Tax Foundation Studies Exemptions for Federal Credit Unions

The Tax Foundation would like to thank the Independent Community Bankers of America for their support of this study.

Tax Loss to the Treasury

Credit unions are growing rapidly, and so is the associated tax loss to the federal Treasury caused by their exemption. Indeed, the tax loss over the five-year period 2004-2008 is estimated in this study to be $12.6 billion. Extended over the typical ten-year federal budget window, the tax loss reaches $31.3 billion. The size of the tax loss is substantially higher than estimates prepared by government arbiters including the Office of Management and Budget or the Congressional Budget Office.

The Tax Exemption's Original Justification

This tax exemption has been in law for almost 70 years because of the original concept of credit unions' cooperative ownership. The original legal "field of membership" restrictions on credit unions were designed to limit their ability to compete by strictly defining who could be a depositor and borrower from a credit union, with the idea that credit unions would use their tax advantage to serve low- income borrowers and depositors. However, over time credit unions have avoided most of the restrictions, and as a result they have competed directly and successfully with other financial institutions in many markets with a major cost advantage, the tax exemption. Moreover, there is no solid evidence that credit unions have turned the subsidy into service for low-income people.

Who Benefits from the Tax Exemption?

Corroborated by other studies of credit unions and banks, the direct and indirect evidence gathered for this study shows that the equity holders of credit unions receive the tax saving as unusual returns. These unusual returns do not show up as relatively high dividends, however. Instead, they occur as unusually large retained earnings accumulated as net worth in their credit unions. The shareholders' extra income reinvested in the credit union provides new capital that allows the credit union to grow faster than other institutions.

There is some evidence that certain types of loans have lower rates at credit unions, especially for loans that have become less profitable and less available at banks, such as auto loans. There is also some evidence that part of the tax advantage is absorbed by costs that are higher than they would have been in a taxed, or more competitive, environment.

Overall, however, the dominant effect of the tax exemption is to boost the equity ratio. Over the past ten years, credit unions have had an equity ratio -- the ratio of equity to total assets -- that is more than 25 percent larger than that of banks.

Of the 50 basis points in subsidy that the tax exemption provides, at least 33 basis points accrue to owners in the form of larger equity and larger assets. Approximately 6 basis points may accrue to credit union borrowers through lower interest rates, and not more than 11 basis points are absorbed by higher labor costs. There is little or no effect on deposit rates or other costs.

By giving a tax exemption to credit unions while taxing their competitors -- banks, thrifts and finance companies, financial institutions that offer the same consumer deposits and loans -- the federal government distorts the allocation of resources. It promotes the employment of deposit and credit resources in the tax-free credit union sector at the expense of all these other financial institutions.

Competitive Advantages Beyond the Exemption

Along with the tax exemption, a steady erosion of limits on credit union membership has allowed credit unions to grow much more rapidly than banks, especially over the past two decades. In 1998, the U.S. Supreme Court struck down the liberalization of membership rules, but the U.S. Congress promptly passed new legislation overriding the court. As a result, credit unions have rapidly consolidated, merged and broadened their geographic markets, all the while maintaining their tax exemptions. Thus, Congress created new tensions by weakening the original case for tax exemption.

Banks currently are subject to extensive costs to insure that they are meeting the credit demands of low-income borrowers. Credit unions were excluded from these provisions because of the presumption that they must be serving such consumers. After all, their charters are rooted in common bonds that seem to assume that credit unions meet these requirements. But the evidence shows that credit unions do not serve low- and moderate-income people to any greater extent than banks. For example, most credit unions have an occupational bond that requires members to be employed, often in industries with relatively highwage jobs.

Proposals for Reform

Today credit unions continue to grow faster than banks, have little practical limitations on membership, and make business loans that increasingly have no limits on who can borrow, how much or for what purpose. Even the limits that Congress has imposed, as they otherwise removed limits on credit union markets and competition, have broad loopholes and remain under serious challenge by the credit union industry.

Today the principal justification for the tax exemption would seem to be that it already exists and, therefore, removing it could adversely impact thousands of institutions and their customers. Under current law, as it is being enforced, there is no good policy argument based on equity or efficiency for maintaining the tax exemption. Some analysts have argued that small institutions (under $10 million in assets) should continue to be tax exempt because of their special character and, perhaps, innate inefficiencies. Notably, the corporate income tax already takes size into account by taxing low-income firms at lower tax rates (15 percent, while larger firms pay rates that range from 34 to 39 percent).

Removing the credit unions' tax exemption would create a more equitable tax system and help level the playing field with other financial institutions. It would also raise about $2 billion in tax revenue each year, either directly from credit unions or from more profitable and more highly taxed banks, where some credit union deposits and assets would migrate in a competitive market. Finally, it would raise the rate of return on some $65 billion of capital that is squirreled away in credit unions, earning lower rates of return than would be the case at taxpaying banks.

The unusually high equity ratio of credit unions would be reduced; and management of capital costs would make credit unions more efficient, perhaps lowering operating costs and interest rates on deposits and raising rates on loans, at least in some markets. Credit unions would be more subject to market control and would manage risk and return more efficiently, increasing the value of their franchises to their owners, despite smaller relative size and slower growth.

Introduction

The competitive and legislative environment of credit unions has changed dramatically in the past several years and a key aspect, their exemption from the federal income tax, has come under increasing scrutiny.¹ Credit unions in the U.S. and elsewhere have been free of federal income taxes for a variety of related reasons, including their cooperative organization and ownership structure, their so-called common bond, and their mission to provide services to small or relatively lowincome savers.

That status has come under increasing question both in the U.S. and abroad, largely because of increasing attention to the principle of neutrality in taxation and the equity principle that calls for a "level playing field" or "treating equals equally."² The special tax status of credit unions has also been increasingly challenged because of questions about the ability and commitment of credit unions to serve low-income people. The competitive and tax status of credit unions were tested by a Supreme Court decision in February of 1998 that insisted credit unions enforce tight "common bond" requirements in order to obtain and maintain their charters and tax exemption.³ Congress reacted swiftly. Within six months Congress had passed the Credit Union Membership Access Act (CUMAA) which authorized the very multiple-group fields of membership that the Supreme Court had just invalidated.

While the issue of membership groups appears limited, it is the key restriction that analysts had used to argue for the special regulatory and tax status of credit unions. The irony is that CUMAA indirectly removed the special status of credit unions, but in their override of the Supreme Court, Congress appeared to strongly endorse the status quo, which includes the special tax status of credit unions. Thus, there is an unresolved tension in the regulatory and tax environment facing the financial services industry. If nothing else, removing the special status of credit unions under regulatory guidelines calls into question the related federal income tax exemption.

Yet recently the Secretary of the Treasury pointedly implied that the issue will not be on the table any time soon. In a question and answer session at the Federal Reserve Bank of Chicago's Bank Structure Conference in May 2004, Secretary Snow was asked if he thought the special tax status of credit unions was fair, if and when it would be reviewed and whether it would be altered. He responded with a unqualified "No" to the idea of reviewing the exemption.⁴

Whether challenges to the tax status of credit unions can be so easily dismissed is doubtful, even if officials would like to ignore the issue. But the response does point to the intransigence of political forces to challenge the status quo, especially when it has political appeal enjoyed by the current tax treatment of credit unions. Whether one favors or is opposed to the current tax treatment, a sound understanding of its origins, history, and effects on credit unions, their customers and their competitors is essential for decisions to maintain or to alter the existing preferential tax and regulatory treatment of credit unions.

Credit unions are among the most rapidly growing financial firms in the country. This unusual growth could possibly be due to special tax breaks or other subsidies, or perhaps simply because they play a unique or special role in the economy. But in recent years, analysts have begun to question the bases for special treatment of credit unions and more research has begun to point to adverse consequences of their special tax treatment. This study reviews the historical basis of the federal tax exemption of credit union income and factors eroding its support. It also looks at the consequences of the tax break and who gains and who loses from credit unions' tax exempt status.

Large and Growing Tax Loss

The losses from the tax exemption are quite large, according to several alternative estimates. The U.S. Treasury estimates that taxing credit unions would raise $1.24 billion in fiscal 2005, according to Bickley (2003) who provides an excellent overview of the issue of removing the credit union exemption from federal income taxation.⁵ The U.S. Congress' Joint Committee on Taxation (JCT) estimates that the loss in federal tax revenue is about $1.2 billion in 2004 and that this will rise to $1.8 billion by 2008.⁶ A recent study by Chmura Economics & Analytics (May 2004) indicates a much larger tax loss of $1.89 billion in 2002. The most recent comparable JCT estimate for the fiscal 2002 tax loss is $0.9 billion.

The most recent estimates computed for this study indicate that the tax loss is even larger, about $2 billion in 2003 (calendar year), and $1.9 billion in 2002. Over the ten-year revenue estimation period 2004 to 2013, the tax loss is expected to be $31.3 billion. Notably, large federal deficits in recent years require serious attention to both the spending and the revenue sides of the budget. The sizable tax revenue loss from the $31.3 billion tax exemption granted credit unions warrants more attention than normal in this environment, especially since credit unions and the tax expenditure are growing rapidly. Where does this tax loss go? Is it received indirectly by credit union savers in higher dividends or interest? Do credit union borrowers receive the tax savings in the form of lower interest rates? Or are wages higher at credit unions? According to the analysis in this study, the most likely result is that the equity holders of credit unions receive the tax saving as unusual returns. These unusual returns do not show up as relatively high dividends, however. Instead, they occur as unusually large retained earnings accumulated as net worth in their credit unions. The shareholders' extra income reinvested in the credit union provides new capital that allows the credit union to grow faster than other institutions.

For many years, both in this country and abroad, industry leaders, policy analysts and policymakers have debated whether credit unions play some special role in the economy and whether they deserve special tax treatment. Spurred by the Supreme Court challenge (1998) to the special treatment, Congress passed CUMAA. This act enhanced the availability of the tax subsidy for credit unions because it widened the field of membership, the group of people that could be customers of a credit union. The looser field of membership requirements also allowed credit unions, especially large ones, to expand their growth opportunities, reinforcing the competitive advantage obtained from their tax advantages. The ability to pass on their tax saving to broader customer bases, in the form of lower interest rates on loans, higher rates on deposits or increased retained earnings growth, warrants a review of the size of the tax subsidy and its incidence, or where the tax saving goes. The study will detail the effects of the tax subsidy on the pricing and/or income advantages accruing to credit unions and its effects on growth, especially for relatively large credit unions. The changing credit union structure has allowed credit unions to extend their competitive advantages to broader groups facilitating increased competition and market share growth, especially for large institutions. Estimates of the tax loss and its effects on credit union profitability, service prices and cost structure are most useful in assessing the benefit and cost of the federal income tax exemption.

I. The Basis of Tax Exemption-- The Common Bond and How It Is Changing

The first U.S. credit union was chartered in 1909 by the state of New Hampshire.⁷ The federal income tax (and federal chartering of credit unions) came later, and credit unions were not exempt under the new tax. An administrative ruling by the Attorney General in 1917 exempted credit unions from federal taxation. Federal chartering of credit unions began under the Federal Credit Union Act of 1934, but federally chartered credit unions were not exempted from federal and state income taxation until the act was amended in 1937. The amendment was based on the services of credit unions to their members, according to Bickley (2003). Since 1937, both federal and state-chartered credit unions have been exempt from federal income taxation and federally chartered institutions have been exempt from all federal and state taxation. States have generally followed the federal lead.

Only five states subject state-chartered credit unions to their state corporate income tax: California, Indiana, Iowa, Oklahoma and Oregon. Iowa's 12-percent rate is the nation's highest; the other four states have more typical rates that vary from 6.0 to 8.84 percent. New Hampshire imposes two taxes, a business profits and a business enterprise tax, that are similar to an income tax and combine for a maximum tax of 9.25 percent. Several more states impose limited tax levies such as franchise taxes on their state-chartered credit unions. About half of all states subject state-chartered credit unions to sales or use taxes and personal property taxes, and most states impose real property taxes as well.⁸ These taxes are all much smaller than the federal income tax, but even these small differences could affect the performance of federal and state- chartered credit unions.

The key factor is that credit unions are exempt from federal income taxation and have been for many years. Savings and loan associations were also exempt from federal income taxation until 1951 when their exemption, granted under the same tax law provision as that for credit unions, was removed. Congress eliminated the exemptions for savings and loans and mutual savings banks on the grounds that they were similar to profit-seeking corporations.⁹ Since then, large credit unions have come to resemble large thrifts and banks. It should be noted again that while all credit unions are exempt from federal income taxes, only federally chartered credit unions are exempt from all taxes at the state level as well. State-chartered credit unions are not exempt from state income taxes, franchise taxes, property or sales taxes in many states. Thus, one way to assess the effect of tax exemption is to compare federal and statechartered credit unions.

The rationale for tax exemption originates in the uniqueness of the "common bond" under which credit unions are chartered. The common bond determines a "field of membership" from which a credit union may draw and the members are the owners of the accumulated reserves (or retained earnings) in proportion to their shares of deposits. Only members may hold deposits or receive loans from the credit union, though credit unions also make loans to other credit unions and credit union service corporations and they can accept deposits and make loans to non-members under some restricted conditions. Such loans and deposits are quite small however. Thus, a credit union is a nonprofit financial cooperative for tax purposes, though they are often organized as corporations for legal reasons. Profit is reinvested or paid to depositors and taxed as interest under the individual income tax. The principal functions of credit unions are to accept deposits (called shares, and actually equity shares for most of the usual purposes) from their members and to make loans to them or investments. Shares include checkable share draft accounts, regular savings shares, money market, certificates of deposit and IRA or Keogh accounts. The principal loans to members, ranked by size, are new and used vehicle loans, first mortgage loans, other real estate loans, and credit card loans. The motto of credit unions summarizes their original goals, "not for profit, not for charity, but for service." Of course credit unions learned long ago that more service can come from operating efficiently, earning profits and expanding their capital base.

Regulators allow four categories of common bonds. The earliest and most common types are the single occupational or single associational bonds. The first limits the group to employees of the same firm or workers in a single occupational class who may work for many different employers. The second group includes members of a social or civic group, which shares common loyalties, mutual interests or mutual benefits and which provides activities where members have contact with one another. Since 1982, when failing credit unions were perceived as being unduly risky because of the heavy risk concentration created by the narrow common bond criteria, regulators have created broader more diverse common bonds. The third type today, the multiple common bond, allows groups with different occupational or associational bonds to join together. This has been the most controversial because it essentially provides little or no limit to membership; indeed this was the issue in NCUA vs. First National Bank and Trust Co. (1998). The fourth category is the community common bond, which limits the field of membership to people who reside in or are employed in a well-defined local community. This is potentially a bond with even less meaningful restriction than the multiple bond because it has been used to encompass such a broad geographic area and open group. Community charters are the fastest growing segment of the credit union industry. One can easily imagine a U.S. credit union restricted to residents of the U.S., little more restricted than a nationwide bank. Indeed, on July 29, 2004, the National Credit Union Administration approved a community charter for the LA Financial Credit Union, which has a field of membership of all 10.1 million residents of Los Angeles County.

The common bond is unique to credit unions. It was intended to restrict who can be a member and benefit from the services of the credit union, but it also limits the credit union from developing potentially more profitable relationships with depositors or borrowers. Its cooperative nature and the restrictions on its business are both reasons given for the tax exemption. The occupational and associational bonds had always provided the support for the notion that a credit union is a cooperative effort by people of more limited means, or perhaps low income, to make available financial services to which they would otherwise either have no access or face prohibitive cost.

Many studies have questioned whether even these more traditionally chartered firms really live up to that standard. For example, Jacob, Bush and Immergluck (2002) have shown that the incomes of credit union members are little different from the incomes of customers of banks and thrifts. This should not be surprising, especially in the case of occupational credit unions, because the members are employed, often in higher wage sectors such as teachers or trade union members. The largest credit unions are those for the airlines, aircraft production, teachers, government workers and unions. Credit unions focus on consumer banking, while most banks depend more on business customers, but there is little reason to think that credit unions are serving a social purpose not served by other readily available financial institutions. The basis then for special tax treatment could not rest on the notion that credit unions serve a disadvantaged class, though this was the original motivation and basis of the tax exemption.

The introduction and increasing number of multiple bond charters beginning in 1982 provided a more important challenge to the notion that the credit unions are damaged by their social objective of serving a particular field of membership. Until then, potential members were arguably a more limited and perhaps unprofitable group to serve than the potential depositors or borrowers from banks, thrifts or other financial institutions. Banking organizations and their trade groups, which have to compete with the special treatment accorded credit unions, campaigned to end the competitive advantage of tax-exempt credit unions. Since the early-1980s, they have strenuously objected to the liberalization of the limits imposed by the common bond, the legal limits that had justified the credit union tax break earlier.

Chmura (2004) suggests that the "moral hazard" of borrowers, essentially the risk that borrowers will engage in risky activities that make it more likely they will default on loans, was boosted by having credit unions that serve predominately people of small means. This greater risk might have justified tax exemption as a cost offset in the past, according to Chmura. One could add a greater "adverse selection" problem as such a basis. Adverse selection means that the highest risk borrowers are more likely to get loans than low risk borrowers. It occurs because institutions that face default risk will price loans to compensate for the risk and these higher interest rates, in turn, will discourage low risk borrowers from borrowing. This could be another factor that would worsen default rates at credit unions if they focus on high risk, relatively poor consumers, and are prohibited from lending to a broader class of borrowers that might include safer borrowers, in particular business borrowers.

The growing evidence that members of credit unions are no different from people who are bank customers would lead one to reject this rationale. Chmura points out that deposit insurance would protect depositors from the risk of their credit union making high risk loans to relatively poorer people, so that the introduction of deposit insurance for credit unions in 1970 eliminated unusual moral hazard as a rationale. Actually business loans are riskier than consumer loans (real estate, vehicle and other loans) so that the higher default risk--deposit insurance relation would apply more to banks than to credit unions, if in fact credit unions were more constrained.

Another reason that is not mentioned in the literature on the common bond is that credit unions, by having a narrow field of membership, might face greater systemic risks associated with concentration of assets among similarly situated borrowers. For example some of the largest credit unions serve employees of airlines, such as United, Delta, U.S. Air, and the plane manufacturing firm, the Boeing Corporation. Industry difficulties due, for example, to downturn in the economy, shifts in demand patterns or other industry specific problems make credit unions likely to face greater industrial or geographic risk than banks in the same areas. Because of a mandated lack of diversification possibilities, credit unions might have a disadvantage compared with banks. Again, however, the existence of insurance is a more effective remedy against such risk than is a tax exemption. Notably, the insurer and regulator, the National Credit Union Administration (NCUA) has an incentive to reduce these concentration risks by allowing multiple bond institutions, and this, in fact, has occurred widely since the 1980s. Before that, such industry-specific and geographically concentrated risks resulted in higher failure rates of credit unions. While diversification reduces risk to the insurance fund, it weakens the justification for a tax exemption based on perceived disadvantages of cooperative enterprise or higher risk.

Today the principal justification for the tax exemption would seem to be that it already exists and, therefore, removing it could adversely impact thousands of institutions and their customers. And these institutions and customers are perceived, incorrectly, to be relatively lower income or associated with the economic security and progress of lower income people. This is a strong argument in terms of practical politics, but it is not supported by fact and is an egregious violation of one of the most fundamental principles of sound tax policy, neutrality. By having such tax discrimination, the exempt sector will be larger than it would otherwise be and more inefficient than the taxed sector, diverting scarce credit resources, in this case, to lower value uses.

The key issue raised by federal income tax exemption is tax neutrality, a fundamental principle of taxation. Taxing some financial institutions that offer the same consumer deposits and loans while not taxing others, in particular credit unions, distorts the allocation of resources. It promotes the employment of deposit and credit resources in the tax-free credit union sector at the expense of their competitors, banks, thrift institutions and finance companies.

II. The Growth of Credit Unions

Credit unions have grown much faster than banks, a fact that often is cited as evidence of their tax advantage. Figure 1 shows the growth rate of credit union and bank real assets since 1973. The excess growth rate at credit unions was especially noteworthy until 1993, but not as apparent since then. Real assets are used because inflation influences the growth rates and was much higher in the 1970s and 1980s when credit union growth was also relatively high.¹⁰ From 1973 to 1993, real credit union assets expanded at a 6.4 percent annual average rate, almost 4 percentage points faster than the 2.5 percent growth rate of real bank assets. Since then, the growth rate of real credit union assets slowed slightly to 6 percent, while real bank asset growth accelerated to 5.2 percent. While credit union asset growth continued to outstrip that of banks, it was only about 0.8 percentage points faster.¹¹

Comparing credit unions with banks must be done with caution.¹² Banks, especially large ones, have a more diverse base of depositors and borrowers than credit unions. On the asset side of the balance sheet, banks are not so dependent on consumers or on their depositors. Business loans are larger and more profitable (and riskier), while business loans have been nonexistent or very small at credit unions.¹³ For credit unions, their members (consumers) are not only their key source of deposits, they are also the borrowers.

However, credit unions are pursuing greater small business lending. Credit unions recently sought and won approval to do business lending through Small Business Administration (SBA) guaranteed loan programs. These loans are not included in business loans for purposes of the legal limit on business loans. Credit unions also advanced the Credit Union Regulatory Improvement Act (H.R. 3579) in the 108th Congress. This act would, among other provisions, raise the limit on business lending to 20 percent from 12.25 percent, double the size limit on such loans that could be excluded from the limit from $50,000 to $100,000, and exclude certain other business loans from the cap. Since 2000, business loans outstanding at credit unions have more than doubled, rising from $4.1 billion to $8.9 billion in 2003, a 28.8 percent annual rate of growth. As recently as 1997, these loans were only $2.9 billion. Banks, which specialize in business loans, actually saw such loans decline from 2000 to 2003. Commercial and industrial loans, farm loans and commercial real estate loans fell 3 percent over the period, or about $47.2 billion.

 

Figure 1

 

 

Real Asset Growth Has Been More Rapid at Credit Unions Until

 

the Mid-1990s

 

 

 

A large bank, under regulatory definitions, is defined to have assets in excess of $1 billion, but only 82 credit unions (0.9 percent) had assets this large at the end of 2003. There were 424 banks, (5.4 percent of all banks) with assets this large. The average size of total bank assets ($968 million) is close to this threshold.¹⁴ The number of large credit unions is growing rapidly, however. At the end of 2000 there were a little over half as many, only 43 credit unions (0.4 percent of credit unions), with assets over $1 billion, while the number of large banks was 397 (4.8 percent), not much different from that at the end of 2003. The number of large credit unions rose 90.6 percent in only three years while the number of large banks rose only 6.8 percent.

Credit unions are also changing their fields of membership and expanding their geographic coverage to become dominant financial institutions in some local markets. For example, Citizens Equity First in Peoria, Illinois, originally the credit union for Caterpillar Corporation and the largest financial institution in the area, has expanded its field to include employees in selected employer groups and people in 14 neighboring counties. It is the sixteenth largest federally insured credit union with $2.6 billion in assets (at the end of 2003). America First Credit Union, in Ogden, Utah, the twelfth largest credit union with assets of $2.7 billion, is also the largest institution in its area. It defines its field of membership to include people who live, work, worship, attend school or volunteer in any of six counties or within a twelve-mile radius of Mesquite, Utah. The pattern of growth and change among the nation's credit unions is characterized by expanded geographic coverage, dominance in some local financial markets, and the adoption of more generic names with no reference to a narrow field of operation. The top ten credit unions by asset size include credit unions based in such small and medium-sized cities or suburbs as: Merrifield, Virginia; Raleigh, North Carolina; Alexandria, Virginia; Sacramento, California; Santa Ana, California; Tukwila, Washington; Tampa, Florida; The Dallas-Fort Worth Airport; San Antonio, Texas; and Manhattan Beach, California. Their assets of $2.9 to $20 billion give them a dominant share in local markets, though many operate state- wide or nationally.

Another feature of credit unions is that there are very large "corporate" credit unions that provide wholesale financial services to their owners, which are smaller credit unions. Credit unions could use the correspondent banking services of larger banks for their deposit and asset services, help with unusually large loans or more specialized banking services, just as small banks do, but banks are taxed and so the prices of their services would include tax cost. To avoid these higher costs of wholesale financial services, corporate credit unions were set up to provide the same services as corespondent banks. At the end of 2003 there were 31 such institutions holding $108.8 billion in assets. The largest and oldest, U.S. Central Credit Union in Lenexa, Kansas, holds $35 billion in assets and is owned by only 72 member credit unions. These assets are largely passed through credit unions, so adding them to the credit union assets in Table 1 would be double counting. However, the earnings of these credit unions are also untaxed and no allowance has been made here or elsewhere for these untaxed assets in estimates of the tax loss from federal income tax exemption.

The main reason for comparing credit unions with banks, however, is that banks are in the same competitive market for credit union depositors and also, though to a lesser extent, for credit union borrowers. In terms of a "level playing field," it is like-sized banks that are the most disadvantaged by the tax exemption available for credit unions of similar size. Banks might grow more slowly than credit unions for a variety of reasons that do not depend on the tax advantage of credit unions. For example, since banks serve business borrowers and business and government depositors, different growth rates in their demands for funds or for deposits relative to consumer demand could result in different growth rates. This issue is not addressed here. Instead the difference in growth rates is offered simply as evidence consistent with a cost advantage at credit unions.

While credit union assets have grown faster than bank assets in the past, the number of credit unions is declining, just as is the number of banks. For example, from 1999 through 2003 the number of credit unions declined from 10,626 to 9,369 (-11.8 percent, an annual rate of decline of 3.1 percent). But this consolidation belies the vibrant growth of the sector. Besides rapid asset growth overall (48.3 percent), there was a large increase in credit union membership from 75.4 million to 82.4 million (9.3 percent or 2.2 percent per year). Bank assets grew by less, a still robust 33 percent, but the number of insured commercial banks fell 6.6 percent to 7,769 at the end of 2003. The consolidation of credit unions is proceeding faster than that of banks. The number of credit unions above $1 billion in assets more than doubled between 2000 and 2004, from 43 to 93.

The consolidation in the number of credit unions is occurring at both federally and statechartered institutions, with many large, aggressively growing credit unions swallowing up small, traditional credit unions. The decline in the number of federally chartered credit unions since 1999 has been 12 percent, only slightly faster than the 11.6 percent decline in the number of statechartered credit unions. Federally chartered credit unions are no longer growing much faster than state-chartered ones. At the end of 2003, 61.7 percent of the 9,369 credit unions were federally chartered, still about the same percentage as at the end of 1999, despite their slightly faster pace of decline. However, asset growth at federally chartered credit unions has slowed relative to others, just as overall growth of credit union assets has slowed relative to bank assets. Figure 2 shows the relative size of federal credit union assets compared with assets of banks and state-chartered credit unions.

 

Figure 2

 

 

Taxed Banks and CUs Grew More Slowly Than Federally-Chartered

 

CUs Until the 1990s

 

 

III. The Sources of Competitive Advantage: Sponsorship and Tax Exemption

The academic literature on credit unions has paid little attention to the federal income tax exemption of credit unions. Instead, the emphasis on credit union competitive advantage has focused on sponsorship (see the review of the literature in Emmons and Schmid (2001), for example). The typical or traditional common bond of credit unions has been either a single occupational bond or an associational bond. In both cases, the company, union or workers sponsor the credit union and provide valuable resources. In a fair tax system, sponsors would account for such subsidies as grants and not as tax-deductible expenses and credit unions would show these subsidies as income. But sponsors are likely accounting for these subsidies as tax-deductible expenses, reducing their own taxes, and compounding the inequities in the tax system. Again in an equitable system, these benefits from sponsors would be taxable income that is not currently measured in credit union net income.

Occupational credit unions are the most numerous. For example, Emmons and Schmid (2001) point out that in 1996 about 75 percent of credit unions had an occupational common bond. One example of the benefit of sponsorship is free or subsidized services. In 2003, there were 2,003 credit unions (21.4 percent) that had an annual occupancy expense of zero or less. Another onethird (34.1 percent) had occupancy expenses of less than $2,000 per month, so that more than half of all credit unions (55.5 percent) had average monthly occupancy costs below that level. Another 11 percent had occupancy expenses ranging from $2,000 to $4,000 per month and 11 percent had rents from $4,000 to $8,000 per month. Thus, 77.5 percent of all credit unions had occupancy costs less than $8,000 per month. The importance of tax exemption in holding down cost has not been the focus, despite the fact that it would swamp the value of all other subsidies of credit unions.

Because of the focus on sponsorship, a key issue in credit union research has been why firms sponsor them and who benefits, with the latter question focusing on whether borrowers benefit through lower loan rates or depositors gain through higher deposit rates or some combination of the two. A related issue is who controls a credit union: the sponsor, members (either as depositors or borrowers), or the market. Another important tax policy issue is the incidence of the cost advantage that credit unions have whether arising from sponsorship, free services from members or tax exemption. In addition, a cost advantage could affect the market for competing providers, especially banks and thrifts. Before taking up the question of incidence however, it is important to assess the value of the tax exemption.

IV. How Large Is the Tax Loss and How Big is its Implied Advantage for Credit Unions?

The easiest way to estimate the tax loss is to estimate the tax that would have been paid if net income were subject to federal corporate income taxes. Based on data for all insured credit unions, net income in 2003 was $5.8 billion and the computed taxes would have been $2 billion. The effective tax rate would have been 33.97 percent. Government budget estimates often include a five-year revenue loss estimate and a ten-year revenue loss estimate. For the period 2004 to 2008, the five-year revenue loss estimate is $12.6 billion and the ten-year estimate is $31.3 billion. These losses indicate a much more substantial subsidy and tax revenue loss than existing government estimates.¹⁵

Chmura (2004) estimates a tax loss of $1.9 billion for 2002, based on a tax rate of about 33.3 percent, and net income was $5.7 billion in 2002. Applying the tax rate to the slightly higher net income in 2003 would result in an estimated tax loss of the same $1.9 billion in 2003.¹⁶ The smaller estimates by Chmura Economics and Analytics and by government sources cited below probably arise because they do not take into account firms that have no earnings or negative net income and therefore would pay no federal income tax. However, this source of difference cannot explain much of the difference with government estimates because the losses are small. In 2003, 12 percent of all credit unions fell in this category and reduced the total net income of insured credit unions by $117 million. Netting their losses against all income would understate the potentially taxable net income of credit unions and understate the total tax liability.

The Joint Committee on Taxation (JCT) regularly estimates the "tax expenditure" associated with the federal income tax exemption.

A tax expenditure is the revenue lost because of a tax break, computed assuming that the beneficiaries of the tax break would not alter their behavior in its absence. It is comparable to the estimate here or in Chmura. The JCT's latest estimate for fiscal year 2004 is $1.2 billion, according to Bickley (2003). Table 2 shows recent historical data for the tax loss from JCT publications.

                         Table 2

The Office of Management and Budget (OMB) in the Office of the President also estimates the tax loss due to the federal tax exemption of credit unions. Their estimates are similar and follow a similar methodology as the JCT. For example, in the Federal budget, OMB estimates that the tax loss was $1.3 billion in both 2003 and 2004.¹⁷ The OMB five-year tax loss estimate for 2005-2009 is $7.88 billion, which is far smaller than our estimate of $13.66 billion for the same period.

Based on an assessment of all credit unions, however, the tax loss is substantially larger, about $2 billion. This is over one- third of net income of credit unions, so the repeal of the tax exemption would add about 50 percent to the costs of credit unions, especially larger ones (credit unions with over $75 million of current net income, in this case, because they face marginal tax rates of 34 to 39 percent).

Chmura (2004) also suggests that their estimate is likely to be too high, largely because credit unions would manage themselves to reduce taxable income, if, in fact, they were subject to federal income taxation. For example, banks tend to use provisions against bad debt to reduce income and boost capital reserves in order to cut taxable income. If credit unions were taxed, they would likely boost their debt provisions to lower taxable income as well.

However, depending on the incidence of the tax, explained in Section V below, credit union taxes might be even higher than the estimates here. For example, credit unions, faced with a higher cost of capital, might be able to raise their interest rates or lower deposit rates to cover the higher cost. To the extent that credit unions were able to keep deposits and loans while passing along their higher taxes, gross and taxable income would be larger than currently projected and therefore their tax payments would be as well.

If credit unions did not have the ability to pass on the higher taxes, credit unions would tend to be smaller and some could be forced to close. In this case, the taxable income of credit unions would be lower than current net income and the estimated tax gain from credit unions reported here would be too large. However, in this case, financial services business would switch from credit unions to banks, thrifts and finance companies, raising taxable income and tax liabilities there. In fact the tax receipts from those larger institutions would be higher on the same income transferred from credit unions because they are generally in higher tax brackets. Thus, removing the federal income tax exemption for credit unions could raise more tax receipts than the estimates here because their behavioral changes would increase the tax base at credit unions or at their competitors, whether the credit union sector shrinks or remains as large as it is today.

V. The Incidence of the Tax Loss -- Who Gets the Tax Subsidy?

Who gets the $2 billion of annual tax subsidy? The answer to this question is important because it reveals who benefits from the tax exemption and who is the key constituency for maintaining it. It also reveals who does not gain from the subsidy, though they may be the intended beneficiaries. And are there effects of the tax exemption on credit union competitors? The possible beneficiaries are:

• credit union depositors (higher rates paid on deposits),

• credit union borrowers (lower interest rates on loans),

• credit unions managers and/or workers (higher wages or other payments to managers, workers or the suppliers of services to credit unions),¹⁸ or

• credit union owners could benefit from larger retained earnings and capital of credit unions providing support for faster growth as well as a larger cushion against risk or mismanagement.

If market prices for deposits, loans or resources used in credit unions are affected, then it is also possible that the effects of tax exemption on credit unions also affect their competitors. Amel and Hannan (1998) argue that non-local banks and non-bank firms, such as thrifts and credit unions, have no effect on bank deposit rates. Moreover, in regulatory analyses of banking competition for merger or other antitrust issues, credit unions are routinely ignored, implicitly assuming that credit unions do not affect pricing of banking services. Yet, as Emmons and Schmid (2000) emphasize, banks continue to oppose the legislated advantages of credit union charters.

Incidence of the Tax Exemption

Figure 3 can be used to summarize the incidence effects of tax exemption. A basic model of the loan market is shown there. The analysis focuses on loans, but it applies more generally to total assets and these terms are used interchangeably in applications of this analysis below. The demand for credit union loans shown implies that credit unions can make more loans if the interest rate on loans is lower. If the loan rate were set in a broader market (including banks, thrifts, loan brokers and finance companies, say) then the demand for loans would be horizontal at a market-determined interest rate and the position of the supply curve for loans would determine the quantity of credit union loans. Two supply curves are drawn that differ according to whether credit union income is taxed. The higher supply price corresponds to the supply price of loans if credit unions are subject to federal income tax.

The supply curve in each case is upward sloping and includes the cost of producing an additional dollar of loans. It includes the additional operating expenses and the additional cost of deposit funds and net worth in order to fund the dollar of loans. If any of these costs rise as loans expand, then the curve is upward sloping. The interest cost of deposits could rise as loans rise because a higher deposit volume could require that credit unions offer higher deposit rates. The latter component of cost could be unaffected by the size of credit union assets if credit unions are a relatively small part of the deposit market and essentially take the deposit rate determined by banks and thrifts as given. This is likely and supported by the data here since credit union assets are little more than one-tenth the size of bank assets. There is some evidence, noted below, that credit unions may pay higher rates to fund larger loan volume and if that is the case the slope of the credit union loan supply curve is steeper than it would otherwise be.

Note that the supply curves in Figure 3 are based on long-run supply price. The distinction here, as opposed to short-run supply, is that it includes the cost of capital since capital is variable in the long run and has to be compensated at the market price of capital in order to remain in this sector instead of moving to other business. Moreover, the supply of loans or assets in credit unions requires that firms hold more equity as a buffer against expected and unexpected losses as their assets increase. Economic analyses often focus on the short run, where capital or equity may be treated as fixed, and conclude that an income tax cannot affect output or price. For financial institutions, capital is not fixed even in the short run. Since income, in the long run, includes a competitive return to owners this element of cost must also be included in the supply price and changes in tax rates affect the cost of equity capital and therefore the supply price.

Higher taxes raise the cost of equity capital or the amount that has to be earned to compensate the owners. More loans require that credit unions hold larger capital and capital has a required rate of return, so that the supply price of loans include this cost of equity capital. Taxation of the return to equity would raise this cost, shifting the supply curve upward and to the left as shown in the chart. If the tax rate is about one-third, then this component of the supply price would rise about 50 percent.¹⁹ In recent years the rate of return on assets, which equals the weighted cost of equity, has been about one percent (see Table 1), so that taxing credit unions would raise this cost by 50 basis points. This is the extent of the tax subsidy per dollar of assets.

 

Figure 3

 

 

The Effect of a Tax on Credit Union Income

 

 

 

On the other hand, if the demand curve for loans were flat, the tax increase would not affect loan rates. In this case, loan value and desired retained earnings and capital would fall more, and loan rates would not rise. So long as loan volume falls, deposit rates could decline if their volume affects credit union deposit rates. In this case, the rise in supply price could be partially offset by a shift down in the cost of deposits. This would lessen the effect of the removal of the tax exemption on loan rates and on the volume of loans. The extent to which the removal of tax exemption would raise loan rates, reduce deposit rates, or affect loans or desired capital depends on how flat the demand or supply curves are.

Another way that loan interest rates would be unaffected by tax exemption is if competition forces the subsidy to be passed on to depositors or to managers or other suppliers of credit union resources. Such a rise in cost would offset the tax subsidy, shifting the "no tax" supply curve in Figure 3 back up toward its "with tax" position. In this case the tax exemption would be absorbed by other resource costs, and if it is completely passed on in this way there would be no net effect on loan volume or the size of the credit union sector due to the tax exemption. There is evidence on these issues that can inform an answer to the question of who gets the $2 billion and how it is passed to them.

Evidence on the Incidence of the Tax Exemption

Early studies of credit unions explored whether credit unions paid higher deposit rates or charged lower loan rates. Other potential incidence effects, such as higher compensation or other costs, lower sponsor outlays or larger net worth, were ignored. The results of these studies are mixed. Flannery (1974) argued that the evidence supported lower loan rates. Emmons and Schmid (2000) point out that a later study found that deposit rates were higher, but a 1986 study found that both borrowing and lending rates were not different than at other institutions. Many studies over the years, including several cited by Emmons and Schmid (2000), support the hypothesis that the more diffuse ownership of credit unions has led to higher expenses at credit unions, an indication of what is called "expense-preference theory." This is the notion that managers at institutions without strong discipline from stockholders and the capital markets will maximize their own well being instead of profits by paying themselves high wages or other less obvious perks. Such an agency problem is especially acute at nonprofit institutions or cooperatives, where equity does not trade and managers may be able to replace the owners as the residual claimant of profit.

Some analysts have found some effects of credit unions on local pricing of bank services, however. Tokle and Tokle (2000) find that local credit union market share of deposits has a significant positive effect on deposit rates (significant for one- and two-year certificate of deposit (CD) rates, but not regular saving deposits, which are called shares at credit unions). This study is based on a small sample in rural Nevada and Montana in 1998, however. More recently, Hannan (2002) found that the credit union deposit-market share has a significant positive effect on deposit rates at banks for the three rates he studied: checking deposits, money market deposit accounts and three-month CDs, respectively. His study focuses on banks in 100 metropolitan areas in 1998, using data from Bank Rate Monitor. Emmons and Schmid (2001) also provide evidence that higher loan demand at credit unions raises deposit rates at credit unions. They take this as evidence of a "sponsor control" hypothesis, however, as opposed to member control, either by dominant credit union borrowers or dominant credit union depositors. The more important point is that it also suggests that deposit rates at credit unions are not determined in competitive deposit markets with banks. Some evidence below questions that implicit result, however.

Few studies have looked at loan rates at credit unions and whether they affect comparable rates at banks. However, Feinberg (2001 and 2002) finds that the credit union market share of deposits has a significant negative effect on new car loans in a study of six years of data (1992-97) for a panel of banks. That is, the larger is the share of credit unions in the deposit market, the lower are new car loan rates at banks. This suggests that the taxexempt credit union market share depresses new car loan rates at all financial service institutions.

Higher Loan Rate or Lower Deposit Rates?

If the incidence of the tax exemption fell on depositors or shareholders, who in this case are the same people, credit unions could be expected to pay higher rates on deposits than non-exempt institutions (banks and thrifts).

Table 3 shows selected rates from a survey by Bank Rate Monitor at the end of 2003. The evidence is mixed for loan rates, but the only deposit rate sampled, that on one-year CDs, is sharply higher at credit unions than at non-exempt institutions. For loan rates, credit unions were charging a slightly higher rate on 30-year mortgage loans, but less on one-year adjustable rate mortgages and much less on home equity loans. The largest difference was for new car loans where credit unions charged much lower rates. This survey suggests that the incidence of tax exemption falls on both borrowers and depositors at credit unions, but relatively more heavily on loan rates, especially car loan and credit card rates, than the more competitive rates on real estate loans, especially first mortgages, or deposits.

If the tax treatment of income affects the pricing of loans at credit unions, it should be apparent in the pricing of loans and deposits at federal and state-chartered institutions. Statechartered credit unions are not exempt from state taxes, while federally chartered institutions are. If there is some incidence of taxes on loan interest rates or on rates paid on deposits at credit unions, the differences should show up in differences in the average interest rates on different types of loans or deposits for these two groups of credit unions. Federal credit unions would be expected to charge lower interest rates as competition induces them to pass along some of their tax saving to borrowers. Federal credit unions could also compete by offering higher rates on deposits than state-chartered ones. The NCUA reports interest rates paid and charged at each type of institution since 1992 in their on-line statistics (2003) for various periods (see Table 4).

                         Table 3

The bottom entries in the table show rates paid by the two groups of four types of deposits. They are identical, so the incidence of tax exemption does not appear to fall on deposit rates either. Again, the same pattern of equal rates shows up in earlier data. The incidence of the tax exemption does not show up in higher deposit rates at federal credit unions.

The pattern shown in Table 4 holds over time as well. In the Appendix, the raw data for these interest rates from 1992 to 2003 are shown in a longer table. The differences between the interest rates at federal and state-chartered institutions are show in Table 5. Where the rates differ, they are shown in bold when they conform to the tax incidence hypothesis, that is, the rates on loans are lower and rates on deposits are higher at federally chartered institutions. Note that most of the entries show no difference. For deposit rates, only two entries are consistent with the incidence hypothesis and nine other non-zero entries are of the wrong sign. The mean difference for the deposit rates is zero, which is consistent with the argument that credit union deposit rates are set in a broader deposit market in which credit unions make up a small market share.

For loan rates there are also more differences that are non-zero and consistent with the taxincidence hypothesis, about 30 percent of the observations, but there are slightly more pairs of observations that have the wrong sign. Only credit card rates are lower at federal credit unions in more than two years and have a mean that is negative for the whole period. This pattern is consistent with the argument above that loan rates at credit unions are not different than in the taxed sector of the financial industry because credit unions are a small part of a very competitive market. For car loans and credit cards, there is more diversity and indeed, the credit card rates appear to be systematically lower at federal as compared with state- chartered institutions. Subjecting all credit unions to the federal income tax is not likely to affect market rates on credit cards outside of credit unions and will have little effect on overall loan rates at credit unions. Credit card loans account for only 5.8 percent of credit union loans and even adding in all other unsecured credit union loans raises the share to 10.4 percent of loans and 6.6 percent of total assets. Thus even if all of the incidence of the $2 billion tax break for credit unions (50 basis points as a percent of assets) fell only on these loans, removing the tax break would have little effect. The supply price of assets would rise about 5 basis points (6.6 percent of 50 basis points).

                         Table 5

Are Employment Costs Higher at Credit Unions?

Emmons and Schmid (2000) also provide a useful review of the literature on credit unions efficiency, noting in particular, that Fried, Lovell and Vanden Eeckaut (1993) find that credit unions exhibit the same degree of widespread operating inefficiency as found at other depository institutions. Employment costs make up about half the operating costs of credit unions and about one-fourth (25.9 percent) of total credit union gross income or cost. Relative to total assets in 2003, this is about 1.54 percent.

Chmura (2004, p. 33) compares the employee compensation costs of credit unions and banks. For six asset classes, ranging from under $2 million to over $500 million of assets, Chmura finds that credit unions' employee compensation cost per dollar of assets is about 11 basis points higher, on average, and higher in all but two asset classes of credit unions. The two exceptions are institutions with assets from $10 to $50 million, where the share is smaller at credit unions by one basis point, and over $500 million, where the share of credit unions is smaller by eight basis points. The highest levels of excessive employee compensation cost occur for credit unions in the $2 to $10 million class where compensation is 26 basis points higher than at comparably sized banks (1.86 percent vs. 1.60 percent). The other is a group of large credit unions those with assets of $100 to $500 million, where credit unions have compensation that is larger by 30 basis points of assets (1.69 percent vs. 1.39 percent). Generally then, it appears that credit unions may have slightly higher employee compensation costs, but there are exceptions and they are not clearly linked to asset size. For the next to smallest and the next to largest asset size classes, credit unions appear to have smaller employee compensation costs. While credit unions appear to have higher employer costs, perhaps reflecting that the tax exemption benefit is partially being captured by the managers and workers in credit unions through higher wages or lower productivity, the difference is not large nor does it vary systematically by size of credit union. While the lower employee cost at the largest credit unions suggests that bank competition may constrain employee costs more effectively, the effect does not continuously diminish with the size of credit unions, casting doubt on this channel of incidence.

Emmons and Schmid (2000) find that there is a positive relation between relative wage levels at credit unions (compared with average local wages) and asset size for occupational credit unions sponsored by employers, at least for credit unions with assets up to about $65 million. This was the average size credit union in 2003. They interpret this to mean that the cost of financial distress rises with asset size and so the sponsor of the credit union pays higher wages to induce managers of the credit union to shirk less. One alternative that they do not rule out, however, is that credit union size, per se, raises the cost of a sponsor monitoring the management, or what is called "agency cost," so that the agents (management) can get away with higher wage premiums. More important, any wage premium declines with asset size for larger institutions, according to their evidence, and this supports the notion that competitive pressures and superior management at large credit unions controls shirking without bribes or premiums from sponsors. It is not likely then that the tax subsidy for credit unions has been absorbed by significantly higher employee and management costs.

Does the Tax Break Accrue to Owners Through Retained Earnings, Larger Assets and Net Worth?

A final possibility is that the tax break is absorbed by the expansion of credit union assets to a size that can be supported by the enhanced excess earnings of the credit unions. A tax break initially raises the rate of return on assets and equity, given the volume of loans, assets and equity. More importantly, it shifts the supply price of loans down or to the right because the cost of equity capital falls.

When the loan demand is flat at an interest rate determined in broader loan markets (including banks and other financial institutions), the incidence of the tax break shows up in the size of credit union loans (assets). In the absence of a decline in loan rates, an increase in deposit rates, or a change in the other costs of new loans, the supply of loans would tend to expand, so that the new supply of loans would equal i0 in Figure 3 at a larger quantity of loans and equity. Equity will rise due to the increased loan (asset) supply and because the equity ratio will rise in response to the higher rate of return on equity.

Given the required rate of return on equity, the cost of equity and of assets (loans) will rise as the equity ratio, the ratio of net worth to assets increases. This brings the supply price of loans back to its initial level at i0 in Figure 3, but at a larger supply of credit union assets. In this case, the incidence of the tax break is confined to the largest possible increase in the size of credit union assets and equity.

Again, in this case, removing the tax break would shift the supply curve up and lower the size of credit union assets and loans. At the initial supply of loans, L0, the supply price rises 50 basis points if the initial rate of return on assets is one percent. Loan interest rates would not be affected, however, nor would the interest rate on deposits (if it is also determined in a competitive deposit market with banks and thrifts and where credit unions are a relatively small share of the market), nor would the other costs of credit unions. The maximum size effect of the removal of the tax exemption is to lower assets and to lower the net-worth-asset ratio by one-third. The ROA would initially be lowered by the tax increase and the decline in the net worth-asset ratio would further depress the ROA of taxed credit unions. Ultimately the ROA would decline to the same extent as the equity ratio, one-third under the assumptions here, but regulatory constraints could keep the equity ratio from declining this much.

Under these assumptions the incidence of the tax exemption is to raise credit union assets. The rate of return on assets would rise due to the tax exemption as well, because of the rise in the equity ratio and its effect boosting the cost of capital. Reversing this process indicates the effects of removing the tax exemption. To the extent that the demand for loans is not horizontal or that other costs are not raised by the tax exemption, the removal of the tax exemption would not reduce assets as much.

To examine whether the incidence of the tax exemption falls on the size of the tax-exempt institution, a comparison of rates of return and equity ratios at federal and state-chartered credit unions could be helpful. If a tax break is made available to federal credit unions only, then it would be reflected in an initial rise in their rate of return on assets and assets. Competition would lead these credit unions to expand assets and the equity ratio with the equity generated by higher earnings. In the end, the tax-favored credit unions would expand relative to the less favored credit unions, the rate of return on equity would be unchanged, but the rate of return on assets and the equity ratio would be higher at federal credit unions than at state-chartered credit unions.

 

Figure 4

 

 

The Rate of Return on Assets at More Heavily Taxed State Credit

 

Unions Is Not Lower

 

 

Rates of Return at Federal and State Chartered Credit Unions

 

 

 

The absence of equality of long-run ROEs means that it is not so likely that the two key incidence conclusions will hold either. First, the federal ROA is higher, but not by enough to suggest that the incidence of the tax-exempt status falls on size alone. In 2003, the Federal credit unions had an ROA of 1.03 percent, 8 basis points above that of state-chartered institutions. However, from 1998 to 2003, the average ROA was 0.99 percent at both sets of institutions. Thus the ROA evidence does not support the notion that the incidence of tax exemption falls on the relative size of the institutions.

The equity ratio implication fares somewhat better, but not by enough to support the idea that incidence only affects asset and equity size. The equity ratio (not shown) of the federal credit unions is not much larger. Since 1997, the mean equity ratio was larger, but not by much. For the past six years, the average federal equity ratio was 11.06 percent, about 23 basis points higher than at state-chartered credit unions. In 2003, the federal equity ratio was 10.80 percent, only 28 basis points higher than the 10.52 ratio for state-chartered credit unions. One possible reason is that the tax burden at state-chartered credit unions may not be large enough to make a difference between the two types of institutions.²⁰

Another consideration is that equity ratios are determined not only by the economics of the institution, especially its tax status, but also by regulatory constraint. Regulators evaluate the equity ratio as a positive factor in determining the performance of credit unions. Thus equity ratios do not simply reflect the influence of economic factors such as the exemption from state taxes. The equity ratio at state-chartered institutions could be higher than expected because of regulatory pressures to hold it up relative to where it would be if tax considerations alone were taken into account. That would, in turn, require the ROA for state-chartered institutions to be higher as well to hold up the ROE.

If the tax break did not fully affect the size of the tax-free sector, then the loan rate would be lower and/or operating costs or deposit rates would be higher to offset the tax break at an unchanged asset size. In all three cases, assets, or the size of the tax-free sector, would be less affected by the tax break. When the rate of interest on loans is depressed or the non-interest operating cost is raised, assets are not as affected, but the equity ratio still rises because the cost of equity capital is reduced relative to the cost of deposits. When the tax break is passed on in the form of higher deposit rates only, there is little or no effect on the equity ratio or asset size. In this case the cost of both deposits and of equity capital rise, providing little or no incentive to boost the equity ratio or assets.

The evidence above for loan rates, especially for credit card and auto loans, suggests that incidence could fall somewhat more heavily on loan rates (about 6 out of 50 basis points according to the analysis there). To the extent that tax incidence falls on deposit rates or other costs [no more than 11 basis points according to the Chmura (2004) evidence above] the supply price rises to offset its reduction due to the tax break. The supply curve in Figure 3 does not shift down as much and credit union assets and equity are not boosted as much by the tax break. Similarly in these cases the removal of the tax break would not reduce the size of the untaxed credit union sector as much.

This evidence indicates that the effect of the subsidy does not fall on the equity ratio or size, which is surprising in light of other evidence. For example, research on government sponsored enterprises (GSEs) shows that the effect of the public perception that debt of these institutions is backed by the full faith and credit of the U.S. government falls largely on the companies (assets or equity ratio) and not on their interest rates or overall operating costs.²¹

There is also some evidence that the effect on the asset structure of credit unions is the dominant one. A comparison of credit unions with banks shows that the incidence of credit union tax exemption falls largely on the equity ratio of the credit union sector. In 2003, the equity ratio of all commercial banks, at 9.10 percent, was well below the 10.7 percent rate for credit unions (see Table 6). This 17.8 percent larger equity ratio is consistent with the tax incidence analysis which suggests that credit unions should have a much larger equity ratio than banks if their tax exemption mainly affects their equity size. While the equity ratio at credit unions is not large enough to justify a conclusion that incidence falls completely on the size of the equity ratio, it does suggest that this is the major effect.

However, banks are riskier than credit unions because they are much more dependent on loans to the more cyclical and secularly risky business sector.²² Thus even on a long-run basis, one might expect banks to have a higher ROE than credit unions. Equality of ROEs applies to equally risky firms; riskier firms should have a higher ROE to compensate for greater risk, so that only the riskadjusted ROEs are equal. On a risk-adjusted basis, the equity ratio at credit unions would be even larger than the actual difference. The smaller equity ratio at banks may also reflect size differences. Larger financial intermediaries are more diversified and hold smaller equity ratios. For example, banks with assets below $100 million held equity equal to 11.27 percent, less than the 11.87 percent held by credit unions in the same size class.

The latest year's data underestimate the usual difference in these ratios. In 2003 the average credit union equity ratio was only 10.72 percent. As Figure 5 shows, however, the equity ratio for all credit unions was unusually low and that of banks relatively high in 2003, 17.8 percent larger than that of the average bank ratio. For the past ten years shown in Figure 5, the credit union equity ratio was 26.6 percent larger than at banks. Credit unions consistently have higher ratios and usually ratios that are much higher than in 2003. In fact, the average excess for the past ten years accords very well with the maximum incidence of tax exemption on the equity ratio, especially when one considers that riskier banks would choose to hold higher ratios even if the tax treatment were the same.

                         Table 6

Figure 5

 

 

Credit Unions Have a Higher Equity Ratio Than Banks

 

 

Credit Unions Hold Higher Ratios of Net Worth to Total Assets

 

 

 

Another strong indication of the effect of the credit union tax subsidy is the chartering pattern of credit union and banks. There are many factors that influence whether a bank or credit union obtains its charter from the state or from the federal government chartering institution (the Office of the Comptroller of the Currency or the NCUA, respectively). There is no difference in the tax treatment of state-chartered and federally chartered banks, but there is a difference for credit unions. Since federally chartered credit unions have more tax advantages than state-chartered credit unions, there should be relatively more of them, given other factors that influence this decision. And there are relatively more of them. For banks at the end of 2003, 30.6 percent (2,814 out of 9,181 banks) were federally chartered, while 61.6 percent of credit unions were federally chartered at the same time, about twice the share of federal bank charters.

A key difference between banks and credit unions is that bank shareholders can make asset decisions that are almost completely separate from their demand for banking services. People who bank at one institution can be stockholders in another, and shareholders in a bank do not have to be depositors or borrowers. Credit union shareholders cannot sell their ownership claims. This puts their equity claims under the full control of management, and this can lead to capital market inefficiency. Active capital market evaluation contributes to enhanced efficiency at banks, holding down the equity ratio and boosting ROA in order to raise the ROE. It also provides strong incentives against sponsor abuse or managerial abuse of owner interests.

While the evidence on the incidence of the tax break is mixed, this largely reflects the equality of data on credit union costs and loan rates. Most of the analysis and inferential conclusions here support the view that the principal benefit of the tax incidence is to boost the size of the credit union sector and not to lower loan rates or boost deposit rates for credit union customers.

Summary

While the tax expenditure arising from credit union tax exemption is large and growing, data limitations, especially on comparable interest rates and cost, preclude a definitive estimate of the incidence of the tax. The overall effect of the tax exemption can be measured as the rise in the rate of return on assets, before taxes, that would be required to keep the rate of return on assets, after tax, unchanged. With a tax rate of one-third and average rate of return on assets of one percent, the ROA before taxes in the absence of the exemption would have to be 1.50 percentage points, or 50 basis points higher. This is the size of the subsidy (per dollar of assets) that accrues to the beneficiaries of the current tax exemption.

Research on GSEs and credit union research support the view that the principal incidence of tax exemption is an increase in the total assets and equity ratio of the credit union sector. However, direct evidence comparing more or less highly taxed credit unions that are all exempt from federal income taxation is not very supportive of this conclusion. Evidence on other incidence effects is also not strong except for some evidence that credit unions charge lower interest rates on loans, especially on those loans where efficient national markets do not constrain credit union pricing, most notably credit card loans and to a lesser extent car loans.

Based on the evidence analyzed here, the conclusions are that some 6 basis points of the 50 basis point subsidy accrues to credit union borrowers through lower interest rates, while no more than 11 basis points are absorbed by higher labor costs. There is little or no effect on deposit rates or other costs. Thus, 33 to 44 basis points of the subsidy accrue to owners in the form of larger equity and larger assets. No allowance has been made for the reduced credit union costs of sponsors, but to the extent that these are large it is a transfer within the firm and has little or no effect on the size of credit unions or tax receipts. Ending tax exemption could be expected to reduce the relative size of credit unions and to reduce their growth rate. Existing taxpaying financial service firms would gain market share, absorbing the lost market share of credit unions and boosting tax revenues even more. The equity ratio of credit unions would be reduced; management of capital costs would make credit unions more efficient, perhaps lowering operating costs and interest rates on deposits and raising rates on loans, at least in some markets. Credit unions would be more subject to market control and would manage risk and return more efficiently, increasing the value of their franchises to their owners, despite smaller relative size and slower growth.

VI. Proposals for Change

Abolish the Exemption or Leave it Only for Small Credit Unions?

One proposal for tax reform that often arises is to leave the tax exemption for small credit unions and abolish it for large, complex credit unions. One specific proposal is to restrict the federal income tax exemption to institutions with total assets of less than $10 million.²⁴ The rationale for this proposal is that only the larger credit unions are similar and competitive with banks and other financial intermediaries, at least from a household or consumer standpoint. In addition, smaller credit unions generally have traditional or narrow common bonds and are more likely to serve the purpose of the cooperative credit union movement. Bickley (2003) and the Joint Committee on Taxation estimate that the $1.2 billion tax expenditure or subsidy to credit unions in 2004 would have been cut to $1.0 billion if only those with assets greater than $10 million are included. This suggests that about $200 million in lost tax revenue is due to the tax exemption of small credit unions. As noted above, however, the tax loss for all credit unions is closer to $2 billion and removing the exemption for credit unions with assets bigger than $10 million would represent the bulk of any tax gain, according to this study's estimates.

The OMB estimates that the ten-year revenue gain for 2004 to 2013 from taxing the credit unions with assets in excess of $10 million is $12 billion. The five-year estimate (2004-2008) is $5 billion, including $0.6 billion in 2004 and $1 billion in 2005. The 2004 estimate is smaller because it is assumed that the proposal is implemented for only half of 2004.²⁵ Our estimate is $12.5 billion for the 5-year period and $31 billion for the ten-year period. The estimates are not much different than the tax revenue gain from taxing all credit unions.

This proposal would leave the exemption for very small credit unions, but it would restore neutrality for larger credit unions. This is called a practical solution because only about half of credit unions (51.1 percent) had total assets in excess of $10 million in 2003. So only 4,792 would become taxable and of these, nearly 6 percent had no income in 2003, so only about 4,512 would have had a tax bill, based on reported net income. However, taxing these credit unions, 48.1 percent of all the credit unions in the country, would tax 98.1 percent of the net income of all credit unions and collect 98.9 percent of the revenue from taxing all of them. Thus, it appears that such a proposal would yield $2 billion as well, subject to rounding error, but would impact fewer than half of all credit unions.

Only $22 million in additional revenue would be raised by applying the corporate tax rate to the net income of credit unions with assets under $10 million. And only $120 million more of net income would be taxed if all credit union net income were taxed instead of the income of the larger institutions. This is much lower than the $200 million gain indicated by the JCT -- about one tenth as much, but the reasons for the difference are not clear. While the estimate here makes the "practical solution" more appealing because of its negligible revenue loss, it still results in a failure of neutrality for small credit unions and other firms. Genuine tax neutrality would require taxing all net income of all credit unions, without regard to size.

Therefore, excluding taxation of small credit unions simply because they are small is largely a policy choice and would doubly compensate them for being small. The income tax system already accounts for size, taxing firms with lower income at lower marginal tax rates. For example, firms with taxable income less than $50 million pay a 15 percent rate. Over 35 percent of credit unions would have faced this rate in 2003. For firms with incomes from $50 million to $75 million, the marginal tax rate is 25 percent, applied to income over $50 million. Some 6.5 percent of all credit unions would have faced this marginal tax rate in 2003. Adding in the 12.8 percent of credit unions that had no net income in 2003, brings the total to over half of all credit unions with a net income of $75 million or less. Larger credit unions, at least with incomes over $75 million, face marginal tax rates from 34 percent to 39 percent. So most credit unions would face lower tax rates than the larger ones, even if all credit unions were subject to the federal income tax, just as firms in other industries are. Is there any reason why generally small, low-income credit unions should pay no federal income tax when small firms in other industries do?²⁶

Neutrality would require that the tax rates of all small firms, not just small credit unions, would have to be lower. This could be accomplished by creating a tax threshold for net income above zero. Then all small firms could face a more favorable tax rate, not just credit unions. For example a zero tax bracket up to $50 million would exempt almost half of all credit unions and similarly situated firms in the rest of the financial sector and other industries as well. But of course this scheme conflicts with the notion that all firms with net income should pay taxes, even if they face a different marginal tax rate because of size.

Another proposal that has arisen to correct the unequal treatment of credit unions is to extend the Community Reinvestment Act (CRA) to credit unions. This would clearly provide more accountability for credit unions to ensure that they serve their communities, especially their lowincome individuals, in a manner proportionate to their tax-exempt status. Banks currently are subject to extensive costs to insure that they are meeting the credit demands of low-income borrowers. Credit unions were excluded from these provisions because of the presumption that they must be serving such consumers. After all, their charters are rooted in common bonds that seem to assume that credit unions meet these requirements. But the evidence shows that credit unions do not serve the poor to any greater extent than banks. For example, most credit unions have an occupational bond that requires members to be employed, often in industries with relatively high wage jobs.

Subjecting credit unions to CRA requirements would compound problematic and costly regulations. Some bank regulators recently redefined large banks to be those with assets larger than $1 billion instead of $250 million. The reason for this change in classification was simply to reduce the set of large banks that are subject to relatively higher regulatory costs related to CRA. Small banks face a lower CRA regulatory burden because they are deemed to be more communityoriented in their competition for earning assets and deposits and there is a regulatory interest in holding down the regulatory burden on small banks. While extending the more burdensome CRA regulatory requirements to credit unions would help level the playing field, it would be a move away from regulatory efficiency. Credit union officials have argued that the NCUA already examines credit unions on CRA-type criteria even though credit unions are not explicitly covered by the act. Certainly extension of CRA requirements to credit unions could await reform of existing CRA requirements, which has been on the regulatory agenda for several years.

VII. Conclusions

Bickley (2003) has emphasized that until 1951, savings and loan associations were not subject to the federal income tax. He suggests that it was the changing financial environment that altered the relationship of thrifts to their members and caused the loss of their special tax treatment. The implication is that the changing relationship of credit unions to their members (changes in the common bond and field of membership) will ultimately lead to the end of the federal income tax exemption of credit unions. Whether that parallel is correct and predictive for credit union taxation remains to be seen. Certainly renewed calls for a new round of fundamental tax reform are indicators that tax system fairness will be addressed. However, entrenched political support for protecting the status quo for credit unions does not bode well for quick tax equity reforms. Neutrality requires that similar firms be taxed equally. But are credit unions the equal of banks or thrifts? Many voters view their credit union as an extension of the employee benefits granted by their employer or employee association, a service of a social organization to which they belong, or perhaps a community service related to a community they live in. But when the bond that defines the customer base extends to a community, like the recent decision upholding the charter for the Los Angeles County Credit Union, it is difficult to see that the bond is more than any other corporate business model defining a customer base.

Fiscal neutrality would require removing the special tax treatment of credit unions. One suggestion is that the special tax status for small institutions be maintained while eliminating it for larger credit unions that have equally broad potential customer bases, and that have commercial operations, expenses and income comparable to large banks and thrifts. This would seem to be a compromise policy path. Large financial firms would be subject to the same tax treatment, but the proposal would protect the tax exemption of small credit unions that adhere to their statutory mission to provide services and benefits to people of low economic status that may not be available otherwise.

However, small credit unions are as likely to be service facilities that are part of an employee benefits package, and like other benefits are efficiently produced. Tax reform has long recognized that untaxed employee benefits result in an inequity for the nation's taxpayers and the direction of change is certain, if slow and halting. Studies show that credit unions have not been particularly effective in reaching audiences that truly have limited access to loans, saving facilities or other financial services offered by credit unions. One of the greatest impediments to access is the lack of a job or occupational connection, and occupational credit unions do not overcome these obstacles. Even associative or community-based credit unions do not serve the jobless with loans or other financial assistance. Few workers would find their ability to obtain car or mortgage loans or to find attractive saving options to be eliminated by the absence of their credit union.

There are broad constituencies that would resist tax reform, some of whom have considerable weight politically. First are the leaders and employees of credit unions that would see the credit union movement, purportedly and perhaps once, as a social movement under threat. Second, many populist groups would see imposing a higher cost on credit unions as an attack on the poor or lowincome voters. Many credit unions are tied to trade unions, and while trade unions are declining and of relatively small numerical importance, they often loom large as potent political forces when acting as a voting block. Thus Treasury Secretary Snow's reluctance to see any issue in the reform of taxation of credit unions, noted at the beginning of this paper, and the Congress' haste to protect the breakdown of limits on the membership of credit unions are understandable from a political perspective. They are not understandable from the perspective of sound tax policy, however.

Tax reform of credit union income taxation is a "no-brainer" when viewed in a broad tax neutrality context. It is also compelling when the either the size of the revenue loss or the ineffectiveness of the tax break for achieving any social goal is considered. Reform would likely have to be part of a more compelling tax reform package if it is to ever be adopted. This is ironic because the tax loss from credit unions is very large, about $2 billion per year currently and over $30 billion over the next ten years. This study could not find any net benefit to members that could not or would not be available in the absence of tax-subsidized credit unions. Most notably, the credit union subsidy, by its very nature, has largely failed to deliver financial services to low-income people.

The origins of the credit union tax exemption reach back to the Great Depression, a time when basic financial services were limited. Today, however, the financial services industry is highly competitive, and its products and services are accessible to all consumers. Credit union members are just like other consumers who use banks instead of credit unions. Credit unions are not compelled by regulators to meet a higher standard in the service of low- and moderate-income customers, and there is no evidence that they do so voluntarily. The $650 billion credit union industry may have outgrown in size and scope its original, tax-exempt mission.

From the perspectives of tax fairness and sound tax policy, where the least economic distortion and inefficiency is desired, taxation of credit unions warrants prompt and genuine consideration in any tax reform debate.

² For example, see New Zealand Treasury (2000) and Department of Finance, Government of Ireland (1998) for discussions of the identical problems in these countries and proposals to end the exemption of corporate (or company) income tax. The Ireland paper also discusses moves in the European Union to promote neutrality by eliminating tax exemptions for credit unions and their customers. The distinguished free market Finance Minister Charlie McCreevy (2000), the recently appointed EU Commissioner for Internal Markets, indicated support for the corporate tax exemption, but opposition to the broader tax exemption of interest income at Irish credit unions. However, he noted the continuing opposition of the EU Commission to state aid to private sector activity and implies that the corporate exemption is precisely such aid. Other recent studies include Bickley (2003), Florida Tax Watch (2003) and Chmura Economics and Analytics (2004).

³ The Supreme Court case was NCUA vs. First National Bank and Trust Co. (1998).

⁴ The author was in the audience when this exchange occurred.

⁵ See Bickley (2004).

⁶ See Joint Committee on Taxation (2003).

⁷ Credit unions originated in the Raiffeisen banks started in 1854 in rural areas of the Rhine by Friedrich Wilhelm Raiffeisen. These were cooperative banks intended to serve farmers and were the first to have a field of membership and use the same operating principles as today. Such banks still exist with this name, and purpose, in Austria and several other European countries. Cooperative banks for artisans date back to 1850 and purchasing cooperatives date back to 1844 when the Rochdale Pioneers organized the first cooperative in Rochdale, England. See Credit Union League of Hong Kong, www.hkcreditunion.org. and Bickley (2003) for the historical information.

⁸ This information is drawn from a survey provided by the Office of the General Counsel, Credit Union National Association, 2004.

⁹ See Congressional Budget Office (March 2003), p. 218.

¹⁰ The personal consumption deflator (2000 = 100) is used to deflate both asset measures.

¹¹ Chmura (2004) points out that this relatively slower growth may suggest that CUMAA did not augment the tax break for credit union and foster their faster growth. The Chmura study suggests that the evidence on relative growth of credit unions is mixed because the average size credit union has been growing faster than that of banks since 1998. This is an overstatement, however. The faster growth of the average size simply reflects a faster pace of consolidation in the number of credit unions compared with banks. This faster pace of consolidation more than offsets the slowdown in the relative growth rate of credit union assets relative to those at banks. Both banks and credit unions have seen booming growth in their assets while their numbers are declining. Both sectors have been consolidating since the 1980s.

¹² See, Gunther and Moore (2004a) for an example of such a cautious study.

¹³ CUMAA limits the size of business loans to 12.25 percent of total assets, but Small Business Administration lending and business loans under $50,000 are not capped at all. According to NCUA data, in 2003 member business loans outstanding were only $8.9 billion, or 2.4 percent of total loans and 1.5percent of total assets. Such loans are expected to become a larger share of assets, especially at large credit unions.

¹⁴ Similar comparisons have been made by Schenk (2004). He and the Credit Union National Association (2004a) also argue that credit unions consistently rate higher in consumer satisfaction than banks. This is another example of a lack of comparability, however, because credit unions are small and more closely tied to their consumer customer base than banks. A proper comparison would compare similar-sized institutions serving comparable customer bases. Why would one expect larger and more business-oriented banks to earn high marks from households that are relatively unimportant as sources of a funds or as borrowers? While banks value consumer business, for many banks consumers are not the core business or the focus of marketing efforts.

¹⁵ The estimate is based on a conservative assumption that real credit union assets grow at the 6 percent pace they registered from 1993-2003 and that prices rise at a 2 percent rate, the same as over the ten-year period 1993-2004. Interest rates are assumed to average levels of the same earlier ten-year period.

¹⁶ Florida Tax Watch (2003) estimates the tax loss in Florida alone is $102 million. In Florida, as in several other states, all credit unions are exempt from state and local taxes, except for state-chartered credit unions that face real and tangible property taxes. The estimate includes all tax savings that arise from special exemptions from taxes for credit unions.

¹⁷ See Office of Management and Budget (2004), Table 18-1, p. 287.

¹⁸ Emmons and Schmid (2003) provide a nice review of the tension between banks and credit unions and explain that the tax exemption could either cause credit unions to be inefficient or it could give them a competitive advantage. They view these effects as mutually exclusive, but tax-exempt status could lead to both results. It is also possible that tax exemption could be the result of greater inefficiency that occurs because of restrictions on credit unions, as well as the cause of greater inefficiency.

¹⁹ The after tax-rate of return on equity is (1-t) times the before-tax rate of return, where t is the tax rate. Thus the supply price of equity capital measured vertically in Figure 3 would rise by 1/(1-t) times the cost of equity capital as t rises from zero to onethird. When the tax rate is one-third the cost of equity capital and the new supply price will rise by an amount that is 50% percent larger than the cost of equity in the absence of the federal tax.

²⁰ The NCUA, the federal credit union regulator, and CUNA, the national trade association for credit unions, do not publicize data on taxes paid by state-chartered credit unions, so it is difficult to evaluate the significance of these taxes directly. Most institutions use the NCUA's call report template to report their costs so that taxes are included in "miscellaneous operating expenses" on in the case of sales taxes sometimes in "operating fees."

²¹ See CBO (2004), for example.

²² In 2003, commercial and industrial loans and commercial real estate accounted for 11.5 percent and 7.9 percent of bank assets, respectively. While credit unions are expanding their business lending, which could be more safely and profitably supported by their relatively high equity ratio, it remains minuscule in comparison.

²³ The average ROA for the ten years 1994-2003 was 1.01 percent, below the average for banks, 1.21 percent. Gunther and Moore (2004b) argue that using summary measures for each firm's performance ratio is superior to ratios computed for the group. Using data for individual credit unions and summary measures for banks prepared by the FDIC does not alter the conclusions here, however.

²⁴ Chmura Economics and Analytics (2004) notes that the Reagan administration proposed that credit unions with assets in excess of $5 million lose their federal income tax exemption as part of their broad tax reform that occurred in 1986. This recommendation was apparently rejected in congressional debate over the otherwise sweeping reforms. Indexed for inflation, the Reagan-era exemption would be about $8 million today. The Carter administration also had proposed removing the tax exemption for all credit unions, according to Chmura.

²⁵ See CBO (2003), proposal 21.

²⁶ The effective or average tax rate of credit unions for all institutions would have been 33.7% of income if all were taxed and this would rise to 33.97% if only larger institutions were taxed.

Amel, Dean and Timothy Hannan, "Establishing Bank Market Definitions Through Estimation of Residual Deposit Supply Equations," Board of Governors of the Federal Reserve System working paper, 1998.

Bickley, James M., "Should Credit Unions Be Taxed?" Congressional Research Service, 97-548E, updated October 6, 2003.

Chmura Economics and Analytics, An Assessment of the Competitive Environment Between Credit Unions and Banks, The Thomas Jefferson Institute for Public Policy, May 2004.

Congressional Budget Office, Budget Options, March 2003.

________, Federal Subsidies and the Housing GSEs, May 2001; updated estimates, April 8, 2004.

Credit Union National Association, "CUs beat big banks in Forrester survey," www.cuna.org/newsnow/04/system061104-12.html, June 11, 2004a.

________, "Snow reiterates support for CU tax exemption," www.cuna.org/newsnow/04/ wash091304-1.html, Sept. 13, 2004b.

Department of Finance, Government of Ireland, "Tax Treatment of the Returns on Credit Union Savings," TSG 98/35, www.finance.gov.ie/ Viewprnt.asp?DocID=1496.

Emmons, William R. and Frank A. Schmid, "Bank Competition and Concentration: Do Credit Unions Matter?" Federal Reserve Bank of St. Louis, Review, May/June 2000, pp. 29-42.

________, and _______, "Banks, Credit Unions and the Dynamics of Competition," Federal Reserve Bank of St. Louis, Working Paper 2000-006A, 2000.

________, and _______, "Credit Unions and the Common Bond," Federal Reserve Bank of St. Louis, Review, September/October 1999a, pp. 41-64.

________, and _______, "Wages and Risk Taking in Occupational Credit Unions: Theory and Evidence," Federal Reserve Bank of St. Louis, Review, March/April 1999b, pp. 13-31.

Feinberg, Robert, "The Competitive Role of Credit Unions in Small Local Financial Services Markets," Review of Economics and Statistics, August 2001.

________, " The Determinants of Bank Rate-Setting in Local Consumer Markets: Comparing Market and Institution Level Results," Working Paper, 2002.

Flannery, Mark J., "An Economic Evaluation of Credit Unions in the United States," Federal Reserve Bank of Boston, Research Report 54, 1974.

Florida Tax Watch, Florida Credit Unions Receive Tax Exemptions Worth $102 Million, April 2003.

Fried, Harold O., C.A. Knox Lovell and Phillipe Vanden Eeckhaut, "Evaluating the Performance of U.S. Credit Unions," Journal of Banking and Finance, (April 1993), pp. 251.65.

________, ________, and Suthathip Yaisawarng, "The Impact of Mergers on Credit Union Service Provision," Union College, Working Paper, March 1998.

Gunther, Jeffrey and Robert R. Moore, "Small Banks' Competitors Loom Large," Southwest Economy, (Jan/Feb. 2004a), pp. 1, 9- 13.

________, and ________, "Small Banks Far From Thriving, Partly Because of Credit Unions," American Banker, August 20, 2004b.

Hannan, Timothy H., "The Impact of Credit Unions on Deposit Rates Offered for Retail Deposits by Thrifts and Banks," Board of Governors of the Federal Reserve System, Working Paper, September 2002.

Hansman, Henry, The Ownership of Enterprise, Harvard University Press, 1996.

Jacob, Katy, Malcolm Bush and Dan Immergluck, Rhetoric and Reality: An Analysis of Mainstream Credit Unions' Record of Serving Low Income People, Woodstock Institute, February 2002.

Joint Committee on Taxation, "Estimates of Federal Tax Expenditures for Fiscal Years 2004-2008," JCS-8-03, December 22, 2003, and earlier issues.

Keating, Barry P., and Maryann O. Keating, "An Empirical Estimation of the Degree of Expense Preference Behavior Between Credit Unions by Common Bond Type," Quarterly Review of Economics and Finance, (Summer 1992), pp. 71-84.

McCreevy, Charlie, "Speech on the Taxation of Credit Unions," Adjournment Debate, February 2, 2000, www.finance.gov.ie/ viewdoc.asp?DocID=1567.

National Credit Union Administration, 2003 Year-end Statistics for Federally Insured Credit Unions, August 16, 2004, and earlier issues available at ncua.gov.

New Zealand Treasury, "The Tax Status of Credit Unions," Issues Paper, Ministry of Economic Development, Policy Advice Division of Inland Revenue, September 6, 2000.

Office of Management and Budget, Analytical Perspectives, Budget of the United States, Fiscal Year 2005, Washington D.C.: Government Printing Office, 2004.

Patin, Roy P., and Douglas W. McNiel, "Benefit Imbalances Among Credit Union Member Groups: Evidence of Borrower-Dominated, Saver Dominated and Neutral Behavior?" Applied Economics, (April 1991), pp. 769-80.

Schenk, Mike, "Commercial Banks and Credit Unions," presentation for the Credit Union National Association (CUNA), www.cuna.coop.

Smith, Donald J., "A Test of Variant Objection Functions in Credit Unions," Applied Economics, (September 1986), pp. 959- 70.

________, "A Theoretic Framework for the Analysis of Credit Union Decision-Making, Journal of Finance, (September 1984), pp. 1155-68.

________, Thomas F. Cargill and Robert A.

Meyer, "An Economic Theory of a Credit Union," Journal of Finance, (May 1981), pp. 519-28.

Tokle, Robert J. and Joanne G. Tokle, "The Influence of Credit Union and Savings and Loan Competition on Bank Deposit Rates in Idaho and Montana," Review of Industrial Organization 17 (2000), pp. 427-38.

U.S. Treasury Department, Credit Unions, U.S. Government Printing Office, 1997.

      Appendix: Interest and Dividend Rates at Federal and