Who benefits from State Revolving Fund Earmarks?

By Walker Grimshaw, Stephanie Vo, and Katy Hansen

The 2023 omnibus spending bill includes hundreds of earmarks from the largest source of federal funds to invest in safe drinking water, wastewater treatment, and stormwater management. Typically, Congress appropriates around $2.5 billion per year for the State Revolving Funds (SRFs), and then the EPA divides the appropriation among the states and territories according to a formula. States then have significant discretion to determine how to allocate the funds. However, Congress started to strip away some of these resources for Congressionally Directed Spending projects–commonly known as earmarks—in 2021. In 2022, the total amount of earmarks increased to 53 percent of total SRF appropriations: Congress earmarked $1.47 billion of the total $2.76 billion SRF appropriation for 715 projects. 

Who benefits from these earmarks?

This analysis shows localities that receive earmarks tend to be less well-off, less white, and larger on average. For example, Santa Rosa, New Mexico, will receive $800,000 for their water system. Santa Rosa has a median household income of $37,052 and qualifies as a severely disadvantaged community under New Mexico’s Drinking Water SRF program. Santa Rosa would likely struggle to finance water infrastructure without federal financial assistance. 

But hundreds of wealthy localities will also receive earmarks. For example, Sharon, Massachusetts, will receive $3.5 million for their water system. Sharon's median household income is $139,385 and they could likely finance water infrastructure without federal support. Providing grants to places that can take loans to finance projects reduces the amount of funds available for communities who are otherwise unable to finance needed projects. Thus, earmarks undermine the commitment to ensure that at least 40 percent of federal funds reach communities most in need. 

Data

We look at the median household income (MHI), poverty levels, racial characteristics, and population of the localities that receive earmarks, similar to our analysis of the allocation of the DWSRF and CWSRF programs. These indicators are decent measures of drinking water and wastewater system needs:

• Median Household Income: MHI indicates the financial capacity of water systems, though it is not a sufficient indicator of economic distress. Many state SRF programs use MHI to identify disadvantaged communities.

• Poverty rates: The proportion of the population living below the federal poverty line indicates the ability of households to pay for infrastructure projects. 

• Race: The percentage of people identifying as non-Hispanic white is often a strong predictor of water quality and access. 

• Population: Localities with small populations often struggle to fund infrastructure projects because the cost per capita for provision may be higher.

We matched 542 of the 715 earmark projects to the localities they benefit. Most of the projects not matched to localities were earmarked for a county, district, or authority. These regional or county-level projects were excluded because the demographics of a county writ large are often different from the demographics of the residents served by a system and we do not know with precision where these systems serve. This exclusion likely ignores many localities in rural areas that benefit from earmarks.

The boxplots, shown in Figure 1, illustrate how the demographics of the localities that received an earmark compare to all others in the state. Each dot represents an earmarked project; the size of the dot is scaled to the funding amount per capita. The vertical black line shows the median value of the MHI, poverty rate, racial background, and population size in each state. 

Figure 1: Boxplots of the demographic characteristics of localities that received 2022 earmarks compared to the state. Some wealthy localities, shown by the pink and purple dots to the far right, receive earmarks.

Results

We use a probit statistical regression model to assess the correlation between demographic characteristics and the likelihood of receiving an earmark. The results, shown in Figure 2, show: 

• MHI: median household income is not correlated with the likelihood of receiving an earmark. The MHI of most localities that received earmarks is below the state median, but some have median household incomes more than twice the state median.

• Poverty rates: localities with higher poverty rates are more likely to receive earmarks. Most localities that receive earmarks have fewer people above the poverty line than the state. 

• Race: Most localities that received earmarks have a lower proportion of white residents than the state writ large.

• Population: Population is positively correlated with the likelihood of receiving an earmark. Most localities that receive earmarks are larger than the median community size in their state. Many large or very large cities receive earmarks.

Figure 2: Probit statistical regression model showing the correlation between demographic characteristics and the likelihood of receiving an earmark.