Going to the Numbers 
By Martin Harris

One of the rewards for getting past the Winter Solstice is seeing the days get longer. Another is getting my annual new (2005) copy of the National Digest of Educational Statistics (yes, I can appreciate that having the 2-inch-thick publication show up in their mail-box isn’t everyone’s make-my-day event) and dropping everything else while I browse through it. This year, particularly, I’ve been hearing a lot of "We’re Number 1" cheerleading by those involved in Vermont public education, and so I thought I’d go to the real numbers to get the real story.

Let me start by observing that getting your own copy of the NDES isn’t exactly an accomplishment. Just call your favorite constituent-service Congressman or Senator, and he’ll get one for you. Or call the National Center for Educational Statistics (1-877-4ED-PUBS) and they’ll send one. If you’re a property-tax payer (and it’s impossible not to be, renter or owner) you’ll doubtless be interested in how your school dollars ( 3/4 of your total tax, typically) are being spent.

Let’s start with the totals for 2002-3, shown by State in Table 164. The national average was $9299 per pupil. Six States spent more than Vermont, which came in at $11,075, New Hampshire at $9,802, Utah at $5,969. There’s also a "Current Expenditures" sub-category, which excludes capital expenditures and debt service, and therefore shows a lower number for Vermont --$10,545--  and some in the education business here prefer to recite that number when asked about costs.

Why is Utah’s per-pupil cost half of Vermont’s? Maybe because average class size (closely comparable to pupil-teacher ratio but sometimes not exactly the same thing) is twice as big. In Table 65 you’ll find that the national average is 15.9. Vermont came in at 11.3, compared to 22.4 for Utah. Ah, but they pay for that stinginess with lower student test scores, you  say.

Well, let’s go to Table 112, which reports State-by-State student results on the mandatory Federal test for 4th grade reading. The national average is 217 (out of 500), which means that about 2/3 of all students nationwide aren’t making "proficient", and aren’t able to function at grade level. Tops in the nation was Massachusetts at 231 (class size 13.6); Vermont and New Hampshire tied with 227, Utah lower at 221. Department of Defense schools came in at 226 (class size 13.3) while District of Columbia came in at the bottom at 191 (class size 13.8). Two observations might be made from these numbers: the first is that "We’re Number 2 or 3" isn’t much to cheer about when the actual numbers are so dismal, and the second is that there’s not much support for the smaller-classes-produce-better-student-achievement argument, in these data. A few States, Vermont included, don’t publicize these National Assessment of Educational  Progress test results;  they purchase, deploy, and publicize the results of local tests, for which scores somehow seem to be higher and opportunities for data-comparison with other states are much lower.

The proficiency percentages are also in Table 112, as straight numbers with no "nearly-achieved-the-standard" statistical tricks mixed in, as is done in some of the special-purchase tests. For 4th grade reading, "proficient or above",  they are:  US, 30%; MA, 44%; VT, 39%, UT, 34%. Results for 4th grade math, 8th grade reading and math, and so on are roughly comparable. Of course, a math proficient 4th grader can figure out that, if 44% of his classmates have made "proficient", then 56% --the majority-- haven’t. Those numbers are in Table 121 and reflect 2005 test scores. More next week.