What's the difference? Many thousands of younger immigrant kids have had the benefit of two years in English-immersion classes, while the bulk of the older limited-English students spent their most formative years in bilingual classes where they never were forced to learn the chief language of this country.
Some analysts might suggest that the smaller classes pupils began enjoying in 1997 are also a factor. But middle-schoolers enjoyed smaller classes for two years while still in the primary grades before the bilingual ban took effect. If that were the chief factor in performance improvement on these tests, they should have improved as much as kids on the elementary levels. They did not.
So Proposition 227 has achieved much of what it set out to do. And the California results — showing reading and math scores up 12 to 20 percentiles among the state's third-graders over three years, and virtually no change among high schoolers — are already encouraging efforts to win bilingual bans in many other states, including Arizona, Colorado and Massachusetts. A fledgling effort is under way in New York.
Improvement among the general primary-level student population ranged around 6 to 8 percentile points on the standardized Stanford 9 test, which is pretty good. Improvement is even greater among English learners than among the general student population. Their reading scores are up 12 points over the last three years; in math, they're up 17 points. That's a pretty good trend.
Even if there's been some cheating by teachers providing students with some questions in advance, as critics of the state testing program loudly claim, the improvements among immigrants are still highly significant. Kids who haven't learned English don't figure to get much benefit from hearing English-language questions ahead of time.
Ron Unz, the author and financial sponsor of 227, says the nice gains shown in official records may actually understate reality.
"Increasing numbers of the most successful English learners have now learned English sufficiently well that they no longer fall into this category," he said.
Their scores are listed among those of ordinary English-speaking students.
"Obviously, if you remove large numbers of the most successful students from any category, you will tend to retard its apparent average gains," he adds.
Unz notes that no news outlet greeted this year's numbers with any great surprise or large headlines.
"The continued rise of California immigrant test scores has become part of the expected norm," he said. "It's greeted with less and less surprise every year. You can be sure if test scores ever dropped significantly for these kids, the headlines would be very large."
That isn't like for the massive curriculum change Proposition 227 forced on California educators won't go away. And the more immigrant children are compelled to learn English in the earliest grades, the better their scores will look as they move up through the educational system.
It's a classic example of the voting public knowing better than state legislators and so-called experts, who opposed the bilingual education ban almost unanimously. No wonder the most important laws in California are passed at the ballot box and not in Sacramento.