I assume its for several reasons, though this is only my own hypothesis. First, EM is much more progressive about candidate and student evaluation than the other specialties of medicine. EM is still the only one that uses a Standardized LOR (the SLOE), and has so for over a decade. No other specialty, that I know of does this. EM was the first to adopt the SVI as well. So to an extent, EM as a specialty and the folks at CORD are always trying to improve the process. I’m not sure other specialties are nearly as progressive on this front.
The other issue is the SLOE. I love the SLOE, and I think it really makes candidate evaluation so much better, because it puts less importance on board scores and more importance on EM performance (which is what SHOULD matter). But if you worked in a field that didn’t have a SLOE, and everyone just had the same generic copy and paste LORs that really say little objective, then there isn’t much to evaluate in terms of deciding who to interview. You’d just use board scores and geography, and ERAS can do all that for you. So on day one, you could basically filter candidates down to the list you want to interview without ever actually looking at their applications. But in EM, where the SLOE is the most important part of the application, you have to open each app and review the SLOEs to know if you want to extend an invite. That takes considerable time. But it beats the alternative, which is just basically deciding strictly off board scores.