Methodology (Last update: January 26, 2021).
Also, please check the guidance of federal response to COVID-19 from
Congressional Research Services, USASpending.gov, USASpending Datalab, and GAO Report.
As of January 25, 2021, 2020, $1.7 trillion has been paid to individuals, organizations, and corporations via various channels, excluding the $900 billion Coronavirus Response and Relief Supplemental Appropriations Act on December 27. This current paid-out amount indicates an outlay in COVID-19 Federal Funds Tracker. The total budgetary resource made available was $2.7 trillion, federal funds that have been promised to be spent was $1.9 trillion, and the remaining balance, which has not yet been declared to be spent, was approximately $0.8 trillion (Check USASpending.gov). The total obligation that the federal government promises to provide will be $2.8 trillion, including the $900 billion Coronavirus Response and Relief Supplemental Appropriations Act. We will keep the tracker updated as soon as possible, employing newly disseminated datasets.
About COVID-19 Federal Funds Tracker:
COVID-19 Federal Funds Tracker measures primarily seven federal fund facilities: Paycheck Protection Program (PPP loans), Unemployment Assistance, Financial Assistance, Contracts, Economic Impact Payment, Health and Human Services Funding, and Medicare Accelerated and Advance Payments. Additionally, we separately show Coronavirus Relief Fund (CRF) in the state hexagon map, with about $150 billion directly delivered to states and municipalities. We also combined economic indicators exploiting proprietary data sources, such as community mobility reports from Google, aggregate consumer spending on credit and debit cards registered from Affinity Solutions, monthly housing values from Zillow, etc. (See the details in Data Descriptions below).
Given that government agencies publish data on their individual platforms rather than on a centralized platform, there are accessibility and comprehension limitations for stakeholders (Check COVID-19 Spending: Known Data Limitations from USASpending.gov). Specifically, different datasets cover heterogeneous geographic regions and industry details. For example, while PPP loans, Financial Assistance, and Contracts datasets can examine the county-level, Unemployment Assistance, Economic Impact Payment, Health and Human Services Funding, and Medicare Accelerated and Advance Payments are only able to explore state-level data (some are city-level).
Data Descriptions (Click data names to see details from their websites):
Paycheck Protection Program (PPP) Loans data (zip code-level) are available to download from the U.S. Small Business Administration (SBA). Using HUD USPS Zip Code Crosswalk File, we generated PPP Loans county-level data. PPP Loans data contain an industry identification code, the North American Industry Classification System (NAICS). We captured the main industry code, the first two digits of the code. COVID-19 Federal Funds Tracker’s PPP Loan Fund Share by Category was segregated using this code.
Note: SBA’s Economic Injury Disaster Loan Emergency Advance (EIDL Loans) data are duplicated in the Pandemic
Response Funding data (See the description of Financial Assistance & Contracts (Pandemic Response Funding) below). We did not combine
EIDL Loans in the amount of PPP Loans in the COVID-19 Federal Funds Tracker.
Unemployment Assistance data (state-level), Families First Coronavirus Response Act and Coronavirus Aid, Relief, and Economic Security (CARES) Act Funding to State, are available to download from the U.S. Department of Labor. The data reveals Section 2102 – Pandemic Unemployment Assistance (PUA) benefits funding authorized under PL 116-36; Section 2103 – Emergency Relief for government entities and non-profits, authorized under PL 116-136; Section 2104 – Federal Pandemic Unemployment Compensation (FPUC) benefits funding, authorized under PL 116-136; Section 2105 – Temporary Full funding of First Week of Regular Compensation, authorized under PL 116-136; Section 2107 – Pandemic Emergency Unemployment Compensation (PEUC) benefits funding, authorized under PL 116-136.
Economic Impact Payment (EIP), also known as Stimulus Checks, data (state-level) is available to download from Internal Revenue Service (IRS). The first round of Economic Impact Payments provides a $1,200 stimulus check to individuals less than $99,000 threshold, head of household less than $146,500.
Note: Once the IRS and Treasury deliver the second round of payments, we will keep updating when the data is disseminated.
Financial Assistance & Contracts (Pandemic Response Funding) data (zip code-level) is available to download from USASpending.gov (Updated regularly).
Two memorandums, M-18-08, Guidance on Disaster and Emergency Funding Tracking (February 2, 2020), and M-20-21, Implementation Guidance for Supplemental Funding Provided in Response to the Coronavirus Disease 2019 (April 10, 2020), guide disaster funding process requirements – especially COVID-19 related spending. The Office of Budget and Management (OMB) segregates the data code based on the need in COVID-19 expenditures, known as DEFC codes. Recipients of disaster funds – particularly in the COVID-19 funds – are responsible for reporting their spending via the Government-wide Treasury Account Symbol Adjusted Trial Balance System (GTAS). The codes allow researchers to examine and confirm how federal funds have transferred exploiting fund categorization. Each recipient possesses one specific identification ID, 'Assistance Award Unique Key,' and data delineate 'when,' 'why,' or 'how much' of the fund has been received in variables. This Pandemic Response Funding data contain variables that reveal the purpose of the federal expenditure, such as why those recipients received their money; for example, 'Program Activities Funding This Award' informs a fund facility and "Award Description Program" refers to the purpose of recipient spending. The collected data related to the COVID-19 crisis from OMB disseminates to the public via USASpending.gov since July 10, 2020.
Note: We called it Pandemic Response Funding data following the definition by Pandemic Oversight. This data was also known as USASpending data.
Note: Pandemic Response Funding data may not be tagged the Coronavirus Preparedness and Response Supplemental Appropriations Act, 2020 or
the Families First Coronavirus Response Act with a DEFC code due to the timing of the Memorandum's in-effect date.
Two significant data limitations from Pandemic Response Funding data:
Coronavirus Relief Fund (CRF): CRF is the amount that state and local governments received from the federal government, approximately $150 billion. Pandemic Response Funding data were not sophisticatedly designed to track spending after distributing funds because of the duplicated county or regional names. For example, the data do not contain a state name that eliminates duplicated county or regional names. We resolve this issue by using 1) other variables that give us hints on the location of duplicates, and 2) searching all information from various proprietary sources (Check more details in Coronavirus Relief Funds (CRF) section below). In the state hexagon map, COVID-19 Federal Funds Tracker checks how states and localities have spent CRF. We will keep it updated as soon as possible when Pandemic Oversight updates the latest version.
Provider Relief Fund: Health and Human Services COVID-19 Funding contains four funding categories: COVID-19 Awards, Provider Relief Fund, RHC Testing Fund, and Uninsured Relief Fund (Check more details in Health and Human Services COVID-19 Funding section below). Currently, Pandemic Response Funding data provides only Provider Relief Fund; this fund is distributed directly to a single entity in Salt Lake City in Utah, making payments to recipients across the country. This causes Utah’s amount of Pandemic Response Funding to be over-stated. COVID-19 Federal Funds Tracker eliminated Provider Relief Fund’s amount; therefore, the amount of financial assistance declined by approximately $100 billion, compared to USASpending.gov. In lieu of abolition, we replaced it using Health and Human Services COVID-19 Funding data.
Note: Business Type Code X and Award Base Action Date was on April 10, 2020, in Salt Lake City in Utah was eliminated, $100.6 billion.
Note: On January 26, 2021, we found that Provider Relief Fund, Rural Health Clinics (RHCs) Fund, and Uninsured Relief Fund were transferred
from Salt Lake City in Utah to Minnesota (9900 Bren Road East, Hopkins, Minnesota). We reflected this change to the COVID-19 Federal
Assistance Award Unique Key: ASST_NON_RHC20200001_7526 ($217M)
Pandemic Response Funding Data Details
Pandemic Response Funding data contains two primary fund types: Financial Assistance and Contracts. Fiscal Assistance is similar to subsidies because the government distributes money to recipients without any transactional purposes, such as purchasing recipients’ goods or services. By contrast, Contracts primarily require transactions between government agencies and vendors so that recipients must deliver appropriate goods or services in a proper time; for example, medical companies that received money from the government agency under COVID-19 related bills provide medical equipment. COVID-19 Federal Funds Tracker segregates two types because the characteristics of the two are very disparate. As of January 25, 2021, the Pandemic Response Funding data consist of ten Assistance_PrimeAwardSummaries CSV files regarding Financial Assistance (approximately 1GB and 1 million observations per each dataset), an Assistance_Subawards file (47,800 observations), a Contract PrimeAwardSummaries (442,500 observations), a Contract Subawards (29,300 observations), and other auxiliary files related to the data, such as the readme file.
Note: We also used the HUD USPS Zip Code Crosswalk File during the cleaning process in order for consistency in the data merging process. Although
the Pandemic Response Funding Data recently added county name variable, formats of those two, HUD USPS Zip Code Crosswalk File and
Pandemic Response Funding files, varied. We combined state and county codes from the data and generated FIPS codes, including State and
County FIPS (5-digit number). Some state and county names that were missing, such as “Northern Marianas”, FIPS code “69100,” “69110,” and
“69120,” we manually added to the dataset.
Note: $1.6 billion Global Assistance was included in COVID-19 Federal Funds Tracker’s total amount of federal funds; however, the amount is not
shown in state or county maps.
Financial Assistance Data consist of a wide range of fund sources. The main concern of Financial Assistance
data was a lack of acknowledgment of coverage, whether the data contained information of a specific fund or
not. During the data cleaning process, we confirmed that Paycheck Protection Program, Unemployment
Assistance, Economic Impact Payment, Health and Human Services Funding, Medicate Accelerated
and Advance Programs, and Contracts need to be merged to measure the magnitude of COVID-19 federal
funds because some datasets, such as Airline Industry Support (Coronavirus Relief – Pandemic Relief for
Aviation Workers, etc.), Education Stabilization Fund, or FEMA Disaster Assistance to Individuals and
Households, and so forth. In sum, there is no need to add such auxiliary datasets from other entities’ datasets.
Note: the full description of fund details can be found in the Pandemic Response Funding Data.
Here are the top 10 Financial Assistance lists in Pandemic Response Funding Data (as of January 25, 2021):
Specifically, Financial Assistance Data consist of Prime Awards and Subawards. The prime awardee indicates a stakeholder that directly receives funds from the federal government, and a sub-awardee is the final recipients, which received funds via prime awardee. As of January 25, 2021, we exploited only prime award datasets because subawardee-related data take time to be collected (lagging). When a fully covered Subaward dataset is available, we will update it.
We categorized financial assistance recipients eight different categories (State Govt, City or County Govt, Other Govt, Regional Organizations, Nonprofit (501C3), Small business, for Profits, Individuals, Higher Education, and Other) employing business type code and business type description variables in the Prime Award dataset. As of January 25, 2021, we saw that the allocation to the state, county, and city governments was approximately $150 billion, similar to CRF's amount. In the later stage, we will segregate details following the main categories of fund purposes, such as airline industry support, educational purpose, etc.
Similarly, Contracts data consist of various variables, and such variables allow to keep track of funds. One crucial difference is that Contracts data do not explain characteristics of awardee information. In contrast, Contracts data contain some additional descriptions that introduce funds’ purposes, such as agency names and the North American Industry Classification System (NAICS). These variables explained which agency awarded funds or what kind of business industries that awardee participate in. Using the NAICS’s first two-digits, we segregate the business industry in COVID-19 Federal Funds Tracker.
Health and Human Services (HHS) COVID-19 Funding data (state and city-level) are available to download from HHS’ Tracking Accountability in Government Grants System (TAGGS). TAGGS updates four types of COVID-19 related spending categories, COVID-19 Awards (emergency supplemental appropriation funding to the COVID-19 vaccine developing companies), Provider Relief Fund (health care provider support, including hospitals, front-line workers, and other health care providers), Rural Health Clinics (RHCs) Fund (helping residents in underserved rural areas have access to care), and Uninsured Relief Fund (attributing to the testing and treatment of COVID-19 for uninsured individuals).
Medicare Accelerated and Advance Payments data (state-level) are available to download from the Centers for Medicare & Medicaid Services. This program is designed to alleviate healthcare providers' financial burdens (Part A providers and Part B suppliers), accelerating cash flow.
Note: see more information at https://www.cms.gov/files/document/accelerated-and-advanced-payments-fact-sheet.pdf.
COVID-19 Cases data (county-level) are available to download from New York Times. We used the 7-day moving average of COVID-19 cases.
Note: In the county-level data, New York city-related observations were not county-level. It was a city-level. Therefore, COVID-19 Federal Funds Tracker in New York County addresses New York City cases, not New York County. Additionally, the data contained duplicates in Joplin and Kansas City in Missouri (same County FIPS). We dropped two duplicates.
Coronavirus Relief Funds (CRF) data in the state hexagon map are available to download from Pandemic Oversight. Given the data limitation that Pandemic Oversight Data do not have any state or county identification variable, it was difficult to track funds. For example, if the data showed a recipient was “ORANGE, COUNTY OF,” we were not able to know whether the observation belongs to Orange County in California or Florida.
However, we figured out how to confirm the county-level information using other variables, such as location Information (most location information is missing but still useful), descriptions of prime recipients, or amounts of the prime award, using text analysis. We generated state name and code if the location information variable contained, ", State Abbreviation (i.e., ", CA"). Among 46,000 observations, 25,000 had location information. The remaining 21,000 observations were confirmed manually. Firstly, we captured the information using the prime recipient description. For instance, if it contained a unique identification, such as "TREASURY, CALIFORNIA STATE," we could add observations to the state variable. Also, we checked all Indian Tribe Governments' website manually and confirmed the location of their tribes. After this process, about 200 observations were still missing because of duplicates i.e. the same county or regional names existing in different states, such as Orange County. To resolve it, we used the prime recipient award. For example, Orange County has two different prime award amounts, $243,146,628.5 and $554,133,764.9. Using this difference and other variables such as the 'subrecipient' and 'locationifo' variables in data, we confirmed the amounts belonging to which state.
Note: as of January 25, 2021, the CRF data have not been updated since September 30, 2020. We will keep it updated when newly updated data are disseminated.
Demographic Information data are available to download from the U.S. Census Bureau. We used the American Community Survey (ACS) 5-year Estimate (2014-2018) to show the population of regions and Per Capita in the state hexagon map.
CDC Social Vulnerability Index (SVI) data are available to download from the Agency for Toxic Substances and Disease Registry. Social Vulnerability Index measures the preparedness of responding to hazardous or disastrous situations using various indicators, such as poverty, access to transportation, crowded housing, etc. The index indicates whether the region is at higher or lower risk given the unprecedented situation (0 to 1 scale, close to 1 means the area is riskier, and the national average was 0.51 in 2018). We attached this information to the county map using the 2018 SVI index.
Work-in-progress (January 25, 2021)
Economic Indicators (Last data update: January 7, 2021):
Consumer Spending: Affinity Solutions
Google Mobility: Google
House Price Index: Zillow
Employment: Paychex, Intuit, Earnin, and Kronos
Job Posting: Burning Glass
Merchant Revenues: Womply
Education Engagement: Zearn
Requests for additional information on the data or questions can be directed to
© Public Finance Institute. 2021
Public Finance Institute is a public finance research collaborative based in New York City seeks to engage with critical public policy and budgetary questions in domestic and international contexts. The collaborative works with researchers and policymakers to improve public finance practice and education.