Surviving the Great Recession: Nonprofit Housing Developers Through the Lens of Organizational Theory

Abstract

This study identified factors that influenced California nonprofit housing development organizations’ (NHDO) survival and financial stability during the Great Recession. NHDO typically develop and manage affordable housing, while providing social services. During the recession, NHDO financial issues were exacerbated and compounded by the elimination of state redevelopment funds. This research tested organizational theories through bivariate and multivariate analyses from Internal Revenue Service 990 tax forms for nearly 800 NHDO. In many ways, the factors that influenced NHDO sustainability and performance were similar to those affecting for-profits and other nonprofits. For example, older and larger organizations with more staff and revenue fared better. Other factors were unique to this sector (e.g., the region and type of housing developed affected outcomes). An important finding was that reliance on government funding was negatively associated with survival and revenue. The lessons learned from NHDO inform other organizations about surviving and thriving during tough economic times.

Keywords:

• Low-income housing
• nonprofit sector
• Great Recession
• organizational theory

Acknowledgments

I would like to thank Drs. Paul Ong and Deirdre Pfeiffer for their input on this manuscript.

Notes

1. The California Department of Finance’s Redevelopment Agency funds were dissolved as of February 1, 2012, which is beyond the timeframe of this study.

2. The LIHTC program is administered by the federal IRS as well as by state allocating agencies. For-profit or nonprofit organizations can apply for tax credits. Projects that serve the lowest income tenants for long affordability periods receive priority for funding. State allocating agencies must set aside 10% of the LIHTC funds for nonprofits (O’Regan & Quigley, 2000).

3. HUD administers the HOME program, which requires that 15% of funds be set aside for CHDO. Typically, CHDO are nonprofit organizations that meet the legal and organizational structure requirements to participate in the HOME program (U.S. Department of Housing & Urban Development, 2015).

4. The NCCS IRS Core Data provide more information, but the Trend Data have been updated and duplicate organizations have been eliminated from them. For California, the National Taxonomy of Exempt Entities-Core Codes (NTEE-CC) most related to housing development were examined: L20, L21, L22, and L24. For these four codes, the Core Data had 984 NHDO in 2000 and the Trend Data had 821 NHDO in 2000. The number of NHDO in both the Core Data and Trend Data was 802 NHDO (including one NHDO outlier from Region 5 Sacramento Metropolitan).

5. Form 990 and Form 990-EZ are used by tax-exempt organizations, nonexempt charitable trusts, and section 527 political organizations to provide the IRS with the information required by section 6033. These forms are required for organizations with annual gross receipts that are normally more than $25,000 (since 2010, the gross receipts amount increased to $50,000) (Internal Revenue Service [IRS], 2015).

6. Form 990-EZ can be filed by organizations with gross receipts of less than $100,000 and total assets of less than $250,000 at the end of their tax year. Since 2000, the gross receipts has changed to less than $200,000 and total assets to less than $500,000 at the end of their tax year (Internal Revenue Service [IRS], 2015).

7. Form 990-PF, entitled “Return of Private Foundation,” is a report that must be filed each year with the IRS by organizations exempt from federal income taxes under section 501 of the Internal Revenue Code. It is an information return and not an income tax return, since the organizations that file it do not pay taxes. Form 990-PF provides financial information, such as sources of income (Foundation Search, 2015)

8. For the full OLS regression models, both the log of revenue and the revenue squared (as the dependent variable) were run, and these produced similar results as revenue. For the OLS regression models, the change in revenue (as the dependent variable) produced similar results. There were two differences: (a) the sign for total revenue in 2000 (in 2010$) changed to a negative; and (b) in the San Francisco Region and other regions 3–7, total revenue did not affect revenue change (and was statistically insignificant). It is expected that the total revenue in 2000 coefficient would be negative because of the following regression formulas for revenue, R_t = α+βR_t-1+γX+ε_t, and change in revenue, (R_t- R_t-1) = α+(β-1)R_t-1+γX+ε_t. In these formulas, R_t = revenue in 2010; α = the intercept; β = vector of coefficients each denoting the slope or the rate of change in the dependent variable for a 1-unit change in Rev_t-1; R_t-1 = revenue in 2000; γ = vector of coefficients each denoting the slope or the rate of change in the dependent variable for a 1-unit change in X; X = vector of covariates; and ε_t = error term. In the model for revenue, the coefficient for total revenue in 2000 is .622 (after adjusting for scaling). If past revenue is subtracted from both sides of the equation, then the equation would be (.622)*(R_t-1–1)*(R_t-1) = (− .378)*(R_t-1). For the revenue change model, the coefficient on total revenue in 2000 is (− .378), which confirms the model is the same except that the coefficient for revenue in 2000 should be less than zero (negative). The full model using the dependent variable, change in revenue, had a lower adjusted R ² at 0.19, compared with the model using revenue as the dependent variable.

9. The age of an NHDO was determined by using the rule date when the IRS provides a determination letter recognizing the organization’s exempt status (NCCS Data Dictionary Core Public Charities, 2000). Age data were changed for some NHDO. There were some negative numbers, and some NHDO reported age as 2000. These numbers were changed to age zero. The total number of NHDO with age zero was 27.

10. The definition of the nonprofit housing development organizations’ total revenue is described in the Appendix according to the Urban Institute’s NCCS data dictionary for IRS 2010 Core Data.

11. NHDO employee salaries reported to the IRS in 2000 were used as a proxy for the number of employees. This is an approximation of the number of employees within an organization, and reflects the size of the organization. Note that this variable would be for paid staff and would exclude volunteer staff. The number of employees is reported in the 2010 NCCS IRS database, but not in the 2000 database.

12. Direct public support is used as an estimate of government funding (NCCS Data Dictionary Core Public Charities, 2000). This direct public (or government) support excludes indirect public support (NCCS Data Dictionary Core Data, 2000).

13. The percentage of public support is the amount of public support divided by the total amount of revenue from the year 2000.

14. Unlike the 10-year census, the U.S. Census Bureau conducts the ACS throughout the year, every year (U.S. Census Bureau American Community Survey, 2015). For the ACS, the Census Bureau randomly samples addresses in every state, the District of Columbia, and Puerto Rico. The ACS “is a nationwide, continuous survey designed to provide communities with reliable and timely demographic, housing, social, and economic data every year” (American Community Survey Information Guide, 2013, p. 2).

15. A total of 26 counties had missing foreclosure data for 2000 from the RAND California data. Nineteen NHDO that survived were in counties with missing foreclosure data, and seven NHDO that were terminated were from counties with missing foreclosure data. The majority of the counties were from Region 7 Northern California Non-Metropolitan, and the remainder were from Region 1 Greater Los Angeles, Region 3 San Joaquin Valley, and Region 5.

16. One NHDO that survived was excluded from Region 5 Sacramento Metropolitan because the total revenue in 2000 exceeded $180 million (converted into 2010 U.S. dollars). This outlier had an extremely high total revenue, which skewed the results.

17. There were 426 new NHDO established during the 2000–2010 timeframe (as indicated by filing a 990 tax form). Thus, approximately 60% NHDO survived and 40% were new in 2010. These new NHDO had 4% greater average revenue compared with NHDO that survived ($1.69 million compared with $1.62 million), but 57% less median revenue compared with NHDO that survived ($282,000 compared with $656,000).

18. Ten NHDO reported negative revenue in their 2000 IRS 990 tax form.

19. Only one NHDO reported negative revenue in their 2010 IRS 990 tax form.

20. This percentage change was calculated for all NHDO that survived, by taking the difference between the average total revenue in 2010 and the average total revenue in 2000 (in 2010$) and dividing this amount by the average total revenue in 2000. For the 618 NHDO that survived, there was an average revenue change variable for each NHDO. Using this variable, the average revenue change for all of the NHDO from 2000 to 2010 was $329,000 and the median was $29,000. The minimum revenue change was approximately − $19 million and the maximum was about $12 million. A total of 224 NHDO had a negative amount of revenue change (or 36% of the 618 NHDO that survived).

21. This percentage change was calculated by taking the difference between the average revenue in 2000 (in 2010$) for NHDO that survived and the average revenue in 2000 for terminated NHDO, and dividing this amount by the average revenue in 2000 for terminated NHDO.

22. Only one NHDO had a negative percentage of direct public support per revenue. Sixteen NHDO had percentages of direct public support that exceeded 100%.

23. According to the U.S. Census, by the end of 2000 the U.S. rental vacancy rate was 7.8%, and in 2010, the rate was 9.4%. In California, the vacancy rate was 4.5% in 2000 and 7.5% in 2010. In metropolitan areas with a high demand for rental units, there may be low vacancy rates and rental prices may increase (Gabriel & Nothaft, 1988). An excess supply of rental housing may result in a high vacancy rate and lower rental prices (Gabriel & Nothaft, 1988).