Monday, October 5, 2015

Basic Textual Analysis of EPA Regulations

In the last post I showed the relationship between the majority party in the House, Senate, and Presidential office and EPA regulatory restrictions on agriculture from 1974 to 2014. As I noted there, the relationships aren't always clear and they definitely aren't what a "naive" understanding of the current political divide would predict. In this post I thought I'd take a step back and provide a more general summary of EPA regulations, without the focus on politics and agriculture.

The RegData 2.2 database provides highly detailed data on regulation at the federal level. Cutting it down to just EPA regulation yields 11,353 observations. Each observation gives a number of restrictions and word count for a given Part in a given Title in a given year in the Code of Federal Regulations. Titles are divided into Chapters, Subchapters, and Parts, but RegData doesn't split the data up by Chapter or Subchapter.

The Title designation basically tells you which agency you're looking at, though agencies like the EPA can be found in several Titles. The primary Title for the EPA is 40. Parts divide the Titles into subject areas. For instance, Title 40, Part 171 includes information about the certification of pesticide applicators. Part 406 details effluent guidelines for grain mills.

The chart below shows the number of Parts in Title 40 from 1974 to 2014. There are more than 3 times as many Parts now as there were in 1974. Most of the growth in Parts since 1974 occurred before the mid-1990s. If you look at my previous post, the number of parts generally tracks with the number of restrictions on agriculture. I don't know whether this is a coincidence or not.

Digging deeper, the chart below shows the total number of EPA regulatory restrictions and word count in the Code of Federal Regulations. The regulations are not restricted to any certain industry regulated by the EPA. Word count is fairly straightforward, but regulatory restrictions certainly isn't. A restriction is an instance of words like "must," "shall," "cannot," etc in the Code of Federal Regulations.

Regulatory restrictions (left axis) and word count (right axis) both grow consistently over time, which is quite different from the general trend in ag regulations by the EPA. Trend growth in EPA regulations is about 3,317 restrictions per year. The two series grow at similar rates indicating that the EPA isn't putting a lot more or fewer restrictions relative to the number of total words.

I don't know what any of this implies other than the EPA is writing a lot more regulations than they did 5, 10, and 40 years ago. I think there's a possibility that the growth in Parts per Title might be a decent proxy for growth in the scope of regulation while the growth in restrictions by a particular agency is a proxy for growth in the intensity of regulation. I'm sure the people at RegData have thought of this and maybe I just need to dig into their literature more. It would be interesting to see if there's a way to separate those two concepts with these data. Their project is opening up all sorts of new avenues of research; it's exciting stuff!

If there's any specific regulatory agency you're interested in, please feel free to drop a request (along with your thoughts) in the comments below and I'll see about summarizing some other data for you. In the meantime, expect more on the EPA and some posts on USDA regs as well.


  1. This is very cool stuff! It really makes me wonder what is driving the increase in regulation. Is it an increase in demand for regulation or supply (framing it as Stigler/Peltzman might)?

    1. You bring up a good point, Dallas. In the research I've completed and have planned, I want to focus more on the effects of regulation, which I think can say something about the drivers of it. Of course, it isn't definitive in that respect.