Here is a guest post from Nate Persily:

In the immediate aftermath of the NAMUDNO decision, I shared the view espoused by many that this was a placeholder decision that urges Congress to act so that the Supreme Court won’t need to declare section 5 of the VRA unconstitutional. After contemplating what the next case might look like, however, I am starting to think that maybe the Court has placed itself in a position where it might never confront the basic constitutional questions surrounding the coverage formula. To understand why, one must work through the four ways such a challenge could get to the Supreme Court.
First, as with the constitutional challenge in NAMUDNO itself, a covered jurisdiction could argue that Congress has exceeded its authority under the Enforcement Clauses by requiring the jurisdiction to submit its voting law changes for federal approval. The response from the Court would appropriately be an exhaustion-style argument: go seek bailout and if you are denied bailout, then come back and see us. The implication here is that jurisdictions unconstitutionally covered by section 5 — because they have not engaged in the requisite predicate unconstitutional acts that would justify coverage and preclearance — should be able to bail out. The coverage formula is only unconstitutional if “good” jurisdictions are unable to bail out easily and that can only be demonstrated once they try.
That leads to a second potential course to the Supreme Court: appeal from a denial of bailout by the District Court for DC. Then the jurisdiction would argue, also somewhat similarly to NAMUDNO itself, that either bailout should be granted or the coverage formula is unconstitutional. Here, again, the Court can avoid constitutional difficulty (and probably can do so more easily than it just did in Monday’s decision). One possibility is that the jurisdiction was appropriately denied bailout and the coverage formula remains constitutional as applied to it (see Tennessee v. Lane for some similar logic on this point). Or the jurisdiction should be allowed to bail out and the bailout statute should be read so as to make it possible for such a jurisdiction, given its voting rights track record, to be able to bail out. The bailout requirements seem individually severable, so I could see the Court rereading them to require, with the most extreme interpretation, for instance, that jurisdictions with a ten year clean record of no intentional discrimination are entitled to bailout. This may contort the bailout statute, but I think severing it in this way (or interpreting it to avoid constitutional difficulty or even vindicating an as-applied challenge to the bailout statute) seems about as easy a course as the one they just took.
The third way a case gets to the Court is from an appeal from a denial of preclearance by the District Court in DC (akin to Georgia v. Ashcroft). The covered jurisdiction alleges that its voting change is allowed by the Constitution, that the new standard for retrogression (Ashcroft and Bossier Parish-fixes) is unconstitutional, and/or that the coverage formula is unconstitutional. Here again I would expect the Court to read the retrogression standard to avoid constitutional difficulty (which is a tough job, as 30-plus pages of mental gymnastics in my Yale Law Journal piece attests) or simply sever it from the rest of the statute, which might leave the retrogression standard, perhaps, as limited to actual violations of the Fifteenth Amendment. In other words, the Court would allow the voting change to go forward, while still not reaching the question whether the jurisdiction is appropriately covered. The new retrogression standard might be the part of the statute that would most bother Kennedy, given his opinions in the Shaw cases etc. By defanging or even deleting the new retrogression standard, though, the Court should not need to reach the question of the constitutionality of the coverage formula.
Finally, a case could come to the Supreme Court because a jurisdiction fails to submit a voting change and is then sued in local district court to force them to submit the change for preclearance, and that decision is appealed all the way up. The jurisdiction might defend its actions by saying that the coverage formula is unconstitutional, on its face or as applied to it. Here again, the response from the Court is akin to exhaustion: if you don’t want to submit your voting changes for preclearance, try to bail out. If the jurisdiction is unconstitutionally denied bailout, then the Court can deal with it then.
Of course, each of these situations presents the possibility that the Court will deal with the larger issue of the law’s constitutionality. And if we view the NAMUDNO decision as simply a warning shot to Congress, maybe the Court will shoot more directly if Congress does not act soon (which was my initial response to the decision). However, it is also possible that the law could perpetually be read to avoid constitutional doubts so a never to force the Court to earn the headline: “Court Strikes Down Voting Rights Act.”