One of the pet projects I’ve wanted to work on for a while is a site that would guide political policy by directly polling all citizens — rather than just elected representatives — on how a government ought to spend its resources. Let’s call it Citizens in Charge… CiC.

Here’s how it would work:
1. Upon registration you are given 100 “policy dollars” to spend.
2. You get to tell government what issues you want government resources spent on. For instance, “I want $15 on health care, $3 on the military, and $20 on public transportation.” You can also nominate new issues if they aren’t already on the ballot, e.g. “Spend all the rest on skateboard parks”.
3. You can change your allocation of policy dollars at any time.
3. The service continually aggregates participants’ responses and calculates averages and other useful statistics. So we might discover, for instance, that the average spend across all participants on skateboard parks is $22.
4. The service continually communicates the results to the public and to government officials.
5. Optionally, you could also use the service to do simple one-off polls, e.g. opinion on participation in the Iraq or Afghanistan wars.

With a few thousand citizens participating I think you would get, at minimum, a public opinion poll that demands attention.

I got to thinking about CiC again after stumbling across a site called ConvinceMe.Net. ConvinceMe bills itself as, “A new way to argue. Debate online, vote online.” In a nutshell it’s a service that hosts online community debates on issues proposed by anyone. ConvinceMe seems more geared at entertainment than serious policy-making, but it is nevertheless the sort of building block you’d need to make the CiC vision come true.

CiC implementation challenges:

• Who gets to frame the issues, and how many issues should there be? For instance, should the issues be worded as “spend $X on issue ABC”, or “spend $X on issue ABC by taking actions X, Y and Z”? I prefer the former, at least as a start, because it’s simple. You could begin with the list of top-level budget headings from the existing government budget, plus an option to propose new issues at the grass roots level.
• How can issues be framed in a relatively unbiased way, with actionable results? This is always a problem in a referendum, where much wrangling occurs over how to word an issue before it is put to a public vote. ConvinceMe suffers from abuse in this regard.
• How to encourage broad and representative participation? (Lots of people dislike polls, or feel they are a waste of time. And would this be a service that only the computer-savvy make use of?)
• How to discourage anti-social behavior such as multiple voting, false impersonation, hacking, and so on? And how to protect participants’ security and privacy in case the service does get hacked?
• How to ensure only people who are allowed to vote can vote (e.g. only citizens of country X can vote on issue Y), and that their vote counts only once.

It seems inevitable to me that at some point such a service will exist, at least as a research tool if not a tool to inform public policy. Any computer science and politics experts out there willing to partner up and take a crack at it?