In the We All Count Data Lifecycle, the first step is Funding.

 

Funding is often overlooked as a factor in data equity because it’s the hardest to change. Data scientists and data project workers often consider how to increase equity within their projects and don’t address the macro-level above them. When we’re working on a project Funding is a lot like the sun to a gardener. It’s one of the most important factors and it’s the one you can’t change. That’s why we’d like to address this introductory checklist to the funders themselves. 

We’ve started a list of the most basic things we’d do if we were funding a data project that we wanted to be fair, equitable, and effective:

 

 

Pre-Project
  • In our competition rubric we would include points for using proper data equity documentation such as a data motivation statement and a data biography. 

 

  • We will consider doing our funding selection process “blind”. We might have the proposals submitted with no identifying information in them as to the individuals or organizations that are submitting. We’re not suggesting this for the entire funding cycle – but maybe for a meaningful portion of it.

 

  • Distribute our calls for proposals more equitably and non-conventionally. Many times these announcements of funding happen entirely within a bubble. People and organizations who are outside that bubble don’t even get a chance to submit or apply.

 

  • In our assessment of risk for your investment, pay careful attention to whose risk we are considering. Whose risk we are prioritizing. What about the risks to vulnerable people of not producing this data rather than the risks to the reputations of the privileged people in producing the data?

 

  • Get the jargon out of our proposal applications. Have someone outside of our sector review your proposal applications. We’ve found this is one of the fastest, cheapest, and easiest steps towards better equity in funding.

Beginning of Project
  • Structure data products that study up. Don’t put most of our investment into collecting data from the most vulnerable. Invest money into studying the most powerful people and what projects might change their lives and systems in a way that would have whatever impact we’re looking for. The ROI is often surprisingly good. 

  • We would require all investments to include outcome, output, and impact measurement that includes equity measures, not just improvement measures or progress measures.

  • And speaking of improvement, we might consider investing in projects that sustain rather than “improve.” The quest for constant improvement is a world view unto itself. 

  • We would set up a structure that allows for participatory funding. We’re talking here about the participation of the communities your investment is intending to help or improve or empower. 

  • We might try to build an investment fund that produces data products from “missing datasets”. Fund the production of data biographies for data you know your sector uses a lot.

  • We’re going to make sure that the timeline of the project is best for getting equitable answers to our question, not just best for our internal timeline. Often it feels like the funders want answers in three months, the researchers want to commit for 3 years, and the participants need the project for the next 30 years. Lack of time is one of the most expressed complaints at the project level.

.

During Project
  • In funding projects we’d include a line item for reimbursement of the people providing data. This is one way to reduce the extractive nature of your investments. Would we take a 3-hour survey for free? Would we provide our data for free with no immediately tangible benefit?

  • Use a data equity steward to advise our fund. This can be a key hire to increase equity and increase the gravitas of our results. 

  • We will establish a payment schedule that supports equity in our project, reducing the financial and psychological burden of the project workers and the project participants by making sure they have the right amount of the funding when they need it. 

  • We will weigh the benefits of different levels of oversight from us as a funder. Will being at arms-length be best for the project, so we don’t interfere and skew the study? Will being involved in how money is allocated or what policies are followed help to structure the project for increased equity? Will being more frequently connected by meetings and updates beyond baseline and endline reports help or hurt the project? Each project is different and each of these management styles has benefits and compromises. 

  • We’re going to be transparent about the funding amount, the funding structure, and the funding source not only with our research team but with the project participants as well. This is a silver bullet for all kinds of equity issues at the participant level. 
End of Project
  • When it comes to the results of a project we’ll consider what we are financially incentivizing. Too often lying is incentivized. Data scientists want to show positive impacts, dramatic results or confirm expectations because that’s the economically rational thing to do. How can we set up a system that rewards transparency and honesty more than just positive results?

  • Consider how the renewal process is impacting the equity of our project. Are compromises being made to secure our next round of funding? 

  • We will have mechanisms for debriefing and feedback that go beyond money and renewal. 

  • We will consider how our own employees are rewarded at the completion of projects including bonuses, title changes etc. and apply it to the people working on our data projects. We will work in a model that aims to retain good employees and build upon institutional success in data science. 

  • We will make how equitable a project was a key metric for success.