Monday, September 28, 2015

Numbers is hard

A friend of this feature sent this one along, in which a professor struggles to explain what should be obvious to the bien pensant It Couple of TPT, Cathy Wurzer and Eric Eskola:
[University of Minnesota Professor Sam Myers] pointed out that the survey only included those who indicated themselves as “black” and not of multi-racial background.  He also said he couldn’t reproduce the finding that there was a $4,000 drop in median income for black households, based on any previous data.  The Star Tribune reported the survey numbers with a dramatic headline, “Black household income plunges in one year in Minnesota.”  Yesterday the paper featured a commentary by Louis J. King III, president of Summit Academy, reiterating the point that “Minnesota now trails Mississippi when it comes to median household income for blacks.”  As a solution, King called for an increase in SNAP (food stamp) benefits and an increase in job training program spending,

But are the statistics valid?  Not according to Professor Myers who stated, “I looked at every single year of the American Community survey 1% sample between 2000 and 2013…… six of those years, black income went up, seven of those years, black income went down.  But, the long term trend was a positive trend.”

“I don’t know anybody who uses that 1% sample…..You should use a 5% sample, there is a 5% sample.”  Myers said.
Pick the numbers you want. Get the results you want. Nice racket. And lest you think Myers is some sort of racist who wants to keep people down, a reminder from the story:
Myers is the Chair of the Roy Wilkins Center for Human Rights and Social Justice at the Humphrey Institute.
And a reminder: Eskola and Wurzer have been hosting Almanac for 20 years.

1 comment:

Bike Bubba said...

I'm no where near as bothered by the fact that a couple of journalists couldn't understand the statistics as I am by the fact that the Census Bureau appears to have had the same problem. Or, worse, knew better but released the study anyways.