Last Friday, as I was about to board a plane, I read an article about an exciting initiative being launched in Washington, D.C.

During the flight, I drafted a long, gushing piece, praising Abigail Smith, the new deputy mayor for education, and arguing that D.C. was becoming the most important city for systemic reform after New Orleans.

Upon landing, I was on the verge of posting the piece when I saw another D.C. schools announcement.

This one took the wind from my sails.

I sadly shelved the paean.

Here’s the story: D.C. has recently undertaken two invaluable reforms that, when combined with the city’s other systemic features, place D.C. on the brink of becoming the urban school system of the future.

But a third announcement shows that some city leaders are still tragically wedded to the old, failed approach.

I’ll start with the good news, then the bad, and end with a recommendation for solidifying D.C.’s place as a national model for systemic reform.

The Mayor’s Office announced that unused district facilities will be made available to charters (with a preference for high performers) and that the city will establish a common enrollment system for district and charter public schools. Kudos to Smith and all involved.

When you add this to what the city already has, you can see the outlines of a true “system of schools.” D.C. has an independent charter sector (charters are authorized by a non-district entity) that is large, growing, and has many standouts. The city also has a small but doughty private school scholarship program.

What comes into focus is a thrilling performance-centered, choice-based, sector-agnostic approach. If this system were portfolio-managed wisely, new schools would be regularly opened, great schools would be expanded, failing schools would be closed, and families would choose from among a variety of options. This is the continuous-improvement process for urban schooling.

What astonished me is that the city coupled these two great announcements with a truly retrograde proposal: giving the district the power to authorize charters. There are numerous reasons to strenuously oppose this.

First, DCPS—via the old school board—had authorizing power before, and it was terrible at it. It ultimately gave it up, and the city’s other authorizer had to adopt the district’s charters, which was a mess that took years to clean up.

Second, districts in general are bad authorizers. Their inherent micro-managing, compliance-oriented tendencies come out, and they are put in the senseless position of overseeing schools that compete with their own.

Third, D.C. is one of the few cities that actually has a key systemic element correct: It has separated school operation and authorization. It has a district and numerous nonprofits that operate schools but do not authorize. It has an authorizer, DCPCSB,* that does not operate schools. This proposal would be a major step backward, embedding these two very different functions in a single organization.

Fourth, both personally and professionally, I enormously admire district leader Kaya Henderson, but she won’t be chancellor forever. A charter-hostile successor combined with the district’s anti-charter instincts would produce problems galore.

Lastly, this is a cynical attempt to prop up the broken district. For decades it has been unable to run great schools—despite waves and waves of reforms—and families have left it in droves. It is withering as the independent charter sector ascends.

Rather than accepting this reality and winding down the district, the city’s leaders are trying desperately to bail it out, utilizing numerical legerdemain to make the district look better. As the Post notes, “test scores posted by chancellor-chartered schools would count toward the school system’s overall proficiency rates.”

If it were the city’s leaders’ duty to protect the district, this would make sense. But that’s not their job; it’s their responsibility to look after the interests of kids, not the interests of an irreparably broken organization.

Under the current arrangements, great charter operators are already here and more are coming. They have access to a strong authorizer and very good funding and can steer clear of the district’s perpetual dysfunction. Why in the world alter this?

If DCPS wants to partner with a great charter operator, it (or a district-affiliated nonprofit) can negotiate an agreement with a CMO, and they can jointly propose a school to DCPCSB.

Despite these objections, the Mayor’s Office was actually half right; the city does need another authorizer. They just got the “who” and the “what” wrong.

Here’s the alternative: Create a new entity, as permitted by the D.C. School Reform Act, and have it oversee all of DCPS’s schools.

In other words, develop an authorizer for the district. DCPS could continue to run its schools as it saw fit, but each would have a contract with the new authorizer, which could close those that persistently fail.

D.C. would then have the most promising and rational system of schools in America: Every public school (district and charter) would have a performance contract and would be overseen by an authorizer. Rather than propping up the district, the city would have it function like a CMO—an operator of schools whose longevity and market share are determined by its performance.

Operators would operate, authorizers would authorize, and the Mayor’s Office could begin to judiciously manage the portfolio of schools.

So let’s give the city a mulligan. We’ll pretend the district-as-authorizer proposal never happened, swap it for an authorizer-of-the-district proposal, and I’ll gladly publish my paean now sitting on the shelf.

Heck, I’ll even lend a hand if they’d like.

-Andy Smarick

* My Bellwether colleague Sara Mead is a member of the DCPCSB; the piece represents my views.

This blog entry first appeared on the Fordham Institute’s Flypaper blog.