There are a lot of things they can do differently and I fully agree with the training complaints. I was THIIIIIIIIIIIS close to be a TFA member myself. But I knew that I just didn't have the heart for teaching like that. But I don't see how the turnover argument works because most students aren't taught by the same person from 1 grade to the next. If the TFA teachers gave those kids a solid base then they'll be better off.
Moreover, TFA tends to get the smartest people to do it and we all know that we've had teachers before where we thought they didn't know shyt. As far as KIPP goes, one of my friends here did that and I can't imagine a more passionate person than him. Dude came from nothing. Every firm we interview with the first question he asks is about their pro bono program. Don't take this as my glowingly endorsing TFA either, but I'd love to look at the study you're talking about.
Regarding the turnover, a lot of new teachers (TFA and regular teachers) admittedly struggle during their first year or two in the classroom and they don't get the hang of it until later. You need experience to improve your craft but if teachers are struggling early, then leaving after a few years, you'll have a constant supply of inexperienced teachers. That's just how I look at it but I know it partly depends on the individual and the amount of work/preparation that they put into it.
I'm not saying that anyone working with KIPP or other Charter Schools are bad people, just that their data can be misleading. Rubinstein addressed KIPP and their data more thoroughly
here and
here.
I didn't get too far but this link will take you to the study I was looking at:
Teach For America: A Review of the Evidence | National Education Policy Center
They refer to a few other studies of the effectiveness of TFA teachers so I'm going to try to read those too. Some reviews of these studies have come to completely different conclusions which is always interesting to me.
This article uses one of those other studies to criticize TFA's effects on students, while
this page on the Teach for All website uses the same study to celebrate TFA's impact on students.