In-Depth Analysis, March 2006
by Robert D. Dietz, Ph.D.
The Internal Revenue Code contains several provisions that promote homeownership. Among these are two related policy tools: Mortgage Revenue Bonds (MRB) and Mortgage Credit Certificates (MCC). MRBs reduce the monthly housing payment for a qualifying homebuyer by reducing the interest rate of the program participant’s mortgage. In contrast, MCCs provide a homeownership incentive by reducing a participant’s federal tax liability. Presently, MRBs are used much more than MCCs. This article examines the features of these two programs in an effort to assess their relative effectiveness in expanding homeownership opportunities.
Mortgage Revenue Bonds
MRBs are tax-exempt bonds that finance mortgage loans targeted to moderate-income first-time homebuyers. Issued by state and local housing finance agencies (HFAs), MRB proceeds are used to purchase or originate mortgage loans at below market rates. To qualify for a mortgage financed by an MRB, the household’s income cannot exceed 115% of area median family income.  The price of a home purchased with an MRB-financed mortgage may not be greater than 90% of the average price of homes in that area.  The participant must not have owned a home in the previous three years. Exceptions to these rules are made for targeted low-income neighborhoods, households with many members, or households living in certain high cost areas.
MRBs are a type of private activity bond, as defined by the Internal Revenue Code.  The volume of private activity bonds that may be issued annually is limited by state volume caps, which have been adjusted for inflation since 2003. For 2006, each state’s volume cap is generally equal to $80 per capita.  For 2003, HFAs had approximately $6.7 billion in MRB authority out of a total private activity bond cap of $24.7 billion.  However, the total authority available for MRB issuance in 2003 was equal to $19.5 billion due to unused bond authority from previous years. 
Due to their tax-exempt status, MRBs allow states to borrow at lower interest rates. The savings can be passed onto homebuyers in the form of below market rate mortgages. The interest rate on mortgages financed by MRBs is typically greater than the interest rate on the MRBs themselves, with the difference used by HFAs to finance the costs of operating the program. These costs include education, outreach, administration, and legal and underwriting services associated with bond issuance. Under present law, the interest on mortgages financed by MRBs may not be greater than 1.125% points above the interest rate on the MRBs.
Qualifying homebuyers in the MRB program thus have smaller monthly mortgage payments due to reduced interest costs. Furthermore, the reduced monthly housing cost increases the ability of the homebuyer to qualify for a mortgage, which in many cases is an obstacle to homeownership.
Consider the following example, which is illustrated in Table 1.
- An MRB-qualifying household with an income of $39,000 (the average household income of a MRB program participant in 2003).
- The household purchases a home with a price of $108,407 (the average price of a home purchased with an MRB-financed mortgage in 2003).
- The homebuyer pays a downpayment of 3% ($3,252). 
- The homebuyer obtains a 30-year, fixed-rate mortgage with an interest rate of 6.04%. 
- For the first year, the household’s mortgage payments are $7,598, of which $6,316 is interest.
Now consider an otherwise similar homebuyer obtains an MRB-financed mortgage at a rate of 4.25%.  With respect to the first year of mortgage payments, the homebuyer pays an amount equal to $6,208, of which $4,435 is interest. Due to the tax-exempt MRB-financed mortgage, the qualified homebuyer has a reduction in annual housing costs of $1,390. This represents approximately a 14.5% reduction in housing costs, a substantial benefit for the homebuyer. 
Mortgage Credit Certificates
HFAs can convert part of their MRB issuance authority into MCCs. MCCs provide qualified homebuyers a nonrefundable tax credit under generally the same income and purchase price limits that apply to MRB-financed home purchases.  The credit must have a rate of at least 10% but no more than 50% of the mortgage interest paid. However, the amount of the credit claimed by the taxpayer in a single year may not exceed $2,000. Once allocated, an MCC is effective until the residence of the homebuyer ceases to be a principle residence.  MCCs are issued by exchanging MRB authority for MCC authority. The tax code generally allows a four-to-one exchange rate. For example, if an HFA gives up $1 million in MRB authority, it may issue $250,000 in MCCs. MCC authority is defined as the amount of indebtedness for which a dollar of interest paid allows a dollar of tax credits. For example, $250,000 in MCCs allows tax credits for interest payments on $250,000 of mortgage debt. If these MCC’s are 50-percent credits, this would be equal to $500,000 of total mortgage debt.
Comparing MRBs and MCCs
To compare the MCC and MRB programs, suppose the household from the previous example qualifies for a 50-percent MCC. The homebuyer’s first-year mortgage payments total $7,598, of which $6,316 is interest. A 50-percent MCC credit allows a credit equal to the lesser of 50-percent multiplied by interest paid ($3,158) or $2,000. In this case, the homebuyer is allowed a tax credit of $2,000. Consequently, the homebuyer’s after-tax housing payment falls by approximately 21%, which is greater than the 14.5% reduction provided by the MRB program. 
This example demonstrates a potential advantage that MCCs possess over MRBs: targeting flexibility via the credit rate. However, the MCC effectiveness is limited by the fact that the credit is nonrefundable. If the homebuyer in this example does not have tax payments of at least $2000, then the full subsidy is not available. According to IRS statistics, the average MRB-program participant has only $640 of tax liability after credits. However, it should be noted that the average MRB-program participant is not identical to the average MCC-program participant. Indeed, given the constraint on tax liability, the programs’ respective rules suggest that the average MCCs are a more effective policy tool for promoting homeownership for households with relatively higher income and tax liabilities. 
The MCC program appears to be relatively underutilized. During 2003, MRB issuance was approximately $3.3 billion. In contrast, MCC issuance was equal to $263 million.  This is puzzling because MCCs, in many cases, seem to provide a larger net benefit per household, as illustrated in Table 1. Moreover, in general, the MCC program requires less administration than the MRB program. For example, issuing MRBs requires underwriter’s costs and bond counsel fees. Of course, HFAs finance such costs through bond arbitrage. In contrast, the MCC program is less costly to operate. Because the tax credits are issued directly to homebuyers, no interaction is required with bond markets. The only overhead infrastructure required of the program is administration necessary to select program participants and education or marketing necessary to attract qualified homebuyers to the program, requirements which also exist for the MRB program. 
Furthermore, the MRB program is interest rate sensitive. As interest rates decline, as has occurred since 2001, the difference between market mortgage interest rates and MRB-financed mortgage interest rates declines while the limitation on arbitrage remains equal to 1.25% points. As many HFAs use the arbitrage to finance their administrative costs, there is reluctance to reduce the spread between the MRB interest rate and the MRB-financed interest rate. Consequently, the decline in interest rates has reduced the benefit provided by the MRB program, which may be a factor in the lack of issuance of housing bond authority cited earlier in this paper.
An important distinction between the MCC and MRB programs is the role of cash flow. As detailed earlier, the MRB tax subsidy reduces a qualified homebuyer’s monthly costs, thereby increasing the likelihood of qualifying for a mortgage. In contrast, the MCC does not reduce the monthly housing payment. As a result, a mortgage lender may not be convinced that the potential homebuyer qualifies for a given mortgage, especially as the tax benefit is uncertain, as it is constrained by final tax liability, which may not be known until the following year. However, there is a great deal of variability with respect to MCC sophistication among lenders. More sophisticated lenders, particularly those in states or areas with large MCC programs, are likely to incorporate MCC tax effects in determining whether mortgage qualifying criteria are satisfied. In any event, credit scores, rather than income-to-loan ratios, are increasingly more important as qualifying tests. 
Notwithstanding the tax liability constraint, the MCC program offers a considerable, if not greater, aggregate subsidy to homebuyers than MRBs. One method of determining the value of the subsidy is to compare the relevant tax expenditures of the two programs. A tax expenditure is the dollar value of a tax benefit if the tax provision were to be repealed and replaced with an expenditure program. As a starting point, suppose the volume cap for MRBs is equal to $100 million. As shown on Table 2, the tax expenditure for this MRB program in the first year is approximately equal to $1.39 million.  Suppose instead HFAs decide to convert all MRB authority into MCCs. A total MCC authority equal to $25 million would be created. The first-year tax expenditure for the $25 million of MCC issuance is equal to $1.52 million.  The tax expenditure, or subsidy value, is therefore 8.7% higher under the MCC program.
As presently defined in the IRS Code, MCCs and MRBs offer different policy tools for promoting homeownership. The differences between the two programs suggest that each may be better suited for distinct policy environments, rather than one dominating the other in terms of effectiveness. MCCs offer flexibility in targeting the size of subsidy by allowing differing credit percentages to be applied. Dollar for dollar, MCCs can be used to give an equivalent subsidy to a slightly greater number of homebuyers. Moreover, in the case shown in Table 1, MCCs can be used to give a particular homebuyer a greater subsidy, at the cost of reducing the number of participating households. However, the MCC subsidy is limited to $2,000 per year. The MCC program is further limited by its nonrefundable nature. The MCC program is less costly to operate than the MRB program and its tax expenditure is larger given the exchange rate defined by the IRS Code.
These characteristics suggest that MCCs may be more effective than MRBs in high cost regions or metropolitan areas with higher household incomes and tax liabilities. This conclusion is consistent with the evidence of program utilization. California’s HFA has the most expansive MCC program in the nation, measured in both the absolute level of credit allocations and as a relative measure with respect to MRB allocation. Furthermore, MCCs may be more effective during periods of relatively low interest rates. However, MCC issuance declined between 2002 and 2003, a period when interest rates were generally falling.
In contrast, MRBs are better suited for the tax policy objective they perform now; namely, reducing the monthly housing payment of qualified homebuyers and indirectly allowing homebuyers to qualify for previously unattainable mortgages. MRBs may also have an advantage because the HFAs have practice and familiarity with MRB programs in terms of administration, community outreach and institutional knowledge. However, MCCs should not be overlooked as an effective low-income housing tax policy alternative due to an excessive reliance on established practices.
ev. Proc. 2005-15 contains the IRS 2005 income limits. Return to the Article
 IRS uses HUD estimates for these limits (per Rev. Proc. 2005-22). HUD estimates are found at http://www.huduser.org/datasets/il.html. Return to Article
 According to the Code, private activity bonds are tax-exempt bonds issued by or on behalf of local or state government for the purpose of providing special financing benefits for qualified projects, most often for projects involving a private interest. Tax-exempt private activity bonds include enterprise zone bonds, mortgage revenue bonds, veterans’ mortgage bonds, student loan bonds, redevelopment bonds, and certain 501(c)(3) bonds. Return to Article
 IRS Rev. Proc. 2005-70. For states with small populations, a minimum level of approximately $247 million per state is provided. Return to Article
 State HFA Factbook: 2003 NCHSHA Annual Survey Results 2004. National Council of State Housing Agencies. Return to Article
 A three-year carryfoward is permitted under present law. Return to Article
 Insufficient data exist to determine the average downpayment required for a MRB-financed mortgage, although conservations with NCSHA, state HFAs, and Joint Committee on Taxation staff indicated that 3% is a reasonable estimate. Return to Article
 Mortgage Bankers Association average for 30-year fixed-rate mortgages for the week ending January 25, 2006. Return to Article
 This is the interest rate charged to low-income households in California for MRB-financed mortgages. Return to Article
 This percentage is calculated by including annual real estate taxes, PMI and insurance costs, totaling for this example approximately $2000 per year. Return to Article
 A three year carryforward is available. Return to Article
 Both MRBs and MCCs have a nine-year recapture period, during which sale of the residence results in a recapture tax being assessed against the homeowner. However, IRS regulations provide rules under which an MCC may be reissued to the new buyer of the residence, provided the buyer is qualified to participate in the MCC program. Return to Article
 Of course, the 50-percent credit is the most generous MCC. This example is constructed to compare a relatively generous MRB and a similarly situated MCC. Conversations with NCSHA and state and local HFA staff indicate that 25 to 28 percent MCCs are average. A 25% MCC allows a 16.4% reduction in annual housing costs, which is greater than the 14.5% reduction provided by the MRB-financed mortgage. Return to Article
 However, the NCSHA 2002 and 2003 Factbooks indicate that MCC program participants had a lower income than MRB-program participants. This may be partially explained by the limited use of the MCC across states and differing policy objectives across HFAs. Return to Article
 State HFA Factbook: 2003 NCHSHA Annual Survey Results. 2004. National Council of State Housing Agencies. Return to Article
 However, this infrastructure has a quasi-fixed cost character. Once the program is in place, an argument can be made that the fixed cost of administration should be spread over as much MRB issuance as feasible in order to minimize the average cost of operating the program. Return to Article
 Conversations with NCSHA and HFA staff. Return to Article
 This is determined by excluding the interest received from bond issuance from gross income. Return to Article
 This is calculated by determining the amount of interest allocable to $25 million of MCC authority, using the same interest rate as in previous examples. Return to Article
See previous In-Depth articles