Abstract
The tax exempt status enjoyed by credit unions is based in part upon such institutions providing financial services to individuals who are traditionally underserved by conventional banking institutions. Rather than relying upon an abstract measure of underserved status, we instead empirically estimate the probability that a household does not have an account with a traditional bank. We then use this information to determine whether individuals with reduced access to banking services are also more likely to belong to credit unions. We find that underserved households are less likely, rather than more likely, to use the services of credit unions.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Furthermore, neither author has any bank or credit union affiliations besides as a customer/member.
The National Credit Union Association (NCUA), the organizing body for credit unions, states on its website mycreditunion.gov, “credit unions provide valuable access to financial services for people underserved and unserved by traditional financial institutions.”
To be precise, the “unbanked” variable in our analysis is employed as a label and predictor for a household being “underserved.” The percent of observations deemed “underserved” in our analysis (17%) closely corresponds to those considered “underbanked” in FDIC surveys (18%). The “unbanked” variable is not well-compared to the variable of the same name in the FDIC due to data-related differences in variable construction.
While most small credit unions may be serving the underserved sector of society (Mohanty 2006), the number of underserved individuals in these small credit unions will be dwarfed by the number of (well-off) individuals in larger credit unions. In fact, the latter group – well-off individuals at larger CU’s – constitute the bulk of CU membership, and the dollar value of transactions made by this group is higher, thus disproportionately transferring any CU tax subsidy benefits to society’s well-off.
Variations in state-level income taxes and gross receipt taxes affect the state-specific tax burden on a financial institution. This state-level variation is an interesting future area of research in determining how CU’s respond to differing incentives. One complicating factor, however, is that states often levy taxes on financial institutions irrespective of their physical presence in the state.
There are two caveats we would note regarding our variable. Specifically, not all credit unions have community charters, and there may be an issue with individuals who live close to a credit union being unable to access the credit union due to common membership requirements. We are not concerned regarding headquarters being in a different location since our variable is meant to capture the physical location of the branch.
As expected, CU member households are somewhat older on average due to the nature of the data collection process. Our data are collected from survey responses and will tend to oversample older individuals as the respondents, and they will be coded as members of a CU member household if even one individual in the household has a CU account.
Due in part to this aspect of the dataset, we later employ robustness tests where we limit our sample to low-income individuals using median, instead of mean income to split the sample. Similarly, we also test our results leaving income entirely out of the analysis. We find our results to be robust to each of these alternative specifications.
There are four key characteristics for which the CFM is not representative of the US population. These are age, income, race, and homeownership status. As a result, the CFM recommends the use of a general raking method to adjust the observational weights to better match other national surveys and provides computer code to this effect. Unfortunately, this weighting requires that all of age, race, home ownership, and income variables are available for each observation. To accomplish this, the CFM randomly fills in missing values from the appropriate bin of the cdf for that factor using a uniform distribution assumption. This introduces additional noise into the analysis. We therefore refrain from using the raking method. Instead, we control for the relevant variables directly in the regression analysis. CFM clearly states that “all raw input can be considered a simple random sample.” Additionally, even though we present descriptive statistics for informational purposes, we refrain from drawing undue support for our hypotheses based upon these statistics.
Notice that the present analysis does not directly consider the issue of loan provision, since, unfortunately, obtaining a representative, client-level, dataset for banks and credit unions is not currently possible. Given the diversity in the size and structure of banks and credit unions, and current market composition, using client data for just one or a few banks and/or credit unions would not allow us to make more general assertions. It is for this reason that we opt for the more limited household survey data with no differentiation between depositing and lending relationships but with the important benefit of being both representative and unbiased – either toward banks or toward credit unions. Nevertheless, our findings regarding the unbanked-to-underserved portion of the argument are serious enough to warrant consideration in their own right since they have implications regarding the distribution of benefits by underserved status.
This is in keeping with those in the literature who have also found that the substitutability or complementarity between banks and credit unions is more complex than it would otherwise seem (Périlleux et al. 2016).
It is also possible that our observed results reflect the fact that credit unions have been so successful in serving the needy in their membership ranks that these individuals have improved their relative income standing to the point that they can no longer be considered underserved. In that scenario, credit unions would have still fulfilled their mandate, and, if those (formerly underserved) individuals now also have traditional bank accounts as of the time our dataset begins, then these positive effect of credit unions from the distant past would no longer be unobservable.
References
Astore, M. 2016. Banks vs. credit unions: A comparative analysis of Massachusetts banks and credit unions before, during and after the recession. Bridgewater State University Working Paper.
Chakrabarti, R., Donghoon, L, van der Klaauw, W., and B. Zafar. 2011. Household debt and saving during the 2007 recession. Federal Reserve Bank Staff Report No. 482.
Chatterji, A.K., Luo, J., and R.C. Seamans. 2017. Too legit to quit? Banks vs. credit unions after the financial crisis. Working Paper.
Clark, L. 1943. Credit Unions in the United States. The Journal of Business 16(4): 235–246.
Cook, T., and L. D’Antonio. 1984. Credit union taxation: Competitive effects. Journal of Economics and Business 36(2): 251–262.
CUNA. 2016. U.S. Credit Union Profile—Mid Year 2016 CUNA Economics and Statistics. www.cuna.org Accessed 24 January 2019.
DeYoung, R., Goddard, J., McKillop, D.G., and J.S. Wilson. 2019. Who consumes the credit union tax subsidy? QMS Research Paper. Downloaded October 21, 2020. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3429208.
DiSalvo, J., and R. Johnston. 2017. Credit Unions’ Expanding Footprint, Federal Reserve Bank of Philadelphia Banking Trends, First Quarter, 17–25.
Dunn, L., and R. Olsen. 2014. US household real net worth through the great recession and beyond: Have we recovered? Economics Letters 122: 272–275.
Dunn, L., and I. Mirzaie. 2016. Consumer debt stress, changes in household debt, and the great recession. Economic Inquiry 54(1): 201–214.
Dymski, G., and W. Li. 2003. The macrostructure of financial exclusion: Mainstream, ethnic, and fringe banks in money space. Espace-Population-Societies 21(1): 183–201.
Ely, D., and K.J. Robinson. 2009. Credit unions and small business lending. Journal of Financial Services Research 35(1): 53–80.
Feinberg, R.M. 2001. The competitive role of credit unions in small local financial services markets. Review of Economics and Statistics 83(3): 560–563.
Feinberg, R., and D. Meade. 2017. Economic Benefits of the Credit Union Tax Exemption to Consumers, Businesses, and the U.S. Economy, prepared on behalf of the National Association of Federally-Insured Credit Unions. 1–16.
FDIC. 2017. 2017 FDIC National Survey of Unbanked and Underbanked Households https://www.fdic.gov/householdsurvey/2017/2017report.pdf. Downloaded 4 April 2019.
Finke, M.S., J.S. Howe, and S.J. Huston. 2017. Old age and the decline in financial literacy. Management Science 63(1): 213–230.
Fulford, S. 2015. How important is variability in consumer credit limits? Journal of Monetary Economics 72: 42–63.
Huston, S.J. 2012. Financial literacy and the cost of borrowing. International Journal of Consumer Studies 36: 566–572.
Jackson, W.E. III. 2003. The future of credit unions: Public policy issues. Filene Research Institute Paper.
Jacob, K., M. Bush, and D. Immergluck. 2002. Rhetoric and reality: An analysis of mainstream credit unions’ record of serving low-income people. Chicago, IL: Woodstock Institute.
Jambulapati, V., and J. Stavins. 2013. The Credit CARD Act of 2009: What Did Banks Do? Federal Reserve Bank of Boston Working Paper 13-7.
Jiang, S.S., and L.F. Dunn. 2013. New evidence on credit card borrowing and repayment patterns. Economic Inquiry 51(1): 394–407.
Jones, L.E., C. Loibl, and S. Tennyson. 2015. Effects of informational nudges on consumer debt repayment behaviors. Journal of Economic Psychology 51: 16–33.
Karels, G., and C. McClatchey. 1999. Deposit insurance and risk-taking behavior in the credit union industry. Journal of Banking & Finance 23: 105–134.
Koku, P.S. 2015. Financial exclusion of the poor: A literature review. International Journal of Bank Marketing 33(5): 654–668.
Marshall, L., and S. Pellerin. 2017. Credit unions: A taxing question, Federal Reserve Bank of Richmond Econ Focus, Second Quarter: 20–34.
Maskara, P. K., Kuvvet, E., and G. Chen, G. 2020. The role of P2P platforms in enhancing financial inclusion in US—An analysis of peer-to-peer lending across the rural-urban divide. Financial Management, forthcoming.
Melzer, B.T., and D.P. Morgan. 2015. Competition in a consumer loan market: Payday loans and overdraft credit. Journal of Financial Intermediation 24(1): 25–44.
Mohanty, S.K. 2006. Comparing credit unions with commercial banks: Implications for public policy. Journal of Commercial Banking and Finance 5(2): 97–113.
NASCUS. 2007. Survey of the state credit union system, report to former chairman of US Ways and Means Committee, House of Representatives. https://www.nascus.org/pdf/website-final-nascus-survey-state-cu-system.pdf.
Nicolini, G., B.J. Cude, and S. Chatterjee. 2013. Financial literacy: A comparative study across four countries. International Journal of Consumer Studies 37: 689–705.
Périlleux, A., A. Vanroose, and B. D’Espallier. 2016. Are financial cooperatives crowded out by commercial banks in the process of financial sector development? Kyklos 69(1): 108–134.
Schuh, S., Shy, O, and J. Stavins. 2010. Who gains and who loses from credit card payments? Theory and calibrations. Federal Reserve Bank of Boston Working Paper 10-03.
Shen, S., A.G. Sam, and E. Jones. 2014. Credit card indebtedness and psychological well-being over time: Empirical evidence from a household survey. The Journal of Consumer Affairs 48(3): 431–456.
Solo, M.T. 2008. Financial exclusion in Latin America-or the social costs of not banking the urban poor. Environment and Urbanization 20(1): 47–66.
Summers, L.K. 2001. Comparing credit unions with other depository institutions. United States Department of the Treasury Technical Report.
US Government Accountability Office. 2006. Greater transparency needed on who credit union serve and on senior executive compensation arrangements, GAO Report, 07-29. Washington DC: Government Accountability Office.
Walker, D.A. 2016. Credit unions’ rising share of small business loans. Journal of Accounting and Finance 16(6): 11–24.
White, R. 2011. If it quacks like a duck: In Light of today’s financial environment, should credit unions continue to enjoy tax exemptions? Georgia State University Law Review 28(4, 13): 1365–1394.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
Race
This variable was recoded into a hierarchy of Hispanic, then Black, Asian, other minority, and finally White. “Other minority” was defined as including any of native Hawaiian, American Indian, and, self-reported “other” race (Tables 7, 8).
Family Structure
Two questions were used to define family structure. These were: (1) the marital status of the household head and (2) the number of children in the household. Marital status was recoded for simplicity to one of either married or single. Widowed or divorced were coded as single and separated was coded as married. The number of children was recoded to missing rather than having a negative value.
Education
This variable was initially coded as years of education; however, due to lack of informational value, we recoded education into four groups. These groups represented whether individuals had less than a high school degree (education in category 0-11), exactly a high school degree (education in category 12), some college (education in category 13-15), or a college degree or better (education in category 16).
Income and Employment
Our “income” measure was a variable created by the survey designers and was meant to capture all legitimate sources of income that should be claimed on income taxes. Specifically, the types of income included were all salaries and wages from the household head and partner, farm or business income, unemployment or disability payments, child support and alimony, social security and various supplemental or income assistance programs, education assistance income, scholarships or fellowships, estate money, trust or inheritance money, and interest on savings, bonds, dividends, pensions, annuities, rent, royalties, or other incomes. Notice that the inheritance portion of income did not seem to matter in affecting our results, since we ran regressions without the inclusion of our income variable as well. The few individuals with negative household income were not used, and household income was recoded in thousands. Because we were interested in the natural log of income in our regressions, we next recoded anyone with income that was strictly between zero and one to be exactly one. We noticed that inflation-adjusting income made essentially no difference for our results. To determine employment status, we examined whether anyone in the family (household head or spouse) is currently employed. Dual-income households were considered to be those with both a household head and a spouse who had earned a wage within the last month.
Banked
This variable was constructed based on whether individuals in the family had either a checking or a savings account. Individuals without either of these were considered to be unbanked.
Credit Union Membership
We created this Boolean from a direct question in the dataset asking if anyone in the household was a member of a credit union.
Mutual Funds and Inheritance
These variables were coded directly from questions in the data regarding the amount of inheritance money that was received last month (separate from information used to determine income) and the total amount of mutual funds (or stocks) an individual owned. We considered them to have mutual funds if the individual had a positive nonzero level of mutual funds (or stocks). Similarly, they had a positive level of inheritance if the money from their inheritance last month was a positive nonzero number. Notice that we grouped stocks and mutual funds together, however, using only mutual funds alone had similar results.
Renter
This variable was constructed from a direct question that asked the respondent if she owns or is in the process of buying her residence. When the respondent stated that she wasn’t, we considered her to be renting. Notice that this definition excludes all respondents who were making any home payments during the process of buying a home.
Age, Gender
These two variables were coded directly from the data. For age, we recoded the variable using the current date of the survey minus the birthdate. Gender was coded in the data in a binary fashion.
Rights and permissions
About this article
Cite this article
Maskara, P.K., Neymotin, F. Do Credit Unions Serve the Underserved?. Eastern Econ J 47, 184–205 (2021). https://doi.org/10.1057/s41302-020-00183-3
Published:
Issue Date:
DOI: https://doi.org/10.1057/s41302-020-00183-3