How a Lie Becomes a Law

Last Friday, November 19th, Bill Gates was in Kentucky giving a speech on education. The PI had the AP summary the next day:

On Friday, billionaire Bill Gates took aim at school budgets and the master's degree bonus.

"My own state of Washington has an average salary bump of nearly $11,000 for a master's degree - and more than half of our teachers get it. That's more than $300 million every year that doesn't help kids," he said.

"And that's one state," said Gates, the co-chair of the Bill & Melinda Gates Foundation, at a speech Friday in Louisville to the Council of Chief State School Officers. Gates also took aim at pensions and seniority.

Also picking up the lede were KOMO TV in Seattle, along with the Seattle Times.

Today, the WashACE blog picked up on the story, calling Gates' comments "perceptive insights" in the broader context of how to handle the state budget crisis.

Bill Gates says it, the media repeats it, it becomes the groupthink of the ed reform class, and merrily we go on our way.

The trouble is, what Gates is saying is flat-out wrong. You can see for yourself by going to the OSPI website and looking at this year's salary schedule. Look at the column for BA+0, then look in the same row at the column for MA+0. The result? A $6,800 difference. Move down to the row for 8 years of experience, and you'll get to about a $7,000 difference. Neither is the $11,000 difference that Gates cited.

How was the lie born? Gates didn't just pull those numbers out of the ether; instead, he's quoting a study done by the Center on Reinventing Public Education (CRPE) at our own University of Washington. The lead researcher is my old friend Marguerite Roza, and the germ of the lie can be found in the appendix:

This analysis used data from two sources. The 2003-­‐04 Schools and Staffing Survey from the National Center for Education Statistics provided state-by-­state figures for both the percentage of teachers with masters degrees, and the average salary of teachers at each degree level—bachelor’s or below, master’s, to name a few—for given years of longevity. This analysis used these data to compute the average percentage salary increase awarded for education credits earned beyond a bachelor’s degree. The analysis then applied the percentage increases to the more recent state-­by-­state average salary figures and total number of teachers from the National Educators Association’s 2008-­‐09 Salary Survey, in order to compute the dollar value of the master’s bump in each state.

(Personal aside: National Educators Association? You're an expert on reforming public schools, and you can't even get the name of the frackin' teacher's union right?)

Look at the process here: using data from 2003, published in 2004, then refracted again through the lens of a different study from 2008, this final product was made. Common sense tells us that the more you play with any set of numbers the farther away from their original meaning they're going to get, and that's exactly what you're seeing here. Instead of using the salary schedule--the simplest, clearest data--Dr. Roza decided to do data gymnastics instead. It calls her results into question.

Things get worse. Quite a bit worse, to my mind. The 2003-2004 Schools and Staffing Survey that the report mentions is available online, and one of the tables that you'll find is the pithily named "Percentage of public school districts and private schools that had salary schedules for teachers and among those that had salary schedules, the average yearly base teacher salary, by various levels of degrees and experience and selected public school district and private school characteristics: 2003–04." Their numbers?

In 2003-2004, the average salary for a teacher with a bachelor's degree and no experience: $29,100
In 2003-2004, the average salary for a teacher with a master's degree and no experience: $31,900.

That's a difference of only $2,800.

So let's look at this lie again: the numbers had to be processed more than the "meat" that makes the average hotdog, and there was a perfectly good data set in the report that was ignored in the process of making the larger point. There were two easy ways to get to the heart of the question, and the CRPE ignored both of them. That's either driven by an agenda or rank laziness; neither is particularly appealing.

So Bill tells this lie. He may even believe the lie, but it's a lie in spirit and in fact because of the absolute crap job that the CRPE did in getting those numbers. The real pisser, if you're Bill Gates and have been made into a liar by this "research", is found on the very last page of the report:

Funding for this work was provided by the Bill and Melinda Gates Foundation.

Money well spent, Bill.

Why let the truth get in the way of a good lie, though? It's much easier to just repeat what Bill said. He's got a foundation, after all, and he talks a lot about education, not to mention the whole billionaire thing, so he must be right. Just watch--in the coming months you'll hear this lie repeated with vigor by people who heard the soundbyte and nothing else and figured "Hell, Gates said it, and there's real University-backed research, so it must be true!"

And that, my friends, is how a lie becomes a law.