Getting back into writing

I should post more often. Previous gaps in my writing were generally because I didn't have much to say. Over the past month or so, though, I have had quite a bit of things to say, and I have drafts of posts or brief notes to myself on my phone to show for it, I just didn't get around to writing up a blog post. So here's one of those drafts of a post, completed. Better late than never, right?

Recently I posted about a project I'm on at work, and how three years is probably the reasonable minimum time to complete a regulatory project like this - instituting new or changing existing safety regulations. Three years from when someone notices a problem that's the responsibility of this agency to when that problem gets fixed. And that's a minimum, assuming no serious research is required, assuming no major changes are made partway through to how the agency wants to handle it, etc. One of my first major projects here related to a certain kind of safety standard. This particular issue has got worse slowly but steadily since the 1960s. It was brought to peoples' attention by two accidents, in 2003 if I remember correctly. The project began soon after those, was well underway when I got here in 2008, and the last piece of it was supposed to be published today but it doesn't go into effect all at once.

Eight-plus years from accidents in which a dozen people died to enacting regulations intended to make that kind of accident less likely. That sounds like a long time, right? That's what I thought when I first started working here; everyone "knows" bureaucracy is slow, but it was striking and worrisome to see it up close. But you know, there's a good reason for the intermediate steps, and most of them are inevitable given American history in this area.

Where does a regulation come from? I don't know much about Congress beyond the "Schoolhouse Rock" version, but in this executive branch agency, it's a long process. Once someone in a position to set the agenda for the agency decides that a certain project would be a good idea, a regulatory team has to get the language of amending the Code of Federal Regulations just the way they want. That team includes at least five people and usually more: a tech writer, an economist, a lawyer, a regulatory development manager, and a subject matter expert, and there's usually more than one of the last. This is not their only responsibility during that time; most are assigned to multiple projects at once. They may often have to refer to their respective bosses, who may not see eye to eye with each other. Then there's the preamble, explaining and justifying what we're doing, which is usually more pages than the actual amendatory language. All this takes a couple months at least.

Once the team is happy with it, it goes to everyones' superiors, and then to other federal agencies for further review and approval, including legal review and economic review. That takes another few months. That is just to publish a notice of proposed rulemaking. The public and industry groups usually get several months to comment on that, as well as other government agencies that aren't officially in the review process. After the comment period, it goes back to the team and the process starts over for a final rule, which has to take into account the comments. Minor and/or easy regulations might take less time than that or be able to skip steps, but more high-priority ones will usually have extra steps or take longer on any of those. And if a project is minor enough to be easy, it's generally minor enough to get pushed aside in favor of higher priorities.

And the thing is, that's all inevitable, or damn near. The economic review during the reg-writing process makes it take a few weeks longer, but if it wasn't included, there's a good chance that other federal agencies later on would shoot the whole project down and send it back to square one entirely. Same for the lawyer. The subject matter expert isn't working at their discretion, they're conferring with other people and have to consider thousands of pages of current regulatory text, independent industry standards, and other factors. Most of the review and approval steps are required by law or executive order (for our purposes, the same thing), which were implemented in response to pressure over regulations that probably really did have too many unintended consequences or were bad ideas from the start.

Everyone wants safety regulations to be effective, harmless and quick. Unfortunately, you have to pick two at most, and you can't even be guaranteed of achieving both of the ones you aim for. (Why is that? In addition to all the obvious general reasons like human error and known unknowns, there are specific reasons as well. More on that later.)