Evaluation of Scaling the New Orleans Charter Restart Model:

Our Research Questions

7. What can we learn from the CRM schools that ultimately did not succeed?
Six CRM schools in this evaluation eventually failed: they lost their charters, exited the CRM cohort, or closed their doors entirely. What can we learn from this subset of schools?
All CRM schools – indeed, all schools in any district – must have the capacity to absorb exogenous shocks as they come. Circumstances beyond a school’s control (enrollment changes, facilities issues, district policy changes, and the like) will necessarily impact a school’s operation but need not negatively impact student learning if the school is equipped to respond. The disruption inherent in taking over a school for turnaround is unavoidable but anticipated: the CRM Theory of Action situates CMOs as strategic agents which will pilot their schools through fraught environments and provide the supports needed to keep schools focused on students first and foremost. Yet student impact results indicate that fewer than half of CRM schools post significant positive effects on student learning (see Impact results here). Further, of the CRM schools posting negative impact on student learning, there are six that have failed profoundly by virtue of the fact that they no longer exist in their CRM formulation:
SchoolOperatorCurrent Status
New Orleans:
ClarkFirstLine SchoolsAbandonment of CRM model
Crescent Leadership AcademyROPExit from CRM cohort for noncompliance
John McDonoghFINClosed after second year of operation
McDonogh 42ChoiceLost charter for academic nonperformance
HumesGestaltTaken over by another CMO
KlondikeGestaltSurrendered charter

While we observe some common factors which contribute to these schools’ challenges, these factors are not necessarily unique to these schools. Rather than attribute a school’s failure to a particular set of unique characteristics, we find that failure among the CRM schools follows the accumulation of a set of common challenges, beginning at selection. These schools did not exit the field because of any single endogenous characteristic or exogenous event; nor can these failures be explained by differences in student populations. The six CRM schools that exited the field did so not because they differ in kind from other CRM schools, but because they differ in degree. The six schools that failed outright did so because they experienced multiple operational shocks, of great degree, over their entire organizational lifespan, while lacking the internal capacity to absorb those shocks in order to recover.

One of the factors we see impinging on school functions is that of grade span match between CRM and Flagship schools. For example, Clark and Klondike both suffered for their grade span mismatch with other schools in their networks. Their operators, Firstline Schools and Gestalt Community Schools, lacked experience and expertise in secondary and elementary education respectively. However, a grade span mismatch is not a guarantee of failure: Freedom Prep and Cohen also serve different grade spans from all other schools in their respective networks; and eight CRM schools face either a grade span mismatch to their flagship or have no flagship at all.

Similarly, leadership turnover is significantly associated with lower performance for CRM schools. The six failed schools each experienced at least one leadership turnover and five of the six experienced more than one change in leadership. But, again, leadership turnover does not guarantee complete failure: six additional CRM schools experienced leadership turnover without closing or losing their charters.

Location moves and facilities challenges also created disruptions to school operations according to respondents. In New Orleans, this was expected: Katrina destroyed or damaged school facilities citywide and many schools were forced to occupy temporary structures or to open in buildings knowing they would eventually move elsewhere. Tennessee CRM schools did not experience moves from place to place, but degraded facilities presented typical challenges of upkeep. Co-location with closing schools also surfaced real conflict between CRM and outgoing school staffs. For example one CRM school reported that the closing school’s administration would deny CRM students access to shared spaces within the school (such as the gym) and alleged that closing school staff pulled fire alarms during testing to disrupt students’ performance in the CRM school. Interestingly, even schools in new facilities found fault with their buildings: schools were not designed to accommodate particular programmatic elements (e.g. maker spaces), were more difficult to keep secure, or were too big for targeted enrollment.

These exogenous factors were not unique to any particular CRM school, but some schools weathered more or more intense shocks than others. But beyond schools that continue to limp along, we must consider the nearly one third of CRM schools no longer exist in the CRM at all.

CLA and John McDonogh, discussed at length below, were identified by NSNO and RSD as problematic organizations from selection onward, yet we see little evidence of systems-level actors remediating the myriad problems they identified. Clark experienced multiple leadership turnovers, failed to establish positive culture, and ultimately abandoned its mission of high quality, college preparatory education. Gestalt surrendered the charters to Humes and Klondike rather than continue running schools that had failed to turn around in a 4-year window. And McDonogh 42, which struggled to recover from leadership and staff turnover, location moves, and limited CMO capacity, ultimately was designated for closure by the state and will be handed to another charter operator to attempt turnaround.

The CRM Theory of Action posits that schools will succeed by virtue of their CMOs; and that CMOs will succeed by virtue of their embeddedness within a system overseen by NSNO and RSD/ASD. Yet we see little evidence of systems-level responses in moments when CMOs work to keep even bad schools open so as to avoid the reputational and financial impacts of school closure and (in a more generous interpretation) to redouble efforts to support students. The system-level partners trusted in their Theory of Action far beyond the point at which CMOs revealed themselves as lacking capacity to keep schools afloat or, perhaps, behaved as bad actors Organizational Capacity Report.

Two Case Studies: John McDonogh and Crescent Leadership Academy

To better illustrate these arguments – that failed schools suffered for an accumulation of shocks; and that the schools themselves, their CMOs, and that systems-level partners failed to absorb those shocks – an examination of two failed New Orleans schools may be particularly instructive: John McDonogh and Crescent Leadership Academy.

John McDonogh High School represents the most dramatic failure of the CRM. John McDonogh was operated by Future is Now (FIN), a new enterprise developed by a nationally-known school advocate who served in leadership positions for both FIN and an existing West Coast-based CMO. NSNO expressed strong concerns at selection about FIN in terms of both CMO- and school-level leadership capacity and quality. NSNO also raised concerns about FIN’s lack of track record as a standalone entity, as compared to the more mature and robust CMO that was FIN’s progenitor. Yet NSNO awarded FIN funding based in large part on its CEO’s reputation in response to political pressure from LDE/RSD.

Upon opening, enrollment challenges – hardly unique to John McDonogh – led to policy and operational decisions aimed at appeasing the CMO’s budgetary concerns and political position rather than putting student needs first. The principal and the out-of-state CMO leadership disagreed profoundly about basic pedagogical strategy. These disagreements were further exacerbated by a lack of understanding among CMO- and school-level leadership of both local context and the operational processes and priorities of the CRM as defined by NSNO and RSD. No single one of these challenges – some exogenous, some endogenous – was unique to John McDonogh, but the accumulation of them placed the turnaround of John McDonogh in an extremely precarious position from the start.

Finally, or perhaps fundamentally, the decision to document the turnaround of John McDonogh by inviting a reality television production into the school severely undermined FIN’s credibility as a youth-centric educational agency and disrupted attempts to install culture change in the building. For example, during an evaluation site visit, a CREDO researcher observed an altercation between a student and teacher in the hallway during a class transition. Attracted by the raised voices, a cameraman and producer rushed over the site of the argument, which was already winding down. The producer then asked the teacher and student if they would be willing to re-enact the altercation so that he might capture it on film. In this instance and more generally, having a production crew embedded in the school disrupted daily operations, distracted students and teachers, glamorized student violence and mental illness, and undermined the existential purpose of John McDonogh as an institution of learning.

Still, other than the TV debacle, none of the challenges faced by John McDonogh were unique to John McDonogh. Other CRM schools grappled with leadership-CMO disagreements, suffered for having an out-of-state CMO, faced low enrollment, struggled to establish positive school culture, and failed to fully grasp their position within the larger CRM ecosystem. But no other schools confronted every single one of these challenges, to the degree these challenges presented in this particular school, from the school’s very inception.

Even at closing, political pressure superseded student need. Upon surrendering their charter, FIN messaged the early closure of John McDonogh as the fault of RSD. FIN suggested that RSD planned to repurpose the John McDonogh building, and that this facilities issue was the real reason for John McDonogh’s early closure. NSNO and RSD were complicit in this messaging, allowing FIN’s leadership to salvage their reputations rather than committing to a transparent accounting of FIN’s profoundly and unconscionably flawed operation of the school. In our estimation, this failed commitment to transparency reinforced community concerns expressed at selection and exacerbated public distrust of the CRM as a viable reform strategy in the following years.

Of perhaps even greater illustrative value is the experience of Crescent Leadership Academy (CLA). Unlike John McDonogh – indeed, unlike all other schools in the New Orleans CRM – CLA was structurally problematic in its very design within the CRM ecosystem, in addition to the more typical challenges it faced. CLA aimed to replace the closing Schwarz Alternative High School, the expulsion center or “educator of last resort” prior to Katrina. But where does an educator of last resort fit into an all-choice ecosystem?

An early misstep was RSD’s decision to treat CLA as a full open-enrollment charter school. The notion that any family would select an expulsion center – which, by definition, serves students who did not choose this educational option but rather were forced to enroll when their school of choice expelled them – was untenable. Financing such a school through the usual per-pupil mechanisms, when student enrollment is volatile by design, makes budgeting for student need nearly impossible. Further, holding an expulsion center to accountability criteria based on a 180-day school year, even though their students enroll continuously throughout the year and often on a time-limited basis, theoretically threatens the viability of the entire system. And, indeed, we heard from both CLA and other schools that bad actors would expel low-performing students shortly before testing merely to manage their own testing populations. This behavior was eventually curtailed by the adoption of a community-wide explusion policy that required all expulsion recommendations to be centrally reviewed and approved.

This particular set of factors was unique to CLA, and represented existential threats to the school’s viability from selection onward. We find that the notion of “educator of last resort” is fundamentally incompatible with an all-choice district. But, these fundamental flaws in CLA’s design in terms of the school’s role within the CRM ecosystem could have been ameliorated if systems-level partners had held CLA to its original promises of opening CTE academies, implementing therapeutic and trauma informed models, and developing student-centered programming that went beyond merely holding students in place until they might be returned to their sending schools. None of these promises were fulfilled, however, reinforcing the image of CLA as merely a holding facility for disruptive students.

Instead, NSNO cut CLA from the grant cohort, allowing CLA’s bad behavior to continue unchecked. The systems-level response to CLA reflected an abdication of responsibility for what NSNO knew had been a flawed selection from the onset. CLA failed to meet the terms of their grant agreement, failed to implement promised program offerings, and failed to establish effective leadership and governance practices. Of particular note was the CMO’s deference to CLA’s local board chair, who insisted on hiring an individual to serve as founding principal who shared none of the pedagogical or behavior values ROP professes. Unsurprisingly, the tenure of the first principal lasted only a few months, which then left the school without a principal for much of its first year. The CMO provided a part-time school leader who split duties between CLA and another (out-of-state) site. The principal hired at the end of the first year of operation also failed to meet NSNO’s criteria, resulting in NSNO divorcing CLA from the i3 funding and the CRM project.

CLA’s unique position within the CRM is not the sole driver of its failure. In fact, CLA’s own responses to its situation exacerbated (and were exacerbated by) the far more typical challenges CLA also encountered. Indeed, NSNO declared CLA noncompliant on their i3 grant not because of these system-design flaws, but because of CLA’s failure to meet the terms of their grant agreement regarding leadership quality, program offerings, and management structure. CLA failed to identify a high quality, full-time principal at selection; suffered repeated leadership turnover over the course of the study; never instilled the student programming they had promised; and insisted on facilities specifications that delayed their exit from rotting modular (at one point, a CLA principal reported that he was running a garden hose from a facility next door into his modular because the bathroom lacked running water).

CLA continues to operate (and has had two more leadership transitions since exiting the CRM program). The school’s ongoing attempts to install a therapeutic model, improve facilities, and make a case to families that CLA is a viable long term placement as well as an expulsion center, have had some positive effect on reputation. Further, RSD’s move to centralize expulsion and schools’ attendant capacity-building around student supports, restorative models, and in-school behavior remediation have reduced CLA’s enrollment volatility according to school-level respondents. Yet CLA continues to confront the set of challenges common to other CRM schools throughout the course of the i3 initiative: human capital (particularly leadership quality) deficits, a dearth of teachers who can be effective with the CLA student population and remain dedicated to student success, struggles with culture, out-of-state CMO management, high student need and lack of community-based supports.

That numerous of these challenges at John McDonogh and CLA were noted by NSNO at selection – and, the fact that numerous of these challenges were not unique to John McDonogh or CLA among the CRM schools – speaks to the importance of holding fast to selection criteria, even if the only alternative to a bad selection is to make no selection at all. That numerous of these challenges increased in degree over the first year of operation speaks to the need for ongoing monitoring, support and targeted intervention. In the absence of both system-level (click here) and CMO support (click here) , John McDonogh closed after only two years of operation. CLA still operates, but has continuously struggled to reformulate both its pedagogical and operational models.