The linings in the dark clouds over urban education are showing flashes of silver.

In Baltimore, they are barely visible, but they are there. Reading scores in the early grades are inching up -- as they have been for a while in many other big city systems -- as educators concentrate on instruction, blotting out distractions.

But there are no miracles here, and Baltimore is behind similar districts with large concentrations of poverty.

Still, the modest increases announced this week in the city's first- and second-grade standardized test scores bode well. They come only a year after the city replaced a hodgepodge of reading approaches -- every teacher a free-lance -- with a citywide program heavily tilted toward phonics.

Similarly, Baltimore County has seen a three-year improvement in reading test scores with an unprecedented concentration on early literacy. It's striking that the county, the nation's 25th largest district, has managed to keep itself well above national averages despite increasing poverty in many schools.

Is this the Lake Wobegon Effect -- a sort of educational inflation that makes every child, every school, above average?

Perhaps, but some national and local experts believe there's something at work unprecedented in the three decades since those heavy clouds gathered over urban systems.

"Generally, cities are doing better than they have in many years," says Paul D. Houston, executive director of the American Association of School Administrators, based in northern Virginia. "There's a lot more focus on achievement. The energy that's going into urban schools is unprecedented in the last 30 or 40 years."

Baltimore, for example, has most of the pieces in place for a full-scale assault on illiteracy.

This fall, some of the city's lowest performing schools will devote three hours a day to activities around reading. New mathematics books, the first in 10 years, will be in classrooms after a summer of concentrated teacher training in math.

Baltimore isn't alone in concentrating on fewer, but vital tasks. As Betty Morgan, Baltimore schools' chief academic officer, says, "We learned from the losers in World War II that it's a mistake to move on too many fronts at once."

Bonnie Copeland, the former city school board member now heading the Achievement First reform program at 28 Baltimore schools, believes that when the city's 1999 Maryland School Performance Assessment Program (MSPAP) scores are announced in winter, they'll outshine the results of last fall's Comprehensive Test of Basic Skills (CTBS), a commercial test in which city kids are graded against a national norm.

The MSPAP results remain bleak: Only 16 percent of city schools performed satisfactorily last year. But Baltimore scores rose for the first time in four years. The city is wise to gear instruction to MSPAP, the test on which its fortunes rise or fall with the state and the only instrument with which to accurately compare city kids with others in Maryland.

Other cities improve

Urban scores around the country are improving, not only on state tests like MSPAP, but on standardized tests like the CTBS. Kids who live in cities with high rates of poverty don't differ much from each other. The cry these days in the cities is that good instruction is better than a good neighborhood -- a new and potentially powerful notion if carried through in actions.

Consider these 1999 statistics:

El Paso: Between 1996 and 1999, passing rates on the Texas Assessment of Academic Skills (TAAS) moved from the 60 percent range to the 80 percent range, largely because of a communitywide effort known as the El Paso Collaborative for Academic Excellence.

Chicago: Students scoring at or above the national norm in reading and math on the Iowa Test of Basic Skills (a nationally standardized test much like the CTBS) increased to the highest levels in a decade.

Dallas: Sixty percent of all students passed the Texas test, up from 55 percent last year. TAAS is an easier test than MSPAP, but what is significant in Texas is that African-American students are doing very well on the state test.

Houston: African-American students this year surpassed the state passing rate on TASS by three percentage points.

Sacramento: The district increased its first-grade reading scores on the Stanford 9 Achievement Test (another standardized test like CTBS) from the 54th percentile in 1998 to the 62nd percentile. Sacramento finished its third year with the same reading program, by the Open Court company, installed last year in all of Baltimore's primary grades, but the district was helped by a $3 million private grant to hire 50 literacy coaches.

Seattle: Scores on the Iowa test increased six points in reading between 1996 and 1999. African-American children in grades three and five scored nine points higher in math this year than in prior years and also made strong gains in reading and language.

Comparing notes

Another factor in the rejuvenation of city districts is that educators are comparing notes on triumphs and failures, and they're doing so not only with each other, but with mayors, civil rights leaders and others outside education.

Baltimore has just joined Chicago, Denver, New York City, Philadelphia and Seattle in a "Cross City Campaign for Urban School Reform," which will meet here this year. The Baltimore schools also will play host to a delegation from El Paso next month.

"We're all learning from each other," says Copeland, whose Achievement First approach is borrowed from a successful model in New York City. "They're saying, 'Come and see what we've done,' and so are we."

Scores overemphasized?

There's a down side, though, Houston of the administrators' group believes.

"Urban districts," he says, "had been dying from test scores. Now they're starting to live on them. I worry about the concentration on scores. We don't measure much in education, so test scores are all we have, but there's more to education than test scores."

The cities are doing better, Houston says, "but at what price?"

In the stampede to improve scores, there is a danger well-thought-out reading instruction will be trampled by "drill and kill" and that schools will become less exciting and welcoming.

And there's still the problem of the upper grades.

Even Maryland's affluent suburban middle schools are turning out mediocre reading scores. In the city, those scores are dreadful.

It's urgent that children learn to read by age 9, but how well they read at high school graduation is just as urgent.