Data Matters: Investing in Infrastructure to Support Diverse Newsrooms

Photo by Michael Hicks, distributed via Creative Commons

As part of ongoing research for Geraldine R. Dodge Foundation exploring how philanthropy can support diverse, inclusive newsrooms, I visited local newsrooms, interviewed experts and community groups, sat with journalists and listened.

On one of these visits, I was working with an editor who was in the process of recruiting for several open positions. I asked her questions about how the job description was created and particular language choices for the role. Towards the end of our conversation, almost as an afterthought, I asked how the organization tracked recruitment, captured the numbers reflecting the diversity of applicants for these roles and if there were specific initiatives attached to specific goals.

The editor opened a drawer at her desk and pulled out a messy stack of Post-It notes and placed them on the desk.

This is how they were tracking diversity data.

I went through the stack: some Post-Its included brainstormed list of places to do outreach, job boards, listservs and minority journalist associations. Another set of Post-Its included names of prospective recruits, potential applicants, people who maybe knew someone worth recruiting. The Post-Its weren’t attached to any specific goals, not shared with the multiple people apart of the interview process and were kept only on Post-Its. These were resources and good intentions scribbled in sharpie. My eyes grew very wide and concerned.

That is when I realized there is a problem with data culture and diversity in journalism.

In a follow-up conversation months later, the editor had moved on to another role at a different organization. I asked her what happened to the data. “What data?” “The Post-It notes…” “Oh, I threw those out when I cleaned out my desk.”

This is when I realized that there is also a problem with infrastructure.

In my research for the Local News Lab, this was an extreme but illustrative example of many dynamics which emerge in discussions about improving newsroom diversity and community engagement: good intentions coupled with a lack of time, resources, siloed networks, siloed information, a lack of transparency, accountability, strategy that is not attached to metrics all result in the oft-repeated expression of valuing diversity without meaningful measurement to account for those values.

Building diverse organizations is a multifaceted challenge but inequity is also inherently a structural problem. In the absence of a container within an organization to hold a power analysis and examine, measure, interrogate and invest in solutions to address the structural elements of inequity the outcomes risk resembling Sisyphus where the ball that repeatedly crumbles apart represents quality journalism that serves the public.

Addressing Structural Inequities Requires Infrastructure

Photo by Matt Popovich, distributed via Unsplash

The process and continuous practice of building thriving, diverse organizations calls for both committed leadership from the top and meaningful, measureable engagement at every level of the organization. More than 30 years of organizational development research support the positive outcomes of this investment: diverse organizations measurably improve decision making, team morale, problem solving, innovation and resilience. To achieve this, diversity must be made a part of the institutional fabric.

Arlene Notoro Morgan is the Assistant Dean for External Affairs in the School of Media & Communication at Temple University and formerly the Associate Dean at Columbia University School of Journalism. While working at the The Philadelphia Inquirer she led the Pluralism Project, a newsroom-wide diversity program spearheaded by Jim Batten, the late Knight Ridder CEO and implemented by The Philadelphia Inquirer’s then-editor Maxwell King.

The Pluralism Committee, Morgan wrote in Diversity: No Matter What was designed with a mandate to build a model for making “diversity a value for everyone from the newsroom clerk to circulation truck drivers.” To execute on this mandate for the newsroom, “King created a policy that required 50 percent of all future newsroom hires would be journalists of color and women. At the time, minorities made up a little more than 10 percent of the newsroom and women were about 30 percent.” Morgan was tasked with moving these figures forward and she applied data to drive the change efforts which influenced the entire newsroom operation from recruitment to how local reporters practiced community journalism. The Pluralism Committee included individuals from all levels of the newsroom and was the structural container for this long-game. It held a “five-year plan to meet a long list of pluralism plan goals, ranging from examining diversity content of certain topics like fashion, food and obituaries to creating a speciality program to bring young journalists of color into beats like science, sports, business, criticism and photography.”

Leadership invested in a vision for diversity to become woven into the institutional fabric of The Philadelphia Inquirer. In her interview with me for this research, Morgan explained how Pluralism Committee hosted the infrastructure for both measurement and accountability for this vision, “The data came first. Once we started and we did the content audits and the image audits, we could see who is doing what and who wasn’t doing what. We realized how deficient we were in certain areas like business reporting for example…”

The audits reexamined how news was delivered to diverse communities. As an example, Morgan recounted a piece presented and dissected in a Pluralism Committee workshop on a problematic hospital system, “So I said ‘Ok a lot of the patients are Spanish-speaking so how did you get this out to them? Did you translate it? Did you get it into the medical centers? Did you get it into the clinics? What did you do with it’… They hadn’t done anything. So [the team] went back and said ‘We have to reprint this.’ And they did. They reprinted everything in Spanish and they got it out to all the clinics, all the centers, they got it out there.” The audits also reshaped who delivered the news, “[For example], we can look at the newspapers and you can look at the population density that is Spanish-speaking in certain areas and those news organizations don’t even have one Spanish-speaking reporter… who is reflecting what is going on in the community? Who even knows what is going on in the community?”

While application of audience engagement is a still taking shape in journalism field today during Morgan’s tenure at the Inquirer applying data to operationalize coverage across diverse areas was a nascent practice. “In the digital world, if you don’t know who your audience is you’re dead. You’ve got to know who is out there, you’ve got to know who you’re working for. Figure out who is in your community.”

Content audits were so exhaustive they included everything from classifieds to obituaries: “Obituaries really tell you a lot. About who is important in your community and who, basically, the community regards… It was a significant portrait because it told us a lot. And then we took all the information we got and we really created change efforts and goals.”

Those goals and change efforts were also attached to measureable outcomes. Morgan explained, “Managers would be judged each year about how well they implemented goals,” and this influenced department budgets, bonuses and promotions. It was accountable, not punitive in nature and the managers professional development was also folded into the change efforts, “In our Pluralism planning we really created something for senior reporters, most of whom were white. And we got them into sabbaticals… we gave them time off to do a special project, so we didn’t make them feel left out as part of the process. Plus, we got them on to various committees. It seemed to me if you gave people work and you gave them a feeling of ownership into this process, if you brought them into recruiting sessions, if you made them part of the hiring process, they bought into it. It’s when they were left out and made to feel like they weren’t valued that problems emerged.”

Decades of research on building diverse organizations support the process Morgan overviews. As, Frank Dobbin and Alexandra Kalev wrote in Harvard Business Review, “It’s more effective to engage managers in solving the problem, increase their on-the-job contact with female and minority workers, and promote social accountability — the desire to look fair-minded. That’s why interventions such as targeted college recruitment, mentoring programs, self-managed teams, and task forces have boosted diversity in businesses. Some of the most effective solutions aren’t even designed with diversity in mind.”

Morgan explains the intentionality invested in these solutions were effective because the endgame was always front of mind: make strong journalism and serve the public. “Our whole premise on the whole issue of recruiting and content audits was to make the journalism better. And when you approached it that way, everyone understood it. Because it was better journalism, it was accurate journalism, it was more interesting, it was fascinating…We had people who were invested in making a difference.”

“And we lost it all…”

Across the industry, newsroom diversity was directly negatively impacted by the media industry’s business model upheaval, the financial crisis and recession. As the Atlantic reported on the ASNE census findings following the recession, numbers of women and journalists of color dropped and “90 percent of newsroom supervisors from participating news organizations were white.” The programs in newsrooms to recruit, train and retain diverse journalists, initiatives like Pluralism Project which Morgan led, were cut as part of cost-saving measures and with those cuts went the talent, processes, networks and infrastructure.

“So you can see the infrastructure collapsing and you did not have the kind of leadership in place, “ Morgan explained.

The institutional fabric unraveled and survival became priority.

Truth As Product: Prioritizing Data-Informed Newsroom Ecosystems

Photo by Andreanna Moya, distributed via Creative Commons
“…we can begin with a notion of data from empirical science, as a set of measurements extracted from the flux of the real. In themselves, such measurements are abstract, blank, meaningless. Only when organised and contextualised by an observer does this data yield information, a message or meaning.” 
Mitchell Whitelaw writes in Art Against Information: Case Studies in Data Practice

With more than 700 participating organizations, the American Society of News Editors (ASNE) is a journalist membership organization which leads the nation’s largest annual census for journalists of color as a tool to measure its goal of having “the percentage of minorities working in newsrooms nationwide equal to the percentage of minorities in the nation’s population by 2025.”

Since 1997, the organization reportedly followed the practice of disclosing participating individual news organizations in survey findings. This practice of data and outlet disclosure created transparency for participating newsrooms, important information for job-seeking minority journalists and provided local context to an industry issue that is usually examined with top-line national data/figures.

In 2016, ASNE ended the practice of listing individual news organization’s diversity data at the request of member organizations. As reported in Richard Prince’s Journal-ism’s column, ASNE President Pam Fine wrote “We wanted to collect and share as much information as possible. We don’t like agreeing to withhold information but felt the greater good was to use the information from as many outlets as we could.”

It is possible to serve both journalists and news organizations alike, but in ASNE’s decision to not disclose individual newsroom diversity data for the “greater good” it is important to ask whose greater good does that ultimately serve? The number of outlets in the survey increased from the previous year the information was still collected in the survey, just not disclosed. The move signaled to many minority journalists as a lack of transparency, something data collection often solves for, not creates. “Disclosing numbers is such a basic and minimal step that I don’t think any newsroom that chooses not to do so can be really serious about improving diversity,” Lena Groeger stated in a Columbia Journalism Review piece on the news.

Mimi Onuoha is a leading researcher and artist focused on missing datasets which she defines as a “term for the blank spots that exist in spaces that are otherwise data-saturated.” In an interview with me, she unpacks the tension in question “What isn’t counted and why isn’t it counted? And what are the benefits or harms with that?”

“For every dataset where there’s an impetus for someone not to collect, there’s a group of people who would benefit from its presence. More data doesn’t always mean better answers, but in cases where data is used as the end-all tool of proof or a definitive measure for change, then it’s clear that lacking it can be a serious structural disadvantage,” Onuoha writes in Missing Datasets.

In this specific case, collecting but not disclosing the dataset of individual news organizations participating in the census makes the missing dataset designation to apply. If improving newsroom diversity is seemingly the goal most modern news organizations in America share, including the explicit goal of ASNE member organizations, then it is necessary to examine how that goal is measured, the limits of this specific dataset and why there is not presently greater capacity in the space to meaningfully measure and collect newsroom diversity data.

This specific census is an annual snapshot and is only one example of many large, nationally-led surveys measuring various aspects diversity in the media. Other surveys include VIDA Count, examining gender disparity in publishing that recently expanded its survey to include data collection on race and ethnicity, gender, sexual identity and ability, Radio & Television News Directors Association’s (RTNDA) annual survey of women and minorities in broadcast journalism and Media Matters for America’s (MMFA) annual survey on racial and gender diversity on the Sunday political talk shows. These examples approach data collection from a specific focus, community or explicit political point of view and are not driven nor owned by the individual news organizations being collected upon.

As the industry recovers from the recession and continues to remake itself, for the large part infrastructure for initiatives like the Pluralism Project were not replaced while the imminent need for newsrooms to be diverse and effectively cover diverse communities remains more imperative than ever before.

There is a reason why, in the context of local newsrooms, there is not a measurement or metric to account for this disconnect, Onuoha explains “Those who have the resources to collect data, lacks the incentive to.”

Through my research, I unpacked that many local newsrooms maintained a stated goal to improve or “value” newsroom diversity and community engagement but lacked the internal infrastructure and processes to collect and maintain data to support efforts. In practice, the data that is collected is done so individually, in an ad hoc manner that is not folded into a larger strategy or analysis which can be used to make the organization as a whole stronger.

“Think about data collection as a relationship. When you think about just data existing in the wild on its own, then it gets to have this veneer of neutrality. It’s just there. It just exists. It just came about and somehow it’s there. But once you think of it as this outcome of a relationship, it’s a term. The data collection is a transaction and the data results from that relationship, but all of a sudden you were forced to think about the different positionality and the power that’s inherent in those separate places. The thing that really helps with this is remembering that there’s always some entity that decides or wants or has the need to collect something, and there’s another entity that will be the collected upon… the different process of positionality reminds you that the power is just inherent in the whole thing. Data is always the result of the power relationship. Always.”
— Mimi Onuoha

There have been several pieces that account for the substantial process and labor required to build diverse media including:

None of these newsroom leaders above required a census to recognize a disparity and take the initiative to design processes to shift for positive change. Their leadership should be extolled as it models for the industry solutions leadership in what Onuoha describes as “data-saturated” news operations. By investing in infrastructure, newsrooms can sustain these processes, compensate labor, provide material resources and also attach accountability to outcomes. Diversity can be valued by actually being valued.

“A lot of people who work with data, they get their data from APIs, from devices. They never have to think about how it’s collected because you just have the data…because of that, you never have to think really about how it was collected. You don’t have to think about who has collected it. You might in a very shallow way be like, ‘Oh, let me make sure it comes from a source that’s reliable.’ But you never have to understand what was the history around that and what are the other things that are collected that you could be using but you’re not. Because that is seen as the beginning of the process, it cuts off all these other stuff that happened before.”
— Mimi Onuoha

In short, for local newsrooms tackling the challenge of improving diversity the lack of incentive and infrastructure to maintain comprehensive datasets reinforces a continued lack of full transparency on the scope and depth of the problem that individual newsroom leaders are investing substantial effort, time and resources to solve. This dynamic creates a labor vacuum and disconnect both for the effectiveness of building solutions and meaningful community engagement.

In my research interviews, I observed that unless individual champions with budget authority are driving the initiative then the lack of infrastructure creates a vacuum wherein individual staff (primarily women and journalists of color) continue to invest substantial labor toward change efforts that are not structurally sustained beyond them or their individual departments.

Building inclusive, diverse cultures require investment of both time and resources and this is what infrastructure reinforces. Presently, the lack of infrastructure to account for and measure the labor and process required to build diverse, inclusive newsrooms makes it difficult to meaningful invest budget and resources to sustainably support this labor through turnovers, business cycle changes and over the arc of time. With infrastructure, processes become best practices, leaders are developed, goals are made explicit and culture is shifted. These systems and practices spread to other ecosystems and communities as staff inevitably move on in their careers.

This is how infrastructure supports cross-pollination. By investing in the container to process change over the long haul, the news ecosystem is made stronger as a whole.

Moving Forward: Invest in Ecosystem Leadership & Intentionality

There are several options for media funders seeking to invest in infrastructures which support diverse, inclusive news ecosystems and opportunities to explore, including learning from other communities and industries struggling with similar power relationships where Onuoha described, “Those who have the resources to collect data, lacks the incentive to.”

According to Onuoha. “You either have to change it and create an incentive where previously there wasn’t one to collect or you try to give resources to empower others to collect or a third-party with incentive and resource steps up.”

In following up reports, I will share incentive model that can be applied toward how data infrastructure can incentivize stronger community journalism with diverse communities. But for explicit investments in newsroom operations, it is important for funders to recognize that the process for building infrastructure is a labor, especially for systems that will allow for the historically marginalized and underrepresented to benefit from spaces equitably.

An example of this labor, Onuoha describes the following scenario: “Sometimes the act of collection involves more work than the presence of the data and it outweighs the benefit of having the data. Sometimes, this happens with sexual harassment data where one can reasonably say ‘Yeah, it would be great to report this, but do you know what it would be like for me to report this?…’

In the technology industry, another data-saturated environment notoriously missing datasets on racial, gender and ethnic diversity, there are several examples of individuals who empowered themselves to collect data technology companies lacked incentive to collect using open source tools. Tracy Chou, a former Pinterest software engineer, led efforts to challenge technology companies to disclose their diversity numbers as she wrote on Medium:

“The actual numbers I’ve seen and experienced in industry are far lower than anybody is willing to admit. This means nobody is having honest conversations about the issue. While companies do talk about their initiatives to make the work environment more female-friendly, or to encourage more women to go into or stay in computing, there’s no way of judging whether they’re successful or worth mimicking, because there are no success metrics attached to any of them.”

The method Chou used for data collection was transparent, open-sourced and can easily be remodeled for newsroom leaders seeking to hold their own census. Leaders in the space built on this work and followed suit with their own internal census and industry leaders and including using the tools to share resources for professional development. These efforts helped lift up initiatives such as Open Diversity Data that modeled both the practice of public transparency and public gratitude for companies that collect and disclose diversity data. Chiu-Ki Chan and Cate Huston created and open-sourced Technically Speaking, a weekly newsletter that delivers diverse, inclusive speaking and training opportunities for women and underrepresented minorities. Efforts like these interlock and reinforce developing an ecosystem working toward diverse, inclusive communities. Today, Chou has joined Erica Joy Baker, Ellen Pao, bethanye McKinney Blount, Laura I. Gómez, Y-Vonne Hutchinson, Freada Kapor Klein and Susan Wu to create Project Include which describes itself as a “group effort to accelerate diversity and inclusion solutions in the tech industry.” In just two years, technology companies disclosing their own diversity data has become expected common practice. Leadership and infrastructure can engine internal and industry cultural change.

“A community coming together collectively is very powerful.”

Another possible step for media funders seeking to invest in diverse ecosystems is to serve as a community convener, a third-party that can step into the challenge and provide the space to support individual leaders stewarding community and building infrastructure.

For example, Onuoha writing for Quartz explored one case where “Broadway won’t document its dramatic race problem, so a group of actors spent five years quietly gathering this data themselves.” The missing dataset here was justified by the actors union because, “the theater industry does not keep data on the race and ethnicity of Broadway actors. It does, however, collect vast amounts of information about everything else… That hasn’t prevented the League, however, from collecting the same type of demographic information about its audiences.”

Pun Bandhu, the Asian-American actor who gave voice over social media to the racial disparity for Asian-American actors on Broadway “… organized a meeting for Asian-American performers to talk about the state of the industry. The response, once again, was overwhelming. Over one hundred people showed up. What emerged was the creation of the Asian American Performers Action Coalition (AAPAC). It’s primary mission: to track racial demographic data in the industry.” It resulted in the publicly available dataset on the state of racial representation in the industry.

“What AAPAC’s data prove is that there is a difference between declaring the importance of more roles for people of color and ensuring that those performers actually get roles,” Onuoha writes. The statement strikes familiar to those who recognize the difference between news organizations which declare that they value diversity and those which ensure there are mechanisms in place for diversity to be achieved.

This case is illustrative of a dynamic where incentive to create infrastructure is driven by those most impacted and in this space there is a role for media funders to support ecosystem community leadership. Onuoha says, “There are examples of groups who have previously been in that situation where they’re like, “We have the incentive but not the resources.” Often a community coming together collectively is really powerful. When people do that, all of a sudden that changes it…There’s something really powerful about people being like, “No, we’re just going to make this happen.”

Overall, this work requires intentionality and a continuous recognition that diversity, as a practice not a target, means we are continuously doing the work toward the world we want and deserve. Through implementing interlocking pieces and working toward a practice that can be strengthened over time, we are ultimately building shared responsibility and shared leadership for ecosystems that inform and sustain our communities.

Sabrina Hersi Issa serves as a Senior Advisor to the Geraldine R. Dodge Foundation, where she leads Local News Lab research on journalism ecosystems and how philanthropy can support efforts to build diverse, inclusive newsrooms.