An Inclusive Vision for Democratic Technology

While we try to understand the lessons of 2016, it is time to deal with a key missing element in the conversation on the future of Democratic technology: the voices of people who are not straight, white men.


For the last decade, Democrats have touted our major data, digital, and technology advantage. We’re justifiably proud of our voter database and our digital tools. The role of analytics on Obama 2012 has taken on the power of myth, but now there’s a rising post-election narrative that questions the significance of the lauded Democratic data, digital, and tech advantages.

While we try to understand the lessons of 2016, it is time to deal with a key missing element in the conversation on the future of Democratic technology: the voices of people who are not straight, white men.

I’ve worked on, supported, managed, and built much of the most innovative technology that wins Democratic campaigns. And I’ve often been one of the only women and people of color in the room (but not the only).

If you’re reading this, I doubt I need justify the importance of diversity. But what you might not realize is how White our technology, digital, and data efforts are on the Democratic side, and how that affects how we win and who we engage. It is critical for us to examine and understand that we need to change this now, in the present, so that we’re ready to actually win in 2018.

I’m speaking as a Brown woman, but also as an organizer, leader, and technologist. This isn’t just about inclusivity, it’s about winning — and it’s critical for all Democrats to understand that in 2016 you can’t have one without the other. Listening to the voices of all Democrats will enable us to practice better diversity and inclusion, but also allow us to fix bigger problems that hinder our ability to reach our most important voters.

Inclusive Democratic technology needs to have:

  1. A focus on Black, Brown, AAPI, poor, and queer folks. The majority of Democratic fundraising is optimized for older, white donors. Our models are built to turn out “reliable” voters, without acknowledging that what makes a voter reliable is consistent voting history. If communities of color are ignored in the TV buys, digital buys, staff allocations, and office space, it’s not surprising that our voting history is less reliable. The data around race/ethnicity is often straight up garbage— which is how my Filipina staffer got emails from Democratic campaigns asking her to join Latinx for a candidate I’ll decline to name. If you don’t understand why Filipinas might have Latinx last names, you’re going to be missing critical voters. We have to start by building tools for the voters, volunteers, and staff we should be trying to attract, not building tools and then adapting them for those communities.

  2. Mobile is the present, not the future. 36 percent of Latinx households are cell phone only. That number will likely grow, and we’re wholly unprepared for a mobile-only world in electoral organizing. Last year saw more focus on mobile than ever before, but we’ve still got a long way to go. What’s more, mobile phones are used at higher rates for a wider range of people of color than their white counterparts. If a campaign is going to invest in a voter contact program, it needs to start with a mobile strategy, not to tack one on as an afterthought. Our analytics need to understand not just who is on mobile, but what they are consuming, what’s the best device to talk to them on, and how to fully integrate mobile interaction with our organizing and messaging. And our technology needs to be mobile friendly. The private tech sector has responsive design at the core of their strategies, but political tech needs to make a heavy investment in understanding how to make our work mobile-first and responsive, not just making sure you use products pretty well on a phone.

  3. Diversify our leadership. We’re getting better at making the pipeline more colorful and diverse with programs like the Kairos Fellowship, but at the end of the day the thought leaders, directors, CEOs, and decision makers are majority white, usually white men. It’s time for us to promote and cultivate diverse tech leadership, not just junior or mid-level staff. And it’s time for white leaders to fund work with diverse leadership at the helm so that people of color can take time to be on panels, write think pieces, and attend summits that typically continue the cycle of white leadership. And frankly, it’s time for some of those white leaders to make room for diverse leadership in their actions. Introduce us to funders. Champion our work early. Share our pieces. Bring us to summits. Mentor us long term. Give up your spot on the panel and suggest one of us instead. Donate to our ventures. Stop using “the pipeline” as an excuse and see what existing mid-level and senior leaders you could promote or support now.

  4. Grassroots fundraising from people of color. The grassroots fundraising done on political campaigns is exceptional, but our (primarily email) strategies are usually about engaging older, white, rich folks. Asking little old ladies to give more money cannot be our strategy. But building lists that understand race and ethnicity is difficult, often because those teams are dominated by white writers and because campaigns need to raise money fast. But high dollar donors raise money from every type of person, and even successfully mount multicultural fundraising committees, so it’s possible for digital teams to do it as well. I have personally been told things like “I think it’s more important for Latinx folks on our email list to do other things,” which meant that we needed Latinx votes, but their money wasn’t enough for us to worry about. This is a faulty assumption about our community. Our purchasing power is serious, growing 70 percent faster than other Americans since the 80s. We’re here, we’re Latinx, and we’re ready to fork out to candidates who ask us for our money in appeals relevant to us. And we’re not the only ones—just ask our AAPI and Black counterparts.

  5. Trans aware, as a start. The Voter Activation Network, which maintains the voter file for most Democratic campaigns, only has two options for gender: male and female. Sure, adding “other” would be a start, but if we think that literally othering people is sufficient voter engagement, no wonder we’re not engaging these folks. As Democratic technologists we should put our heads together to figure out how to put the gender spectrum in our models and understand how to use that data to do outreach. At the very least, we need to make our products, signup forms, and databases trans aware if we’re going to be the party of LGBTQIA folks.

  6. Deal with and try to understand transience. If you’ve ever knocked on doors or called voters, you’ve probably experienced outdated voter data. Call sheets of older voters can sometimes feel like a funeral march of people who died years ago—not even within the last year. Knock on doors in a place like Nevada or Nashua, NH, and you’ll find people who moved 5 or 10 years ago. That kind of error isn’t just about cleaning up the voter file, it’s bad data. Economic insecurity is real for our voters, and that means people move. Our smartest minds should be thinking about how to deal with transience in our community, how to make it easier to update the voter file to understand who is moving and when, and how to harness the fact that your cell phone number moves with you across voter registration lines.

  7. Integrate social into our data. We’ve barely scratched the surface on how to integrate a person’s social media profile into our data, and our transactional relationship with people’s data rarely delves into the listening you can do with social conversations. The next two election cycles are going to be about how we’re not only driving message on social, but understand who is consuming those messages, how they are driving them, and the network effects of their shares. This isn’t an easy problem, but we’re not even attempting to solve it! A first step would be understanding how to systematically collect social media profile data from supporters that we can access in key moments. At Hillary for America, we were successful in getting data through our mobile app, our email list, and more. Obama 2012 was able to access data through the Facebook Open Graph, now sadly defunct. It’s up to practitioners to solve this problem outside of platforms.

  8. Technology power is people power. It’s time to push this power down into the hands of organizers and volunteers. At the core of our work is voter contact, and the people who actually talk to the voters are the ones who need to flex their digital chops and hone in on key data points. The point of building all of this infrastructure shouldn’t be to put it in a black box where elite (mostly white) data scientists and software engineers and digital strategists have the magic to win. It’s not enough to train organizers how to use social media or pull lists without training everyone in collective decisions at the volunteer or organizer level. It’s on us to train the next generation of organizers to be smart local strategists by building the tools to help make decisions and training them to develop that strategy.

  9. Build a working feedback loop. Accountability for data/digital/technology teams is imperative to enable us to understand what folks are really experiencing on the ground. Anecdotal feedback has its place — but it can’t govern large, national decision-making. Similarly, there is no substitute for being on the ground and actually organizing with the model, or the talking points, or the software. Our feedback loop is broken. If a volunteer comes into an office and tells us her lived experience, such as her Trump voting neighbor is regularly on call lists, we need to have a rubric for when one volunteer’s anecdote becomes an avalanche of anecdata when volunteers in states all across the country are saying the same things. Understanding how best to utilize anecdotal data would be a big first step in knitting together national and local strategy.

Whether micro or macro, it’s true that the Democratic Party has work to do at every level to pace our data, digital, and technology to the challenges we face. But we need to internalize the lesson of 2016 that we’re glossing over, in big ways and small ways: when the Democratic Party doesn’t focus on Black, Brown, AAPI, poor or LGBTQIA voters, we’re leaving our voters behind. So much has been said about the need to focus more on white working class voters, but the present and future of our party is the exact opposite. Engagement with marginalized communities isn’t just about visibility, but about the strategy, tools, and resources that we use to win. It means we need to elevate leadership who have the lived experience to make those decisions.It means we need to build with, not for. And it means that we need to start now—2018 is just around the corner.

  • Patrick Atwater

    I would add that there’s also a key need to better deliver public services for everyone. It’s on those of us that believe that honest and effective government matters and that the digital revolution offers tremendous potential to improve how government actually operates need to show in a way that’s obvious to people that don’t geek out on this on Civic Hall (aka the vast, vast majority of marginalized communities and those with very real needs for quality public services).

  • Mariana Ruiz

    I couldn’t agree with you more Jess which is why the Kairos Fellowship doesn’t just believe in building a pipeline of junior leaders of color but we’re also equally focused on tearing down the barriers to leadership within digital and challenging the underlying assumptions of the dominant digital model which privileges white cis gendered men.

  • 1. IP LAW AND CAMPAIGN FINANCE LAW. The 2008 Obama for President campaign collected fresh data and built state-of-the-art software. But laws governing the value of that intellectual property mean that the valuable data and code never becomes a living asset of the Democratic party after the election.

    2. DNC BOOM AND BUST OPERATIONS. Even if that transition of talent and IP to the DNC was allowed or had happened, the DNC would have had to continuously fund its large ongoing budget between elections, and commit to massively improving and extending the data and software.

    3. DATA TREATED LIKE A HOSTAGE INSTEAD OF A SHARED ASSET. Another barrier is that the DNC and state parties expects local candidates and campaigns to pay for access to the latest data/code. So candidates have every reason both to shop around and to not update regional or central databases. Built in disincentives make our data worse, create barriers to entry for new participants, and make it harder to win elections.

    4. DATA/APPS DON’T SUPPORT THE ROOTS BETWEEN ELECTIONS. There is no support from the DNC or state parties for helping very local Democratic clubs, discussion meetups, or ally groups. The same data used to prepare for our next GOTV could help build the Drinking Liberally community or a VFW post to pull together to help local vets. Turning the party database into a tool for political organizing, community organizing, and rapid team building is part of what keeps a political party stay connected and relevant between elections.