When a court needs to give interim relief (such as issuing a preliminary injunction or granting a stay of a lower court order), it is acting where there is a risk of error.  For this reason, the standards for determining whether interim relief is warranted are necessarily probabilistic, taking into account the risk that the court will err and the harm to each side if the Court does err.  (I discuss the general Remedies standards in more detail in chapter 9 of this book.)

When the Texas court issued its interim maps in the Texas redistricting case, it did not take into account how likely the Voting Rights Act challenges to those maps were likely to be.  As I explained in my earlier post, the Supreme Court held today that this was an error.  But from the point of view of Remedies, what’s most interesting about the opinion is the splitting of hairs that the Court commands between dealing with a potential section 2 violation (a likelihood of success on the merits standard, familiar from the Supreme Court’s Winter case) and a “reasonable probability” standard for Section 5 (which the Court says it puts in place because earlier caselaw says that only the Washington DC court, and not the local three-judge court crafting the interim plan, can decide whether or not a plan should be precleared). The more I look at the language, the harder it is for me to understand what the Court was trying to accomplish.

Pages 5-6 of the opinion have three paragraphs on this question.  But let me highlight the language specifically on the difference between the two standards.

Likelihood of success:

Where a State’s plan faces challenges under the Constitution or §2 of theVoting Rights Act, a district court should still be guided bythat plan, except to the extent those legal challenges areshown to have a likelihood of success on the merits. Plaintiffs seeking a preliminary injunction of a statute mustnormally demonstrate that they are likely to succeed onthe merits of their challenge to that law. See Winter v. Natural Resources Defense Council, Inc., 555 U. S. 7, 20 (2008).

Reasonable probability (my emphasis):

The need to avoid prejudging the merits of preclearanceis satisfied by taking guidance from a State’s policy judgments unless they reflect aspects of the state plan thatstand a reasonable probability of failing to gain §5 preclearance. And by “reasonable probability” this Court means in this context that the §5 challenge is not insubstantial. That standard ensures that a district court is not deprived of important guidance provided by a state plan due to §5 challenges that have no reasonable probability of success but still respects the jurisdiction and prerogative of those responsible for the preclearance determination. And the reasonable probability standard adequately balances the unique preclearance scheme with the State’s sovereignty and a district court’s need for policy guidance in constructing an interim map. This Court recently noted the “serious constitutional questions” raised by §5’s intrusion on state sovereignty. Northwest Austin, 557 U. S., at 204. Those concerns would only be exacerbated if §5 required a district court to wholly ignore the State’s policies in drawing maps that will govern a State’s elections, without any reason to believe those state policies are unlawful.

Why the splitting of hairs?  To a non-technical reader, “likelihood of success,” “reasonable probability,” and “not insubstantial” chance of success sound like they are the same.  Yet in the context it is clear that they are not the same standards.

It seems, as Michael Li suggests, that this language was some kind of political compromise.  From the Supreme Court oral argument, it was clear that Justice Sotomayor did not believe the Texas court could give any deference to the Texas maps while the preclearance case was pending.  It was also clear that Justice Kennedy was concerned about how this interim map (and section 5 generally) infringed on Texas’s sovereignty.  Was this language crafted to fudge the issue, so that some can claim the “not insubstantial” is a tougher standard than likelihood of success on the merits (admittedly my first read of the language) and others suggesting that is an easier standard (as in, “not insubstantial” means “not frivolous”).

The more I read the language the less sure I am I know whether reasonable probability is meant to be greater or lesser than the likelihood standard.  It is even worse, because the Court instructs that the three judge court “should presume neither that a State’s efforts to preclear its plan will succeed nor that it will fail.”

I’ll be interested in hearing from others about the meaning of these two standards.