Standard & Poor's Puts Michigan Data Under Microscope
The Wall Street giant Standard & Poor's has released the first of what will be an annual report on the condition of Michigan's public schools, providing a taste of the kind of data-driven analysis it hopes to market to other states.
"Statewide Insights," issued last month, is based on the company's analysis of some 100 million pieces of district-by-district information, ranging from test scores to spending patterns to per- capita income, for school years 1996-97 through 1998-99. The company first put those data on the Internet last May. ("Mich. Districts Get Report Card From Wall St.," May 30, 2001.)
The summary of trends across districts found that while the composite passing rate on the state test—the Michigan Educational Assessment Program—increased by 6.7 percent over the three-year period, the participation rate dropped 5.5 percent, "drawing into question the true magnitude of the perceived improvement in test scores."
The report also found large gaps in the passing rates by race and economic status. In 1999, 59.9 percent of white, non-Hispanic students met state standards on the MEAP, compared with 31 percent of black students, 36.8 percent of Hispanic students, and 33.2 percent of Native American students. Similarly, only 40 percent of test scores for students from poor families met state standards, compared with 56.2 percent for students from wealthier families.
"The gaps are, in some cases, quite substantial, overall, across the state," said William J. Cox, the project manager for the school evaluation division of the New York City-based company. "It's the first time that the information is being published for every single school district in Michigan."
Under the newly revised federal Elementary and Secondary Education Act, districts and schools nationwide are to be held accountable for improving the test scores of students by subgroup, including race and income.
In the 16,500-student Ann Arbor, Mich., public schools, the overall passing rate was 67.4 percent, well above the state average. But the proportion of MEAP tests passed by economically disadvantaged students was only 24.9 percent, compared with 69 percent for better-off students. That achievement gap of 44 percentage points is much greater than the state average of 12.7 percent, and higher than the achievement gap for districts with comparable student populations.
Rossi Ray-Taylor, the district's superintendent, said the system had been working to close the achievement gap for several years. "We are part of a nationwide coalition of districts that are very much like Ann Arbor, very high-performing districts with large gaps," she said. "So it's helpful to know, but it's not exactly new to us."
She said that, overall, she found the S&P analysis helpful. Ann Arbor is interested in benchmarking its performance against that of other districts in the state, she noted, "so this assists us in doing that." But she cautioned that some of the data still needed to be refined so that they were truly comparable and timely across sites.
Release of Standard & Poor's report came the same month that state Superintendent of Public Instruction Thomas D. Watkins Jr. unveiled a new accountability framework for Michigan's public schools.
"One of our main focuses within the state's new accountability system is to make sure that it's aligned with the information provided by S&P," said T.J. Bucholz, a spokesman for the Michigan education department, "in that S&P provides multiple measures of school district success."
While money matters, the state summary concludes, money alone won't enable the state to meets its achievement goals.
The report found that districts with the same operating expenditures per student had widely varied passing rates on the state tests. For example, in one of every four districts, fewer than half the MEAP test scores met state standards in 1999. Yet per-pupil expenditures were above average in one-third of them.
"More fundamental and perhaps structural changes may be needed to maximize the state's return on incremental spending," the report says.
"For the first time, an independent source has put together vital information showing that increased funding alone will not meet our educational goals," Gov. John Engler, a Republican, said in a statement.
But while the report provides some useful information about the state's public schools, it's also limited by the gaps in state record-keeping. For example, the report notes that because Michigan does not track the progress of individual students over time, and tests students only periodically in each subject, "MEAP performance trends, whether positive or negative, are somewhat questionable."
The report also raises serious questions about the state's ability to break down test scores by race and income, because many students neglect to identify their race and ethnicity on test-response forms, as requested. Similarly, administrators often fail to provide information on student characteristics or fill out such information inconsistently.
Mr. Cox of S&P speculated that if the data were available for all students, gaps in achievement for different subgroups could be bigger.
Despite the achievement gaps between and within districts, the analysis found that many schools and districts were already "beating the odds." The report identifies 292 schools in 144 districts that posted above-average MEAP results, even though they serve above-average levels of students from poor families.
Many of those schools were in Detroit and other urban areas. The report suggests that those sites are worth a closer look in the search for effective practices that might be copied.
Standard & Poor's plans to unveil school-level reports in Michigan in the first quarter of this year. The company had been waiting until financial data were available on individual schools. But so many requests for building-level information have been made, Mr. Cox said, that S&P decided to go ahead without such data.
The company will release a similar annual report on public education in Pennsylvania later this year, and will post 1999-2000 data on Michigan districts on its Web site later this month. It also is negotiating with several other states to provide the service, which will cost Michigan $11 million over five years.
Vol. 21, Issue 16, Pages 16,19