By Joanna Woronkowicz and Doug Noonan
In the wake of disappearing federal data sets and increasing volatility in public funding, the question of how to preserve and expand access to cultural data has become more urgent. What would it take to build a centralized, national arts and cultural data infrastructure? Is such a thing even possible—or desirable—outside of government?
As long-standing advocates for data-informed policy in the arts, we’ve been grappling with these questions—and the trade-offs, challenges, and opportunities that come with rethinking where, how, and by whom data should be collected, managed, and shared. What follows is a dialogue between the two of us as we try to map the terrain—and maybe even sketch out what comes next.
Doug:
This is a classic public good problem. Centralization brings economies of scale, helps standardize formats, and enables interoperability. If everyone builds their own siloed dataset, we lose coherence—and a lot of value. But centralizing everything, especially within government, also introduces risks. Just look at what’s happening with NCES. In today’s political climate, federal data repositories feel increasingly fragile.
Joanna:
Absolutely. We’ve seen that fragility firsthand. There’s huge turnover for the external team building the NEA’s new data tool — NASERC. Will it even survive next year’s budget? Meanwhile, the data repository NADAC is still federally supported at ICPSR. Yet ICPSR relies on NIH and NSF, and decisions out of DC hold its fate (and half its revenues) in the balance. The infrastructure is vulnerable.
Doug:
But public funding isn’t the only instability. Philanthropy has its own volatility. But private funding offers more flexibility—especially around what kinds of data get collected. Public agencies can’t even ask certain questions. A privately funded system could. I wouldn’t say private funding means “de-politicized” data, but priorities still shift.
Joanna:
Right—and even in ideal scenarios, foundations tend to fund short-term projects, not long-term systems. If a major backer walks away, a data effort can collapse overnight. The real challenge is sustainability, no matter the source.
Doug:
This is why I lean toward “crowdsourcing” the infrastructure and voluntariness with positive ROI. Federal systems rely on passive, administrative data. Fine. But what we need now is a proactive culture of voluntary data sharing—by organizations and individuals. If we want scale and relevance, we need participation.
Joanna:
Participation only happens when contributors see value. Look at DataArts. It wasn’t perfect, but it gave us an example of privatized, centralized data collection at scale. The problem? People contributed because funders required it. When that mandate weakened, so did engagement. It’s not that centralization doesn’t work—it’s that value matters. If contributors get nothing back, they’ll stop.
Doug:
Exactly. That’s why I like the “take a penny, leave a penny” model. You contribute anonymized data, and you get access to comparative insights—benchmarking, trendlines, useful feedback. It’s reciprocal and low-risk.
Joanna:
And it doesn’t have to be government-run. But it does require infrastructure: standards, privacy protocols, plug-and-play tools that integrate with the systems people already use—Tessitura, Submittable, Spektrix. If contribution isn’t near-automatic, it becomes just another burden.
Doug:
The Remuseum project is trying to do this in the museum world—real-time, anonymized data sharing with built-in feedback. It’s promising. But the sector hasn’t fully embraced the idea of data as a shared resource.
Joanna:
Trust is an important piece. Part of the problem is cultural. We’ve conditioned organizations to see data as a compliance tool, not a strategic asset. Changing that mindset is as much about culture as it is about code.
Doug:
And all of this requires funding. I don’t see user fees or subscription models as the solution. Arts organizations are already stretched thin. This has to be collectively supported. We need a culture of evidence more than just evidence of culture.
Joanna:
Yes. Foundations need to stop curating creators and start investing in systems. The last decade has seen a shift toward individual fellowships and project grants. That matters—but it doesn’t scale. If you want lasting impact, fund the scaffolding. Fund the infrastructure that enables learning, adaptation, and collaboration. If you want innovation, fund the systems that make experimentation visible.
Doug:
Especially now. With federal investment in arts data disappearing, there’s an opportunity—not just to fill the gap, but to rethink the whole structure. Better priorities. Better tools. Better access. This is the moment for private funders to step in—not just to rescue, but to reimagine. I’m not talking about ArtsAnalytics.org. We started this project years ago. The immediate crisis and the future solutions must include pieces—like data standards, support for collecting data, an actuall data repository—that this project isn’t trying to provide.
Joanna:
We’ve always worked within the constraints of what the feds had already built. Maybe now we can start fresh—with something more usable and future-proof. Most arts organizations don’t have the capacity to use platforms like NADAC. A new system needs training, lightweight tools, and services that help organizations apply the data meaningfully.
Doug:
Right. Deep microdata is great for researchers, but many users just want a clean chart or a quick insight. Think of the American Community Survey or American Time Use Survey. Their data portals are clunky—but they exist. We can do better.
Joanna:
So maybe the real question isn’t “federal or private?” It’s: what does a durable and functional infrastructure look like—and how do we share responsibility for building and sustaining it?
Doug:
It’s not either/or. It’s both/and. We need the scale of centralization and the flexibility of decentralization. We need standards and participation. Structure and openness. Investment and culture change.
Joanna:
And we need to start now—before the next data system disappears. Before we’re left with nothing but anecdotes.