Birmingham, AL has a population of 212k and a yearly budget of 390 million. (https://ballotpedia.org/Analysis_of_spending_in_America's_la...) When you consider the USA govt spent about 70 Billion on R&D in 2016 (http://www.aaas.org/page/historical-trends-federal-rd). That would be about 0.5% of total R&D spending.
One of my personal hopes with Basic Income is that it addresses this issue by providing a safety net so that people who's livelihood relies the existence of the status quo can be liberated from the incentive to keep the status quo. For example, our government creates a Department for Homeland Security to address threats. It employs thousands of people. What happens if the threats go away and the department is no longer needed? Suddenly thousands of people are out of a job. That's thousands of people with an incentive to keep the department in business, which requires threats to our homeland security. Multiply this example across not just government institutions but private as well. People have an incentive to keep their gravy train running, even if it does irreparable harm. We need to remove this incentive so we can have faster iteration cycles, not just for socio-political ideas, but all ideas. Hopefully, a basic income can remove these perverse incentives by giving people a safety net. They're not afraid to lose their job, or move on from a shitty job, because they know they and their family won't go hungry, homeless, or without access to medical care.
I think we need a way to have a faster iteration cycle for trying sociopolitcal ideas. The problem is, its almost impossible to change the laws necessary at a governmental level in any reasonable time scale. This is exacerbated by the fact that you are inside a "monolithic legacy codebase" in which you can only afford to make minor changes over many years to take you from point A to some goal which is point B, and you still can't verify that point B works in any practical way.
Ideally, you could have at least an entire city (population ca. 50k minimum) where you could just try out all this stuff, and continually carry out sociopolitical experiments. I suppose that's not feasible, but any idea that goes in the direction of increasing iteration times and improving the testability of ideas like this would be a good direction I think.
What do other HNers think about this?