What’s Behind Standardized Graduation Rates? Data System Investment

Last week Secretary of Education Margaret Spellings announced that the Department of Education will begin requiring all states to use the same method to calculate high school graduation rates. NCLB already requires states and high schools to report graduation rates, but it allows states to craft their own formulas to do so. The result: states inevitably found ways to inflate graduation statistics. And the state-by-state patchwork of methods used makes it impossible to compare graduation rates across states.

Spellings’ announcement is an important, smart move following years of pressure from education and civil rights organizations to improve graduation rate data. Without comparable, meaningful data to expose low graduation rates, states can continue to ignore the drop-out crisis that is plaguing low-income communities—especially in urban areas—around the country.

But Spellings’ announcement also raises some important questions: Do states have in place the data systems they need to calculate new, standardized graduation rates? And, if not, how will they pay for new state data systems? So far, neither Spellings nor news articles covering the new regulations have addressed these issues in any detail.

Where State Data Systems Stand

The first step toward accurate graduation rates is a student unit record data system that can track individual students. These systems assign a unique identification number to each student in the state, so that the state can follow individual students from the time they enroll in ninth grade through high school graduation—even if they transfer between schools or school districts.

An accurate graduation rate formula (such the one adopted by the National Governor’s Association, which appears below) measures the percent of students from an entering ninth grade cohort who graduate with a standard diploma in four years. This is commonly known as a “cohort” method of calculating graduation rates.

According to the results of a 2007 survey by the Data Quality Campaign, 36 states have the full data system capability to calculate this formula. Seven other states have student unit records in place, but lack a robust data audit system. Robust audit systems ensure that districts report valid dropout data, and they are necessary to remedy past documented problems with unreliable district reporting.

All 43 states that have already invested in data systems to track individual high school students should be prepared to comply with new federal graduation rate measures within the next four years. Some states that implemented these data systems only recently may need time to accumulate four years of student data (from the time the first ninth grade class enrolls until they complete 12th grade) to calculate a cohort graduation rate.

But seven states that don’t have the necessary data infrastructure to calculate accurate graduate rates are going to have to play catch up—and that will likely be expensive.

Laggards Face Data System Costs

The exact costs of establishing a student unit record data system depend on the data infrastructure a state already has in place, but they are substantial. The Data Quality Campaign estimates that putting in place the structure for these systems costs between $1 million and $3 million annually over several years of development—not including ongoing maintenance costs after the system is in place or staff time at the district level.

These seven states may not be eager to spend money on K-12 data initiatives. One reason they don’t have them already is that they’ve been unwilling to pony up the necessary funds in the past. But soon they may not have a choice—although if the history of NCLB and IASA implementation is any guide, they will likely drag their feet. And they may complain about more “unfunded mandates” from the federal government.

NCLB does provide funding to help states cover the costs of developing and administering state assessments and standards—$409 million in fiscal year 2008, distributed to states based on their share of students ages 5 to 17. Once a state meets all of NCLB’s assessment requirements, it can use the grant money for “support for data reporting.” But given the limited amount of money available in this pot (average state grant: $7.6 million), many states don’t have a lot left over after annual testing costs. The Bush Administration’s 2009 budget request did not propose any increase in funding for state assessments. And Secretary Spellings hasn’t indicated that the Department of Education will provide additional funding to help states implement new graduation rate standards.

A Worthwhile Investment

States that haven’t already invested in student unit record data systems may not want to pay the costs to do so. But in the long run, student unit record data systems are an incredibly valuable investment that every state should make. If new, standardized graduation rate standards require lagging states to improve their data systems, that will be an additional benefit of these requirements.

Even states that already have the framework in place for graduation rates could do a lot to further develop their systems—for example, tracking all students from pre-kindergarten through higher education, or performing more extensive coding of student characteristics and outcomes. This type of data is invaluable in education research and accountability efforts.

That is, if states actually use it. According to the Editorial Projects in Education Research Center, only 16 states used a cohort formula for NCLB accountability purposes in 2006-07, while others preserved their old formulas even if they had accumulated the data to use a more accurate cohort formula. In many cases, the less accurate measures produce inflated graduation rates that make states look like they’re doing better than they actually are. Standardizing graduation rate measures will result in more accurate data—and make sure state taxpayers are getting their money’s worth from new data systems.