Friday, September 20, 2013

Job Post: Create the Future of Law

Do you love law, but hate bluebooking? Do you think we should be able to apply version control to law, but know better than to just dump the U.S. Code into git?

Then you might be the person for us. We are looking for a few strong programmers with an interest in and (hopefully) understanding of law. Web stack (html/css/js) + XML + document parsing experience are all big plusses. As is ambition, integrity and an itch to change the world.

Who are we?

Xcential has built legislative drafting and repository systems for more than a decade. In the last couple of years, we have taken on a number of exciting projects, working on legislation and codification projects from Hong Kong and Chile to the U.S. House. What you see online about our work is just the tip of the iceberg.

Interested? Send a resume or a link to me at ari.hershowitz at xcential dot com. Or tweet me your interest (@arihersh) and a link to something interesting about you.

Monday, September 9, 2013

From 1789 to #OpenData2013

The First U.S. Law
I'm looking forward to the Data Transparency conference today in Washington, DC, organized by +Hudson Hollister. You already know this if you've been following this blog, or my tweets (@arihersh), or have had the good fortune to speak with me about legislative data recently.

I will be presenting on [participating in] a panel on legislative data, moderated by +The Cato Institute's Jim Harper ( says: if you like this blog, you might like Jim's Open Bills project).

The core of my presentation will be the proposal from my previous post for Congress to launch Operation Clean Desk. Here, I want to give some morsels of historical perspective on the codification process.

The first law enacted after adoption of the Constitution was "An Act to regulate the time and manner of administrating certain Oaths" in 1789. (Check wikipedia or the Statutes At Large, 1789-1799, page 23, if you don't believe me.) Parts of that Act still survive in Title 2, Sections § 21, § 22, and § 25, and Title 4, § 101 and § 102, of the United States Code.

If you take a look at section 25, for example, it is drawn from Chapter 1, section 2 of that first Act, and was amended by an act in 1948. If Congress wants to amend it again (say to allow the Oath of Office to be administered electronically), it would likely have to cross-reference that original act, and the 1948 amendment. That is true even though in practice the law is almost always referred to as 2 U.S.C. § 2. So in effect you would have 3 references in the amendment (1789, 1948 and U.S. Code). By contrast, if Congress wanted to amend Title 4, § 101, it could do so directly without reference to the earlier act. Why? Because Title 4 has been enacted by Congress. How do I know?  Look for the asterisk in this list of U.S. Code titles.

There have been many attempts at codification of U.S. law over the years, starting with the first official codification, the Revised Statutes of the United States (1876, wikipedia link). The U.S. Code structure was first established by Congress in 1926 and the Law Revision Counsel-- an official office to compile the code and prepare codification bills, was established in 1974 (2 USC 285). In nearly 100 years since the Code was first created, about half of the titles have been enacted.

There are currently 8 codification projects prepared or in preparation for Congressional enactment. Technology, and political will, can accelerate this historical project. That would be a tremendous legacy for Congress and a great step forward for government transparency.

Friday, September 6, 2013

Data Transparency 2013: Operation Clean Desk

Mayor John Dore, Seattle Municipal Archives
I've been meaning to organize these papers
I have been invited to speak [join a breakout panel on legislation] at the Data Transparency conference (2013) in Washington, D.C. next week. My message will be quite simple so that everyone can then focus on the other, very impressive, speakers: Now is the time for Congress to launch Operation Clean Desk.

Cleaning up

As I have discussed before, the process of legal drafting has built up a lot of cruft over the years. Amendments on top of amendments on top of original Acts. Even politicians who want to reduce the size of government end up doing so by adding to this growing body of overlapping law.

Congress saw the fundamental solution to this problem years ago, and created the Office of the Law Revision Counsel, which prepares the U.S. Code (see 2 U.S.C. 285). This is a single, clean, compilation of most of U.S. Federal public law. But until Congress formally enacts the Titles of the Code, these compilations live in parallel with the growing sea of historical laws. The LRC also compiles draft bills that would enact individual Code titles.  This process has been called positive law codification, though some (e.g. +Harlan Yu ) have proposed to refer more directly to "enactment of U.S. Code titles". Passing these codification bills would be step 1 of Operation Clean Desk.

Step 2 is related: make adjustments to Congressional drafting workflow, so that new laws fit more directly into the existing structure of the U.S. Code. This process is already underway: the Office of Legislative Counsel is well aware of the value of drafting consistency and the new House Rules have focused on consistency and transparency. For example, citations in a bill should make direct references to U.S. Code sections wherever possible. There are also terrific drafting guidelines (e.g. pdf from House Legislative Counsel, 1995) about the language, form and structure of amendments.

These drafting guidelines can be updated in practice to work even better with structured data formats (e.g. the U.S. Code in XML, announced by the Speaker's office earlier this summer). I don't mean to prescribe the details of drafting rules here; there are a lot of great minds at the Office of Legislative Counsel and other places in government, who work on this.  Some ideas are captured in +Grant Vergottini's blog post on variations of how to represent amendments, including the redlining model used in California. Other ideas (e.g. plain language drafting) can be brought to bear. What I recommend here, is that Congress should give priority to a further standardization of the drafting process itself.

Now is the time

I hope you will agree that these are good ideas, albeit not really new ideas. What is new is the timing. The speakers list at the Data Transparency conference suggests why. While no one might have objected before, now there are key leaders in Congress who genuinely care about the technical details. Operation Clean Desk will not get them a lot of press or a lot of invitations to the Sunday talk shows. But they know it needs to happen in order to facilitate a common agenda of transparency and government efficiency.

The low profile nature of the LRC's codification bills has often also made them low priority. But these bills are extremely important for the project of government transparency and they need a push. I believe that the high level participants in the Data Coalition conference can come together to give these bills the priority in Congress that they deserve-- despite the lack of lobbying or special interest pressure.

At the same time, Congressional offices can be drawn in to the project of drafting standardization. I envision a combination of drafting technologies and education of Congressional staff.

Jeffrey Beal's desk (CC license)
I know we have a law on that...just give me a moment
This two step process defines Operation Clean Desk for me. And there is no time like the first Data Transparency conference to begin this work.

I will be in good company at the conference: House Speaker John Boehner (R), Congressman Darrell Issa (R), Senator John Warner (D), and many others from both sides of the aisle. The conference keynote will be given by U.S. Deputy CTO Nick Sinai.