Tuesday, May 13, 2008

Teachers Unions Still Needed

Kevin Carey talks about a new Education Sector report which Carey touts as standing for the proposition that even among teaching newcomers, union protection is essential.
I suspect there's a hope among some union antagonists that teachers unions will fade in power and importance over time, much as their private sector counterparts have. That hope is often based on a generational theory of change: as the teachers who remember or participated in the initial struggle for unionization retire in large numbers in the next few years, they'll be replaced by a younger generation that grew up in a more de-unionized society, people who don't see teaching as a 30-year career leading to a comfortable retirement and will thus be less supportive of what unions have to offer.

According to a new survey of teachers published this week by Education Sector, these hopes are unfounded. From 2003 to 2007, the percentage of teachers who see teachers unions and associations as "Absolutely Essential" increased from 46% to 54%, a statistically significant change. And the change among teachers with less than five years of experience was even more striking: 30% to 51%.

Interestingly, this is despite the fact that most teacher are quite open to reforms of traditional labor arrangements that many teachers unions fail to actively support at best, and oppose at worst.
While one can quibble with Carey's and Ed Sector's conclusion, the fact is that more and more younger teachers think that the positions that the union opposes regarding the profession, i.e. recruiting teachers, salary and compensation changes, may make for a change of heart as the progress through the ranks.

Just as an aside, it is worth noting that the number of "newcomers" with less than five years of experience was 110 and the number of "veterans" with greater than 20 years was 363. By contrast a similar study in 2003 pegged those numbers at 211 and 484 respectively. Based on this, while the changes may be "statistically significant" the sample size between the two studies alters the statistics.

Like Carey though, I don't object to the unionization of teachers on a conceptual level. My beef with the unions is their professed stance of being there "for the kids." Such a stance simply chafes me to no end. The unions are not there for the kids, indeed when push comes to shove in school politics, the kids are usually the last group to be considered. No unions have a duty to protect the interests of their members--the teachers. The teachers themselves may care about their students, but the union doesn't, shouldn't and can't. Unions have a duty to protect the teachers, first, last and only. "Protecting" students isn't their job.

When the union sticks to the business of protecting teachers, I will be willing to allow them a little more latitude in making their case (this is not say that I won't call them on studid assertions), but they can make their case. What I won't stand for, and what no one else should either, is the premise that their concern is for the students.


Anonymous said...

The change in the sample size does not "alter the statistics." That's the whole point of statistics, to account for things like different sample sizes.

Matt Johnston said...

Actually reducing the sample size does alter the statitistics. At the very minimum it increases your sampling error. Reducing your "newcomer" sample size by nearly 50% alters your margin of error by a significant factor (I can't remember the general increase).

Matt Johnston said...

Ooops forgot to include this. Let's assume that a sample size of 200 has a margin of error of +/- 4%(which may be accurate) and a sample size of 100 has a margin of error of +/- 7% (which also may be accurate) that means a difference in result on one question from one survey to the next could be as little as 3% or as much as 11%. When the response you get is say 35% believe in a position, you could have a range of between 24% and 46%, that is a very, very big range of error.

See, sample size does play a role, that is why samply size matters.