Why Education Innovation Tends to Crash and Burn
Having opined a good bit about "innovation" (check out Ed Unbound for much of my current thinking), I'm sometimes asked about why it's so hard to scale promising programs, models, pilots, and notions. On that note, I just had the chance to spend a few days with a bunch of terrific folks discussing just this topic at a Kauffman Foundation retreat. Kaufmann will be issuing a synthesis with the collected wisdom that emerged. Meanwhile, I figured I'd share my own thinking with you.
There are at least two big sets of obstacles when it comes to "scaling" innovation. First, innovative models often rely on tough-to-replicate elements. Second, there are key structural conditions that impede efforts to grow even more replicable models.
First, seemingly successful pilots often depend more on the conditions that attend their adoption and execution than the model itself. Pilots tend to benefit from a number of advantages that disappear when these efforts start to "scale," namely:
• Philanthropic support- Dollars are often available to fund new initiatives. Such funding allows CTE or remedial programs to offer services and opportunities that prove unsustainable when the program expands to new sites that lack the extra resources.
• Expertise- Pilot efforts are, by design, promoted and supported by the experts who have conceived of the model (or intervention). They benefit from intense, sustained, loving attention by those who are most knowledgeable about and invested in the idea. Later sites have less access to that talent.
• Enthusiasm- Pilot efforts are inevitably launched where the leadership (and/or the faculty or instructors) are enthusiastic enough about the venture that they're willing to invest all the energy necessary to launch it. That passion and sense of ownership are enormously helpful in making early iterations successful.
• Accommodating policies- Pilot efforts are frequently launched where they are because the local leadership has the wherewithal to get the waivers, leeway, or support to launch the effort. So a new academic program benefits from special treatment when it comes to staffing rules or funding. When the same models are implemented in less accommodating settings, the models frequently fail to deliver the anticipated results.
Second, even promising models run afoul of several structural impediments, including:
• Reliance on entrenched institutions- In nearly every sector, transformation is the product of new ventures emerging with new ways of leveraging tools, talent, and technology. Remember, the average lifespan of a Fortune 500 company is only fifty years. Three-fourths of America's 25 biggest firms didn't exist in 1965. Most transformation is not the work of "system change" but a natural consequence of new providers replacing the old. In education, however, there's a presumption that staid institutions (in K-12 or higher ed) can and will remake themselves with the right encouragement and support. It's a noble aspiration, but it's a hell of a challenge.
• Lack of price competition- The immense subsidies and public underwriting in education, along with the fact that most public school families are insulated from questions of cost because they pay nothing (other than taxes) for their child's schooling, blunt sensitivity to cost. The problem is that the vast majority of innovation in the world is not "disruptive" but is focused on modest, incremental boosts in productivity. These frequently show up in price competition. In education, however, the inelasticity of the demand curve means that institutions have spent little time or effort trying to build or scale increasingly cost-effective models; instead, most energy has been devoted to models that are bedecked with bells and whistles.
• Lack of outcome comparability- Because there are not good or comparable measures of outcomes, it's difficult for even highly effective models to glean the benefits from their performance. This makes it hard for new models to displace familiar providers, and places a premium on marketing and perception while limiting the rewards for quality.
• Discomfort with for-profits- For-profits typically have much more ability and incentive to scale than non-profits. Whereas non-profits are driven by the passion and commitment of their leadership, for-profits are propelled by the logic of maximizing returns. This leads for-profits to take risks in pursuit of growth and to aggressively seek opportunities for expansion. Absent pressure to maximize returns, non-profits tend to default to the innate attraction of focusing on existing clients; they therefore tend to grow much more slowly. Non-profits also have a much more limited capacity to attract the capital necessary to fuel growth. Yet public discomfort with for-profits means that our hopes frequently rest on non-profits or public entities.
What can be done about any of this? It's hard to be sure. Most of what's tripping up innovators is either inherent in the enterprise or a product of deep-seated norms and arrangements. That said, three thoughts:
• Put a premium on innovations that scale easily- The most difficult innovations to scale are those that rely heavily on talent and complex implementation. The easiest to scale are those that leverage technology or other tools to provide services with few moving parts. For instance, Amazon.com or Facebook are remarkably easy to scale, because most of the quality of the experience is almost identical for thousands or even millions of users. Similarly, Tutor.com is easier to scale than is a program which depends on recruiting and training local tutors.
• Resist the notion that innovative models can readily be housed in existing institutions- Established institutions have established norms, cultures, policies, and routines. No matter how energetic and enthusiastic are those who would adopt innovative models, the difficulties of maneuvering around these realities makes innovation a bad bet. Innovations may be adopted successfully here or there when backed by committed leadership, but they can be quickly bent into unrecognizable forms when adopted by others that are less committed.
• Focus on cost and outcomes in allocating public dollars- Encouraging the successful scaling of innovative models is going to depend in large part on whether the larger environment supports such ventures. An environment dominated by formula funding, hefty subsidies, and few useful measures of quality is designed to accommodate the status quo. Changing that requires changing public policies at the federal, state, and local levels.
All of this helps to explain why "innovation" is a term of endearment in an Apple store, but more likely to sound like an epithet in the nearest teacher workroom or faculty lounge. Resolving this state of affairs will be hard, but probably no harder than watching waves of promising new ideas crash and burn.