los angeles - California Policy Center

CALIFORNIA POLICY CENTER | RESEARCH | JUNE 2015

LOS ANGELES

PUBLIC SCHOOLS VS. CHARTER SCHOOLS:

A cost per-pupil and educational achievement comparison

BY MARC JOFFE AND ED RING

SUMMARY This study examines cost per-pupil for high school students in the Los Angeles Unified School District (LAUSD), comparing its traditional public schools to those attending its larger charter school network, the Alliance College Ready Public Schools. It also examines the educational achievement outcomes for traditional and charter high school students, focusing on nine LAUSD traditional schools and nine LAUSD Alliance charter schools. The schools were chosen because of their close proximity to each other and similar demographic makeup, limiting extraneous variables from corrupting the study, and making for as close to an “apples-to-apples” comparison as is possible. CHARTER SCHOOLS HAVE A LOWER COST PER-PUPIL THAN TRADITIONAL SCHOOLS: Based on an analysis of relevant school costs and the number of enrolled high school students, the data shows the per-pupil costs for Alliance charter high school students to be $10,649 per year, compared to $15,372 per year for students at traditional public high schools within LAUSD, that is, we find a per-pupil cost differential of 44% in favor of Alliance charter schools. CHARTER SCHOOLS HAVE HIGHER API SCORES AND GRADUATION RATES THAN TRADITIONAL SCHOOLS: Academically, comparing LAUSD Alliance charter high schools to LAUSD traditional high schools located in the same neighborhoods, we found the Alliance schools have decisively higher Academic Performance Index (API) scores, 762 vs. 701, and higher graduation rates, 91.5% vs. 84.1%. CHARTER SCHOOLS HAVE HIGHER NORMALIZED SAT SCORES THAN TRADITIONAL SCHOOLS: With respect to SAT scores, when we normalized the comparison between the LAUSD Alliance charter and LAUSD traditional schools under consideration to equalize the rate of test participation, we found that the Alliance charter students outperformed the LAUSD traditional students with average scores of 1417 vs. 1299 – a significant difference. Among college bound students, an SAT score of 1299 puts the student in the bottom 27% nationally. A score of 1417, by contrast, places the student at 41% nationally. We have reached these conclusions by collecting district and school level data, and making common-sense adjustments where appropriate, and as noted in the text. Throughout this report, and especially in the footnotes, we have explained and documented our sources and calculations. We invite other researchers to conduct similar analysis and believe they will come to similar conclusions. The academic comparisons we have made involve 18 high schools; nine LAUSD Alliance charter high schools, and nine LAUSD traditional high schools. Our financial comparisons are somewhat broader in scope – involving the cost per high school pupil attending one of the largest public school districts in the nation and those attending a large charter school system. We can conclude from this data that an effective charter school operator can better learning outcomes at lower cost than traditional public schools serving a similar population. Ultimately, this finding supports the educational choice concept: by replacing a “one size fits all” solution with an array of educational choices, we can provide better results for California’s children and taxpayers.

1

Los Angeles Public Schools vs. Charter Schools: A cost per-pupil and educational achievement comparison

INTRODUCTION The purpose of this study is to help assess to what extent charter schools have the potential to provide a higher quality, more cost-effective educational solution for California K-12 students over traditional schools, particularly those in low income communities. Our study is not meant to provide a comprehensive assessment of the comparative quality of all traditional public schools vs. all public charter schools, but rather to focus on high school education in one of the most challenging parts of California for delivering quality educational outcomes, the low income communities of Los Angeles. The students in these communities fall within the Los Angeles Unified School District (LAUSD). The charter schools we selected for analysis are part of the Alliance College Ready Public Schools (Alliance), a nonprofit company that operates 26 charter middle schools and high schools, all of them a part of LAUSD, and most of them located in south-central Los Angeles. The first section of this

study will analyze the per-pupil costs for LAUSD traditional schools vs. LAUSD Alliance charter schools. The second section of this study will analyze the educational performance of LAUSD traditional schools vs. LAUSD Alliance charter schools. We will look at SAT scores, API scores, dropout rates, and rates of college admission. A critical element of this analysis will be to reduce the data for LAUSD to those schools that are in the same neighborhoods as the LAUSD Alliance charter schools. A district-wide LAUSD performance score would be distorted upwards by (1) the performance in many of the wealthier neighborhoods in the San Fernando Valley and elsewhere, and by (2) the upward skew represented by the presence of the three charter school operators within LAUSD (many of them in wealthier neighborhoods), enrolling nearly 40,000 students in 59 senior high schools [1], since they are formally part of LAUSD.


2

Previous Studies of Per-pupil Costs As Tom Rutten and Richard Riordan argue in a 2013 Los Angeles Daily News column entitled “The mysterious case of LAUSD’s finances,” LAUSD perpupil spending estimates vary widely [2]. They quote a district “baseline” estimate of $7,000, going on to discuss how that baseline is supplemented by numerous earmarked funds and categorical grants aimed at various segments of the student population. And at the other extreme, Adam Schaeffer, in a 2010 Cato Institute study entitled “They Spend WHAT? The Real Cost of Public Schools,” reported that the “real” cost of public education per-pupil at LAUSD was $25,208 [3].

Perhaps the closest thing we have to an official number is the 2013-14 “Current Expense per ADA” reported by the California Department of Education [6], which is $10,442. ADA stands for Average Daily Attendance which means the number of students marked “present” at all LAUSD’s schools on an average day. ADA is less than enrollment because some students are absent on any given day. To us, enrollment seems to be a more appropriate denominator for a cost per-pupil calculation. If a student is absent 18 days during a 180-day school year, the district is still educating a student – and not nine-tenths of one.

We believe that Schaeffer’s number is overstated for three reasons. The first factor is the simple passage of time. His estimates were based on 2008 data. Between the 2008-2009 school year and the 20132014 school year, authorized expenditures in the LAUSD budget fell from almost $18 billion [4] to just over $13 billion [5]. Although part of this reduction can be attributed to declining enrollment, much of it is due to state spending cuts in the aftermath of the Great Recession that are just now being reversed. Second, Schaeffer uses authorized budget figures rather than actual expenditures. As we will see in the next section, LAUSD spends far less than its budgetary authority, and so it is more accurate to use actual instead of budgeted amounts. Finally, the budget figures referenced by Schaeffer includes some double-counting as we also discuss below.

On the other hand, “Current Expense” does not include all the costs school districts incur. Among the items excluded from current expense are facilities construction and retiree health benefits [7]. Since schools require buildings and teachers unions often demand post-employment benefits, public education cannot be provided by LAUSD without these types of expenditures. We thus find that both the numerator and denominator of the “Current Expense per ADA” are understated if the purpose is to calculate the cost of educating a single pupil.

3

In the analysis that follows, we use fully loaded numerators and denominators, finding that the two alterations largely cancel out. However, as will be seen, further adjustments needed to make LAUSD’s costs comparable to those of Alliance, significantly raise our LAUSD cost per-pupil estimate.


PER-PUPIL COSTS FOR LAUSD TRADITIONAL VS. CHARTER In this section, we estimate per-pupil costs for LAUSD traditional vs. Alliance charter. Because LAUSD is a complex entity and because we wish to fully justify our calculations, this section is quite long. While we believe policy specialists will find this discussion rewarding, more casual readers may wish to skip to the end of this section to see our bottom line. As discussed in the previous section, cost analysis is more properly based on actual expenditures in audited financial statements rather than on budgets. Actual expenditures may vary substantially from budgeted amounts in either direction. In the case of LAUSD, the budget total substantially overstates actual expenditures. For the fiscal year ending June 30, 2014, LAUSD’s budget (available at http://laschoolboard.org/ sites/default/files/LAUSD2013-14FinalBudget.pdf) shows total authorized amounts of just over $13 billion [8]. This contrasts with actual expenditures of less than $8 billion [9] in the district’s audited financial statements. The audited statements are included in the district’s Comprehensive Annual Financial Report (CAFR) available at http://emma. msrb.org/EA696060-EA545171-EA941437.pdf. This large discrepancy between the budget and audited financial statement amounts has two main causes. First, the budget total double-counts over $1 billion of Internal Service Fund expenditures. Quoting from page 19 of the budget: Internal Services Funds, which total approximately $1.08 billion, account for the payment of employee health & welfare benefits, workers’ compensation, and liability insurance. These funds are for accounting purposes as required by State law. They serve as “pass-through”

accounts. In other words, the $1.08 billion in expenditures here already show up in other funds, and to count them in addition to the other funds would be counting them twice. For this reason, Internal Service Funds should not be considered as part of the funds that help operate District schools. Second, actual spending in all major governmental funds was substantially below authorized amounts in the budget. The biggest discrepancy occurred in district capital funds which had total authorized expenditures of $3.2 billion and actual expenditures of $0.7 billion. Operating and debt service funds also saw substantially less spending than authorized [10]. The audited financial statements contain two expenditure totals – one in the Statement of Activities and the other in the Governmental Fund Statement of Revenues, Expenditures and Fund Balances. Although they are not substantially different – $7.98 billion and $7.83 billion respectively – the relatively small discrepancy masks fairly large differences in how they are calculated. These differences warrant our attention. Expenses in the Statement of Activities are presented according to the accrual basis of accounting while Governmental Fund presentation uses the modified accrual basis – an accounting style that is much closer to the cash basis. Cash accounting focuses on the actual expenditure of funds during the fiscal year, while accrual accounting focuses on the financial obligations accrued by the district during the year. A reconciliation of the modified accrual and full accrual expenditure totals for LAUSD follows (amounts shown are in thousands of dollars):


4

Table 1 – LAUSD 2013-14 Expenditure Reconciliation, Modified Accrual vs. Full Accrual Bases [11] 7,833,503,000

Total Expenditures (Modified Accrual Basis): Deduct: Principal repayment

-358,744,000

Capital outlay

-589,514,000 -16,225,000

Accrual of compensated absences and other retirement benefits

-514,000

Other adjustments Add: Unfunded OPEB Expense

563,992,000

Depreciation

494,457,000 40,716,000

Interest accrued but not paid Total Expenditures (Accrual Basis): Under the accrual basis, the repayment of bond principal is not considered an expenditure nor are capital outlays. However, the accrual method involves depreciating assets, which largely offsets the absence of capital expenditures. It may be argued that capital outlays and/or bond principal should be added back to LAUSD’s total. The cost of building facilities is part of the overall cost of education and should not be excluded from per-pupil costs. However, including both capital outlays and bond principal repayments in the total would be a form of double-counting, because the capital outlays are made with borrowed funds. The money should be captured either when it is borrowed and spent, or when it is repaid – not both times. Inclusion of the depreciation amount in the accrual expenditures serves the purpose of incorporating capital expenditures. Since LAUSD does not issue pension obligation bonds, virtually 5

$7,967,671,000 all of its borrowing supports capital expenditure. Thus everything financed with LAUSD’s bonds will be depreciated on the district’s books. Inclusion of the depreciation amount thus provides a reasonable proxy for capital expenditure, and we do not see a strong case for altering LAUSD’s accrual accounting expenditures for our calculations. LAUSD’s accrual basis expenditures also include retiree healthcare benefits that current employees have earned but will not claim until they retire. These OPEB accruals are excluded from the modified accrual totals. Alliance does not offer retiree healthcare benefits – providing it with a significant cost saving relative to LAUSD. (Both LAUSD and Alliance offer pension benefits largely through CalSTRS – so accruals for cash retirement benefits should not be a major differentiating factor between Alliance and LAUSD).


Because Alliance only reports expenditures in accordance with the accrual method of accounting, we use LAUSD’s accrual method expenditure total of $7,967,671,000 for our comparisons. Alliance’s total expenditures for the fiscal year ending June 30, 2014 were $102,789,813 as shown on page 5 of its audited financial statements available at http://emma.msrb. org/ER822375-ER640652-ER1042507.pdf. LAUSD reports total enrollment of 726,371 on page 157 of its CAFR, yielding a per-pupil cost of $10,969. Alliance reports total enrollment of 10,020 students in its 2013-14 performance dashboard available at http://www.laalliance.org/performance/13-14/ dashboards/Alliance-wide%20Performance%20 Dashboard%202013-2014.pdf, yielding a per-pupil cost of $10,258. Thus, before making adjustments, we find that perpupil expenditures using the accrual basis, are $711 higher for LAUSD traditional than for Alliance charter – a difference of almost 7%. The calculations are summarized below. LAUSD’s enrollment total of 726,371 includes 95,381 students [12] enrolled in Independent Charter Schools. However, LAUSD expenditures shown in its financial audit do not include charter school expenses. As noted on page ii of the CAFR: “This report includes all funds of the District with the exception of the fiscally independent charter schools, which are required to submit their own individual audited financial statements”. Thus the 95,381 charter school students should be removed from the denominator of the cost per-pupil calculation.

We now make adjustments to these reported totals to increase their comparability. LAUSD provides adult education while Alliance does not. By removing the adult education cost and enrollment from LAUSD data, we can more nearly achieve an apples-to-apples comparison. LAUSD’s audit reports adult enrollment of 32,267 and adult education costs of $75,993,000 [13]. Similarly, LAUSD provides early education services to pre-kindergarten children and infants. Alliance does not provide similar services. Consequently, we remove 12,829 early education enrollees and $128,407,000 in Child Development Fund expenditures [14] from the LAUSD totals. In addition to Adult Education, LAUSD offers Career Technical Education (CTE) and Regional Occupational Programs (ROP). These two programs had combined enrollment of 29,779 and a total budget of $43.6 million [15], implying a per-student cost of $1464. The CAFR does not provide actual expenditures for these two vocational programs which are not offered by Alliance. But unless actual expenditures were far greater than budget, the two programs are much less expensive on a per student basis than LAUSD’s traditional K-12 program. Since LAUSD’s actual spending overall was below budget, using budgeted expenditures for the two programs is a conservative approach. The adjustments made to LAUSD to eliminate enrollees and costs not associated with K-12 traditional public school students are summarized below.

Table 2 – Alliance Charter/LAUSD Traditional Expenditure per Enrolled Student Comparison (Initial) $7,967,671,000

$102,789,813

Number of Enrolled Students

726,371

10,020

Expenditure per Enrolled Student

$10,969

$10,258

Total Expenditures


6

Table 3 – Adjustments to LAUSD Amounts to Improve Comparability with Alliance

Before Adjustments

Expenditures

Enrolled Students

$7,976,671,000

726,371 -95,381

Remove Charter School Students Remove Adult Students and Expenditures

-75,993,000

-32,267

Remove Vocational Students and Expenditures

-43,600,000

-29,779

-128,407,000

12,829

$7,719,671,000

556,115

Remove Pre-K Enrollees and Expenditures Adjusted Totals (K-12 Traditional Public School Students Only) The Adult Education and Pre-K expenditure totals are based on the modified accrual accounting method; governmental financial statements do not break down accrual accounting data to the

fund level. After these adjustments, LAUSD per-pupil costs rise to $13,881 or about 35% higher than the Alliance amount. The adjusted cost comparison is summarized in the table below.

Table 4 – Alliance/LAUSD Expenditure per Enrolled Student Comparison (Adjusted) LAUSD

Alliance

$7,719,671,000

$102,789,813

Number of Enrolled Students

556,115

10,020

Expenditure per Enrolled Student

$13,881

$10,258

Total Expenditures

A further adjustment to these numbers is needed to reflect differences in grade profiles and differences in educational costs between elementary and high schools. LAUSD and Alliance have substantially different grade profiles, because Alliance does not operate elementary schools. In the 2013-14 school year, all of Alliance students were in grades 6-12. Because most Alliance facilities are high schools, 72% of Alliance students were in grades 9-12 [16]. By contrast, 49% of LAUSD students were enrolled in elementary schools and only 23% were enrolled in high schools [17]. 7

High school education is more expensive on a per-pupil basis than elementary school education, largely due to the many extra programs that high schools offer. Although, per-pupil costs within LAUSD are hard to obtain, we were able to find comparative data elsewhere [18]. Our review of Texas public school data showed per-pupil expenditures averaging $6,654 in elementary schools and $10,323 in high schools [19]. The California Department of Education reports current expense per ADA for elementary and high school districts, i.e. districts which consist of all elementary schools or all high schools,


respectively. For 2013-2014, California elementary school districts had an expense of $8,336 per ADA while high school districts had an expense of $9,569 per ADA [20]. This last observation suggests that educating a high school student in California is about 15% more expensive than educating an elementary school student. If we apply this differential to LAUSD and Alliance and if we assume that elementary and middle school students have

equivalent costs, we can infer costs per high school student at LAUSD and Alliance from the overall per student costs presented in Table 4 above. The results of this calculation are shown in the table below. Taking into account the grade profile and assuming that high school students are 15% more expensive to educate than elementary and middle school students, we find a per-pupil cost differential of 44% in favor of Alliance charter.

Table 5 – LAUSD Traditional/ Alliance Charter Expenditure per Student by Grade Level LAUSD

Alliance

$13,881

$10,258

74%

28%

26%

72%

115%

115%

Per-pupil Expenditure per Enrolled K-8 Student

$13,367

$9,260

Per-pupil Expenditure per Enrolled High School Student

$15,372

$10,649

Per-pupil Expenditure per Enrolled K-12 Student Estimated Proportion of Elementary/Middle School Students (K-8) Estimated Proportion of High School Students (9-12) Ration: Per-pupil High School Expenditure/ Per-pupil K-8

One other potential adjustment would benefit LAUSD in per-pupil cost comparisons. Although we do not have enough data to quantity this adjustment, we describe it here. The issue involves special education students – those who require special accommodations due to various disabilities, and are thus more expensive to educate than children learning in traditional classroom environments. A recent report on KPCC [21] states that LAUSD has 82,000 special education students and a $1.4 billion special education department budget – implying a per-pupil cost of about $17,000. The LAUSD CAFR shows a somewhat lower expenditure amount – $1.318 billion [22] – and does not provide a figure for total special education enrollment [23].

Assuming the 82,000 special education enrollment figure is correct, that number accounts for about 15% of the district’s K-12 enrollment [24]. By contrast, Alliance reports that 9% of its students are in special education [25]. The organization does not break out costs for special education students in its financial reporting. Thus, we can conclude that a comparison of costs aside from those associated with special education would be more favorable to LAUSD, but it is not clear how large the adjustment would be. In conclusion, we find that LAUSD spends 44% more to educate each high school student than Alliance, but that a part of this differential may be explained by the greater proportion of special education


8

ACADEMIC PERFORMANCE COMPARISONS While cost per-pupil is a key variable in comparing the effectiveness of differing approaches to public education, it is not the whole story. The quality of various educational solutions, measured in terms of outcomes for students, taking into account various degrees of preexisting adversity faced by the students, forms an essential part of any comparative analysis. When comparing academic performance in LAUSD traditional high schools vs. LAUSD Alliance charter high schools, district-wide comparisons are easy enough to gather, and offer some insights. But a more meaningful set of comparisons may be derived by presenting academic results between individual LAUSD traditional and LAUSD Alliance charter high schools that are located in close proximity to each other. This is based on the assumption that schools in the same neighborhoods will have student bodies that have similar characteristics, which we will attempt to verify using available data.

Accordingly, we have identified 18 high schools, nine that are part of the LAUSD Alliance charter network and nine that are LAUSD traditional public schools, organized into seven matches, where schools that are in the same neighborhoods are considered to be a match. In four cases, one Alliance charter high school is compared to one traditional high school located in the vicinity. In two cases (#4 and #5 on the list below), there are two traditional high schools that are both located very close to an Alliance charter high school, so both of them are considered. In one case (#3), there are two Alliance charter high schools that are both located very close to a traditional high school; both Alliance schools are considered. The maximum distance for any comparison is 3.4 miles, in most cases the schools being compared are about 2.0 miles apart.

List of LAUSD Traditional and LAUSD Alliance Charter High Schools in Close Proximity: 1. A lliance High School, Cindy and Bill Simon Technology High School, 10770 Wilmington Ave., Los Angeles. LAUSD High Schools, 1.6 miles away: South East Senior High, 2720 Tweedy Blvd., South Gate, and, 2.9 miles away, South Gate Senior High, 3351 Firestone Blvd, South Gate. 2. Alliance High School, Patti & Peter Neuwirth Leadership Academy, 4610 South Main St., Los Angeles. LAUSD High School, 0.7 miles away, Dr. Maya Angelou Community High School, 300 E 53rd St, Los Angeles.

9

3. Alliance High School, Collins Family CollegeReady High School, 2071 Saturn Ave., Huntington Park. LAUSD High Schools, 2.8 miles away, Bell Senior High, 4328 Bell Ave, Bell, and, Maywood Academy Senior High, 3.3 miles away, 6125 Pine Ave., Maywood. 4. Alliance High Schools, Environmental Science & Technology High School, 2930 Fletcher Drive, Los Angeles, and, Tennenbaum Family Technology High School, 2050 North San Fernando Road, Los Angeles. LAUSD High School, 2.4 miles and 3.4 miles away, respectively, John Marshall


Senior High, 3939 Tracy St., Los Angeles. 5. Alliance High Schools, Health Services Academy High School, 10616 South Western Ave., Los Angeles, and, Judy Ivie Burton Technology High School, 10101 South Broadway, Los Angeles. LAUSD High School, Middle College High School, 0.8 miles and 2.7 miles away, respectively, 1600 Imperial Highway, Los Angeles. 6. Alliance High School, Marc & Eva Stern Math and

Science High School, 5151 State University Drive, Los Angeles. LAUSD High School, 1.6 miles away, Woodrow Wilson Senior High, 4500 Multnomah St., Los Angeles. 7. Alliance High School, Media Arts and Entertainment Design High School, 113 South Rowan Ave., Los Angeles. LAUSD High School, 3.2 miles away, James A Garfield Senior High, 5101 E Sixth St., Los Angeles.

Alliance (green) and LAUSD (red) High Schools


10

Demographic Comparisons for LAUSD Traditional and LAUSD Alliance Charter High Schools in Close Proximity The fact that two high schools are in the same neighborhood doesn’t prove they have a demographically similar student body. Before making academic comparisons it is important to review other variables. The next two tables examine the demographics of the 18 high schools, comparing the Alliance charter high schools to the LAUSD traditional high schools.

As can be seen in Table 6, the ethnic makeup is very similar between the LAUSD Alliance and LAUSD traditional schools. Both groups of student bodies are 91% Latino. Of the remaining students, the Alliance schools are 5% African American compared with 2% in the traditional schools, and the traditional schools have slightly more Asian (2% vs. 1%) and White (2% vs. 1%) students compared to Alliance [26].

Table 6 – Ethnic Composition of Alliance vs. LAUSD Students Latino

African American

Asian

White

Other

South Gate Senior High School

99%

1%

0%

0%

0%

South East Senior High School

98%

1%

0%

1%

0%

Maya Angelou Community High School

88%

11%

0%

0%

1%

Bell Senior High School

98%

0%

0%

2%

0%

Maywood Academy

99%

0%

0%

1%

0%

Marshall Senior High School

60%

2%

13%

12%

0%

Middle College High School

68%

31%

0%

0%

0%

Wilson Senior High School

93%

2%

3%

1%

0%

Garfield Senior High School

99%

1%

0%

0%

0%

LAUSD Weighted Averages (same neighborhoods)

91%

2%

2%

2%

0%

LAUSD Traditional Schools

11


Latino

African American

Asian

White

Other

Cindy and Bill Simon Technology High School

87%

13%

0%

0%

0%

Patti & Peter Neuwirth Leadership Academy

90%

9%

0%

0%

0%

Collins Family College-Ready High School

99%

0%

0%

0%

0%

Environmental Science and Technology High School

86%

2%

3%

2%

0%

TennenBaum Family Technology High School

92%

2%

1%

1%

0%

Health Services Academy High School

81%

17%

0%

1%

0%

Judy Ivie Burton Technology High School

93%

6%

0%

1%

0%

Marc & Eva Stern Math and Science High School

96%

1%

1%

1%

0%

Media Arts and Entertainment Design High School

96%

0%

0%

3%

0%

Alliance Weighted Averages (same neighborhoods)

91%

5%

1%

1%

0%

Alliance/LAUSD

1.00

2.25

0.25

0.43

0.0

Alliance Charter Schools

The other primary demographic variables that are relevant when assessing the similarity of student bodies are those that may present challenges to individual students. Using available per high school data, the next table attempts to make this assessment by examining the percentages in each group of schools of students who have learning disabilities, students who are English language learners, and students who are on the free or reduced lunch program (in both Alliance and LAUSD the percentage of students on the lunch program who get a free lunch vs. a reduced lunch are around 90%).

As shown in Table 7, LAUSD traditional has slightly more English language learners, at 17% vs. 16%. LAUSD Alliance has more students on the free/ reduced meal program, 90% vs. 85%. LAUSD traditional, on the other hand, has significantly more students who are identified as having learning disabilities, 13% for LAUSD traditional vs. 8% for LAUSD Alliance charter. It is difficult to conclude too much from this data. It would probably be reasonable to conclude that overall the student bodies at these 18 schools are very similar ethnically and demographically [27].


12

Table 7 – Other Demographic Characteristics of LAUSD Alliance vs. LAUSD Traditional Students English Language Learned

Students with Disabilities

Free/ Reduced Lunch Program


17%

13%

90%


18%

12%

93%


26%

14%

64%


20%

12%

85%

Maywood Academy

16%

12%

90%

Marshall Senior High School Middle College High School

14% 1%

11% 0%

79% 85%


15%

17%

88%


16%

13%

87%


17%

13%

85%

English Language Learned

Students with Disabilities

Free/ Reduced Lunch Program

21%

10%

99%

17%

10%

95%

14%

6%

92%

9%

7%

79%

16%

10%

64%

Health Services Academy High School

18%

10%

91%


16%

8%

99%

12%

9%

90%

21%

7%

95%

16%

8%

90%

0.93

0.67

1.06


Alliance Charter Schools Cindy and Bill Simon Technology High School Patti & Peter Neuwirth Leadership Academy Collins Family College-Ready High School Environmental Science and Technology High School TennenBaum Family Technology High School

Marc & Eva Stern Math and Science High School Media Arts and Entertainment Design High School Alliance Weighted Averages (same neighborhoods) Alliance/LAUSD 13


Educational Outcomes: LAUSD Alliance Charter Schools vs. LAUSD Traditional Schools To evaluate the academic performance of students in LAUSD Alliance Charter schools compared to traditional LAUSD traditional high schools located in the same neighborhoods with similar student demographics, we focus on four variables, each school’s rate of attendance, their API (Academic Performance Index) score, the average SAT score per school, and the graduation rate per school. These comparisons are displayed on Table 8. As can be seen, the reported average attendance at

the LAUSD traditional schools, 95%, was nearly identical to the reported average attendance at the Alliance schools, 96%. On the other hand, the LAUSD Alliance schools display a distinct edge in their average API scores, 762 vs. 701, and in their graduation rates, 91.5% vs. 84.1%. With respect to average SAT scores, the raw data for the LAUSD traditional schools actually shows them outperforming the Alliance schools, 1299 for LAUSD vs. 1122 for the Alliance schools [28]. But this bears further analysis.

Table 8 – Academic Performance Indicators, Alliance vs. LAUSD Daily Attendance

API Score

Average SAT Score

Graduation Rate


95%

709

1271

94%


97%

689

1298

83%


94%

578

1067

68%


91%

706

1276

82%

Maywood Academy

97%

705

1277

80%


97%

758

1443

81%


99%

807

1336

99%


95%

653

1217

85%


97%

714

1255

87%


95%

701

1299

84.1%



14

Alliance Charter Schools

Daily Average API Score Attendance SAT Score

Cindy and Bill Simon Technology High School Patti & Peter Neuwirth Leadership Academy Collins Family College-Ready High School Environmental Science and Technology High School TennenBaum Family Technology High School Health Services Academy High School Judy Ivie Burton Technology High School Marc & Eva Stern Math and Science High School Media Arts and Entertainment Design High School Alliance Weighted Averages (same neighborhoods) Alliance/LAUSD To accurately compare the SAT scores of students in LAUSD Alliance Charter schools with those attending traditional LAUSD traditional high schools located in the same neighborhoods with similar student demographics, it is necessary to consider the relative proportions of students taking the test. While the LAUSD traditional students averaged 1299 on their SAT vs. 1122 for the Alliance charter students, outperforming them by 177 points, a much lower percentage of LAUSD traditional juniors and seniors actually took the test. As reported on Table 9, only 31% of the juniors and seniors enrolled in the LAUSD traditional schools took the SAT, vs. 72% of the Alliance charter school students. To appreciate what an apples-to-apples comparison might reveal, imagine what the average SAT score would be in the LAUSD traditional schools if 72% of the juniors and seniors took the test. In this hypothetical example, also consider the fact that because the LAUSD traditional schools have a drop-out rate of 15.9% vs. 8.5% at LAUSD Alliance, an apples-to-apples comparison would have to actually assume a cohort of juniors and seniors whose 15

Graduation Rate

96%

715

1134

86%

95%

798

1154

92%

98%

815

1198

98%

96%

860

1381

91%

97%

694

1200

86%

96%

717

1119

85%

98%

754

1210

88%

96%

776

1280

97%

95%

694

1156

96%

96%

762

1122

91.5%

1.01

1.09

0.86

1.09

numbers have not succumbed to a much higher rate of attrition. Put another way, taking into account attrition, more than 72% of LAUSD’s traditional still-enrolled juniors and seniors would have to take the SAT to properly compare their performance to LAUSD Alliance’s. While data cannot exist for tests that were not taken, we can achieve something closer to an apples-toapples comparison by excluding a portion of the LAUSD Alliance test results. If we assume that LAUSD traditional students who didn’tw take the SAT would have had lower scores than the LAUSD students who did take the test, we can match up the LAUSD traditional and LAUSD Alliance charter samples by dropping the lower scores from Alliance. Since the LAUSD traditional schools had a test participation rate of 31%, we can get a comparable Alliance SAT average by only considering the top 686 scores – which corresponds to 31% of the 2,189 enrolled juniors and seniors (see Table 9). According to data provided by Alliance, the top 686 SAT scores at the eight Alliance schools was 1417 – 118 points higher than the LAUSD traditional average [29].


Table 9 – SAT Performance Comparisons, Normalized Based on Rate of Participation Average SAT Score

# of JRs and SRs

# Taking SAT

% Taking SAT


1271

1,279

369

29%


1298

819

237

29%


1067

274

50

18%


1276

1,533

390

25%

Maywood Academy

1277

504

112

22%


1443

1,091

444

41%


1336

194

100

52%


1217

637

224

35%


1255

960

359

37%


1299

7,291

2,285

31%

Average SAT Score

# of JRs and SRs

# Taking SAT

% Taking SAT

1104

202

115

57%

1114

323

306

95%

1218

296

291

98%

1496

281

95

34%

1198

174

108

62%

1144

223

97

43%


1228

275

197

72%

Marc & Eva Stern Math and Science High School

1268

274

252

92%

1213

141

111

79%

1122

2,189

1,572

72%


Alliance Charter Schools Cindy and Bill Simon Technology High School Patti & Peter Neuwirth Leadership Academy Collins Family College-Ready High School Environmental Science and Technology High School TennenBaum Family Technology High School Health Services Academy High School

Media Arts and Entertainment Design High School Alliance Weighted Averages (same neighborhoods)


16

# of Alliance students taking SAT based on LAUSD participation ratee: Average SAT score for top 686 Alliance Students Alliance/LAUSD

686

31%

1417 1.09

CONCLUSION Overall, we conclude that LAUSD Alliance charter high schools provide better outcomes at lower costs than comparable LAUSD traditional operated public schools in the same area. We estimated the per-pupil costs for Alliance charter high school students to be $10,649 per year, compared to $15,372 per year for students at traditional public high schools within LAUSD, that is, we find a per-pupil cost differential of 44% in favor of LAUSD Alliance charter schools. Academically, comparing LAUSD Alliance charter high schools to LAUSD traditional high schools located in the same communities, we found the Alliance schools to have decisively higher API scores, 762 vs. 701, and measurably higher graduation rates, 91.5% vs. 84.1%. With respect to SAT scores, when we normalized the comparison between the LAUSD Alliance and LAUSD traditional schools under consideration to equalize the rate of participation, we found that the LAUSD Alliance students outperformed the LAUSD traditional students with average scores of 1417 vs. 1299. This differential is significant. Among college bound students, an SAT score of 1299 puts the student in the bottom 27% nationally. A score of 1417, by contrast, places the student at 41% nationally [30]. Alliance high schools have succeeded in producing academic achievers in the upper 1/3rd of their student body whose SAT scores are within striking distance of the national average, despite operating in some of the most disadvantaged communities in the United States.

17

We have reached these conclusions by collecting district and school level data, and making commonsense adjustments where appropriate, and as noted in the text. Throughout this report, and especially in the footnotes, we have explained and documented our sources and calculations. We invite other researchers to conduct similar analysis and believe they will come to similar conclusions. Finally, we want to reiterate that the academic comparisons we have made are fairly narrow, involving 18 high schools, nine LAUSD Alliance charter high schools, and nine LAUSD traditional high schools. Our financial comparisons are somewhat broader in scope – involving the cost per high school pupil attending one of the largest public school districts in the nation and those attending a large charter school system. Because the sample size is still relatively limited, our study thus cannot support a categorical conclusion that charter schools are always better than public schools, or vice versa, but we can conclude from this data that an effective charter school operator can deliver better learning outcomes at lower cost than traditional public schools serving a similar population. Ultimately, this finding supports the educational choice concept: by replacing a “one size fits all” solution with an array of educational choices, we can provide better results for California’s children and taxpayers.


About the Authors: Marc Joffe is a policy analyst for the California Policy Center. He is also the founder Public Sector Credit Solutions, established in 2011 to educate policymakers, investors and citizens about government credit risk. PSCS research has been published by the California State Treasurer’s Office, the Mercatus Center and the Macdonald-Laurier Institute among others. Prior to starting PSCS, Marc was a Senior Director at Moody’s Analytics. He has an MBA from New York University and an MPA from San Francisco State University.

Ed Ring is the executive director for the California Policy Center. Previously, as a consultant and fulltime employee primarily for start-up companies in the Silicon Valley, Ring has done financial accounting for over 20 years, and brings this expertise to his analysis and commentary on issues of public sector finance. Ring has an MBA in Finance from the University of Southern California, and a BA in Political Science from UC Davis.

Footnotes 1 – For a list of LAUSD Charter Schools, refer to Los Angeles Unified School District, Charter Schools Division, “2014-15 Charter Schools Directory: http://notebook.lausd.net/pls/ptl/docs/PAGE/CA_LAUSD/ LAUSDNET/SCHOOLS/ADD_SCH_DOCS/CHARTER%20 SCHOOLS%20DIRECTORY%202014-15%20PARENTS.PDF For enrollment by LAUSD High School refer to California Department of Education, Analysis, Measurement, and Accountability Reporting Division (this table also includes SAT scores by school): http://dq.cde.ca.gov/dataquest/satactap/sat. aspx?cyear=2013-14&cchoice=SAT3b&year=1314&cdscode=1 9647330000000&clevel=District&ctopic=sat&level=District 2 – Richard J. Riordan and Tim Rutten (November 1, 2013). The mysterious case of LAUSD’s finances. Los Angeles Daily News, http://www.dailynews.com/opinion/20131101/themysterious-case-of-lausds-finances-richard-j-riordan-andtim-rutten 3 – Adam Schaeffer (March 10, 2010), They Spend What? Cato Institute Policy Analysis Number 662. http://object.cato.org/ sites/cato.org/files/pubs/pdf/pa662.pdf 4 – LAUSD 2008-2009 Final Budget. Page I-51.http:// notebook.lausd.net/pls/ptl/docs/PAGE/CA_LAUSD/ LAUSDNET/OFFICES/CFO_HOME/ALL%20SECTIONS%20 091108.PDF. 5 – LAUSD 2013-2014 Final Budget. Page 49.http:// laschoolboard.org/sites/default/files/LAUSD201314FinalBudget.pdf. 6 – California Department of Education, 2013-14 Cost Per ADA spreadsheet,http://www.cde.ca.gov/ds/fd/ec/documents/ currentexpense1314.xls. 7 – California Department of Education, Current Expense of

Education.http://www.cde.ca.gov/ds/fd/ec/currentexpense. asp. 8 – LAUSD 2013-14 Budget, Page 49. 9 – LAUSD 2013-14 Comprehensive Annual Financial Report, Pages 15 and 18. 10 – The CAFR contains budget to actual comparisons for most funds at Page 20 and Pages 66-89. 11 – Adjustments described here were gathered from Pages 15, 18 and 19 of the LAUSD CAFR. 12 – This and other LAUSD enrollment numbers cited below are on page 157 of LAUSD’s CAFR. 13 – LAUSD CAFR page 74. 14 – LAUSD CAFR, Page 165. 15 – LAUSD Budget, Pages 25 and 29. 16 – Derived from the Alliance Performance Dashboard referenced earlier. 17 – Enrollment data by type of school is included in the LAUSD CAFR on Page 157. About 11% of LAUSD students are enrolled in K-12 schools, so the actual proportions of elementary and high school students are slightly higher than the numbers provided above. Later, we use an estimate of 26% high school students in LAUSD by proportionately allocating back those enrolled in K-12 schools. 18 – As this study was being completed, the authors found a set of LAUSD School Accountability Report Cards that contain per-pupil costs by school site. Our preliminary analysis of these report cards shows average per-pupil costs of $7,854 for elementary schools and $8,407 for high schools. These amounts generally exclude special education schools which have much higher costs and typically offer grades K-12. 19 – Author’s analysis of data downloaded from Texas Education Agency’s Academic Excellence Information


18

System.http://ritter.tea.state.tx.us/perfreport/aeis/2012/ xplore/DownloadSelData.html. 20 – California Department of Education, Cost per ADA Spreadsheet 2013-2014, ref. summary tab “Average by LEA Type.” http://www.cde.ca.gov/ds/fd/ec/documents/ currentexpense1314.xls 21 – Annie Gilbertson, Costs for LAUSD special ed services climb as parents feel the pinch, 89.3 KPCC, December 19, 2014.http://www.scpr.org/blogs/ education/2014/12/19/17698/costs-for-lausd-special-edservices-climb-as-paren/ 22 – LAUSD CAFR Page 121. 23 – The CAFR only shows the number of children in special education schools, but does not include the number of special education students enrolled in traditional schools. 24 – K-12 enrollment totaled 556,115 as shown on Page 157 of the LAUSD CAFR. This is the 585,894 LAUSD students net of early education, adult education and independent charter schools shown in Table 3 above minus the 29.779 students in Career Technical Education and the Regional Occupational Program. 25 – 2013-14 Alliance-Wide Performance Dashboard,http:// www.laalliance.org/performance/13-14/dashboards/Alliancewide%20Performance%20Dashboard%202013-2014.pdf. 26 – The ethnic breakdown of the enrolled students at each of LAUSD high school, including the charter high schools, can be found from the CA Dept. of Education, Educational Demographics Unit: http://dq.cde.ca.gov/dataquest/Enrollment/EthnicEnr.asp x?cType=ALL&cGender=B&cYear=2013-14&Level=School& cSelect=Alliance+Gertz-Ressl–Los+Angeles+Uni–19647330106864&cChoice=SchEnrEth 27 – English Language Learner: CA Dept. of Education, Selected District Level Data, Los Angeles Unified for the year 2013-2014 http://dq.cde.ca.gov/dataquest/cbeds3.asp?FreeLunch=on&P

19

ctEL=on&cChoice=DstProf2&cYear=2013-14&cTopic=Profile& myTimeFrame=S&cSelect=1964733–LosÂngelesÛnified Free/Reduced Lunch Program: CA Dept. of Education, Selected District Level Data, Los Angeles Unified for the year 2013-2014 http://dq.cde.ca.gov/dataquest/cbeds3.asp?FreeLunch=on&c Choice=DstProf2&cYear=2013-14&cTopic=Profile&myTimeFra me=S&cSelect=1964733–LosÂngelesÛnified Students with disabilities data was provided by Alliance and LAUSD’s Office of Data and Accountability http://notebook.lausd.net/portal/page?_ pageid=33,131762&_dad=ptl 28 – CA Dept. of Education, API Reports http://www.cde.ca.gov/ta/ac/ap/apireports.asp CA Dept. of Education, SAT Scores http://dq.cde.ca.gov/dataquest/satactap/sat. aspx?cyear=2013-14&cchoice=SAT4b&year=1314&cdscode=19 647330111658&clevel=School&ctopic=sat&level=School CA Dept. of Education, Cohort Graduation Rate http://dq.cde.ca.gov/dataquest/CohortRates/GradRates.aspx ?Agg=S&Topic=Graduates&TheYear=2013-14&cds=196473301 06864&RC=School&Subgroup=Ethnic/Racial Daily attendance data was provided directly by Alliance and LAUSD’s Office of Data and Accountability http://notebook.lausd.net/portal/page?_ pageid=33,131762&_dad=ptl 29 – Individual SAT scores for the nine schools under analysis were provided by Alliance. The overall averages were corroborated with CA Dept. of Education data. The top 686 scores, as calculated, were then extracted from that data. 30 – National percentile rankings for SAT 2014 for College Bound Seniors https://secure-media.collegeboard.org/digitalServices/ pdf/sat/sat-percentile-ranks-composite-crit-reading-mathwriting-2014.pdf
