Evaluating state funding for gifted education programs.

In a recent article in Roeper Review, Baker and Friedman-Nimz (2002a) proposed that supplemental funding from state governments should be the primary leverage point for policy advocacy in gifted education. Baker and Friedman-Nimz explained how pursuit of federal legal protection has been unsuccessful and why it is likely to remain unsuccessful for the foreseeable future. Further, they explained that "a critical problem with litigation-based strategies for achieving equity for special populations is that they ultimately create divisive competition among defined student populations for access to finite educational resources, creating an unhealthy and ultimately nonproductive systemic tension, diverting attention from the central issue--providing suitable (ability-appropriate) educational opportunities to all students" (p. 7). Ultimately, it is the role of state legislatures to establish what constitutes a "suitable" or "thorough and efficient" education for children within their borders and to ensure that local school districts have access to the necessary resources to provide such an education.

As a follow-up to the previous article, this article presents an overview and analysis of current state funding for gifted education programs. It begins with an overview of state school funding approaches for supporting programs for children with a variety of special educational needs. Next, the article offers some insights into the relative effectiveness of each method via ex ante policy analysis. The next section includes an overview of the funding methods presently (1998-99) used by states for providing supplemental aid for gifted education programs. Finally, data from the Financial Survey of Local Governments (F-33) of the U.S. Census are used to evaluate the level and distribution of aid to local districts for gifted education in fiscal year 2000 and changes in that aid from fiscal year 1996 through fiscal year 2000.

Overview of State School Funding Methods

Following are descriptions of methods used by states for providing aid for educational programs that serve special populations. Each description also includes some comments on the perceived advantages and disadvantages of the method. These mechanisms are most often discussed in the context of programs for students with disabilities. A substantial body of literature on these approaches is available through the Center for Special Education Finance of the American Institutes for Research (http://csef.air.org/). Recent research has begun to explore, in greater detail, state funding for serving at-risk, limited English proficient, and gifted and talented students (Baker, 2001a; Baker & Friedman-Nimz, in press-a, in press-b; Baker & Markham, 2002). Funding methods generally fall into one of the following five categories:

1. Weighted funding (pupil weights). Under a weighted funding system, state supplemental aid is allocated on a per student basis where the amount of aid is based on the funding weight associated with each student. For example, a gifted child might be weighted 1.12 (Texas), such that a district receives 12% additional aid (12% x Base Aid Allocation per Pupil) per each gifted child identified. The perceived advantages of weighted aid include the ability to accommodate cost differences by student need (e.g., districts with more students in need may access more funding) and the ability to attach the pupil weights directly to general fund allocations (weights may be added directly to district general funds, giving district administrators flexibility in how funds are used).

Where children with disabilities are involved, the primary disadvantage of pupil weights is that they tend to stimulate over-identification of subjectively based disabilities (learning disabled and behavioral disorders) and increased restrictive placements (where weights are tied to placement). This limitation is less relevant in gifted education where pupil weight programs often include identification limits. (1) One "problem" or "difficulty" of using weights and identification limits for gifted education is that smaller districts may have too few students to generate useful levels of aid (Baker, 2001b, 2001c). Another is that when supplemental funds are added to general funds for "flexible" use, they may or may not end up benefiting the intended recipients (e.g., there is no guarantee that funds brought in through pupil weights will actually be spent on gifted programs, unless specified in state accounting requirements). An advantage of pupil weights over some other options is that when base aid allocations are increased, weighted aid is necessarily increased unless the legislature consciously chooses to decrease the weight.

2. Flat-grant funding. Flat-grant funding is based on a fixed funding amount per student, and may be allocated either per identified student, or per total student population, assuming fixed portions of special needs students across districts. In the latter case, flat-grant funding is sometimes referred to as census-based funding. Where flat grants per identified pupil are capped at a specific percentage, they are the equivalent of a census-based grant. For example, the New York flat grant of $196 per 3% pupils is the equivalent of $5.88 per total pupils. The Wyoming legislature recently shifted from a $150 per 3% pupil grant to a $9 per total pupil grant, adding funding to the program and changing the units.

A perceived disadvantage of census-based approaches is that they fail to accommodate differences in need across districts (higher rates of giftedness in some districts than others). Flat grants, like pupil weights, also fail to provide sufficient aid to districts lacking critical mass. Finally, unlike pupil weights, a legislature can neglect to increase the level of a flat grant on an annual basis, even in years when other aid is increased. An advantage of a census-based approach is that it provides program flexibility. Schools may implement a school wide enrichment program, may opt to provide more costly intervention for highly gifted students, or use any other model allowable under state policy.

3. Resource-based funding. Resource-based funding is based on an allocation of specific education resources, usually teaching staff, but sometimes classroom units (the costs estimated to provide for the total resources of one class, including teacher, instructional materials, supplies and equipment). One perceived advantage of this approach in special education is that it can accommodate varied resource needs across districts. In gifted education, however, resources (personnel or classroom units) are typically allocated on a census basis, like Virginia's funding program that allocates one gifted specialist per one thousand enrolled pupils. A potential advantage of personnel-based funding in gifted education is that the approach does not necessarily dictate or even imply the numbers or percentages of children that should be served by the personnel. A district may choose to use their personnel either to coordinate school-wide enrichment, or to coordinate diagnostic testing and programmed instruction. A perceived disadvantage, however, is that aid allocated only for personnel reimbursement ignores resource costs beyond personnel (materials, supplies, equipment, transportation, etc.).

4. Percentage reimbursement funding. Under a percentage reimbursement system, the amount of state supplemental aid a district receives is directly based on its prior year expenditures for the program. States generally reimburse only the percent of expenses the state can afford in a given year, and states generally set guidelines as to which expenses are reimbursable. One perceived advantage of this approach is that reimbursements may accommodate significant cost differences across districts. A perceived disadvantage is the compliance burden of accounting for allowable costs, potentially stimulating inefficient segregation of resources within districts.

Prior to 1992, Connecticut used a percentage reimbursement program that originally reimbursed gifted program expenses on a 30 to 70% sliding scale, which was subsequently reduced to a 15 to 30% sliding scale, based on district wealth. A possible danger is that a state legislature might choose to retroactively cut funding and not reimburse expenses already incurred. Although this did happen to Connecticut school districts in the early 1990s, this is politically difficult.

Recent research by Baker (in press) indicates that districts in states using percentage reimbursement programs for special education generally have more funding available to use on their core instructional programs. That is, special education is less of a drain on general education funds in states that use percentage reimbursement.

5. Discretionary grants. Some states use discretionary grants to fund gifted education or grants rewarded on an application basis. Discretionary grants are rarely used, and generally considered unacceptable as the primary basis for funding programs for students with special needs. Discretionary grants are usually available above and beyond a basic set of resources for purposes of creating and evaluating innovative programs. A disadvantage of this approach is that in order to receive grants, districts must have sufficient staff to prepare a proposal and enough local resources to provide a basic program that can be improved with an infusion of state aid. Further, because discretionary grants tend to be relatively small, districts with greater local capacity to supplement the grants are most likely to apply.

Ex Ante Evaluation Framework

Ex ante policy analysis involves the analysis of written policies to discern their meaning, logic, and expected outcomes. This section outlines the relative effectiveness of alternative approaches to funding gifted education programs for achieving goals of equity and adequacy, based on the limited body of existing empirical research on this topic. This section also addresses desirable and undesirable local behaviors stimulated by alternative state funding mechanisms (Baker, 2003).

Equity is defined for purposes herein as the comparable availability of opportunities for gifted children to access and participate in gifted and talented programs. This definition includes the concept of neutrality discussed by Baker and Friedman-Nimz (2002b) in their recent research on the determinants of the availability of opportunities for gifted children. Baker and Friedman-Nimz argue that the availability of educational opportunities should be statistically unassociated with student population characteristics that are not directly associated with educational needs. For example, educational opportunities should not be statistically associated with race, ethnicity, or socioeconomic status unless it can be shown that specific educational needs are associated with race, ethnicity, or socioeconomic status.

Adequacy is defined by the authors for purposes herein as whether the aid program in question is likely to provide local school districts the ability to cover the costs of operating gifted and talented programs. In this case, costs refers to the costs of providing the necessary supplemental resources to fully accommodate a child's special educational needs. Each child in a school district with a special need also typically brings general education revenue to the district. As a result, it is the role of the supplemental aid to cover the difference between the additional resource costs and the general education revenue for that child. That difference is the "marginal cost" (sometimes referred to in policy circles as "excess cost"), or the cost of meeting the child's needs, that is above and beyond basic education costs (assuming that the general education revenue meets basic education costs).

Table 1 provides a summary of expected equity and adequacy outcomes of alternative aid allocation methods. Each of the previously discussed mechanisms may be implemented with the intent to meet the full costs of operating programs (at cost) or with the intent to supplement locally available resources (below cost). The effectiveness at achieving equity or adequacy goals varies across mechanisms when aid is allocated at less than full cost. For example, a flat grant allocated at less than full operating costs per gifted child often provides districts with greater capacity to supplement already existing programs, while failing to provide sufficient resources to districts with limited local capacity (Baker, 2001b; Curley, 1991). The effectiveness of each mechanism also interacts with other policies, for example, a mandate for serving students with girls and talents (Baker & Friedman-Nimz, in press-a).

Whether or not supplemental funding provides districts with comparable or fully equalized abilities to purchase services for their students also depends on how that supplemental funding is integrated with general funding. Table 2 provides some guidelines for the integration of alternative supplemental funding programs under differing conditions of general funding. Although there are a variety of general funding formula types, there are really only two relevant scenarios to this discussion--either general funds are equalized or they are not.

Revenues and expenditures of school districts may be highly equalized, falling in a very narrow range due to a combination of state aid to achieve basic funding levels and strict limitations on local spending. Few, if any states fall into this category, and such models, while seemingly "ideal" from an equity perspective, can be more problematic than less equitable general funding solutions. Some economists argue that state funding focused too narrowly around equity objectives can lead to "equitable inadequacy" over time as state legislators fail to provide sufficient basic funding (Campbell & Fischel, 1996; Downes & Figlio, 1997). In most states, revenues and expenditures of school districts are partially, or even substantially equalized by raising the revenues of low capacity districts up to a specific level (foundation level), and then allowing districts to raise additional revenues either with local taxes alone or with some matching aid from the state. Under such circumstances, supplemental aid must be equalized by local districts' ability to supplement their foundation aid.

Review of State Policies 1998-99

This section provides a brief summary of current state policies for funding gifted education programs (see Table 3). The majority of data for this section were drawn from the National Center for Education Statistics (Seilke, Dayton, Holmes & Jefferson, 2001) Public School Finance Programs in the United States and Canada 1998-99 (www.nces.ed.gov/edfin). Some changes were made to the representation of state policies in response to a survey administered in 2002 by the State Legislative Policy Task Force of the National Association for Gifted Children.

In 1998-99, eleven states used discretionary grant programs as the sole basis for funding gifted education, thirteen states used flat grants, eight states used pupil weights, seven states used resource-based funding, and one used a percentage reimbursement program. The most common use for discretionary grants was to promote new or innovative programs or staff development. Discretionary grants are rarely intended to support ongoing operations. Missouri's "discretionary grant" program, which reimburses a percentage of approved costs, is more consistent with a "percentage reimbursement" classification.

Pupil weights ranged from. 12 (Texas) to .64 (Georgia) and higher (Florida). Flat grants ranged from $55 (Arizona) to $320 (Washington) per eligible pupil. Both pupil weight and flat grant programs were likely to be implemented with limitations on numbers of pupils eligible for funding, ranging from 2% (Washington) to 5% (Texas). Colorado took a unique approach to flat grant implementation to address the problem of insufficient critical mass in small school districts. In Colorado, small districts were guaranteed a minimum flat grant of $6,500.

Resource-based funding programs were used primarily to help districts meet the costs of staffing gifted programs. Virginia's formula, which provides funding for one specialist per 1,000 pupils (all enrolled pupils), is a common approach to resource-based funding. The use of a per pupil ratio or "caseload" serves to constrain state aid allocations much the same way eligible pupil limits in flat grant and pupil weight programs do, but avoids the implication that only X% of pupils can be classified as gifted. Tennessee's resource-based model, which more broadly addresses allowable costs, is more typical of percentage reimbursement programs.

Ex Post Evaluation Framework

Ex post evaluation involves the analysis of data to determine actual conditions or outcomes of policies. Baker (2001 a) developed a framework for evaluating the equity, adequacy, and rationality of state aid allocations and applied that framework to aid allocations for limited English proficient, at risk, and gifted education programs using national data from the 1995-96 school year. Adequacy, rationality, and equity are measured as follows:

Adequate: Aid allocation per expected need pupil as a percent of core expenditures exceeds minimum reported, though not necessarily empirically cost based, adequacy weight from existing literature.

Rational: Aid allocation per pupil and total allocation significantly correlated (p<.05) with expected prevalence.

Equitable: Aid allocation per pupil significantly correlated in the expected direction (p<.05) with two of three context measures (median family income, core expenditures per pupil, state revenue share).

Adequacy is difficult to measure for gifted education programs for two reasons. First, there is limited existing information on the "marginal" costs of adequate services for gifted children. Baker and Friedman-Nimz (in press-b) provide ballpark estimates of 30 to 60% above regular education expenditures per pupil, based primarily on costs of providing qualified personnel. (2) Second, precise adequacy measurements across districts require that we know how many children require special services (actual prevalence of giftedness). Program participation rates or local gifted identification rates are a poor proxy for actual prevalence because identification and participation rates are influenced by the availability of resources (Baker, 2003). That is, where either more local resources or more state funding are available, districts tend to identify and serve more children in gifted programs. As a result, the best one can do is to assume uniform prevalence across districts. Rationality of gifted education aid, the relationship between aid levels and prevalence rates, also cannot be measured because actual prevalence of giftedness cannot be known.

Only the equity of gifted education aid may be measured effectively. In any state where general education resources vary, and especially in states where those resources vary with respect to local fiscal capacity, state aid allocations for gifted education should be higher in lower wealth (property wealth and income) districts and in districts receiving higher overall percentages of state aid (usually the same districts). It may also be appropriate that supplemental aid be higher in districts with generally lower core instructional expenditures per pupil, except where core expenditures are lower as a function of local choice rather than local capacity.

Ex Post Evaluation of State Funding 2000

For this section, data from the Annual Financial Survey of Local Governments administered by the U.S. Census Bureau were used. Data were compiled for fiscal years 1996 through 2000. Financial data were merged with data on district characteristics, including numbers of children in poverty (1997 estimates) from the Common Core of Data of the National Center for Education Statistics. The F-33 survey includes district level data on local, state, and federal revenues received and on expenditures on instructional and noninstructional functions. Analyses herein focus specifically on revenues received by local districts from state governments for purposes of supporting gifted education programs (variable C08). Fiscal year 2000 data are analyzed in detail. The complete panel is used only to evaluate relative increases or decreases in local revenues for gifted education by state from 1996 through 2000.

Table 4 summarizes gifted education revenues of districts in the 19 states reporting those revenues in fiscal year 2000. Table 4 includes funding levels, 5-year changes in funding levels, and correlations with selected district indicators. All measures are weighted by district enrollment. For comparison purposes, state revenues are expressed per 5% enrollment. In other words, if 5% of children in a district or state were gifted, how much funding was allocated per each of those children? State revenues per 5% enrolled pupils ranged from approximately $1 (allocated through Minnesota's since-defunct discretionary grant) to an astounding $2,738 in Florida. Other respectably high funding states included North and South Carolina with $650 and $850 per pupil.

Gifted education revenues per 5% enrolled pupils are also expressed with respect to mean current expenditures per pupil to reflect the marginal value of gifted education revenues, assuming that average current expenditures represent the average cost of a general education. Again, Florida is highest at 47% and only North and South Carolina in addition to Florida exceed 10%. Why are these percentages much lower than pupil weights expressed in finance policies? Pupil weights are typically multiplied by a "base state aid" value, or value that represents a minimum bar of general education funding to be achieved by all districts in the state. In this case, actual gifted education revenues are expressed with respect to average per pupil expenditures, which in most states significantly exceed basic aid levels. For example, in 2000-2001 base state aid in Kansas was $3,820 per weighted pupil, and the state used a weight of .10 for at-risk children, yielding $382 per at-risk child. Mean per pupil expenditures in Kansas, in that year, were approximately $5,500 per pupil. The $382 per at-risk pupil turns out to be only about 7% of the $5,500.

Table 4 also shows that in most states, from 1996 through 2000, gifted education aid allocations increased, and those increases were statistically significant. Such increases are expected during a time of rapid economic growth. One might expect a downturn in gifted education revenues for the past two years given the dramatic change in economic climate. In four states (Arkansas, Arizona, Illinois, and Montana), districts received less aid for gifted programs from 1996 through 2000. Increases or decreases shown in the table are the average slope or trend for the period.

The next section of Table 4 addresses whether state aid was allocated equitably, according to Baker's (2001a) ex post evaluation framework for supplemental state aid. Recall that in general, an equitable allocation of state aid would involve higher levels of aid being received by districts with lower current expenditures per pupil and/or higher percentages of revenue from state sources. Regarding the current expenditures correlation, the assumption is that districts with higher current expenditures per pupil have greater ability to supplement gifted education aid. As a result, they should receive less supplemental aid for gifted education. An exception to this rule is the case where a district has relatively high expenditures per pupil not because of local capacity to raise more revenues, but because of higher costs due to higher concentrations of students with special needs. Low capacity, high need districts typically receive the majority of their funding from state sources. In general, low fiscal capacity districts are more dependent on state aid. As such, districts receiving larger portions of the revenues from state sources should generally receive higher allocations of supplemental aid.

An additional exception occurs where general education funding variance is negligible. In this case, it is reasonable for gifted education aid to be allocated in flat amounts across districts, or at least in patterns unassociated with variations in current expenditures, which do not exist, and unassociated with general state aid dependence. States qualifying for this "General Equity Waiver" include Florida, Hawaii, and Washington. (3)

Finally, correlations between percentages of children in poverty and gifted education aid per 5% enrolled pupils are also included. In general, one might assume that higher concentrations of children in poverty exist in districts with lower fiscal capacity and districts that are more dependent on state aid. While this is partly the case, this additional correlation does raise other questions about the extent to which state aid programs support gifted program access for children in low-income communities. With respect to poverty, an equitable distribution is a neutral distribution, one with no relationship between gifted education aid and poverty concentrations, except where poverty concentrations are associated with district capacity in the expected direction. In that case, a desirable distribution is a positive correlation between poverty concentration and gifted education funding. A negative correlation between gifted education aid and poverty concentration is generally unacceptable.

Only one state, Virginia, strongly equalizes gifted education funding with respect to both current expenditures per pupil and state aid dependence. This finding is consistent with findings of Baker (2001a) and Baker and Friedman-Nimz (in press-a), using data from previous years (1993-94 and 1995-96). As one might expect, gifted education aid in Virginia is also allocated in positive correlation with poverty concentration. Florida also presents a generally positive equity profile, allocating more aid to higher poverty districts and to districts more dependent on state aid.

A handful of states possess questionable, negative combinations of correlations. For example, Arizona, California, and Nebraska possess combinations of positive correlations between gifted education aid and current expenditures per pupil and negative correlations between gifted education aid and state aid dependence, implying that higher capacity districts are receiving more aid. Most of these correlations, while significant, are relatively small. Illinois and South Carolina possess relatively strong (in the context of this analysis) negative correlations between gifted education aid and poverty concentration, along with negative correlations between gifted education aid and state aid dependence. The finding for South Carolina, like the finding for Virginia, is consistent with previous analyses performed by Baker and Friedman-Nimz (in press-a, in press-b).

Conclusions and Policy Recommendations

This article presents both ex ante and ex post frameworks for evaluating supplemental state aid for gifted education programs. Further, this article applies those frameworks for evaluating state school finance policies as of 1998-99 and state aid allocated to local districts for gifted education in 2000. Only Florida provided both sufficient and equitable support for gifted education, assuming general education conditions to be adequate as well as equitable in that state (a big assumption). Gifted education funding in Virginia, while less adequate than supplemental funding in Florida, continues to serve as a model of equitable distribution, especially for states where general education funding remains inequitable. Further, Virginia's personnel-based approach may provide districts with greater flexibility regarding program implementation than identified pupil-based approaches. Personnel-based allocations may also ensure that even small school districts have sufficient resources to hire qualified personnel, arguably the most important and most expensive resource in delivering quality services to gifted children.

In light of the findings herein, and with the hope of continued interest in and support for supplemental state funding for gifted education, we propose the following set of policy recommendations:

* Determine, based on state school finance policies and their interaction with other policies, how to calculate equity and adequacy.

* Conduct an annual evaluation and reporting of state school finance policies, to include measures of equity and adequacy of resources for gifted education.

* Conduct frequent, if not annual, evaluations of the influence of state school finance policies on the distribution and quality of opportunities.

* Develop "model policies" and frameworks based on current and evolving understanding of the effectiveness of existing policies for promoting equitable access to high quality opportunities.

On the first point, we propose that state officials and academic researchers collaborate on the annual collection and analysis of data on state aid allocations and state regulatory policies for gifted education. Data included in the U.S. Census Bureau's annual fiscal survey of local governments may also prove valuable, but may not be precise enough for disaggregating revenues allocated for gifted education under state special education policies or identifying other sources of funds which may positively affect gifted children (targeted support for AP courses, etc.). Further, data should include not only measures of revenues received by local districts from states but also local district expenditures on gifted education and numbers of pupils served.

On the second point, it is important that researchers, policymakers, and advocates continue to follow up on whether state aid allocations are meeting their objectives of improving the availability and quality of opportunities for gifted children. Some early findings using national data sets suggest that some states' aid policies have stimulated more disparities than they have resolved (Baker & Friedman-Nimz, in press-a, in press-b; Curley, 1991). That is, the presence of state aid has been shown to increase differences between percentages of children served in gifted programs in high and low income districts. Such outcomes should be considered unacceptable.

Finally, each of these types of ongoing analyses should be used to directly inform policy development. Annual evaluation of state policies and periodic analysis of the influence of various policy types may be used to establish policy benchmarks and measure state progress toward those benchmarks. For the time being, progress toward equity benchmarks is most easily measured, as there continues to be a lack of sufficient information regarding what constitutes adequate funding for gifted education. One possible resolution to this problem is to develop a set of indicators for evaluating the quality as well as the distribution of services for gifted children and to compare quality indicators in states operating under different models and with different levels of financial support for gifted education.

Table 1

Evaluation Framework for Alternative Funding Mechanisms

Funding          Funding
Approach         Level      Equity

Pupil Weight     Below      Only increases opportunity for districts
                 cost       that already have sufficient general
                            education resources.
                 At         May help some low resource districts, but
                 cost       generally stimulates program expansion in
                            high resource districts. Same amount of
                            state aid may be used to create greater
                            equity than flat grant if weights integra-
                            ted into general fund formula.
Flat Grant       Below      Only increases opportunity for districts
                 cost       that already have sufficient resources.
                 At         May help some low resource districts, but
                 cost       generally stimulates greater program expan-
                            sion in high resource districts.
Resource         Below      Generally provides opportunity for dis-
Based            cost       tricts that already have sufficient finan-
                            cial resources to supplement aid, especial-
                            ly where aid allocations are flat across
                            districts (e.g. $14,000 toward a specia-
                            list's salary).
                 At         Highly equitable if allocated to districts
                 cost       on sliding scale basis according to local
                            capacity. Still depends on equity of
                            general education resources.
Percentage       Below      May still improve equity if allocated on
Reimbursement    cost       steep sliding scaleaccording to local
                            resource availability.
                 At         Highly equitable if allocated to districts
                 cost       on sliding scale basis according to local
                            capacity. Still depends on equity of gene-
                            ral education resources.
Discretionary    Below      Inequitable.
Grant            cost
                 At         Still generally inequitable because disc-
                 cost       retionary grants typically rely on the
                            motivation of local district administrators
                            (which may reflect local parental pressure)
                            to pursue the funding to provide a program.

Funding          Funding
Approach         Level      Adequacy               Local Behaviors

Pupil Weight     Below      Inadequate.            Some districts with
                 cost                              few local resources
                                                   may opt out.
                 At         Only adequate if       Stimulates identifi-
                 cost       general education      cation if allocated
                            resources are uni-     on per pupil basis.
                            formly adequate.       Usually capped.
Flat Grant       Below      Inadequate.            Some districts with
                 cost                              few local resources
                                                   may opt out.
                 At         Only adequate if       Stimulates identifi-
                 cost       general education      cation if allocated
                            resources are uni-     on per pupil basis.
                            formly adequate.       Usually capped.
Resource         Below      Inadequate.            Some districts with
Based            cost                              few local resources
                                                   may opt out.
                 At         Adequate if re-        Districts may tend
                 cost       sources include all    to over-segregate
                            necessary resources    resources (restric-
                            for program opera-     tive placements,
                            tion, not just         etc.) in order to
                            personnel (though      maintain clear
                            that's a good star-    accounting records
                            ting point).           of which resources
                                                   are used for gifted
                                                   education. May con-
                                                   flict with school-
                                                   wide approaches. May
                                                   also increase
                                                   compliance burden.
Percentage       Below      Generally              Some districts with
Reimbursement    cost       inadequate.            few local resources
                                                   may opt out.
                 At         Adequate. Research     Districts may tend
                 cost       indicates that when    to over-segregate
                            states use this        resources (restric-
                            approach for           tive placements,
                            special education,     etc.) in order to
                            fewer funds are        maintain clear ac-
                            drawn away from        counting records of
                            general education.     reimbursable expen-
                                                   ses. May conflict
                                                   with schoolwide
                                                   approaches. May also
                                                   increase "com-
                                                   pliance" burden.
Discretionary    Below      Inadequate.            May stimulate some
Grant            cost                              districts sitting on
                                                   the fence. Most will
                                                   opt out.
                 At         May be adequate for    Relies on district
                 cost       those who choose to    self selection to
                            pursue the funding.    participate. May
                                                   stimulate some dis-
                                                   tricts sitting on
                                                   the fence, but un-
                                                   likely to do more
                                                   than that, even if
                                                   funding is allocated
                                                   at program costs.

Table 2

Integration of General and Supplemental Funding

 General Funding
    Situation              Pupil Weight               Flat Grant

All districts have    A standard weight may     A uniform flat grant
same general ed.      be provided for all       may be provided for all
revenue per pupil.    districts that repre-     districts that repre-
                      sents the "marginal"      sents  the "marginal"
                      cost above the uniform    cost above the uniform
                      general ed. revenue.      general ed. revenue.

State supplements     Pupil weights are         A flat grant could
all districts to a    typically applied to      conceivably be alloca-
given level (Foun-    the foundation level      ted on a sliding scale
dation Aid), but      (first tier) in these     according to a dis-
allows spending       programs, but dis-        trict's fiscal capacity
above that level.     tricts have varied        to supplement that aid
                      capacity to raise         (then it would not be
                      revenue above that        "flat" however).
                      level.

                      Pupil weights can         In general, flat grants
                      never be fully            face the same problem
                      equalized under these     as pupil weights with
                      circumstances.            respect to equaliza-
                                                tion.

 General Funding                                      Percentage
    Situation             Resource Based             Reimbursement

All districts have    State may purchase        State may fully reim-
same general ed.      (full cost) uniform       burse uniform set of
revenue per pupil.    set of resources for      allowable costs for all
                      all districts.            districts.

State supplements     States may use same       State may use same
all districts to a    sliding scale, based      sliding scale, based
given level (Foun-    on same fiscal capa-      on same fiscal capacity
dation Aid), but      city measures (if they    measures (if they are
allows spending       are good ones) as used    good ones) as used for
above that level.     for general aid sup-      general aid supplement
                      plement to assist         for reimbursing dis-
                      districts in              trict expenses.
                      purchasing resources.
                                                For example, if a dis-
                                                trict is 70% dependent
                                                on state aid to achieve
                                                the foundation level of
                                                general education spen-
                                                ding, the state may
                                                reimburse 70% of their
                                                supplemental expenses.

Table 3

State Funding Policies for Gifted and Talented Students

       State             Formula Type (a)        Allocation Basis (a)

       Alaska               Pupil Weight            Eligible Pupil
      Arizona                Flat Grant          $55 per 3% pupils or
                                                $1000 (greater of two)
      Arkansas              Pupil Weight             Up to 5% ADA
     California        Flat Grant (varied by        Eligible Pupil
                       pupil count and fiscal
                             capacity)
      Colorado               Flat Grant          $6,500 per district +
                                                   Flat grant per 7%
                                                enrolled pupils + other
                                                special purpose grants
District of Columbia        Pupil Weight
      Florida               Pupil Weight            Eligible pupil
      Georgia               Pupil Weight            1.64 or 64% per
                                                    eligible pupil
       Hawaii              Resource Based
       Idaho                 Flat Grant               Enrollment
      Illinois          Flat Grant/Resource       Either grant per 5%
                                                 FTE pupils or $5,000
                                                      per teacher
      Indiana           Discretionary Grant       Support for program
                                                       planning,
                                                   implementation or
                                                     continuation
       Kansas              Resource Based           Number of staff
      Kentucky             Resource Based            Teachers and
                                                     Coordinators
     Louisiana              Pupil Weight         .6 x eligible pupils
       Maine                 Percentage             Approved Costs
                           Reimbursement
      Maryland             Discretionary          Program improvement
                             Grant (1)                   plans
   Massachusetts        Discretionary Grant
      Michigan             Resource Based       Teachers in districts;
                                                  Summer institutes;
                                                     Comprehensive
                                                    service centers
     Minnesota             Discretionary
                             Grant (2)
    Mississippi              Flat Grant
                       (general fund add-on)
      Missouri          Discretionary Grant        Reimbursement of
                                                approved program costs
      Montana           Discretionary Grant       Support for teacher
                                                training, innovation &
                                                 program continuation
      Nebraska          Discretionary Grants       Approved programs
     New Mexico              Flat Grant
      New York               Flat Grant             $196 per 3% ADA
       Nevada           Discretionary Grant
   North Carolina            Flat Grant               For 4% ADM
    North Dakota        Discretionary Grant
        Ohio               Resource Based        Salary allowance and
                                                    per pupil cost
                                                       allowance
                                                   (Classroom Unit)
      Oklahoma              Pupil Weight        .34 per eligible pupil
       Oregon           Discretionary Grant
   South Carolina            Flat Grant         per eligible pupil for
                                                     a) academic,
                                                    b) artistic and
                                                 c) advanced placement
                                                       programs
    South Dakota              Repealed
     Tennessee             Resource Based         Allowable costs of
                                                service option selected
       Texas                Pupil Weight            .12 per 5% ADA
        Utah                 Flat Grant             Eligible Pupil
      Virginia             Resource Based          1 instructional
                                                  positions per 1000
                                                1000 eligible students
                                                     (cost share)
     Washington              Flat Grant             $320 per 2% FTE
   West Virginia        Discretionary Grant
      Wyoming                Flat Grant

(1) Phased out to be integrated into district general funds in
future years.

(2) No longer available

Note: ADA=Average Daily Attendance; ADM=Average Daily Membership;
FTE=Full Time Equivalent Pupils

(a) See Seilke, Dayton, Holmes & Jefferson (2001).

Table 4

Summary of State Gifted Education Revenues to Local Districts FY 2000

State                Districts           State Revenue
                                         per 5% Pupils

                                      Mean          SD

Arkansas                 310            $ 7          $ 49
Arizona                  215           $ 29          $ 16
California               982          $ 193         $ 135
Colorado                 176          $ 148          $ 92
Florida                   67        $ 2,738       $ 1,262
Hawaii                     1          $ 450
Illinois                 894          $ 154         $ 104
Indiana                  292          $ 127         $ 134
Maryland                  24          $ 118         $ 125
Michigan                 719           $ 41          $ 31
Minnesota                343            $ 0           $ 1
Montana                  455           $ 15          $ 35
North Carolina           117          $ 650          $ 71
Nebraska                 570          $ 213         $ 138
Ohio                     611            $ 3          $ 48
South Carolina            86          $ 850         $ 290
Utah                      40          $ 302         $ 131
Virginia                 132          $ 488         $ 224
Washington               296          $ 122          $ 15

State                Gain/Loss      Percent        Percent
                      per 5%       Above Mean      Poverty
                     1996-2000      CUREXPP

Arkansas              -7.59 *         0.1%         -0.02
Arizona               -0.96 *         0.6%          0.07
California            14.05 *         3.2%         -0.05
Colorado              45.05 *         2.4%          0.15 *
Florida              576.29 *        47.9%          0.28 *
Hawaii                17.79 *         6.9%
Illinois              -5.04 *         2.3%         -0.53 *
Indiana                5.29 *         1.9%         -0.10
Maryland              21.89 *         1.5%          0.13
Michigan               2.49 *         0.6%          0.32 *
Minnesota              3.56 *         0.0%         -0.04
Montana               -1.06           0.3%         -0.01
North Carolina       133.99 *        10.8%          0.02
Nebraska              62.29 *         3.4%         -0.02
Ohio                   0.73 *         0.1%         -0.05
South Carolina        34.25 *        14.1%         -0.51 *
Utah                  32.85 *         7.0%          0.06
Virginia              32.71 *         7.5%          0.42 *
Washington            11.08 *         2.0%         -0.16 *

                         Correlations

State               Current       Percent        General
                    Expend       of Revenue       Equity
                   per Pupil     from State     Waiver (a)

Arkansas             0.02          -0.01
Arizona              0.21 *        -0.14 *
California           0.11 *        -0.15 *
Colorado             0.35 *         0.08
Florida              0.21           0.27 *          Y
Hawaii                                              Y
Illinois            -0.04          -0.10 *
Indiana             -0.22 *         0.15 *
Maryland             0.81 *        -0.34
Michigan             0.31 *         0.07
Minnesota            0.00           0.02
Montana             -0.05           0.06
North Carolina       0.11          -0.07
Nebraska             0.28 *        -0.14 *
Ohio                -0.04           0.02
South Carolina       0.11          -0.25 *
Utah                 0.12           0.21
Virginia            -0.40 *         0.61 *
Washington          -0.29 *        -0.02            Y

Note: Data Source: U.S. Census Bureau Fiscal Survey of Local
Governments 2000 (F-33)

(a) States with coefficient of variation for general instructional
expenditures per pupil approximately 10% or less

* p<.05

(1) Alternatively, states control expenditures by enacting stringent eligibility guidelines.

(2) Estimates were based on costs of providing one FTE teacher per elementary school of 400 pupils, assuming caseload of primary beneficiaries of approximately 20. Prices of qualified teachers were estimated from salaries of teachers with a Master's degree in gifted education as reported in the 1999 Schools and Staffing Survey. Marginal costs were estimated with respect to mean current expenditures per pupil.

(3) Note that an oversimplified method for judging equity is applied in this article. Ideally, a "relative adequacy" standard should be used, whereby existing revenues are adjusted for various cost-related needs of school districts (economies of scale, varied regional labor costs, special student needs, etc.). That is, in a perfectly equitable funding system, revenues should vary across districts, but that variance should be directly associated with differences in costs. More thorough analyses and discussions of the integration of supplemental aid and general aid are provided in Baker and Friedman-Nimz (in press-a) and Baker and Friedman-Nimz (in press-b).

REFERENCES

Baker, B. D. (2001a). Living on the edges of school funding policies: The plight of at-risk, limited English proficient and gifted children. Educational Policy, 15(5), 699-723.

Baker, B. D. (2001b). Gifted children in the current policy and fiscal context of public education: A national snapshot and state level equity analysis of Texas. Educational Evaluation and Policy Analysis, 23(3), 229-250.

Baker, B. D. (2001c). Measuring the outcomes of slate policies for gifted education: An equity analysis of Texas school districts. Gifted Child Quarterly, 45(1), 4-15.

Baker, B. D. (2003). An empirical test of bureaucratic responses to state categorical aid: The case of services for limited English proficient children. Manuscript submitted for publication.

Baker, B. D. (in press). State policy influences on the internal allocation of school district resources: Evidence from the common core of data. Journal of Education Finance.

Baker, B. D., & Friedman-Nimz, R. C. (2002a). Is a federal mandate the answer? If so, what was the question? Roeper Review, 25(1), 5-10.

Baker, B. D., & Friedman-Nimz, R. C. (2002b) Determinants of the availability of opportunities for gifted children: Evidence from NELS '88. Leadership and Policy in Schools, 1(1), 52-71.

Baker, B. D., & Friedman-Nimz, R.C. (in press-a). State policies and equal opportunity: The example of gifted education. Educational Evaluation and Policy Analysis.

Baker, B. D., & Friedman-Nimz, R. C. (in press-b). Gifted children, vertical equity and state school finance policies. Journal of Education Finance.

Baker, B. D., & Markham, P. (2002). State school funding policies and limited English proficient children. Bilingual Research Journal, 26(3), 659-680.

Campbell, C. D., & Fischel, W. A. (1996). Preferences for school finance systems: Voters versus .judges. National Tax Journal, 49(1), 1057-1073.

Curley, J. (1991). Financing programs for education of the gifted and talented in New York state. Journal of Education Finance, 16(3), 332-347.

Downes, T. A., & Figlio, D. N. (1997). School finance reforms, tax limits, and student performance: Do reforms level up or dumb down? Unpublished Manuscript, Tuffs University, Medford, MA.

Seilke, C. C., Dayton, D., Holmes, C. T., & Jefferson, A. L. (2001). Public School Finance Programs of the United States and Canada. National Center for Education Statistics. Retrieved October 1, 2002 from www.nces.ed.gov/edfin

Authors' Note

This article does not reflect positions of either the Texas Association for the Gifted and Talented or the Council for Exceptional Children.

Manuscript submitted: October 21, 2002

Revision accepted: January 27, 2003

Bruce Baker is an associate professor in the Department of Teaching and Leadership at the University of Kansas, specializing in school finance policy and the economics of education. His recent research has been published in the Journal of Education Finance, Educational Policy, Economics of Education Review and Educational Evaluation and Policy Analysis. He regularly consults with state legislatures on issues involving school funding policies and has worked as an expert witness in finance equity litigation in state and federal court. E-mail: [e-mail].

Jay McIntire is executive director of the Texas Association for the Gifted and Talented in Austin. Jay was previously with the Council for Exceptional Children, working primarily as a policy specialist on the education of children with exceptionalities. E-mail: [e-mail].