More than 130 peer-reviewed studies have concluded that stadiums do not generate net economic benefits for their host communities.Bradbury, Coates, and Humphreys (2023), in their Journal of Economic Surveys meta-analysis, describe near-universal consensus on the point. Coates and Humphreys’s earlier survey asked, more pointedly, whether economists had reached a conclusion.. The answer, they found, was yes.The literature is right.

It's just answering a different question than the one private capital is asking. The academic finding measures whether a stadium creates net new economic activity for a metropolitan region. The investor question asks whether an integrated owner-developer has the potential to capture attractive risk-adjusted returns on the spending that occurs within their owned campus. Those are two different units of analysis, and the answer to one does not determine the answer to the other.

This is the analytical distinction that separates sports-anchored real estate from every other real estate category and the one most pitch decks skip past. This issue does not. This issue does not.

Any credible investment case for sports-anchored real estate has to start by confronting the most persistent finding in sports economics. Three decades of applied microeconomics, cataloged most recently in Bradbury, Coates, and Humphreys's 2023 survey in the Journal of Economic Surveys, concludes that sports venues do not produce the economic impact their public-finance proponents promise.

The mechanism — the substitution effect — is one of the most durable findings in sports economics: consumer spending at a stadium district is approximately offset by reduced spending at alternative entertainment venues elsewhere in the metro. The net change in regional economic output is roughly zero. A dollar spent at a new ballpark restaurant is, on average, a dollar not spent at a restaurant three miles away.

This literature cannot be dismissed. The academic finding is correct, and we'll go further: it's correct at the unit of analysis it's measuring. That's the key. The error was not in the economists' work. It was in how that work was translated — by stadium advocates, by municipalities, and sometimes by team ownership itself — into the claim that private investors should expect the same dispersed zero-sum dynamic in their own projects.

It does not.

The academic literature answers the public policy question: should taxpayers subsidize stadium construction? The answer, correctly, is no. The investment question is different: when a private owner-developer captures the spending that occurs on their owned campus, does the return clear the market cap rate?

Here is the inversion, stated cleanly. The metropolitan-level framework asks whether a stadium creates net new economic activity for a region. The investor question asks whether an integrated owner-developer has the potential to capture attractive margin on the spending that occurs within their owned campus. The composition of redirected spending is what creates the analytical distance between the two questions.

When consumer entertainment spending is dispersed across independently owned restaurants, bars, and retail scattered throughout a metro, the marginal revenue flows to thousands of operators with no single-entity margin capture. When that same spending occurs within a privately owned, integrated district, the development entity may earn rent, percentage rent, and direct operational margin on a meaningful share of those transactions, across multiple revenue channels within the same physical footprint.

The substitution finding, in other words, is not a disconfirmation of the private-investment case. It is a precondition for understanding it. If metropolitan consumption is approximately zero-sum, then sports-anchored capture is structurally a zero-sum bet against the owners of the displaced alternative venues, and the thesis requires two additional claims:

Bradbury's 2022 synthetic control study of the Battery Atlanta, using county-level sales tax data and a constructed counterfactual, is the most methodologically rigorous test of the sports-anchored thesis at a specific site. The finding: the net increment to county-level sales was small and not statistically significant, with approximately one-third of district sales appearing to derive from displacement of other local activity.

That finding deserves engagement. It is the strongest counter-evidence to our own case, and anyone pitching this category without addressing it is not being straight with their allocators. So here is how we engage it:

Bradbury is measuring county sales tax — a metro-level accounting of whether new net activity arrived. Our thesis measures entity-level capture — whether the developer-owner receives above-market yields on the assets they built. A finding that one-third of district sales are displaced is not a finding that the district fails as an investment. It is a finding that up to two-thirds of district sales may be genuinely new, and that entity-level margin is being captured on the full flow either way.

When you measure entity-level capture directly at a specific project, you can see what the metro-level studies cannot. The Battery Atlanta is the only sports-anchored district with publicly disclosed, SEC-audited financials sufficient for institutional underwriting, and those financials document a single-project data point that reframes the conversation.

Approximately 9 million annual visitors spend money at tenant restaurants, retail, entertainment, and hospitality venues within the owned campus. The Braves' development entity earns rent, percentage rent, and direct operational margin on those transactions. The 2025 reported financials showed the following:

At this specific project, real estate OIBDA exceeded baseball operations OIBDA in both 2024 and 2025. The same consumer spending that produces a zero net increment at the county level produced, at the entity level, real estate profit that exceeded the profit of the anchor team itself. This is a single-project data point — not a category generalization — but it is the only audited evidence we have, and it is what the metro-level framework cannot capture. The analytical distinction expressed in one project's dollars. Same spending. Different unit of analysis. Different measurement.

One important honest caveat on the 2025 headline: approximately $27-30M of the year's $97M total RE revenue reflected the April 2025 Pennant Park acquisition, not organic growth. The cleaner same-store comparison during the Braves' losing season cannot be isolated from publicly disclosed data. The directional reading, that district economics partially decouple from franchise win-loss at sufficient activation density, remains plausible, but the headline growth rate conflates organic performance with inorganic acquisition. We flag it. Next week's issue does the full Battery autopsy.

The entity-level capture argument has a second structural dimension that does not show up in metro-level analysis: how that capture is treated under league revenue-sharing rules. In the NFL, roughly 70% of total league revenue derives from national media contracts shared equally across all 32 teams. Team-level revenues are subject to additional sharing formulas. But real estate development returns from surrounding mixed-use assets sit entirely outside current league revenue-sharing frameworks.

Read the CBAs directly, and the language is unambiguous:

Under current CBA treatment, an owner who develops a mixed-use district is entitled to the associated NOI, appreciation, and refinancing proceeds from those assets, none of which currently flow through league redistribution. Several owners have publicly discussed this structural feature in recent earnings calls. The incentive alignment is straightforward: under current CBA language, development returns from surrounding real estate sit outside the revenue-sharing pool that applies to most other franchise-level income streams.

The exclusion is documented. It is not permanent. Allocators who underwrite this category should assume nothing about the durability of the current exclusion. The NBPA has signaled interest in expanding the definition of shared revenue to capture real estate income and franchise value appreciation. In 2018, then-NBPA Chief Financial Officer Gary Arrick stated on the record that district-level real estate income represents revenue that should be shared with players, arguing for recognition of the players in additional revenue streams, including real estate income and franchise value appreciation.

Revenue definitions have historically expanded in every successive CBA across all three leagues. The NFL's own trajectory traces from Designated Gross Revenue, to Total Revenue, to All Revenue, each cycle capturing more categories than the last. Teams are deliberately structuring developments through separate corporate entities (Atlanta Braves Holdings reports baseball and mixed-use development as distinct segments, for reasons that should now be obvious) to create legal separation between sports operations and real estate. The current exclusion is sturdy. It is not eternal.

Our view: the probability of leagues expanding revenue-sharing definitions to capture district-level income appears low before 2030, when the current NFL CBA expires. After that, it becomes a live underwriting assumption that allocators should treat as an unpriced variable rather than a settled exclusion. Monitor CBA negotiations closely and size accordingly.