Whether you consider yeserday’s New York Times article on K12.com a “hit piece” (Tom Vander Ark) or a “blockbuster” (Dana Goldstein), there’s little doubt that it will have a long-term impact on the debate around digital learning. Polls show that the public and parents are leery of cyber schools, and this kind of media attention (sure to be mimicked in local papers) will only make them more so.

But just as these criticisms aren’t going away, neither is online learning itself. The genie is out of the bottle. So how can we go about drafting policies that will push digital learning in the direction of quality?

This is something we at Fordham are thinking a lot about, and we’ve published three papers (so far) in our series, Creating Sound Policy for Digital Learning: Rick Hess on quality control; Paul Hill on funding; and Bryan and Emily Hassel on teachers. And in January, we’ll publish an analysis by the Parthenon Group of what high-quality fulltime online learning really costs.

I’ll leave it to others to rebut the Times’ extremely selective use of data, expert opinion, and evidence. Where the article landed a punch, in my view, was around the perverse incentives at play today. Clearly K12, and its well-paid CEO, Ron Packard, face strong incentives to boost enrollment at their schools. Unfortunately, states haven’t figured out a way to create similar incentives around quality. And that needs to change.

First, a short digression. I worked at K12 a long, long time ago, just after its creation. (I believe I was employee number 10.) I needed a job, and I convinced Bill Bennett to create a role for me (the august-sounding Vice President for Community Partnerships) in which I would figure out how to take K12’s rich resources and make them available for poor kids. Our basic assumption was that K12’s model—which relied on parents or other caretakers doing most of the instruction—wouldn’t be feasible for kids living in poverty, most of whom would need the custodial care offered by traditional public schools.

To be honest, I didn’t make much progress. The learning materials weren’t even created yet, and so I had few “partnerships” to offer to communities. I left after 9 months for an appointment in the George W. Bush administration.

But what a difference a decade makes. One of the real surprises of the online learning movement is that lots of poor families are choosing to give it a try, and that explains (to a large degree) why K12’s test scores are lagging. (Yes, poverty and achievement are linked, at least for now.)

And the impression painted by the Times article is that online education companies like K12 have every reason to sign up as many parents as possible—poor, rich, whatever—regardless of how prepared they are to tackle the challenge of home-based instruction. Because of some states’ sloppy finance systems, the schools can keep the money if the families change their minds and head back to traditional schools. And, as has been true for all public schools since the beginning of time, the online schools get paid whether their students are learning or not.

Fixing the payment problem is a no brainer. (Schools of all kinds should only get paid for the days of instruction that kids actually show up for.) But is it time to consider performance-based funding, too? To pay companies like K12 more or less depending on how their students perform on state tests or depending on their graduation rates?

In his paper for Fordham, Paul Hill dismisses the idea, arguing:

Pay for performance would create a harsh environment for all education providers. Conventional, virtual, and hybrid schools might spend money on a student’s instruction for a whole course or semester yet receive nothing in return. Online vendors of all kinds, who have little control over their students’ effort or persistence, could be even more at risk. In general, this approach would limit the unproductive use of public funds and quickly destroy any vendor that could not demonstrate good results. It would favor providers with deep pockets, e.g., district-run schools and online vendors supported by large foundations. Performance-based payment as defined here could create a lethal environment for smaller-scale innovators.

He’s probably right about smaller-scale innovators, but I still think it’s worth a try, at least for full-time online schools. (It might be harder in the “blended learning” setting, where a child might be taking just one or two subjects online.)

What if K12 only got paid for every student that made at least a year’s worth of progress on the state test? Some argue that this would create its own perverse incentives, encouraging the company to cherry pick students who are most likely to succeed. But if the measure is student growth, and the test being used is a good one (a big if, admittedly), then all kids but those with severe cognitive disabilities should be seen as contenders.

Instead of indiscriminately signing up students willy-nilly, K12 would then have a reason to vet each family’s situation to make sure they are ready for the rigors of online learning. They would invest, up-front, in assessing whether the child’s parents or other caretakers are up to the task of instructing the student, and whether they have a home situation conducive to success. And then K12 would work like the dickens to make sure every student was making strong progress over the course of the year.

Personally, I’d like to see performance-based pay for all schools. That won’t fly anytime soon, but performance-pay for online learning (at the least the full-time, virtual charter school version) could. Which state is ready to give it a try?

-Mike Petrilli