![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
Native communities in the United States have been persistently underserved by traditional financial institutions. To fill this gap, in recent years both Native and Non-Native Community Development Financial Institutions (CDFIs) have emerged on or near American Indian reservations. Yet to date, no comprehensive evidence exists on the effect of CDFIs on credit outcomes in Indian Country. We combine a large-scale dataset on individual-level credit bureau records with Census block group-level measures of CDFI activity to explore how the presence of Native and Non Native CDFIs affects Indian Country residents’ credit risk scores, a key summary measure of individuals’ credit performance and creditworthiness. Using multiple empirical approaches and addressing endogeneity concerns, we uncover a positive association between Native CDFI presence and credit risk scores for the subsample of individuals initially deemed least creditworthy. We do not find consistent evidence of a similar effect due to Non-Native CDFIs. To the extent that CDFIs have indeed causally impacted individuals’ credit risk scores, our findings indicate that the corresponding improvements have arisen primarily because of Native, rather than Non-Native, CDFI activity. Our analysis thereby offers the first systematic empirical evidence suggestive of the importance of a cultural fit for credit outcomes in Indian Country.
Notes
1 Jorgensen (Citation2016, 19) points out that “no comprehensive research shows the Native CDFI field’s true economic and social effects.”
2 An alternative definition of Native CDFIs, used by First Nations, is that “75% of recipients of financial, technical assistance, training are Native plus one of the following: organization wholly owned or controlled by the tribe; subsidiary of a tribally owned corporation; owned by enrolled members of a Native group; founders are Native/tribal admin/government; at least 75% of managing board of directors is Native or permanent resident of reservation; at least 60% of staff is Native or permanent resident of the reservation” (First Nations 2007, 135). Since we use data from the CDFI Fund, for the purposes of our empirical analysis, we use the CDFI Fund definition of Native CDFIs.
3 The Hopi Credit Association is even older, but obtained CDFI Fund certification later than the Lakota Fund (Jorgensen Citation2016, 13–14).
4 To obtain CDFI certification, “Native Community-serving organizations must show that they are a legal entity at the time of application, have a primary mission of promoting Native Community development, are a financing entity (lender), primarily serve the Native Community market, provide development services as well as financial services, and are accountable to their target market(s)” (Jorgensen Citation2016, 14fn11)
5 There exist alternative reports concerning the exact number of certified Native CDFIs in 2001; see Jorgensen (Citation2016, 14fn10). Regardless of the precise count, however, “there has been tremendous sector growth since 2001” (ibid.).
7 For instance, “when the concept of savings is introduced, the discussion could begin with questions concerning the local historical patterns of storing and saving of foods and resources that would sustain tribal life during the winter season” (First Nations Development Institute Citation2007, 12).
8 Over the course of its existence, the Lakota Fund changes its name into Lakota Funds.
9 For quantitative information on the Four Bands referred to in this section, see Four Bands (n.Citationd.).
10 (See e.g., Dewees and Sarkozy-Baniczy [2008, 5, 8–10, 27] and CDFI Fund [2001, 15–16]). For a more general overview of the history and evolving legal status of Native communities in Alaska, Hawaii, and Oklahoma, see, e.g., Brown (Citation2004); Canby (Citation2009, Ch. XIII); Kauanui (Citation2005); Pevar (Citation2012, Ch. 15); and Quinn (Citation1990).
11 The Native CDFIs that are excluded are First American Capital Corps, Indian Land Capital Company, Citizen Potawatomi Community Development Corporation, Native American Bank, and First Nations Oweesta Corporation.
12 While no systematic data exist on uncertified CDFIs and certified CDFI that do not receive funding from the CDFI Fund, these CDFIs tend to be relatively young and small, and thus their aggregate impact on individuals’ credit outcomes is, at least on average, likely to be comparatively limited.
13 The CDFI-ILR data also contains information on CDFI financial clients, another potentially valuable measure of CDFI activity. This variable, however, is evidently inconsistently reported and thus we do not utilize it in our analysis.
15 More precisely, a CDFI’s service area is constructed by using the Census TIGER-provided internal point of the Census zip code tabulation area (ZTA5) that corresponds to the zip code of the CDFI’s address and identifying all of the Census 2010 block groups whose internal points lie within a 100 km radius.
16 The number of individuals in each third is equal for year 2013. However, the regression sample sizes are not equal across thirds due to that fact that an individual is included in the estimation sample only if they remain visible in the CCP with a non-missing Equifax Risk Score for multiple years. Individuals from the lower third sample tend to have lower survival rates in that respect and thus we observe a smaller sample size for the lower third than for the middle and upper thirds.
17 We, however, do not include block group-by-year fixed effects in our estimated models because their inclusion would fully absorb the variation in our focal variables, NCDFIg(i),t−j and NNCDFIg(i),t−j. We also do not include county-by-year fixed effects due to very limited variation in NCDFIg(i),t−j and NNCDFIg(i),t−j within county-year cells.
18 The magnitude of our estimated effect of cumulative exposure to Native CDFIs closely resonates with Jorgensen and Taylor’s (Citation2015) estimate of the size of temporal changes in average risk scores of one Native CDFI’s clients. Drawing on data for 30 borrowers of the Four Directions Development Corporation, a Native CDFI serving the Penobscot Indian Island Reservation, ME, Jorgensen and Taylor find that the borrowers’ average risk score over a six-year period when increased by fifty-three points.
19 Congruent with this conjecture, Balboni and Travers (Citation2017, 10), for example, note that CDFI banks and CDFI credit unions in general indeed have higher delinquency rates than all banks and credit unions.
20 We also estimated the model using the Arellano and Bond (Citation1991) and Arellano and Bover (Citation1995) general method of moments (GMM) approach that in comparison with the 2SLS approach often better mitigates the trade off between lag length and sample size. None of our findings change as a result.
21 The corresponding first stages for the regressions when bank and credit union branch presence are viewed as endogenous, and are thus instrumented for, are similar and hence omitted.
22 Detailed results on the robustness checks are available upon request.
23 Prior to 2010, the NCUA only provided institutional data on credit unions, but not at the level of credit union branches.
Additional information
Notes on contributors
Valentina Dimitrova-Grajzl
Valentina Dimitrova-Grajzl is in the Department of Economics and Business at Virginia Military Institute and the Federal Reserve Bank of Minneapolis. Peter Grajzl is in the Department of Economics at Washington and Lee University, the Federal Reserve Bank of Minneapolis and CESifo, Munich. A. Joseph Guse is in the Department of Economics at Washington and Lee University and the Federal Reserve Bank of Minneapolis. Michou Kokodoko is in Community Development at the Federal Reserve Bank of Minneapolis.
The authors are grateful to Richard Todd for facilitating this project and for his support and useful feedback. Miriam Jorgensen provided us with a wealth of knowledge about CDFIs and offered constructive criticism. For many helpful comments and suggestions on an earlier draft, we also thank Chrystel Cornelius, participants at the CDFI Symposium webinar organized by the Federal Reserve Bank of St. Louis and partners, and an anonymous reviewer. The views expressed here are those of the authors and not necessarily those of the Federal Reserve Bank of Minneapolis or the Federal Reserve System.
The authors declare that there are no known conflicts of interest associated with this manuscript and that there has been no financial support or other influence that could inappropriately impact or otherwise bias this work.
Data availability statement: The Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP) data that support the findings of this study are not publicly available. The CCP data were used under the license of the Federal Reserve Bank of Minneapolis.
Peter Grajzl
Valentina Dimitrova-Grajzl is in the Department of Economics and Business at Virginia Military Institute and the Federal Reserve Bank of Minneapolis. Peter Grajzl is in the Department of Economics at Washington and Lee University, the Federal Reserve Bank of Minneapolis and CESifo, Munich. A. Joseph Guse is in the Department of Economics at Washington and Lee University and the Federal Reserve Bank of Minneapolis. Michou Kokodoko is in Community Development at the Federal Reserve Bank of Minneapolis.
The authors are grateful to Richard Todd for facilitating this project and for his support and useful feedback. Miriam Jorgensen provided us with a wealth of knowledge about CDFIs and offered constructive criticism. For many helpful comments and suggestions on an earlier draft, we also thank Chrystel Cornelius, participants at the CDFI Symposium webinar organized by the Federal Reserve Bank of St. Louis and partners, and an anonymous reviewer. The views expressed here are those of the authors and not necessarily those of the Federal Reserve Bank of Minneapolis or the Federal Reserve System.
The authors declare that there are no known conflicts of interest associated with this manuscript and that there has been no financial support or other influence that could inappropriately impact or otherwise bias this work.
Data availability statement: The Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP) data that support the findings of this study are not publicly available. The CCP data were used under the license of the Federal Reserve Bank of Minneapolis.
A. Joseph Guse
Valentina Dimitrova-Grajzl is in the Department of Economics and Business at Virginia Military Institute and the Federal Reserve Bank of Minneapolis. Peter Grajzl is in the Department of Economics at Washington and Lee University, the Federal Reserve Bank of Minneapolis and CESifo, Munich. A. Joseph Guse is in the Department of Economics at Washington and Lee University and the Federal Reserve Bank of Minneapolis. Michou Kokodoko is in Community Development at the Federal Reserve Bank of Minneapolis.
The authors are grateful to Richard Todd for facilitating this project and for his support and useful feedback. Miriam Jorgensen provided us with a wealth of knowledge about CDFIs and offered constructive criticism. For many helpful comments and suggestions on an earlier draft, we also thank Chrystel Cornelius, participants at the CDFI Symposium webinar organized by the Federal Reserve Bank of St. Louis and partners, and an anonymous reviewer. The views expressed here are those of the authors and not necessarily those of the Federal Reserve Bank of Minneapolis or the Federal Reserve System.
The authors declare that there are no known conflicts of interest associated with this manuscript and that there has been no financial support or other influence that could inappropriately impact or otherwise bias this work.
Data availability statement: The Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP) data that support the findings of this study are not publicly available. The CCP data were used under the license of the Federal Reserve Bank of Minneapolis.
Michou Kokodoko
Valentina Dimitrova-Grajzl is in the Department of Economics and Business at Virginia Military Institute and the Federal Reserve Bank of Minneapolis. Peter Grajzl is in the Department of Economics at Washington and Lee University, the Federal Reserve Bank of Minneapolis and CESifo, Munich. A. Joseph Guse is in the Department of Economics at Washington and Lee University and the Federal Reserve Bank of Minneapolis. Michou Kokodoko is in Community Development at the Federal Reserve Bank of Minneapolis.
The authors are grateful to Richard Todd for facilitating this project and for his support and useful feedback. Miriam Jorgensen provided us with a wealth of knowledge about CDFIs and offered constructive criticism. For many helpful comments and suggestions on an earlier draft, we also thank Chrystel Cornelius, participants at the CDFI Symposium webinar organized by the Federal Reserve Bank of St. Louis and partners, and an anonymous reviewer. The views expressed here are those of the authors and not necessarily those of the Federal Reserve Bank of Minneapolis or the Federal Reserve System.
The authors declare that there are no known conflicts of interest associated with this manuscript and that there has been no financial support or other influence that could inappropriately impact or otherwise bias this work.
Data availability statement: The Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP) data that support the findings of this study are not publicly available. The CCP data were used under the license of the Federal Reserve Bank of Minneapolis.