Wednesday, January 26, 2011

Numbers Don't Lie (but Michele Bachmann Does)

In her rebuttal to last night's State of the Union Address, Minnesota congresswoman and entertaining crank Michele Bachmann showed the following graph:

If you look beneath the graph's horizontal scale, you'll see that these data are actually monthly unemployment rates for October of each year. (The graph doesn't show annual employment, as one might have expected from a year-by-year display.)

Presumably the reason for picking October is that
it allowed Bachmann to highlight the 10.1% spike in unemployment that occurred in October of 2009. (The annual unemployment figure for that year was a less dramatic 9.3%.)

Noreen Malone at Slate has noted the lack of labels for the even years, which helps to create a sense of separation between the two administrations. But personally, I can't see that the data support such a clean separation. Here is what you get when you look month by month:

I was also amused to notice that in Bachmann's graph, President Bush's bars were squeezed to the left of his yearly slots, while Obama's bars were squeezed to the right of his yearly slots. The effect of this was to enlarge the gap of white space between the red bars and the blue bars - again enhancing the sense of separation between the two administrations. With a regular pattern of small gaps to the left, and a regular pattern of small gaps to the right, the large gap "does the work" of distinguishing between the two presidents. Nice of her team to save us the trouble.

Recently I laughed out loud when I saw Chris Matthews losing it while interviewing Tea Party strategist Sal Russo about a speech Bachmann had given, in which she suggested that the Founders worked tirelessly to eliminate slavery. Bachmann had said:

But we also know that the very Founders that wrote those documents worked tirelessly until slavery was no more in the United States.

Matthews went ballistic about this apparent attempt to whitewash the story of the founding of the nation. The Constitution was ratified around 1788, and slavery was abolished around 1865; so given that 1865 - 1788 = 77, it's pretty hard to imagine that the Founders "worked tirelessly until slavery was no more in the United States." Many have suggested that Bachmann is ignorant of history; maybe she's just ignorant of subtraction.

Bachmann's speech posed another math puzzle:
Do you realize it's been 21 generations that America has survived?
As soon as I heard this, I started scratching my head because I normally think of a generation as being about 25 years long - and 21 times 25 is going to be something like 500 years. Yet I had always thought of America as being more like two hundred and some years old.

Could Bachmann be referring to Columbus? The fuller quote is:

Do you realize it's been 21 generations that America has survived? For 21 [generations], we've passed the torch of liberty from one generation successfully to the next.

It's hard to believe this refers to Columbus, who as I understand it was an Italian. But then again, I don't know what the alternative is. Wasn't "the torch of liberty" lit sometime around 1776? If so, then (2011 - 1776)/21 = 11 or thereabouts...and a generation can't be 11 years long.

So I don't know if she's fuzzy on history or fuzzy on math. She's certainly fuzzy on something.

Sources: I found annual unemployment numbers at For monthly numbers, I used the query tool at to find the following (paste into Word and use the Convert Text to Table command):


Monday, January 10, 2011

Word Puzzle, Word Problem

1. A word puzzle:

"Downshifting" a word means replacing its initial letter with the immediately preceding letter of the alphabet. For example, to downshift the word BOIL, replace the B with an A to obtain AOIL. (Unfortunately, AOIL is not a word.)

Some words are "downshiftable," meaning that when you downshift them, you obtain a new word. For example, TIP is downshiftable, because it downshifts to SIP.

In fact, SIP is itself downshiftable, because it downshifts to RIP.

But unfortunately, RIP is not downshiftable, because QIP is not a word.

Altogether then, we might say that TIP was downshiftable twice: TIP -> SIP -> RIP.

Can you find a word that is downshiftable three or more times?

The best word I could think of before I fell asleep last night was 3x downshiftable. When I woke up this morning, I wrote a program to see if my computer could do any better. It found six words that were 3x downshiftable (mine was in the list), as well as a single 4x downshiftable word! Happy hunting.

Fine print: In case you were wondering what to do with words that begin with A, let's agree that A always downshifts to Z, as in AERO -> ZERO. (Not that AERO is a legitimate word, it's just the best example I could think of to illustrate the concept.)

2. A word problem. This one is for math-major types (solution here).

John picked up a notebook and saw that there was a positive real number written on each page. The product of all the numbers was e^(S/e), where S was the sum of all the numbers. Prove that S is not an integer, and find the average of the numbers in the notebook.

Friday, January 7, 2011


Seen today on Google News:

BBC caught the error quickly; if you click the link, here is what you see:

Since we're doing journalism again today, I'll end by linking to something funny I read that relates to my earlier post about science journalism.