In this post, we reflect on our recent work with PFC, CFC, EFC and The Circle, to build PhilanthropyResponds.ca, a data sharing initiative meant to increase transparency in the sector, and to shine a light on how philanthropic organizations in Canada have responded to the COVID-19 pandemic.
By April 2020 it had become increasingly obvious that the COVID-19 pandemic was going to bring about significant disruption to the daily functioning of Canadian society. As the pandemic increased the need for the services of charities and nonprofits, the sector also experienced a reduction in fundraising revenue, with 1 in 5 nonprofits in Ontario alone closing due to a lack of funding by April.
To help support and sustain the ongoing work of nonprofits, the Canadian philanthropic sector immediately began to mobilize new and existing resources to help nonprofits continue to provide urgent community support. Individual foundations began creating COVID-19 emergency response funds, removing restrictions on existing funding, and innovating their grantmaking practices to better support grantee partners.
While some of these actions were publicly announced, and others were shared directly between peer funders, it was difficult to get a holistic perspective on how Canadian philanthropy was responding to the pandemic. In a sector that traditionally relies on in-person events and word-of-mouth networks which were suddenly dispersed (and in many cases frantically shifting their activities to virtual environments), getting a sense of how the sector as a whole was adapting to emerging challenges was particularly difficult.
Philanthropic Foundations Canada (PFC), Community Foundations of Canada (CFC), The Circle and Enivronment Funders Canada (EFC) quickly identified the need to build a holistic picture of philanthropy’s response to the pandemic, and to reflect the overall response and emerging best practices back to the sector, and invited Grantbook to join the project as a technical partner.
We were inspired by efforts to monitor the response of philanthropy around the world (Candid in the USA, 360Giving in the UK, Italian Nonprofit Network in Italy, and others), but as we started to envision a solution that would address gaps in the data, it became clear that Canada lacks the data sharing infrastructure that those organizations spent years building. To make the best use of our limited resources, we adopted an iterative, agile approach, in order to quickly create something that would support the sector in Canada.
We chose to focus on collecting high-level data directly from funders through a survey. To reflect the findings of the survey back to the sector, we spun up a simple website to host data visualizations and other resources. Where possible, we repurposed or extended the partners’ existing technology stack to keep costs to a minimum.
We quickly cycled through the following steps:
- We started by focusing on rapidly developing and iterating on a survey that would capture the most information while requiring the least amount of effort from funder respondents. To this end, we provided a short bilingual survey which used multiple-choice questions, wherever possible.
- To remove barriers to response, we gave respondents transparency and control over how their responses would be displayed and further shared (e.g anonymously or named).
- To ensure the data was re-usable, we made use of existing standards and definitions (for example, Imagine Canada and Ontario Trillium Fund definitions).
- Given the rapidly changing nature of the pandemic response, we made the survey updatable so that users could submit a response that reflected their current state, and also return to update their data as their response evolved.
- We piloted the survey in both English and French with key funder partners, and subsequently held a focus group to test the survey for usability and understandability.
- Taking this feedback into account, we launched the survey through partner networks, garnering responses from dozens of funders.
- As the pandemic progressed, the Government of Canada announced the Emergency Community Support Fund (ECSF), which CFC along with United Way Centraide Canada and Canadian Red Cross administered the fund. Given their role in the project, we were rapidly able to map this data to our data model and integrate it into the data collection process.
With this high-level data, we sought a quick and flexible way to present the aggregated responses back to the sector in a useful way—building a new bilingual website, PhilanthropyResponds.ca, through which we served live interactive data visualizations, updated in real-time based on survey results to tell the story of the response to the pandemic.
Alongside the live visualizations, we provided additional research and analysis of the survey carried out by PFC’s dedicated research team. We also made anonymized survey responses available as open data to enable others to reuse the data we collected to find their own insights. We sought to enable peer learning by amplifying the learnings that funders were communicating via the survey, and to enable collaboration between funders by providing a Collaborator Hub, where foundations (who chose to) were listed by their funding priorities.
To encourage funders to continue submitting their data as the response to the pandemic progresses, we have shortened the initial survey and included questions about later stage responses. We will continue to make the visualizations, research, and underlying data available to the sector and future researchers.
Launching a new, reactive collaborative project in the midst of a global pandemic was always going to be a challenge. Like everyone else, we spent time negotiating the complexities of our changing work environment and were constantly confronting the unknowable unknowns. In the process, we learned a lot about the challenges and benefits of collecting and sharing data across the Canadian philanthropic sector.
At the core of each of these lessons is recognition that approaches to data collection and sharing should be fundamentally human-centric. This manifests itself in the notion of respecting the time and expertise of those providing, described by, or affected by the data being collected and shared.
- Make it as easy as possible for data providers to share their data. You need to ensure that data collection does not place an unnecessary burden on the data provider, or they will not participate. While our initial survey allowed us to collect useful high-level data, it was very time consuming for users and did not provide detailed information on individual grants. Reusing data that has already been collected for other purposes (i.e. including grants administration data, like that provided by ECSF), would be more granular and less time-consuming for the data providers. However, to effectively aggregate all this data from across different funders would require them to be collected according to some common standards, which would require a lot of upfront coordination and effort by funders.
- Leverage existing networks and work, as much as possible. When spinning up a project so quickly, we benefited greatly from access to existing networks and communities. Being able to reach out to PFC’s member network allowed us to quickly engage with funders to collect feedback and iterate on our survey design. We were also able to generate a far greater response when promoting the survey and website through our partners' various networks. We also made use of existing efforts to define data categories, wherever possible. This allowed us to work quickly and enabled future data mashups, including helping to integrate the ECSF data.
- Move from community consultation to community engagement and ownership. Time and again it is clear that top-down imposed collaboration and data sharing efforts are not effective. While there is a need for centralized coordination, the ultimate goal of data sharing initiatives should be to decentralize power to the community providing and affected by the data being shared. Through this project, we attempted to consult with funders as much as possible, as well as give them granular control over how the data they submitted would be used and shared. While we did take a consultative approach, we could have enabled more structured feedback points and would have benefited from the input of additional voices outside of our core networks. Future data sharing projects should aim to move towards building greater community engagement and aim to give that community a more active role in directing the project and its evolution.
- Beware of the implications of quick decision making in creating additional work and constraining future decisions. All technology decisions are strategic choices, however, when faced with a crisis it is easy to make quick decisions about technology without full consideration. Within this project, we rapidly evolved our technology stack using the tools available to partner organizations that didn’t require significant additional investment. Because of the iterative approach we took and the low investment implications, we were able to make and reverse technology decisions relatively quickly. However, in some cases our experimentation with different technologies created undue complexity which eventually required additional work to resolve (described as accruing “technical debt”) and unnecessarily limited our future approach (leading to “path dependence”). While experimentation can be a useful part of agile technology projects, it is important to build interoperable and well-documented systems that allow for rapidly switching out technologies while maintaining functionality.
- Maximize reuse and minimize barriers to contribution by being open. An important part of any sector-wide data sharing initiative is to maximize the potential benefits. In this project, we released aggregated anonymized data as open data that anyone could reuse for any purpose. While we have specific use cases for the data being collected—providing easy to understand visualizations and monthly analysis—it is possible that the data could also have many other uses (for example in academic research). By giving others the opportunity to build on the data collected, we minimize the need for others to replicate the work we have carried out. Any future projects should also make use of open community-defined, technology-agnostic data standards, and open source tools to help minimize the costs of participation. Any such projects should work actively to promote these openly licensed outputs to all members of the community, not just expect them to actively seek them out. As part of this, projects should actively work to document decisions and outputs transparently by working in the open.
In a short period of time and with relatively few resources, the PhilanthropyResponds.ca project helped reflect the approaches taken by Canadian philanthropy in responding to the pandemic back to the sector. By holding up this mirror, we hope to have enabled better decision-making and helped funders contextualize their role within the sector’s response.
The pandemic has increasingly exposed an underlying need for better data sharing between funders in Canada that can be mobilized, not only in response to global crises, but in response to the many issues philanthropy aims to tackle. During our discussions for this project, we repeatedly talked about the need to share individual grant level data, in order to allow funders to collaborate more effectively, and ensure that there are no significant gaps in funding. Working out how to share this data would enable funders to understand where funding is going and where there are gaps, bringing significant benefits for coordination within the sector and thus to nonprofits and the communities they serve.
We are not the first to identify this need and we are not going to be able to build a solution overnight. To build the robust infrastructure required to meaningfully track the important work that philanthropy does will require coordination, standardization, and upskilling. It will require a centralized coordinator to bring the various different actors in the sector together, not only funders but technology providers and researchers. It will require the development of community-defined open data standards for sharing data. It will need to build data skills and data culture within funders from across the spectrum of Canadian philanthropy. And all this work will require a significant level of investment and involvement from within the sector.
While this may initially seem overwhelming, the COVID-19 pandemic has created the impetus to solve these challenges. Additionally, Canadian philanthropy is well placed to benefit from the lessons learned from existing emerging international initiatives (e.g. Candid and 360Giving).
In many areas, the pandemic has revealed and intensified existing challenges. As the philanthropic sector continues to respond to this crisis, and to support the work of charities and nonprofits, we’ll need to continue to address the underlying issues. We know that better access to data will help the sector become more resilient, to collaborate more effectively, and to make data-informed decisions about funding priorities. Continuing this work will require convening a diverse set of funders, tech providers and researchers to identify immediate priorities. It will also require identifying and committing resources to fund initial research and requirements gathering exercises under the direction of this group.
Canadian grantmakers—submit your data to PhilanthropyResponds.ca to help the sector better understand how our crisis response is unfolding.