10 Where this paper sits in the literature
The papers below are grouped by relevance to the contribution. The full bibliography is in the References.
10.1 Canonical NMTC empirical literature
Freedman (2012). “Teaching New Markets Old Tricks.” The canonical regression-discontinuity application on the NMTC LIC-eligibility cutoff. Modest positive effects on neighborhood employment and demographics. Sets the methodological template for the planned RDD extension. Does not split metro/non-metro; does not use CDE-level analysis; does not use leverage as an outcome.
Harger and Ross (2016). “Do Capital Tax Incentives Attract New Businesses?” RDD by industry; finds NMTC attracts business activity in some sectors but not others. Demonstrates the LIC RDD is statistically powered. Same caveats as Freedman 2012 with respect to the questions this paper asks.
Theodos et al. (2022). “Place-Based Investment and Neighborhood Change.” The dominant recent NMTC paper. Estimates a tract-level event-study difference-in-differences with tract and year fixed effects only. No CDE FE in any specification; project heterogeneity is absorbed via 13-category project-purpose dummies, not CDE identity. Outcomes are neighborhood-level (jobs, poverty, demographics), not capital-stack.
10.2 CDE-level descriptive precursor
Theodos et al. (2021). “Which Community Development Entities Receive NMTC Funding?” The closest existing work on CDE heterogeneity. A 12- page Urban Institute brief that classifies CDEs into five mutually exclusive types — CDFIs and other mission lenders, for-profit financial institutions, governmental and quasi-governmental entities, for-profit nonfinancial institutions, and nonprofit nonfinancial institutions — and tabulates allocations and project-type associations. Contains zero regressions. From the conclusion (p. 7):
“CDE type also associates with project types, with important correlations among CDE types and certain types of projects.”
This paper’s contribution is the regression extension of that descriptive observation: we absorb the CDE heterogeneity Theodos et al. document as a fixed effect and decompose its contribution to the rural- leverage gap.
10.3 CDFI Fund’s own program evaluation
Abravanel et al. (2013). NMTC Program Evaluation Final Report. The official commissioned program evaluation. Reports descriptive leverage statistics and qualitative case studies of early-cohort projects. Source for descriptive leverage facts in the broader literature.
10.4 Cross-program comparison
Corinth et al. (2024). “The Targeting of Place-Based Policies: NMTC vs. Opportunity Zones.” Pairwise comparison of NMTC and Opportunity Zone targeting — which tracts get which program — at the tract level. Outcome is targeting characteristics, not leverage or mobilization. NMTC vs. OZ only — no LIHTC or IRA. No metro/non-metro interaction. Adjacent territory; we cite as a contemporaneous comparison framework but the present paper’s contribution is non-overlapping.
10.5 Investor-side analysis
Gurley-Calvez et al. (2009). “Do Tax Incentives Affect Investment? An Analysis of the NMTC.” Tax-data panel on investor response. Looks at the QEI-side, not the QLICI-side. Cited as the complement to our project-side analysis.
10.6 Methodological anchors
Chetty et al. (2011). “Adjustment Costs, Firm Responses…” The foundational bunching paper using Danish administrative tax records. Identifies taxable-income elasticities at kink points. Source for the excess-mass estimator we apply to the cross-CDE non-metro share distribution at the 20% mandate.
Kleven and Waseem (2013). “Using Notches to Uncover Optimization Frictions and Structural Elasticities.” The canonical notch-design paper. The 20% NMTC mandate is, in form, a notch — and Kleven & Waseem’s methodology is the right framework for it.
Kleven (2016). “Bunching.” The standard review of bunching estimators across applications.
Imbens and Lemieux (2008). “Regression Discontinuity Designs: A Guide to Practice.” Methodological reference for the LIC-eligibility RDD extension.
10.7 Place-based-policy and LIHTC anchors
Diamond and McQuade (2019). “Who Wants Affordable Housing in Their Backyard? Equilibrium Effects of LIHTC.” The canonical LIHTC paper. Different program, different identification, but the spirit of place-based- subsidy capitalization is the same. Useful as an anchor for the LIHTC literature one might compare to.
Neumark and Simpson (2015). “Place-Based Policies.” The handbook chapter. Foundational survey of the U.S. place-based policy literature, with NMTC, LIHTC, Opportunity Zones, and EZ/EC programs in scope.
Kline and Moretti (2014). “People, Places, and Public Policy.” Contemporary review framing the welfare economics of place-based policy.
Glaeser and Gottlieb (2008). “The Economics of Place-Making Policies.” Foundational essay on place-based policy economics in the Brookings Papers tradition.
10.8 European parallel literature
Becker et al. (2010). “Going NUTS: The Effect of EU Structural Funds on Regional Performance.” Implements an RDD at the 75%-of-EU-average GDP eligibility cutoff for Objective 1 Structural Funds. The European methodological parallel to Freedman (2012).
Becker et al. (2012). “Too Much of a Good Thing? On the Growth Effects of the EU’s Regional Policy.” Nonlinear treatment-intensity extension; finds diminishing returns above ~1.3% of regional GDP. Useful “absorptive capacity” framing for EU rural-finance settings.
10.9 How to summarize the literature in one sentence
The NMTC empirical literature — Freedman, Harger and Ross, Theodos, Abravanel — has overwhelmingly focused on neighborhood outcomes (jobs, demographics, prices), occasionally on investor behavior (Gurley-Calvez et al. 2009), and once descriptively on CDE-level heterogeneity (Theodos et al. 2021). This paper is the first to model project-level leverage as the central outcome variable, and the first to use CDE fixed effects to econometrically decompose the rural mobilization gap into intermediary-selection and within-CDE deployment components.