From the start, accountability has been the biggest issue in this latest fight over reauthorization of the Elementary and Secondary Education Act. And it’s promising a bumpy road ahead, too, if the deal that sailed through a conference committee becomes law.

Some background: Everyone, from civil rights groups to state chiefs, agreed it was time to remake the federal role in the current version of the law, the No Child Left Behind Act. And everyone agreed that states and districts needed more running room, as long as there were equity provisions in place for the students that No Child Left Behind was designed to help, such as English-language learners, special education students, and those in racial minorities (aka “subgroup students”).

It’s taken aides and lawmakers months of behind-the-scenes negotiations to arrive at what they say is a good balance.

So exactly where did the deal land on accountability, including when it comes to the very tricky issue of just how much tests will count for accountability purposes? (We previewed that issue for you here.)

This is where it gets even trickier. It seems that even congressional aides have different interpretations of what the agreement says.

Here’s what seems obvious from the language of a late stage version of the deal: The agreement is clearly stronger on accountability than either the House or Senate versions of the bill.

For one thing, it would appear to get rid of so-called “supersubgroups"—a statistical technique in the Obama administration’s No Child Left Behind waivers that allows states to combine different groups of kids for accountability purposes that are usually considered separately. States, such as New Mexico, like these subgroups because they allow for flexibility in accountability. But civil rights advocates—and former Rep. George Miller, D-Calif.—say they mask gaps.

And the deal calls for states to incorporate a mix of academic factors into their accountability systems, including: graduation rates, English-language proficiency, achievement on standardized tests and one other academic factor. That additional indicator could be growth on standardized tests—many states consider both growth and achievement in their accountability systems already under the Obama administration’s NCLB waivers.

In addition to those academic factors, states must also choose another, very different kind of indicator to gauge school performance—things like school climate, educator engagement, postsecondary readiness, or school safety. Importantly, this indicator has to be something that can be reported by subgroup.

Here’s where things get even trickier. The deal says that each of the academic indicators must carry a “substantial weight” relative to the others and relative to at least one more factor, which could be something like school climate, educator engagement, access to advanced coursework, or another factor that works for the state. And the academic factors have to count for more, as a group, than these less traditional factors.

Here’s where aides diverge:

A GOP aide sees that language as meaning that the secretary couldn’t tell states that academic factors have to account for more than 51 percent in a schools rating. (States of course, could go far beyond this.) That’s clearly the message that advocates for practioners are getting, as you can see from this summary of the bill prepared by AASA, the School Administrators Association. And the idea is giving Sandy Kress, an original architect of NCLB some serious heartburn.

A Democratic aide sees it differently. The deal gives the secretary the authority to set ranges for how much each of the four academic factors would count towards a school’s rating, this aide said.

Under that interpretation, the secretary could determine that the weighting of each academic indicator on its own must be a certain range. That situation could give academic factors a much heavier weight overall than the 51 percent academic/49 percent other-factor threshold.

(Politics K-12 translation: Under that aide’s logic, it would seem the secretary could say, for example, that proficiency on tests must count as 20 percent to 40 percent of a state’s acccountability system. And the secretary could peg each of the other academic factors at, say, 10 percent to 30 percent. That would mean that old-school academic factors would get more weight than other, new-school factors.)

But what about those prohibitions on secretarial authority, you say? In this aide’s view, there’s language in the deal making it clear the secretary can indeed regulate in this one key area.

Confused? I don’t blame you. It’s clear as mud. (No wonder civil rights groups are biding their time before weighing in on this deal until they see real, live legislative language.)

And no wonder this was the first take from Charlie Barone, the policy director at Democrats for Education Reform, who served as an aide to Miller, who was the top Democrat on the House education committee when NCLB was written:

“Based on what I’ve seen, for the next [education] secretary, interpreting the new law will be like looking at a Rorschach with one eye closed and with both hands tied behind their back.” (Barone had only seen preliminary drafts when he said that.)

In the end, that complexity could actually be a boon to state and local control, especially since the compromise includes nearly all of the restrictions on the secretary’s authority that were in the House and Senate versions, a GOP aide said late last week.

“The complexity helps,” a GOP aide said. The agreement “leaves a lot of this to states to figure out and the secretary’s ability to interfere with those state decisions is astonishingly limited.”

It won’t be a surprise if some of the big winners in this ESEA reauthorization are education lawyers, said Chad Aldeman, an associate partner at Bellwether Education Partners, who served in the U.S. Department of Education under President Barack Obama.

“The fact that there is that question does not suggest that it is strong policy,” Aldeman said. “You wouldn’t want to draft a bill where there is very foggy uncertainty about what that means and be put into immediate legal limbo. What can the secretary do and not do? I think that’s where the lawsuits will be.”

Or maybe it won’t matter, in the long run. State chiefs say they are not backing away from accountability, no matter how much new leeway they get. (The theme of this year’s Council of Chief State School Officers annual meeting, after all, is equity.)

“I’m bothered when I hear people say that school chiefs won’t hold schools accountable,” said Brenda Cassellius, Minnesota’s education chief. “That’s not been evident with the waivers. ... We’ve supported our schools and we’ve held them accountable. I hope America can see that.”

Politics K-12 co-blogger and Assistant Editor Andrew Ujifusa contributed to this post.