The Rural Mobilization Gap

Intermediary selection vs. market structure in the U.S. New Markets Tax Credit

Author

Dr. Ian Helfrich

Published

May 2026

The U.S. New Markets Tax Credit, enacted in 2000, is the largest place-based tax-credit program in the United States. Over its first two decades it deployed $66.6 billion of federal-credit-subsidized capital into 8,024 projects whose combined cost was $120.9 billion, implying a program-wide mobilization ratio of 0.82 — every federal dollar pulled in $0.82 of additional non-federal capital. This site is the companion to a working paper that asks one thing about that program:

Does the credit mobilize private capital differently in rural versus urban markets — and if so, is the gap a market-structure phenomenon (rural markets are structurally less leverageable) or an intermediary-selection phenomenon (the CDEs allocated to rural deals are systematically less effective at mobilization)?

The headline finding, established by a layered fixed-effects regression on every project the program has ever funded, is that roughly 80% of the aggregate rural mobilization gap is intermediary selection. When we hold the CDE constant — when the same intermediary deploys both metro and non-metro deals in the same year and project type — the rural penalty collapses to statistically zero. The rural gap is real; the market-structure framing for it is mis-specified.

Program years

2001–2022

19,907 QLICIs · 8,024 projects

Total QLICI deployed

$66.6 B

nominal USD, CDE → QALICB

Mobilization ratio
0.82×
private $ per federal $

Non-metro $ share

19.6%

vs. 20.0% statutory

Median leverage, metro

1.19×

Median leverage, non-metro

1.07×

Within-CDE rural penalty
≈ 0
β = −0.05, p = 0.64

% of gap that is selection
~80%
between-CDE component

0.1 How to read this site

The chapters below are sequential but each stands alone. If you have limited time:

For the punchline, read Results and Interpretation — about 12 minutes.
For the empirical strategy in detail, add Empirical strategy — another 15 minutes.
For the explorer’s view of the data, open The map — an interactive globe of every geocoded project, with year-range and metro filters.
For the full paper as a structured argument, read What is NMTC through Where this goes next in order.

0.2 Reproducibility

Every figure and every regression on this site is regenerated by a single command from the SHA-256-anchored raw data:

git clone https://github.com/ihelfrich/us-nmtc-viewer
cd us-nmtc-viewer
pip install -r requirements.txt
python3 scripts/describe_nmtc.py        # raw .xlsx → cleaned CSVs
python3 scripts/make_figures.py         # → 6 PNG figures
python3 scripts/run_regressions.py      # → main_table.csv, bunching, etc.

Total runtime: about two minutes on a laptop. Source files, including the original CDFI Fund release, live at https://github.com/ihelfrich/us-nmtc-viewer.

0.3 Citation

If you cite or reference this work, please use:

Helfrich, I. T. (2026). The Rural Mobilization Gap in U.S. Place- Based Tax Credit: Intermediary Selection vs. Market Structure in the New Markets Tax Credit. Working paper. https://nmtc.ianhelfrich.com