Most CDBG-DR projects must meet both an eligible activity and national objective to be allowable. The process of developing and/or reviewing a proposed or potential project for the ways in which it will meet a CDBG-DR eligible activity and national objective can feel overwhelming. To reach the goal of restoring and revitalizing communities that have been hit hard by natural disasters, grantees need to first navigate the ins and outs of the CDBG-DR requirements to determine if their projects are eligible.
When planning CDBG-DR programs and projects necessary for longer-term disaster recovery, it is critical to maximize the use of the low- and moderate-income (LMI) national objectives, as most grantees are required by HUD to spend at least 70 percent of the total grant on activities that primarily benefit low- and moderate-income households, depending on the requirements of the applicable Federal Register.
Using the LMI Area Benefit (LMA) national objective is one way to secure CDBG-DR funds by meeting one of the national objectives that primarily benefits low- and moderate-income households. However, it can be a challenge to meet criteria when faced with:
- Clearly defining the service area; and
- Calculating the percentage of low- and moderate-income residents in the defined service area
This post is designed to help CDBG-DR grantees understand the LMA activity requirements and run the necessary calculations to find out if their projects qualify for this national objective.
The residents located in the defined service area and who benefit from the project are often referred to as “beneficiaries.”
An important—but often misunderstood—way to meet the LMI requirement is through the Low- and Moderate-Income Area Benefit (LMA). To find out if a project is eligible for LMA, grantees need to fully understand the area that an activity is serving, and then gather information—either through census data or through a direct survey—about the income status of residents in the defined service area. When using the LMA national objective, every CDBG-DR grantee must ensure that at least 51 percent of the residents in the project’s defined service area who are benefitting from an activity are low- or moderate- income (LMI).
How do grantees determine an activity’s service area, and how do they apply survey and/or census data to arrive at a reliable LMI calculation? A close reading of the citation will help us answer these questions.
We start by unpacking the LMA activity requirements…
To better understand LMA, we should unpack the requirements of 24 CFR 570.483 (a)(1) for states and their subrecipients and 24 CFR 570.208(a)(1) for entitlement communities. The two citations are essentially identical, except that the state CDBG citation includes the option for states to allow units of general local government (UGLGs) to use survey data in addition to census data. From 24 CFR 570.483(a)(1) an LMA activity is:
An activity, the benefits of which are available to all the residents in a particular area, where at least 51 percent of the residents are low and moderate income persons. Such an area need not be coterminous with census tracts or other officially recognized boundaries but must be the entire area served by the activity. Units of general local government may, at the discretion of the state, use either HUD-provided data comparing census data with appropriate low and moderate income levels or survey data that is methodologically sound. An activity that serves an area that is not primarily residential in character shall not qualify under this criterion.
This paragraph contains six key elements to bear in mind when using LMA as the national objective. We’ll discuss them one by one:
Activity — The first criteria is that LMA can only be used if the activity is eligible for CDBG funding. This seems obvious but is sometimes overlooked — particularly when a grantee tries to make the activity fit the LMA service area and loses track of the purpose of the activity. In the context of CDBG-DR, not only does the activity have to be a CDBG-eligible activity, but this also includes a requirement for the activity to clearly tie back to addressing an unmet need from the qualifying disaster.
Available to all the residents — In order for an activity to be LMA, it must be available to all residents in the service area. So, for example, a wastewater treatment plant cannot qualify under a city-wide census unless it serves all the city’s residents. Likewise, assisting a marina that restricts membership could not be funded under the LMA national objective.
At least 51 percent LMI — At least 51 percent (not 50.5 percent—no rounding) of the persons in the service area must be low- or moderate-income, based on the average median income for the area served by the activity. We will discuss how to calculate the LMI percentage for a service area below.
Entire area served by the activity — In order to calculate the LMI percentage, one must first identify the service area for the activity. As the regulation states, all persons within the entire defined service area must be included in the calculation, even if the service area is not coterminous, meaning it does not share the same boundary with census tracts. It is incumbent on the grantee or its subrecipients to clearly support and document how the boundaries of a project’s service area are defined. Grantees must exercise caution when using city- or county-wide census data to qualify an activity under LMA—not all activities serve all the residents of an area. For instance, the service area for a neighborhood park probably does not include all city residents, unless there is some feature (e.g., a water park) that other parks do not contain and that attracts residents from across the city.
Census or Survey — HUD allows grantees to use census data or methodologically sound survey data to calculate the LMI percentage, or a combination of the two. Use of the census data is rather straight forward with the help of this HUD dataset and map service. As for the use of survey data, HUD has released the linked CPD Notice to describe how to conduct a methodologically sound survey.
Primarily residential— This is one of the most overlooked requirements when determining whether LMA is appropriate. Some grantees have wrongly tried to use LMA as the national objective for an activity located in industrial or commercial areas with little or no residential components. LMA activities must first and foremost benefit residents—not business owners or county judges.
Armed with a solid understanding of the LMA activity requirements outlined above, we are ready to proceed.
….next we define the service area….
Defining the service area is the first critical step in determining whether a project will qualify under the national objective of Low- and Moderate-Income Area. Typically, this national objective is used for repairs to public infrastructure projects that have a defined area of benefit, such as:
- Public parks
- Water or sewer projects
- Drainage or street improvements
- Community centers
- Public schools
- LMA can also be used for some economic development projects, such as providing infrastructure or direct assistance to a grocery store that primarily serves a residential neighborhood where at least 51 percent of the residents are LMI.
When defining the service area, grantees should identify and document answers to the following questions before determining whether the project can qualify under LMA:
- Who is currently using or who will use the public facility or service: a neighborhood, village, city, county, or residents on a handful of streets?
- Is the area primarily residential?
- Do the geographic boundaries of the service area of the project coincide with a census tract?
…and then we see if we can use census data, survey data, or a combination of the two
Census data for LMA:
If a grantee is able to document and support that a service area coincides with one or more census boundaries, the grantee may use HUD-provided LMI Summary Data for particular census tracts to determine whether the project will primarily benefit low- and moderate-income individuals. Some words of caution if attempting to coincide a service area boundary with a census tract boundary:
- Ensure the service area is truly available to all the residents in the census tract; this may be difficult to support, particularly on location-specific urban projects (e.g. street improvement) or in rural areas where a public infrastructure project is not expansive enough to service the entire population in the larger census tracts; and
- As always, double check that the coinciding census tract and service area are primarily residential
Survey data for LMA:
If a service area is not coterminous (i.e. does not coincide with a census boundary) or if the service area includes parts of multiple census tracts, the grantee may elect to perform a survey of an entire area. Additional information on how to perform a survey on new projects can be found here and a detailed example is provided below, under Census Tract 2, Step 1.
Census and survey data for LMA:
Grantees can also combine census and survey data to calculate LMI for a service area—using census data for tracts entirely within the service area coupled with a survey for the rest.
For example, a grantee determines that a service area includes all of Census Tract 1 containing 3,000 people and a section of Census Tract 2.
Census Tract 1:
HUD’s LMI dataset for Census Tract 1 shows that 1,800 of the 3,000 people in Census Tract 1 are low or moderate income (or 60 percent).
Census Tract 2:
Based on initial estimates, the grantee determines the service area within Census Tract 2 consists of 400 families. For Census Tract 2, the grantee uses the Sample Size Calculator cited in HUD’s CPD Notice on Survey Methodology to determine that it needs to survey 196 families (at 95 percent confidence level and a confidence interval of 5). After surveying all adult persons within those 196 families, the grantee determines that in the service area within Census Tract 2:
- The sampled 196 families are made up of 420 individuals;
- On average the sampled families contained 2.1 persons (420 divided by 196 = 2.1);
- 200 of the 420 sampled individuals are low- or moderate-income;
- This means 47.6 percent (200 divided by 420= .476, or 47.6%) of the 420 sampled individuals in the service area within Census Tract 2 are low- or moderate-income
Since the survey and sample was methodologically sound, the grantee may:
- Step 1: Extrapolate the results from the Census Tract 2 sample population (196 families) onto the total population (400 families) in the Census Tract 2 defined service area; and
- Step 2: Combine the extrapolated Census Tract 2 data with the HUD-provided census data from the entire Census Tract 1
To do this, the grantee must extrapolate the number of people from the sample of 196 families.
- In the sample, each family contained an average of 2.1 persons (420/196)
- So, the extrapolated population count from the 400 families in Census Tract 2’s service area would be (400*2.1) = 840
- Using the same percentage as the sampled population, 47.6 percent of the total Census Tract 2 service area, or 399 persons are low or moderate income (840 * 47.6%) = 399
Now we are ready to combine the two populations to get the LMI percentage for the entire service area:
Since the project serves an area where at least 51 percent of the residents are low- or moderate- income persons, the activity qualifies for the LMA national objective!
By taking the time to understand the LMA activity requirements and by following the census + survey data calculation example above, grantees in pursuit of CDBG-DR funds can find out if their projects are eligible for LMA–without feeling overwhelmed by the process.
Each federal appropriation carries specific alternative requirements and/or waivers which may impact grantees’ methodology for determining whether a residential area is primarily low- or moderate-income. Be sure to review and understand the LMA requirements in the Federal Register Notices applicable to your grant prior to undertaking activities that aim to meet the low- or moderate-income area national objective.