12/04/2007 07:33 pm ET Updated May 25, 2011


There sure is an eerie silence emanating from the U. S. Department of Education these days. When NAEP reading trends appeared earlier, Margaret Spellings told everyone (over and over and over) that "scores for 9-year-olds have risen more in the last 5 years than in the previous 28 years combined." She claimed NCLB as the cause. The fact that NCLB didn't exist for most of that time didn't figure in her claim.

But more recently, a new round of PIRLS (Progress in International Reading Literacy Study) has appeared and while American kids scored the same as they had five years earlier, a number of other countries had advanced and whereas the Americans had been 8th in 2001, they were now 15th (of 39 nations). Some of the scores are suspect. Do you think Russian kids are the best readers in the world? Me neither. But there they are atop the ratings. The official explanation is that Russia has added another grade at the beginning of school (starting at age 6, not 7), but still, their 37 point gain since 2001 is, well, unusual.

"Clearly, as the world becomes flatter, it's becoming more competitive. We need to do better than simply keep pace." That's Spellings statement on the matter in a very brief press release. Wait. These are 9-year-olds. We should judge our competitiveness on the reading test scores of 9-year-olds? The World Economic Forum doesn't do that and in its latest Global Competitiveness Report, the U. S. jumped back to first place, a position it has held most of this century. Education is only one of 12 factors ("pillars," the WEF calls them) that go into the 131 rankings (Amazon advises that the latest report won't be shipped until mid January--it appeared December 1--but you can see the rankings here). But as I haven't written often, and to little avail, schools are the whipping boy for any perceived problem, and facts are irrelevant.

The press release also says, "the U. S. score has not changed measurably from 2001. While we're seeing progress under No Child Left Behind, we can do better." Is the sound you just heard the Department of Education building a massive wall against reality? If we're making progress, how come it doesn't show up on PIRLS?

I imagine that Spellings and other NCLB advocates must also be wringing their hands over the gaps among ethnic groups in this country. In addition to having 100% of all children reach proficiency, NCLB was supposed to reduce or eliminate these gaps. Yet, 4 years into the law (the tests were administered in 2006), the gaps remain large. The scores:

Whites 560

Blacks 503

Hispanics 518

Asians 567

Am. Indian 468

(International Average = 500)

It is true that the largest 4 ethnic groups finish about the international average, but this is not as hopeful a sign as it first appears. The highest scoring nation scored 64 points above the average, the lowest scored 198 points below. The extreme low scorers pull the average down. A nation that actually scored 500 would rank 29th out of 39.

(None of my analyses include the scores for Canadian provinces. Typically, not all provinces participate so that there is no national Canadian score, but it seems inappropriate to use one unit of analysis (the nation) for most scores and another unit of analysis (the province) for some.

We can get further idea of the size of the gaps by imagining the various groups as nations and seeing what ranks they would hold matched up against the 39 countries that took part in the study:

Whites 3rd

Blacks 28th

Hispanics 25th

Asians 1st

Am. Indian 35

From this table of ranks we can see that any document that speaks about the performance of "American Schools" is addressing an institution that does not really exist.

For some reason, NCES abandoned its usual procedure of reporting scores by the level of school poverty and reported a grosser indicator on whether the schools had no children receiving free lunches, some receiving or all receiving. Albeit not a very refined measure it shows huge differences, too: None = 586, Some = 543, All = 486. The proximity of the "Some" category to the overall U. S. average (540) reflects the fact that this category contains the overwhelming bulk of the schools.

The trumpeting of NAEP and the shushing of PIRLS makes USDOE look like an organization of cherry pickers.