ED Response to 60-Day Public Comments

Public Comment Received During the 60-day Comment Period

and NCES Response

November 2016

Common Core of Data (CCD) School-Level Finance Survey (SLFS) 2016-2018

ED-2016-ICCD-0101-0001 Comments on FR Doc # 2016-22302

Submitter Information

Name: Christine Davis
Address: Sacramento, CA, 95814

Document: ED-2016-ICCD-0101-0006

General Comment

To Whom It May Concern:

The California Department of Education (CDE) appreciates the opportunity to review and comment on the Common Core of Data School-Level Finance Survey (SLFS) 2016–2018 as proposed in the Federal Register, Volume 81 on September 16, 2016.

One of the CDE’s primary concerns with the proposed regulations is that California neither collects nor requires its local educational agencies (LEA) to record school-level financial data for reasons delineated below. Although the Every Student Succeeds Act (ESSA) requires the state and LEAs to include in their report cards certain per-pupil expenditures for each school in the state, that requirement does not begin until a data collection for school year 2017–2018, which may be extended by one year if proposed 34 Code of Federal Regulations § 200.30 (Federal Register, Volume 81 on May 31, 2016), is enacted. Further, the SLFS data collection requires a greater granularity of expenditure tracking than ESSA requires, resulting in a greater reporting burden on LEAs.

QUALITY AND UTILITY OF SCHOOL-LEVEL FINANCIAL DATA

Financial data is not kept at the school site level: The CDE’s primary reason for believing that reliable school-level financial data does not exist is that most California LEAs simply do not track expenditures at the school level for a variety of good and valid reasons. As examples:

• There are certain economies of scale in the practice of administering public education by grouping schools into districts. Requiring districts to account for costs at the school level defeats this economy.

• While every district takes the needs of each of its schools into account when allocating its resources among those schools, it is not necessarily beneficial to the district to track the allocation of those resources through its general ledger accounting system. In many districts, not even teacher salaries are tracked at the school level. This is not to say that districts do not know where their teachers teach; the districts just use methods outside the accounting system.

• Direct instructional costs do not always lend themselves to being tracked at or attributed to the school level. While it would seem that teacher salaries could be easily attributed to individual schools, not all can be or should be. As examples, some itinerant teachers teach at multiple schools on an as-needed basis or on schedules that vary day by day or week to week. Substitute teachers, in particular, normally teach at many different schools over the course of a week or month. There is no particular advantage in tracking substitute teacher costs to particular schools.

• For many districts, the system and human resource costs to track and collect financial data at the school level would be prohibitive.

Developing protocols for attributing district level costs to the school level does not reflect real costs: Standardized protocols must be implemented to attribute the appropriate district level costs to the school level. However, no matter how sophisticated the method, the resulting costs, while consistently determined, are not an accurate measure of the actual cost incurred at the school site. Thus, there is a flawed basis upon which conclusions regarding comparability may be made.

The fact that school-level financial data were collected previously does not mean that the data were valid or that the collection should continue: For example, most California LEAs reported school-level financial data for the original American Recovery and Reinvestment Act (ARRA) data collection. However, many acknowledged that they had difficulty complying because of the reasons stated above, resulting in data that were not likely to be accurate and not representative of the actual services provided at the school site.

Although ESSA will require LEAs to begin reporting certain expenditures at the school level, the SLFS data collection requires a greater level of granularity than ESSA requires, resulting in a greater level of LEA burden and a greater level of unreliability due to the reasons stated above. Specifically, the SLFS comprises thirty data elements, whereas ESSA requires only six data elements to be reported in state and LEA report cards for school-level per-pupil expenditures. Additionally, the SLFS requires certain amounts, such as plant maintenance, to be excluded from the data collection, whereas ESSA requires the amounts to be reported.

NECESSITY OF DATA COLLECTION

The CDE questions whether school-level financial data is the appropriate measure for policy makers to ascertain equity among districts and schools. Equality of spending is not a valid measure of educational opportunity because schools are not necessarily expected to be comparable due to their unique characteristics. As examples:

• Some schools provide a “magnet” curriculum that attracts pupils with particular interests. That curriculum may involve greater or lesser costs than the curricula in other schools.

• A school might incur higher substitute teacher costs because of a health epidemic that impacted only that school. That additional cost does not reflect a better educational opportunity for pupils.

• Even equal spending does not assure equal value. As an example, a school could hire ten teachers earning $120,000 each for the same cost as hiring fifteen teachers earning $80,000 each, but class sizes would differ considerably.

Without the context regarding local expenditures, which is available at the local level, the financial data could be misleading and may result in flawed education policy decisions. The high-level financial data collected by the SLFS lacks the context that would be necessary to make reliable assumptions about the data.

Further, once LEAs and the state begin to publish school-level per-pupil expenditure data in their report cards, as required by ESSA, the data collection will be duplicative and unnecessary.

ACCURACY OF ESTIMATED BURDEN

It takes the CDE’s staff just over 200 hours to compile and validate the data for the state-level National Public Education Financial Survey report. It will take CDE significantly more than the estimated 125 hours to compile and validate data for over 10,000 schools and to complete the SLFS in the F-33 format. In addition, as California LEAs currently do not report school-level financial data to the CDE, the time and expense incurred by CDE and LEAs to develop systems and collect the data at the level of detail required in the SLFS would be prohibitive.

RECOMMENDATION

If the U.S. Department of Education (ED) proceeds with the school-level data collection, to minimize reporting burdens, the level of detail should be at most what is required by ESSA and should not begin until the ED can use the published state report card data once it becomes available.

Should you have any questions regarding these comments, please contact Christine Davis, Education Fiscal Services Administrator, School Fiscal Services Division, by phone at 916-322-1770 or by email at [email protected].

Sincerely,

Nick Schweizer, Deputy Superintendent

Services for Administration, Finance, Technology and Infrastructure Branch

RESPONSE:

Dear Mr. Schweizer and Ms. Davis,

Thank you for your feedback posted on November 15, 2016, responding to a request for comments on the School-Level Finance Survey (SLFS) 2016-2018 published in the Federal Register. The National Center for Education Statistics (NCES) appreciates your interest in the SLFS. The Paperwork Reduction Act (PRA) provides an opportunity for an open and public comment period where comments on collections can be made. We are grateful for this process and your comment.

SLFS Reporting Burden

The SLFS is consistent with ESSA regulations that require State and LEA report cards to contain current expenditures per pupil. The point that the SLFS data collection requires a greater granularity of expenditure tracking than ESSA requires, resulting in a greater reporting burden on LEAs does not take into account that “current expenditures” is actually a composite variable comprised of the data items found in the SLFS.

As a point of background, President Obama signed the Every Student Succeeds Act (ESSA) into law on December 10, 2015, which reauthorizes the Elementary and Secondary Education Act of 1965 (ESEA). ESSA requires state education agencies (SEAs) to produce report cards for the 2017–18 school year that include “per-pupil expenditures of Federal, State, and local funds, including actual personnel expenditures and actual nonpersonnel expenditures of Federal, State, and local funds, disaggregated by source of funds, for each LEA and each school in the State for the preceding fiscal year.” 20 U.S.C. 6301 et seq §1111 (h)(1)(C)(x).

Specifically, the regulations proposed by the Department of Education pursuant to ESSA provide that:

“Each State report card must include the following:

(i) Current expenditures per pupil from Federal, State, and local funds, for the preceding fiscal year, consistent with the timeline in § 200.30(e), for each LEA in the State, and for each school served by each LEA—

(A) In total (Federal, State, and local funds); and

(B) Disaggregated by source of funds, including—

(1) Federal funds; and

(2) (2) State and local funds combined (including Impact Aid funds), which must not include funds received from private sources.” 34 Code of Federal Regulations §200.35 (Federal Register, Volume 81 on May 31, 2016).

The definition of current expenditures clearly reflects that it is a composite variable. Current expenditures are comprised of expenditures for the day-to-day operation of schools and school districts for public elementary and secondary education, including expenditures for staff salaries and benefits, supplies, and purchased services. General administration expenditures and school administration expenditures are also included in current expenditures. Expenditures associated with repaying debts and capital outlays (e.g., purchases of land, school construction, and equipment) are excluded from current expenditures. Programs outside the scope of public prekindergarten through grade 12 education, such as community services and adult education, are not included in current expenditures. Payments to private schools and payments to charter schools outside of the school district are also excluded from current expenditures.

As a further point of background, current expenditures per pupil are calculated by dividing current expenditures by membership. As previously mentioned, current expenditures is actually a composite variable. Specifically, the SLFS is consistent with ESSA regulations in that there are seven primary variables and five exhibit items on the SLFS that are actually components of current expenditures per pupil.

There are two types of expenditures collected for the SLFS –“personnel” expenditures and “nonpersonnel” expenditures. For the SLFS collection, four personnel expenditures variables consist of gross salary and wage expenditures (including overtime, incentive pay, and bonuses) for school-level staff¹:

1. Instructional staff salaries,

2. Student support services salaries,

3. Instructional staff support services salaries, and

4. School administration salaries.

All four personnel variables on the SLFS are included within current expenditures. In order to report expenditures per pupil at the school level on state and LEA report cards under ESSA, the data for these four variables are found in the numerator of the current expenditures per pupil calculation. The SLFS also contains two personnel exhibit items that are included in the current expenditures per pupil calculation - Teacher salaries and Instructional aide salaries.

On the SLFS, there are three nonpersonnel variables that are also components of expenditures per pupil:

1. Instructional staff support,

2. Nontechnology-related supplies and purchased services, and

3. Technology-related supplies and purchased services.

In addition, the SLFS also contains three non-personnel exhibit items that are components of current expenditures - Improvement of instruction, Library and media services, and Books and periodicals.

According to interviews with a sample of SEA and LEA fiscal coordinators, all four personnel expenditures in SLFS can be directly tracked to the school-level. NCES and Census Bureau analysts invested substantial time and effort in consultation with SEA and LEA fiscal coordinators in order to determine which non-personnel expenditures could be effectively tracked to the school-level. There was a general consensus that both Nontechnology-related and Technology-related supplies and purchased services could effectively be collected and reported at the school level.

There are also three capital expenditure data items on SLFS that are not included within the current expenditures variable that is necessary to calculate expenditures per pupil for state and LEA report cards under ESSA. These capital expenditure data items─ Nontechnology-related equipment, Technology-related equipment, and Technology software are relatively easy to collect and report according to the consulted sample of SEA and LEA fiscal coordinators.

On balance, the SLFS is consistent in great degree with current expenditures per pupil to be reported in state and LEA report cards under ESSA because current expenditures is actually a composite variable that specifically includes and thus requires four personnel variables, three non-personnel variables, and five exhibit items from the SLFS. Therefore, the SLFS does not impose a significantly greater reporting burden on LEAs or SEAs than the state and LEA report cards required by ESSA.

Timing of Report Cards Proposed Under ESSA and SLFS

The proposed ESSA regulations provide that

“(1) Beginning with report cards based on information from the 2017-2018

school year, a State must annually disseminate report cards required under

this section for the preceding school year no later than December 31.” 34 Code of Federal Regulations § 200.30 (Federal Register, Volume 81 on May 31, 2016).

In contrast, participation in the SLFS is voluntary. The collection period for FY 16 SLFS data will be complete by the end of the calendar year, 2017; for FY 17 it will be complete by the end of the calendar year 2018; and for FY 18 it will be complete by the end of the calendar year 2019. NCES and the U.S. Census Bureau analysts are amenable to requests for extensions of time to submit SLFS data beyond the end of the respective calendar years set forth above.

School-Level Finance Data

In response to the point that financial data is not kept at the school-site level, the vast majority of SEA fiscal coordinators have indicated that they are able to report school-level finance data for the personnel variables in the SLFS. The pilot SLFS revealed that within seven of the nine reporting states, FY 14 data for all four personnel items were submitted for over 93 percent of their operating schools. NCES and Census Bureau analysts conducted extensive research on the feasibility of collecting the non-personnel variables, which included consultations with SEA fiscal coordinators and expert panels of LEA fiscal coordinators, researchers, and practitioners.

NCES and the Census Bureau analysts are aware that some finance amounts cannot be easily attributed to individual schools, such as itinerant teachers that teach at multiple schools. Many states have been able to allocate or prorate the salaries of teachers that teach at multiple schools to the school level. If a state is unable to do this we would note this in our documentation as a data anomaly.

Standardized Protocols

Your point pertaining to the development of standardized protocols is very helpful. We will further explore the need for and, if deemed helpful, create school-level finance codes and publish them in future editions of the accounting handbook entitled Financial Accounting for Local and State School Systems: 2014 Edition (NCES 2014-347)(Allison 2015). At a date to be determined, in 2017, we also plan to convene a national expert panel of school finance practitioners to provide advice on variable definitions, protocols, and methodology.

SEAs currently participating in the SLFS are receiving technical help and expertise from NCES and the Census Bureau during annual training sessions held with SEA fiscal coordinators, which facilitate their ability to provide expenditures per pupil on the state and LEA report cards. These training sessions may also facilitate the development of standardized protocols for collecting school-level finance data. NCES may be in a position to conduct future interactive webinars for state and local fiscal coordinators.

Data Editing

In response to the point that the resulting costs, while consistently determined, are not an accurate measure of the actual cost incurred at the school site, in accordance with NCES statistical standards, SLFS data are edited through an iterative and interactive process that includes procedures for detecting and correcting errors in the data. Data editing minimizes errors and ensures the data are complete, accurate, and consistent across the data file. In accordance with NCES Statistical Standard 4-1, SLFS data are checked for “credibility based on range tolerances to determine if responses fall within a pre-specified reasonable range” and “consistency based on checks across variables within individual records for non-contradictory responses” (NCES 2014).

After an SEA submits SLFS data, SLFS staff conduct a comprehensive review of the data, which includes numerous edit checks and, if necessary, follow-up with the SEA respondent. These edit checks include but are not limited to:

• “consistency” edit checks (e.g., teacher salaries cannot be greater than instructional staff salaries);

• outlier per pupil and per staff expenditure amounts;

• unreasonable zero dollar amounts;

• comparison of school district-aggregated SLFS data with F-33 data to ensure the SLFS data are within a reasonable range at the school district level; and

• comparison of state-aggregated SLFS data with NPEFS data to ensure the SLFS data are within a reasonable range at the state level.

SLFS staff analysts prepare follow-up questions for SEA respondents based on the results of these edit checks. SEAs are asked to explain all undocumented data anomalies and correct any data errors. If the SEA is unable to provide an explanation or revision for these anomalies, SLFS staff may edit the data as appropriate based on a set of established business rules.

Specifically, five data items are collected by NCES across all three CCD fiscal surveys (i.e., collected at the school level for the SLFS, at the LEA level for F-33, and at the state level for NPEFS): instructional staff salaries, student support services salaries, instructional staff support services salaries, school administration salaries, and teacher salaries.²

The amounts reported for these expenditures in the SLFS are compared to F-33 data. The median percentage difference of these expenditures between SLFS and F-33 is examined. The amounts reported on SLFS for these expenditures aggregated to the state level are also compared with NPEFS data.

The Department of Education’s Office of Civil Rights conducts the Civil Rights Data Collection (CRDC) on a biennial basis. The CRDC can be compared to the SLFS. For example, the CRDC collected the following school-level expenditures for SY 2013–14 (FY 14) that can be directly compared with SLFS data items: instructional staff salaries, total personnel expenditures, and teacher salaries. All CRDC school-level expenditures are reported with similar exclusions as the data items reported with exclusions on the SLFS.

Finally, the mean teacher salary calculated from SLFS can be compared with the average teacher salary reported by National Education Association (NEA) for the matching school years.

Necessity of Data Collection

The Department of Education’s mission is to promote student achievement and preparation for global competitiveness by fostering education excellence and ensuring equal access. Policymakers, researchers, and the public have voiced concerns about the distribution of school funding within and across school districts. School-level finance data address the need for reliable and unbiased measures that allow for comparison of how resources are distributed among schools within local districts. The development of the SLFS comports with the mission of the Department in that it provides data to compare how resources are distributed among schools within local districts and across districts.

There is a significant demand for finance data at the school level. NCES has consistently fielded questions about the availability of school-level finance data from researchers and the public over the past 10 years. In response to the growing demand, NCES developed a new collection of finance data at the school level. The SLFS is an extension of two existing collections being conducted by NCES in collaboration with the Census Bureau: the School District Finance Survey (F-33) and the state-level National Public Education Financial Survey (NPEFS). The SLFS is essentially an expansion of the F-33 to include some of the F-33 variables at the school level.

The SLFS is not Duplicative of ESSA Report Card Requirements

The demand for finance data on salaries for teachers, student support services, staff support services, and administration at the school level is very high and policymakers, researchers, practitioners, and the public are driving the high level of demand for the personnel expenditures set forth in SLFS. The SLFS will provide the needed information and not become duplicative or unnecessary once states and LEAs begin to publish school-level per-pupil expenditure data in their report cards, as required by ESSA.

SLFS is distinguished from the expenditure per pupil requirement in ESSA in that the four personnel variables, three non-personnel variables, and five exhibits on SLFS cover a portion of current expenditures at the school level, but not all of current expenditures. Current expenditure functions that are not collected on the SLFS include general administration (administration at the district level), operation and maintenance of plant expenditures, student transportation expenditures, other/business/central support services, and food service expenditures. These data items are components of expenditures per pupil at the state and district level. Furthermore, the personnel items on SLFS only cover salaries and wages and do not collect employee benefits, which are also a significant portion of current expenditures at the school level.

Additionally, in contrast to the ESSA report card requirements pertaining to current expenditures per pupil, the SLFS collects data on two non-personnel variables and one exhibit item that are capital expenditures, which are not components of current expenditures per pupil. Specifically, the Nontechnology-related equipment, Technology-related equipment, and Technology software data items are capital expenditures and not a factor in calculating current expenditures per pupil.

The Common Core of Data (CCD) is the primary NCES database on public elementary and secondary education in the United States. The SLFS is not duplicative to ESSA data in that the focus of CCD is to collect comparable data across all states as part of the comprehensive national statistical database of all public elementary and secondary schools and school districts. The challenge for NCES as a statistical organization is to inform the policy conversation by providing objective and comparable data that can be used to measure differences among schools and school districts based on the demographic characteristics of those schools and school districts. School-level finance data have been identified as one of the important ways of measuring how school-level education resources are distributed across and within school districts and examining the funding distributions across population characteristics (e.g., race/ethnicity make-up, poverty level, urban/rural, etc.) of those schools and school districts.

Accuracy of Estimated Burden

The response burden time is an estimated average time across all jurisdictions, which is based on input from previous and potential future respondents to the SLFS data collection. We expect that response times for some SEAs will be significantly higher and for some significantly lower than the estimated average. We recently completed the FY14 SLFS data collection and the results will be released in a research and development (R&D) report in the near future. Based on preliminary results, we believe that the response time estimates are generally reasonable. Also, Census Bureau analysts are willing to work with states to accept alternative reporting methods that may reduce respondent burden.

The following references apply to this response:

• Allison, G.S. (2015). Financial Accounting for State and Local School Systems: 2014 Edition (NCES 2015-347). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education, Washington, DC.

• U.S. Department of Education, National Center for Education Statistics. (2014). NCES Statistical Standards (NCES 2014-097). Washington, DC: U.S. Government Printing Office. Retrieved May 22, 2014, from http://nces.ed.gov/statprog/2012/.

Thank you again for your comment.

Sincerely,

Stephen Q. Cornman

Project Director

National Public Education Financial Survey (NPEFS)

School District Finance Survey (F-33)

School-Level Finance Survey (SLFS)

US Department of Education

1 Personnel expenditures exclude employee benefits (e.g., employer health insurance and retirement contributions) and exclude expenditures for staff that would typically be considered school district-level staff (e.g., student transportation and operations and maintenance staff).

2NPEFS and F-33 amounts are compared to state-aggregated and school district-aggregated SLFS data amounts “without” exclusions (as opposed to “with” exclusions).

7