Wow

... that we're able to match between the proposal, the answer, and what gets adopted or dropped.

That's around a quarter or a fifth of our rules in our sample did not get adopted.

Okay? So, and what we're doing in our analysis, this was a very long answer, sorry, to your question, but this is a very long way of saying that having done all this, what we do is we match the answers to these requests for comments and see, depending on the identity of the writer, based on the four groups we just discussed, if some groups are systematically having their preferences, as stated in their answers, reflected both at the proposal, adoption, or drop, uh, uh, scenarios of the, of the rulemaking process.

And, uh, in a nutshell, we find that at the proposal and the adoption levels, w- firms are more likely to have their interests reflected in the rules versus the average commenter.

The effect is stronger at the proposal level.

Something happens during the whole rulemaking process that the adoption level, that effect is still there, but slightly weaker.

Y- you're not going that far in your paper, but for Tom and me looking at this from the outside, it sort of looks like that, so.