Under the statute, funds were distributed to geographically defined districts, known as planning and service areas (PSAs). The court found a disparate impact when it compared the amount of funding directed under the two formulas to the four highest and four lowest percentage minority PSAs in Florida. The high percentage minority PSAs received less funding under the State's formula than they did under the alternate formula (Meek, 724 F Supp, at 906). The Meek court ultimately found that this disparate impact could not be justified by any valid State policy and so held that Florida's distribution of funds under the Older Americans Act violated title VI regulations promulgated by the United States Department of Health and Human Services. Plaintiffs' complaint herein frames their title VI regulation claim by reference to the State education financing system as a whole, alleging that "the State education financing scheme" causes a disparate impact on minority students. 45 However, plaintiffs' posttrial submission attempts to demonstrate disparate impact by focusing entirely on evidence of disparities solely in the allocation of State aid. Plaintiffs may have focused solely on the State portion of school funding in recognition that both the Supreme Court and the Court of Appeals have held that disparate funding caused by wealthier, higher spending, school districts within the State is an inevitable—and constitutionally permissible—consequence of a State-local partnership in education funding (San Antonio Ind. School Dist. v Rodriguez, 411 U.S. 1, [187 [187 Misc.2d Misc.2d 104] 104] 50-51, supra; Board of Educ., Levittown Union Free School Dist. v Nyquist, 57 N.Y.2d 27, 45-46, supra). 46 Focusing then entirely on the State component of education funding, the court finds that from 1994-1995 to 1999-2000, New York City has consistently received less total State aid than its percentage share of enrolled students. In those years, New York has approximately 37% of the State's enrolled students and has received a percentage of total State aid ranging from 33.98% to 35.65%. This is evidence of disparate impact. Defendants' argument that transportation aid and building aid should be omitted from this analysis is without foundation. These two funding components are obviously essential in delivering core educational services. Defendants also argue that since State aid is generally awarded according to average student attendance, not enrollment, the correct comparison is to measure State aid by the City's attendance, not by its enrollment. Because City schools tend to have greater problems with truancy, their average attendance tends to be a smaller portion of total enrollment than that of other schools around the State. Accordingly, comparing State funding to a student roster measured by average attendance would tend to decrease the City's historic pro rata shortfall. The court holds that disparate impact may be measured by comparing funding to enrollment figures and that the data presented by plaintiffs demonstrates a disparate impact in this case. Defendants' policy justifications for distributing funding by attendance are more appropriately considered in the second stage of the disparate impact burden-shifting analysis mandated by title VI regulations. In other words, when the burden of proof shifts to defendants, the court will consider their argument that using an average attendance measure to distribute funds is supported by a "substantial legitimate justification." The parties' other nonstatistical evidence concerning disparate impact in State aid distribution is not persuasive. [187 [187 Misc.2d Misc.2d 105] 105] Plaintiffs' expert, Dr. Berne, examined a weighted student unit, "State Aid per TAPU for expense," a unit that is used annually by SED to compare funding among districts in New York City. He found that New York City received less State aid per this pupil unit than the average in the rest of the State. In 1997-1998 New York City received approximately $3,007 in State revenue per TAPU for expense. The average for the rest of the State was approximately $3,325 per TAPU for expense. By this measure, the City receives approximately $318 less per pupil unit. Dr. Berne did not testify as to whether this disparity was subject to a test of statistical significance. Such a test would provide some indication whether the observed disparity was explainable by chance. By contrast, defendants' expert Dr. Wolkoff, in his study of State-wide per capita funding disparities, which is described below, did conduct tests of statistical significance of the comparatively small disparities he observed using a difference of means test. Dr. Berne's analysis of TAPU per expense funding disparities is undercut by his failure to address whether the disparity observed in funding allocated by TAPU per expense was attributable to chance. Additionally, Dr. Berne's failure to conduct a similar analysis in other school years limits the probative value of the conclusion that he drew from this single year's disparity. On the other hand, Dr. Berne's analysis may be considered to be conservative in that it ignores two factors that would tend to exacerbate disparate funding. First, it does not take into account New York City's higher regional costs in providing educational services. As noted above, New York City's higher regional costs depress its buying power for educational services. When examining funding disparities among geographic districts, differences in costs within those geographic districts are relevant in a disparate impact analysis (see New York Urban League v State of New York, 71 F.3d 1031, 1038, supra). Second, it does not account for the higher proportion of STAR tax relief afforded school districts outside the City. As noted above, the STAR program has the effect of directing more education dollars to districts outside the City. Plaintiffs also attempt to rely on certain data introduced during the testimony of defendants' expert Dr. Wolkoff. This data purportedly compares New York City's per pupil spending with the rest of the State excluding New York City. However, the exhibits relied upon by plaintiffs for this point all contain [187 [187 Misc.2d Misc.2d 106] 106] State-wide averages that include New York City. 47 In neither his direct testimony nor his cross-examination did Dr. Wolkoff break this data out into a State-wide average that excluded New York City. During cross-examination, plaintiffs showed Dr. Wolkoff a memorandum of law submitted by defendants in support of a pretrial motion for summary judgment that purported to perform this division between the City and the rest of the State for data from 1995-1996. However, the relevant portion of the memorandum cites an affidavit submitted by Dr. Wolkoff which does not contain this calculation. 48 The court was unable to ascertain whether the calculation contained in the memorandum received the imprimatur of any expert. 49 Accordingly, the court declines to rely on this evidence proffered by plaintiffs. To disprove disparate impact defendants relied on State-wide per capita funding comparisons conducted by Dr. Wolkoff. For two school years, 19951996 and 1996-1997, Dr. Wolkoff examined the per pupil (unweighted) expenditures of all of the State's school districts, and then made the assumption that minority students in each district are funded at the district average. By averaging minority and nonminority per pupil expenditures in all districts across the State, Dr. Wolkoff found that average State aid per minority pupil as measured by attendance was $4,097 in 1995-1996 while average State aid per nonminority student in attendance was $4,019. In other words, as measured by students in attendance, minority students in the State actually received an average of $78 more per pupil than nonminority students. In