The recent 2016 Code for America Summit was full of inspiring stories of civic tech successes. And, as our country responds to the presidential election, political leadership also matters. The right political leadership is needed to take advantage of our civic tech successes and to enable more in the future.
Is Donald Trump the right leader to do this? Would Hillary Clinton have been better? And, why were our choices for President so limited?
Andrew Rasiej [blog] told the CFA Summit audience that he thought more people should be working on using technology to improve our political and electoral system, and I agree.
Significant thought and effort have gone into developing better political and electoral systems. These include ranked voting [Time; N.Y. Times], deliberative democracy, politically oriented social networks [Brigade; Countable], delegated electronic voting [Liquid Feedback; Adhocracy], and other novel systems [Flux; Democracy.Earth]. Despite some impressive successes such as the crowd sourced constitution projects of Iceland and Mexico City, none of these systems have yet reached what I would call escape velocity and gained broad acceptance to the point where they are able to have a big impact on the political balance of power.
So, what is necessary to build a system that will have a big impact? I think there are 4 requirements:
- User-centric non-partisan design: a successful system needs to be both easy to use (think of your mother, your son, or a single parent with two jobs) and useful. Although many people are entertained by politics, most people want a voice and a belief that they can have a real impact on the political decisions that affect their lives. Whatever our own political beliefs, the system also needs to be welcoming and responsive to all political views.
- Designed for scale and endurance (have a vision): a successful system (or system of systems) should be capable of engaging a significant fraction of the U.S. voting population. It should be adaptable for other populations around the world, which also means that concerns about voting privacy should be addressed sooner rather than later. A large-scale system will require a way to delegate decision-making and an automated (e.g. algorithmic) way to extract consensus from a wide variety of expressed opinions. My ultimate vision would be for:
● Everyone age 10 and up to regularly complete a survey of their political views, legislative goals,
● Everyone to choose decision-making delegates for national, state, and local issues.
Transparent algorithms would:
● dynamically monitor agreement between users and their delegates,
● predict voter support for current legislation, and
● produce outlines of new legislation with support of broad majorities.
- Start small and develop agilely (have a plan): ask yourself, what is the smallest and simplest system that will bring us closer to the vision above? I believe that the smallest, but still useful prototype system would develop and test a set of scalable deliberation and consensus tools with 9-12 beta users in 3 groups. Lessons learned from this testing could be applied develop consensus within a set of approximately 100 college students on 10 university campuses with real political issues. Success here could be applied to a crowd-sourced U.S. city or state proposition or initiative. A difference between this development path and the admirable accomplishments by teams working on the Mexico City and Finland constitutions would be that this system would not only allow popular opinions to rise to the top, and it would also include ranking of different proposals and optimize among different opinions.
- Be consistent with the remainder of Tim O’Reilly’s lessons for open government: In short, these can be abbreviated as: based on open standards, simple, agile, participatory, responsive to users, open to experimentation, encouraging measurement, and supportive of developers. In particular, open standards are important for user adaptation and trust in the system. Although there is money and momentum behind for-profit systems, I believe the most successful systems will allow users to export personal data, disclose algorithms used to determine consensus, and allow reasonably priced access to aggregate data by interested parties.
What kind of impact would a successful system have? This system should be a tool to document differences between the voters’ will and government actions, it should enable better laws to be proposed and passed, and it should pressure members of government to support the interests of all of their constituents. Specific benefits could include minimum wage changes; immigration, criminal justice & education reform; and better community/police relations. And, hosting platforms would benefit from a larger user bases and associated revenue opportunities.
What assumptions does this proposal depend on? Nothing particularly radical or unfounded. This proposal assumes that some volunteers will participate and that there are common goals and beliefs across the middle ~80% of our country. If the second assumption is not true, we are in serious trouble and fixing our society will take much more time and education. But, electronic democracy systems (and existing polling infrastructure) can test this hypothesis.
What are the next steps? Personally, I plan to talk to my neighbors and local Code for America Brigade teams, to continue to develop simple ranked voting and optimization codes, and to reach out to existing electronic democracy developers such as Democracy Earth. Personally, I also believe that organized discussions at venues such as the next Personal Democracy Forum in New York City or Code for America Summit in Oakland will be invaluable.
To conclude, as Chester Bowles said, “Government is too big and important to be left to the politicians”. Let us work together to help improve our politics.