This group will be discussing the use of virtual worlds in government. The discussion is related to the National Defense University iCollege’s Virtual World’s Initiative: The Federal Consortium for Virtual Worlds
Using virtual worlds to investigate policy
May 17, 2010 at 3:00 pm #100605
Ted Castonova proposed at the Federal Consortium for Virtual Worlds conference (http://www.ndu.edu/irmc/fcvw/fcvw10/agenda.html) that using virtual worlds and a game design could be used for social science research. In particular, he suggested that this iterative approach could be taken for testing and refining public policy. What do you think?
May 17, 2010 at 3:20 pm #100616
Did he give any specific examples, Paulette?
May 17, 2010 at 3:49 pm #100614
He gave examples of how he has been using this method to test economic influences. He gave some examples of how he thought this could work with health care policy. He suggested that Farmville could perhaps be used by USDA to get at information to inform their policy. The compelling part for me was that we could use virtual worlds to iterate and test public policy data, theories or approaches before it is fully implemented.
May 18, 2010 at 1:38 pm #100612
I thought his suggestion made sense with the examples he used and with other policy decisions we would try to implement at a local level.
In local government, this would help to visualize impacts of zoning and planning policy decisions. Many times there are several options, and by using a virtual build, those trying to decide can actually see how each option would work. And this would help elected officials make a more confident decision which could then be shown and conveyed to constituents.
Here is one example where this would have saved a lot of time, money, and frustration: We had a developer in the last community where I worked who kept creating lots that were about 90 feet in width and established 14 foot side yard setbacks – this meant the widest home that could be built was 62 feet which included the garage since there was no alley. Since many wanted 3-car garages, this takes off about 36 to 40 feet leaving only a 26 to 22 foot wide house. The homes were in a high-end neighborhood with residents who wanted homes measuring much more than this, and a lot of the designs just did not fit. This is a perfect example of how a virtual build would have proven during the approval phase that this type of neighborhood needed either wider lots or smaller setbacks. Instead we dealt with people and builders who bought a lot not realizing their dream home would not fit and who then became upset and frustrated.
May 21, 2010 at 2:55 pm #100610
With current technology, you could create a farm game like Farmville (80 million users) and put in a price support system. We know from existing experiments that people in games react to prices just like people in the real world do (when prices go up, people buy less). Thus one could test out a new price support policy by manipulating the price support policies in the game.
The point is not to *simulate* but rather to *experiment*. A rat in a maze provides a generic problem-solving situation that psychologists and neuroscientists have used for decades to learn more about generic issues in cognition. What I am proposing is the equivalent of a rat maze for public policy. In the given example, we would be studying a generic price support system using the same kinds of controlled experimental methods that scientists use for cognition.
Another example: The Earned Income Tax Credit is an outgrowth of the Negative Income Tax, a concept first floated in the late 1960s. It was tested at great expense and without solid results in the SIME-DIME tests of the later 1970s. I am proposing that we can now do this kind of generic testing of income support in a much quicker, much cheaper, and ultimately much more believable way. A virtual game with an income support system could be used to quickly and easily answer generic questions, like “What happens to labor supply /in general/ when you switch from a minimum-income support plan to a wage-supplement support plan?”
The claim here is not that we should simulate all government inside a virtual world, but rather that we use VR technology to create quick, cheap tests of broad propositions in policy. Also: The claim is not that we would be able to test all questions that a decision maker may have, but only those for which a proper virtual test can be built. Now, what questions might fit this criterion? I don’t know. But I am sure there are many, just as I am sure that *any* regime of quick, cheap policy testing is certain to improve outcomes and policy delivery from what we do now.
May 21, 2010 at 3:05 pm #100608
After my talk I consulted briefly with a program admin who realized suddenly that the exercise of creating a policy in virtual space was *by itself* likely to be valuable. That is, *before* doing any experiment, simply making the effort to implement the policy using real people who care about what’s going on (because there’s a game going on that they care about), is going to produce some obvious learning points.
Think about Monopoly. Suppose there’s a big Monopoly game going on with 1,000 players for two weeks. Now you, as Government, decide you’re going to change how the rent system works. Say, you’re going to cut the rent of RRs by 50%. Uh-oh! People are going to give you some rapid and loud feedback about that. You have to be careful about what you are trying to do, and how you actually decide to implement it. A compensation issue immediately arises from people who already own RRs. Maybe in the planning stage you would never have thought about how critical this compensation issue was going to be. As a senior administrator, you can remember how much time you have spent teaching new hires about things like this, about the nasty way the world responds to even the littlest policy changes. If you could have your gung-ho new hires go through this experience themselves, maybe they’d learn the ropes more quickly. And then we could do experiments on the Monopoly game to figure out how you *would* actually be able to lower the RR rent level without touching off a firestorm.
A silly example. Monopoly! What could we learn from Monopoly?!?! But think about it. Replace ‘Monopoly’ with a big VR game about real estate, land ownership, housing, and development. I know people would love to play a game like that. And I firmly believe that HUD could make good use of it for real world policymaking.
You must be logged in to reply to this topic.