Monday, March 16, 2015

Almost Digital Congress: ToDo List

When fitness hero Jack Lalanne, then in his nineties, was asked about his sex life, he said that he and his wife had sex almost every night. "Almost Monday, almost Tuesday, almost Wednesday...". Similarly, if you ask me about progress in modernizing the legislative process, I would say that Congress is almost digital. Not bad, considering the age of the institution, but still leaves something to be desired.

The confluence of last week's panel at SXSW, organized by the Congressional Data Coalition, and this week's #Hack4Congress show we've come a long way in the last couple of years [1]. In this, and the next few posts, I'll discuss what needs to be done to eliminate the "almost" and reap the full benefits of a digital Congress.

My focus is on the key documents that Congress produces: bills, amendments and ultimately, laws. Whether your guide is the Library of Congress or Schoolhouse Rock, the process should be familiar. An introduced bill goes through a number of changes before it becomes law. This is a change-tracking problem and at each stage we need to know (1) the original document, (2) the proposed change and (3) how that proposed change will affect the original.

I was recently asked, for example, if it is possible to show an automated comparison of the impacts on the law of two patent bills now before Congress: Coons's STRONG Patent Act (S. 632) and Goodlatte's Innovation Act, H.R. 9 (114th). The short answer is no. As a pre-requisite, each bill needs to be available in machine-readable form. H.R. 9 is available in XML, but Coons's Act is currently available only in pdf, not text or XML.

As a second pre-requisite, we would need a tool to apply the language of the bill directly to the law which it amends, in this case Title 35 of the U.S. Code. Take the example of section 4 of Goodlatte's bill. It currently reads:

(a) Amendments.—Section 290 of title 35, United States Code, is amended—
(1) in the heading, by striking “suits” and inserting “suits; disclosure of interests”;
(2) by striking “The clerks” and inserting “(a) Notice Of Patent Suits.—The clerks”...

These changes would be displayed in the context of the target law (section 290 of Title 35) in redlining, e.g.:
§ 290 Notice of patent suits suits; disclosure of interests
The clerks (a) Notice Of Patent Suits.—The clerks of the courts of the United States, within one month after the filing of an action under this title shall give notice thereof...
I've mocked up these changes by hand above, but if this process were automated, it would be possible to show the impacts of this patent bill side-by-side with other proposed patent bills. The 1929 "Ramseyer Rule" requires this kind of redlining when a bill is reported out of committee, but that is often late in the process. Ideally, the textual impact of a bill could be seen as soon as it is introduced.

In California, my company, Xcential, has worked with the legislature to automate this kind of tracking in its "As Amends the Law" feature. For any bill before the Assembly, the public can see the additions and deletions that the bill would effect in the context of current law. You can see how this works with the bill AB 26 Medical cannabis. You can see the bill's impacts on the current law, as well as the changes between different versions of the bill as it is amended.

The other major change-tracking challenge is the application of amendments to bills. This is the aim of the Amendment Impact Program, which Xcential is working on as part of the U.S. House's Modernization Project. Progress on this project was reported by the Committee on House Administration and in House Legislative Counsel Sandra Strokoff's presentation at the 2014 Legislative Data Transparency Conference. [While I work on these projects for Xcential, the information in this blog all comes from public sources and my opinions here are my own.]

In total, to follow a bill from introduction to law, Congress needs to:
  1. Ensure bill transparency: 
  2.  Bill text in consistent machine-readable formats. Many bills, particularly in the Senate, are still only available in pdf until later in the process.
  3. Automate bill-to-law change tracking:
  4. Apply bills directly to the U.S. Code and to non-positive law statutes.
  5. Pass codification bills:
  6. Make all titles of the U.S. Code into "positive law". This is primarily a political challenge, which I have discussed previously.
  7. Ensure amendment transparency:
  8. Amendment text in consistent machine-readable formats.
  9. Automate amendment-to-bill change tracking:
  10. Now being tackled in the Amendment Impact Program, discussed above.
In future posts, I will discuss these goals in more detail: the benefits that each of them brings, the progress that has been made toward achieving them, and some of the challenges that remain.

[1] Almost two years ago, I declared that legislative data has a posse and since then, that posse has been getting results. Congress now publishes machine-readable forms of most bills and amendments. The U.S. Code is also digitized, and updates are made available on the Law Revision Counsel's website shortly after new bills are passed. These and other advances were highlighted in testimony submitted (pdf) by the Congressional Data Coalition to House appropriators.

The House Majority is making digital transparency as a priority, adding this section to the House Rules for the 114th Congress:
The Committee on House Administration, the Clerk, and other officers and officials of the House shall continue efforts to broaden the availability of legislative documents in machine readable formats in the One Hundred Fourteenth Congress in furtherance of the institutional priority of providing public availability and use of legislative information produced by the House and its committees.
Fortunately, more and more citizens are interested and educated about what it will take to fulfill this commitment.