By John Lettieri, Adam Ozimek, and Ben Glasner
On Wednesday, researchers Kevin Corinth and Naomi Feldman published an op-ed in the New York Times assessing the impact of the Opportunity Zones (OZ) incentive. Their piece is noteworthy for one simple reason: even longtime critics now acknowledge that OZs have had a significant impact on investment in designated communities.
That shift reveals how mounting evidence continues to undermine early critiques of the policy.
For example, in 2022, Corinth and Feldman themselves had come to a starkly negative conclusion about Opportunity Zones: “Our results suggest that OZs were indeed not effective at spurring new investment.”
By 2024, they had softened to describing their findings as “consistent with OZ investment flowing to places that may have received the investment in the absence of the program.”
Now the authors acknowledge that the OZ incentive has successfully driven investment into low-income communities, but pivot to a new critique: the policy has failed because it is helping poor areas that were already showing signs of progress.
We disagree with Corinth and Feldman’s revised case against OZs. They and other critics continue to discount the scale of the OZ incentive’s effect on designated communities, the importance of housing development in the process of local revitalization, and the need for policies that boost the economic momentum of distressed areas showing signs of growth.
Below, we examine the op-ed’s key claims against the latest evidence.
Corinth and Feldman write: “Our research also suggests a potentially strong impact on investment in multifamily housing, but that doesn’t translate into the kind of broad economic development opportunity zones were meant to foster.”
Corinth and Feldman now acknowledge that Opportunity Zones are having a sizable impact on housing investment. But rather than grappling with how this contradicts common early critiques of the policy, they simply dismiss the role of housing investment in the cycle of local revitalization altogether.
This is especially strange in the context of a national housing shortage.
Housing development is a vital source of employment in construction and the building trades. New housing supply is essential for reducing displacement pressure on vulnerable residents as neighborhoods improve. And housing investment brings with it a myriad of other benefits to local residents and businesses. As we wrote in a recent piece in the Agglomerations newsletter:
On top of all this, new housing brings new people, new amenities, new demand for local business, and new tax revenue to local communities. These changes should be understood — as they are in other contexts — as an important early step in a broader cycle of local revitalization. Allowing for more time to pass for the benefits of greater housing supply to start feeding into other parts of local communities, future research should focus on measuring those eventual downstream effects: poverty reduction, employment, and other indicators of reinvigorated economic vitality.
It’s of course true that local revitalization requires more than just real estate development. But housing is an obvious and important “upstream” ingredient in the broader cycle of neighborhood improvement. Furthermore, the authors fail to mention early evidence that OZ designation in metropolitan areas caused a significant boost in local employment and establishments as well.
The authors even go so far as to claim that the increase in housing supply in Opportunity Zones actually proves that they didn’t need help to begin with, writing that “builders add housing units when a local economy can already support a growing population.” Wouldn’t a similar critique apply to business investment? Businesses, after all, only launch in markets where they assume local demand and labor supply are strong enough to support their operations.
In this way, Corinth and Feldman have tied themselves to an argument that treats the very fact of neighborhood improvement as evidence that the neighborhood didn’t need to be improved in the first place.
Corinth and Feldman write: “There is nothing wrong with making a profit. But we don’t need government handouts in the form of tax breaks for things investors are already lining up to do on their own. These subsidies should be used only when the private return is low but the social return is high.”
Here, Corinth and Feldman appear to contradict their own findings as well as those of other researchers, including EIG.
First, far from simply rewarding activity that would have happened anyway — or, as they put it, “things investors already are lining up to do on their own” — the housing boom in designated communities was caused by the OZ incentive. The authors acknowledge as much in the op-ed.
OZs as a cohort chronically lagged behind the rest of the country in housing growth prior to designation. They are now far ahead of non-OZ communities on this measure, and the change is directly attributable to the policy — not to coincidence.
Our research shows that the OZ incentive nearly doubled the amount of new housing added to designated communities over a five year period. Since 2019, more than 350,000 new units have been added above and beyond what would have happened without the incentive — an effect that continues to grow as of the latest data.[1]
Second, the idea that federal incentives are inappropriate in cases where private capital can achieve a return on investment is fraught with obvious downsides. Taken to its logical conclusion, Corinth and Feldman’s argument would suggest that federal incentives should exclusively encourage unproductive allocations of capital. While this is perhaps an appropriate standard for certain federal grants and tax-advantaged philanthropic dollars, it makes little sense as a general rule for incentives seeking to motivate private capital.
In contrast, the OZ design encourages investors and entrepreneurs to find and create value in places where socioeconomic needs are high and local potential has not been maximized, but it doesn’t restrict qualifying activities to those with zero underlying market logic.
Corinth and Feldman write: “While the opportunity zones that received most of the investment were poor enough to qualify in the first place, they had already been showing signs of economic improvement before the policy took effect… the money disproportionately flowed to places that didn’t really need it.”
The authors’ central critique is that places where the OZ incentive caused investment to flow are largely ones that didn’t merit support.
Why? Because, they write, these areas “qualified as poor but were already growing economically.”
“Qualified as poor” is strange framing, as if these communities were only eligible thanks to some loophole or technicality. Not so. Corinth and Feldman’s own research convincingly illustrates that OZ investment is going to places that are, in fact, overwhelmingly poor and low-income. They also find that OZ investment is heavily targeted to communities with low rates of upward mobility.
The figure below gets to the heart of the matter.
It is worth noting that early critics predicted just the opposite results. They believed that the incentive’s open-ended design would inevitably result in OZ investment clustering in the least distressed areas. Corinth and Feldman’s own research directly refutes this idea:
“For example, a common critique of OZs is that it provides too much discretion to private investors who may simply invest in the least needy areas where the return on investment is expected to be highest (e.g., Theodos et al. 2023; Corinth and Feldman 2024). Across a number of static measures of distress, that does not appear to be the case, with OZs providing a large amount of investment to distressed areas, reflecting the decisions of state governors to select relatively more distressed areas as OZs, and the decisions of investors to allocate substantial investment to selected areas.”
But crucially, even focusing only on the places that were already showing signs of incipient growth, it is clear that the vast majority of OZ investment was still going to genuinely poor places.
In our work, we measure pre-OZ investment using growth in net housing supply. As our data shows, even the OZs with the fastest rates of housing growth before the policy passed suffered from very high poverty rates and very low incomes. Nascent growth and high need are not incompatible, nor is it obvious why their combination should be disqualifying.
It’s worth noting that the longstanding concern — an entirely legitimate one — about place-targeted incentives is that using taxpayer dollars to support investments in poor, unproductive places is often wasteful and fails to move the needle. Corinth and Feldman abandon this concern and, in fact, seem to argue for the opposite approach. For them, place-based policy should avoid distressed places that demonstrate potential for improvement.
We strongly disagree. Regardless, what is not in dispute is that OZ investment is overwhelmingly going to places with high poverty rates and low incomes, and has caused a substantial boost in housing supply in those areas.
Corinth and Feldman write: “…David Wessel in his book “Only the Rich Can Play” documents the opportunity zone designation of neighborhoods in downtown Portland, Ore., in 2018 despite the downtown already being seen as a prime area to build apartments, office buildings and retail stores. Pretty clearly, some investors took advantage of opportunity zone tax breaks for projects that would have moved forward anyway.”
The authors attempt to bolster their case by pointing to an OZ tract in downtown Portland, Oregon, which they claim was “already being seen as a prime area to build apartments, office buildings and retail stores” at the time of designation in 2018. This one area out of thousands of OZ designations nationwide has long been a go-to anecdote for critics because of a mixed-use development planned there that included a Ritz-Carlton hotel (the apparent logic being that poor areas should not be improved with nice hotels).
But far from making their point, the Portland tract vividly illustrates how difficult it is for lawmakers or academics — or even investors — to correctly predict the trajectory of a given high-poverty community. What seemed to some like a slam dunk investment in 2019 has turned into a disaster, as downtown Portland was hit hard by the pandemic and the lingering social and economic disruption that followed. As one outlet put it, “It appears [the developer’s] dream tower will instead be a 460-foot tombstone for his career.”
It’s critical to understand that the design of the OZ incentive reserves the biggest tax benefits for investments that are held for 10 years or more. This ties the interests of investors and developers to the long-term interests of OZ communities in a way that is often overlooked by outside observers. And in the case of failed investments like Portland’s cautionary tale, the most significant OZ tax benefits never materialize.
The Portland example gets to one of the great failures of American policymaking over the past 40 years: High-poverty places rarely experience a dramatic turnaround. Corinth and Feldman acknowledge this fact in their op-ed, but miss its significance in relation to the design and targeting of the OZ incentive.
Genuine, durable neighborhood turnarounds are vanishingly rare. Only 14 percent of census tracts that had a poverty rate of 30 percent or greater in 1980 had reduced their poverty rate to 20 percent or lower nearly 40 years later. In contrast, roughly two-thirds of neighborhoods experiencing concentrated poverty in 1980 remained stuck in the same condition decade after decade.
What should all of this tell us about the targeting of place-based incentives? We think the answer is obvious: since true turnarounds are rare, helping poor places cement their transition from nascent growth to durable prosperity is an important policy goal.
Instead, federal policy interventions aimed at struggling communities often don’t kick in until a place has fallen beyond repair, and they too often fail to distinguish areas that are primed to respond to a market incentive from those that aren’t yet ready. The OZ incentive avoids such pitfalls by relying on broadly distributed market knowledge to identify high-need/high-potential areas rather than the clairvoyance of lawmakers.
Corinth and Feldman write: “And the money has poured in — a reported $89 billion by the end of 2022, over six times the amount invested in the red-tape-heavy New Markets Tax Credit, a smaller federal program established in 2000 with similar goals of spurring investment in low-income areas, over a comparable period.”
The authors correctly point out that the OZ incentive has succeeded in one of its primary goals: scale.
Legacy programs have no doubt done some good for low-income areas, but their reach has been limited by bureaucratic complexity and narrow design. The OZ incentive was conceived as a simple, fast, and large-scale tool that could attract private capital to needy areas in ways that other policies cannot.
Its streamlined, uncapped design has enabled OZs to reach an unprecedented number of low-income communities in a short period of time. Case in point, by the end of 2022, OZ investment had reached over 5,600 census tracts in a span of less than five years, dwarfing the number of communities that saw New Markets Tax Credit investments over a 20 years period.
This, alone, is a remarkable achievement — one that has come at low cost to taxpayers.
Corinth and Feldman acknowledge that “poorer areas tend not to catch up to better-off areas.” This makes it all the more strange that their op-ed is an argument against Opportunity Zones helping communities that have the potential to do just that.
Notes
- Our previous working paper draft, released in March 2025, included the most up-to-date data available at the time, through the third quarter of 2024. Since then, we have obtained data through the first quarter of 2025. In addition to extending the time series, we have improved our methodology by excluding tracts that border Opportunity Zones, accounting for potential spillover effects.[↩]
