Thinking in numbers

I just heard a discussion of the numbers of people on the U.S. terrorist watch list. There was a recent story that the list included one million names, which represented some 400,000 people (with names plus aliases). Of these we are told, 95 percent are foreigners. Thus we can assume that there are perhaps 20,000 U.S. citizens on the watch list.

Now my name is a common one, so I am aware that if there are one million names on a list, and the list is managed reasonably well so as to allow for alternative spellings and alternative versions of first names (e.g. John, Jack, Johnny and Jackie, not to mention Jon and Sean) there are a lot more people with their names on the list.

But is 20,000 a large number of Americans for the list. It clearly is if you think about how many FBI agents it takes to keep track of 20,000 people. It is also a huge number if you consider how many of these people must be innocent (remember Senator Edward Kennedy was on the list as was Nelson Mandela) and having their liberty limited and privacy limited by government surveillance.

The Census Bureau tells me that there are just over 300 million Americans. Thus 20,000 people represent one in 15,000 of our citizens. Would you be surprised to think that one in 15,000 people would be worth watching in case they committed a mass murder or other act that inspired terror? Maybe not.

After all, we keep 2.3 million people in prison (according to the Washington Post). In theory, I assume we imprison them because we suspect that they might otherwise victimize the public with new crimes (for if they were no threat to society, would it not make more sense to put these folk to work to pay restitution to society for the crimes they committed in the past). So maybe 20,000 Americans is not too long a list of possible terrorists.

The point, in terms of the focus of this blog on knowledge for development, is that what we think of a number depends on the context in which it is framed.