A recent U.S. News & World Report article by Andy Rotherham made a bold claim: school districts do too much. His piece very convincingly argues that school systems spend inordinate resources and mental energy on providing student services, like food and transportation, to the determinant of focusing on their most valuable job: instruction. If we want to drive learning, the argument goes, school districts simply don’t have the capacity and wherewithal to do everything. Schools instead should start unbundling—outsourcing nonacademic services to other agencies and providers across the system—and redouble their commitment to teaching and learning.
In Rotherham’s words, “this approach—do everything—is an anomaly among successful American enterprises. Yet we abandon specialization in the design of school districts.” But it turns out, in fact, that across industries, when firms are trying to boost performance dramatically, they actually do tend to “do everything.” In order to innovate toward new heights of performance, oftentimes firms actually need to wrap their arms around the various and sundry components that could contribute to performance.
For example, in the early days of the mainframe computer industry, IBM could not have existed as an independent manufacturer of mainframe computers because manufacturing was unpredictably interdependent with the design process for the whole mainframe system: the machines, operating systems, core memory, and logic circuitry. Specifying how these discrete features of the system should fit together was essentially impossible at the outset. If the company had tried to separate each of the components—for example, subcontracting operating systems to one outlet and logic circuits to another—the product would have suffered. The company would have been forced to back away prematurely from the frontier of what was possible in raw performance by specifying these constituent parts too early. And in any given industry, so long as the performance of products and service underperforms customers’ needs, companies cannot back off of that frontier. It was no coincidence then that, by and large, independent suppliers of those component computer parts did not even exist initially when IBM designed the earliest mainframes.
The same goes for schools. There are all sorts of interdependencies among the factors that contribute to learning, especially for schools serving students who may not be receiving the supports they need outside of school in order to arrive ready to learn. As Michael Horn and I wrote last year in “The Educator’s Dilemma,” schools struggling to boost achievement likely need to integrate across academic and nonacademic services as far as possible, so long as we are still figuring out the precise supports we need in place to combat the effects of poverty on students’ ability to learn.
In other words, innovation theory suggests that in some circumstances—in which academic performance is reliably high or interdependencies of nonacademic and academic factors are well understood—the suggestion that schools “unbundle” might be well-founded. But unbundling is unlikely to help drive outcomes in schools where students are underperforming and the precise services that they need are not abundantly clear cut.
Although still sparse, there is some research to suggest that keeping these services tethered to schools can prove a powerful antidote to achievement gaps. For example, in their 2009 study of Harlem Children’s Zone, economists Will Dobbie and Roland Fryer studied outcomes of students attending charter schools in the Zone with those only receiving services from the Zone. The researchers concluded that “high-quality schools or community investments coupled with high-quality schools drive these results, but community investments alone cannot.” It bears noting that some use this study to suggest that only in-school instructional quality matters; however, Promise Academy itself offered a range of non-academic services to students in addition to high quality instruction including but not limited to free medical, dental and mental-health services, student incentives for achievement, nutritious cafeteria meals, support for parents in the form of food baskets, meals, and bus fare.
Another study highlighted in Brookings’ Chalkboard blog last week showed similar promise out of Boston College’s City Connects program, an initiative that helps schools to develop, secure, and coordinate supports that target academic and nonacademic barriers to achievement. City Connects’ results suggest, to Rotherham’s point, that schools may not have to be the sole provider of all services. But it also suggests that arm’s length outsourcing won’t get the job done: without a strong role in coordinating the types and dosages of supports dolled out to each student, schools risk missing the benefits those supports could yield in boosting academic performance.
The point then is not that schools should do everything, all the time. But we need to acknowledge that outsourcing and innovation can work at odds with one another. School districts can’t expect to outsource services that are crucial for their students without sacrificing critical control over the frontier of possible innovations that could tailor those supports to drive outcomes. And even factors like bus schedules and diet can contribute to the unique mix-and-match of supports and new approaches that schools—particularly those increasingly focused on tailoring personalized learning approaches to each individual student—may need to offer a given student or family.
Of course, in-house service provision or coordination is more expensive than awarding contracts to the lowest bidder and walking away. But by integrating first, school districts won’t have to take on everything forever. Across numerous industries, seemingly less-efficient integrated systems typically, over time, give way to unbundled modular ones that are more affordable. This occurs because as they continuously integrate to drive performance, interdependent approaches eventually overshoot their performance goals, ultimately leading to an unraveling of the higher cost structure behind a fully integrated organization. This unraveling leads to a more modular world in which an organization can prosper by outsourcing or supplying just one element, or subcomponent, of a product or service.
In other words, the call for schools to do less is not unfounded, but premature, at least among those struggling to innovate toward boosting performance. Although narrowing schools’ onus of responsibilities is certainly attractive on paper, without a theoretical underpinning to this line of argument, these measures don’t hold necessarily such strong odds to improve outcomes at schools serving low-income students who disproportionately fall on the wrong side of the achievement gap.
Full disclosure: I sit on the advisory board of InterconnectED, an initiative of the Boston College Lynch School of Education’s Center for Optimized Student Support, which is informed in part by the City Connects model.
For more, see: