Opening Government Data – the San Francisco Way

The City of San Francisco recently unveiled a one-stop clearinghouse for all sorts of data generated and maintained by city agencies. Its an exciting first step in an effort that is high on the agenda of San Francisco CIO Chris Vein and Mayor Gavin Newsom – a tangible and visible commitment to open source technologies, open government data and citizen engagement.

San Francisco is reportedly taking a different approach to encouraging the use of this data than has been tried in other jurisdictions, notably the District of Columbia. Instead of running a contest and offering cash prizes, San Francisco will explore the idea of an “App Store,” where developers can have their work highlighted and showcased. This is an interesting idea, and the time may be right for a new approach to encouraging developers to use data from governments.

As a participant in the first Apps for Democracy contest, I feel strongly that at the time it was the best approach to engaging developers and getting them excited about building applications using civic data provided by DC. It was so successful that it spawned a second Apps for Democracy competition, two similarly focused contests sponsored by the Sunlight Foundation and the forthcoming “Big Apps” competition in New York City. But I think San Francisco might be on to something here – the “development contest” paradigm for encouraging civic coders might not be as successful going forward as it was when DC first tried it.

For the open government data initiatives of cities like San Francisco, DC and New York to flourish long-term the motivating factors for developers to build applications have to be more sustained than can reasonably be accomplished within the confines of a “contest.” Contests end – local governments need mechanisms to encourage development with their data beyond the awards ceremony.

The notion of an Apps Store is an interesting one, with longer-term potential – I think it will resonate with developers and provide more sustained motivation for developers to build applications that use data from San Francisco agencies.

Looking forward to seeing this idea develop – and to (potentially) getting an app in the App Store.

Looking for Collaborators

I’m thinking about starting a project to develop an application that uses crime incident data from the City of San Francisco and allows people to identify crimes that occur close to their homes and/or places of business. I worked on a similar project for the original Apps for Democracy contest last year.

Crime incident data from the SF Police Department is available in KML format here. I’d like to take the data in these files – which appears to be updated daily – and populate a relational database with information on specific crime incidents, including locational information.

I’d also like to develop an application that citizens can access to identify the crime incidents that occur closest to their specific locations (home, work, or wherever they happen to be at the moment). The application would be accessible through a range of modalities – standard cell phone, micro-browser, Twitter, IM client, etc. – and would query the relational database storing crime incidents.

If anyone is interested in working on such a project, give me a shout. I’ll post updates on any progress made on such a project going forward.

Looking forward to hearing from other civic coders!

Identi.ca Looking Pretty Good!

For those government agencies using Twitter for notification, 311 and other services I’ll bet its been a tough morning.

They may want to have a look at Identi.ca – a competing (albeit much smaller) micro-blogging site that even piggybacks on Twitter’s API. I’ve said before that government’s need to spread the love among different social networking sites.

I’m guessing that today there are more people who will be receptive to such advice…

It’s About the Crowd

Last night, when the polls closed for the special election in Delaware’s 19th Senate District, people across the state fired up their browsers and pulled up the web site for the Department of Elections to check the results. And then they waited… And waited… And waited some more.

The waiting is the hardest part

The unofficial results for the election never really made it to the web. There was much grousing on Twitter about the lengthy delay in posting the results, and some reports that Department of Elections officials did not know why they technology they were using wasn’t posting to the web. This morning, a temporary page listing the results (which appears to have been manually thrown together) appears on the Elections web site.

Before I really get into my groove here, let me say this. There are scores of dedicated, hardworking people that volunteer their time on election day in this state. I know some of them – they are good friends. They almost always go unappreciated for the hard work they do to make sure that our elections run smoothly. We owe them our thanks for their service.

Having said that, if there are experienced, hardworking people investing their time into our elections process why are some aspects of it so broken? With all of the expense and effort that goes into our elections process, for something as simple as posting the results online (particularly in an election where only about 6,000 people actually vote) to fail clearly shows that the system is broken.

Tapping into the crowd

There is a different way to do this – one that is cheaper, faster and better in almost every sense. How can the state fix this broken system? Tap into “the crowd.”

Delaware officials should issue an open call to suggest alternatives to the current system of posting election results on the web – one that is cheap, simple and reliable.

Delaware is full of talented, enthusiastic, civic-minded technologists who would be only too happy to help the state devise a solution. I’ll prove it.

Since I know a little bit about the voting machine systems in Delaware, and the process used to tabulate vote totals, I wanted to propose a simple, straightforward way to collect voting totals from polling places around the state (of course there are phones involved).

Here’s an idea…

The State of Delaware makes almost exclusive use of the ELECTronic 1242 voting machine from Guardian Voting Systems. This machine captures voting totals on several memory modules, one of which is removable. The removable module is taken from the machine (after polls have closed) to a centralized tabulation site. I’m not exactly sure how the voting totals make it from the tabulation sites (I believe that there are a number of these across the state, but I’m not sure how many) to the web, but this appears to have been the issue last night.

One of the nice things about the ELECTronic 1242 is that it also provides a paper printout of voting totals – the paper printouts, along with the memory cartridges, become part of the official voting record. This is what got me thinking.

Why not employ a simple, convenient and secure phone application that would let polling place workers call in the voting totals after the polls close? Building such an application would not be hard to do – polling place workers could use a cell phone or a phone provided at the polling location to call their vote totals in. Access to the system could be tied to a workers’ ANI and would require the use of access codes. More robust security (to ensure that only poll workers can submit vote totals) could entail a call back mechanism (if ANI spoofing is a concern) or even voice biometrics – there are a number of options available in this regard.

This would allow the voting totals to be called in immediately after the polls close. These unofficial results could be stored in a web-ready database, and bypass the cumbersome, unreliable process of taking data out of the centralized tabulation centers and moving it out to the web.

The State of Delaware already has lots of experience building telephone applications — some with comparable security requirements (e.g., checking the status of a tax refund). In addition, the Department of Elections itself has fielded a phone application in the past – one built on a mature platform using open standards — to allow people to check the status of a provisional voting ballot.

I’ll even volunteer to bang out a prototype that Elections officials can use to test this approach, if there is any interest expressed by the State. Just give me a call and I’ll get cracking.

A call for more ideas

I know that there are others in our state that have thoughts on how the current system could be improved. In addition, there may be some who know of solutions working in other places across the country. If you have an idea on how Delaware can improve it’s current system of reporting voting totals, leave a comment here or Tweet it with the #Delaware hashtag.

Here’s hoping that Delaware officials start leveraging the crowd to fix this broken system.