Back in June, I made a bet with Dr. Stephen Davies, Head of Education for the Institute of Economic Affairs (IEA) in London. We had got into a discussion on Facebook about whether there was much chance of a civil war in the United States. I pooh-poohed the idea and tried to formulate a bet with Steve. I first offered 10 to 1 odds and he estimated the probability of a civil war in the United States in the next 10 years to be under 10 percent and so turned down the bet. So I made a more-generous offer of 20 to 1 and he accepted. Because of co-blogger Bryan Caplan’s difficulty in getting someone he bet with to pay up, Steve and I agreed to come up with an arbiter. I suggested Bryan, Steve agreed, and Bryan has generously agreed to be the arbiter.

Here are the terms of our bet, with a clarification by Steve at the end:

1. If there is a any 365-day period between June 15, 2018 and June 15, 2028 in which 1,000 or more people are killed in what appears to be a civil war in the United States, you win the bet and I owe you $400.

2. If #1 does not obtain, you owe me $20.

3. If one of us dies or develops dementia before the time period is up, the bet is cancelled.

4. We agree that we will choose a mutually agreeable arbiter to make a judgement in the event that we don’t agree on the data. We will make this choice by July 14, 2018. [Note: we missed that deadline by about 6 weeks.]

Steve added:

I think the descriptive criterion is the one we already agreed on right? Good idea to cover the contingencies. The only clarification I would add is that ‘civil war’ in this context means “widespread and serious violence of a political nature” so breakdown in law enforcement or something like a major gang war would not count but the kind of confused political violence we have seen in places like Syria, Lebanon, and Somalia would. That ok with you?

It is ok with me and I appreciate Steve’s clarification because it not only is on point but also reduces the range of outcomes for which I would have to pay up. 🙂