As we see the growth of virtual school options, I anticipate that more states will contemplate performance-based funding schemes. Michael Horn noted this trend in his “Digital Roundup” in the most recent issue of Education Next. Florida, for example, reimburses the Florida Virtual School (FLVS) using a completion-based funding formula. This means per-pupil course funding is provided only upon a student’s successful completion of the course. Utah’s Senate Bill 65 stipulates that providers receive 50 per cent of funding upfront and the other half upon student completion. These early examples hold promise as online learning spreads and concerns about appropriate accountability mechanisms grow with it. As Michael Horn also noted, the next wave in performance-based funding will hopefully include actual measures of student performance rather than just paying providers based on course completion.
Performance-based funding is an encouraging shift from inputs to outputs in online education. We want to align online providers around the right incentives—not just collecting education dollars based on raking in enrollments, but actually aiming to keep students engaged throughout the duration of the course. In the case of for-profit providers, such funding mechanisms may also temper critics’ fears that commitment to the bottom line will eclipse student learning as top priority.
Although I embrace this shift, I worry that this solution feels too tidy. As is often the case when education reformers try to figure out easy fixes to messy accountability challenges, there may be vicious byproducts to a virtuous policy. Here are a few unintended consequences that performance-based funding could produce:
Cherry-picking higher performing students. Having recently graduated from law school, I find myself tempted into legal metaphors, so bear with me. Among plaintiff’s lawyers, there is a common model called contingency fee lawyering; lawyers only get paid if they win in court. This arrangement helps poorer clients gain access to legal recourse when they might otherwise not be able to afford representation and holds lawyers accountable for their performance. Critics point out, however, that contingency fee lawyers may cherry-pick only the most promising and least risky cases. This limits access for plaintiffs on shakier legal ground. In turn, this may also limit the development of rich legal precedent as riskier cases are not heard and corresponding case law doesn’t evolve.
What does this have to do with online education providers? My concern is that if we emphasize performance outcomes, providers will leave the most challenging or high-risk students—those least likely to complete their courses—out of their business model. It’s true that performance-based funding formulas can take into account the fact that some students will demand more support by paying more for the success of those students (much how schools serving high free-and-reduced lunch populations receive Title I funding). It’s unclear, however, whether those additional funds will be sufficient to compel providers to cater to more demanding students’ needs—or how much is the right amount. Such formulas also tread on the slippery slope of the soft bigotry of low expectations for at-risk students.
Only paying for things we can measure. Performance-based funding may also marginalize content areas where performance is harder to measure. A theory about what is and should be measured lies at the heart of a performance-based funding scheme. Current policies suggest that course completion is the right metric. Designing a system, however, in which all courses only prove their worth when students pass might marginalize courses that resist binary measurements of pass/fail. I worry that performance-based funding could treat all courses as created equal, without considering how this might shape course content down the line.
Should we be paying for service or for outcomes? On this point, it’s worth keeping in mind that business models accord with how companies offer value: solution shops are businesses that diagnose problems and recommend solutions. For example, intuitive medicine and most (non-contingency) law firms fall into this category, as do high-end consulting firms. Solution shops are almost always fee-for-service because their results are based on the speculation and intuition of experts, rather than on hard and fast rules. Solution shops rarely use standardized processes because their value is resident in the experts they employ, rather than in processes.
An alternative business structure is a value-adding process (VAP) business: unlike solution shops, VAP’s use a fee-for-outcome model based on clear metrics for performance. A VAP’s ability to deliver value is embedded in strong standardized processes. Manufacturing and food services are good examples of VAP businesses—they bring inputs into one end of their premises, transform them by adding value, and deliver higher value products to customers on the other end. Our research shows that disruptive innovations can harness standardized procedures that move parts of a solution-shop type industry into VAP model. For example, in medicine, retail health clinics use a fee-for-service model for simple solutions like flu shots, at a lower cost than a solution-shop style visit to the doctor’s office.
I don’t think policymakers are considering this distinction as they build accountability systems that pay providers purely based on course completion or performance. Performance-based funding policies will relegate online providers to a strictly VAP business model. For certain types of courses, this sounds right. For example, when teaching fundamentals, we can argue that students unambiguously do or do not master basic literacy and math, so paying for outcomes makes sense. I worry, however, that there are some types of concepts or courses that won’t fit so tidily into a VAP paradigm; higher order thinking, analytical writing, or courses that involve more subjective analysis may belong in a solution shop where we pay online providers not based on their outcomes, but for their expertise.
As online education continues to grow, policymakers may need to think harder about how to treat certain subjects differently within their funding formulas because not all courses will necessarily have clear-cut metrics for success. Stanley Fish, reflecting on Derek Bok’s writing, described these unintended consequences in a column last month in the New York Times:
[W]e’re probably measuring the wrong things and the right things are not amenable to measurement. If this is true and it is also true that the culture of measurement is in the ascendancy, we might expect that things that resist measurement—quality, poetry, insight—would be dismissed and set aside, on the reasoning that if it can’t be measured, what good is it? A new technology typically turns its limitations into a mechanism of evaluation and consigns phenomena outside its capacities to the margins, not merely to its margins but to the margins of what is generally significant and worth worrying about.
The fear that poetry will die out might be dismissed as a cliché in 21st-century education reform, but the next time you consider the virtues of performance-based funding, keep his warning in mind. Performance-based funding is certainly an appealing approach to control quality in online education and fits well into courses where the outcomes for which we’re aiming are clear. But we don’t know yet whether these funding systems will relegate such “immeasurables” to the margin; it’s one risk down the line if we apply this promising policy too bluntly.