Introducing NLRB Edge

An idea for a new labor law publication.

A few months ago, I became interested in whether it would be possible to track legal developments in an automated way using some relatively new technologies. A variety of publications, mostly aimed at legal practitioners, already track some of these developments. But these publications are expensive and appear to take a lot of work to produce. I thought that I may be able to provide better coverage than these publications while using less labor and charging a lower price.

To implement this idea, I trained my attention on the National Labor Relations Board (NLRB) for three main reasons: I used to work at the agency, I am a practicing labor lawyer, and the NLRB has a relatively closed universe of legal developments. The closed universe makes the agency easier to track than laws that are mostly developed in the courts and my experience in this area makes it so that I know what all needs to be tracked and how to get ahold of it.

In the process of thinking this through, I also realized that, if I paired this legal tracking with more broadly-interesting commentary and analysis about NLRB developments and the labor movement, I may be able to put together a publication that could attract a few thousand paying subscribers. NLRB Edge is my effort at doing this.

Who Am I?

I am Matt Bruenig. I have been a labor lawyer since 2014 and have worked for the National Labor Relations Board (NLRB), the Machinists Union, SEIU, UFCW, and UAW. I am a practicing labor lawyer (available for hire). I mostly represent individuals and labor unions at the NLRB, but I also do grievance/arbitration and contract bargaining.

Publicly, most people know me as the founder of People’s Policy Project, a think tank that promotes socialist and social democratic policy. At 3P, I aim to provide accessible policy analysis on technical topics, something I also aim to do with this publication.

The Database

To begin this project, I wrote software that could automatically build and update a comprehensive database of thirteen different types of documents that the NLRB releases publicly. The documents include all published and unpublished Board decisions, administrative law judge (ALJ) decisions, and regional election decisions. They also include all GC, OM, and (released) advice memos, as well as all appellate court briefs and 10(j) injunction filings, among other things.

At present, the database has 86,125 documents in it and is increasing in size every day.

The database itself is a tremendous research tool and would be especially useful for NLRB staffers, union representatives, and labor lawyers. The documents contained in it can be accessed as pdfs or text files and can be queried through full text search as well as things like case number, citation, case name, date, and document type.

To give an example of what kind of power this gives you, imagine that you wanted to find all of the ALJ cases that have applied the Stericycle standard for coercive work rules.

The first thing you might want to do is find the actual Stericycle decision. Searching all document names for the word “Stericycle” provides 31 results and includes everything from published and unpublished Board decisions, regional election decisions, ALJ decisions, injunctions, and advice memos.

Because we are looking for a specific Board decision, we can narrow the search to published Board decisions. This query gives us just six results and it is easy to quickly determine that the Stericycle decision on August 2, 2023 is the one we are looking for.

To find the ALJ decisions that have applied Stericycle, we can construct a new query that looks for the word “Stericycle” in any ALJ case decided after August 2, 2023. When we do that, we get these 14 ALJ cases.

When we click to open the text of the decisions, we see that they are exactly what we are looking for. For instance, this is from the most recent ALJ case that contains the word “Stericycle” in it.

You can do something like this using expensive legal research services like Westlaw and Lexis. But, when it comes to the NLRB at least, my database is more comprehensive than those services, is updated automatically, and is free (at least, for me). For now the database will remain exclusive to me, but I am open to ideas others may have about how to make it available to others, e.g. through some kind of arrangement with unions or law firms.

Legal Tracking

In addition to being a good research tool, the process of updating the database makes it possible to identify every new document that the NLRB puts out when it puts it out. This means that, without any direct labor from me, the software is able to generate a list of everything that comes out of the agency.

These new documents can and will be manually reviewed by me as part of doing the day-to-day legal tracking of the agency. But in doing this manual review, I will also feed the documents into Large Language Models (LLMs) such as GPT4 in order to get quick summaries that will make the process of legal tracking a lot faster.

I know there is some skepticism about this technology, but, in my experience, LLMs are very good at document summary, even if they are lacking in some ways as universal chatbots. In any event, for this publication, the LLMs will only be used to speed up document summary, especially less significant documents. They will not be used to summarize documents completely unsupervised.

Commentary, Analysis, and Answers

In addition to automated legal tracking, I also intend to provide commentary on labor law and organized labor more generally, which will hopefully make the publication interesting to a much broader audience of people who follow the labor movement and US politics. This will include providing in-depth analysis of recent developments and even some explanations of the basics of how the NLRB and labor law work. It will also hopefully include answering questions from readers about problems they may be having in their workplace or other things they may be curious about in the area of labor law.

I think my commentary will be unique in some respects due to my particular political perspective. I also think my analysis will just be better than much of what’s out there because it will be less shallow and less hedged.

This part of the content bundle will likely evolve over time as I get a sense of what people want to read.

Please Subscribe

As I get the publication off the ground and populate some initial content, everything will be free. At some point, I will probably implement a paywall for most, if not all, of the future content. But if I get enough paying subscribers even without a paywall, I may just keep everything free, as I have with People’s Policy Project.

If you are interested in supporting this experiment, please subscribe so that you can receive all of the new pieces via email. There are free and paid options for subscription, but I’d especially appreciate it if you would sign up as a paid subscriber so that I can gauge how much interest there actually is in this kind of publication and thus whether it is worth doing.