Union Membership and Test Scores

Bill Gates, like many other educational Deformers, makes the assumption that having a strong teacher union is a Bad Thing for students, because the union supposedly runs the evil bureaucracy that does its level best to avoid teaching anything to students.

From my viewpoint, having taught for over 30 years in Washington DC, that assumption is so wrong it’s laughable. At best, the union can negotiate pay scales and can represent teachers (good or bad) so that they get some sort of due process when the school administration goes after them. However, much of the time, teachers are too frightened even to speak up, and even with due process, many teachers do get canned or forced out quietly.

In any case, I wanted to look at a particular paragraph where the supposedly highly intelligent Gates compares Florida and Texas, on one hand, with Massachusetts and New York on the other.

Gates makes the assumption that Florida and Texas, which are “right-to-work” states and whose politicians have famously taken a leadership role in the current educational Deform movement, should have students who do better on various standardized tests.

Here’s the quote:

“Asked if the National Education Association and the American Federation of Teachers have any incentive to back school reforms that help kids but also diminish union power, Mr. Gates responds by questioning the scope of that power. “We have heavy union states and heavy right-to-work states, and the educational achievement of K-12 students is not at all predicted by how strong the union rules are,” he says. “If I saw that [right-to-work states like] Texas and Florida were running a great K-12 system, but [heavy union states like] New York and Massachusetts have really messed this up, then I could draw a correlation and say it’s either got to be the union—or the weather.””

Well, in fact, in general, Texas and Florida generally score LOWER on the NAEP, the only national test we have outside of the SAT and ACT, than do New York and Massachusetts. In fact, students in Massachusetts consistently scores well above all other states in all of the reading and math NAEP tests given in 2009 at the 4th and 8th grade level.

I prepared some graphs using publicly available data comparing 2009 math and reading NAEP scores and percentage of union membership among all employees in each of the 50 states plus DC. I have picked out some of the very highest-scoring states and some of the very lowest-scoring states. (No surprise, when you compare the city of DC versus the states, DC is last.)

Here is a graph showing the 4th grade reading NAEP scores in the y-axis versus the average total percentage of union members, in each state.

Dots that are close to the right-hand edge of the graph represent states with relatively high percentages of workers that belong to labor unions. In each of these graphs, New York has the very highest percentage of union members of all of the states. The state with the very lowest percentage of union members is North Carolina, at about 3.1%, but I didn’t mark it. Can you figure out which one dot represents NC? It’s the dot a bit to the left of the one for Texas.

Dots that are towards the top edge of the graph are the ones with the highest average state-wide 4th grade reading NAEP scale scores, and the ones towards the bottom are the ones with the lowest scores. I labeled the dots for Massachusetts and New Jersey, which are two of the highest-scoring states; they also have pretty high rates of union membership.

The two lowest dots are for Louisiana and the District of Columbia, which also have relatively low percentages of union membership as well.

I drew two red lines. The vertical one represents the median of the percentages of union members among all employees in each state: about 10.8%. So any dot to the right of that line has a higher-than-median union membership level, and conversely for those to the left of that line. The horizontal line represents the median 4th grade scale score for all of the states and DC: 221. Any state scoring above that line has higher-than-median NAEP scale scores for 2009 in reading for the 4th grade, and vice-versa.

When I look at this graph, it almost looks as if higher percentages of union members in a given state seems to be correlated with somewhat higher grades on the NAEP. But the correlation is not very strong.

Here is another graph, this one for 2009 fourth-grade math NAEP scale scores versus union membership in 2009.

Once again, the two lowest-scoring states are ones with low-to-middling percentages of union members: DC and Alabama. The two states with the highest scores are those with high-to-middling percentages of union members: Massachusetts and New Hampshire. NY, FL and TX have middling scores, but are mostly differentiated by their very different percentages of union members. (Frankly, I am not sure how teachers in rural Texas can pay their rent or mortgages without second jobs…)

Here’s another graph, for 8th grade NAEP reading against percentages of union membership in 2009:

Once again, you see that Bill Gates’s assumption that unions are bad for students doesn’t appear to hold water. Both Florida and Texas, states where very few workers belong to labor unions, scored below the median on the 8th grade reading NAEP. Once again, Massachusetts and New Jersey, with higher-than-median memberships in labor unions, scored at or near the top. New York scored about the same as Florida. The bottom scoring states were DC and Mississippi.

Last graph, for 8th grade 2009 math NAEP scores versus union membership:

Again, if anything, the assumption that he made, that union membership is bad for students, is shattered.

Source for union membership is here.

For the various NAEP scores, you have to go to the main data exploration page for the National Assessment of Educational Progress and look around.