Perspectives on Economic Education Research 9(1) 20-35 Journal homepage: www.isu.edu/peer/ Financial Literacy and Banking Affiliation: Results for the Unbanked, Underbanked, and Fully Banked 1 Elizabeth Breitbach a,2, William B. Walstad b a Department of Economics, University of South Carolina, 1014 Greene Street, Columbia, SC 29208, United States b Department of Economics, University of Nebraska Lincoln, United States Abstract Checking and savings accounts are frequently used by most U.S. households. These financial accounts become important for economic well-being because they affect the ability of a household to make efficient financial transactions and safely store financial assets. Households without these accounts cannot participate fully or effectively in the economy and often incur additional costs when making financial transactions. The study draws on a large national data set, the 2009 survey of Financial Capabilities in the United States by the Financial Industry Regulatory Authority (FINRA), to analyze the segment of the U.S. population who are unbanked (do not have a commercial transaction account) and underbanked (those who hold a transaction account, but supplement the account with alternative financial services). The study investigates how demographic characteristics and financial literacy affect this banking behavior. The results show that unbanked and underbanked households have significantly lower levels of financial literacy. More importantly, the inclusion of financial literacy variables reduces the effect of race and ethnicity variables on banking participation. Key Words: banking behavior, unbanked and underbanked households, financial literacy, financial institutions JEL Codes: G21, D14 1 The authors thank an anonymous referee for helpful comments on a previous version. 2 Corresponding author.
1. Introduction Individual financial decisions in the United States are involved with the banking industry even if an individual chooses to avoid banking institutions altogether. These banking relationships are significant for several reasons. Individuals with traditional transaction accounts are found to have higher levels of savings than their unbanked counterparts. Not only do bank accounts promote increased saving, they often offer check cashing and bill paying services at a lower cost than alternative financial products, such as non-bank money orders (or bill pay) and non-bank check cashing services. In the United States, fully-banked households are those that take advantage of these traditional transactions accounts for banking and are the largest share (about 73 percent) of households when classifying the banking relationships. This high percentage is not surprising because for most households, it would seem impossible not to have a transaction account to make day-to-day financial payments, obtain cash for purchases, or deposit a paycheck and other checks. Nevertheless, there are about five percent of U.S. households do not have any checking or savings accounts and operate largely outside the banking system when making financial transactions. These households are classified as unbanked. In addition, 22 percent of households are classified underbanked. 3 These households have a bank account, but supplement the account with alternatives to traditional banking services such as non-bank money orders, non-bank check-cashing services, payday loans, rent-to-own agreements, or pawnshops. The financial behavior of both the unbanked and underbanked warrants further analysis because this approach to banking matters can often be costly and time-consuming for these households compared with fully banked household that exclusively use traditional banking services. This study investigates the financial literacy of households across unbanked, underbanked, and full banked households in the United States. Financial literacy has become an important economic topic for analysis since the Great Recession and because of the complexity of financial decisions facing households (Lusardi and Mitchell 2014). Since obtaining a transaction account requires a relatively low level of financial literacy, it is expected that the respondents score on a set of financial questions will have a small but significant effect on banking participation. The most common reason a household gives for being unbanked is they do not have enough money to make it worthwhile. 4 For this reason it is expected that financial literacy will have a smaller effect on whether a household has a transaction account relative to indicators of income and wealth. As banking participation moves to the underbanked level it is expected the financial literacy effect will become stronger, as using costly alternative financial services may indicate individuals are failing to budget or lack information and understanding of banking. Overall, it is hypothesized that these households will be less financially literate than fully banked households. Educating households on the benefits of a properly maintained transaction account could translate to significant time and cost savings for these households who most need it. 3 The estimates for the three groups are based on data analysis describe in a later section of the paper. 4 Using the FINRA data, 73% of unbanked households give this reason. 21
2. Literature review Current literature on banking participation has found a relatively consistent definition of an unbanked household. This group is defined as not having any type of commercial bank account, including checking and savings accounts (Grimes et al. 2010; Hogarth et al. 2005; Rhine and Greene 2006; Rhine et al. 2006; Paulson and Rhine 2008). Many papers studying banking participation have focused on demographic and socioeconomic variables. Unbanked individuals were significantly more likely to be single (Rhine and Greene 2006; Rhine et al. 2006; Hogarth, et al. 2005), Hispanic and African American (Hogarth et al. 2005; Rhine and Greene 2006; Rhine et al. 2006; Grimes, et al. 2010), and have increased family size/presence of dependent children (Hogarth et al. 2005; Rhine and Greene 2006). Results for gender and age were mixed or not significant (Amuedo-Dorantes and Bansak 2006; Grimes, et al. 2010; Paulson and Rhine 2008; Rhine and Greene 2006). Socioeconomic variables such as education, work force participation, income, and wealth have been found to be stronger than demographic variables in determining whether a household is unbanked. Education level has been tied to banking participation in nearly all current literature; those with a high school degree or less were significantly more likely to be unbanked (Hogarth et al. 2005; Grimes et al. 2010; Rhine and Greene 2006; Rhine et al. 2006). Previous studies that used a dummy variable to indicate if the respondent was in the work force have found mixed results. Hogarth et al. (2005) took an in-depth look at work status, including working, retired, unemployed looking, and unemployed not looking. Relative to head of households who were unemployed not looking, respondents who were working and retired were significantly more likely to be banked, while unemployed looking were more likely to be unbanked. Other studies have found the result to be not significant (Grimes et al. 2010; Rhine and Greene 2006). Many studies have concluded that having low income is not only a significant factor effecting banking participation, but it is the primary determinant in predicting whether a household is unbanked (Grimes et al. 2010; Hogarth et al. 2005; Paulson and Rhine 2008; Rhine and Greene 2006; Rhine et al. 2006). Researchers have not only investigated income, but a household s overall financial situation, including variables controlling for net wealth and access to credit. Results indicate that those with positive net worth (Rhine and Greene 2006; Hogarth et al. 2005), who own their home (Hogarth et al. 2005; Grimes et al. 2010; Rhine et al. 2006), and own their vehicle (Hogarth et al. 2005) were significantly less likely to be unbanked. Rhine and Greene (2013) found that households who have experienced a drop in income, loss of employment, or loss of health care were significantly more likely to get rid of their transaction account. The majority of current literature is focused in the area of unbanked households. However, there is another category of individuals who do have a bank account, but like the unbanked, make costly financial decisions that warrant further investigation. The Federal Deposit Insurance Corporation (FDIC) (2009) defines an underbanked household as those that have a checking or savings account, but rely on alternative financial services. Specifically, underbanked households have used non-bank money orders, non-bank check-cashing services, payday loans, rent-to-own agreements, or pawnshops at least once or twice a year, or refund anticipation loans at least once in the past five years. Much of the research on the 22
underbanked has focused on specific individual alternative services. The underbanked as a whole, relative to the research completed on the unbanked, is limited. Much of the research has focused on how to reach these individuals, not necessarily the characteristics of these households (Beard 2010, Gross, et al. 2012). The Center for Financial Services and Innovation (CFSI) has also complied information on this group, including descriptions of the underbanked and how they can be reached (Seidman 2005). This paper will work to improve upon the research on the underbanked and help those targeting education programs to better understand their audience. There have been a couple of studies that have explored the relationship between financial literacy and banking participation, focusing primarily on the unbanked. Grimes et al. (2010) included a set of economic literacy indicators in their regression analysis. The first measure included a set of seven economic questions. These questions ranged in topic from the current unemployment and inflation rates, to how prices are determined in a competitive market. The authors found that a higher scores on the set of economic questions lead to a lesser likelihood of the household being unbanked. The second method the authors used was to include indicators for courses taken in high school; including economic, business, or personal finance courses. These courses were combined into one indicator for exposure and used separately in various regression analyses. However, only respondents who had taken at least one or more economic, business, or finance course (the combined indicator) or taken at least one business course were significantly less likely to be unbanked. Another study implemented a pilot program offering financial counseling targeted a set of individuals considered underserved by the mainstream financial services (Wiedrich 2014). The authors of the study used a control and treatment group to determine the effect increased information had on these individuals. While the treatment group saw a decrease in the percentage of debt past due, there were no statistically significant increases in banking participation after the counselling, suggesting that establishing effective programs or strategies to encourage more bank participation is difficult. 3. Data The data set used for this study is the 2009 Financial Capability in the United States survey created by the Financial Industry Regulatory Authority (FINRA). 5 There were a series of three data sets created, including a National Survey, Military Survey, and State-by-State Survey. These surveys had three main objectives: (1) to benchmark key measures of financial capability, (2) evaluate how those key measures vary with underlying demographic, behavioral, and attitudinal characteristics, (3) provide data and estimates to inform public policy toward financial capability. The State-by-State data set was chosen for this analysis due to its large number of observations, approximately 28,146 American adults. A sample of at least 500 respondents from each state and the District of Columbia was obtained by an online survey between June and October of 2009. The data set was weighted to match Census distributions based on age by gender, race/ethnicity, and level of education (Applied Research and Consulting LLC 2009). 5 The data set, along with the questionnaire can be found at: http://www.usfinancialcapability.org/ 23
Included in the data set is information on the household s use of banking services and alternatives to traditional banking options. Together this information will be used to create the three levels of banking participation: unbanked, underbanked, and fully banked. The FINRA data set requires two questions to determine whether the respondent is unbanked. First, respondents are asked if they or their household has a checking account. The second relevant question is whether they or their household has a saving account, money market account, or CDs. If a respondent answered no to both of these questions, they are considered to be unbanked. If they responded yes to at least one of these questions they are banked. 6 Of the banked respondents, it must be understood whether these households are underbanked or fully banked. The FDIC definition of underbanked will be followed as closely as possible using the FINRA data set, though there are a few differences. The first discrepancy will be use of nonbank money orders and check cashing services. The FINRA survey did not ask banked households about their usage of these services, which may underestimate the number of underbanked households. Another difference between definitions is the inclusion of auto title loan usage. While the FDIC definition does not include this alternative financial service, the FINRA survey does ask respondents whether they have used this product. Since the information is available, it will be included as a service used by the underbanked. 7 The final difference is the lack of information about frequency of use. Distinguishing between frequent and infrequent users is not possible, so underbanked households will be defined as those that have taken out or used these services in the past five years. 8 The primary reason the FIRNA survey was used for this analysis was the richness of the information on the respondent s financial literacy. The questionnaire asked respondents a set of five financial questions: (1) Suppose you had $100 in a savings account and the interest rate was 2% per year. After 5 years, how much do you think you would have in the account if you left the money to grow? (a) More than $102* (b) Exactly $102 (c) Less than $102 (2) Imagine that the interest rate on your savings account was 1% per year and inflation was 2% per year. After 1 year, how much would you be able to buy with the money in this account? (a) More than today (b) Exactly the same (c) Less than today* (3) If interest rates rise, what will typically happen to bond prices? (a) They will rise (b) They will fall* (c) They will stay the same (d) There is no relationship between bond prices and the interest rate (4) A 15-year mortgage typically requires higher monthly payments than a 30-year mortgage, but the total interest paid over the life of the loan will be less. (a) True* (b) False (5) Buying a single company s stock usually provides a safer return than a stock mutual fund. (a) True (b) False* 6 Determining whether 373 households were unbanked or banked was not possible due to don t know/refused responses. 7 Excluding households who only use auto title loans from the underbanked, the percentage of underbanked households falls from 23.2% to 18.95% (of banked households). 8 Of the banked households, the underbanked or fully banked status could not be determined for 398 households do to don t know or refused responses. 24
Respondents had the opportunity to respond don t know or could refuse to answer each of the questions. 4. Who are the unbanked and banked households? Table 1 presents the descriptive statistics for the full sample of respondents and a breakdown between unbanked and banked households. 9 All statistics have been weighted so the results represent characteristics of the population of the United States. Previous literature indicated that the percentage of unbanked is around ten percent (Grimes, et al. 2010, Hogarth, et al. 2005). The FINRA data set suggests a much lower percentage of unbanked, 5.3 percent. One possible reason for the decrease in the number of unbanked households is the price of traditional checking accounts has decreased. In 2009 the number of free checking accounts was at its peak, meaning little to no money was needed to open and maintain an account (Bruce 2009). Financial literacy was assessed both as an aggregate score, the number of questions answered correctly out of five questions, and at the individual question level. Incorrect answers were combined with don t know and refused responses. The average number of questions answered correctly by a respondent was three. Breaking down the sample, unbanked households answered one less question correctly, on average, relative to the banked sample. The apparent trend of the unbanked being less financially literate follows through to the individual questions. For all questions, a higher proportion of banked households answered the question correctly, relative to the unbanked. The greatest variance was seen with the mortgage question, with a nearly 28 percent difference between banked and unbanked households. The savings and mortgage questions were considered to be the easiest questions, with the highest percentage of the full sample answering these questions correctly. It is expected, since banking participation is considered a relatively basic form of personal finance, these questions will have the strongest predicating power when determining the effect of financial literacy on whether a household chooses to hold a transaction account. The bond question is considered the most difficult, with less than 30 percent of the full sample answering correctly. This is also the question with the smallest spread between the proportion of unbanked and banked households answering it correctly. It is expected that this question, due to its complexity, will have little effect on whether a household is unbanked. Previous literature has indicated that some demographic and socioeconomic variables are important in analyzing banking participation. For example, race/ethnicity has been found to be a significant determinant of banking participation. Since a larger percentage of the sample in this data set is White non-hispanic, a larger percentage of the unbanked also fall into this category. While it may be true that the majority of the data set is composed of White Americans, a relatively large percentage of unbanked households are Black American and Hispanic. 9 Banked households include all households who report holding a transaction account. This includes the underbanked and fully banked households. 25
Table 1: Mean Unbanked and Banked Characteristics Full Sample Unbanked Banked Unbanked 0.053 Financial Literacy Score (Out of 5) 3.007 2.026 3.062 (1.433) (1.445) (1.412) Savings Question-Correct 0.782 0.619 0.791 Inflation Question-Correct 0.649 0.408 0.662 Bond Question-Correct 0.278 0.195 0.283 Mortgage Question-Correct 0.761 0.498 0.775 Stock Divers. Question-Correct 0.537 0.306 0.550 Female 0.512 0.514 0.512 Male 0.488 0.486 0.488 Age 18 to 34 0.306 0.521 0.293 Age 35 to 54 0.379 0.377 0.379 Age 55 and older 0.315 0.101 0.327 White, non-hispanic 0.687 0.519 0.697 Black, non-hispanic 0.114 0.227 0.107 Hispanic 0.133 0.212 0.129 Asian 0.046 0.019 0.047 Native American/Alaska 0.016 0.024 0.016 Other Races 0.008 0.008 0.009 Married 0.535 0.235 0.551 Single, never-married 0.282 0.543 0.267 Divorced/Separated 0.139 0.200 0.136 Widow 0.044 0.022 0.045 Dependent Child Present 0.384 0.425 0.382 Less than High School Education 0.034 0.175 0.027 High School Degree or Equiv. 0.292 0.462 0.282 Some College, no degree 0.420 0.305 0.427 College Degree 0.159 0.048 0.166 Post College Education 0.094 0.009 0.099 Self-Employed 0.080 0.069 0.081 Employed Full-Time 0.362 0.172 0.373 Employed Part-Time 0.097 0.105 0.097 Not in the Workforce - Homemaker 0.089 0.116 0.087 Not in the Workforce - Student 0.058 0.076 0.057 Not in the Workforce - Disabled 0.042 0.074 0.040 Unemployed 0.098 0.339 0.084 Not in the Workforce - Retired 0.173 0.049 0.179 Income less than $15,000 0.144 0.510 0.123 Income $15,000-25,000 0.131 0.201 0.128 Income $25,000-35,000 0.129 0.131 0.129 Income $35,000-50,000 0.161 0.077 0.166 Income $50,000-75,000 0.189 0.045 0.197 Income $75,000-100,000 0.108 0.019 0.113 Income $100,000-150,000 0.088 0.009 0.093 Income greater than $150,000 0.049 0.006 0.052 Drop in Income 0.405 0.563 0.397 Homeowner 0.592 0.166 0.615 At Least One Credit Card 0.751 0.204 0.781 Observations 27,773 26
It is hypothesized that education level and employment status will be strongly correlated with banking participation. Those with a high school degree or less represent a larger percentage of unbanked households, while the reverse is true for some college education or higher. Respondents who are employed full-time, self-employed, or retired represent a higher percentage of banked households, while those who are employed part-time, homemakers, disabled, or unemployed are unbanked. Income is expected to be one of the most significant determinants of being unbanked. As income increases it is expected that a household would have a greater need for an account. Those earning less than $15,000 make up less than 15 percent of the full sample, but represent more than 50 percent of the unbanked. Not only is income significant, wealth and access to credit is also important in determining need for an account. The breakdown between levels of banking shows a higher percentage of banked households are homeowners; only 17 percent of unbanked households own a home, compared to 62 percent of banked households. Credit card ownership is also of interest when exploring banking participation, 20 percent of unbanked households, compared to 78 percent of banked households, hold at least one credit card. 5. Probit results: unbanked A probit model was specified and estimated to identify the likely influence of financial literacy on banking participation. A dummy variable was created to indicate whether a household is unbanked, equal to one if the household does not have a checking or savings account, and zero otherwise. The results for the financial literacy variables are reported in Table 2, along with all demographic and socioeconomic variables. The first regression, presented in Column I includes a financial literacy indicator equal to the number of questions answered correctly out of a set of five. The more questions the respondent answered correctly, the less likely the household was unbanked. For each question the respondent answered correctly, the household was 0.25 percent less likely to be unbanked. Columns II to VI present the results for the individual question analysis. Indicators for whether respondents answered the question correctly is included, the omitted category are respondents who answered incorrectly or responded don t know/refused to answer. The expected sign of answering a given question correctly is negative, meaning those households are less likely to be unbanked. This result is found across all questions, the results are also significant in all cases with the exception of the bond question. The bond question was expected to have the smallest effect on whether a household was unbanked due to the sophistication of the question. Answering the other financial literacy questions correctly leads to one-half to one percentage point decrease in the likelihood the household is unbanked. While these results do not appear to be large in magnitude, they are in line with other strong indicators of banking participation including race, education, and homeownership. However, the findings indicate that unbanked households could benefit from increased information on transaction accounts. 27
Table 2: Probit Regression Results for Unbanked Financial Literacy Controls I II III IV V VI Financial Literacy Score (Out of 5) -0.0025*** Savings Question-Correct -0.0060*** Inflation Question-Correct -0.0051*** Bond Question-Correct 0.0002 Mortgage Question-Correct -0.0075*** Stock Divers. Question-Correct -0.0031** Female -0.0021-0.0013-0.0015-0.0007-0.0013-0.0011 Age 18 to 34-0.0001 0.0005-0.0004 0.0005 0.0005 0.0003 Age 55 and Older -0.0062*** -0.0065*** -0.0062*** -0.0065*** -0.0064*** -0.0063*** African American/Black 0.0064** 0.0074*** 0.0069*** 0.0078*** 0.0065** 0.0076*** Hispanic 0.0030 0.0034 0.0032 0.0036 0.0029 0.0035 Asain -0.0059* -0.0060-0.0057-0.0056-0.0061* -0.0054 Native American/Alaskan -0.0024-0.0027-0.0028-0.0030-0.0027-0.0027 Other Races -0.0015-0.0016-0.0013-0.0014-0.0017-0.0015 Single Never-Married 0.0076*** 0.0080*** 0.0078*** 0.0080*** 0.0076*** 0.0079*** Divorced/Separated 0.0092*** 0.0094*** 0.0095*** 0.0095*** 0.0093*** 0.0093*** Widow -0.0035-0.0033-0.0033-0.0032-0.0035-0.0033 Dependent Child Present 0.0037** 0.0038** 0.0038** 0.0040** 0.0039** 0.0040** Less than a High School Education 0.0202*** 0.0209*** 0.0214*** 0.0223*** 0.0216*** 0.0216*** Some College but no degree -0.0071*** -0.0077*** -0.0078*** -0.0084*** -0.0076*** -0.0080*** College Degree -0.0088*** -0.0095*** -0.0094*** -0.0101*** -0.0095*** -0.0097*** Post College Education -0.0100*** -0.0108*** -0.0107*** -0.0114*** -0.0108*** -0.0109*** Self-Employed 0.0090** 0.0088** 0.0091** 0.0090** 0.0091** 0.0092** Employed Part-Time 0.0013 0.0015 0.0012 0.0012 0.0011 0.0013 Not in the Workforce-Homemaker 0.0108*** 0.0115*** 0.0117*** 0.0120*** 0.0114*** 0.0116*** Not in the Workforce-Student -0.0022-0.0025-0.0024-0.0027-0.0026-0.0025 Not in the Workforce-Disabled 0.0036 0.0039 0.0041 0.0044 0.0040 0.0042 Unemployed 0.0221*** 0.0225*** 0.0227*** 0.0229*** 0.0218*** 0.0228*** Not in the Workforce-Retired -0.0021-0.0021-0.0021-0.0020-0.0020-0.0020 Income less than $15,000 0.0363*** 0.0387*** 0.0388*** 0.0401*** 0.0370*** 0.0390*** Income $15,000-25,000 0.0168*** 0.0179*** 0.0179*** 0.0184*** 0.0172*** 0.0178*** Income $25,000-35,000 0.0125*** 0.0136*** 0.0134*** 0.0139*** 0.0128*** 0.0135*** Income $35,000-50,000 0.0038 0.0041 0.0042 0.0044 0.0040 0.0042 Income $75,000-100,000 0.0001 0.0001 0.0000-0.0001 0.0000-0.0002 Income $100,000-150,000 0.0028 0.0025 0.0027 0.0024 0.0026 0.0026 Income greater than $150,000 0.0074 0.0074 0.0069 0.0066 0.0066 0.0069 Large Drop in Income 0.0030* 0.0029* 0.0028* 0.0028* 0.0030* 0.0028* Homeowner -0.0142*** -0.0146*** -0.0148*** -0.0151*** -0.0144*** -0.0149*** At Least 1 Credit Card -0.0440*** -0.0455*** -0.0453*** -0.0459*** -0.0446*** -0.0451*** Pseudo R-squared 0.3144 0.3130 0.3127 0.3111 0.3142 0.3118 Observations 26,585 26,585 26,585 26,585 26,585 26,585 Standard errors in parentheses * p<.1 ** p<.05 *** p<.01 28
Consistent with findings from previous literature, income is a significant indicator for whether a household has a transaction account. Those in the lowest income brackets are three percentage points more likely to be unbanked. Black Americans, those with low levels of education, and unemployed respondents are also significantly more likely to be unbanked. 10 6. Who are the underbanked and fully banked households? The sample size for the underbanked analysis will not match the unbanked analysis because the level of participation within the banked households is unknown for some respondents. Those individuals who do not know or refused to disclose their level of participation, or for which the survey was terminated before their status was determined, were dropped. 11 Of the banked households for which it could be determined whether they were underbanked or fully banked, 23.2 percent are underbanked. 12 Table 3 presents the descriptive statistics for underbanked and fully banked households, along with all banked households for which their level of participation could be determined. The aggregate score on the set of financial literacy questions shows a significant gap between the underbanked and the fully banked households. Underbanked households answer an average of 2.7 questions correctly while fully banked households answer 3.2 correct. A greater proportion of the fully banked households answered all individual questions correctly, relative to underbanked households. The inflation and stock diversification questions have the largest difference between the percentages of underbanked who answered correctly compared to fully banked, with a spread of 15 percent. This is particularly alarming since these are relatively easy questions. Again, the bond price question had the smallest spread, with a seven percent difference between the underbanked and fully banked households that answered the question correctly. While the decision on whether to open a transaction account is a relatively basic one, the decision to use an alternative financial service is more complex. This suggests that the more difficult questions may be better indicators of whether a household is underbanked or fully banked. A relatively small amount of research has looked at the characteristics of the underbanked. The descriptive statistics indicate White and Asian respondents represent a significantly higher percentage of fully banked households, while Black American, Hispanic, Native American/Alaskan, and respondents who report multiple races/ethnicities compose a higher percentage of underbanked households. 10 The regressions were also run without the financial literacy variables. Without the inclusion of financial literacy controls the coefficients on race and education increased in magnitude. 11 Households who answered yes to at least one alternative service are considered underbanked, even if they did not know/refused other services. If the respondent answered they did not use any of the alternatives but didn t respond or refused one question they were dropped from the analysis. This change decreases the sample size from 26,544 to 26,146 households. 12 This percentage is slightly smaller than the number presented in the introduction. The previous percentage was relative to all households. The current percentage is in terms of only those with a transaction account. 29
Table 3: Mean Underbanked and Fully Banked Characteristics Proportion Banked Sample Proportion Underbanked Proportion Fully Banked Underbanked 0.232 Financial Literacy Score (Out of 5) 3.078 2.673 3.200 (1.403) (1.356) (1.393) Savings Question-Correct 0.794 0.734 0.812 Inflation Question-Correct 0.666 0.550 0.701 Bond Question-Correct 0.285 0.232 0.301 Mortgage Question-Correct 0.779 0.719 0.797 Stock Divers. Question-Correct 0.554 0.439 0.589 Female 0.512 0.520 0.510 Male 0.488 0.480 0.490 Age 18 to 34 0.290 0.413 0.252 Age 35 to 54 0.381 0.421 0.369 Age 55 and older 0.329 0.166 0.379 White, non-hispanic 0.698 0.593 0.730 Black, non-hispanic 0.107 0.181 0.085 Hispanic 0.128 0.163 0.118 Asian 0.047 0.034 0.051 Native American/Alaska 0.016 0.027 0.013 Other Races 0.008 0.010 0.008 Married 0.552 0.475 0.575 Single, never-married 0.265 0.327 0.247 Divorced/Separated 0.137 0.163 0.129 Widow 0.046 0.036 0.049 Dependent Child Present 0.382 0.525 0.339 Less than High School Education 0.026 0.045 0.021 High School Degree or Equiv. 0.282 0.348 0.262 Some College, no degree 0.427 0.450 0.420 College Degree 0.166 0.114 0.181 Post College Education 0.100 0.043 0.117 Self-Employed 0.081 0.078 0.082 Employed Full-Time 0.374 0.384 0.370 Employed Part-Time 0.096 0.096 0.096 Not in the Workforce - Homemaker 0.087 0.108 0.081 Not in the Workforce - Student 0.056 0.063 0.054 Not in the Workforce - Disabled 0.041 0.062 0.034 Unemployed 0.084 0.126 0.072 Not in the Workforce - Retired 0.181 0.083 0.210 Income less than $15,000 0.122 0.154 0.112 Income $15,000-25,000 0.128 0.187 0.110 Income $25,000-35,000 0.129 0.164 0.119 Income $35,000-50,000 0.166 0.194 0.158 Income $50,000-75,000 0.197 0.171 0.204 Income $75,000-100,000 0.114 0.070 0.127 Income $100,000-150,000 0.093 0.045 0.107 Income greater than $150,000 0.052 0.015 0.063 Drop in Income 0.397 0.541 0.353 Homeowner 0.616 0.414 0.677 At Least One Credit Card 0.782 0.608 0.835 Observations 26,146 30
It was expected that underbanked households will be less educated and less likely to work full time, relative to fully banked households, creating a greater need for non-bank, short term loans. The trends follow as predicted, the proportion of underbanked is highest for those with some college experience or less. Unemployed respondents make up only eight percent of all banked households, when breaking down the banked category to underbanked and fully banked households unemployed respondents represent thirteen percent of the underbanked. The income variables do not appear to be as strong for the underbanked analysis when compared to the unbanked. However, underbanked respondents are nearly twenty percent more likely to report an unexpected drop in income. This was expected as the services that define the underbanked are high cost, short-term loans. 7. Probit results: underbanked Table 4 reports the probit results for the underbanked households. The dependent variable, underbanked, is a dummy variable equal to one if the household uses at least one alternative banking service. If the household has not utilized any of these services the variable is made equal to zero and the household is concerned to be fully banked. Column I of Table 4 presents the result for the aggregate measure of financial literacy. This result is significant and in the expected direction: for each question answered correctly the household is 1.5 percent less likely to be underbanked. An indicator for whether the respondent answered a specific financial question correctly is presented in the next columns. Results show that all signs on the coefficients for answering the questions correctly are in the expected direction and significant, except the mortgage question. The coefficients on the financial literacy questions range from one percentage point (bond question) to four percentage points (inflation question). The question with the largest impact is the inflation question. As expected, the more difficult questions were better indicators for whether a household was underbanked or fully banked. This indicates that these households could benefit significantly from increased financial information regarding the use of non-bank, short-term loans. Black Americans and Native Americans were also significantly more likely to be underbanked. These coefficients are large in magnitude, around eleven and seven percentage points, respectively. Education, wealth and income, and whether the household experienced an unexpected drop in income also had a large impact on whether the household used alternative financial services. 13 13 When financial literacy variables were removed from the regression analysis some changes in the demographic and socioeconomic coefficients occurred: slightly increasing the impact of race/ethnicity variables. While this result was unexpected, it suggests that race alone is not the main determinant of whether or not a household uses the alternative services. Education and income variables also see a slight increase in the magnitude of their effect. 31
Table 4: Probit Regression Results for Underbanked Financial Literacy Controls (Relative to Fully Banked Households) I II III IV V VI Financial Literacy Score (Out of 5) -0.0148*** Savings Question-Correct -0.0290*** Inflation Question-Correct -0.0379*** Bond Question-Correct -0.0140* Mortgage Question-Correct -0.0050 Stock Divers. Question-Correct -0.0311*** Female -0.0229*** -0.0173** -0.0203*** -0.0160** -0.0148** -0.0192*** Age 18 to 34 0.0477*** 0.0517*** 0.0455*** 0.0509*** 0.0514*** 0.0495*** Age 55 and Older -0.0767*** -0.0784*** -0.0763*** -0.0780*** -0.0783*** -0.0768*** African American/Black 0.1137*** 0.1187*** 0.1153*** 0.1206*** 0.1206*** 0.1186*** Hispanic 0.0027 0.0048 0.0031 0.0051 0.0051 0.0046 Asain -0.0379** -0.0358** -0.0364** -0.0350** -0.0346* -0.0346* Native American/Alaskan 0.0751*** 0.0749*** 0.0742*** 0.0741*** 0.0746*** 0.0752*** Other Races 0.0180 0.0182 0.0177 0.0188 0.0183 0.0190 Single Never-Married -0.0125-0.0115-0.0118-0.0112-0.0113-0.0115 Divorced/Separated 0.0255** 0.0256** 0.0259** 0.0258** 0.0256** 0.0251** Widow 0.0391** 0.0402** 0.0419** 0.0392** 0.0396** 0.0388** Dependent Child Present 0.0969*** 0.0972*** 0.0967*** 0.0985*** 0.0983*** 0.0977*** Less than a High School Education 0.0347 0.0369 0.0367 0.0392* 0.0393* 0.0373 Some College but no degree -0.0135* -0.0177** -0.0164** -0.0201** -0.0204** -0.0169** College Degree -0.0598*** -0.0660*** -0.0632*** -0.0679*** -0.0687*** -0.0638*** Post College Education -0.0719*** -0.0791*** -0.0756*** -0.0809*** -0.0821*** -0.0764*** Self-Employed -0.0162-0.0173-0.0161-0.0174-0.0175-0.0169 Employed Part-Time -0.0369*** -0.0373*** -0.0373*** -0.0375*** -0.0377*** -0.0368*** Not in the Workforce-Homemaker -0.0142-0.0127-0.0128-0.0123-0.0122-0.0132 Not in the Workforce-Student -0.0510*** -0.0522*** -0.0509*** -0.0520*** -0.0526*** -0.0514*** Not in the Workforce-Disabled 0.0559*** 0.0576*** 0.0575*** 0.0577*** 0.0585*** 0.0575*** Unemployed -0.0125-0.0119-0.0121-0.0119-0.0120-0.0115 Not in the Workforce-Retired -0.0493*** -0.0498*** -0.0491*** -0.0494*** -0.0494*** -0.0489*** Income less than $15,000-0.0263** -0.0218* -0.0231* -0.0208-0.0208-0.0231* Income $15,000-25,000 0.0492*** 0.0537*** 0.0524*** 0.0541*** 0.0545*** 0.0516*** Income $25,000-35,000 0.0353*** 0.0384*** 0.0380*** 0.0387*** 0.0390*** 0.0367*** Income $35,000-50,000 0.0370*** 0.0391*** 0.0385*** 0.0391*** 0.0394*** 0.0379*** Income $75,000-100,000-0.0399*** -0.0405*** -0.0401*** -0.0410*** -0.0413*** -0.0408*** Income $100,000-150,000-0.0552*** -0.0565*** -0.0564*** -0.0570*** -0.0573*** -0.0558*** Income greater than $150,000-0.0909*** -0.0923*** -0.0927*** -0.0925*** -0.0935*** -0.0916*** Large Drop in Income 0.0890*** 0.0881*** 0.0886*** 0.0888*** 0.0885*** 0.0885*** Homeowner -0.0979*** -0.1001*** -0.0998*** -0.1003*** -0.1004*** -0.0993*** At Least 1 Credit Card -0.1188*** -0.1207*** -0.1205*** -0.1209*** -0.1209*** -0.1191*** Pseudo R-squared 0.1539 0.1528 0.1536 0.1523 0.1521 0.1532 Observations 25,174 25,174 25,174 25,174 25,174 25,174 Standard errors in parentheses * p<.1 ** p<.05 *** p<.01 32
8. Conclusion Banking participation is often the first experience an individual has with the financial world. It is also one of the simplest forms for engaging in the financial community. Avoiding banking institutions is costly, both in terms of time and money. Those with low levels of banking participation fall into the lowest income brackets and would benefit most from any cost savings associated with a properly maintained transaction account. When comparing households with no transaction account to those with, financial literacy was a significant indicator for whether the respondent s household had a transaction account both when financial literacy was measured as an aggregate test score and an individual topic. As previous literature suggested, income is the primary determinant for whether a household has a transaction account. However, it may be the case that since the decision on whether to have a transaction account is a basic one, the financial literacy questions presented here may not create a good measure. A future questionnaire could include financial questions focused primarily on banking and budgeting. Questioning whether the unbanked are knowledgeable about balancing a checkbook and budgeting may offer a better indicator for financial literacy in this context. While the effect of the financial literacy measure was relatively small for unbanked households, it was much larger for underbanked and fully banked households. More importantly, the inclusion of financial literacy variables decreased the significance of race and ethnicity variables. This outcome is important because it suggests that that education can help improve banking participation among racial minorities. The results of this research suggest that providing financial information to households with low levels of banking participation could be beneficial. The findings also suggests that those in the lowest income brackets and minorities may benefit most by targeted information on the benefits of transaction accounts. Specifically, increasing knowledge on the proper maintenance and cost of transaction accounts can help increase banking participation, thereby saving these households money. References Amuedo-Dorantes, Catalina, and Cynthia Bansak. 2006. Money Transfers among Banked and Unbanked Mexican Immigrants." Southern Economic Journal 73 (2): 374-401. Applied Research and Consulting LLC. 2009. Financial Capability in the United States: Initial Report of Research Findings from the 2009 National Survey. Beard, Martha. 2010. Reaching the Unbanked and Underbanked. Central Banker. Winter 2010. Bernheim, B. Douglas, and Daniel M. Garrett. 2003. The Effects of Financial Education in the Workplace: Evidence from a Survey of Households." Journal of Public Economics 87 7-8. 33
Bruce, Laura. 2009. Bank Fees Take Flight. Bankrate. http://www.bankrate.com/finance/checking/2009-checking-study.aspx. Carswell, Andrew T. 2009. Does Housing Counseling Change Consumer Financial Behaviors? Evidence from Philadelphia." Journal of Family and Economic Issues 30 (4): 339-356. Cole, Shawn, and Guri Kartini Shastry. 2008. If You Are so Smart, Why Aren t You Rich?" Presented at the 2009 Federal Reserve System Community Affairs Research Conference: Washington DC. Federal Deposit Insurance Agency. 2009. FDIC National Survey of Unbanked and Underbanked Households." Federal Deposit Insurance Corporation. FINRA Investor Education Foundation. 2012. National Financial Capability Study. http://www.finrafoundation.org//programs/capability/index.htm. Fox, Jonathan, et al. 2005. Building the Case for Financial Education." Journal of Consumer Affairs 39 (1): 195-214. Grimes, Paul W., et al. 2010. High School Economic Education and Access to Financial Services." Journal of Consumer Affairs 44 (2): 317-335. Gross, Mathew, et al. 2012. Use of Financial Services by the Unbanked and Underbanked and the Potential for Mobile Financial Services Adoption. Federal Reserve Bulletin 98 (4). Hathaway, Ian, and Sameer Khatiwada. 2008. Do financial education programs work?" Hogarth, Jeanne M., et al. 2005. Who Has a Bank Account? Exploring Changes over Time, 1989-2001." Journal of Family and Economic Issues 26 (1): 7-30. Lusardi, Annamaria and Olivia S. Mitchell. 2014. The Economic Importance of Financial Literacy: Theory and Evidence. Journal of Economic Literature 52 (1): 5-44. Paulson, Anna, and Sherrie L. W. Rhine. 2008. The Financial Assimilation of an Immigrant Group: Evidence on the Use of Checking and Savings Accounts and Currency Exchanges." Journal of Family and Economic Issues 29 (2): 264-278. Rhine, Sherrie L. W., and William H. Greene. 2013. Factors That Contribute to Becoming Unbanked. Journal of Consumer Affairs, 47: 27 45. Rhine, Sherrie L. W., and William H. Greene. 2006. The Determinants of Being Unbanked for U.S. Immigrants." Journal of Consumer Affairs 40 (1): 21-40. 34
Rhine, Sherrie L. W., et al. 2006. The Importance of Check-Cashing Businesses to the Unbanked: Racial/Ethnic Differences." Review of Economics and Statistics 88 (1): 146-157. Seidman, Ellen, et al. 2005. Getting to Know Underbanked Consumers: A Financial Services Analysis. The Center for Financial Services Innovation (CFSI). Wiedrich, Kasey. 2014. Financial Counseling & Access For The Financially Vulnerable. Corporation for Enterprise Development (CFED). 35