Homeownership has long been a part of the American dream, but it has eluded nearly 40 percent of low- and moderate-income people who currently rent but would prefer to own.¹ This need not be so. A federal tax credit to expand homeownership opportunities by spurring no-interest second mortgages² can make the dream of homeownership a reality for a half a million lower-income families over a ten- year period.

Homeownership is the primary path to the middle class for many families. It represents the best possible opportunity for disadvantaged groups to build family wealth and economic security. Home equity provides a cushion that allows a family to borrow to start a small business, send its children to college, and handle unexpected events such as job loss or medical bills. Homeownership gives working families the opportunity to accumulate economic resources that create stability and expand opportunities for future generations while simultaneously improving the social stability of low-wealth communities.

Reduced crime, greater school retention, reduced teenage pregnancies, higher life- satisfaction, greater civic engagement, and improved property upkeep are widely known positive effects of homeownership.³ According to a recent University of Tennessee study, children of homeowners are 25 percent more likely to graduate from high school and 115 percent more likely to graduate from college than similar children of renters. As a result, according to the study, children of homeowners earn $150,000 more over a lifetime.⁴

However, homeownership remains out of reach for millions of Americans. Rates among low-income families, minorities, and central city residents lag far behind other groups. Only 45 percent of low-income households live in owner-occupied homes, as opposed to 86 percent of high- income households.⁵ Minority homeownership rates are also well below the rest of the country: 45 percent for minorities versus 72 percent for white households. Among households with incomes 20 percent to 50 percent higher than the area median, nearly 80 percent of white households are homeowners compared to less than 65 percent of black households. And central city homeownership rates are 23 percent behind suburban rates.⁶

Unequal homeownership rates cause disparities in wealth since renters have significantly less wealth than homeowners at the same income level. In fact, the median wealth of non-elderly low-income homeowners is 12 times greater than the median wealth of non-elderly renters of the same income.⁷ Minorities face similar problems. Although African-Americans now earn 62 percent of white income, the average black family has a net worth one-tenth that of the average white family. In the 1990 census, the median wealth of African-American families was $4,400, compared to $44,000 for white families. And averages fail to tell the whole story. Almost two-thirds (61 percent) of all black households and over half (54 percent) of all Hispanic households have no net financial assets at all.

To address wealth disparities in the United States and therefore make opportunities more widespread, the homeownership rates of disadvantaged groups must rise. Since the largest expense for most low-income families is housing, homeownership provides the greatest opportunity for a family's resources to be dedicated both to present needs and future goals. The family home functions much as a savings account that increases each month as the mortgage is paid down and the dwelling appreciates in value. As a form of savings, housing appreciation has historically been an excellent investment.⁸ In fact, 63 percent of all the wealth owned by minority families and 67 percent of the wealth owned by non-elderly low-income families consists of the equity in their home.⁹

Our analysis of federal policy concludes that current policies fall short of the important national goal of increasing homeownership opportunities for low-income families. We show that no-interest second mortgage programs are extremely effective at addressing the two major obstacles renters face in achieving homeownership: borrower affordability and lender risk constraints.

To make homeownership a more attainable goal, we propose a $1 billion federal homeownership tax credit that would induce the private market to provide no-interest second mortgages to low-income families who otherwise could not purchase a home.¹⁰ Instead of receiving interest from the home buyer on the second mortgage, the lender would get a tax credit to be applied against other income. A family that could only afford 75 percent of a market-rate first mortgage's monthly payments on a modest home would now be able to buy the house. The cash provided by a first mortgage of 75 percent of the purchase price combined with a no-interest second mortgage of 25 percent of the purchase price would be sufficient to buy the house. The borrower would make payments on his first mortgage every month and pay off the second when he sells the house or 25 years later, whichever came first. Through the efficient workings of the marketplace, we estimate that on a nationwide basis this tax credit would help 500,000 families become homeowners over a ten-year period and leverage $23 billion of private lending.¹¹

The tax credit can be seen as a binding contract between the new homeowners and the government. The government helps hard-working, low-income people take an important step into the middle class by buying homes. The new homeowners pay the mortgages they were able to obtain through the program and invest in their home and community. Once the government has played its role, the contract enforces itself: homeowners who do not keep up their end of the bargain will lose their homes, and any home equity built up along the way, through foreclosure.

Such defaults, however, will not likely be common. Contrary to popular belief, loans to low-income borrowers are not significantly riskier than home loans as a whole. Indeed, evidence from within and without the banking community suggests that losses from "community development" loans are often no greater than losses from conventional loans.¹² The exceptional repayment history of many community lenders-- including our organization, Self-Help--reinforces this experience. The president of the Minneapolis Federal Reserve Bank cited several such examples in arguing that "Low Incomes Do Not Mean High Defaults."¹³ Because of this strong track record of loan repayment, the loans generated by the tax credits should perform quite well and thus make the tax credits themselves that much more attractive on the open market.

Federal Policy and Homeownership

For such an important national goal, our tax policy does surprisingly little to help those groups whose homeownership rates are low. Many assume that since mortgage interest costs and real estate taxes are tax deductible, the federal government already provides strong incentives to advance homeownership. Although it is true that these two deductions amount to $58 billion--twice the amount allocated to all direct housing assistance programs in the United States--tax deductibility does virtually nothing to help low-income families buy or own a home. In fact, because most low- income families take the standard deduction instead of itemizing their tax returns, over 90 percent of the total benefits of the mortgage interest deduction accrue to homeowners with more than $40,000 in annual income.¹⁴

The programs offered by the U.S. Department of Housing and Urban Development (HUD) have largely focused on providing rental assistance for poor tenants to pay landlords on a monthly basis. Of the over $2 billion FY 1999 funding increase for HUD, over half of the increase ($1.3 billion) was designated for the Section 8 program, which provides vouchers to low-income Americans to supplement their rent payments. Helping families with the lowest incomes afford rental housing is a crucial national goal and should be continued--many families are not ready for homeownership and rental housing is their best option. However, many other families are ready to assume the responsibility of owning a home but cannot afford it. Because the rental assistance programs do not help low-income families buy homes and build wealth, they complement but do not replace the tax credit we propose in this paper.

HUD does have a number of important programs to promote low-income homeownership. It has recently begun experimenting with programs that allow modest numbers of tenants to use Section 8 rental subsidies to pay mortgages when buying a home. In addition, the HOME and Community Development Block Grant (CDBG) programs permit states and localities to undertake homeownership initiatives. And the Federal Housing Administration's (FHA) single- family mortgage insurance program has been extraordinarily effective in assisting low-income, minority, and other under-served families to become homeowners for the past 65 years. A second mortgage, as described in this paper, could effectively be combined with an FHA-insured (or conventional) first mortgage to increase income and wealth affordability for families who could not otherwise qualify for the loan.

The Low-Income Housing Tax Credit program (LIHTC) successfully induces investors to provide funds for low-income multi-family rental housing projects. LIHTC has leveraged billions of dollars of private investment in a million units of much-needed rental housing. It is not a homeownership program, however. A small number of LIHTC tenants may use the program to buy their houses or apartments, but they would have to wait an unreasonably long 15 years before being able to buy, in addition to surmounting administrative complexities.

In short, federal housing assistance largely helps high-income families pay their mortgage or low-income families pay their rent. Federal policy's focus is therefore not on helping low-income families become homeowners.

Public-Private Partnerships Rooted in Communities

We believe that federal aid should also be directed to increasing low-income homeownership. In promoting homeownership among low-income American families, policymakers should consider four principles:

1. Empower individual families and communities. Helping individual families realize their homeownership dreams empowers the families and the communities in which they live. Low-income, first-time homebuyers are entrepreneurs willing to put their own economic future on the line by investing in their neighborhood. As such, they are a good bet for public investment. Low-income families are committed to keeping their homes once they buy them because they realize it is their best chance to rise to the middle class. They become stakeholders, buying into a neighborhood's future and working to improve it.

2. Utilize private delivery channels. Market incentives make the private sector vastly superior overall to the public sector in delivering products to consumers. Mortgages are no exception. Private banks and other mortgage lenders have an efficient delivery system of thousands of branches and hundreds of thousands of loan officers spread throughout the country. This system should be used--not replicated--in promoting homeownership. However, the private sector cannot provide products for people who cannot afford them.

3. Use private funding. Mainstream capital markets--investors such as pension funds, insurance companies, and mutual funds-- invest trillions of dollars in mortgages in America for a market return. Wall Street funds mortgages on Main Street throughout the country by purchasing mortgage-backed securities from Fannie Mae and Freddie Mac, which buy the mortgages from lenders. The challenge for federal policy is to link the millions of families who need mortgage loans with the capital markets having the investment resources to fund them.

4. Put government and the private sector in partnership through non-bureaucratic and decentralized means. The government is effective at supplementing existing resources in national priority areas but not as efficient at directly implementing decentralized programs. In a time of scarce federal resources, any government subsidy should be limited and highly leveraged to catalyze private activity.

The Effectiveness of No-Interest Second Mortgages

There are two major obstacles to homeownership. The first is borrower affordability. According to a recent study, only 12 percent of families who currently rent could afford to buy a modestly priced home (defined as one that is half the regional median house price). There are two distinct affordability problems that keep those 88 percent of renters from buying a modest home: wealth and income constraints. One-third face only the wealth constraint: they have income sufficient to pay the monthly mortgage payments but lack the wealth to cover a down payment and/or closing costs. The other two-thirds are constrained by both inadequate wealth (they cannot afford down payments and closing costs) and lack of income (they cannot afford the monthly mortgage payments).¹⁵

Beyond the affordability problem, a second obstacle thwarts increasing homeownership: lender risk. Low-wealth borrowers by definition cannot come up with large down payments. A lender is reluctant to make such loans because the ratio of loan amount to home value is too high, increasing the risk that the lender will lose money if there is a default.

The best solution to wealth and income constraints and the risk obstacle is a no-interest second mortgage. A family would receive two mortgages: a market-rate first mortgage from a conventional lender and a smaller, no-interest second mortgage. This second mortgage is subordinate to a private lender's larger first mortgage and does not require interest payments by the home buyer.

A no-interest second mortgage opens up the private marketplace dramatically. The size of the first mortgage is reduced by the second mortgage. A lower loan amount can solve the borrower's affordability problem by reducing his monthly loan payments, as well as by significantly reducing down payment requirements. It can also solve the lender risk problem by lowering the lender's loan-to-value ratio; if the borrower defaults, the lender has a much greater chance of recouping its smaller loan through selling the house at foreclosure. As a result, the lender can be more forgiving of a low-wealth family's minor credit problems. In sum, with a no-interest second mortgage a low-wealth borrower can participate in the mainstream mortgage finance market, and this market has every incentive to encourage this participation.

Here is how a no-interest second mortgage would work for the Jones family of four in Asheville, North Carolina. They earn $24,145 a year--55 percent of area median income--and want to buy an $85,000 house. Without a second mortgage, the family could not afford to buy the house since monthly payments would take up too much of their income. With a 3 percent down payment of $2,550 and an $82,450 first mortgage, the monthly payments of $677 would take up 34 percent of their income too much to qualify for a conventional loan.¹⁶

However, if the Jones family could obtain a no-interest second mortgage worth 19 percent of the cost of the house, they could qualify for a first mortgage and buy the house. They would put down 3 percent ($2,550), obtain a market-rate first mortgage of $66,217, and use a $16,233 second mortgage to pay the seller of the house the full $85,000. Because of the no-interest second mortgage, their monthly payments would be reduced by $114 to a manageable $563, which is 28 percent of their income.

The Jones family would make this $563 first mortgage payment until they moved or paid off the mortgage 30 years later. They would not have to make payments on the $16,233 second mortgage while it is outstanding but would have to pay it off in its entirety when they moved or when the loan became due in 25 years. If they paid it off when they moved, they would use the house sale proceeds to pay off the second mortgage after paying off the first mortgage. If they stayed in their house for the full 25 years, they would have to raise the cash to pay the second mortgage off at that time. They would pay for the loan using savings or through pledging some of their accumulated home equity and obtaining a new loan. At year 25, the loan would be worth roughly 20 percent of its present value ($3,246) and would not be as much of a burden to pay off.

The fact that borrowers of balloon mortgages have to raise the full loan amount at one time is an inherent disadvantage of this type of loan. If real estate values decline significantly or the family is unable to raise sufficient funds, paying off the balloon could present a problem and potentially cause a family to lose its home. We believe, however, that this problem is mitigated for the vast majority of homebuyers with even 1 percent house price inflation because the mortgage term is so long and the loan does not amortize or charge any interest. In other words, as time passes the real value of the loan decreases while the value of the equity in the home increases, allowing the vast majority of families to repay the loan without hardship. Therefore, we believe that the advantages of making the mortgage more affordable to families during its term makes the tradeoff worthwhile.

Many municipalities have recognized the power of no-interest second mortgages and begun offering them with money from federal grant programs such as HOME or CDBG or local bond issuances. These programs have proven to be extremely successful. These resources are limited, however, and must serve many other purposes as well. And rural areas and many poorer cities lack the economies of scale and/or resources to provide this assistance, effectively shutting people out of homeownership exactly where it is most needed.

No-Interest Second Mortgage Case Studies

No-interest second mortgages effectively allow people with no other options to become homeowners. Self-Help's Walltown Homeownership Project catalyzed the Durham, North Carolina, neighborhood of Walltown. The neighborhood is historically African-American and adjoins Duke University. It is also blighted and very poor; absentee landlords own 80 percent or more of the properties. The project demonstrates how homeownership not only increases wealth for the home-owning family but also stabilizes disinvesting neighborhoods.

Self-Help purchased 45 vacant or deteriorated rental duplexes and so far has converted 30 of them into virtually brand-new single-family homes. It then resold the houses to low-income families--some of whom earn as little as $17,000 a year--to increase neighborhood homeownership rates. The median income of these homebuyers is $21,000, about 43 percent of area median income. Over 95 percent of the buyers are African-American; most are headed by single mothers; all are first- time homebuyers; and one-fifth were previously renters in the same neighborhood.

After repairing the properties to meet model home standards, the sales price on the houses is about $75,000. However, everyone who wants to live in these houses needs a no-interest second mortgage to bridge the borrower affordability and lender risk gaps. As a result, subordinate, no- or low- interest mortgages from the City of Durham, Federal Home Loan Bank of Atlanta and/or the North Carolina Housing Finance Agency are packaged together, adding up to almost half of the sales price. The buyers can afford the smaller, market-rate first mortgage. Moreover, conventional lenders, at a low loan-to-value ratio, are now willing to make the first mortgage to borrowers who would not have otherwise qualified. These families are therefore able to buy houses nicer than they ever considered, significantly improve their economic futures, and contribute to the revitalization of a neighborhood.

On a much larger scale, the New York City Housing Partnership has succeeded in revitalizing areas of New York that many thought were beyond reclaiming. Over 15,000 families have invested in these communities by buying their own homes through the Partnership. This activity represents private investment of more than $1.5 billion in 50 low-income communities across the five boroughs of New York City. New York faced the same issues as Walltown: how to make homeownership possible for the low-wealth households eager to buy but only able to afford first mortgages significantly below the houses' selling prices. As in Walltown, no-interest second mortgages enabled the affordability and risk problems to be overcome.¹⁷

A Single-Family Federal Homeownership Tax Credit

The federal tax code is an ideal mechanism to achieve decentralized, market-driven implementation of a policy goal. Through the tax code, we propose establishing a $1 billion annual homeownership tax credit, which we estimate would help more than 500,000 low-income families become homeowners over a ten-year period. In addition, we believe that the credit would leverage $23 billion of private lending and significantly increase the stake that low-wealth families have in their communities.¹⁸

The program would induce investors to fund no-interest second mortgages to low-income families. In lieu of receiving interest payments from the families, the investors would receive a tax credit on other income from the government. State housing finance agencies (HFAs) would auction off the tax credits to lenders who would compete to provide the second mortgages (25-year, no- interest balloon loans). Combined with a market-rate first mortgage, the no-interest second loan would ease borrower income and wealth constraints and make the difference between families renting and owning a home.

The LIHTC program works in a similar fashion for investors in low-income multi-family rental housing projects. As a supplement to the successful rental credit, our proposal seeks to do the same for homeownership. Banks would probably be the most likely users of the tax credit. They are used to making home loans and would already be aware of low-income borrowers that need the second mortgages. Banks are also big investors in LIHTC projects and are thus familiar with this type of program. Employers and other for-profit businesses in search of investment opportunities through tax credits are also likely participants. Finally, community-based nonprofit groups and regional or national intermediary nonprofits could also play a significant role. They could match homebuyers with participating lenders, provide pre-purchase and default counseling, act as lenders themselves, or help develop a secondary market in the second mortgages and accompanying tax credits.

Homeownership Tax Credit Details

We believe that the homeownership tax credit program could be administered simply and in a straightforward manner, using proven delivery mechanisms. The implementation details follow:

Evaluating the Credit

The advantage of tax credits being applied against an entire pool is that risk is diversified across the pool, not concentrated in a particular loan. Because a secondary market can easily develop where the originator can sell the loan and associated tax credits to investors, the ultimate investors can be removed from knowing about particular homebuyers. In addition, the auction process allows market pricing to develop, so the bid prices should increase over time as the whole process becomes more efficient. For example, under a different structure, the LIHTC bids rose from 50 cents for a dollar of credits to 80 cents today as efficiencies increased.

The lender would determine its bid based on its evaluation of various criteria including: (1) how well it can initially lend out the second mortgage pool to eligible borrowers, (2) how effectively it can relend prepayments in the first ten years, (3) expected losses, (4) expected prepayments after the ten-year tax credit period, and (5) its desire to make the second mortgages for Community Reinvestment Act ²⁰ reasons or to also obtain the first mortgage.

Our tax credit proposal meets the four principles for effective policy described earlier. The proposal promotes widespread low-income homeownership, thus fulfilling our first requirement of empowering families and communities. Private mortgage lenders continue to make the first mortgages, and have the ability alone or in partnership with a nonprofit group to make the second mortgages, thus meeting our second requirement of private product distribution. Capital to fund both first and second mortgages comes from mainstream private sources seeking a market return, so our third requirement, an emphasis on private funding, is achieved. Finally, the modest amounts of government capital put into this program would allow homeownership to reach deeper than ever before, dramatically leveraging public resources through the efficiencies of the private market. A nationwide federal program would correct geographic imbalances and give every American working a full-time job the opportunity to own a modest home.

Conclusion

Homeownership increases family wealth, builds strong neighborhoods, and reduces social problems. It is the single most powerful tool available for helping hard-working families reach the middle class.

This aspect of the American dream is not available to everyone who works hard and wants to buy a home, however. As a result, nearly 90 percent of renters cannot afford to buy a home that is half the median sales price in their area. Low-income households own homes at roughly half the rate of high-income households, and minorities own homes at less than two-thirds the rate of white households.

Increasing low-income homeownership needs assistance to succeed. While the federal government already spends billions of dollars on housing, this money--on the whole--does not promote low-income homeownership. It is largely targeted through the tax code to high-income homeowners through $58 billion of homeownership deductions and to low-income renters through HUD grants and the LIHTC programs. These rental assistance programs serve an important need but do not increase homeownership or build family assets.

Our proposed homeownership tax credit addresses these problems. While it would cost less than two percent of the existing homeownership tax deductions, it would target those who need help the most: families who are working hard but lack the wealth and income to move to the middle class through homeownership. The tax credit would empower individual families and communities. Through the efficient, decentralized delivery of mortgages to the families that need them, it would leverage private resources to achieve important national goals without creating new bureaucracies.

The homeownership tax credit will prove to be an efficient, cost-effective tool to help place all families within reach of the American dream.

Appendix:
Comparing Brookings Model and Self-Help Proposal

Endnote 10 describes work done by Collins, Belsky, and Retsinas of Harvard's Joint Center for Housing Studies that was published by the Brookings Institution. The Brookings paper presents an illustration of a homeownership tax credit that shares many similarities with our proposal (hereinafter, "Brookings model").

In several respects, however, the Brookings model and our proposal differ. The most significant difference is the form of the second mortgage. The Brookings model allows HFAs to decide the loan term and type. Their illustrative example is a 3 percent interest, 30-year amortizing second mortgage, with a servicing fee of 0.38 percent. The 3 percent interest rate (less servicing costs) is intended to help spur investor interest. An amortizing second loan ensures that a borrower will not face a balloon payment. Self-Help, on the other hand, recommends a 25-year, no-interest, non-amortizing balloon mortgage. The benefits of Self-Help's plan are as follows:

1. Affordability. The no-interest balloon reduces the monthly total housing payment for the borrower by eliminating the monthly payment on the second mortgage. As a result, the program can reach significantly lower-income families.

2. Efficiency. The no-interest balloon eliminates the necessity of servicing the second mortgage. Servicing the loan is not only costly (lenders charge $120 per year per loan in our experience), but burdensome for a borrower to pay with a separate check and for a lender to receive and track. Such small amounts of interest after deducting for servicing costs have a low value to investors--funds must be allocated between principal and interest and declared and accounted for as income while not materially adding to their income statement. Pool administration also becomes much simpler if the amount of loans outstanding remains constant until there is a loss or payoff, rather than declining slightly over time and requiring relending of small amounts of repaid principal.

3. Time. A 25-year balloon should not be a burden for a family to pay off: Sixty percent of first time homebuyers move before their 15th year of living in a home to begin with, and even modest inflation and house appreciation will reduce the real value of the second mortgage considerably by the time the loan balloons. Using a 7 percent discount rate, the loan would be worth just 20 percent of its original amount at year 25. In addition, if a borrower pre-pays the loan just before year 10, for example, the lender can make another loan to an additional family for another long-term 15-year balloon, which allows the tax credit to have greatest impact.

4. Protection. By requiring the borrower to pay off the second mortgage when obtaining a cash-out refinance of the first mortgage, the program helps ensure that sufficient equity remains to pay off the balloon. Also, the risk of default increases if the borrower is responsible for making two different monthly payments--one for the first mortgage and one for the second. A no-interest balloon eliminates this risk by eliminating payments on the second.

Table: Additional Differences follows:

Feature	Brookings Model	Self-Help Proposal
Servicing Fees	38 basis points (bp), equivalent to FHA requirement on the second. An additional servicing fee of 28 bp would be allowed on the first loan to induce lenders to originate both the first and the second mortgage.	None, since there is no servicing necessary with a balloon payment. Self-Help does not believe that the amortizing second loans could be serviced as cheaply as the Brookings Model assumes, which forces the homebuyers to pay interest that does not significantly benefit the program. A servicer is not necessary because, upon sale of the property, the new buyer's attorney will see the second deed of trust while doing a title search and the seller's proceeds will pay the lender back at that time.

* * * * * * * * * * * * * * * * * * * *

Loan Amount	18 percent - 22 percent of home value. Calibrated to avoid mortgage insurance by bringing first mortgage to 80 percent loan-to-value ratio. The loan can cover up to 4 percent in closing costs (origination fees are capped at 2 percent) plus the difference between the down payment and 80 percent of the lower of the sales price or appraised value. The maximum loan size would be $25,000 (indexed to FHFB's house price index), except for mortgages used as part of neighborhood rehabilitation strategies (which would have a $40,000 cap).	10 percent--40 percent of home value. Calibrated to make a home affordable to the borrower by bringing housing expense-to-income ratio to a conventional limit for purchase of home (28 percent). Therefore, the amount of the second mortgage is tied to the minimum amount necessary for the borrower to purchase the home. This provides certainty that the second mortgage amount is necessary under the Self-Help proposal for the family to buy the home and allows the program to reach lower incomes while spreading the subsidy over the largest possible number of homeowners. Borrowers who can afford it, then, would receive private mortgage insurance.
Tax Credit Loss and Relending Conditions	Investors would lose the credit when a loan is prepaid within 10 years. Investors with prepaid loans would have first priority for annual credit allocations in order to swap in new loans to substitute for prepaid loans.	Investors receive, each year, the percentage of their tax credit allocation equal to its utilization percentage: the amount of eligible loans outstanding divided by the original pool amount. Thus, investors are penalized pro rata for originally failing to make the promised loans, for losses due to defaults, and for not relending funds when the borrower pays off during the first 10 years.
Loan Number Minimums	100 loans originated to allow enough risk pooling to increase the price investors will pay for the credit.	To be determined by state housing finance agency (HFA). While in a highly populated state 100 loans is probably a reasonable minimum, in small, rural states the HFA may need to break down the credit blocks to a smaller number.
Eligible Geographic Areas for Loans	Under-served areas as defined by HUD. This approach could help ensure the credit targets low-income areas.	Unlimited. Self-Help believes that the limits should be income and purchase price. Limited service areas often discourage loan officer participation while unlimited service areas that focus on income can spur activity in under-served areas without an explicit stipulation. However, HUD definitions may be broad enough to reduce geographic targeting concerns.
Tax Credit Recapture	Mortgage Revenue Bond rules apply: if the borrower sells his home within 10 years and his income has risen above 5 percent over 80 percent of AMI, the borrower must pay the Treasury up to one-half of the profit from the house appreciation.	Self-Help believes that this recapture obligation is too strict for potentially small increases in family income over a long period of time.

Endnotes

1. According to the 1995-1997 Fannie Mae National Housing Survey, 77 percent of Americans prefer to own their home rather than rent, and owning a home in the future is a top priority for 38 percent of low- and moderate-income renters.

2. A second mortgage is a mortgage that has rights secondary to the first mortgage. In other words, the proceeds from a foreclosure sale must pay the first mortgage before any funds can go to repay the second mortgage.

3. See sources collected in J. Michael Collins, Eric S. Belsky, Nicolas P. Retsinas of Harvard's Joint Center for Housing Studies, "Towards a Targeted Homeownership Tax Credit," January 1999, at 3, published by the Brookings Institution's Center on Urban and Metropolitan Policy at http://www.brook.edu/ES/Urban/nic.pdf (hereinafter, "Brookings paper"). See also Michael Collins, Neighborhood Reinvestment Corporation, "The Many Benefits of Home Ownership," November 1998.

4. Thomas P. Boehm and Alan Schlottmann, "Does Home Ownership by Parents Have an Economic Impact on Their Children?" University of Tennessee, Department of Finance and Economics, 1999.

5. Low-income is defined as those at less than 80 percent of area median income. High-income is defined as those at greater than 120 percent of median income.

6. Statistics from 1995 Consumer Finance Survey. The overall national homeownership rate is 66.8 percent.

7. The median net wealth of low-income homeowners under 65 is $57,000; for low-income renters, the figure is less than $5,000.

8. Melvin L. Oliver and Thomas M. Shapiro, Black Wealth/White Wealth: A New Perspective on Racial Inequality (1995), 108. (The average housing unit tripled in value from 1970 to 1980.)

9. See Oliver at 64 and Brookings paper at 2.

10. Self-Help developed this proposal in conjunction with the work of Collins, Belsky, and Retsinas of Harvard's Joint Center for Housing Studies, whose work was published in the excellent Brookings paper (see endnote 3 above). The Brookings paper presents an illustration of a homeownership tax credit that shares many similarities with our proposal (hereinafter, "Brookings model "). Both induce investors through a $1 billion federal tax credit to provide no-or low-interest second mortgages to low-income borrowers to address income and wealth constraints. As in the Low- Income Housing Tax Credit program, both allocate the tax credits to state Housing Finance Agencies, which auction the credits to lenders. However, there are a number of differences between our proposal and the Brookings model that are described in an appendix to our paper. In addition, the Neighborhood Reinvestment Corporation has proposed a pooled homeownership credit as part of the LIHTC program that was helpful in developing our proposal.

11. See endnote 18 below.

12. See, for example, Larry Meeker and Forest Myers, "Community Reinvestment Act Lending: Is it profitable?" Financial Industry Perspectives, (1996), 13-20. The authors found that all banks studied "indicated that losses on their CRA loans were comparable to losses on their conventional loans " (p. 22). Marci Mills, vice president of Bank of America Community Development Bank (writing in Community Developments, February 1998, 3-5), argues that the bank's experience demonstrates that "community development loans and non- CRA loans can perform similarly " (p. 3). Also see Edwin S. Mills and Luan Sende Lubuele, "Performance of Residential Mortgages in Low-and Moderate-Income Neighborhoods," Journal of Real Estate Finance and Economics 9, (November 1994): 245-260. The authors find that single-family mortgages made in low- and moderate-income neighborhoods of US cities actually perform better than comparable national data.

13. See Gary Stern, "Suspending Disbelief: Low Incomes Do Not Mean High Defaults," The Region (June 1994). The experience of Self-Help bolsters the examples offered by Stern: After lending or purchasing $225 million in loans to almost 4,000 families targeting minority, low-income, female and first-time homebuyers--our loan losses are well under 1 percent.

14. See Richard Green and Andrew Reschovsky, "The Design of a Mortgage Interest Tax Credit: Final Report Submitted to the National Housing Institute," September 1997. See Brookings paper, 4-12.

15. Howard Savage, Who Can Afford to Buy a House in 1993? U.S. Census Bureau, 1997. Cited in Brookings paper, at 11.

16. Monthly payments include principal and interest charges plus $100 for property taxes and hazard insurance. Example assumes mortgage is a 30-year fixed rate loan at a 7.5 percent interest rate.

17. The Partnership works because the city provides the buyer with the land for a nominal fee and provides the family with a $10,000 subsidy, and the state provides a final subsidy of approximately $15,000 to make a two- to three-family house affordable to low- to moderate-income families. See Charles J. Orlebeke, New Life at Ground Zero: New York, Home Ownership, and the Future of American Cities (1997) and www.nycp.org.

18. 500,000 families. $1 billion of credit authority is $100 million of credits available each year for ten years; $1 billion of credit authority allocated each of ten years would total $10 billion of credits. Assuming second mortgage lenders purchase credits for 80 cents on the dollar (see Brookings paper, pages 26-28, for related discussion of internal rates of return and tax credit purchase rates), $8 billion of second mortgage funds would be available to be lent. A modestly priced home can be defined as one that is half of the national median house price, or $62,050 (American Housing Survey 1995). If an average second mortgage is 25 percent of the house price, that would be a $15,513 mortgage. 516,000 second mortgages of this size would total $8 billion of lending. $23 billion. This lending would leverage an additional $23 billion of privately funded first mortgages at 75 percent of the purchase prices.

19. "Sale" of credits would occur through creating a pooling structure, such as a limited liability company (LLC), limited partnership, or real estate investment trust (REIT). In the case of an LLC, the for-profit tax credit "purchasers" become members of the LLC. The "seller" of the credits the for-profit or nonprofit lender becomes the managing member of the LLC, and the tax benefits pass through to the "purchasers."

20. The Community Reinvestment Act, enacted in 1977, requires regulated financial institutions to help meet the credit needs of the local communities, including low- and moderate- income neighborhoods, in which they are chartered.

Eric Stein and Martin Eakes are vice president and president respectively of Self-Help one of the largest nonprofit community development financial institutions in the country. See www.self-help.org for more information.