Note: Jal Mehta, associate professor at the Harvard Graduate School of Education, is guest blogging this week.

One problem with the school reform debate today is the way we cut the debate. One common way of pitching the problem is as “hard” versus “soft.” In this telling, valor is awarded to those willing to make “hard” decisions: these people support merit pay, firing bad teachers, holding schools accountable, and closing failing schools. On the other side, from this point of view, are those who are “soft": people who are opposed to measuring outcomes, who in theory want to empower teachers but in practice want to support a failing status quo. Alternative certification and charter providers are good; traditional preparation and traditional public schools are bad.

Conversely, critics like Diane Ravitch take precisely the opposite view. In their view, the villains are those who want to “privatize” the system through expanded charters, increased merit pay, vouchers, union-busting, and other market-oriented schemes that challenge the fundamental nature of public education. Ironically, despite taking this stance, Ravitch casts the debate in much the same way as do her opponents: market-oriented reformers on one side and democratic traditionalists on the other.

I think we should see the problem in a different way - the key difference is between “thin” and “thick” theories of change. Policies like top-down accountability, No-Child-Left-Behind style, are “hard,” but they have a “thin” theory of change--namely that if we set standards and hold people accountable for meeting them, they will figure out how to hit those benchmarks. Similarly, two decades of experience with charter schools (writ large) suggests that that is also a “thin” theory of change - namely that if we remove regulation and give people autonomy, they will create better schools. The idea of “relinquishment” today is another example: we assume that if we deregulate and give parents choice, schools will magically improve - again, possibly, but it is all dependent on who is running those schools, what they know, what faculty they can recruit, and so forth. Thin theories are attractive, because you can put them on a bumper sticker, but producing good education is hard, and thus most things that work take a “thicker” theory of change.

What would a “thick” theory need to encompass? Consider that any field has the following four components: 1) knowledge--developing the knowledge that will be used in the field; 2) human capital--attracting, selecting, training, and retaining the people who will work in the field; 3) organizational processes at the site of delivery--developing effective processes that govern the work where it is going to be carried out; and 4) overall performance management and accountability. All fields have to make some choices about these four elements; ideally, they form an effective process in which the components are linked to produce good practice.

Given the way that the educational field has developed, it has, in recent years, become very strong on the last element (accountability) but has continued to be weak on the first three. By contrast, in stronger professions, such as medicine, there is considerable attention to the first three--extensive basic and applied science; high selectivity in determining who can enter the field, plus extensive clinical training; strong quality control processes at the site of delivery--but relatively little with respect to back-end accountability (there is malpractice, but there is no “No Patient Left Behind”). No Child Left Behind, in contrast, is strong on back-end accountability but weaker on the first three, an imbalance which partially explains why it has not been able to achieve its ends.

All four pieces of the puzzle (practice-relevant knowledge, strong human capital, school-level processes of improvement, external support and accountability) need to be present and integrated in order to generate quality practice. Devolving power to schools, for example, is likely to yield at best inconsistent results unless there are other processes in place to supply good people and knowledge to those schools. Similarly, developing practice-relevant knowledge is unlikely to matter if there aren’t school-level processes to incorporate this knowledge. If the stool is missing one of its legs, it tends to topple over, even if the other legs are strong.

Teach for America is an example of an organization whose teacher-focused strategy is delimited by the fact that it only partially controls the four levers of the process. TFA came up with an ingenious way to get more smart college students to teach (human capital), and they have increasingly studied the most successful of their teachers to see what practices they should teach their new recruits (knowledge). They have also tried to give their graduates real-time feedback about their teaching and lesson planning over the course of the year in an effort to break down the isolation of the profession and move towards collective standards (organizational processes). However, they are sending their young teachers into highly troubled schools, which lack their own effective collective organizational processes and whose notions about teaching may be at cross-purposes with TFA. Because TFA has not been able to control all the levers needed to create a fully functioning system, its recruitment success has, in many cases, been overwhelmed by the dysfunctions of the system that its teachers are entering.

Some charter schools and particularly effective CMOs have had considerably more success because they do control the various levers of the process. (Not coincidentally, many TFA alumni head these CMOs, in part because of a desire to create better conditions than those they experienced as teachers.) These CMOs recruit mission-aligned teachers, develop knowledge that is consistent with their pedagogical beliefs, and have clear school norms and cultures that are consistent with their mission. Many of these networks, like KIPP, shutdown or eliminate schools that are not performing to their standards (back-end accountability). These processes are common across schools with widely varying pedagogical beliefs--project-based schools like Expeditionary Learning as well as more traditional ones like KIPP and Achievement First. In recent years, both the “no excuses” and the pedagogical progressives have started their own graduate schools of education in part because they wanted even greater alignment across the chain--they wanted to be able to train teachers in the pedagogical approach employed in their schools.

At a national level, Finland, one of the strongest countries on the PISA results, is a good example of a highly successful school system that integrates elements in the inverted pyramid model outlined above. Finland has moved away from centralized control, vastly limiting the role of the central ministry and centralized assessments and accountability. Foreign visitors to Finland often ask about the nature of the testing and accountability system, and are surprised to hear that there are no centralized assessments, consequences for failing schools, or any of the other hallmarks of American efforts to rationalize schools. Instead, Finland selects its teachers carefully, invests heavily in training, and creates time and opportunity for teachers to collaborate The central authority’s role is limited to setting some general standards that outline what students are expected to achieve at different grade levels. These pieces fit together--the delimited role of back-end accountability works because of the strength of the upfront investment in selection and training of teachers.

Usually we put Finland on one side of the debate and CMOs on the other, because Finland is “soft” and CMOs are “hard.” But I think both are worthy models, in their ways, because both represent “thick” theories of good schooling.

-- Jal Mehta