assigning an IP address to a post because the p-value thresholds we employ are of the order of approximately 10−11. For example, even though there are 7,098,111 posts on EJMR, we assign exactly zero posts to the large set of bogon IP addresses. Despite this very conservative approach our procedure recovers 47,630 distinct IP addresses of EJMR posters and assigns 66.1% of the roughly 7 million total posts to these IP addresses. We then describe aggregated features of posting behavior on EJMR. Based on the geographic content of the IP addresses we identify and the origin of the associated internet service providers, we show that posting on EJMR is pervasive and very common in academia. Over 10% of posts originate from universities including all top-ranked universities in the United States. A substantial number of posts also come from government agencies, companies, and non-profit organizations employing economists. However, the vast majority of EJMR posts comes from residential IP addresses located in the United States and in particular in cities with elite universities.
Like other online platforms posts on EJMR are very concentrated across posters. A mere 5% of IP addresses generate over 50% of posts and 20% generate more than 80% of posts and this concentration is even more pronounced for toxic posts.
---
I clipped out the presentation from the NBER youtube stream:
https://youtu.be/pFCApVPawI4?t=27354
PDF of slides from presentation (archive.org)
Last edited:
