Credit Union Common Bonds: Participation & Policy

Credit unions, cooperative financial institutions, originated in 19th-century Europe and arrived in the United States in the early 20th century. They've maintained a significant presence despite the rise of corporate financial institutions, even increasing market share in some developed countries. In the US, credit unions are distinct due to the legal requirement that members share a common bond—a principle enshrined in the Federal Credit Union Act of 1934, defining the bond as 'occupation or association, or [belonging] to groups within a well-defined neighborhood, community, or rural district.' This 'one-person-one-vote' structure ensures equal control among members. The paper notes that historically, credit union members often came from groups underserved by traditional banks, but today, the demographic profile of members is more representative of the general population. The growth of credit unions, from 1% of the adult population in 1935 to 33% in 1989, highlights their increasing significance in the financial landscape.

The common bond requirement became a focal point of national debate in the late 1990s. A Supreme Court case involving the AT&T Family Credit Union and the National Credit Union Administration (NCUA) challenged the NCUA’s interpretation of the 1934 Act, which allowed multiple common bonds within a single credit union. The court ruled against multiple common bonds, siding with banks who argued this violated the spirit of the original legislation. This decision, however, was quickly overturned by President Clinton signing into law the 1998 Credit Union Membership Access Act, essentially reversing the Supreme Court's ruling. This swift legislative response demonstrated the significant political weight of the credit union movement and the complexities surrounding its regulation. The opposing viewpoints highlighted the tensions between promoting competition and preserving the integrity of the credit union model.

The debate surrounding the common bond requirement reflects contrasting views on credit union efficiency and their role in the financial marketplace. Some argue that credit unions operate efficiently, offering better terms to members than traditional banks and thrifts, which is a viewpoint often emphasized by bank and thrift managers. Conversely, concerns exist about the inherent efficiency of credit unions due to their one-member-one-vote governance. Critics point to potential free-rider problems, lack of professional management, and weak member incentives as factors that hinder efficient operation. These differing perspectives contribute to the ongoing debate about credit union effectiveness and the potential impact of their expansion, with the fear of large institutions undercutting traditional banking institutions. The tension between these competing arguments lies at the core of the common bond discussion, influencing the regulations on the size and structure of credit unions.

The theoretical model employs a Hotelling (1929) linear city framework. A line segment represents a city, with households distributed along it. Each household's location reflects its preferences for banking services, simplified to a single index from zero to one. Credit unions are modeled as scarce or differentiated entities, located at specific points on the line, while commercial banks are ubiquitous and offer services at a fixed price anywhere. This setup emphasizes the importance of household preferences in determining credit union participation. Households are further grouped by 'affinity' or common bonds, focusing on occupational bonds for simplicity. The model limits the economy to three firms (A, B, and C), each potentially sponsoring a credit union. The spatial distribution of households reflects the common bond structure, ensuring that, for example, employees of firm A are clustered together on the line.

The model simulates credit union formation and operation across two periods. In period one, only single-common-bond credit unions are allowed. Households vote on a management team, with the winning team minimizing the sum of member travel costs, effectively locating the credit union centrally within the group's preference spectrum. The model assumes that travel costs increase quadratically with distance, which is why credit unions will locate in the center of the preference spectrum. Households then decide whether to join the credit union or use a commercial bank. If the potential membership is too small, the credit union may not be viable. Period two introduces a regime allowing multiple common bonds, enabling the formation of multi-group credit unions.

The model is simulated 10,000 times using randomly generated parameters for household distribution and credit union costs. The simulations compare outcomes under single and multiple common bond scenarios. Results show a dramatic increase in the number of households and employee groups served by credit unions when multiple common bonds are allowed. The simulation highlights that multiple-group credit unions comprising geographically disparate groups may have lower participation rates due to increased travel costs for some members, while geographically contiguous multiple-group credit unions show high participation. The model also explores comparative statics, examining the impact of changes in household travel costs and commercial bank service prices. Increased commercial bank prices lead to a rise in credit union usage, suggesting that credit unions are more attractive when bank competition is weak.

The empirical analysis utilizes data from federally chartered and insured occupational credit unions in 1996. The sample includes 4,733 credit unions, categorized by membership type (single or multiple common bonds). A significant portion (2,753 or 58.2%) of the credit unions in the sample had multiple common bonds among their members, while the rest (1,980 or 41.8%) operated under a single common bond. These credit unions are further classified by membership type such as educational, military, government, manufacturing, and services classifications. The data allows for a comparison of credit union characteristics and performance across different common bond structures. The dataset, therefore, provides a robust foundation for analyzing the impact of the common bond requirement on credit union operations. This large sample size increases the reliability of the findings and allows for a more detailed examination of credit union characteristics and behaviors.

The study uses a semi-parametric empirical model with two dependent variables: 'PARTICIPATION' (actual members/potential members) and 'COST' (operating expenses/total assets). Key independent variables include the number of members (or potential members for participation rate analysis), total assets, and the Herfindahl index (measuring local bank deposit market concentration). The model incorporates lagged values of membership, potential membership, and total assets, which are included in the non-parametric part in logarithmic form. The parametric part includes a dummy variable for multiple-group credit unions, the county-specific Herfindahl index, the growth rate of the credit union's home state's real gross state product, and dummy variables for specific membership types. This comprehensive set of variables allows for a detailed analysis of the factors affecting credit union participation rates and operating costs. The use of a semi-parametric model offers flexibility in capturing both linear and non-linear relationships between the variables.

The empirical results show a significant negative relationship between potential membership size and participation rates, which aligns with the theoretical model’s findings. Credit unions with multiple common bonds demonstrate higher participation rates compared to those with single common bonds, holding other factors constant. This finding applies to a comparison of otherwise identical credit unions. However, the analysis does not apply to the scenario where a single-common-bond credit union expands to incorporate another group, increasing potential membership and violating the ceteris paribus assumption. The study also finds a positive correlation between local banking concentration (as measured by the Herfindahl index) and credit union participation rates. The cost ratio analysis shows that costs tend to rise before declining with membership size, which potentially contradicts the initially assumed economies of scale, and suggests other influencing factors like the presence of subsidies and service variability across credit unions are at play.

The empirical analysis reveals a key finding: credit unions with multiple common bonds exhibit higher participation rates than comparable credit unions with a single common bond. This result holds true when controlling for other factors such as the size of the potential membership, total assets, and local banking market concentration. The study uses a large dataset of federally chartered occupational credit unions in 1996. The researchers emphasize that this finding applies to a comparison of two essentially identical credit unions differing only in the number of common bonds. It does not apply to a situation where a single-common-bond credit union expands its membership to include additional groups, thereby increasing its potential membership and altering other key variables. The observed higher participation rates in multiple-common-bond credit unions suggest that broadening membership eligibility, while seemingly counterintuitive, can positively impact overall member engagement. This is a key insight that challenges previously held assumptions about the relationship between common bonds and credit union success.

The study also investigates the influence of local banking competition on credit union participation rates. Using the Herfindahl index as a measure of market concentration, the analysis finds a positive association between higher banking concentration and higher credit union participation. This supports the structure-conduct-performance paradigm, suggesting that credit unions become more attractive when banking competition is weaker. In markets with less competition from banks, credit unions may offer a more compelling alternative. The findings on cost ratios are less conclusive. While the model assumed economies of scale, the data suggest a more complex relationship between operating expenses and membership size, possibly due to the influence of factors such as unrecorded subsidies, varying service levels, and the overall competitiveness of the local banking market. The cost ratios illustrate the importance of considering non-economic factors in assessing credit union performance.

The researchers caution against interpreting the findings as unqualified support for multiple-group credit unions. The positive correlation between multiple common bonds and higher participation rates is observed while holding other factors constant. Expanding a single-group credit union to include multiple groups will alter the size of the potential membership, which is found to have a negative impact on the participation rate. The analysis focuses on the impact of common bonds holding all else equal, a condition not necessarily present in real-world credit union mergers. The impact of increased potential membership is negative, meaning that the benefits observed from multiple common bonds are conditional on the size of the credit union. The study highlights the nuanced relationship between credit union characteristics and the broader economic and competitive environment, calling for continued research to further explore these dynamic interactions. Subsidies and variations in service offerings across credit unions add to the complexities of interpreting cost data.