Clay Collins

Are credit scores gender-neutral? Evidence of mis-calibration from alternative and traditional borrowing data This study investigates whether credit scoring systems inherently disadvantage women within the subprime borrowing context, where alternative credit d…

Forthcoming in JBEF: “Are credit scores gender-neutral? Evidence of mis-calibration from alternative and traditional borrowing data“ by Z. Liu and H. Liang. www.sciencedirect.com/science/arti...

Frontline Union Army captains cut desertions and boosted cohesion through leading by example, earning postwar-wage gains and greater recognition, from @andyferrara.bsky.social, Christian Dippel, and Stephan Heblich https://www.nber.org/papers/w34057

Politics at Work (Forthcoming Article) - We study how individual political views shape firm behavior and labor market outcomes using new micro-data from Brazil. We first show that business owners are considerably more...

Forthcoming in the AER: "Politics at Work" by Emanuele Colonnelli, Valdemar Pinho Neto, and Edoardo Teso. www.aeaweb.org/articles?id=...

The Beautiful Dataset (Forthcoming Article) - Natural experiments, high stakes, expert subjects, knowledge of the precise objectives and exact rules, clean measurement, observability of strategies, incentives, actions and ...

Forthcoming in the JEL: "The Beautiful Dataset" by Ignacio Palacios-Huerta. www.aeaweb.org/articles?id=...

Work from Home and Disability Employment (Forthcoming Article) - There has been a dramatic rise in disability employment since the pandemic. At the same time, work from home (WFH) has risen four-fold. This paper asks whether the two are caus...

Forthcoming in AER: Insights: "Work from Home and Disability Employment" by Nicholas Bloom, Gordon B. Dahl, and Dan-Olof Rooth. www.aeaweb.org/articles?id=...

Distinguishing Causes of Neighborhood Racial Change: A Nearest Neighbor Design (Forthcoming Article) - We study neighborhood choice using a novel research design that contrasts the move rate of homeowners who receive a new differentrace neighbor immediately next-door versus slig...

Forthcoming in the AER: "Distinguishing Causes of Neighborhood Racial Change: A Nearest Neighbor Design" by Patrick Bayer, Marcus Casey, W. Ben McCartney, John Orellana-Li, and Calvin Zhang. www.aeaweb.org/articles?id=...

🚨NEW PAPER 🚨 Are police more right-wing and biased against marginalized groups than the general public? If so, why? My new article in
@pnas.org w/ @tylerreny.bsky.social, Newman, and Sears provides some answers. 🧵1/n

Big thanks to Dan Bernhardt and QREF, two anonymous referees who provided great feedback, Steve Salaga, Chenlong Ma, and James Zhang for their help on this.

If you like this paper, Sam is one of our brilliant PhD students in Sport Management. He has a lot of cool other projects he’s working on (including a Nature pub!) that he features on his LinkedIn. linkedin.com/in/sam-shuoyu-chen-833760177 Ivy is also a PhD student in Statistics.

Key takeaways for those who stuck around: skin tone algorithms like CASCo can be a great source for doing research related to racial biases or colorism. And Hupu is a really interesting data source for finding data on numerical consumer evaluations in sports.

This effect isn’t huge (in the biggest specification, a 2-unit difference in skin tone has a similar effect as a 1-unit change in game score- which can be one additional made free throw), but it’s incredibly persistent. Of course, there’s a bunch of nuances to this that I’ll leave the paper for.

It appears that Hupu users (who are almost exclusively Chinese) overwhelmingly believe that players with darker skin tones are of greater athletic ability than their lighter-skinned counterparts.

This even applies when we only include players with exceptional performances, when we up the sample to only include the starting rotation, and when we only include players who did not appear in the game. What’s driving this? Well we have a guess.

We run a ton of robustness checks and find the results hold when looking at the number of highest rating votes and lowest rating votes (darker skinned players receive more 10 votes, fewer 1 votes).

They also enjoy less of a boost when their team wins. Interacting skin tone with upset outcomes is a bit messy, but we generally find for darker-skin players, a small positive effect for upset losses and a small negative effect for upset wins (relatively symmetric).

We find that, controlling for player performance, players with darker skin tones are consistently rated higher than their lighter-skinned peers. Interacting skin tone with wins and losses finds that darker-skinned players are insulated slightly from the negative effect of losses.

CASCo sorted players into 13 different skin tone categories, with 1 being the lightest skin tone, and 13 being the darkest. This gives us a lot of variation. We use a straightforward linear regression to measure the effect of a players’ skin tone on their game rating. No fancy metrics here.

Classification Algorithm for Skin Color (CASCo): A new tool to measure skin color in social science research Objective A growing body of literature reveals that skin color has significant effects on people's income, health, education, and employment. However, the ways in which skin color has been measured ...

To avoid trying to hand-categorize skin tones, we use CASCo. A really cool algorithm that processes skin tone using images, and it’s relatively easy to use too. onlinelibrary.wiley.com/doi/10.1111/...

Notice I say “skin tone” instead of “race”. That’s because we investigate a colorism argument. It may not just be your race that affects fans’ opinions, but how dark or light your skin is within your race. This may matter, as each player’s headshot appears on the Hupu app.

This win-loss effect is quite big, and applies even to players who didn’t appear in a minute of the game. That’s cool, but it gets interesting when we look at the effect of skin tone.

We find that players with more votes are rated more positively and players who put up better performances (measured using Hollinger game score) are rated higher. Winning and losing a game is a big driver. Win- score goes up, lose- goes down.

Players can be scored an aggregate rating from 1 to 10. And players are voted on a lot. We find an average of 4,800 votes for each player, each game in 22/23- including players who never appeared in each game. Lots of cool data to work with.

But the problem has always been data. It’s hard to tell how fans rate players. Enter Hupu, a Chinese social media app that allows players to rate the performance of athletes. In the 22/23 NBA seasons, they could rate every player after every game.

A big question in sports: how do fans perceive player performances? Are stats all that matter? Do star players receive additional boosts or punishments? Is there a race effect where players of a certain skin tone are rated higher/lower? And we can make labor market extrapolations as well.

Discrimination and subjective player ratings: Evidence from China This paper provides an empirical examination of how fans perceive and evaluate athlete performances. Using NBA games during the 2022–23 season, this p…

First, a link to the paper. TLDR: cool data, cool approach, racial biases found. www.sciencedirect.com/science/arti...

New Paper Alert! Really proud to announce my paper with Sam Shuoyu Chen and @ivycollins.bsky.social‬, Discrimination and subjective player ratings: Evidence from China, has been published in The Quarterly Review of Economics and Finance. A thread.

Venue tax hike for new Spurs arena goes to Bexar County voters in November Bexar County voters will have their first — and likely only — chance to vote on a new Spurs arena in the Nov. 4 election.

As a researcher I'm excited for another sports facility vote. The city should have one too. The hotel tax IS a tax on residents. City+county officials think raising hotel tax rates will attract more visitors. Nothing says "Visit San Antonio" like a 17% hotel tax.
www.ksat.com/news/local/2...

DC council okays record-shattering $6.6B+ stadium subsidy for Commanders in exchange for local hiring promise The D.C. city council voted on Mayor Muriel Bowser's Washington Commanders stadium plan on Friday, and it went as expected, approving the revised agreement 9-3. Needing five votes to block a two-third...

For a team with no other tangible offers in place from other local governments, and a site that was a potential gold mine for DC but will now be in Commanders owner Josh Harris's control on a 90-year sweetheart lease, $6.6B is a pretty incredible get.

The National Association for Business Economics (NABE) strongly condemns the baseless removal of Bureau of Labor Statistics (BLS) Commissioner Erika McEntarfer and the unfounded accusations leveled against the work of the agency. This unprecedented attack on the U.S. statistical system threatens the long-standing credibility of our economic data infrastructure.

NABE strongly condemns the baseless
removal of BLS Commissioner McEntarfer & the unfounded accusations leveled against the work of the agency. This unprecedented attack on the US statistical system threatens the long-standing credibility of our data

files.constantcontact.com/668faa28001/...

Passing on the flame: do mega sports events promote health behaviours?: Krekel, Christian; Kavetsos, Georgios; Ziebarth, Nicolas R.