It's not often that we share Antonin Scalia's highly idiosyncratictake on Supreme Court business. But, on the matter of Chief JusticeJohn Roberts's emerging jurisprudential style, we find ourselvesnodding violently in agreement with Scalia's recent judgment.Signing onto a Roberts decision that helps gut the McCain-Feingoldcampaign finance law, Scalia scolded the chief justice foreffectively overturning an earlier ruling without owning up to it."This faux judicial restraint is judicial obfuscation," Scaliaprotested.

Nowhere does Scalia's critique of Roberts's style apply moreforcefully than with regard to the Court's recent ruling in ParentsInvolved v. Seattle. Roberts's plurality opinion in thehigh-profile desegregation case elbows aside some 50 years ofprecedent while affecting a posture of doe-eyed innocence. If thephrase "What, little ole me?" doesn't appear anywhere in the text,that's only because it wasn't necessary. It is implied inpractically every word.

The case weighs plans by two public school districts, one in Seattleand the other in Louisville, to ensure a modicum of racialintegration within their schools. The way the districtsaccomplished this was by regulating the proportion of white tononwhite students each school could enroll. So, for example, if awhite student wanted to attend a school where white students werealready significantly overrepresented relative to the districtpopulation, that student might have to attend another schoolinstead. The arrangement had helped the districts buck recentschool resegregation trends even though it only directly affected atiny fraction of students.

Not surprisingly, the arrangement also relies on a form of racialclassification--how else to keep track of the proportion of whitesto nonwhites?--which is what Roberts finds unacceptable. Or, if notunacceptable, then at least so "inherently suspect" as to only bejustified under exotic circumstances. (And, even then, only ifpracticed in a highly circumscribed way. ) Under Roberts's readingof the Constitution, pretty much the only reason a district shouldbe able to classify students by race is to correct pastdiscrimination by the district itself. Sounds reasonable enough. Buthere's the rub: Since Seattle never formally practiced segregation,and since a federal court has proclaimed Louisville desegregated,neither program currently passes constitutional muster, accordingto Roberts.

Suffice it to say, we're skeptical. At the very least, as JusticeStephen Breyer points out in his angry dissent, the distinctionbetween formal and informal segregation is not nearly so neat asthe hyper-literal Roberts suggests. For example, although it's truethat no local law ever mandated segregation in Seattle, thedistrict's policies--which allowed whites to transfer out ofheavily black schools but restricted blacks from transferring intowhite schools--historically achieved the same goal. More to thepoint, the Supreme Court has repeatedly affirmed that, whiledistricts aren't necessarily required to adopt integration plansabsent a history of de jure (i.e., legally mandated) segregation,they're certainly allowed to. Apparently sensing that precedentruns against him on this point, Roberts simply ignores it, arguingthat the relevant portions of an oft-cited Court opinion appear inan ancillary passage and are therefore moot. It's the equivalent ofoverturning a landslide election on a technicality.

But to make this a debate about legal minutiae would be to overlookthe truly radical implications of the ruling. At heart, Roberts'sopinion is an assault on Brown v. Board of Education, the landmark1954 desegregation case, masquerading as an attempt to redeem it.Under its mainstream interpretation, Brown required school boardsto actively integrate themselves, rather than merely abolish rulesthat prevented black students from attending white schools. InRoberts's view, however, the problem with the pre-Brown regimewasn't racial isolation per se, merely that "schoolchildren weretold where they could and could not go to school based on the colorof their skin." To appreciate how significant a reinterpretationthis is, ask yourself the following: If, post- Brown, a schoolboard in Alabama had ended formal segregation while nonethelesspreventing black students from attending white schools--say, bybarring children from predominantly black neighborhoods fromenrolling--would the Court have been satisfied? Of course not. (Infact, the Court more or less said as much in 1968, after Southernsegregationists pursued a variation on this strategy.) But suchefforts would have been consistent with Roberts's reading of thedecision.

As it happens, it's a reading that's been percolating for decades,though it has never achieved currency outside a relatively smallgroup of conservative legal activists and scholars. Today, the viewlives on in elite organizations like the Federalist Society, withwhich Roberts has long been affiliated. Indeed, the much-cited codato Roberts's opinion--that "the way to stop discrimination on thebasis of race is to stop discriminating on the basis of race"--islifted almost verbatim from a 2005 dissent by circuit court judgeCarlos Bea, also a Federalist Society booster, which itself recallsa slogan favored a decade ago by former solicitor general TheodoreOlson, another Federalista.

It is precisely the mainstream consensus on Brown's meaning thatdrives Roberts and like-minded conservatives to be so coy in theirdiscussion of the opinion. When asked during his 2005 confirmationhearing whether he accepted "both the holding and the reasoning inthe Brown case," Roberts responded with a less-than-transparentmusing about the "genius of the decision"--a decision he has justhollowed out. Justice Scalia is right: There's a yawning gapbetween modesty and obfuscation. The more you see of John Roberts,the more you appreciate the difference.

Chip Shot

Say this much for our president: He is consistent. Back in the1990s, when he was still governor of Texas, he had an opportunityto help some of his neediest constituents get affordable medicalcare. The federal government had just created the State Children'sHealth Insurance Program, or s-chip, making billions in newspending available to states that created public insurance programsfor poor kids and their families. While even some of his fellowRepublican governors were jumping at the opportunity, Bush balked.The program might get expensive in the long run, he feared. And, ohdear, it would mean more government. So Bush fought efforts tocreate an expansive s-chip program in Texas, arguing a minimalistversion would be better, even though the state had one of thehighest proportions of uninsured residents in the country.

The Texas legislature, though hardly a bastion of socialism, didn'tsee things Bush's way--and bullied him into supporting a biggerprogram. But now, as president, Bush is waging the same fight allover again. S-chip is up for reauthorization this year. With evenmore Americans uninsured than in the '90s, Congress seems inclinedto expand it so that it can cover more people. Under a newbipartisan proposal in the Senate, s-chip funding would increase by$35 billion over the next five years, allowing it to reach many ofthe children who still lack insurance.

But Bush, backed by some of his more conservative allies, wants nopart of this. He's willing to reauthorize the program, but he alsowants to restrict it- -by, among other things, limitingeligibility to only those people who are below 200 percent of thepoverty level. Under his counterproposal, at least 17 states wouldactually lose s-chip funding, meaning that more kids and familiesin desperate need of medical insurance would go without.

Why the resistance? Money, for one thing. To pay for the expansions,Democrats have proposed raising taxes on cigarettes and rescindingsome of the unnecessary subsidies Bush's Medicare drug planfamously throws at insurance companies--steps that are no morepopular with this administration than with its supporters in thetobacco and insurance lobbies. But the fight is also philosophical.Bush and his allies object that, for every ten people who gaininsurance through s-chip expansions, between two and five fewer willget private insurance--since employers, particularly those withlow-income workers, will be less likely to offer coverage once thepublic alternative is available.

But this hardly matters as long as the net effect is an expansion ofinsurance--which clearly is the case here, as the CongressionalBudget Office recently concluded. Thanks to s-chip, the number ofpeople without health insurance today is much lower than it mightbe otherwise; if the program expands, it will be even lower in thefuture. To put it in human terms, that means millions of more kidsgetting their regular checkups and--when they need it--more seriousmedical care.

Of course, what really spooks Bush and the right is the possibilitythat s- chip could be a stepping stone to universal coverage. It'sa legitimate fear. If Americans notice what a good job governmenthas done insuring kids, they might clamor to have it cover moreadults, too--which, come to think of it, sounds like a pretty goodidea as well.