How World of Warcraft could help test government policies

This post is written by Robert Packer, Senior Policy Advisor, UK Department of Health and Social Care.

• The problem: Experimentation in policymaking is difficult.

• Why it matters: Without experimentation, we’re reliant on correlations and case studies to inform our decisions. Experiment, the gold standard of causality, is painfully absent.

• The solution: Use virtual worlds as economic laboratories.

In January, Pia Andrews discussed ways in which we can translate government regulation into code in order to improve policymaking. She explains that “to learn to code rules well, we need to experiment. We’ll need to devote some time and money to trying different approaches. We’ll need to be able to fail, and fail safely”.

Many policy-makers will know that the above is easier said than done – ministers want results, and are rarely incentivised to invest in experimentation. And then there’s the legal context; in many areas of policymaking “experiments” may be legally risky – in service provision, those who don’t enjoy the “finished product” would have grounds to complain, and in some areas (think tax law), equal treatment is imperative.

"While modelling is a very useful decision support tool, in the complex world of social and economic policy, it is no substitute for experiment as it will only tell us about cause and effect in a very simplified world."

Citizens react in complex ways to government interventions, including government experimenting – often we have to make “big bang” announcements, or unanticipated changes, in order for our policy to be effective. Finally, policy changes often go through laborious (often justifiably so) parliamentary scrutiny that makes iterative changes in response to experimental findings challenging.

All this all too often leaves little space for the comparison of different “treatments” across different groups that defines an experiment.

A new way forward

To overcome the above obstacles, and make policy in a scientific way, policy-makers often rely on modelling – we create an economic model of the factors that will lead to policy outcomes, and run the numbers to inform our decision as to the optimum intervention.

While modelling is a very useful decision support tool, in the complex world of social and economic policy, it is no substitute for experiment as it will only tell us about cause and effect in a very simplified world. Economists often take advantage of natural experiments – real-world events that “naturally” give rise to a test and a control group, thereby permitting us to draw reliable conclusions about cause and effect. This year’s economics Nobel laureates did exactly this, answering important questions for society with empirical evidence. However, if we’re to build systematic experimentation into our policymaking we cannot sit around waiting for one of these natural experiments to present itself.

If only we had a laboratory full of people interacting with one another in order to achieve clear objectives where we could manipulate the determinants of their behaviour, observe the outcomes, and fail safely…

Games as policy experiment platforms

In theory, this exists: EVE, an online interstellar space exploration game draws over hundreds of thousands of players and, as of 2017, boasted an economy of $55 million U.S. dollars a year. World of Warcraft, another “mass online multi-player game” attracts over eight million players and has an economy larger than that of Venezuela.

Both these “virtual worlds” behave like real-world economies: they consist of millions of real people buying, selling, competing and cooperating, and the administrators have the power to manipulate the rules of the game and prices to their heart’s content.

Could these virtual worlds be used as “economic laboratories”? Could we create our own virtual worlds to these ends? If the answer is yes, this has the potential to fill a longstanding gap in the canon of economics and policymaking: experiment.

Virtual worlds could have a lot to teach us about cause and effect in policymaking.

👋 You can create a post like this one! Share your thoughts with a community of public servants. Learn more

(Image Credit: Unsplash)