(concurring).
After considerable debate and reflection all members of this panel have agreed that (with one minor exception) the Rhode Island statute is constitutional. The careful and conscientious opinion of the majority reflects the fact that we have all found this a difficult case. Given the difficulty and the work and thought that the majority’s opinion reflects, I am tempted to join it. But I do not do so because there remains a disagreement that I believe is significant. In essence, the majority holds that, while it is technically bound by a minor precedent to reverse, the law, in logic and spirit, really runs to the contrary. Indeed, the majority writes that, were it not for the letter of this unargued Supreme Court case, namely, Springfield School District v. Department of Education, 483 Pa. 539, 397 A.2d 1154, appeal dismissed for want of a substantial federal question sub nom. School District of Pittsburgh v. Department of Education, 443 U.S. 901, 99 S.Ct. 3091, 61 L.Ed.2d 869 (1979), it would affirm the decision of the district court. Members of the Jamestown School Committee v. Schmidt, 525 F.Supp. 1045 (D.R.I.1981). But, I do not agree that this panel’s result rests upon so slender a precedential reed. Moreover, I believe the Establishment Clause calls for a more *14“practical” approach to this type of problem than the comparatively “theoretical” approach taken by the district court and the majority of this panel. Since the court’s opinion invites reconsideration of Springfield, it seems appropriate to explain why I believe the district court is wrong in holding the Rhode Island statute unconstitutional.
The Supreme Court in Everson v. Board of Education, 330 U.S. 1, 67 S.Ct. 504, 91 L.Ed. 711 (1947) specifically held that the Establishment Clause of the United States Constitution does not forbid a state (“as a part of a general program,” id. at 17, 67 S.Ct. at 512) to pay for transporting school children to Catholic, as well as to public, schools. The Court reasoned that a law that provides for transporting school children is a law that promotes “the welfare of the general public,” id. at 14, 67 S.Ct. at 511, much like a law that provides “ordinary police and fire protection, connections for sewage disposal, public highways and sidewalks,” id. at 17-18, 67 S.Ct. at 512. As such, the Court held that it was not the type of law “which is designed to support institutions which teach religion.” Id. at 14, 67 S.Ct. at 511. This principle is clearly enunciated in Everson. It has been clearly applied in many later cases (see, e.g., Cromwell Property Owners Association, Inc. v. Toffolon, 495 F.Supp. 915 (D.Conn.1979) (very similar issue to that in case at bar); Bennett v. Kline, 486 F.Supp. 36 (E.D.Pa.) (same), aff’d, 633 F.2d 209 (3d Cir.1980) (without published opinion); Snyder v. Town of Newton, 147 Conn. 374, 377-79, 161 A.2d 770, 772-73 (constitutionality of busing upheld even though transportation to private schools created “some additional expense to the town”) appeal dismissed for want of a substantial federal question, 365 U.S. 299, 81 S.Ct. 692, 5 L.Ed.2d 688 (1961); West Morris Regional Board of Education v. Sills, 58 N.J. 464, 279 A.2d 609 (busing within specified cost limit for students to private schools up to twenty miles away upheld), cert. denied, 404 U.S. 986, 92 S.Ct. 450, 30 L.Ed.2d 370 (1971); Springfield School District v. Department of Education, supra; Rhoades v. Abington Township School District, 424 Pa. 202, 220, 226 A.2d 53, 64 (1967) (constitutionality of busing for private school children upheld even if parochial schools receive indirect financial benefit from busing), appeal dismissed for want of a substantial federal question, 389 U.S. 11, 88 S.Ct. 61, 19 L.Ed.2d 7 (1967); State ex rel. Hughes v. Board of Education, 154 W.Va. 107, 174 S.E.2d 711 (1970), appeal dismissed and cert. denied, 403 U.S. 944, 91 S.Ct. 2274, 29 L.Ed.2d 854 (1971)). In my view the district court’s decision is wrong, not simply because it conflicts with the disposition of the Springfield case, but because it is inconsistent with several aspects of this more firmly established principle.
First, the record does not reveal any significant favoritism of “religions as against nonbelievers” in theory or in practice. Torcaso v. Watkins, 367 U.S. 488, 495, 81 S.Ct. 1680, 1683, 6 L.Ed.2d 982 (1961). The district court found a theoretical favoritism (presumably sufficient to warrant a finding of unconstitutionality, see, e.g., Wolman v. Walter, 433 U.S. 229, 254, 97 S.Ct. 2593, 2608, 53 L.Ed.2d 714 (1977) (field trip program held unconstitutional because of “unacceptable risk of fostering of religion”)) in the fact that Catholic school students have a greater right to transportation outside “school districts” than do public school students. But this fact is irrelevant for the purposes of finding a theoretical favoritism, for the “school districts” that the court discussed are public school districts, the boundaries of which were drawn with public schools rather than parochial schools in mind. It is hardly surprising that each Catholic school is not located in the center of each public school district. A program that seeks approximately equal or even absolutely identical treatment for Catholic and public school students will likely transport some Catholic school students across public school district boundaries. Presumably, some public school students will be taken across Catholic school boundaries (if such boundaries exist) as well. It is simply not possible for a theoretical inequality to arise only from the fact that busing will take some private school students across public school boundaries.
The matter would have been very different had the district court found, as a practi*15cal instead of a theoretical matter, that Catholic schools obtain a disproportionate benefit because they serve students from regions that are significantly larger than the local school districts. But the record suggests the contrary. The defendants offered evidence which suggests that no matter how the problem is viewed, the Catholic schools receive no practical advantage. According to this data, in 1978-1979 the state transported nearly 90,000 public school students and about 8,300 religious school students. It apparently paid over $10 million for public school transportation, but merely a little more than $850,000 for religious school transportation. In fact, if one focuses only upon students who were bused outside of their local public school districts, one finds that more public school students were bused (2,864) than religious school students (1,154). And (perhaps because many public school students transported across district lines were handicapped children) the total inter-district public school transportation bill was $2 million while inter-district parochial school busing cost less than $125,000.
The most relevant figure, however, is the transportation cost per pupil. The defendant’s figures suggest that the cost per pupil for transportation within public school districts was $100 per public school student and $102 per parochial school student; the cost per pupil for transportation outside the public school districts was $108 per parochial school student but $702 per public school student (perhaps reflecting the cost of busing handicapped students). Taking all (inter- and intra-district) transportation costs together, the state’s costs amounted to about $120 per public school pupil and $103 per religious school pupil.
These figures are not inherently implausible. In fact, they are fully consistent with the hypothesis that Catholic students live, on average, a bit further from their schools than public school students. After all, in a small state like Rhode Island per pupil transport costs may be more a function of the extent to which an individual bus is filled with students than of the distance traveled.
The district court did not “accord these figures great weight,” primarily because the “survey provide[d] no information as to how the reported costs were calculated.” Members of the Jamestown School Committee v. Schmidt, 525 F.Supp. at 1050. The court also noted that in some areas a school had to pay much more for transporting students to a particular Catholic school than to a public school. But the district court did not decide this case on the ground of practical advantage or significant difference in transport costs. To the contrary, it apparently accepted the defendant’s figures arguendo, and held that, while it was somewhat more expensive to bus religious school students outside public school districts than to bus public school students within public school districts, how much this difference might be was “unimportant.” Id. at 1051.
The statewide dollar differences in per pupil expenditures of the sort introduced in this case, however, could not support a conclusion that this law, as a practical matter, impermissibly favors Catholic schools over public schools. The differences are de minimis. It would be impossible for a state to create a system guaranteed to produce absolute equality of costs. Indeed, the state’s figures could reflect far greater disparities before they came close to the type of favoritism that has been held to cause a constitutional problem. See, eg., Springfield School District v. Department of Education, 483 Pa. at 558 n. 9, 397 A.2d at 1164 n. 9 (“A constitutional consideration would arise only if the cost of transportation for the students attending sectarian schools was so disproportionate that it became apparent that the transportation provided to the public school youngster was merely a ruse to confer a benefit to the sectarian school pupil.”); Bennett v. Kline, 486 F.Supp. at 39 (adopting Springfield rationale); Cromwell Property Owners Association, Inc. v. Toffolon, 495 F.Supp. at 923 (“At some point, the cost of inter-district transportation for students attending sectarian schools may become so grossly disproportionate compared to the ordinary expense of public school busing that the ‘indirect’ benefits in regionalization accruing to sectarian institutions will rise to a constitutionally significant level.”). The cost differentials in isolated *16cases, while possibly relevant, prove little in and of themselves, for in other isolated instances public school busing is presumably more expensive than private. The issue at stake is the effect of the plan as a whole; police protection does not become unconstitutional because in one precinct it is unusually expensive to protect a church, nor fire protection because in one neighborhood the church alone is made of wood. The district court’s uncertainty about comparative costs would warrant at the most further factual development were the matter more questionable.
Second, the transportation statute is neutral on its face. It divides the state into regions and provides that:
A pupil attending a school, including a public school, vocational school, special education program ..., a consolidated school ..., a regional school ..., or a non-public non-profit school ... shall be provided with bus transportation to the school or facility which the pupil attends, within the region in which the pupil resides ....
R.I.Gen.Laws § 16-21.1-2 (1981). Can a court find this statute unconstitutional simply because it can find another statute (setting up public school districts) which, in combination with the transportation statute, creates what it (wrongly, in my view) conceives as a theoretical advantage for Catholics? Does a fire department violate the Establishment Clause if it buys equipment tall enough to reach a church dome if zoning laws make the equipment unnecessary for other buildings? If so, and more important for purposes of this case, does the Establishment Clause strike down general laws that would authorize such expenditures even if in fact the dome is no higher than other buildings? My point is that, if the Establishment Clause strikes down such an “expenditure” law, it must do so because in fact the law will lead to disproportionately large expenditures favoring religious institutions, not because there is a theoretical possibility it could do so at some other time or in some other place.
Third, there are other practical considerations. As long as Everson is the law, proportionate state expenditure on school transportation is support for child safety rather than support for the church. See Everson v. Board of Education, 330 U.S. at 18, 67 S.Ct. at 512. If so, there ought to be a feasible way for a state to do what it has a right to do — namely, to draft a statute that pays for transportation proportionately. Rhode Island’s experience thus far suggests that this is easier said than done. And, one might test the more “theoretical” approach by asking what, if this statute were unconstitutional, the legislature here ought to have done? Should it have confined Rhode Island’s Catholic school students to public school districts, even though many districts contain no Catholic schools? Should it have bused a student to the border of such a district, where the child can walk across the boundary and board the next district’s bus? Should it have provided that Catholic students can be bused only to those Catholic schools that happen to be located in the middle of a public school district? Should it have said that school children who attend a Catholic school that happens to be located on the boundary between public school districts A and B can be bused if they live in A, but not B; or that they can be bused if they live in either district but must travel on separate (and hence more expensive) buses so that neither bus crosses the district line? The fact that it is difficult to think of a statute that would satisfy the approach of the district court and the fact that plausible statutory solutions seem arbitrary or only tangentially related to any serious Establishment Clause objective suggest that the more “theoretical” approach is wrong. If Ever-son is right, there should be a way to apply it in Rhode Island.
Fourth, I am concerned, here as elsewhere, about the practice of combining several different statutes to find in the combination a theoretical preference, which is then held unconstitutional. In Usher v. Schweiker, 666 F.2d 652, 661-62 (1st Cir.1981), we pointed out that the Equal Protection Clause does not automatically invalidate a statute simply because that statute can be combined with another to produce a result that seems unfair or even irrational. Rather, given the vast number of laws, *17rules and regulations, any absolute “rational total net result” requirement, as applied to possible combinations of different laws, is unworkable and would make legislating too difficult. The same is true here. It is easy to think of laws that might combine with transportation, police or fire protection statutes to produce theoretical advantages for religious institutions. A state law may end the public school day at 4:00 p.m., giving Catholic school students the theoretical right to take buses at different (and more expensive) times. A state law may place public schools in safer areas of the city, making police protection for private religious schools potentially more costly. A state law may require public schools to be built of fire-proof material, making fire protection for religious schools potentially more expensive. The problems engendered by combining different statutes and then asking whether, in theory, the combination gives Catholic students something their public school counterparts are denied are not ones that rise in this case to a constitutional level. The ease with which statutes can be combined to produce possible differences suggests that a court should examine the significance of these differences in practice.
For these reasons, unlike the majority of this panel, I do not find it surprising that the Supreme Court dismissed Springfield for want of a substantial federal question. Nor am I surprised that other cases have followed Springfield. See e.g., Cromwell Property Owners Association, Inc. v. Toffolon, supra; Bennett v. Kline, supra. Indeed, Justice Black in Everson noted that the state could have required “a local transit company to provide reduced fares to school children including those attending parochial schools, or ... a municipally owned transportation system [could have] undertake[n] to carry all school children free of charge.” These schemes would have been fully as constitutional as the plan the state did adopt, he explained, for the denial of such services was “obviously not the purpose of the First Amendment.” Everson v. Board of Education, 330 U.S. at 17-18, 67 S.Ct. at 513; see Wolman v. Walter, 433 U.S. at 253, 97 S.Ct. at 2608 (reaffirming validity of illustration). Yet, the Justice made no calculation of per pupil costs, and mentioned no maps of school district lines— even though local trolley lines rarely pay heed to school district boundaries. The First Amendment, after all, is impartial. It plays no favorites between public and private school students, and that impartiality implies that it not penalize parochial school students on the basis of public school district lines. It is possible for expenditures for a neutral purpose like school transport to be weighted too heavily in favor of religious school students and thus to offend the Constitution. However, the question of constitutionality in such a case is primarily a question of practical effect, measured in terms of costs incurred by the state and actual benefit conferred on the parochial school student. No unconstitutional practical effect has been demonstrated here. For this reason, and not simply because of the Supreme Court’s dismissal of the Springfield case, I agree with the court’s result.
I also agree that § 16-21.1-3 (providing for busing to a school outside a student’s region where “there is no similar school within the region”) is unconstitutional. In giving the Rhode Island commissioner of education the authority to examine the religious content of a private school’s curriculum and then to make judgments about the propriety of busing based on that examination, the section entangles church and state in a manner that the Establishment Clause was designed to prevent.
The Supreme Court has severely limited a court’s power to intervene in religious disputes, even to decide property matters. See Presbyterian Church in the United States v. Mary Elizabeth Blue Hill Memorial Presbyterian Church, 393 U.S. 440, 89 S.Ct. 601, 21 L.Ed.2d 658 (1969) (courts may not consider theological issues when resolving property disputes between religious bodies). See Freund, Public Aid to Parochial Schools, 82 Harv.L.Rev. 1680 (1969). Of course the “wall between church and state,” Everson v. Board of Education, 330 U.S. at 18, 67 S.Ct. at 513, is not absolute. A state may, for instance, inspect religious schools to see that they meet the state’s educational standards. See Board of Education v. Al*18len, 392 U.S. 236, 246 n. 7, 88 S.Ct. 1923, 1928 n. 7, 20 L.Ed.2d 1060 (1968). Similarly, it may provide textbooks to religious school students, despite the consequent risks of entanglement. See id. at 244-45, 88 S.Ct. at 1927; Wolman v. Walter, 433 U.S. at 236-38, 97 S.Ct. at 2599-2600; Meek v. Pittenger, 421 U.S. 349, 95 S.Ct. 1753, 44 L.Ed.2d 217 (1975). See generally Freund, supra. Yet, it is difficult in this case to find any strong educational or other reason to have the state’s commissioner of education examine the theological orientation of one religious school to determine whether a student is entitled to inter-regional busing to attend another. The Rhode Island legislature should be able to redraft this severable section to limit inter-regional busing in a less intrusive way.