The Resilience Network Manifesto

I have been writing here on e-Literate about the need for higher education to create a resilience network in response to COVID-19 and future continuity crises (like hurricanes, wars, and so on). In the course of evangelizing this idea with various parties, I have discovered two barriers. First, as with every other aspect of the COVID-19 crisis, people are at different stages of understanding with the first-, second-, and third-order effects. We are not yet all on the same page regarding how different this crisis is and how different our response needs to be. Second, even with people who are far along in that journey, many struggle to envision what an appropriately different response could look like. COVID-19 is revealing the reflexes and habits of mind that have made us vulnerable. But recognizing that our most reliable strategies are failing us is not the same as being able to envision new strategies. We are confronting a cascade of systemic failures, one of which is a failure of imagination.

Time is running out. We are not moving quickly enough and I have apparently not been communicating clearly or urgently enough. This is a call to action.

What’s at stake

Let’s start with what we know.

We know that, while online education can be as effective for students as face-to-face, it often isn’t. We know that delivering quality online education is hard, requiring specifically trained faculty, specifically designed courses, specifically optimized technology services, specifically designed guidance programs, and a whole host of other difficult operational skills. We know that, when colleges don’t get all of these hard operational details right, more students fail out or drop out. Sometimes dramatically more. We know that underserved and neurodivergent students are disproportionately hurt during these failures. We know that students who drop out are at higher risk of not returning.

We know that the most effective online programs have taken years or even decades to hone. We know that the vast majority of colleges and universities do not have years or decades of experience and do not have sufficiently knowledgable staff in sufficient numbers to be able to build such programs quickly. We know that many of the people upon whose shoulders a transition will fall are not active participants in the associations that help people like them find the knowledge and resources that they need.

We also know that even the tried-and-true formulas of the most experienced online programs may not fit the environment we will face several months from now. I am not aware of any university that designed entire degree programs to be adaptable on the fly to multiple possibilities of online and blended. I am not aware of any university that has trained all its faculty and prepared its physical plants for the kind of flexibility we may very well need for an unknown period of time. The vaccine is at least a year away, and that estimate is extremely optimistic.

The other day I was on a call with an administrator I respect from one of the very best and most effective online programs that I know of. One that operates at significant scale. He said he is very worried that his own administration is overconfident in their ability to move the rest of the university online quickly and effectively. He said they think they are a lot better prepared than they actually are. I can’t name that university here, but if I did, you would be appropriately afraid. I certainly am.

We are on the cusp of losing a generation of students. But that is not the worst of it. Because if we lose a generation of students, then many colleges and universities, including ones that we have thought of as safe and stable, will fail. The University of Michigan is forecasting a $1 billion budget shortfall. Every college that emphasizes the benefits of residential education is at particular risk. All of their previous investments in that idea are now costs that they have no obvious way of sustaining. Everything that was a mark of distinction is now a disadvantage.

The Cal State system has announced that all campuses will be closed in the fall and all courses will be online. Their more prestigious sibling, University of California system, is having a harder time making that call. Here’s an excerpt from the article linked above:

The state’s other four-year public university system, the 10-campus University of California, was not prepared on Tuesday to immediately follow CSU’s example, a spokeswoman said.

“We are exploring a mixed approach with some material delivered in classroom and labs settings while other classes will continue to be online,” Claire Doan said in an email to EdSource. “Our campuses will reopen for on-site instruction when it is safe to do so — in coordination with federal, state and local health departments and authorities.”

Classes across the UC system have been online since March. The UC Board of Regents is meeting next week and is expected to discuss plans for the fall. So far, UC San Diego has announced a massive coronavirus testing plan for students, faculty and staff. UC San Diego this week is starting testing for the 5,000 students currently on campus, with plans to expand that to all 65,000 students, faculty and staff in the fall. That campus is aided by having its own medical center.

First in nation, California State University to close campuses for in-person instruction this fall

Think about that billion-dollar shortfall that Michigan is facing. Now think about how the UC system is struggling with the idea of giving up the fees that pay for student dormitories, cafeterias, gyms, and so on. And struggling to imagine how they convert the kind of residential education that they have held up as an ideal into an online format that students will recognize as being of equal quality—and equal financial worth—in three months. Think about how many colleges and universities are in this situation.

We are facing more than one pandemic.

All of those crazy predictions of 25% or 50% of colleges going out of business are suddenly a lot more plausible. We will lose institutions. Some will be ones that you don’t expect. Many that do survive may have to make draconian cuts that, particularly if they are made out of a sense of panic, could fundamentally alter the missions of these institutions while putting them at greater risk of failure. This could happen very quickly.

As with COVID-19 itself, every day counts. We will permanently lose students and universities. Careers will die. Institutions will die. The question is how many. The answer to that question depends on the adequacy of our response.

Our current response is inadequate

I have written before about how the POD Network has published a Google Sheet with 400+ links to university continuity web sites, and that it overlaps to an unknown degree with a separate Google Sheet organized by Bryan Alexander that also has several hundred entries. This is laudable work. And it is impressive, particularly given how much else the organizers have to deal with in the rest of their work and personal lives. Four hundred institutions is a non-trivial percentage of all colleges and universities in the United States. Almost a double-digit percentage. (Not all of the institutions on the spreadsheet are US colleges and universities, but that doesn’t take anything away from the accomplishment.) I have spoken with many educators who are involved in one sort of effort or another to create a database or a web site or a Google Sheet that lists resources. I thank you all. I admire you all.

Please understand that I mean no disrespect when I say that all of these efforts amount to a drop in the bucket relative to the current need. Those people who found time in their busy days to add a link to their web site in the Google Sheet? Each of those people is in the middle of a separate institutional hurricane. It is chaos. OK, they have some content. Is it the right content? Is it everything they need? If not, will they have the time they need to click through each of the 400+ links in that Google Sheet in an effort to find what they need? Are they able to get the information effectively to all the institutional stakeholders who need it, when they need it?

You already know the answers to these questions.

Some of you have pointed out to me that there are years—decades, even—of high-quality resources that people have been creating to help faculty learn how to teach online. That’s true. Let’s imagine that everybody knows about this content. (They don’t.) Imagine that they know where to find it. (They don’t.) And that they know what to look for. (They don’t.) Ask yourself this question: As somebody who knows about all of this great existing content intimately, are you able to fill all of your campus’ informational needs simply by plugging in resources that already exist? Or are you finding yourself creating new documents, videos, and training interventions despite the fact that all these great resources exist, and that you know about them, and you know where many of them are, and you know what you need? As your institution goes through successive waves of the crisis—from online to blended, or blended to online, through the budget crisis, through the thrashing about for new kinds of programs that can keep you afloat, through teaching everybody just what the heck a stackable credential is, through the unknown series of twists and turns—as you go through all of the waves that are coming, what are the odds that you will continue to need to create new resources?

So we’re all in #pivotonline now, and online is finally getting its due. Yay rah. Nobody expected us to pivot this way. Nobody prepared for this. Nobody can fully prepare for what comes next. Because we don’t know what comes next.

We are not ready for this.

Nobody will save us

Will the state of Michigan fill the university’s billion-dollar budget hole? Could they do it even if they wanted to? I don’t know. Will universities that had what they thought were good cushions in their endowments be able to live off their rainy day funds now that the stock market has tanked and revenues are circling the drain? I have no idea. Will the Federal government ride to the rescue? I’m not even sure that the Federal government will continue to deliver my mail six months from now.

No central agent will coordinate the production of COVID tests. Nobody is procuring and distributing PPE, balancing supply and demand across 50 states. Nobody has maintained the stockpile of ventilators and made sure that they are in good working order. There will be no grand rescue plan. Nobody is coming to save us. We must save ourselves.

The corollary is that whatever we can do that is different and more effective than what we are doing now, it will not be beautifully architected and meticulously orchestrated. It will be messy. Inefficient. Chaotic. Redundant. Exhausting. In other words, similar to the current situation, but incrementally and iteratively better. It will not be an adequate response to the crisis either. We are beyond the point where adequacy can be achieved in the foreseeable future. We must now aspire to produce the least inadequate response possible and to make it less inadequate as quickly as possible. This is where we are. The sooner we come to terms with that, the more careers and institutions we will be able to save.

What a resilience network could be, conceptually

When I have spoken about a resilience network with various people, I have repeatedly gotten the sense—sometimes said explicitly, but often not—that people think it is something that I intend to own or control. Let me be clear: the Resilience Network is not trademarked. It is not proprietary to me. This is not about me. It’s about mustering a more effective group response to immense suffering.

In terms of ownership and central coordination, I think of the resilience network as something like a hashtag and something like a potluck dinner. It’s like a hashtag in the sense that it is a way to crowdsource organization with intentionality. Creating lists of useful resources is good. Creating discoverable resources is better. Creating discoverable resources that are tagged with something that has semantic weight is better still. There is a difference between the way the #adorables hashtag works and the way the #MeToo hashtag works. Also, #pivotonline is different than #resiliencenetwork. There’s nothing wrong with either of them, but one focuses intentionality on the problem of the moment—getting courses online—while the other focuses on building a supply chain that can help us respond to the larger crisis as it evolves over time in a way that leaves us more capable rather than less. But the main point is that anybody can label their resource, their list, their workshop, or whatever as #resiliencenetwork.

The resilience network something like a potluck dinner in the sense that the tagging with intentionality enables us to conduct some lightweight collaboration. If we know who is coming to the potluck dinner and what they are planning on bringing, we can begin to identify redundancies and holes. If we have four salads and no proteins, then maybe somebody should bring a protein. Note the level of specificity. Realistically, everybody is producing and sharing what they are capable of producing and sharing. No amount of intentionality will make me capable of cooking salmon without producing an inedible mess. But I can probably manage some burgers if I know that people need protein and already have plenty of greens. We can incrementally but meaningfully improve our coordinated production and sharing of resources.

This description may give you a sense of what a resilience network could feel like, but it’s probably still too vague to give you a clear vision for what it actually does. We need to get more specific and operational.

What a resilience network could do

I have a few ideas about what we could realistically be doing to improve the effectiveness of our crowdsourced response at this moment of great chaos.

Use technologies to help with curation

Between the two aforementioned Google Sheets alone, there are hundreds of web sites that purportedly fulfill the same purpose. It would not be hard to have a web crawler index all of these sites and apply machine learning to begin classifying the documents it finds. It could make a list of all resources that look like instructions for creating effective lecture videos, all the ones that look like they are about facilitating online discussion, and so on. I suspect the level of effort for the first iteration of this is somewhere on the order of one engineer working for somewhere between a few days and a few weeks. Several iterations would likely be required to make the software as useful as it could be. For example, we might cross-reference URLs with IPEDS data so that somebody browsing the database could filter documents for ones being used by institutions that are like their own. We could also increase its value by having it crawl more data sources, like databases that are being maintained by various associations for different purposes. But even the first iteration could greatly increase the value of the resources that are being shared by increasing their discoverability and making their curation easier.

Importantly, this application of technology would enable us to detect changes in the resources being produced over time. As we roll through successive phases of the crises, needs will change. The people on the ground who are managing their respective institutions’ sites will be the first ones to see the new needs as they arise. If we can watch their responses as they are evolving, then we can galvanize faster and more effective collective action.

By itself, a crawler with some intelligence behind it would be a major improvement over what we have now. It doesn’t supplant the current efforts by human experts. Rather, it augments them. Even better would be to have a collaboration space where some of the people who are creating, curating, or looking for content can collaborate across organizational boundaries. The crawler’s classification algorithm will need to be tuned and retuned over time. To do this, we will need people to correct its mistakes and to enrich the results with information that only the human experts can apply—like, for example, if the resource is grounded in evidence-based practices—in order to make the resource lists maximally useful. So having a large-scale but porous collaboration space, like Slack or Yammer, could complement the machine classification by supporting better human curation across the organizational silos that normally confine this sort of work. At some point, we all need to talk about what we’re bringing to the potluck, what we have too many of, and what we need more of.

Use technologies to help with dissemination

The strategy above would help us develop better resource lists, but that’s only part of the problem. Another part is that most educators struggling to get through this crisis don’t have time to sort through even well-curated lists. Support staff at colleges and universities are doing heroic work organizing content into structures that help faculty walk through course design and delivery processes with step-by-step and just-in-time guides. But they are overwhelmed. Even if all faculty came for help (which they don’t), there simply aren’t enough staff to answer all the questions that aren’t on the generic guides.

We have technology that can help with this too. Once we have the curated and evolving list created with the help of our crawler, we can feed it to a chatbot. Educators could ask natural language questions for exactly what they need and get targeted answers. As with the crawler, a first iteration of a chatbot could be built quickly and then refined iteratively over time. Rather than answering the same questions 100 times over, the human experts could identify the increasingly frequent questions, train the chatbot to recognize and answer them, and then focus their time on answering the tough or unusual questions. A cloud-based chatbot could be embedded anywhere and everywhere that people might have questions in the moment. It could be put in an LMS. Or in a courseware product. Or on a university web site. Or an association web site. It could be in all of those places at once. Since different points of use would likely have different audiences, we can use the information about where the chatbot is embedded to scope the answers and produce more reliably useful ones. If the chatbot is embedded in a chemistry text, then users are more likely to be asking questions about teaching or learning chemistry. If it’s on the site of an association for instructional designers, then we can expect different questions.

With these technologies, we can assemble a supply chain. We can find resources wherever they are, track changes in production of those resources, organize and curate them so that the right resources are more likely to be routed to their point of need, and create a delivery system that is ubiquitous.

When people talk about the Fourth Industrial Revolution—a phrase that I confess I am ambivalent about—this is what that term can mean in the best sense. Machines helping humans to be more fully human. Creating tools that let us apply our expertise and good intentions more effectively, wasting less time on work that could be automated so that we can invest more time in high-value work that requires human judgment. This is not the hard part. There is absolutely no reason I can think of why we can’t or shouldn’t do this.

Facilitating collaboration

But by itself, technology won’t save us. We need to do the work. And we need to do it together. Collaboration is hard in even the best of times, but the fact that it is hard does not make it any less necessary. We need to find lightweight yet effective ways to coordinate our potluck efforts.

I am offering to run a traveling collaboration roadshow. The first organization to schedule such an event is the Apereo Foundation, although I am in conversations with a number of other organizations and expect to have more such announcements soon. In Apereo’s case, my roadshow will be part of their 2020 virtual conference next month. I’ll describe how we intend to structure the interaction, as I expect it to serve as the first iteration of a template for similar events.

I will record a brief video talking about the concept of a resilience network as a hashtag and a potluck. The video will end with two discussion prompts. First, what does your community have to offer to such a network now that it wishes to share more effectively? Second, what could your community create, consistent with its strengths and own immediate needs, that it could contribute to a larger effort? This video will be embedded in a discussion thread before the event. Community members will be invited to comment.

At the two-hour session itself, I will open the conversation with a brief expansion on the idea of a resilience network, including some of the ideas in this post. I will then have a conversation with a community leader, responding to the comments that we gathered in response to the discussion prompt. (In this case, my discussion partner will be Aria Chernik from the Open Source Pedagogy Research + Innovation network (OSPRI), a Red Hat-supported project housed at Duke University, about how institutions can process the lessons learned from the pivot to emergency remote instruction this spring.) The intention here to set the stage for the second round of brainstorming, this time in the synchronous session.

From here, we will go to breakout groups. Participants will add to the original discussion thread as they develop ideas in their breakout groups. We will end the session with a final discussion led by Aria and me to process the ideas.

I hope to conduct a number of these sessions with different organizations, and to encourage each organization to follow up on the brainstorming by identifying the best ideas that their group can follow through on. By the fall, I hope to have had enough of these conversations that representatives from these groups can gather for a potluck planning session. What is everybody cooking up? Are there ways in which we can adjust our plans to increase the complementarity and reduce redundancy in the work that each group is doing?

I don’t know for certain what we should do after that. It will depend on what everyone wants to do and is able to do. The way we will figure out how to better collaborate going forward is by collaborating better now.

It’s on us

I am working to bring about all of the ideas listed above, as well as helping out with some other initiatives that I believe are important to the idea of a resilience network (whether or not they choose to label themselves as such). In many cases, I am doing this work for free. In some cases, I may actually end up contributing money as well as time. I’m doing this not because I believe that the resilience network is my project but because I believe that helping in this time of crisis is my responsibility. If you want to do something that you believe is relevant to #resiliencenetwork and think I can help, then ask me. I will help if I can. Whether or not you need my help, you definitely don’t need my permission. We need a resilience network. Please build one. Do your part. Take the name. Take the ideas. Take whatever you need. Just do it.

I love the people who work in education. I don’t mean that in the casual sense. I love you. The work you do—particularly now—inspires me. It pains me greatly to tell you that what you are doing is not enough. We are not rising to the challenge, collectively, in the way that we need to be. Everybody is working incredibly hard. Everybody is working in good faith. Those are not enough. We also must work effectively. And we must work differently. Change and efficiency are not higher education’s strong suits.

But we are good at learning. It’s time to learn quickly and act quickly.

I know that we can do this. But I don’t know if we will.

Are you with me?