David Rodeback's Blog
Local Politics and Culture, National Politics,
Tuesday, January 19, 2010
I Love Numbers, but What Do They Mean?
Numbers mean things, and I adore them. But they don't always mean what they appear to mean at first glance.
I love numbers. I've loved them as long as I've known them. I work with them, off and on, and I like to relax with them, too. In my youth, I would buy a calculator, learn all the math it could do, then buy a more advanced calculator and repeat the process. More recently, this trend has petered out; I've had the same calculator for about ten years, and I'll probably never learn everything it does.
Numbers are wonderfully useful, insightful things, as long as you use them right. But they themselves tempt us to misuse them, and there are also people out there who are perfectly willing to encourage the abuse.
Consider today's special election in Massachusetts to fill the late Senator Edward Kennedy's vacated seat. Polls before the election can give us some sense of what to expect and why, but things change from day to day, and the polls themselves are easy to skew, if you're trying to achieve a particular result. Exit polls are interesting too, but they, too, have their limitations. They're often poor at predicting actual outcomes.
In the end, the numbers that really matter are the actual votes cast by the people who actually vote. Knowing those numbers, we are poised to look back at the exit polls, for example, for some explanations. Exit polls are much better at explaining something we already know happened than they are at predicting what will happen.
From time to time, we hear of increasing troop counts in -- at the moment -- Afghanistan. But they are a poor measure of our national commitment there, not to mention our intentions and prospects. Significantly, the numbers tell us nothing about our troops' rules of engagement, which in the modern world is often where a war is won or lost.
For many months we've been hearing counts of people who have H1N1, the Malady Formerly Known as Swine Flu, as well as the death tolls. But stories which put those numbers in perspective are relatively rare. For all the news reports I've heard and read on the subject, only about a dozen have compared H1N1 infections and fatalies to other flu epidemics. Are more people getting it, or fewer? Are more people dying, or fewer? Is the percentage of people infected who actually die lower or higher than with other influenza strains and epidemics? In other words, should we be more worried about H1N1 than about other flus? Body counts don't tell us much. When more numbers come out, yes, swine flu seems a little deadlier. And the really interesting numbers show a major difference in who is dying: with H1N1, it tends to be young people, which is the opposite of ordinary flus.
I heard a story just today that there are 116,000 dead voters on Massachusetts' voter rolls. What does this number mean? Is there massive voter registration fraud? Will today's results be illegitimate, no matter what they are? Or is this simply a matter of local governments having more urgent matters to which to devote their limited funding than keeping the voter rolls squeaky clean? In truth, I've never seen a voter registration roll that was completely current, and I never expect to. The crucial numbers are the ones we don't yet know: How many of those dead registered voters will vote today? And how many times each, on the average? That's where the harm is, and it is preventable in other ways. In the end, it's a lot cheaper to require photo identification at the polls than it is to update the voter rolls every time a newspaper runs an obituary.
Every so often we hear reports of rampant hunger in the United States. There was a story two or three months ago which claimed that one-sixth of Americans "don't know where their next meal is coming from." Let's take a moment to think about conditions in Haiti right now, where the problems are massive. Now, if you please, let's look back at the United States. One-sixth?
The stories are spun as showing the need for even more generous donations to food banks and other similar charities -- which would be perfectly fine with me. There are real needs out there, especially lately. But am I supposed to believe that one-sixth of Americans (a) have completely bare cupboards, pantries, and refrigerators; and (b) have no money to buy any food at all; and (c) aren't receiving food stamps; and (d) have no honest way of earning a little money for food or of otherwise obtaining food today?
In slightly better economic times, I'd be willing to believe that one-sixth of Americans don't know where their next meal is coming from because they haven't decided where they'll eat out (or take out) this time. If one-sixth is a real number at all, in this case, I might be willing to believe that they asked a lot of children, who don't know where their next meal will come from, not because their parents don't feed them, or at least provide food to eat, but because they're children and don't have to worry much about these things. The news story in this case didn't say anything about whom they asked, what questions they asked, what the answers were precisely, and how they evaluated the answers. Without that information, the number is just hype.
A final thought: My readers are probably too intelligent for this, but there are others out there who would take my skepticism about specific numbers as evidence that I don't think anyone's hungry out there, or that I don't think we should help them -- or similar things about H1N1 victims, voter fraud, or other significant matters. For some reason, some people think that we who refuse to swallow every proferred dram of hype lack compassion for real people who suffer real problems. I'm glad my readers are smarter than that.
As regards compassion, well, that's a topic for another day. I estimate there's about a 72 percent chance that it will be next week, tee hee.
Copyright 2010 by David Rodeback.