Properly Counting Indiana’s Voucher Students, Step by Step

The Indiana Department of Education (IDOE) reported that 39.3 percent of voucher recipients (7,779 students) were never previously enrolled in public schools in Indiana, implying that these vouchers are an additional fiscal cost to the state because, as many are interpreting it, “those kids would have paid their way to a private school with or without a voucher.” This is simply inaccurate, and here’s why.
Buried in the 7,779 students reported by the IDOE as “never attending a public school in Indiana” are voucher students in the following four groups:

1. Kindergarten Eligible Tax Credit Scholarship Students – Students who first enrolled in kindergarten at a private school using the School Scholarship Tax Credit program, then shifted to the voucher program. This group is a very large portion of the total “Previous SGO Award” pathway voucher students. The rest of the “Previous SGO Award” pathway voucher students (all other grades) had to first attend a public school to be eligible to receive their original tax credit before shifting to a voucher.
2. Kindergarten (and some first grade) “Failing School” Pathway Students – Incoming students who were never previously enrolled in public school in Indiana, but were zoned to a failing one. A portion of these first-grade students did not attend kindergarten at all (not mandatory) or were previously enrolled at a subsidized daycare center serving children of pre-K and kindergarten ages, but no further.
3. Kindergarten (and some first grade) “Sibling” Pathway Students – Incoming students who were never previously enrolled in public school, but whose sibling is already part of the voucher program. A portion of these first-grade students did not attend kindergarten at all (not mandatory) or were previously enrolled at a subsidized daycare center serving children of pre-K and kindergarten ages, but no further.
4. Grades 2-12 “Failing School” Pathway Students New to Indiana – Some incoming students in grades 2-12 have moved to Indiana from other states and were never enrolled in public school in Indiana, but were zoned for a failing one.

In all of these cases, some portion of these students would now be enrolled in the public school system if not for the voucher they are now receiving to attend a private school.

Furthermore, it is fair to presume most of the students in all four groups have been diverted from the public school system, whether it can be documented or not. Remember, the Choice Scholarship Program is means-tested, with vouchers awarded only to students from low- and some middle-income families.

Here’s why that 39.3 percent figure (7,779 kids) is actually more like 19.1 percent (3,779 kids)—and even that is an overestimate.

Calculations

Start with sub-group #1 from above, Kindergarten Eligible Tax Credit Scholarship Students, by far the largest of the four sub-groups buried in IDOE’s calculations of 7,779 voucher students or 39.3 percent “never attending a public school in Indiana.” Then, peel away the noise to estimate how many of these students now would likely be enrolled in a public school if their voucher was revoked.

Our calculation of this estimate for sub-group #1 is laid out in substantial detail via STEPS 1-5 below:

Over the past three years, a total of 4,927 students first qualified for a Choice Scholarship (voucher) using the “Previous SGO Award” pathway. Table 10 in the IDOE report shows that 532 students followed this path in 2011-12, and 1,972 more in 2012-13. Table 11 shows 2,423 more students used this path in 2013-14. So, using those three years, the arithmetic looks like this:

STEP 1a: 532 + 1,972 + 2,423 = 4,927 “Previous SGO Award” pathway students (original count)

Applying the 77 percent retention rate (see IDOE’s Table 17) to the 2,504 (532 + 1,972) students using this path in the first two years, we can estimate that 4,351 of the 4,927 “Previous SGO Award” pathway students are still in the current Choice Scholarship recipient count. Here’s the math:

STEP 1b: 2,504 x 77 percent = 1,928 –> 1,928 + 2,423 (year three count) = 4,351 “Previous SGO Award” pathway students (retention count)

Of that total, the IDOE report already excludes 640 students from its 39.3 percent “No Prior Public School” share (see Table 14). Thus:

STEP 2: 4,351 (from STEP 1b) – 640 = 3,711 “Previous SGO Award” pathway students are counted in IDOE 39.3% “No Prior Public School” share

In other words, about 3,711 of the 7,779 voucher students IDOE reported as “never attending a public school in Indiana” are students who switched from one of school choice program (tax credit scholarship) to the other (voucher).  They first enrolled in a private school with the financial help of a tax credit scholarship then subsequently switched over the voucher program when they became eligible. It’s fair to say that most of these kids would have been enrolled in a public school, if they had not receive this financial assistance.

Now, to be cautious, we must still estimate how many of those 3,711 students might have still attended a private school even without this financial assistance. To derive that estimate, we must isolate the count of “Previous SGO Award” pathway students who entered the tax-credit scholarship program initially as kindergarten students. It is reasonable to expect that some small share of these students might have still enrolled in a private school even without their tax-credit scholarship, then subsequently been able to continue on in a private school without their Choice Scholarship (voucher).

Since the IDOE reported that 640 of the 2,423 “Previous SGO Award” pathway students in 2013-14 were previously enrolled in a public school (see IDOE’s Table 14), we can then reasonably conclude the other 1,783 students received their first tax-credit scholarship award in kindergarten.1This assumption is possible because, through 2012-13, the only eligibility paths for the School Scholarship Tax Credit program were: 1) switching from a public school, 2) entering kindergarten, or 3) switching from a private scholarship program. Thus, since they didn’t qualify for the scholarship tax credit program by virtue of switching from a public school, these remaining 1,783 students must have come through one of the other two paths.2 This result is reached as follows:

STEP 3a: 2,423 (from Table 14) – 640 (from Table 14) = 1,783 “Previous SGO Award” pathway students from Year 3 entering system in kindergarten

Based on another Friedman Foundation analysis, about 60 percent of the “Previous SGO Award” pathway students, during 2010-11 and 2011-12, were initially kindergarten-eligible students. Applying this ratio, we can estimate that 2,939 of the 3,711 “Previous SGO Award” pathway students, counted in the “No Prior Public School” share, were initially kindergarten-eligible tax-credit scholarship students. First applying the 60 percent adjustment, then adding in the result from STEP 3a:

STEP 3b: 1,928 (pulled from Year 1 and Year 2 cohorts in STEP 1a) x 60 percent = 1,156 “Previous SGO Award” pathway students, retained from Year 1 and Year 2 cohorts, entering system in kindergarten

STEP 3c: 1,783 (from Step 3a) + 1,156 (from STEP 3b) = 2,939 “Previous SGO Award” pathway students entering system in kindergarten

Historically, about 7 percent to 8 percent of Indiana’s K-12 students have enrolled in private schools. Recognizing that existing private school options are more heavily concentrated in the areas where voucher-eligible students tend to be clustered (i.e., urban centers), we’ll use a 15 percent baseline private school enrollment rate for this analysis. In other words, we are cautiously assuming Choice Scholarship (voucher) recipients are twice as likely to attend a private school, without financial assistance, as the statewide average, despite being from low-income families. Applying that 15 percent baseline private school enrollment rate:

STEP 4: 2,939 (from STEP 3c) x 15 percent = 440 “Previous SGO Award” pathway students assumed to continue in private school without their voucher

Now we must exclude these 440 students if we wish to zero in the number of “Previous SGO Award” pathway students assumed to return to public school if their vouchers were revoked. Making this adjustment:

STEP 5: 3,711 (from STEP 2) – 440 (from STEP 4) = 3,271“Previous SGO Award” pathway students assumed to return to public school if their voucher is revoked

Thus, we estimate that 3,271 “Previous SGO Award” pathway students, included in the IDOE count of 7,779 Choice Scholarship recipients that “never attended public school in Indiana”, would now be enrolled in a public school if they their Choice Scholarship was revoked.

As for the student counts of other three sub-groups, the data needed to generate a detailed estimate similar to our calculations for sub-group #1 were not provided in the IDOE report. Though not included in the report, the IDOE should have the data needed to identify the student counts for sub-group #2 and sub-group #3, and may also be able to do the same for sub-group 4. At the very least, it would be reasonable for the media and policymakers to request that the IDOE examine its database to determine if count data for those other groups can be isolated.

Since the necessary data detail was not readily available to mirror the analysis done for sub-group #1, I employed a much less precise approach to generate a very rough estimate for the number of voucher students that fit in sub-groups #2, #3, and #4 combined as explained in this single step below:

Table 11 of the IDOE report shows that 4,525 voucher students qualified via the either the “Failing School” pathway (1,796) or the “Sibling” pathway (2,729).  To determine how many of these students might fall into sub-groups #2, #3, and #4, collectively, I had to fall back on my professional judgment to apply some reasonable assumptions.  If enrollment were distributed evenly across all thirteen (13) grades of K-12, each grade would represent about one-thirteenth or 8% of the enrollment.  However, actual enrollments tend to skew more heavily toward the lower grades as more students drop out along the way toward graduation.  Furthermore, sub-groups #2 and #3 include both kindergarteners and some 1st graders.  Sub-group #4 included students across 11 grade levels.  Considering all of these soft factors together, I settled on a simple assumption that 16% (a 2 grade level enrollment share) of the “Failing School” pathway and “Sibling” pathway students would fall into one of these three sub-groups. Apply this logic:

STEP A: 4,525 (from Table 11) x 16% = 724 voucher students either in entering Kindergarten, 1st Grade, or new Indiana residents qualify via the “Failing School” or “Sibling” pathway

Because this estimate is already so rough, I made no further adjustment to disaggregate this count into students likely to return to public school if their voucher were revoked versus students likely to persist in private school without any financial assistance.  That next step seemed to be more futile than valuable.  Instead, I’ve just simplistically assumed that all 724 students would need to return to public schools if their Choice Scholarship were pulled.

So, if we pull all of this analysis together, the 7,779 voucher students reported by the IDOE as “never attending a public school in Indiana” can be rationally recast as 3,784 voucher students “not diverted” from public schools”.  The simple math looks like this:

7,779 Voucher students “never attending a public school in Indiana”
– 3,271 “Previous SGO Award” pathway voucher students assumed to return to public school if their voucher is revoked (from STEP 5)
–    724 Voucher students either in entering Kindergarten, 1st Grade, or new Indiana residents qualify via the “Failing School” or “Sibling” pathway (from STEP A)
______
  3,784 Voucher students “not diverted” from public schools

If IDOE’s Table 13 were reconstructed to differentiate between: a) voucher students diverted from public school and b) voucher students not diverted, using our estimates above, then the breakdown would look like this:

ENDNOTES

1. The only possible exceptions would be students who received their first tax credit scholarship award at an older age because they had previously used financial assistance from a private scholarship program.

2. The number of tax credit scholarship recipients who were eligible by switching from a private scholarship program is de minimis.  Furthermore, it is also reasonable to conclude that the profile of this small group is very similar to the entering kindergarten group with regard to its propensity to enroll in a private school with any financial assistance.