Austin's Open Government Online Charter Amendment

There has been a petition filed to place on the May ballot an amendment to the City's charter that would force the city to put some of it's day-to-day business online. This charter amendment aims to bring accountability and transparency to the higher levels of municipal government, where the primary targets are City Council members, the City Manager, Assistant City Managers, division level managers and their respective staffs. The amendment requires some very specific conditions (I'll highlight the technical ones below) but first a disclaimer.

Although I currently work for the City of Austin, I am in no way representing or advocating *any* official position of the City in this matter. I am writing this as a concerned citizen of Austin who would like to see our government as transparent and efficient as possible and I am using information that is publicly available elsewhere to make my points.

With that in mind, this is a theoretical plan so assumptions have been made which could be incorrect. At this time, I am neither for nor against the proposed Charter amendment. My focus with this post is to try and bring a technical solution to the mandated goals of the Charter amendment in which to serve as a base for discussion with the Austin community. This opinion is put forward in the hopes it will facilitate discussion and will bring other concerned citizens into the mix and to showcase the alternative technical ways to satisfy the Charter amendment if passed. So please, do not read anything into this post that is not explicitly stated.

And with that, off we go. Here are the relevant sections (italics are direct quotes from the amendment; regular text is my opinion):

• The City must, as expeditiously as possible and to the greatest extent practical, make all public information available online in real time and accessible to the public.
  OPEN ACCESS TO CITY BUSINESS
  • Within one year of the date this Amendment takes effect, applications and proposals for any permit or contract of significant value must be provided to the City in an electronic format.
  • The City must maintain a system for electronic notification (such as email lists) to interested persons of any event or new information relating to the matter.
  • All written communications between the City and the applicant relating to the matter must be posted online in real time in a manner searchable by the public.
  OPEN ACCESS TO CITY CALENDARS
  • For all matters involving City business, the following people must maintain calendars of all meetings and maintain logs of all telephone calls: (a) City Councilmembers and their staff; (b) City Manager and his or her staff; (c) Assistant City Managers and their staff; and (d) all department heads.
  • Calendars and logs must be posted online in real time and be accessible to the public.
  • This provision must be implemented within six months of approval of this amendment.
  OPEN ACCESS TO CITY ELECTRONIC COMMUNICATIONS
  • In order to better preserve written electronic communication for public disclosure, the City must establish a system that automatically archives all incoming and outgoing electronic communication that deals with City business to and from the following people in their official capacity: (a) City Councilmembers and their staff; (b) City Manager and his or her staff; (c) Assistant City Managers and their staff; and (d) all department heads.
  • The above people are prohibited from discussing City business via any form of written electronic communication, such as a private email account, that is outside of the City’s automatic archiving system.
  OPEN ACCESS TO CITY FUNCTIONS
  • All public information that has previously been released to someone making a public information request and which, because of the nature of the subject matter, the City determines is or is likely to become the subject of a subsequent public information request for substantially the same information
  EFFECTIVE ACCESS TO INFORMATION
  • The City must create and maintain online tables of contents and indexes to enable the general public to easily find and access online City documents and public information. Information must be searchable, and be able to be located by author/submitter, individual recipient, date, and subject matter.

These are some pretty tough technical requirements but they are possible. If the city stays with its current technology (articles are located here, here, and the City's Cost Estimate) which is comprised primarily of proprietary, closed technology, this will be a *very* expensive endeavor. But, using Free Software and open standards, the city can put togther a long-term, cost-effective solution that would meet all of the above technical requirements.

The primary requirement of this plan is that all data will be created/stored/transported in XML formats. Once in XML, there are ways to work with the data that will satisfy the amendment.

• Jabber - This will be the core transport piece and the legacy application data connector. Jabber is essentially an XML router and can act as the "bridge" from older client/server applications to the charter mandated goals.
• Zimbra - This is the email/calendaring piece. At its heart, it's an XML processing engine.
• Zope/Plone - This provides the public presentation framework as well as the internal data workflow paths for data classification and dissemination.
• OpenOffice.org - This would be the city's standard productivity application for creating/editing data. All data would be stored in OpenDocument format which is basically a zipped XML file (well, actually, it's 3 files zipped into one). All citizen correspondence with the city would also be in OpenDocument format.

The Charter Data Flow diagram below shows a stylized data path and where the repositories reside in the model. It is important to remember this same model allows for a real-time and stored data repository that is controlled by the users of the system in the Plone workflows. This allows classification/declassification of data to be handled by the data custodians themselves, effectively streamlining the process. That very same system is also the frontend used by the citizens to gain access to the data. So let's break down the pieces one-by-one using the above amendment goals as an outline...

OPEN ACCESS TO CITY BUSINESS
• Within one year of the date this Amendment takes effect, applications and proposals for any permit or contract of significant value must be provided to the City in an electronic format.
• The City must maintain a system for electronic notification (such as email lists) to interested persons of any event or new information relating to the matter.
• All written communications between the City and the applicant relating to the matter must be posted online in real time in a manner searchable by the public.

If the city decides to commit to an open standards based format like OpenDocument immediately, the one year timeframe is a doable, although still a difficult timeline. As long as the city continues to accept binary, closed formats, it only pushes the timeframe of conversion and subsequent Charter amendment compliance futher into the future. OpenOffice is the perfect choice for an end-user tool as the city is not requiring the purchase of a particular company's software package to facilitate communication between its citizens and their government. Purchasing an office suite, even though it's the "market leader," puts an arbitrarily high hurdle in the way for those citizens who lack the funds to purchase the software. These disenfranchised citizens are now compelled to violate copyright laws in the attempt to communicate to their own government. With OpenOffice, the city could customize both the programs and templates required for effective communication. The city can create the exact functionality needed internally as well as customizing the functionality for its citizens externally while offering those tools to the public for no charge (for example, putting the software on a cdrom for those w/o Internet connections; creating live CDs that contain all the software and templates that can run on old, recycled machines, etc).

Further, to satisfy the electronic notification clause, a mail list management system is being integrated into Plone and there is already software available. Mailman is the most widely used mail list manager on the Internet and can be easily extended with just a bit of python code (Mailman, Zope, Plone all use python as the development language. Zimbra and Jabber can extend functionality using python so the city can standardize on one internal development language.)

And, as mentioned before, if the city standardizes on OpenDocument, incorporation of the contents of all submitted documents can be easily indexed once entered into the system (by entered I mean "saved to the storage medium.") Because OpenDocument files are just gzip'd text files, an index process can be used to unzip the files and pull just the text content (one file of the 3 contain just the content of the document w/no formatting) and make searching the repository more efficient and immediately available to the public.

OPEN ACCESS TO CITY CALENDARS
• For all matters involving City business, the following people must maintain calendars of all meetings and maintain logs of all telephone calls: (a) City Councilmembers and their staff; (b) City Manager and his or her staff; (c) Assistant City Managers and their staff; and (d) all department heads.
• Calendars and logs must be posted online in real time and be accessible to the public.
• This provision must be implemented within six months of approval of this amendment.

Access to city calendars through Zimbra should be facilitated quite easily. Either using replication within the system to provide a daily look in a downloadable format (like iCal) or the system can be setup to show real-time calendars on demand for key personnel. Public access would be facilitated through the Zope/Plone system while the actual data resides in the Zimbra Collaboration server. In regards to the phone logs, the members of government that the Charter applies to could be switched to a VoIP system in which the calls are stored in the repository with XML descriptions and metadata which would facilitate near real-time access to the actual phone calls for public review.

OPEN ACCESS TO CITY ELECTRONIC COMMUNICATIONS
• In order to better preserve written electronic communication for public disclosure, the City must establish a system that automatically archives all incoming and outgoing electronic communication that deals with City business to and from the following people in their official capacity: (a) City Councilmembers and their staff; (b) City Manager and his or her staff; (c) Assistant City Managers and their staff; and (d) all department heads.
• The above people are prohibited from discussing City business via any form of written electronic communication, such as a private email account, that is outside of the City’s automatic archiving system.

Once again, using the Zope/Plone products, the system can be designed in a way that all information is kept within the repository and using Plone workflows, access can be facilitated.

In respect to the prohibition of those City personnel using external means to facilitate communication, there is really no way to stop this unless it is made into a law that carried some form of legal consequence. Currently, the city has too many ways to bypass the restriction (information security staff can't block all webmail sites [common sense dictates that]; councilmembers and other staff targetted in the amendment can bring in their own personal laptops and attach to the public wireless network [reference here], etc) but I would propose that the staff in which this amendment is targeted be moved on to a centralized thin-client based system to allow for easier accountability. One way to approach this would be to provide these employees with wireless tablets/laptops that tie-in to a backend thin-client system where the installation of unapproved software and traffic can be monitored. With a centralized thin-client system in place along with a legal mechanism, the system could better provide the compliance and accountablity the Charter amendment strives to achieve.

OPEN ACCESS TO CITY FUNCTIONS
• All public information that has previously been released to someone making a public information request and which, because of the nature of the subject matter, the City determines is or is likely to become the subject of a subsequent public information request for substantially the same information

This goal is already an inherent part of the envisioned system. Plone workflows will allow for the classification of data so once an Open Records request is made and the results known, the data from that request can be entered into the system and classified as public data. It could also be tagged with metadata that would enable classification of the data into Open Record groups so subsequent searches could use the narrowed down datasets.

EFFECTIVE ACCESS TO INFORMATION
• The City must create and maintain online tables of contents and indexes to enable the general public to easily find and access online City documents and public information. Information must be searchable, and be able to be located by author/submitter, individual recipient, date, and subject matter.

Again, this goal is already envisioned within the Plone content management architecture. Metadata tags can be added at will to provide the criteria needed for efficient searches.

PROS & CONS

I think the Pros of this system are self-evident as it is basically a transparent interface into the people's government. Redundancy of data and repetition of action have been minimized in the current design making the system as efficient as possible for the lowest possible cost. There are also savings by using products such as OpenOffice which can function in multiple roles thoughout the data collection/generation process as well as the obvious benefit of no ongoing software licensing fees.

Also, this plan does not replace any existing or upcoming applications like those listed in the City's cost estimate. This plan works with those systems to bring a centralized and efficient workflow with data classification to the already existing data collection process.

But there are a few Cons of the proposed design:

• Cost - There would be a hefty upfront cost for hardware and existing data conversion. The current official estimate is $24 million one-time cost and $11 million ongoing. Without getting into the specifics of cost, I feel the official costs might be slightly high as I think there could be some savings in one-time consulting fees, software licensing costs and ongoing staffing (possibly as much as $6mil in the one-time category). But the costs to implement a system like this would be high, make no mistake about it. This nut can't be cracked cheaply.
  One thing that I'd like to bring up which isn't specifically called for in the Charter amendment is data conversion costs to the OpenDocument format. Although the Charter amendment does not call for all city data to be accessible online, I would like to mention that it would be time and money well-spent. Although I couldn't find an "official" conversion efficiency rate for Microsoft Office to OpenDocument, I have found numerous articles (here's one) hinting the conversion rates are quite high. My own testing of the few hundred Microsoft formatted documents on my hard drive showed about a 90% conversion rate (by that I mean 9 out of 10 did not complain about conversion and the docs I opened up looked formatted correctly).
  So working on the assumption there is an effective 85% conversion rate, if the city has 2.5 million documents, that leaves 375,000 documents that would need some form of "hand-holding" to convert. If one employee can convert 20 documents a day, they would get through the problem documents in 51.37 years. To bring this down into the timeframe required by the Charter amendment (just for argument's sake: 1 year), the city would contract close to 50 people for document conversion (50 x $35,000/yr for a temp = $1.75 million). To me, this is a relative bargain as the city would have the bulk of it's data in a non-proprietary, easily convertible and manipulative format.
  Now, there are much better people than I to figure out total costs for a plan like this (and I would love to work with someone on this, just email me) but suffice to say, the city's official $24mil estimate would be sufficient to cover the costs of this plan. The city's official cost estimate provides for new positions (which would be needed for more python programmers, etc) so I think the cost estimate is a good attempt to "herd cats" but could be just a bit high. The silver lining is that the city should be able to recoup some (not all) of the one-time costs over the long-term with the savings of having the data in machine manipulative formats, an efficient and defined data workflow path (enhancing employee productivity) and the ongoing savings in license fees for the proprietary software replacements.
• Momentum - By that I mean that getting any large organization to change takes time. There are sure to be groups/departments/managers that are hostile to these types of ideas. But the Charter amendment demands specific goals within specific timeframes so this is the best type of pressure for a government agency (external pressure is always the most efficient). Having worked in government for the bulk of my career, I can make a safe assumption that a plan this agressive will meet with much resistance internally. It may very well take upwards of the first proposed year just to convince the parties targetted by this amendment to agree to the changes of workflow. But with public pressure and hopefully, media coverage, a change like this can be successful.

So, in summary, I believe the City of Austin can achieve the goals set forth by the Charter amendment although some additional time will be needed to allow for such massive infrastructure changes. It's important to keep in mind this proposal is just a theoretical model and may not even be possible with the aforementioned products. I'm assuming quite a bit (like ZODB will scale or can be broken into manageable chunks, the city's internal departments agree on a common plan, etc) so please take this post as a suggestion and not a roadmap. My experience in working with and for different government agencies as well as with Free/Open Software leads me to believe that the city can satisfy the requirements of the Charter amendment in a timely fashion and can provide to the public the most transparent and efficient government possible.

As always, I can be reached through comments or by email at scott(dot)brown(at)opennetworks(dot)org.