In the Tax Reform Act of 1969, Congress created the concept of program-related investments, or PRIs. As defined in that law, a program-related investment is an investment (usually a loan) that meets three requirements:

• Its primary purpose must be to accomplish charitable objectives.
• It cannot have, as a significant purpose, the production of income or the appreciation of property (i.e., a prudent investor seeking a market return would not enter into the investment).
• It cannot be used for lobbying or to support lobbying.¹

As long as a PRI meets these requirements, it can be counted, as grants are, toward meeting the 5 percent payout required by law. PRIs enable foundations to (1) stretch their resources (since they can recover the principal and sometimes receive interest); (2) fund projects, such as building construction, that are not usually considered appropriate for grant funding; and (3) invest in high-risk charitable ventures or borrowers that commercial banks tend to avoid.

Although a PRI may generate interest income, it is conceptually very different from a foundation's regular investment portfolio. The objective of a PRI is to further a foundation's charitable purposes, while the objective of a foundation's investment portfolio is to preserve capital and to produce market-oriented returns. Over the past 30 years, a number of foundations have utilized PRIs, although mainly on a limited basis. Among foundations that regularly make PRIs, the Ford Foundation has invested 2 percent of its assets, the John D. and Catherine T. MacArthur Foundation 1.5 percent, and the Irvine Foundation 1 percent.²

Four different circumstances can trigger the consideration of PRIs as a funding mechanism.

First, activities that will produce income for the borrower that can be used to repay the loan are prime candidates. Most commonly, PRIs are used to finance the construction or rehabilitation of buildings that can be rented. Since The Robert Wood Johnson Foundation generally will not provide grant funds for bricks and mortar, PRIs offer a way to finance building construction. They have also been used to provide student loans that can be repaid upon graduation and to support medical practices in underserved areas that can be repaid out of revenues from patient-care.

Second, the lack of credit-worthiness of many nonprofit organizations makes them suitable candidates for PRIs. Banks are often unwilling to make loans for real estate transactions to nonprofit organizations because of their slim operating margins, uncertain funding, inexperience with loans (or finances generally), and lack of resources that can be used as collateral. Foundations can step in and make loans to nonprofit organizations that commercial lenders consider too risky.

A third circumstance in which a PRI can make sense is as a source of gap financing. A dollar gap can appear at any stage of a project. For example, at an early stage, foundations can provide the seed money needed to shape an idea, explore its feasibility, and begin pre-development work such as hiring an architect-elements that often must be in place before banks will consider financing a project. Or, once a project is under way, a PRI can be used to provide bridge financing that will take it from one stage to another.

Fourth, where a program calls for loans to a number of community organizations, it might be appropriate to make a PRI to an intermediary organization that in turn will make smaller loans to fund a number of projects. In this way, a foundation can stretch its limited dollars, spare itself the administrative burden of supporting and evaluating a multitude of small borrowers, and tap into existing development and financial expertise.

The Foundation's first PRI, made in 1982, was a $3 million loan to the United Student Aid Funds, to be used as collateral for scholarships for needy women and rural and minority medical students. Its most recent PRI was made in 1997-a $1 million loan to help the Minneapolis Foundation establish a program expanding medical services in underserved rural and urban areas.

The Foundation has made 26 loans, totalling $35 million. Of this, $16 million was awarded to organizations participating in national programs that included both a grant and a loan component. The remaining $19 million was in loans for programs that had no grant component. A significant amount of the loan portfolio went to establish statewide or national loan programs that then provided credit to many smaller nonprofit organizations. Over the years, the Foundation's experience with PRIs has been more the product of seizing specific opportunities than of forward planning.

PRIs to Finance Building Construction and Renovation

Most PRIs have been made to enable the recipient to put together financing needed to construct or renovate buildings, such as community mental health facilities, hospitals, skilled nursing and long-term care facilities, and day-care housing.

THE COMMUNITY HEALTH FACILITIES FUND

The Community Health Facilities Fund, or CHFF, is a not-for-profit organization that enables community-based behavioral health care organizations to gain access to tax-exempt debt financing by pooling their loan applications. The idea grew out of a national survey conducted in the late 1980s that found an unmet need of nearly $2 billion for the capital costs required to build and renovate treatment facilities for people with mental illness, substance abuse problems, and developmental disabilities. Most organizations providing services were nonprofits whose limited experience, low cash reserves, and tenuous revenue streams hindered their ability to obtain capital financing. It seemed logical that if the organizations' loans could be grouped and the risk spread among them, then investors would be more likely to buy the bonds the group would issue.

Building on this suggestion, in 1991 the Foundation awarded a $5.5 million PRI to establish the Community Health Facilities Fund. The loan was to be paid back with 2 percent interest over a 25-year period. A second loan for $2.8 million was made in 1995.

The loan program went through two phases. The first, between 1991 and 1995, blended traditional bond financing and the concept of group risk-sharing. In Illinois and Florida, bonds were issued to pay for capital improvements and a 10 percent reserve was set aside to cover bad debt. Government agencies reviewed the applications to assure program eligibility and the financial strength of the applicants. The problem was that some of the organizations applying for loan funds-for example, a small clinic needing to expand-were not financially strong enough to attract investors. The solution was to put both strong and weak applicants into a single pool before making the bond offering. The Robert Wood Johnson Foundation's PRI to the Community Health Facilities Fund was used to add an additional 10 percent to the reserve fund to cover defaults, thus making a total of 20 percent set aside to cover bad debts. This additional 10 percent (known as "credit enhancement") gave investors the security of knowing that their risk of losing money was further reduced. Florida raised $23 million and Illinois $7 million in municipal bond offerings.

The second phase, from 1995 to the present, dealt with two problems that were recognized early by CHFF and Foundation staff members. First, organizations needing loans were not usually ready for pooling at the same time, yet the concept required that they all had to be presented to potential investors simultaneously. Second, the approach taken in the first phase was limited to the states in which the offering was made. This precluded national offerings or placing organizations from different states in the same pool. In response, CHFF created a trust instrument that grouped loan applications from a number of organizations throughout the country as they became ready and placed them into a single loan pool. Within the pool, loans were then divided according to risk. The less risky loans were packaged and sold to traditional investors. The riskier loans were held and backed by the Foundation PRI.

To date, the PRIs, totaling approdimately $8 million, have enabled CHFF to help 30 community-based behavioral health organizations throughout the United States secure approximately $85 million in loans to construct new buildings, renovate facilities, and buy equipment. The tax-exempt debt, which nonprofit organizations are able to obtain through the CHFF, provides them with lower interest rates, more capital, and longer repayment periods than commercial lenders are able to offer. It is one of the few programs to give nonprofits access to this kind of flexible, low-cost, and long-term credit. Moreover, according to Chris Conley, the president of Connecticut-based Nonprofit Capital LLC and manager of the CHFF, "PRIs have created a model for delivering a certain type of capital to the nonprofit behavioral care market. The tax-exempt bond market is not set up to finance large numbers of unrated nonprofit corporations that have relatively small capital needs. The CHFF program has created a structure and process to do that."

CHFF loans are paid back through the patient revenue streams. To date, there has not been a single default. The program has demonstrated that nonprofit organizations, even those without a significant credit history, do repay their loans.

THE PROGRAM ON CHRONIC MENTAL ILLNESS

In 1986, the Foundation launched a national Program on Chronic Mental Illness. It funded the creation of Local Mental Health Authorities in nine locations to coordinate the care of people with serious mental illness. The needs of this population include health care, support services, and housing so that they can remain in the community. The Program on Chronic Mental Illness was funded largely with grants, but there was also a PRI component. Loans of $1 million, payable over 10 years at 4 percent interest, were made available to each of the program's sites.

All nine sites took advantage of PRIs. Since the Local Mental Health Authorities were concerned mainly with providing clinical services, and did not, as a rule, have expertise in housing or mortgage financing, Local Housing Development Organizations, linked with the Local Mental Health Authorities, were set up to acquire and rehabilitate housing, to place clients, and to manage the properties. The loans were made to these Local Housing Development Organizations. They, in turn, entered into partnerships with state mental health or financing agencies and used the loans to buy and renovate properties. Most were small family units; even residential complexes were small, with the largest being 4 to 6 units. After the properties had been acquired, the states provided grants and long-term mortgage financing. State funds helped to retire the loans so that additional properties could be acquired. The participation of the U.S. Department of Housing and Urban Development in the program lowered the risk of default. It agreed to provide 125 rental subsidies through so-called "section 8 certificates" for each of the nine sites.

Through the PRI mechanism, the Program for Chronic Mental Illness was able to leverage $9 from federal, state, and private sources for every dollar lent by The Robert Wood Johnson Foundation. To date, over 4,500 units have been developed. The former national deputy director of the program, Martin Cohen, observed that the PRIs made it possible "to sustain large amounts of affordable and safe housing for people with serious mental illness. Already, other organizations and groups (and others without experience) that serve an array of disabled and disadvantaged people who lack needed housing and supports are using this model." All loans due have been repaid.

AIDS HOUSING OF WASHINGTON'S BAILEY-BOUSHAY HOUSE

In 1990, the Foundation awarded AIDS Housing of Washington a $1.5 million PRI, payable over 10 years at 3 percent interest, to help finance its Bailey-Boushay House. Located in Seattle, Bailey-Boushay House was the nation's first newly constructed skilled-nursing and day-health facility for people with HIV and AIDS.

When the 3 percent interest rate of the PRI was combined with the higher market rate of a commercial bank loan, the overall long-term borrowing rate decreased to a manageable 8 percent. This saved Bailey-Boushay House about $500,000 in interest payments over its first ten years of operation, and helped assure the long-term viability of the facility. Bailey-Boushay House has served more than 2,000 people with HIV/AIDS in its 35-bed facility and day health program. The loan has been fully repaid.

OUR COMMUNITY HOSPITAL

A PRI to Our Community Hospital, in Scotland Neck, North Carolina, provided the financing needed to transform itself into a long-term care facility and offer the kind of services needed by its aging population.

Our Community Hospital, in rural northeast North Carolina, was in danger of being closed because the number of patients had decreased and the amount of red ink had increased. In 1989, the citizens of the Scotland Neck decided to convert the hospital to an extended care facility, complete with senior housing, swing beds (which could be used for short- or long-term care), and a primary care facility.

Lacking experience in this kind of transaction, the community turned to the North Carolina Office of Rural Health for assistance. The Office of Rural Health sent an advance team to determine the feasibility of a conversion and to identify sources of financing. The feasibility study yielded a "go" decision. Revenues were expected to come primarily from federal and state governments once the facility was designated as a rural health clinic under federal law.

In 1990, The Robert Wood Johnson Foundation made a $500,000 PRI to Our Community Hospital. The loan was used to fill various gaps in financing the conversion. The state of North Carolina provided matching funds, and the community conducted an old-time fundraising campaign, complete with sales of books, candles, and cookies. Our Community Hospital continues to thrive and to meet the health care needs of its community. The loan from The Robert Wood Johnson Foundation was repaid in 2001.

Loans have been used to finance physicians and other health professionals who agree to set up practices in underserved rural or inner-city areas. Two PRIs are particularly noteworthy: Practice Sights, a national program that encourages health professionals to work in underserved areas, and the hospital-based rural program of West Alabama Health Services.

PRACTICE SIGHTS: STATE PRIMARY CARE DEVELOPMENT STRATEGIES

In the 1990s, the Foundation began to address some of the structural and social factors that discourage health care professionals from practicing in rural areas and inner cities: poor reimbursement, isolation, and the lack of social amenities and professional opportunities for health professionals and their spouses. In 1992, the Foundation initiated Practice Sights: State Primary Care Development Strategies, a five-year program to increase the number of primary care providers in medically underserved areas. Under this program, ten states were awarded grants to train physicians, nurses, and other providers, and to give them incentives (such as forgiving loans and providing assistance in practice management) to work in rural areas and inner cities. In addition, a loan component was added and earmarked largely for the construction, renovation, and equipping of facilities.

Every Practice Sights grantee was eligible to apply for a PRI. After a competitive application process, the Foundation made 10-year, 3 percent interest loans ranging from $700,000 to $1.5 million to applicants in four states. Each of the loan recipients was required to raise five dollars from other sources for every dollar of PRI funds it received. Two states, Virginia and Idaho, designated nonprofit agencies to receive and manage PRIs. They, in turn, placed the loan proceeds in bank certificates of deposit. The banks committed themselves to make loans under their most flexible terms and to manage this portfolio of loans themselves. The two other states, Nebraska and Minnesota, designated nonprofit agencies to receive and manage the PRIs directly. In Nebraska, an area economic development agency agreed to match the Foundation loan on a 5-to-1 basis and to underwrite and manage the loan portfolio. In Minnesota, a community foundation agreed to a similar matching, underwriting, and loan management agreement.

According to an assessment conducted by the Practice Sights National Program Office, the PRIs produced mixed results.³ The Healthy Communities Loan Fund of the Virginia Health Care Foundation, which received the first and smallest loan of $700,000, in 1995, has made 16 loans totalling nearly $1.6 million. These have helped recruit 24 primary care providers (physicians, dentists, nurse practitioners, and physician assistants) to areas needing health professionals. In November, 2000, the project's banking partner, First Virginia Bank, received the Main Street Award from the American Bankers Association's Center for Community Development for its work in this project. Minnesota, which received a $1 million PRI in 1997, has made $1.1 million in loans to date, largely for capital improvement projects.

The two other states-Nebraska and Idaho-have since closed their loan programs and repaid the loans. Both states indicated that the loans were too stringently structured: They felt that the 5-to-1 matching requirement was too high, that the 3 percent interest rate charged by the Foundation was not competitive, and that the loans should have been made for more than 10 years.

WEST ALABAMA HEALTH SERVICES, INC.

In 1991, The Robert Wood Johnson Foundation made a $500,000 PRI to West Alabama Health Services, located in one of the more medically underserved areas in the country. Attracting health professionals to the region is a great challenge. The West Alabama Health Services, a nonprofit organization dedicated to providing health care for people in the region, took an innovative approach to the workforce problem. It decided to borrow money and capitalize a leasing company. Unlike, say, an automobile leasing company, which rents a product, the West Alabama Health Services leasing company employed people-doctors, nurses, respiratory therapists, midwives, and other professionals-whom it rented to underserved communities. The idea was that the leasing company would be able to guarantee a salary to its health professionals, thus attracting them to West Central Alabama. In turn, the community clinic would use patient revenues to make lease payments to the West Alabama Health Services leasing company. Lease payments would then be used to repay the Foundation loan.

The concept was unique to the area, and no other funders or lenders would participate. However, the project had two crucial factors working for it. First, the organization and its director, James Coleman, have a long history of delivering on projects. Second, the organization had an unusual form of collateral. In the event that West Alabama Health Services could not meet payment on its note, it pledged revenues from the Greene County Racing Commission, which oversees the dog tracks that are extremely popular and lucrative in Alabama.

In earlier days, the Foundation used PRIs to guarantee loans that enabled deserving women and minority students to attend medical school to become health professionals. The United Student Aid Funds, a recipient of grants from The Robert Wood Johnson Foundation since the mid-1970s, made loans to needy African-American, Native American, Hispanic, and women medical students. By 1982, there were enough defaults on these loans that the United Student Aid Funds was worried that its reserve fund might be depleted. It requested a grant from the Foundation to increase the reserves. However, the Foundation did not feel that grant funds were appropriate, and agreed to provide the United Student Aid Funds a $3 million interest-free PRI to be used as a reserve against default on the student loans. The PRI was repaid and closed.

Although mechanisms that combine elements of grants and loans are not specifically loans, they do bear a sufficient resemblance to PRIs to merit discussion. These hybrid options are relatively new to the Foundation's work. "Recoverable grants"-non-interest-bearing funding instruments-are one of these mechanisms. The Primary Care Development Corporation received a recoverable grant. Similarly, the Foundation has made grants to organizations that have, in turn, used the money to establish revolving loan funds. The NCB Development Corporation does this to provide pre-development and bridge financing to organizations constructing affordable assisted living facilities, and the Southern Rural Access Program does it for the purpose of increasing access to primary care in rural areas of the South.

THE PRIMARY CARE DEVELOPMENT CORPORATION

In 1993, the Foundation awarded the New York City-based Primary Care Development Corporation, or PCDC, a $1.5 million grant and, in 1996, a $300,000 recoverable grant as part of an ambitious multi-foundation effort to establish a new agency whose goals included developing 40 new or expanded primary care centers and increasing the ability of health care providers and staff to effectively run these centers. A total of $35 million from a variety of public and philanthropic sources supported PCDC's operations, $17 million of which came as a grant from New York City to establish a pre-development revolving loan fund. PCDC would eventually have access to more than $250 million in municipal bond financing for the primary care centers; in addition, PCDC provides a wide range of technical support services to the staffs of the centers.

The recoverable grant helped PCDC pay personnel and other costs while it waited to receive financing and fees it had charged its clients for its services. The fees were to reimburse PCDC for the staff time and other expenses associated with helping clients obtain loans. Under the terms of the recoverable grant, PCDC would begin repayment of the principal 16 months after its award. It would pay the Foundation 10 percent of its positive cash flow starting in June, 1998, and the monthly payments would continue until the principal was fully repaid.

By the end of 1999, PCDC had helped to expand or build 28 health centers, representing a total investment of $100 million. However, the nature of the financing it carried out had changed considerably. Most notably, there was an increased demand for smaller loans. Loan requests from smaller, less experienced organizations require much more assistance in helping clients understand and prepare financing applications. Also, smaller agencies, with less of a track record, have less chance of getting financing than larger, more established applicants. Thus, more PCDC staff was needed to spend more time providing technical assistance to agencies that had less chance of actually getting funded. Since fees are paid only on projects that are approved, the revenues received fell well short of projections. In 1999, the Foundation agreed to forgive the recoverable grant and thereby waive the repayment of the funds.

NCB DEVELOPMENT CORPORATION

In 1992, the Foundation authorized a national program called Coming Home, and it gave the NCB Development Corporation, the national program office, a $6.5 million grant to develop model affordable assisted-living facilities for low-income and frail seniors in rural areas. The NCB Development Corporation used $4.3 million, or about 70 percent of its grant, to establish a revolving loan fund. This fund had originally been conceived as a source of permanent financing capital that organizations would then leverage with other forms of permanent financing.

As the Coming Home program developed, however, it became clear that the loan funds would be better spent in a project's early stages on market research, planning, and other predevelopment work. Banks generally are reluctant to provide loans for these purposes, especially to nonprofits with little real estate experience. Yet they are critical. David Nolan, the director of the Coming Home program, noted, "These costs can easily reach $125,000 and very few not-for-profit community-based organizations have that kind of money to risk. But without spending that money, no project would ever happen. It was the classic Catch-22 situation, but one that the revolving loan fund could remedy."⁴ So the plan shifted and the revolving loan fund was used to enable nonprofits to borrow money in order to pay for feasibility studies, marketing surveys, other pre-construction costs, and bridge financing to prevent delays between different stages of a project. According to Nolan, the revolving loan fund has helped produce 330 affordable assisted-living units and leveraged nearly $32 million in equity, secondary financing, grants, and conventional debt for eight projects.

SOUTHERN RURAL ACCESS PROGRAM

In 1997, the Foundation funded the Southern Rural Access Program to help build the institutional and leadership capacity necessary to improve access to basic health care in eight of the nation's most medically underserved rural states. In part, the program provides grant funds that can be used to establish revolving loan funds designed to help rural doctors, other providers, clinics, and hospitals gain better access to capital financing sources.

Virtually all the projects started in early 1999, so information about the program is limited. However, two states, South Carolina and Arkansas, have already approved more than $4 million in loans, and the projects have been able to secure matching support from a variety of sources, including the U.S. Department of Agriculture (Arkansas, Louisiana, South Carolina, and West Virginia); state government (Louisiana and West Virginia); private philanthropy (West Virginia), and community development financial institutions (Arkansas and Mississippi). All the projects intend to work closely with the private banking community, and to leverage additional capital resources from it.

An interesting feature of the Arkansas and Mississippi projects is the selection of not-for-profit intermediaries-the Arkansas Enterprise Group and the Enterprise Corporation of the Delta-that know economic development and capital financing in distressed rural communities and are now adding health care lending. These organizations' knowledge of capital financing, and their strong working relationship with federal financing sources-including the Department of Agriculture, the Small Business Administration, the Department of Treasury's Community Development Financial Institution program, and the New Markets Tax Credit Initiative, which was just passed by Congress-could create some interesting models in the next several years.

Twenty years and $35 million later, what can be said about program-related investments as a way of carrying out The Robert Wood Johnson Foundation's mission-or, for that matter, any foundation's mission? What lessons have been learned, and what changes should be considered?

First, PRIs are an appropriate, and underused, mechanism for foundations.

Loans allow a foundation to finance nonprofit organizations that want to build or renovate real property or to attract health professionals to underserved areas. Although traditional lenders are often not interested in taking a chance on such high-risk borrowers, this is exactly the kind of risk-taking for the public good that philanthropic organizations were designed for.⁵

In addition to making capital available for projects that are otherwise unattractive to commercial lenders, PRIs have been shown to enhance the financial skills of the nonprofit organizations that borrow money. Negotiating and managing PRIs have provided them the experience on which to build subsequent projects. PRIs that enable small nonprofit organizations to buy a building also give them a sense of security, since they no longer have to worry about rent increases or finding new space if a lease runs out.

Second, the Foundation's experience has demonstrated that nonprofit
organizations, even struggling ones, do repay loans on time.

The repayment record of the nonprofits has been excellent. As commercial lenders see that nonprofits pay back their loans on time, they are likely to be more willing to support projects at earlier stages of development.

Third, PRIs are difficult for both foundations and the organizations seeking loans.

For a foundation, PRIs are time-consuming, require a great deal of paperwork, and have legal and financial implications (such as due diligence requirements) that grants do not require. Foundations are not generally set up to operate as banks, but PRIs put them in the position of acting as bankers. PRIs also raise ethical issues: for example, can a foundation, in good conscience, foreclose on a failed loan to a nonprofit organization doing good work?

For an applicant, gaining access to foundation loans can be difficult. Negotiating a loan agreement requires time and skills that many nonprofits may not possess. On top of that, on any single project an applicant may have to juggle financing from several sources, each with different terms and conditions.

Fourth, the terms, such as interest rates and length of a loan,
are sometimes onerous; flexibility should be built in.

In a number of cases, interest rates fell nationally afer PRIs had been made, and loans could be obtained on more favorable terms than the Foundation had been able to offer. However, borrowers were locked in to the terms that had been negotiated earlier-and had since become non-competitive. In addition, a 10-year repayment period might have been too short, given the length of time it takes to develop the loan programs and the high cost of maintaining them. Making loans that bear no interest and have longer repayment periods-20 to 25 years, say-would alleviate these problems. In addition, consideration should be given to reducing stringent matching requirements, without compromising the Foundation's interest in leveraging other funds.

Fifth, PRIs should include some grant funding for marketing, technical
assistance to borrowers, and loan administration activities.

Generally, the cost of aggressively marketing the loan program, hiring a staff to help organizations apply for loans, and servicing the loans was not taken into account when the repayment terms were negotiated. These are legitimate expenses that benefit borrowers and insure that the PRIs are maximized and used as intended.

In its two decades of experience with PRIs, The Robert Wood Johnson Foundation has tried a variety of approaches. They range from the relatively simple, such as gap financing for Our Community Hospital in rural North Carolina, to the more complex, such as pooling applicants for tax-exempt bond issues in the Community Health Facilities Fund and supporting a company that leases health professionals in rural Alabama. Even now, the Foundation is trying new hybrid approaches that combine grants with loan financing. There does not appear to be a single model that works best, and perhaps that is an advantage. PRIs can, and should, be flexible enough to meet the needs of borrowers and the marketplace.