view

Bad Math That Passes for Insight

We live in an era of technological advancement. Whether it's genomics, nanotechnology or software algorithms, the world is driven by mathematical solutions to complex problems.

Yet at the same time, we are surrounded by what I like to call Bad Math. It seems as if the average person has little familiarity with the fundamental workings of mathematics. Statistical errors are rife. Even the classic error of confusing correlation with causation seems to be impossible to vanquish.

Some of the blame for this lays in our education system. We teach by rote, instead of explaining critical reasoning and analysis. Instead of teaching children what to think, we should be teaching them how to think. It is a fundamental failure of our education system.

Which brings us to today's subject.

As you might imagine from a review of any of our morning reads, I plow through a lot of media, much of it good, some of it excellent. What you don't see linked are the poorly reasoned, statistically amateurish, logically defective articles that don't make the cut. Too many writers seem to have a dominant right hemisphere. While that may be great for creativity and language skills, it means they are deficient in logic, numbers and analysis.

Perhaps a few examples might help to explain my ire.

Consider this recent Forbes headline: "5.5 Million Americans Eye Giving Up U.S. Citizenship, Survey Reveals," and its annual twin, "Record Numbers of Americans Are Renouncing Their U.S. Citizenship." And I just learned that the number of Americans renouncing U.S. citizenship increased 39 percent in the third quarter! Wow, that is a huge number. People must be leaving the U.S. in droves. We have to do something about this, right?

The first headline is based on sloppy thinking; the second is technically accurate, but feeds into the misleading narrative of the first.

Let's start with the actual numbers: In the third quarter, a grand total of 776 people renounced their US citizenship. Most of them live abroad, and no longer want to pay taxes to the U.S. AND the nation where they lived. Although a 39 percent increase sounds like a huge change, in actual numbers, it's modest. And the underlying reason is obvious: If you are an American living abroad, with no intention of returning to live in the U.S., why would you pay income taxes to both the U.S. and your resident country? The U.S. is the only Organization for Economic Cooperation and Development nation that doesn't exempt its citizens for taxes when they reside overseas and pay taxes to their domicile country. Hence, this is a question of basic accounting, combined with a resident's assessment of whether they might return to live in the U.S. or not.

So far in 2014, 2,353 have renounced their citizenship. In a nation of 316.1 million people, this isn't even a rounding error.

When I read a sentence like this in the Forbes article, "73% of Americans abroad are tempted to give up their U.S. passports," I know I am entering bad-math territory. I don't know how anyone can quantify the word "tempted" or how anyone can derive any significance from it. However, if temptation mattered, considering how many times over the past year I was tempted to throttle journalists for abusing math, I should be on the FBI's most-wanted list.

Perhaps more amusing was a paper that looked at "The Relationship between Wedding Expenses and Marriage Duration," by two economists at Emory University. Of course, all of the subsequent headlines focused on the splashy but dubious correlations outlined in the paper:

CNN: "Want a happy marriage? Have a big, cheap wedding"

New York Post: "The pricier the ring, the likelier the divorce"

The cost of your engagement ring or your wedding is unlikely to tell us if you might have a happy marriage. It does, however, let us know if you are likely to be able to afford to divorce.

Another example comes from a site called Music That Makes You Dumb, which at least says it recognizes that correlation doesn't equal causation. You have to wonder if same can be said of the Washington Post's Know More blog (perhaps a name change is in order). By comparing the 10-most-popular bands at each college with their average SAT score, we end up with a correlation between musical tastes and "dumbitude and smartitude." See graphic below:

Rather than dissect this, I'll leave it to readers to do that.

To be blunt, I don't really care what your musical tastes are or if you actually like Nickelback or Lil Wayne. Experience shows that musical tastes are a function of factors unrelated to intelligence: Upbringing, where you live and culture. What I do care about are overbroad generalizations, poor reasoning or spurious correlations.

Dissecting the logical flaws of this thesis and all the others is easy. I hope readers learn to do the same with everything they read.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Barry L Ritholtz at britholtz3@bloomberg.net

To contact the editor on this story:
James Greiff at jgreiff@bloomberg.net

Before it's here, it's on the Bloomberg Terminal.