Don’t charter schools perform about the same as regular public schools?

We hear the question in this blog’s title a lot. It’s based on findings in a number of reports about charter schools such as the now famous (infamous?) 2009 CREDO study, “Multiple Choice: Charter School Performance in 16 States.”

The claim that only 17 percent of charter schools outperform traditional public schools comes from that 2009 study. This assertion is often made by school-choice opponents in Kentucky, including anti-charter opponents in the General Assembly.

But there is good evidence that most charter school reports are not studying charter schools properly, which biases the findings against these public schools of choice. Believe it or not, some of that evidence can even be found in CREDO’s 2009 study. Once we look at this evidence, the charter school advantage for students becomes quite apparent. In fact, because the quality of charter-school programs varies considerably between states, if we focus on a state with a notably better charter program, the advantage to students in those schools of choice becomes quite dramatic.

Critics certainly jumped on CREDO’s original 2009 study when it came out. They gleefully pointed out that the findings show 83 percent of charter schools don’t do any better than traditional public schools, and a large proportion of charters actually do worse.

Charter opponents again claimed backing for their position when CREDO issued an updated national charter study in 2013.But, the critics are only telling a part of the tale. The “Rest of the Story,” as the late Paul Harvey would have put it, helps explain why a lot of reports on charter schools –and even many sections in the CREDO studies – draw the wrong picture.

The basic issue isn’t complicated. Very simply, the majority of students who enter charters are often already several years behind academically. Since no one has a magic education wand that can turn those lagging kids into stellar scholars overnight, it simply takes time for charter schools to get lagging kids on track.

This basic fact isn’t rocket science. It isn’t a surprise, either. But most reports on charter schools simply ignore the basic truth that charters need time to work with students before benefits really show.

Not every report on charters has such flaws, however. Also in 2009, the Boston Foundation in conjunction with the Massachusetts Department of Education and researchers from Harvard University and the National Bureau of Economic Research issued “Informing the Debate, Comparing Boston’s Charter, Pilot and Traditional Schools.” This report compares charter school lottery winners to lottery losers in a sort of random-sample approach to develop a suitable comparison group of students in traditional public schools for those who got into charter schools in Boston, Massachusetts. Among other interesting findings, this report shows in Figure 4 that as students stay and progress through multiple grades in their charter school, their performance notably outstrips the comparison sample of students who lost the lottery and had to attend traditional public schools in Boston.

Another 2009 report from Carolyn Hoxby on charter performance in New York City, “How New York City’s Charter Schools Affect Achievement,” also uses the lottery approach. Hoxby also looked at the charter advantage over time. Like the Boston Foundation’s study, Hoxby’s Figures IVa and IVb show remarkable advantages in both math and English language arts accrue to charter school students once those kids spent a couple of years in their new schools of choice.

Now, here is something really amazing. The findings of much better charter performance over time are actually echoed in CREDO’s reports – even the first one from 2009. However, for unknown reasons, CREDO buries this finding deep within their report bodies. Most people who cite CREDO don’t seem to have noticed this information.

Figure 10 from the original 2009 CREDO report shows the relative performance of students in charter schools over time compared to average performance of students in traditional public schools. According to this part of the CREDO report, first-year charter students performed notably worse than students in traditional schools. The charter school students performed basically about the same as students in traditional public schools at the close of the second year of charter attendance. However, even in the 2009 study, once students had three years under their belt in a charter, they started to move out ahead. If we use a conversion for the Standard Deviation numbers shown in Figure 10 to days of extra learning found in later CREDO studies, it appears that charter students were about 65 days behind in math and 43 days behind in reading in the first year. By the third year, however, charter students were performing about 14 days ahead in reading and 22 days ahead in math.

This is a very important finding. It is data-based evidence that charter schools indeed do need some time to work with kids before real improvement can be expected. CREDO’s 2009 finding shows that charter studies that don’t allow for this time-is-needed fact of life will almost undoubtedly underestimate true charter-school performance.

Unfortunately, almost every report on charter schools makes this mistake, lumping all charter students together in their analysis, which guarantees bias in those studies against an accurate depiction of real charter school impacts. It’s an easy mistake to make.

Ironically, most of the findings even in the latest CREDO 2013 study include those first-year kids. But the 2013 CREDO study on charters shows dramatic evidence that the students-need-time principle still applies.

Furthermore, the charter impact over time, shown in Figure 42 in the 2013 nationwide CREDO study, is a lot more impressive compared to the 2009 study’s findings. CREDO writes in that 2013 report of charter students across the nation:

“Once a student is enrolled for four or more years, their learning gains outpace TPS by 50 days in reading and 43 days in math per year.” (Page 79)

In the 2013 update report, CREDO again shows that the first-year charter students are about 43 days behind in reading and 57 days in the hole in math. That’s almost identical to the 2009 findings for first year charter students. However, by year two in the 2013 report, charter students already are ahead of the TPS students, and they are actually ahead by amounts similar to those that charter students in their third year posted in the 2009 study.

Things only get better as time goes on. As previously noted, by year four or more of charter school attendance, even when averaged across all the states in CREDO’s study, the schools-of-choice students are about 43 days ahead in math and 50 days ahead in reading.

Thus, when the 2013 CREDO national update study averages performance across all states evaluated – some of which have done a poor job of implementing charter schools – the charter advantage over time is notable.

But, what happens when we look at impacts in a state that is well-known for better charter school laws and programs? Yet another CREDO 2013 report, “Charter School Performance in Louisiana, 8/8/2013,” shows truly dramatic results, as depicted in Figure 9 from that report.

Figure 9 in CREDO's 2013 report on Louisiana charter schools shows that state's charter students leaping ahead of their traditional-school counterparts even in the second year of attendance in their new school of choice, gaining an extra 85 days in math and around 78 more days of learning in reading.

In fact, the results for Louisiana’s charter students in year two are MUCH stronger than the results for year three students in the original CREDO 2009 study that covered 16 “states” (actually 15 states plus Washington, DC).

Things just keep getting better and better for Louisiana charter school students who hang around for even longer periods in their school of choice. CREDO estimates that by the time students have spent four or five years in Louisiana’s charters, their relative performance equates to about 180 extra days of school compared to Louisiana's traditional public school students. That’s a full extra academic year!

To sum up, the answer on charters’ better performance actually is dramatic, but it takes good research to find the truth.

Even if we rely on findings in CREDO reports, charters most certainly outperform if they are allowed enough time to work with the generally under-performing students who come to them from traditional public schools.

If we look at a state with better charter policies, the charter advantage over time is truly awesome. Clearly, the message in the Louisiana story is that Kentucky would be wise to adopt better-than-average charter school legislation and can probably expect superior results if that happens.

In closing, those who deny that charter schools have notable performance advantages are relying on out-of-date information or poorly crafted reports that don’t recognize something most parents and others understand very well – kids can’t change their academic performance overnight. It takes time and patience on the part of talented and enthusiastic people who really want and have the flexibility – to do the right thing to make it happen for those students.

And, flexibility to get the right teachers and to do the right thing is an advantage that charter schools have always enjoyed.(Editorial Updates January 14, 2015)