J-PAL's response to COVID-19

Many of our affiliates and invited researchers had already heard by March 9th from the J-PAL office they work with on a field project to discuss adjustments to field survey and project activities. Depending on the situation in each country, several offices have already taken steps to limit the exposure of subjects, surveyors, and field staff. For some projects, J-PAL had the week prior halted the start of new data collection or interrupted ongoing field work. For field work that was still continuing—mostly in areas that are not currently strongly affected—J-PAL offices took steps to ensure good hygiene and monitoring. Till now our focus had been to provide broad guidance to regional offices with flexibility to adapt measures to local conditions. 

By March 16th, it was clear that the much faster and wider spread of the virus, and the increased advice of our host universities and local governments necessitated the immediate halt all project activities that create contacts between staff or program implementation staff and human subjects that would not have occurred in the absence of the project. As I mentioned, many project activities have already been curtailed or suspended already. All survey activities at J-PAL offices on remaining projects were halted on Tuesday, March 17th AM (local time zones). Field staff had up to Thursday, March 19th PM (local time zones) to hand in all materials. J-PAL research staff is actively work with primary investigators (PIs) to determine if survey work can be shifted to online or phone surveys. Work that can be done from home by RAs continues.
 
Planned or new projects will continue to stay on hold until further notice. If the project is being implemented by a J-PAL office, the project staff is in touch with PIs to work on how field activities should be adjusted and how they can use phones or online tools. J-PAL’s senior management team is convening every Monday morning EST to review the situation on the ground and the official advice from local health authorities and host universities to update our guidance on a rolling two-week basis. 

One of our top priorities is to try and do the best we can for our 1,200 project surveyors whose dedicated work and effort is such an important part of what we do every day. These surveyors are contract employees hired to work directly on individual field projects. To mitigate the impact of this shut down that is beyond not just our, but anyone’s control, we mobilized additional resources to provide them with six weeks of wages beyond the suspension of field work.

J-PAL manages initiatives (e.g. GI, ATAI, PPE, IGI etc.) wherein we receive an umbrella grant from donors. These grants almost always have a host university PI listed for grant administration purposes in the host university system. These initiatives then call for proposals from J-PAL’s affiliated researchers, and in many cases, the sub-awardee PIs run their field work through non-JPAL organizations (e.g. their employer universities, IPA, CERP, etc.) In many cases, J-PAL is only one of the co-funders of the research. In such cases where the PI is not a researcher at the initiative’s host university (e.g. MIT), the default has been for MIT to cede to the IRB guidance clearances from PI's employer university.

In response to COVID-19, many universities including MIT have asked all researchers that "Human subject research which requires in-person contact needs to be suspended. This applies not only to campus, but also in field work. Virtual interaction is permitted. Human subject research should be limited to analysis of data already collected and researchers must cease work that requires interactions with human subjects.” Thus, any projects for which MIT’s IRB is the IRB of record or any of the co-PIs is a researcher at MIT must halt any project activities that create in-person contacts with human subjects that would not have occurred in the absence of the project. You can read more about this here.

For projects for which the IRB of record is not at MIT, and no MIT faculty or staff are involved as co-investigators, the rules and guidance of the PI’s employer university and the project’s IRB continue to guide their activities rather than being subject to MIT's IRB or rules for MIT's PIs. PI’s are requested to actively obtain guidance from the IRB of record and from their employer university and follow that guidance on whether field work should continue or cease immediately. 

But even in these cases, we strongly urge all researchers to follow J-PAL’s practice and immediately halt all activities that could contribute to the spread of COVID-19. If for some reason field work is continued, we request that the project obtain the advice of a public health expert on mitigating efforts that reduce the risk of infection, such as instructing surveyors on social distancing (e.g. replacing handshakes with other greetings), providing sick leave for surveyors with a fever or cough, halting surveying if the respondent appears sick, etc. 

We requested that the PIs of all projects funded by a J-PAL initiative based at either Global or any of our regional offices report by Friday March 20th on IRB guidance for their project and the actions they have taken in response to the COVID-19 threat. We also requested that all researchers and staff on such projects follow the guidance of their respective governments and employer institutions on travel, working from home, and holding any meetings or events. We have asked our PIs to ensure that all staff on J-PAL funded projects follow this guidance in all project locations. 

We are keenly aware of the tremendous disruptions our affiliates and invited researchers are facing due to closed universities, delayed proposal approvals, additional work needed to transition classes online, research staff who may no longer be available, and field work that stands suspended. In response, we took four steps effective as of March 17th:

• J-PAL initiatives with upcoming spring 2020 proposal deadlines extended their timelines by at least thirty days (see new deadlines below).
• All reporting requirements for ongoing initiative-funded projects due in the next sixty days were extended by an additional sixty days.
• J-PAL will be flexible on requests on no-cost extensions for projects currently in the field
•  We will follow up on donors’ ability to provide cost extensions for these projects as well.

We will keep this page updated should circumstances require pushing RFP due dates back even further.

In response to the COVID-19 pandemic, many J-PAL initiatives opened dedicated off-cycle rounds for proposals from J-PAL affiliates and invited researchers. The initiatives sought to fund rapid, useful, and responsible research that is relevant to COVID-19 response, mitigation, and recovery. Initiatives are accepting proposals through a common application here. Please note that funding is only available for J-PAL affiliates and invited researchers. Examples of projects eligible for funding are for:

• Evaluate interventions aimed at COVID-19 responses or the recovery from the pandemic.
• Add-on modules to existing randomized evaluations to collect data on COVID-19.
• Add a treatment arm to an existing study related to responses to COVID-19.
• Conduct a survey following up on a completed randomized evaluation to measure the impacts of the intervention on interactions with COVID-19 responses.

As of September 21, 2020, J-PAL initiatives have received 59 off-cycle proposals and 5 regular cycle COVID-19-related proposals for a total of $3,038,059. We have funded 34 proposals for a total of $1,546,176. Many more are currently under review. The breakdown of proposals received by initiative is the following:

• CaTCH - 2 proposals received for $39k in funding
• CRRI – 1 proposal received
• CVI - 7 proposals received for $356k in funding; 6 proposals funded for $263k
• DigiFI - 3 proposals received for $489k in funding; 2 proposals funded for $373k
• GI - 1 proposal received for $70k in funding (note: The proposal corresponds to GI’s regular round; GI does not have a COVID off-cycle round)
• IGI - 13 proposals received for $378k in funding requests; 7 proposals funded for $179k 
• JOI - 22 proposals received for $881k in funding; 9 proposals funded for $372k
• PPE – 14 proposals received for $678k in funding; 9 proposals funded for $319k (note: three proposals are from PPE’s regular round)
• SPRI - 2 proposals received for $98k in funding

While our problems are small compared to what the world is facing, we are far from immune to the economic impact of COVID-19. Almost all our field operations have stayed shut worldwide since March and the research pipeline has also been adversely impacted.  All our in-person training courses and our scale-up work that requires in-person contact have been suspended. This has benched dozens of staff worldwide and also threatened the completion of many ongoing field projects (either because the underlying program implementation may not happen again, or the budget may not last a prolonged shut down).

And we have had some significant committed funding being put on hold by donors due to the economic impact of the Great Lockdown on them. This includes funding for scale up and research in many critical sectors in developing countries.

As a result of all this, we have had to freeze staff salaries, delay many long-planned promotions, and leave many key open positions unfilled. We were able to identify emergency funding to keep more than 1,600 J-PAL survey staff on payroll for eight weeks after the shutdown started and project funding was suspended. We also moved many of our projects to phone surveys wherever possible Unfortunately not all projects can do remote surveying and the funding eventually ran out as the lockdowns have dragged on way past 8 weeks. We were therefore forced to lay off more than 1,000 incredibly valuable and talented members of our team worldwide until field research can safely re-start. This is the largest layoff in J-PAL’s history. 

As the full economic effect of the Great Lockdown comes to bear over the next year or longer, we expect more aspects of our organization to be impacted. 

All events, meetings, and visits related to policy outreach, training, and communications also stand suspended for now throughout all J-PAL offices worldwide. This includes standalone activities of J-PAL’s policy and training verticals (e.g. Executive Education in June), as well as on individual research projects (e.g. training of implementation partner staff on a research project, or a matchmaking event committed to as part of a grant). Almost all our donors have written to us offering support and flexibility in managing program deliverables and we are grateful for that.  

In response to the indefinite postponement of our in-person Evaluating Social Programs executive education course in Cambridge, we offered a free webinar series during the originally scheduled course dates. These webinars provided an introduction to why and how randomized evaluations can be used to rigorously measure social impact, and recordings of each session are available here.

We are working to take as many of our trainings and events online as possible. For example, rather than hosting the Innovations in Data Experimentation and Action (IDEA) Initiative Conference that was to take place in April, we will instead relaunch it in fall 2020 as a series of webinars to accompany the release of an upcoming handbook for working with administrative data. Chapters will be presented in a weekly series of webinars open to researchers, data practitioners, and students who work with or are interested in administrative data. We hope to drive an increase in participation and attendance through transitioning these presentations online. 

In response to the postponement of our in-person Evaluating Social Programs Executive Education course and the demand we’ve seen for online training materials, we have moved up the next run of our online Evaluating Social Programs course (J-PAL 101x) to begin on April 28. For the first time, the course was self-paced, allowing learners to complete the course at their own speed.

We have a library of hundreds of randomized evaluations, many of which could provide important insights in the fight to slow COVID-19 or mitigate its impact. Evidence from existing research can help inform preventive health measures and decisions in the interim—for instance, in Peru, J-PAL affiliated researcher Paul Gertler (UC Berkeley) and co-authors tested a large-scale handwashing intervention that introduced an innovative mix of mass media campaigns along with more intensive community activities. They found that the program improved knowledge related to handwashing and modified behaviors. 

We have asked our policy staff to go through our database of studies and policy insights and map them to key challenges being faced in slowing the spread of COVID-19 and mitigating its impact on communities.

For example, our Health sector prepared an evidence note with some general recommendations for responding to the COVID-19 pandemic. The note provides lessons on increasing uptake of healthy behaviors and on improving the delivery of key health products and services. Our Health sector manager, Anupama Dathan, is available for follow-up conversations on incorporating the evidence into policy decisions.

There is much more to be learned from our database of studies. For instance, are there insights from our portfolio of education randomized evaluations about how best to teach children online? Or insights from our research on summer programs on how best to keep teenagers safe when school is off? Or how can governments quickly design targeted cash transfer programs to help poor households in a slumping economy?

In addition, we are committing the time of our specialized sector teams to respond to policymakers, institutions, and governments seeking expert knowledge to develop response strategies, and where appropriate, linking them to academic researchers in our network. Our hope is to operationalize our affiliates’ deep well of relevant experience, knowledge and actionable ideas through these rapid-response partnerships, as well as leverage our extensive library of evidence synthesis. 

As universities around the world shift courses online for the remainder of the semester, we are sharing our insights from developing and running our online MicroMasters in Data, Economics, and Development Policy with instructors and students. We’ve tried different approaches to online education and identified two key lessons that are essential for keeping students engaged—and actually learning. Please read our blog post.

Given the incredible uncertainty of how long the current shutdowns will last, a growing worry that this is the new norm till a vaccine is developed and widely available next year, the need to do urgent anti-COVID-19 research, and to make sure we find opportunities for our survey staff to be safely and productively engaged, it is important that we not wait but innovate on how we can resume our research operations, wherever possible, in a way that is safe for the surveyors, respondents and the communities we live in. 

We are crowdsourcing from our network of researchers and staff, and partner organizations, best practices/tips on switching from in-person to surveying online or via phone (especially in asking surveyors to call from their home rather than a central location) and sharing them in a “living” resource on our website, accessible here. If you have experience conducting surveys online or via phone we request that you please fill out this very short form. From this page, you will be able soon access the shared learning. 

On Tuesday, March 24th, Tavneet Suri (MIT; Scientific Director, J-PAL Africa; Co-Chair J-PAL Agriculture sector), shared insights via webinar on transitioning a large-scale survey on the effects of universal basic income in Kenya from in-person to phone interviews, with a focus on shortening and adapting questionnaires to be conducted via phone.