People don’t cherish the things that they obtained without a cost. Here’s an interesting research from the 2025 AI in Social Science Conference, by Jason Sockin from Cornell University. The paper’s not online yet, but an abstract can be found at NBER
Interviews (which seems to be a tentative title…?)
Elliott Ash, ETH Zurich. Soumitra Shukla, Harvard University. Jason A. Sockin, Cornell University
Interviews allow employers to learn about workers, but do they also enable workers to learn about firms? Using data on hundreds of thousands of interview reports from Glassdoor, we find that candidates for high-paying jobs are 9 percent more likely to reject a job offer if they believe the interview was easy.
Easy interviews appear to convey poor “fit’’ as those who accept offers after easy interviews are one-half of a standard deviation less satisfied with their jobs and 14 percent less likely to remain with their employer for at least one year. Analysis of interview narratives using a large language model reveals that difficult interviews signal colleague ability and opportunities for skill development, whereas easy interviews convey a nonselective process. Interviews offer workers a preview of match quality, highlighting a unique channel through which labor markets may become less efficient if firms fully automate their hiring processes with AI.
Apart from the main finding, one result from the paper is really interesting to look at:
The firms with the hardest interviews are dominated by consulting (McKinsey, BCG, Bain, Oliver Wyman, Kearney) and finance/tech (Bridgewater, Jane Street, D. E. Shaw, Akuna, AWS, Google, SpaceX). These industries set the bar high with selective and challenging processes.
The easiest interviews are concentrated in retail, food service, and insurance. Examples include Taco Bell, Pizza Hut, Dollar General, Family Dollar, Menards, Aflac, and Primerica.