For a decade, the news from the city schools was good. Buildings might be dilapidated, deficits might bring schools to the brink of bankruptcy, and superintendents might be fired, but every summer, educators released test results standing next to charts that showed steady improvement. Baltimore was no longer the worst school system in the state.
But for the past three years, progress — as measured by test scores — has virtually stalled.
Critics of CEO Andrés Alonso say the lack of continued improvement shows that he has failed to make the nuts and bolts of teaching his focus. Another problem, they say, is the constant change within the school system — including high principal turnover — that has marked his tenure.
"He's just not an instructional reformer, and has done nothing to improve teaching since he's been here," said Jessica Shiller, a professor of education at Towson University. "There's been a lot of flip-flopping, moving things around, and turmoil in the system, all in the name of educational improvement. But structural changes, governance changes are just not going to improve achievement."
But other education experts say that such a plateau was expected. Early successes came after the school system made easy fixes, and now leaders must confront the hardest problems of educating children from high poverty neighborhoods with little home support, they say.
School board president Neil Duke said it may be time to assess whether Alonso's reforms have moved the system in the right direction.
"From a policy perspective, everything is going to be on the table," Duke said. "And it is going to be a broad discussion of what we've done, and ... whether our policy decisions have worked."
Alonso acknowledged that his reforms had more impact in his earlier years, and anticipated that sustaining it would be a challenge. He said some initiatives, such as focusing on the most-at-risk students, have tapered off and educators are in a time of transition.
"It's the cycle of improvement," Alonso said. "Once you do radical things that get you a bump right away, and that's not going to continue unless you do things in a completely different way. And we are now at a point where people have to learn to do much harder things."
The reform of Baltimore's schools began in 1997 with a city-state partnership and an infusion of state dollars. Almost immediately there was hope. By 1998, 70 percent of the schools were improving, and third-graders were making progress for the first time in four years.
That progress was documented by test scores nearly every year, on an array of exams from national standardized tests, the Maryland School Performance Assessment Program and then the Maryland School Assessments. In addition, the city's graduation rate improved from 43 percent in 1996 to 72 percent in 2011.
In the early years, educators used increased state funding to make basic changes, such as reducing class sizes, training every teacher in how to teach phonics, and instituting a new math and reading curriculum.The system also standardized instruction so a student who moved from the west to the east side wouldn't be lost.
Alonso took over in 2007 and and for the first two years, the system saw large gains in test scores. He closed 26 failing schools, negotiated a progressive teachers agreement, and allowed principals to control their own budgets and educational programs.
Alonso has acknowledged that some of the gains in test scores were inflated by cheating under his watch. For the past two years the system deployed an army of monitors to ensure test security and accuracy. The result was a decline in scores in 2011, and scores failed to rebound this year.
"The fact that we didn't get a bump this year was disappointing, but it was real, and the progress over time still held," he said.
But some believe that Alonso's first years of gains were actually the fruits of his predecessors' labor.
Jimmy Gittings, president of the city's principals union, said that most superintendents' results are realized two to three years after their reforms. Alonso, he said, is now confronting the impact of his own.
The two superintendents before Alonso — Charlene Cooper Boston and Bonnie Copeland — put programs in place that provided continuity and stability for Baltimore's highly mobile, and challenged students, he said.
"Yes, in the first three years of Dr. Alonso's administration test scores increased, but the reason for those dramatic increases was due to the fact that programs that [Boston and Copeland] implemented were having a positive impact on our children's achievement."
Gittings said that successful programs, such as a required phonics-based reading curriculum for grades pre-K through third, were abandoned under Alonso's plan to give principals more autonomy. Gittings said principals were told recently that they'd be moving back to streamlined curriculums in language arts and math.
"They are going back to what was already working," he said. "But it was at the expense of the cohort of those students who spent three years in the gap while they were experimenting."
Alonso said the idea that he was responsible for others' test scores was "nonsense." The progress made during his tenure — such a doubling of the number of students scoring advanced on reading tests over the last five years — cannot be downplayed.
Alonso also ordered an analysis of students' raw test scores, which he said shows that "kids might be missing the proficiency mark, but they're still moving." For example, in every grade, among students who didn't pass math and reading tests, scores improved from 2007 to 2012 — even though they didn't become proficient.
Some say the system's stall reflects a constant churn in leadership under Alonso.
Gittings said 85 percent of principals were forced out by Alonso, leaving the system with a deficit of experience, and a lack of leadership in the toughest schools.
Although national research shows that principal turnover can hurt student achievement, Alonso called it a "myth" in Baltimore. He produced data showing that on middle school math tests, principals with one year of experience posted larger gains than those with six or more years of experience.
A recent report on Baltimore said that the foundation of the system's reforms — accountability and autonomy — have been implemented inconsistently.
While Baltimore was identified as a model for cities attempting similar school reforms, Sarah Yatsko, who researched the district for the Center on Reinventing Public Education, said the system still has kinks to work out before results are seen.
"For Baltimore, autonomy has been the Achilles heel — it's been the toughest thing to implement," she said. "When it comes to the nuts and bolts of running a school, the messaging is very confusing across the district."
Although levels of poverty have not been held up as excuses for poor achievement in recent years, several experts said they cannot be ignored.
Abell Foundation president Robert Embry said the backgrounds of students may be an important factor in the plateau in test scores.
He noted that studies have shown that a major factor in how well students do is the income and education of their parents, and whom they share a classroom with. Low-income students in classes with students of higher socio-economic level generally do better.
"We have tougher kids to teach," Embry said. "As our students have gotten poorer, the challenge has gotten greater."
Shiller also suggested that the city's political leaders need to work with the school system to tackle issues that impact students and their parents, such as poverty and jobs.
Having test scores level off for several years is not unusual in cities nationwide, other experts say.
Jack Jennings, former president of the Center on Education Policy, said test scores tend to stall once teachers have gotten initial gains from implementing a curriculum that matches the state's and sets students up to do better on statewide tests.
"In a way it is the low-hanging fruit, it is the easy part. But then it gets harder," he said. "It doesn't mean it can't be done. It is like a never-ending struggle."
Melanie Hood-Wilson, a city schools parent, said the scores could reflect recent school evaluations, which showed substantial deficiencies in instruction.
"We, as parents, place too much emphasis on test scores and too little on understanding what quality teaching is and what a well-run school looks like," she said. "We, as parents, have to set the bar higher and make our expectations clear to our schools, our principals, and our teachers."
Tests scores in city school
•In 2003, 42 percent of third-graders passed the statewide math test. That rose to 78 percent in 2009 but dropped to 74 percent this year.
•In 2003, 44 percent of fifth-graders passed the reading tests. That rose to 82 percent in 2009 but dropped to 76 percent this year.
•In 2003, 31 percent of fifth-graders passed the math test. That rose to 75 percent in 2009 but dropped to 70 percent this year.