Academic Intervention Services

Karl Widerquist, Staff Economist

The Educational Priorities Panel

Introduction

The creation of Special Needs/Academic Intervention Services (SN/AIS or simply AIS for short) is the largest restructuring of the allocation of funds to New York City’s Community School Districts (CSDs) in more than a decade. It replaces forty-five[1] individual allocation programs recognizing various special needs of pupils with one comprehensive, block-grant allocation. According to the Board of Education’s Division of Budget Operations and Review the goals of the new formula are to:

· Simplify the formula allocation process.

· Add flexibility in combining resources to design programs and services.

· Ensure Stable Funding to allow provision/intervention as well as mandated programs.

· Enable Financial Planning through retention of tax levy accruals…

· Break Down Barriers between different service categories.

· Broaden Focus from special education pupils to pupils with special needs.[2]

Much of the reason for creating AIS was to support special education reform so that students could receive the services they need while removing the financial incentive to classify students as “special education” if it is not absolutely necessary. Another reason for AIS was to give districts more flexibility to use their resources.

Replacing a large number of small spending mandates with one block grant will certainly add flexibility. One would hope it would also increase equity. This report examines the equity effects of AIS and demonstrates that more needs to be done to increase equity in the allocation of funds to CSDs. AIS may help special education reform but it does not provide enough resources for high-needs districts to prevent continued academic failure. Part one describes how the new Special Needs/Academic Intervention Services works. Part two examines the effect of AIS on the distribution of funds toward districts that have large concentrations of high-needs pupils. The major finding of this analysis is that the creation of AIS itself does not increase the portion of funds going to needy districts, and in many cases, it actually redirects funds toward less-needy districts. The formula does not make it easy understand why funds are directed toward particular districts. To some extend the allocation formula still rewards districts for poor performance. The report concludes by discussing how the Board of Education could achieve its goal of greater flexibility and simplicity while also increasing equity and eliminating rewards for poor performance.

Both state and federal aid to school districts recognize poverty as a factor that requires greater education funding. But within New York City, the distribution of aid to school districts does not make poverty as central criterion for the allocation of funds. Reimbursable funds are more than 59% higher in the eight districts with the largest percentage of students receiving free lunch (an indicator of poverty) and Special Needs/Academic Intervention Services are more than 37% higher. But tax levy funds, by far the largest portion of funds, are basically flat per pupil, making the overall budget less than 19% higher in the eight districts with the largest percentage of students receiving free lunches than in the eight districts with the lowest percentage of students receiving free lunches. As Table 1 shows, the eight districts with the highest level of poverty receive on average $1,103 more than the eight districts with the lowest level of poverty.

Table 1: Per pupil CSD allocation and budget by percent receiving free lunch

	Percent
	free lunch	AIS	Tax levy	Reimbursable	budget	Title 1
1st Quarter	54.25%	1,216	4,546	1,372	5,918	245.51
(lowest % free lunch)
2nd Quarter	74.63%	1,294	4,458	1,542	6,000	450.00
3rd Quarter	87.89%	1,626	4,796	2,016	6,812	605.30
4th Quarter	93.78%	1,675	4,832	2,189	7,021	673.34
(highest % free lunch)
Simple Average	77.63%	1,453	4,658	1,780	6,438	494
Percentage difference
between 1^st and 4^th quarters		37.74%				174.26%

This is a small difference if it is compared to state Operating Aid, which directs about six times as much funds to the least wealthy one-fifth of independent school districts than to the wealthiest one-fifth, but it is quite progressive compared to the total budgets of independent districts statewide. Wealthier districts nearly always spend more than less wealthy districts. The wealthiest one-fifth of independent school districts spend on average nearly $16,000 a year per pupil—well more than double the funds that are available to any CSD in New York City.[3] No amount of redirection of funds within New York City can make up for the fact that city schools are greatly under-funded compared to less-needy suburban districts, but the direction of funds within New York City is still an important issue that needs to be considered.

Part One: The rules of Special Needs/Academic Intervention Services

According to the FY 2001 BOR Allocation Memorandum No. 1, the methodology for the AIS formula is as follows:

The Special Needs/Academic Intervention Services allocation formula allocates teacher positions and a fixed dollar amount based on a “virtual” special needs register. The special needs register is comprised of a percentage of the following populations:

· Student Population: 8%. October 31, 1999 audited register, grades kindergarten to 12 general and special education, adjusted for long term (sic) absentees (no shows).

· Academic Criteria: 11%. From FY 1999 testing, general and special education pupils scoring at or below the 25^th percentile on the Reading or Math CTB, grades 3, 5, 6, and 7; Level 1 on the State Reading or Math exam, grades 4 and 8. The percent below is applied to the October 31, 1999 audited register to produce the estimated number of pupils below academic criteria. This action compensates for untested pupils, i.e. grades kindergarten, 1, 2, and 9 and above.

· Low Income: 6%. October 31, 1999 free lunch eligible pupils—general and special education.

· Limited English Proficiency: 2%. The number of English language learners scoring at or below the 40^th percentile on the Language Assessment Battery test.

· Mobility: group of 1,500 pupils. Expected pupil mobility in a group of 1,500 pupils. Mobility is expressed as a percentage. It is the number of pupils entering and/or leaving a school’s register after October 31^st, divided by the October 31^st register plus all students who entered after October 31^st.

· Self-Contained Special Education: 25%. Audited October 31^st, 1999 register of pupils enrolled in self-contained special education classes excluding inclusionary pupils…and long term (sic) absentees.

· Teacher Certification: one pupil for every two uncertified teachers. Division of Human Resources’ FY’00 uncertified teacher data.

Applying the needs factor to the respective populations produces a systemized special needs (sic) “virtual” register of just under 150,000 pupils. This represents 20% of the entire CSD pupil enrollment.

The Special Needs register drives the district allocation of teacher positions and dollars. These amounts were calculated using the current special education program, but applied to all the students in the special needs “virtual” register.

· Teacher Positions: 1 position per 9.85 special needs pupils.

· Fixed Dollar Amount: $2,190 per special needs pupil.[4]

Although this formula is still rather complex, it would certainly be a great simplification if it were to fully replace the old system of 45 different methodologies. However, the AIS formula would require $155 million more than districts would receive under each of the 45 old methodologies, and the Board of Education has only been successful in identifying $85 million of new and redirected funding and was, therefore, not able to fully put the new system in place. Instead, the Board opted for a maximum increase (of 13.85%) and a minimum increase (of 1.5%) over what the districts would have received had the old system remained in place.

Of the 32 Community School Districts, 26 of them are at one of the two limits—4 at the lower limit and 22 at the upper limit. Certainly, there is a substantial difference between an increase of 1.5% and an increase of 13.85%, but the fact that more than three-fourths of the CSDs are at one limit or the other demonstrates that the old system of 45 different allocation formulas is still partially in effect and is still having a very substantial impact on the level of aid received by each district. Great simplification for CSD officials is provided by the fact that the mandated funding of the old system is gone and replaced by a block grant, but the complex method of calculating the amounts of funding each CSD is eligible for is still partially in effect.

Table 2 (parts A, B, and C) summarizes the rules of the old allocation system. Most of the allocation programs that have been consolidated into the new system (30 out of 45) were based on the number of special education students. The others are based on some other concept of need[5] (5), per capita (1), random[6] (3), discretionary (1), or some combination these (3). The criteria for two of the allocation programs [Attendance Child Abuse/Neglect (5) and AIDP alternatives to suspension (40)] were not explained in the FY2000 BOR Allocation Memorandum.

Aside from special education, the needs-based programs are all based on low achievement rather than on socio-economic or categorical need. Allocating funds in such a manner rewards failure and punishing success. If a district with greater socio-economic needs has low test scores, it will be allocated extra funds. If the district successfully uses those funds to bring test scores up, the funds will be taken away until test scores drop again. Thus, for over a decade the Educational Priorities Panel has called for replacing allocation programs based on low achievement with programs based on socio-economic need. AIS makes at least some attempt to do this by allocating 6% of its funds based on the number of low-income students in each district. This explicit incorporation of this factor is a substantial improvement, but this is still a relatively small portion of funds. It is only a little over half the funds given to districts with low-achieving students.

The Board of Education argues that the larger context is that high-poverty districts already receive more funds and that over the past 3 years there have been substantial increases in funding for high-poverty districts. These increases, however, reflect a mix of federal, state, and city funding sources. This report examines whether the shift to AIS shifts more funds toward students that are at risk of being held back. And, as part two demonstrates, the new AIS system does not direct more funds to CSDs with greater socio-economic needs.

Table 2 (parts A, B, and C): The rules of the old allocation system

2A: GENERAL EDUCATION-TAX LEVY
1. Special Needs, Module 2B	Per student falling below a specific score in reading or math.
2. Special Education Class Coverage	Based on the number of special education students.
3. Special Education Pupil Mainstreaming	Based on the number of periods a student is mainstreamed
4. Section 504 Paraprofessional	Based on recommendations of the Division of Student Support Services, or FY99 needs.
5. Attendance Child Abuse/Neglect	Not Explained in BOR
6. Kindergarten Paraprofessional	Random, based on school utilization.
7. Latchkey	Random, reimburses districts that happen to have a latchkey program.
2B: SPECIAL EDUCATION-TAX LEVY
8. Classroom Teacher	Based on the number of special education students.
9. Crisis Intervention Teacher
10. Supervisor
11. Industrial Arts Teacher
12. Teacher Trainer
13. Language Coordinator
14. Basic Paraprofessional
15. Counselor, Related Service
16. Speech Teacher, Related Service
17. Teacher Absence
18. Basic Paraprofessional Service
19. Clerical Support for Supervisors
20. OTPS & Other Support Services
21. OTPS Industrial Arts
22. OTPS Related Service, Speech/Counsiling
23. OTPS Supervision and Support
24. IEP Driven Paraprofessional & Absence
25. Bilingual Paraprofessional & Absence
26. Health Coordinator
27. Budget Reduction
28. Restoration Prevention/Intervention
29. Fractional Pool
30. Pupil Suspension Hearings (New)

2C: REIMBURSABLE
31. ERSSA, Educational Related Support Services, Per Capita	1/3 Per capita, 1/3 low test scores, 1/3 on the number of declassified students
32. IDEA, Individuals with Disabilities Education Act Per Capita	Based on the number of special education students
33. IDEA, Least Restrictive Environment Pilot
34. IDEA, IEP Equipment
35. School Age Plus
36. IDEA, School Age Plus
37. PCEN, Pupils with Compensatory Educational Needs	Based on number of students who fail reading and math exams
38. AIDP, Attendance Improvement/Dropout Prevention, Elementary & Middle Schools	Based on attendance with more funds being given to schools with low attendance
39. ERSSA, AIDP Transitional Program	Based on number of students decertified as special education.
40. AIDP, Alternatives to Suspension	Not explained in BOR, see Division of Student Support Services Memorandum #67, June 10^th, 1998.
41. Chapter 53 Reading, Per Capita	Partly per capita, partly based on low reading, math, and LAB scores, and party based on the number of teachers with less than five years experience.
42. SIG/IPP, State Incentive Grant-Improving Pupil Performance, Per Capita	Per capita
43. SIG/IPP, Quality Improvement Program Plan, QUIPP	Partly per capita, partly based on the number of resource room teachers, speech teachers, crisis intervention teachers and basic paraprofessionals in each district
44. Title VI, Consultant Teacher	Based on the number of special education students
45. AIDP Welcome Centers	Random, reimburses districts that have welcome centers

Part Two: How well do needy students fair under AIS?

The New York State Education Department identifies poverty, limited English ability, and racial/ethnic group identity as factors associated with need[7] among general education students:[8] Although state and federal aid to schools recognize these need factors and at least partially allocate funds accordingly, these factors have played a relatively small role in determining the level of funds allocated to CSDs as Table 3 illustrates. This section examines how much AIS helps districts with high concentrations of students in four needs categories: the percentage of students receiving free lunches (an indicator of the level of poverty in the district), the percentage of students who are English Language Learners (ELL), the percentage of nonwhite students, and the percentage of special education students. Data for the first three of these comes from the Citizens’ Committee for Children,[9] and data for the percentage of special education students was calculated from the number of special education students and the total enrollment figures reported in the FY 2001 BOR allocation summary.

To a large extent, AIS gives the largest increases to districts with lower needs, but mostly the increases are random with some high-needs and some low-needs districts receiving large increase and with some high-needs and some low-needs districts receiving small increases. For additional statistics on the allocation of funds by student needs see Appendix Tables 1 and 2.

Table 3 compares the 4 districts receiving the minimum increase to the 22 districts receiving the maximum increase in terms of the average percentage of students in each of these needs categories. On average, those receiving the maximum increase have a lower percentage of students in each of these four needs categories than those receiving the minimum increase: Districts receiving the minimum increase have an average of 82.3% of their students receiving free lunches, while districts receiving the maximum have an average of only 78.9%. Districts receiving the minimum have an average of 17.5% of their students who are English Language Learners (ELL), while districts receiving the maximum have an average of 17.4%. Districts receiving the minimum increase have an average of 91.6% nonwhite students, while districts receiving the maximum have only 85.9% nonwhite. This may sound like a small difference and clearly it shows that most districts in New York City have a large percentage of nonwhite students, but that 5.7 percentage point difference means that districts receiving the maximum increase have about 67% more white students than districts receiving the minimum. Districts receiving the minimum increase have an average of 17.8% of their students classified as special education while districts receiving the maximum have an average of only 11.2% special education. That is, districts receiving the minimum increase have an average of nearly 60% more special education students than districts receiving the minimum.

Table 3: Comparison of districts receiving the minimum and maximum increases

	Average percent receiving free lunch	Average percent ELL	Average percent nonwhite	Average percent special education
Districts receiving the minimum increase (1.5%)	82.3%	17.5%	91.6%	17.8%
Districts receiving the maximum increase (13.85%)	78.9%	17.4%	85.9%	11.2%

Looking at the data for each district reveals the same pattern or lack of pattern. Table 4 ranks the CSDs by the percentage of students receiving free lunch, and demonstrates that the level of poverty in a district seems to have little relationship with whether a district receives the minimum or maximum increase. The district with the highest percentage receiving free lunches receives the maximum increase, but the second-ranked district receives the minimum increase. The district with the second-lowest percentage of students receiving free lunches receives just above the minimum increase, but the district with the third-lowest percentage receives the maximum increase.

Table 4
CSD	Percentage change	Percent free lunch	Rank by percent free lunch
9	13.85%	96.20%	1
7	1.50%	95.00%	2
6	13.85%	94.70%	3
16	13.85%	94.30%	4
32	13.85%	93.80%	5
12	10.05%	92.50%	6
23	13.85%	92.30%	7
19	13.85%	91.40%	8
5	13.85%	91.10%	9
4	2.69%	90.50%	10
1	1.50%	89.90%	11
14	13.85%	89.70%	12
17	13.85%	89.10%	13
10	11.32%	89.00%	14
8	8.64%	84.80%	15
21	13.85%	79.00%	16
11	1.50%	76.90%	17
30	13.85%	76.70%	18
15	13.85%	76.60%	19
18	13.85%	75.10%	21
27	13.85%	75.10%	20
24	13.85%	73.40%	22
29	13.85%	71.80%	23
28	13.85%	71.40%	24
13	13.85%	69.10%	25
3	1.50%	67.50%	26
20	13.85%	66.20%	27
22	13.85%	65.20%	28
2	13.85%	54.70%	29
25	13.85%	49.70%	30
31	1.59%	40.40%	31
26	5.15%	21.20%	32

Table 5 ranks the districts by the percentage of students who are English Language Learners (ELL). Again whether a district has a high or a low percentage of ELL students has little relationship to whether the district receives the minimum or maximum increase or something in between. The district with the highest percentage of ELL students receives the maximum increase, but so does the district with the lowest percentage of ELL students. Those receiving the minimum increase, tend to be in the middle of the distribution.

Table 5

CSD	Percentage change	Percent ELL	Rank by percent ELL
6	13.85%	49.80%	1
24	13.85%	29.60%	2
10	11.32%	28.90%	3
9	13.85%	28.10%	4
30	13.85%	27.60%	5
20	13.85%	27.00%	6
7	1.50%	25.10%	7
32	13.85%	24.70%	8
12	10.05%	22.60%	9
1	1.50%	20.40%	10
25	13.85%	19.10%	11
15	13.85%	18.40%	12
21	13.85%	18.30%	13
14	13.85%	18.00%	14
2	13.85%	17.80%	15
19	13.85%	16.60%	16
28	13.85%	15.20%	17
8	8.64%	14.70%	18
4	2.69%	14.60%	19
3	1.50%	14.20%	20
5	13.85%	12.90%	21
22	13.85%	12.10%	22
27	13.85%	10.70%	23
11	1.50%	10.10%	25
26	5.15%	10.10%	24
17	13.85%	9.80%	26
29	13.85%	8.30%	27
13	13.85%	5.80%	29
23	13.85%	5.80%	28
18	13.85%	5.50%	30
31	1.59%	4.30%	31
16	13.85%	2.40%	32

Table 6 ranks the CSDs by the percentage of nonwhite students in the district. Again there is little relationship between the percentage of nonwhite students and whether the district receives the minimum or maximum increase. The district with the fewest nonwhite students receives just above the minimum increase, while the district with the second-smallest portion of nonwhite students receives the maximum increase. The district with the largest percentage of nonwhite students receives the minimum increase while the district with the second largest percentage of nonwhite students receives the maximum increase.

Table 6

CSD	Percentage change	Percent nonwhite	Rank by percent nonwhite
7	1.50%	99.8%	1
16	13.85%	99.7%	2
9	13.85%	99.6%	3
23	13.85%	99.6%	4
5	13.85%	99.5%	5
17	13.85%	99.4%	6
12	10.05%	99.2%	7
4	2.69%	98.8%	8
6	13.85%	98.5%	9
13	13.85%	98.3%	10
19	13.85%	98.3%	11
29	13.85%	97.6%	12
32	13.85%	97.40%	13
1	1.50%	94.1%	14
10	11.32%	93.4%	15
18	13.85%	90.7%	16
14	13.85%	90.5%	17
8	8.64%	90.0%	18
11	1.50%	88.1%	19
3	1.50%	84.3%	20
30	13.85%	83.1%	21
27	13.85%	82.7%	22
28	13.85%	81.1%	23
24	13.85%	80.5%	25
15	13.85%	78.2%	24
25	13.85%	70.7%	26
22	13.85%	68.8%	27
2	13.85%	67.7%	29
26	5.15%	58.8%	28
21	13.85%	55.6%	30
20	13.85%	52.9%	31
31	1.59%	36.5%	32

Table 7 ranks the CSDs by the number of students classified as special education. There is some randomness in this category as well, but it is apparent that districts with higher percentages of special education students generally receive smaller increases. All of the districts receiving the minimum increase are ranked in the top 8 in terms of the number of special education students. Seven out of the top 8 districts receive less than the maximum increase, while 14 of the 15 lowest ranked districts receive the maximum increase. The shift of funds away from students classified as special education to those classified as general education is part of the Board of Education’s effort to reduce the financial incentive to classify students as special education in order to receive funding. The Board of Education argues that this effort has succeeded in reducing the number of students classified as special education while providing services to meet those student’s needs.

Table 7

CSD	Percentage change	Percent special ed	Rank by percent special education
7	1.50%	21.06%	1
1	1.50%	20.44%	2
12	10.05%	20.06%	3
4	2.69%	16.14%	4
16	13.85%	15.65%	5
10	11.32%	15.34%	6
11	1.50%	15.23%	7
3	1.50%	14.66%	8
14	13.85%	14.37%	9
23	13.85%	14.32%	10
5	13.85%	13.91%	11
8	8.64%	13.81%	12
9	13.85%	13.65%	13
32	13.85%	13.24%	14
27	13.85%	12.69%	15
19	13.85%	12.67%	16
31	1.59%	12.57%	17
21	13.85%	12.53%	18
15	13.85%	12.24%	19
13	13.85%	11.08%	20
17	13.85%	11.00%	21
20	13.85%	9.96%	22
28	13.85%	9.92%	23
18	13.85%	9.70%	25
22	13.85%	9.58%	24
6	13.85%	9.57%	26
25	13.85%	9.36%	27
26	5.15%	9.26%	29
30	13.85%	9.02%	28
29	13.85%	8.94%	30
2	13.85%	8.87%	31
24	13.85%	4.56%	32

Comparison of six districts

Table 8 compares 6 districts in terms of the size of their allocation increase and in terms of the same 4 need characteristics. The districts compared are the 3 districts with the lowest percentage of students receiving free lunches (districts 25, 31, and 26) and the districts with the highest percentage of students receiving free lunches (districts 6, 7, and 9). As one can see from the table, districts with high percentage of students receiving free lunches tend also to be high in other need categories as districts with a low percentage of students receiving free lunches tend also to be low in other need categories.

District 26 has the fewest students receiving free lunches of any CSD. It ranks 24^th (out of 32 CSDs) in the percentage of students with limited English proficiency, 29^th in percentage of students who are nonwhite, and 29^th in percentage of students classified as special education, yet it receives a higher increase than district 7 both in percentage and per pupil terms, despite the fact that district 7 ranks highly in every need category. It ranks 2^nd in percent free lunch, 7^th in percent ELL, 1^st in percent nonwhite, and 1^st in percent special ed. Arguably district 7 has the most at-risk pupils of any district in the city, yet it receives the minimum increase in funding, apparently held there only by the hold harmless provision of the AIS rules, and it would be receiving even less if the new AIS allocation formula were to go fully into effect.

District 31 ranks low in all 4 needs categories. It ranks 2^nd to last in percent free lunch and percent ELL, last in percent nonwhite, and 18^th (just below the median) in percent special ed. District 31 receives a relatively small increase, but it is still larger in percentage terms than district 7’s increase. What then does the shift to AIS have to do with broadening the focus of categorical funding to serve all students with special needs?

Districts 25, 6, and 9 all receive the maximum increase of 13.85%, but district 25 is a relatively low-needs district while districts 6 and 9 are high-needs districts. District 25 has relatively few at-risk students in terms of poverty, race, and special education, although it does have more than the median number of ELL students (ranking 11^th) it ranks below districts 6, 7, and 9 in this category as well. Despite being ranked lower in every needs category than district 7, district 25’s increase is 8 times as large in absolute terms, 4 times as large in per pupil terms, and 5 times as large in percentage terms. What then does AIS have to do with getting more money to at-risk students? It is difficult to tell from the AIS formula exactly why district 25’s increase is so much larger than district 7’s.

Another factor to be aware of is the randomness of the changes. District 7 has similar needs characteristics to districts 6 and 9, but receives less than one-fifth of the increase in percentage terms than either of those districts. Again, it is difficult to tell from the AIS allocation formula exactly why this is happening and what goal it serves.

Table 8: A comparison of 3 relatively advantaged with 3 relatively disadvantaged districts

CSD number	26	31	25	6	7	9
Total increase	787,388	709,044	3,102,394	4,599,286	387,527	5,281,958
Per pupil increase	50	19	132	170	32	198
Total per pupil budget	5,426	5,737	5,672	6,324	7,582	6,796
Percentage increase	5.15%	1.59%	13.85%	13.85%	1.50%	13.85%
Relative size of increase	middle	near minimum	maximum	maximum	minimum	maximum
Percent free lunch	21.20%	40.40%	49.70%	94.70%	95.00%	96.20%
Percent ELL	10.10%	4.30%	19.10%	49.80%	25.10%	28.10%
Percent nonwhite	58.80%	36.50%	70.70%	98.50%	99.80%	99.60%
Percent special ed	9.26%	12.57%	9.36%	9.57%	21.06%	13.65%
Rank by percent free lunch	32	31	30	3	2	1
Rank by percent ELL	24	31	11	1	7	4
Rank by percent nonwhite	29	32	26	9	1	4
Rank by percent special ed	29	18	28	27	1	13

How does AIS compare to an across-the-board increase in funding?

AIS gives an average of a 10.95% increase in funding. Thus, we can easily compare the practical effect of the switch to AIS for each district as Table 9 does below. 23 districts do slightly better with AIS, but of the 9 districts that would do better with an across-the-board increase, many of them would do substantially better. Most districts receive the maximum increase of 13.85% and therefore would receive a 2.90% less if the old allocation system were retained with across-the-board increase of 10.95%. Because some districts receive much smaller increases, some districts would have as much as 9.45 percentage points more with an across-the-board increase. A quick look at Table 9 shows that the districts that would do substantially better with an across-the-board increase are spread through the distribution in terms of percentage of students receiving a free lunch—from the second to the last ranked district. One of the key problems with AIS, is a lack of transparency. Neither the rules describing AIS allocations nor the rules describing the 45 allocation programs it replaced give one a good understanding why these particular districts receive increases that are so much smaller than other districts with similar needs.

Table 9: How does AIS compare to an across-the-board increase?

CSD	Percent free lunch	Rank by percent free lunch	Percentage change in funding	Difference with an across the board increase
9	96.2%	1	13.85%	-2.90%
7	95.0%	2	1.50%	9.45%
6	94.7%	3	13.85%