Potential Challenges to Targeting Low- and Moderate-Income Communities in a Time of Urgent Need: The Case of CDBG-DR in New York State after Superstorm Sandy

Abstract

New York State received $4.5 billion in Community Development Block Grant-Disaster Recovery (CDBG-DR) funds after Superstorm Sandy. A major CDBG-DR requirement is to prioritize assistance to low- and moderate-income (LMI) populations. The state is spending two fifths of funds on community-wide (e.g., infrastructure) recovery activities. For these activities to be documented as LMI, a specified percentage of residents benefiting from them must be LMI. We explore the potential tension between addressing community recovery needs and prioritizing LMI assistance. Specifically, we develop a series of scenarios to estimate the likelihood that any community-wide activities will be documented as LMI in New York State. We find that documenting these activities as LMI is largely dependent on the underlying demographics of disaster-impacted areas. Additionally, as recovery activities increase in size, thereby impacting larger populations, they are less likely to be documented as LMI, potentially disincentivizing larger, more impactful investments. We recommend empirically based LMI targets for CDBG-DR grantees.

Keywords:

• CDBG-DR
• CDBG
• Sandy
• recovery
• resiliency
• HUD

Acknowledgments

We would like to thank Maria Jessa Cruz and Kelly Richardson for their invaluable contributions to this research.

Notes

1. As amended in 24 CFR 570 (https://www.hudexchange.info/resources/documents/24-CFR-Part%20-570-CDBGs.pdf); more information is available at https://www.hudexchange.info/resources/documents/Basically-CDBG-Chapter-21-CDBG-DR.pdf

2. The program was originally developed by the White House and HUD on two separate tracks, and each allocation scheme gave a different weight to the three factors. In the final formula, poverty received double weight compared with population and overcrowding (Nathan et al., 1977).

3. Also see HUD’s video presentation, min 9:00, available at https://www.youtube.com/watch?v=CW3GwKSOuTM; and FDS report, available at https://www.huduser.gov/portal//publications/pdf/cdbg_redis_eff_v2.pdf

4. HUD lowers this threshold in communities with “no or very few areas in which 51 percent of the residents are low and moderate income.” (HUD, 2017) For example, Nassau County block groups in 2015 qualified as LMI if at least 38.05% of their population, rather than 51, earn less than 80% of AMI. A detailed table is available in Appendix A, as well as at https://www.hudexchange.info/programs/acs-low-mod-summary-data/acs-low-mod-summary-data-exception-grantees/

5. These notices are published in the Federal Register. Examples include the Gulf Coast hurricanes in 2005 (GPO, 2006, p. 7666), the multiple disasters in 2008 (GPO, 2008, p. 52870) and 2011 (GPO, 2012, p. 22583), the severe storms and flooding that happened in 2010 (GPO, 2010, p. 69097), Hurricane Sandy in 2012 (GPO, 2013, p. 14329), and Hurricane Joaquin in 2015 (GPO, 2016, p. 39687).

6. Because of a sequestration order from the President on March 1, 2013, pursuant to Section 251A of the Balanced Budget and Emergency Deficit Control Act, funding was reduced to $15,180,000,000. The Act is available at https://www.congress.gov/bill/114th-congress/house-bill/4512/text.

7. We also had access to the FEMA Public Assistance (FEMA PA) data, that provide information on community-level recovery activities such as debris removal, emergency protective measures, and the repair, replacement, or restoration of disaster-damaged, publicly owned facilities and the facilities of certain private nonprofit (PNP) organizations. However, the geographic information accompanying this data set has reliability issues; therefore, we decided to use the FEMA IA data because of the greater reliability and accuracy of its geographic data.

8. As a robustness check for both of these categories we also considered (a) the top two quartiles, and (b) the top decile of both general and major/severe damage. The results, confirming the same general point, are provided in the following note.

9. Our robustness analysis indicates that the percentage of damaged properties located in LMI areas with concentrated damage ranges between 16 and 21, and the proportion of LMI block groups ranges between 15 and 24% depending on geographic scale and damage concentration. Results are summarized in the table below:

Note. LMI = low- and moderate- income; NYC = New York City; MID = most impacted and distressed.

10. An example is the Village of Babylon Carlls River Tributary project, which entails comprehensive study of the watershed and implementation of the improvements identified. This project serves an area of over 70 block groups in Babylon, Suffolk County. In contrast, the service area for the Municipal Complex in the Town of Blenheim, Schoharie County, which houses the Fire Department, Emergency Operation Center (EOC), Town Hall, Highway Department and an emergency shelter, is only one block group. This information is obtained from the State’s Q2 2017 Quarterly Performance Report, on pages 61 and 79, respectively. Available at: https://stormrecovery.ny.gov/sites/default/files/crp/community/documents/Q2_2017_QPR_Sandy_NYS.pdf

11. The exception thresholds are only applied when all block groups belong in the same county. In cases where the service area consists of block groups from exception and nonexception counties, the higher threshold of 51% is applied. In cases where service areas consist of block groups from different exception counties, the highest of all thresholds is applied.

12. It should be noted that by spatially adjacent we refer to census tracts with consecutive 11-digit Federal Information Processing Standard (FIPS) codes, which in rare cases may not be immediate neighbors. The methodology entails selecting one census tract, and then selecting the next one, two, four or nine census tracts that immediately follow it in the data.