The OpenAQ Ethos
This post is written by Christa Hasenkopf, co-founder and Director of OpenAQ.
OpenAQ has evolved over time. At first, it was a tiny project Joe Flasher and I started in our living room because we were frustrated that no entity was surfacing — or could be persuaded to surface — the primary data layer of global, timely government air quality information at a station-level and fine temporal scale in a transparent and open/free manner.
As the data and the open-source project grew, they attracted a larger community to help build and build on top of the platform. This let us get a glimpse of the power of open data to convene disparate sectors and geographies around a common purpose. To channel that convening power, we developed online portals to connect with this forming community and to connect them with each other. Eventually, this morphed into holding in-person workshops that pulled disparate air quality actors in local communities together via the open data platform, but it often resulted in other powerful collaborations and projects beyond the data itself. And the workshops provided a pipeline of awesome local air quality actors and examples to connect with our existing global community.
Now, we are entering a new phase that still involves open data and community-building, and also: working with organizations to find and share more meta-data, helping groups convene communities for workshops, and aiding governments that wish to share their air quality data openly. Importantly, it also means turning our organization into a 501(c)(3) non-profit.
Looking at where we are now and where we started from, I see an implicit ethos behind our work. As we begin a transition to this next phase, it’s worthwhile articulating this OpenAQ ethos in the following points:
1. Money and staff-size do not equal impact
I don’t believe our impact has or will track with staff size or annual budget. In fact, I would argue that the smaller we can be and achieve our mission, the more awesome (e.g. efficient and sustainable) we are.
In reality, we fill a very small and specific gap in the air inequality community — and this enables a much larger ecosystem of sectors and geographies to do the actual important and impactful policy, science, media, and other environmental and public health work. Filling this small and specific gap to date has been pretty cheap — and more than anything, the work has relied on individuals’ donations of skills and time, rather than their pocketbooks. In fact, I sincerely doubt we would have gotten nearly as far if we had just gotten a large pile of money to build the project and no community help.
For a very long time and for several reasons, a non-profit model appealed much more to me than a for profit model for organizationally structuring OpenAQ. Our mission is entirely non-profit oriented, and I think many entities we wish to engage with will be more at ease working with us a full non-profit. Most for profit business models I can think of also center on one particular product built off of an open data system, and this seems like it could be distracting to the main mission of opening up air quality data and fostering a community around it. That said, I did have reservations about what I understand to be typical non-profit models (in addition to also worrying about the legal and financial logistics of operating a small-scale non-profit).
For instance, approximately 70% of non-profits in the US require board members to provide yearly monetary donations to the organization they serve. But from what I’ve observed, it seems like it can (a) lead to a board that is brought on more for its members’ deep pockets and financial connections than their actual qualifications to successfully set the overall strategy of the organization, and (b) push the organization to a larger and larger money-generating board that requires even more infrastructure and time to support it rather than the mission itself. And regardless of the board, the more folks you bring on to the organization, in general, the faster the funding-hamster wheel spins for salaries, office space, and typical non-profit fundraising events. This just doesn’t seem to jive with the same ethos that helped us build a grassroots, global open-source open data community.
So my thought is: If we are an organization that may be able to achieve our mission by running leanly — relying on a working board with awesome skill-sets, committed to our mission, keeping our staff super low, not having an office, keeping our mission simple and focused, yet still impactful — why would we not try that first? It’s probably a whole heck of a lot more sustainable in the long run. And I think personally way more fun, too.
2. Support the community humbly
We play the role of cheerleader for groups fighting air inequality in their communities. We help enable activities and amplify the voices of those doing them, but we, at the end of the day, aren’t the one doing the hard work of measuring the data nor acting on it. Humbly approaching this work means doing it with appropriate attribution, transparency, and in collaboration with the community — and recognizing that what we’re doing is likely still pretty imperfect and a work in progress.
3. Done is better than perfect
Speaking of imperfect, many open data projects (including ours) have unclear boundaries of what data-types they could ingest versus what they don’t. Also, these datasets often don’t have well-defined (or any) guidelines or standards — in fact, the messier the data landscape, the more interesting and useful it can be for a group to wrangle the data in the first place. And you can spend forever and a day trying to design the perfect system and architecture. But we believe done is better than perfect. This is why we didn’t wait to launch our project until we had air quality ground monitoring data from all available countries — we started with three and grew as we could. This is why we didn’t wait to aggregate all possible data types (e.g. large volumes of low-cost sensor data, accompanying station weather data, or a lot of meta data that is out of reach). And that is why by not waiting to find massive funding or crafting schematics at our best guess of the perfect system, we’ve been able to build the only system available to date that lets you access open station-level air quality from 62 countries and counting. We’ve been able to get closer to perfect by putting something out there and getting community feedback about it as we go.
4. Build a system that improves other systems
We don’t want to aggregate and share data as an end in itself. Rather, we want to help organizations — especially governments — share their data better or at all. Our community has done this through contacting their environmental agencies to improve their APIs. Or from community members noticing errors — and eventually helping correct the way real-time air quality data is reported from a government agency. We think this is awesome because it improves air quality access to anyone utilizing that primary data source, not just those who access it through our system.
5. Don’t fall too much in love with fancy metrics
I’m a scientist. I like charts, graphs, and trend-lines. And there’s a certain sense of comfort in building a strategy that lets you systematically tick off quantitative milestones toward success. But real impact and influence often don’t work that way — we just approximate them as best we can with these artificial frameworks. For instance, I often think of that lone air quality monitor that the U.S. Embassy launched in Beijing several years ago. There’s a pretty strong narrative that connects the dots from that monitor to year-on-year improvements in annual-averaged air quality levels in Beijing since 2014. But you can’t pin that story down with a quantitative analysis showing how that monitor directly influenced the media, activist, science, and policy communities in Beijing, across China, and across the world. And if that monitor had not been launched or did not stay put because one couldn’t quantitatively define its short-term success, we would all be the worse off for it.
This is not to say there is no benefit to tracking success through setting long-term goals, defined with quantitative metrics. For instance, I certainly do look at our platform’s usage, its data coverage, website analytics, our workshop attendance, etc. But I recognize these quantitative metrics are poor proxies for the true impact of our system and community, especially in the long-term.
Instead, I find deep faith in our effort from talking to researchers who share the new types of public health research they can do with access to real-time data, the government workers who explain they are so happy that they can finally access their own government’s air quality data through our system, or our workshop participants who describe later banding together to lobby their parliament for change or seeking to conduct a policy-relevant field study. And going back to numbers: If we are able to influence one national-level policy or spur one large-cohort, long-term epidemiological study that ultimately impacts a country’s air quality standards, we are doing good enough in my book.
6. One tutorial is worth a thousand thought leaders
Joe said this to me recently. I love this phrase. Our community is awesome because they do stuff. If someone mentions, “I’d really like to see data from Country X,” that’s nice to hear. But it’s even better to point our community to a website where that data is shared. Best of all? Actually writing an adapter to ingest that data. Recently, a community member contacted me to ask if there was an open-source wrapper available that would calculate air quality index values from the raw data in our system. I said no, not yet, but it was on our Community Wishlist. He responded, “Actually, I’ll build that.” Now, that is how a grassroots, disparate open-source group gets the job done.
7. Assume the best of others
Community-building and data sharing of any sort revolve around trust. I have noticed there are usually two things lacking when groups don’t trust one another: assumed mutual good intentions and mutual respect. If those things can be remedied, so many other logistical or social obstacles can be overcome. It may sound simplistic, but the number one way to help instigate mutual assumed good intentions and respect between two groups — or between your group and another — is to be the first to articulate clearly that you assume their good intentions and to state your respect for those groups. It has helped us do concrete things like ingest more — and more accurate — data into our system, create better outcomes for our workshops, and given us opportunities to share our views on best open data practices with wider policy-relevant audiences.
Frankly the vast majority of people working on air quality issues — from government workers to activists to researchers to journalists — truly work to have a positive impact within their means and mandates. And at the same time, a lot of folks in this space probably don’t get many accolades or much respect for it, and some may even feel attacked in their community for their efforts. Extending some good will goes a long way toward building the trust necessary to share information and form collaborations.
8. Exist only for as long as needed
Don’t get me wrong: I adore OpenAQ. But as it stands today, I would consider it a huge win if a university, non-governmental organization, or other long-standing entity took on the project and committed to its long-term existence. It’s probably a bit more aspirational, but it’d also be wonderful if we became obsolete because countries adopted a standard real-time data sharing protocol and data-sharing mechanism. And if some organization pops up, doing identical work in a manner better than us, well, then, we should have the humility and grace to bow out. There is so much work that remains to be done in the air inequality space across the world — we’ll surely find something else worthwhile to do.
The precise shape of OpenAQ’s work will evolve over time — and likely even this ethos will, too. But we believe it is important to expressly articulate the ethos under which we aim to achieve our mission to help others fight air inequality.