NCLB Graduation Accountability
This brief reports the results of a new Editorial Projects in Education Research Center analysis of state policies related to the implementation of high school graduation accountability under the federal No Child Left Behind Act. Graduation rates are a required academic indicator under the law’s provisions for defining and measuring adequate yearly progres, or AYP, for high schools and districts. States must calculate and report graduation rates for all students and required subgroups, must establish performance targets for graduation, and must incorporate this information into their procedures for determining AYP.
Results of this analysis reveal that states are beginning to respond to mounting pressure to produce more accurate graduation rates. However, states continue to exercise substantial autonomy when deciding exactly how they will respond to the mandates of the federal law.
That latitude, along with sometimes-sizeable differences between state-reported rates and estimates from independent researchers, has raised questions about whether state-reported graduation rates are accurately capturing the magnitude of the dropout problem. Research has shown that many states significantly underestimate the number of students graduating each year.
In 2003-04, the average state reported a graduation rate of 83 percent to the U.S. Department of Education. By comparison, the EPE Research Center’s Cumulative Promotion Index estimated that only 70 percent of public school students graduated with a standard diploma that same year. The CPI rates for minorities are lower still, with about 53 percent of black students and 58 percent of Hispanic students graduating in 2004.
In 2005, the National Governors Association unveiled its Graduation Rate Compact. Signed by the governors of all 50 states, the Compact served as an important impetus encouraging states to produce more realistic and consistent graduation rates. The Compact requires states to work toward implementing a graduation rate calculation that tracks individual students from 9th grade to graduation and accounts for transfers in and out of the system. It also sets forth a strict definition of a high school graduate and rules for assigning certain students to different cohorts.
Calculating Graduation Rates
For the 2006-07 school year, an EPE Research Center analysis of state accountability workbooks and supplemental sources finds that 16 states are using a cohort rate for calculating graduation rates under NCLB. This method allows for the accurate determination of an entering high school cohort and movement into and out of school systems. A cohort rate based on a statewide data system typically satisfies the NGA Compact’s requirements for calculating a graduation rate. Statewide data are employed by 14 of the 16 states that have implemented a cohort method.
During the 2005-06 school year, 10 states were employing a cohort rate. This indicates a substantial amount of progress during the past year. In addition, several states moving to cohort rates between 2006 and 2007 replaced methods generally viewed to overestimate the graduation rate by considerable margins.
For example, Indiana adopted a cohort rate for 2007, replacing its persistence-rate calculation. That earlier method estimated the percent of students not dropping out from grade to grade during high school. It should be noted that a persistence rate does not actually measure high school completion. Using the cohort calculation, Indiana reported a 76.5 percent graduation rate for the class of 2006, compared to a rate of nearly 90 percent calculated for the 2005 graduating class using the persistence rate.
This comparison illustrates a major challenge states may face as they institute new graduation-rate methods. State policymakers may be reluctant to adopt new calculation approaches if there is a perception that such a step could negatively impact state-reported graduation rates.
Louisiana and North Carolina’s move to a cohort rate this year marks the extinction of two graduation-rate methods that had come under scrutiny. In 2006, Louisiana was the only state using a dropout rate for purposes of NCLB accountability, the provisions of which explicitly require states to employ a graduation rate. Similarly, North Carolina abandoned its on-time rate calculation, which reported the proportion of all high school graduates in a given year who receive a standard diploma in the expected amount of time (as opposed to taking more than four years to finish high school). Sizeable shifts in publicly reported graduation figures can be expected in both states as a result of their new methods, an expectation already confirmed in North Carolina. The state reported a cohort rate of 68.1 percent for the class of 2006, a precipitous drop from the state’s on-time rate of 96.1 percent in 2005.
While many states are working to implement a cohort based calculation, the leaver rate remains the most common approach to measuring graduation rates, with 32 states employing this method for 2006-07. The leaver rate expresses the percent of students leaving high school with a standard diploma as a proportion of all those documented as leaving with a diploma or other completion credential or as a dropout.
Not only do the states use a wide variety of methods to calculate graduation rates, but differences also exist in the specific ways that states may implement the same method. As a result, it can be difficult if not impossible to compare graduation rates across states or to ascertain the magnitude of the dropout problem using state reported statistics. A key goal of the NGA Compact and the work of such groups as the Data Quality Campaign is to increase the comparability of state graduation rates and improve the overall quality of educational data systems.
Setting Goals for Graduation
As noted above, some states are working to implement more reliable methods for calculating graduation rates. Under the terms of NCLB, states are also allowed to decide how high or low to set their sights for graduation rates and to determine how much progress is sufficient for a school to make AYP. Accordingly, graduation-rate goals have been established at very ambitious levels in some states and at less-than ambitious levels in others. Some critics have argued that the discretion states possess over goal-setting for graduation rates considerably weakens this provision of the law.
As defined in this brief, a state’s “current target” is the graduation rate high schools must achieve to make AYP on the graduation rate indicator for the 2006-07 school year. The average graduation target for the current year is 75 percent, down from 77 percent in 2005-06. Targets ranged from a high of 95 percent in Indiana to a low of 50 percent in Nevada. Graduation targets remained stable in 36 states between 2006 and 2007.
Some states, however, substantially reduced their target compared to last year. North Dakota had the most drastic drop, setting the 2006-07 target at 73.09 percent compared to 89.9 percent for the previous year. Massachusetts dropped its target from 70 percent to 55 percent. Such reductions were sometimes the result of changes in a state’s method for calculating graduation rates. For instance, states implementing a cohort method might consider recalibrating targets, since graduation rates calculated using this approach are often lower than those produced by other calculations.
The EPE Research Center analysis also identified the final graduation target that schools will be expected to achieve by 2013-14. Only three states set this goal at 100 percent, the ultimate target for the percent of students academically proficient required under NCLB. Nevada has the lowest final graduation target, maintaining its current target of 50 percent through the 2013-14 school year. In all, 32 states do not plan to raise their graduation goals between 2006-07 and 2013-14. Ohio shows the most ambitious increase in targets, raising its current goal of 73.6 percent to 100 percent by 2013-14.
Under NCLB, states are allowed to set criteria for the amount of annual improvement that schools failing to meet the current graduation target must demonstrate in order to make AYP. Those criteria are generally quite lenient. In 30 states, any amount of improvement, no matter how small, allows a school to make adequate yearly progress. An additional four states require annual improvement of just one-tenth of one percentage point annually for schools below the graduation goal. To make AYP in Delaware, New Mexico, and South Carolina graduation rates do not have to improve, they must simply hold steady from one year to the next.
Although the No Child Left Behind Act does not require states to implement sophisticated data systems capable of following individual students over time, such systems have been advocated by a number of national organizations. As a result, many states are in the process of developing such capabilities, in part as a way to generate more accurate information on high school dropout and graduation.
To assess the state of state data systems, the EPE Research Center examined data from a 2006 survey conducted by the Data Quality Campaign. According to the DQC, 29 state data systems currently have the four key elements necessary for calculating a graduation rate consistent with the NGA Compact: student identifiers; student-level enrollment, demographic, and program participation data; student-level graduation and dropout data; and a state audit system.
For many states, however, progress on data-system development is quite recent. In fact, only 18 of those 29 states have had student-identifier systems in place for at least four years, which would allow them to calculate a true cohort rate for the class of 2007. The EPE Research Center analysis also finds that fewer than half (eight) of those states intend to employ a cohort-based rate for purposes of NCLB accountability for 2006-07.
Prior to the passage of NCLB, a small number of states calculated graduation rates and used those indicators for their own accountability or public-information purposes. Some states currently report multiple graduation-rate indicators that might, alternatively: employ different calculation methods (e.g., a leaver versus a cohort rate); involve different time frames (e.g., a four-year versus a five-year rate of completion); or count recipients of different credentials as graduates (e.g., diplomas, alternative certificates, or GEDs).
Minnesota, for example, uses a leaver rate to comply with NCLB but also recently reported a statewide graduation rate for the class of 2006 that is consistent with the NGA Compact. Texas reports a variety of dropout and graduation indicators, such as alternative completion rates that include students who were still enrolled in school or had received a GED at the expected time of graduation.
The past few years have witnessed rising pressures on states to produce more accurate graduation rates. While states are responding to those pressures, a review of state progress in implementing policies on graduation accountability shows that there is still a significant amount of work to be done.
Increasing numbers of states are now using a cohort based method to more accurately calculate graduation rates. However, some states capable of such a measure have not yet taken the step to employ a cohort rate for NCLB accountability. Apart from the measurement of graduation rates, the great variability in the goals states set for high school graduation continue to raise concerns regarding the practical consequences that graduation rates may have when determining AYP.