Reports that 88% of California's 6,700 elementary, middle and high schools failed to meet the state's Academic Performance Index (API) goals stirred new calls for reform. Yet, the results conceal even more troubling issues. As the "new economy" spawns unprecedented disparities in wealth, social class increasingly determines academic achievement. Public schools cannot hope to improve unless the markedly unbalanced, socially divisive economic-development patterns transforming society are also corrected.

At first blush, the API rankings, which are based on last year's Stanford 9 test results, appear to be more about ethnicity than economics. Statistically, school ratings most closely correlate with the ratio of white and Latino students, the two groups that make up more than 82% of the K-12 population. API scores increase sharply as the proportion of white students rises, but fall as the ratio of Latinos grows.

But closer inspection shows that such ethnic data reflect the growing socioeconomic divide between white and Latino populations. White students tend to come from wealthier families with a long histories of education and economic achievement. Most Latinos are recent immigrants. Many are just beginning the multigenerational struggle for advancement.

API results are profoundly shaped by these facts. Surveys used to compute the rankings show that nearly 70% of the parents of Latino students have a high-school education or less; just 30% attended college. In contrast, 80% of all white pupils have a parent with at least some college experience. Half of all Latino elementary students have limited English proficiency, compared with just 2% of white pupils. About 65% qualify for subsidized student lunches versus 19% for white students.

These socioeconomic differentials translate into hugely disparate academic outcomes. The average API for socially disadvantaged students--those qualifying for subsidized lunches or whose parents didn't graduate from high school--was 499, a staggering 118 points below the statewide average. Combined reading and math test scores for 4th-grade Latinos whose parents didn't attend college were about 50 points lower than for students from families with such experience. Language-challenged or poorer Latinos and whites scored from 30-50 points worse than their counterparts with better English skills and higher family incomes.

Why do affluent students from well-educated families typically do better? Wealthier children face far fewer pressures to work during school or to quit before graduation to earn a living. Better- educated parents understand schoolwork and thus are able to help with homework.

Numerous studies also show that education spending is skewed in favor of wealthier communities. API rankings markedly drop, for example, as a school's percentage of noncredentialed teachers rises. Students from poorer families are far more likely to be taught by such teachers. A greater proportion of socially disadvantaged and Latino pupils are shoehorned into year-round schools, whose average API was 22 points lower than for more affluent campuses.

In the past, balanced economic development allowed America to accommodate similar wealth and education disparities. Early 19th-century European immigrants, for example, were absorbed by, and later controlled, many of the country's once-flourishing manufacturing industries. As their economic status rose, their children and grandchildren became progressively better educated, eventually shifting into white-collar service and professional occupations.

Recent immigrants have generally followed this pattern. But during the 1990s, California's political and economic climate, as in many other regions, shifted in favor of white-collar, cyber-economic development. Urban policies neglected, if not shunned, other forms of economic growth, including inner-city school construction, on which the state's aspiring working classes historically depended. Influential politicians from privileged coastal and Bay Area enclaves resisted legislation making it easier to redevelop former industrial sites throughout immigrant-rich communities. Eventually, almost all the state's critical urban revitalization needs were subordinated to environmental and other new-economy concerns.

The California economy sharply changed course. During 1990-99, the state generated 1.6 million jobs, a 13% growth rate far below the national norm. Only 23% of this expansion occurred in socially diverse cities--Los Angeles, San Francisco and San Jose--and Orange County, by far the most anemic urban growth in California history. More than 65% of all new jobs were divided among very high-end professional and extremely low-wage personal service employment; 120,000 manufacturing positions evaporated from the state.

Working-class wages stagnated relative to those of elite, white-collar occupations. The personal incomes of security and commodity brokers in California, for example, rose three times faster in 1993-1997 (the latest years for which data are available) than in the state's manufacturing and construction sectors. While dot-com fluff filled the news, California generated far fewer of the manufacturing, construction and trade employment opportunities so critical to its blue-collar communities than ever before.

All this comes at a time when academic performance is touted as the foundation of financial success. Yet, the economic prerequisites for rising academic attainment from generation to generation are being eroded in California. New-economy wealth disparities are generating an increasingly intractable education and social divide.

There is, however, another possibility. For several years, Texas, a state with similar education demographics, has outperformed California in academic testing. The biggest difference is that Texas' Latinos do far better as a group than in California. In recent national reading tests, for example, white Texas eighth-graders scored just four points higher than white California students, but Latinos scored 13 points higher. About 66% of Texas' Latinos met basic reading standards, compared with 48% in California, the largest disparity of any group.

One reason why Texas does well is its greater proportion of native-born Latinos. Another may be the state's supposedly tougher academic standards. Perhaps the most crucial, however, is Texas' incomparably more equitable and robust economic development.

In the last decade, Texas grew nearly 2.5 times faster than California, producing nearly 500,000 more jobs from a much smaller base. Services accounted for 41% of all Texas new employment, much less than in California. The state's blue-collar manufacturing and construction sectors added more than 274,000 new positions versus just 39,000 in California. Per-capita personal incomes rose much faster in Texas than in California in every major employment category except financial services.

Texas' core urban communities, Dallas-Ft. Worth, Houston and San Antonio, absorbed 63% of the state's total growth, nearly three times the proportion of new jobs generated in California's major urban communities. Even Austin, the Texas city cherished by new-economy elites, produced about as many new construction and manufacturing jobs as service positions.

Texas' experience raises difficult questions. If California's economic imbalance continues to worsen, most of the popular nostrums for improving public schools will almost certainly fail. Already, the API figures show that reducing class size, a widely touted solution for poor student performance, had no discernible effect on academic achievement scores. Economic status simply overwhelms classroom innovation.

Should present trends continue, moreover, California will have to invent a mechanism for assuring working-class social mobility. Previously, education achievement and broad-based, multisector economic development went hand in hand. Today, the "old economy" is scorned in the belief that time has passed it by. Consequently, an entire generation who once relied on its expansion to improve their lives may be stranded.

Perhaps this can't be helped. Maybe Texas is just an economic anachronism that the United States will soon outgrow. That still leaves the riddle of how aspiring classes will compete in a world in which academic achievement is paramount but that lacks the material prerequisites for doing so. As California's sobering API data demonstrate, it is this gulf, much more than computer or Internet access, that defines the true digital divide we must somehow bridge.