Over two decades, the New Markets Tax Credit has deployed $66.6 billion into distressed American communities, financed more than 8,000 businesses, and supported over 850,000 jobs. By any measure, it works.
But then you pull up the investment map. The funded projects don’t scatter evenly across eligible tracts. They cluster along the edges of gentrifying downtowns, near hospital systems, beside university districts. Go deeper into isolated poverty, the rural counties losing population, the urban cores without an anchor institution, and the tribal lands no conventional bank ever reached, and the dots thin out. What the map shows is capital seeking the safest qualifying distress available: tracts distressed enough to qualify but close enough to a transitioning neighborhood, a medical corridor, or municipal money to make a conventional lender comfortable. Through fiscal year 2022, 19 congressional districts across ten states had received zero investment in the program’s history. Native communities have received about 3% of all the capital.
I manage New Markets Tax Credit portfolio compliance at NCIF, a Chicago-based CDFI — which means I’ve spent years watching these deals. I notice which kinds of communities they land in and which kinds of communities they don’t. Managing and underwriting these deals, I’ve learned to read this skewed map as a structural outcome the program’s mechanics produce, not a failure of effort.
The investment map matters even more now because Congress made the credit permanent in 2025. Permanence offers the chance to redesign the mechanics that drain capital from the hardest places, not because it changes market forces, but because it gives the field a stable runway to fix them instead of fighting just to keep the program alive. But permanence won’t redraw the map by itself. Without deliberately changing the mechanics, the geography will stay the same.
Capital finds the “safest distress”
A casual reader sees the clustering on the investment map and assumes proximity to distress is enough. A practitioner sees something more specific: The deals that close are the ones where a conventional lender could still get comfortable with transitional tracts, not terminal ones.
These tracts are distressed on paper, but they are not the hardest places to reach. Because the program makes its intermediaries raise conventional debt for most of each deal, capital drifts toward where conventional lenders still show up, and the map voids out where need runs deepest.
Why the hardest places stay empty
Roughly 43% of US census tracts qualify, but only about 4,300 — fewer than 6% of all tracts, or roughly 14% of eligible ones — had received a dollar through 2020. The credit only fills a gap; it does not replace the conventional financing that makes up most of every deal’s capital stack. Three deal realities keep commercial capital out of the hardest tracts:
Appraisals: Urban deals have comparable sales; a $10 million facility in a mid-size city can support its value. The same building in a rural market might appraise at $4 million for lack of comparables, wrecking the loan-to-value ratio and the senior debt needed to close.
Fixed costs: Closing runs $200,000 or more regardless of size. On a $15 million urban project that’s a line item; on a $4 million rural deal it devours the credit equity, so less reaches the community.
Collateral: Consider Oglala Lakota County in South Dakota, which sits entirely within the Pine Ridge Reservation and is the poorest county in the country, with per capita income around $11,000. A facility there carries the same need as any urban project and none of the conditions a lender wants. On tribal trust land, sovereignty means standard foreclosure doesn’t apply, and a conventional lender financing the leverage loan cannot take the land the way it could fee-simple property. That keeps out the very lenders these deals require. A 2024 Philadelphia Federal Reserve study found that 46% of people in majority–Native American communities live in banking deserts, twelve times the national rate. The 3% of New Markets capital that reaches Native communities is less a scandal than a predictable and fixable result.
Permanence won’t redraw the map by itself
The community development world celebrated permanence, and rightly. Certainty helps attract institutional investors, build intermediary capacity, and sustain the multi-year pipelines these deals require. But certainty alone likely locks in the existing map. The same leverage lenders will favor the same transitional tracts. The same fixed-cost math will sink the same small rural deals. The same collateral problem will deter the same conventional debt on tribal land.
But permanence does change one thing that matters: It gives the field runway to redesign the mechanics instead of just renewing them.
Use the levers that already exist
The fixes don’t require a new program. Three levers already exist — they just need to be pulled harder:
- Make the rural target binding: The program has used a 20% non-metropolitan benchmark since 2008, and the latest round added a 20% rural increase. But a benchmark is not a floor. Turning deep-rural and persistent-poverty-county deployment into a scored, enforced allocation condition would move capital past the transitional tracts that currently absorb it.
- Fund the intermediaries built for the hardest places: The Native American CDFI Assistance program has put more than $220 million into Native lenders that underwrite where mainstream banks won’t. Pairing New Markets allocations with that capacity, plus credit enhancement that substitutes for the foreclosable collateral that trust land can’t offer, attacks the tribal-lands gap at its root.
- Put catalytic capital where conventional debt won’t go: The binding constraint on the deepest-distress deals is the leverage lender, not the tax credit. First-loss or guarantee capital from foundations and impact investors that absorbs the risk a bank won’t price would open markets the credit alone cannot reach. This is the clearest alpha here: mispriced risk in overlooked geographies.
If I had to pull one lever first, it would be pooling the small deals. The roughly $200,000 it takes to close a New Markets transaction barely moves with deal size, so it punishes small and rural projects and rewards large urban ones. Aggregate a dozen sub-$5 million deals under a single standardized closing and a shared legal and accounting structure and that fixed cost spreads across all of them instead of devouring the credit equity of each. The same math that makes a $4 million rural deal uneconomic on its own makes it viable in a pool. This fix is the fastest of the three because it needs no act of Congress, just a community development entity willing to build the structure.
The program has proved it works where capital lands. The permanent era’s real question is whether the mechanics that drew this map are right for the next twenty-five years.
An ask for investors
My call to action for impact investors and catalytic capital providers is concrete: Take the first-loss position on the leverage loan.
A conventional bank won’t lead a deal in Oglala Lakota County because it can’t foreclose on trust land and can’t find comparable sales to support the appraisal. But an impact investor who prices that risk differently and accepts a below-market return on the first-loss tranche in exchange for being the capital that makes the deal possible unlocks the rest of the stack. The tax credit equity follows. The deal closes. The community gets the facility. This is not philanthropic capital filling a gap out of generosity. It is investment capital finding a market the mainstream hasn’t priced correctly yet.
The dots that aren’t on the map are there because no one has been willing to go first, not because the underlying need isn’t real. The deals that won’t pencil out for a bank are exactly the ones catalytic capital exists to make. Go where the dots aren’t.
Jay Patel is an impact investing and community development finance professional at the National
Community Investment Fund, a Chicago-based CDFI. The views expressed are the author’s own and do not represent the views of NCIF or any other organization.
Guest posts on ImpactAlpha represent the opinions of their authors and do not necessarily reflect the views of ImpactAlpha.
