In an article on Bankrate, “Teachers Lose Battle With Industry Giants” the leader of NTSAA was quoted with regard to a TIAA study on fees in open vendor states, the article says:
NTSAA’s DeGrassi says that study is flawed because it’s based on a limited data set. “The only other statewide data points for the study are Texas and Iowa. Data from three very different states is not a national study,” he says.
I understand Mr. DeGrassi was attempting to debunk the study, which can be read here, however he inadvertently provided the fodder to debunk a “study” that his own organization put out last year whose premise was that reducing vendors in public school 403(b) plans led to reduced participation. The study was based on a county (with 55 school districts) in California, and single school districts in both Colorado and Pennsylvania.
Evidently, NTSAA doesn’t see the contradiction. NTSAA put out a “study” that had absolutely no statistical significance since the sample population was nearly non-existent (a county and two school districts does not significance make in stats) and they have led people, congressional leaders and state legislatures to believe that the study is significant. I again challenge the NTSAA to submit their study to a refereed journal and note how quickly it is rejected.
I’d like to point out that the TIAA study does use a sample size that is likely statistically significant because it use the entire population – there are only three states that have open vendor laws and all three were included, how exactly could more be included if they didn’t exist?
As if the devastating letter from the consultant to Jefferson County Public schools wasn’t enough to debunk the NTSAA “study” (read here), now the NTSAA leadership itself has confirmed the study should not be taken seriously, after all, it only included data from “..three very different states…” and is thus, “..not a national study.”
Scott Dauenhauer, CFP, MSFP, AIF