1. Public Finance Seminar Spring 2019, Professor Yinger
State Aid

2. The Education Finance System Types of State Aid Research on State Aid
Class Outline
The Education Finance System
Types of State Aid
Research on State Aid

3. The U.S. Education Finance System
State Aid
The U.S. Education Finance System
State elected officials make the rules for elementary and secondary education.
State elected officials also design the education finance system:
State aid (≈ 47% of total in average state)
Property taxes and perhaps other local taxes (≈ 45%)
Compensation for homestead exemptions
The federal government provides a little funding (≈ 8%), tax breaks for property taxes, and some incentives (through NCLB).

4. State Aid
Data Note
It is hard to pin down the sources of education spending.
The Census collects data only for independent school districts.
Many districts, including Syracuse, are dependent, that is, they are departments in a city government.
One effort to account for this (McGuire, Papke, and Reschovsky in the Handbook of Education Finance and Policy, 2nd Edition) finds that in :
Local sources provided 43.1% of school K-12 school revenue,
Local property taxes supplied 82.1% of this local revenue,
States provided 44.1%,
The federal government provided 12.5%.

5. EDUCATION FINANCE IN THE U.S.
State Aid
EDUCATION FINANCE
IN THE U.S.
School District
Local Property Taxes
State Aid
Federal Aid
Local Voters
State Regulators
Authority
Revenue

6. The Education Finance System, 2
State Aid
The Education Finance System, 2
The broad rules are laid out in a state constitution, which has phrases such as “a system of free public schools’ or “a sound, basic education.”
Elected officials design a system that meets their objectives, which usually (but not always!) do not involve much re-distribution.
People in low-performing districts bring suits into the state courts, and the state courts rule on the constitutionality of the system designed by elected officials.

7. The Education Finance System, 3
State Aid
The Education Finance System, 3
“The education of children is a fundamental value of the people of the State of Florida. It is, therefore, a paramount duty of the state to make adequate provision for the education of all children residing within its borders. Adequate provision shall be made by law for a uniform, efficient, safe, secure, and high quality system of free public schools that allows students to obtain a high quality education and for the establishment, maintenance, and operation of institutions of higher learning and other public education programs that the needs of the people may require.” Fla. Const. art. IX, § 1(a).
As a result of the 2002 amendments, Florida’s constitution also requires the legislature to make adequate provision for reduced class sizes, and provides that every four-year-old child in the state have access to a "high quality pre-kindergarten learning opportunity." Fla. Const. art. IX, § 1(b).
Prior to 1998, the constitution simply required the state to make “[a]dequate provision…for a uniform system of free public schools.”

8. State Aid
The Role of the Courts
1971: Serrano decision by California Supreme Court rejected California’s education finance system based on U.S. and California Constitutions.
It is unfair, the court said, for a child’s education to depend on the wealth of his school district.
1972: Rodriquez decision by the U.S. Supreme Court ruled out education claims based on U.S. Constitution.

9. The Role of the Courts, 2 Since 1971: State Aid
43 state courts have heard challenges to their state’s education finance system.
20 more education finance systems have been declared unconstitutional by a state supreme court.
Court decisions have led to major education finance reforms in many states, including California, Kentucky, New Hampshire, New Jersey, Texas, Vermont.
Several states have implemented major reforms without a court mandate, including Kansas, Maryland, and Michigan.

10. State Aid
The Role of the Courts, 3
One careful accounting of all the court cases can be found in:
S.P. Corcoran and W. N. Evans, “Equity, Adequacy, and the Evolving State Role in Education Finance,” Handbook of Research in Education Finance and Policy, 2015.

11. FLORIDA SUPREME COURT REJECTS ADEQUACY CLAIMS
State Aid
The Role of the Courts, 4
areas/schoolfundinginfo/
FLORIDA SUPREME COURT REJECTS ADEQUACY CLAIMS
In a close 4-3 decision in which the deciding judge joined in the result but not the opinion of his colleagues, the Florida Supreme Court earlier this month rejected a ten- yearlong attempt by public school students, parents and citizen organizations to have the state's education finance system declared unconstitutional. Citizens for Strong Schools, Inc v. Florida State Board of Education.
The Florida constitution provides that education is "a paramount duty" of the state and requires that "[a]dequate provision" shall be made by law for a "high quality" education system. The plurality ruled that these terms do not provide "any manageable" standard for the court to use in assessing the adequacy of the amount of funding that the schools receive. The Court rejected plaintiffs' position that the state's educational standards provide such a standard, because
Adopting State standards as constitutional minima would have the perverse effect of encouraging the weakening of curriculum standards in order to achieve the higher passage rates and to satisfy court-imposed requirements. 

12. Educational Disparities
State Aid
Educational Disparities
What the courts are responding to are enormous disparities in both resources and student performance across school districts within a state.
This disparities arise because districts vary widely in both available resources and educational costs.
Large cities, which often contain many poor families, have particularly low student performance.

13. Sources of Student Performance Gaps State Aid Demand Supply
Poverty, Limited English, Wages
Disparities in Income & Wealth
Gaps in Education Resources
Gaps in Education Costs
Gaps in Student Performance
Demand
Supply

14. Educational Disparities, Example
State Aid
Educational Disparities, Example
NAEP 4th Grade Math, 2015
Percent Proficient
City
State
City/State
Albuquerue 28 27
103.7%
Atlanta 26 35
74.3%
Austin 47 44
106.8%
Baltimore 12 40
30.0%
Boston 33 54
61.1%
Charlotte 51
115.9%
Chicago 30 37
81.1%
Cleveland 13 45
28.9%
Dallas 34
77.3%
Detroit 5
14.7%
Fresno 14 29
48.3%
Houston 36
81.8%
Los Angeles 22
75.9%
Miami 41 42
97.6%
NYC
Philadelphia 15
33.3%
San Diego 31
106.9%
Source: U.S. Department of Education

15. Educational Disparities, Example Cont.
State Aid
Educational Disparities, Example Cont.
These 21 cities are the ones who volunteered to participate in the NAEP (National Educational Assessment Program) tests in 2015.
Many disadvantaged cities (e.g., Buffalo, Rochester, Syracuse, Milwaukee, Newark) are not included.
The state averages include the cities; if the city is a large share of state population, the score for “rest of state” could be very different from the score for the state as a whole.
In the case of NYC, e.g., which has 39.3% of New York State’s students, the city proficiency rate is only 63.7% of the rate in the rest of the state.
Similar disparities exist for 8th grade math and for 4th and 8th grade reading.

16. State Aid
Source: New York State Education Department

17. State Aid
Source: Yinger, J. “Unequal Access to Good Schools in New York State.” .

18. Key Questions in Education Finance Reform
State Aid
Key Questions in Education Finance Reform
1. What is the best way to measure the education provided by a school district?
What do we want schools to provide?
2. What is the appropriate equity standard?
How would we recognize a fair outcome if we saw it?

19. State Aid
Measuring Education, 1
Spending per pupil is a poor measure of education
Easy to measure, but hard to interpret
Rejected by most courts
Spending ignores variation in the cost of education due to
Concentrated disadvantage among students
The high cost of attracting teachers to some districts

20. State Aid
Measuring Education, 2
Pupil performance, such as test scores and drop- out rates, provides a better measure of education
It corresponds with what parents want
It is consistent with the trend toward setting higher standards
It is consistent with accountability programs (discussed next class)
An education cost index provides a bridge between performance and spending
An equivalent approach is using higher “weights” for disadvantaged students

21. State Aid
Educational Adequacy
Adequacy is the equity standard emphasized in most recent court decisions and by most policy makers:
Every student should be in a school that delivers an adequate average student performance.
An adequacy standard does not eliminate all disparities:
Districts are allowed to provide above-adequate educations if they can.
Adequacy is achieved by a foundation aid program (discussed below).

22. Other Equity Standards
State Aid
Other Equity Standards
Access Equality
The education provided by a district should depend only on its property tax rate.
Refers to fairness for taxpayers, not students.
Is achieved by power-equalizing aid (discussed below)
Was the main issue in Serrano.
Wealth Neutrality
Educational outcomes should not be correlated with school district wealth.
Difficult to achieve.
Equality
All school districts should provide the same level of education.
May require state provision, as in Hawaii.

23. The Foundation Aid Formula
State Aid
The Foundation Aid Formula
40 states use a foundation aid formula, designed to achieve educational adequacy.
The foundation aid formula is
Aj = aid per pupil to school district j.
E* = foundation spending per pupil (state-selected; the same in every district).
t* = minimum required property tax rate (state-selected; the same in every district).
Vj = actual property tax base per pupil in district j.

24. The Foundation Aid Formula, 2
State Aid
The Foundation Aid Formula, 2 E*
Spending per Pupil
Tax Base per Pupil
Vmin
Vmax
Required Local Taxes
Foundation
Spending
Level
Spending Above Foundation Level
Aid

25. The Foundation Aid Formula, 3
State Aid
The Foundation Aid Formula, 3
A foundation aid formula can easily be adjusted for educational costs (that is, focused on performance):
S* = foundation spending level per pupil in a district with average costs
Cj = educational cost index for district j.
Pupil weights can also be used.
This formula is equivalent to offsetting fiscal disparities across schools.

26. The Foundation Aid Formula, 3
State Aid
The Foundation Aid Formula, 3
A foundation formula must address four issues:
1. How much spending is “adequate”?
2. Should the foundation level be adjusted for variation in education costs across districts?
3. Should a minimum local property tax rate be required?
4. How should burden of funding an adequate education be distributed?

27. What Spending Is Adequate?
State Aid
What Spending Is Adequate?
Rose decision in Kentucky:
“sufficient oral and written communication skills to enable students to function in a complex and rapidly changing civilization”
“sufficient understanding of governmental processes to enable the student to understand the issues that affect his or her community, state, and nation”
“sufficient levels of academic or vocational skills to enable public school students to compete favorably with their counterparts in surrounding states, in academics or in the job market”
CFE decision in NY:
“meaningful high school education, one which prepares them to function productively as civic participants”

28. Adjust for Costs? State Aid
Districts with higher costs must spend more to achieve any given performance level.
So a cost adjustment is needed to combine a performance definition of education with an adequacy standard.
According to Verstegen and Knoeppel (J. Ed. Fin. 2012), 36 states adjust aid for student poverty and 42 states adjust for English language learners.
But no state does a comprehensive, estimated cost adjustment.
Full cost adjustment leads to extensive re-distribution, which is politically difficult, and requires statistical procedures.

29. Require Minimum Tax Rate?
State Aid
Require Minimum Tax Rate?
To reach foundation spending, E*, a district must levy at least the selected tax rate, t*.
But when a school district receives state aid, it only spends some of the money on education – the rest goes to relief from local taxes.
Thus, the foundation level of spending will not be achieved unless a minimum rate of t* is required!

30. Foundation Aid without Minimum Tax Rate Requirement
State Aid
Foundation Aid without Minimum Tax Rate Requirement E*
Spending per Pupil
Tax Base per Pupil
Vmin
Vmax
Local Taxes
Foundation
Spending
Level
Spending Above Foundation Level
Aid
Lost
without
Required
Minimum
Tax Rate

31. State Aid
Power-Equalizing Aid
Power-equalizing (or guaranteed tax base, GTB) aid, is the main program in 3 states and a supplementary program in 10 others.
The idea behind GTB aid is that a district’s spending should depend on tax effort, not tax base:
where V* is a policy parameter set by the state.

32. State Aid
Power-Equalizing Aid, 2
To determine the associated aid formula, note that
Combining this with the GTB formula leads to:

33. State Aid
Power-Equalizing Aid, 3
GTB is matching aid; the state share of spending decreases with Vj.
If V* < maximum Vj, then matching rates are negative in rich districts.
This is called recapure.
Vermont is an example.
Raising V* raises the cost of GTB aid.
To offset this effect, set matching rates at a fraction of the value in the above formula
That is, flatten the line in the following picture.

34. Power-Equalizing Aid, 4 State Aid Tax Base per Pupil Matching Rate =
Matching
Rate =
State
Share
Tax Base per Pupil
Vmin
Vmax
Recapture V*
mi = (1 – Vi /V*)

35. State Aid
Power-Equalizing Aid, 5
GTB aid can be adjusted for costs (but rarely is):
Using GTB as a supplement to foundation aid is misguided:
Foundation aid already requires poor districts to set tax rates above their desired level.
Adding GTB aid will not induce any further tax rate increases.

36. Comparing Foundation & GTB Aid
State Aid
Comparing Foundation & GTB Aid
GTB aid is often thought to be more equalizing than foundation aid.
In fact, however, price elasticities are small, so the response to GTB aid is small, even for poor districts. (See D/Y, NTJ, 1998.)
So:
Foundation aid is much more equalizing at the bottom of the property value distribution (and is the only way to ensure adequacy).
GTB aid is more equalizing at the top of the property value distribution—at least if it includes recapture

37. Comparing Foundation and
State Aid
Comparing Foundation and
GTB Aid, 2 E*
Spending per Pupil
Tax Base per Pupil
Vmin
Vmax
Recapture
Initial Spending
Spending with Foundation Aid
Spending with Power-Equalizing Aid V*
Small Impacts Due to Small Price Elasticity
With Required
Min. Tax Rate

38. State Aid
Wealth Neutrality
Wealth neutrality is defined as a zero correlation between property wealth and performance outcomes.
This is the same as a flat regression line for performance as a function of wealth.
When it was first proposed, many people thought GTB aid would lead to wealth neutrality.
But a famous paper by Feldstein (AER 1975) shows that this would be true only by coincidence because outcomes depend on the behavioral responses to the matching rates.
A low response by low-wealth districts (which seems to be the case) would lead an upward sloping line.
A high response by low-wealth districts would lead to a downward sloping line.

39. Research on State Aid 1. Bradford/Oates Equivalence Theorem
2. The Flypaper Effect
3. Matching Aid
4. Aid and Efficiency

40. The Bradford/Oates Equivalence Theorem
State Aid
The Bradford/Oates Equivalence Theorem
Recall from the last class that augmented income is:
where A is lump-sum aid per pupil.
The B/O theorem says that one should not compare aid and income, but should instead compare tax-share weighted aid and income (AER 1971).
This approach accounts for the fact that the value of aid to a voter depends on the voter’s tax share.
If her tax share is 0.5, she only pays $0.50 for a $1.00 increase in property taxes; equivalently, $1 of aid only saves her $0.50.

41. This Set-Up Leads to Two Key Points
State Aid
This Set-Up Leads to Two Key Points
First, the expression for augmented income in the budget constraint (and hence in the demand function) is inherently nonlinear.
A specification with ln{Y } and ln{A} (let alone Y and A) makes no sense!
But a close approximation is available to make the term linear.

42. State Aid
The Approximation

43. This is the flypaper effect, f. The income term with f is:
State Aid
The Flypaper Effect
Second, even after the Bradford/Oates correction, income and aid do not have the same impact on education demand.
This is the flypaper effect, f.
The income term with f is:

44. Estimates of the Flypaper Effect
State Aid
Estimates of the Flypaper Effect
This specification is by no means universal.
Studies that use it or something equivalent usually find a value of f in the range.
Eom et al. (2014) claim to find much larger flypaper effects with this specification.
But Nguyen-Hoang and Yinger (2018a) show that Eom et al. make a scaling error in their aid term.
When this error is corrected, their approach leads to flypaper effects in the same range.
The estimates in N-H/Y (2019) are between 12 and 14.

45. Interpreting the Flypaper Effect
State Aid
Interpreting the Flypaper Effect
The value of f is the dollars of voter income that are needed to have the same impact on education spending (or performance in some models) as a dollar of state aid with the B/O weights.
Another approach is to calculate how much of each dollar of aid is spent and how much is turned into tax relief.
Nguyen-Hoang/Yinger find that $0.013 of every dollar of voter income is spent on education.
They also find that, on average, $0.15 of every $1 of state aid is spent on education; the rest goes to property tax relief.
The spending share is higher when voters’ tax share is higher.
The comparable average spending share for federal aid is $0.28.

46. Interpreting the Flypaper Effect, 2
State Aid
Interpreting the Flypaper Effect, 2
In a recent working paper, Nguyen-Hoang and Yinger raise two further issues about the flypaper effect.
A demand model based on student performance has two flypaper effects, which are similar in magnitude:
The impact of aid on the selected performance measure and
The impact of aid on efficiency, that is on the demand for school quality measures other than those in the performance measure.
The magnitude of the flypaper effect is affected by the salience and framing of the factors that influence the value of aid to voters (i.e., the B/O weights).
See: Nguyen-Hoang, Phuong, and John Yinger “The Flypaper Effect: Methods, Magnitudes, and Mechanisms.” Working Paper, Syracuse University, January.

47. With matching aid, the community budget constraint is
State Aid
Matching Aid
With matching aid, the community budget constraint is
So the household budget constraint
And tax price is

48. Matching Aid, 2 The matching rate:
State Aid
Matching Aid, 2
The matching rate:
Note that in this formulation, the matching rate is the state share
Some studies use the state match per dollar of local or
This leads to:

49. Matching vs. lump-sum aid
State Aid
Matching Aid, 3
Matching vs. lump-sum aid
There is a well-known theorem that matching aid has a more stimulative effect than equal-cost lump- sum aid.
This is an application of the classic microeconomic theorem that price subsidies have larger effects (per dollar) than cash grants.
The relevant graph follows:

50. Cash vs. Price Subsidy State Aid Clothing Food F1 F3 F2
Budget Line with
Cash Grant
Cost of Both Programs
(in Units of Food)
Tangency Point
with Price Subsidy I3 I2 I1
Budget Line with Price Subsidy
with Cash Grant

51. State Aid
Matching Aid, 4
This theorem fails to consider two key elements of the demand for public services
According to the Bradford/Oates equivalence theorem, matching aid alters the value of existing lump-sum aid (and, to be specific, makes it less stimulative).
Both matching and lump-sum aid may alter efficiency, but these effects are not well understood.

52. With matching and lump-sum aid the household budget constraint is
State Aid
Matching Aid, 5
With matching and lump-sum aid the household budget constraint is
Matching aid affects the value people place on lump-sum aid (because it alters their tax price),
This effect could reverse the standard theorem because it lowers the stimulative impact of matching aid.

53. State Aid
Aid and Efficiency
Both matching and lump-sum aid appear in augmented income, which affects efficiency.
Moreover, matching aid appears directly in tax price, which also affects efficiency.
Hence both type of aid may alter efficiency, but in different ways; these effects might reverse the standard theorem.

54. The STAR property tax exemption is like matching aid.
State Aid
Matching Aid, 6
The STAR property tax exemption is like matching aid.
With STAR, T = t(V – X) = tV(1 – X/V); the matching rate is X/V. Hence
That’s what we estimate in the STAR paper (Eom et al., EF&P 2014). The price elasticity on the STAR term is -0.6.

55. State Aid
Salience and STAR
Inspired by an experiment by Chetty et al. (AER 2009), many scholars not look for the impact of salience or visibility.
Large changes are more salient than small changes.
Highly publicized tax changes are more salient than others.

56. State Aid
Salience and STAR, 2
STAR (equivalent to matching aid) was highly publicized in its early years.
STAR was phased in, with three years of large changes, followed by many years of small adjustments.
So the STAR price elasticities should be largest at the beginning, which is what Nguyen-Hoang and Yinger find in a 2018 working paper on STAR price elasticities (2018b; a follow-up to Eom et al., 2014).

57. What Causes the Flypaper Effect?
State Aid
What Causes the Flypaper Effect?
Many hypotheses have been offered to explain the flypaper effect.
Hines and Thaler (JEP 1995) argue that it arises from the behavioral economics concept of framing.
Money that flows into a government’s budget is framed by households as support for public services, whereas money that flows into a household budget is framed as support for general consumption.
The working paper by Nguyen-Hoang and Yinger on the flypaper effect (2018a) finds empirical support for this hypothesis.

58. What Causes the Flypaper Effect? 2
State Aid
What Causes the Flypaper Effect? 2
STAR exemptions appear on a homeowner’s tax bill, so it is reasonable to think that they are framed as part of their “local public education” budget, with a primary impact on education demand decisions.
In , STAR exemptions were supplemented by STAR rebate checks which were 30% of STAR exemption savings in 2007 (with adjustments base on income in 2008 and 2009); these checks arrived in the mail.
It is reasonable to suppose that these rebate checks are unlabeled income, and hence appear in a household’s general budget, with minimal impact on the demand for school quality.

59. What Causes the Flypaper Effect? 3
State Aid
What Causes the Flypaper Effect? 3
Although this framing hypothesis is reasonable, the STAR rebates are formulated as an adjustment to the STAR tax share; in effect they increase the STAR exemption amount by 30% (with some variation in 2008 and 2009).
Another possibility, therefore, is that voters view the rebates as an adjustment to their tax share—not as an addition to unlabeled income.
The main tests in N-H/Y compare an econometric specification in which rebates are treated as part of the STAR tax share and a specification in which the tax savings from the rebates are treated as income.

60. What Causes the Flypaper Effect? 4
State Aid
What Causes the Flypaper Effect? 4
Let t be the school district’s tax rate and τ be the rebate percentage. Then in the rebate years:
Specification 1 (STAR tax share with rebates):
Specification 2 (Income plus STAR tax savings):

61. What Causes the Flypaper Effect? 4
State Aid
What Causes the Flypaper Effect? 4
N-H/Y conduct specification tests for these two specifications.
The specification consistent with framing (with the tax savings from rebates in income) dominates the other specification.
N-H/Y also have some additional evidence supporting framing; check out the paper if you are interested.

62. What Causes the Flypaper Effect? 6
State Aid
What Causes the Flypaper Effect? 6
The N-H/Y working paper also goes over many other theories about the causes of the flypaper effects.
In the case of state aid to education, the available evidence does not support any of these other theories.