NYC Opportunity
Published in

NYC Opportunity

More Equity, but also More Assistance Funds Going to Banks: the 2021 Paycheck Protection Program

This is the seventh post in the COVID Data Analysis series by NYC Opportunity’s Poverty Research Team. This post assesses whether the 2021 PPP funds have been distributed equitably in New York City using the U.S. Small Business Administration’s Paycheck Protection Program loan-level data. It follows our January, 2021 report on the first round of PPP loans.

The Paycheck Protection Program (PPP) is a forgivable loan program created by Congress in March 2020 to respond to the economic crisis small businesses faced due to the COVID-19 pandemic. Through August 2020, the program, administered by the U.S. Small Business Administration (SBA), provided approximately 5.2 million loans, totaling just over $525 billion, across the United States. Though PPP helped many small businesses stay afloat during COVID-19 shutdowns, a sizable number of small businesses were ignored by bank lenders or could not receive a loan before the funds were exhausted.

As documented in our previous post, the first round of PPP relief failed to reach many of the city’s vulnerable small firms in the hardest hit sectors and in low-income minority neighborhoods that were deeply wounded by the pandemic. Inequities in the distribution of the first round of PPP funds were an essential byproduct of systemic injustice. This is because PPP loans were originated through the existing banking infrastructure, but small businesses owned by people of color have historically had difficulties getting capital from traditional banking systems due to a decades-long history of racially discriminatory lending practices.

In late December 2020, the PPP program was reauthorized and revised with the Economic Aid to Hard-Hit Small Businesses, Nonprofits, and Venues Act (Economic Aid Act hereafter). The program reauthorization was accompanied by reforms to improve targeting of supports to smaller and minority-owned businesses, including upscaling PPP loan size for food service sectors from 2.5 to 3.5 times of average monthly payroll costs; limiting eligibility for “second-draw” PPP funding to firms with up to 300 employees that could demonstrate 25% revenue losses¹; allowing only smaller community banks — i.e., those with up to $1 billion in assets — to take the first crack at second round PPP loans from January 11 to 19, 2021; and an expansion of permissible costs.²

To further promote inclusive and equitable relief for small business, the Biden administration announced the following PPP reforms³ in late February 2021:

  • A 14-day exclusive application period during which only small businesses with fewer than 20 employees could apply for a PPP loan
  • Revising the PPP loan calculation formula for sole-proprietors, independent contractors and self-employed applicants, to base it on gross income, rather than net profit, which resulted in many micro enterprises being effectively excluded or receiving loans under $100
  • Setting aside $1 billion for businesses in the sole-proprietorship, independent contractor, and self-employed category that had no employees and were located in low-to-moderate income areas
  • Expanding PPP eligibility to small business owners with non-fraud felony convictions, unless the owner was incarcerated at the time of the application
  • Allowing business owners with federal student loan delinquencies to apply for a PPP loan
  • Ensuring access for non-citizen small business owners who were lawful U.S. residents by clarifying that they could use Individual Taxpayer Identification Numbers (ITINs) to apply for a PPP loan.

So, how did these PPP reforms work out for smaller and minority-owned businesses in New York City? Our analysis of the data found that PPP reforms did justice to the city’s micro-businesses, and firms in hard-hit sectors and majority-minority neighborhoods. However, this apparent success came at the expense of PPP’s efficiency since the program relied on a delivery mechanism in which middlemen could needlessly drive up taxpayers’ costs.

More PPP loans directed to NYC in 2021 than 2020

Table 1. Summary PPP Approved Lending, New York City

Through June 30, 2020, there were 244,850 PPP loans disbursed to the city’s small firms, totaling over $12.3 billion (see Table 1 above). That is on top of 160,407 loans approved during the program’s first round of funding from April 3, 2020 to August 6, 2020. The average PPP loan was $50,395 in the second round of funding. That is substantially lower than the first round, when the average loan was around $117,739. This suggests that in the second round, more loans may have gone to smaller businesses.

The number of loans less than $50,000 nearly doubled in the second iteration of the program (see Figure 1).

Chart with the amount of PPP Loans issued: 2020 vs 2021
Figure 1: PPP Loans by Amount

The reforms helped even the playing field for the city’s micro businesses who were excluded from the first iteration of PPP relief

Figure 2. Time Trends in Round 2 PPP Loan Approvals: New York City

While many of the city’s vulnerable small firms received a relief payment in the first round, many mom and pop shops and sole proprietors were left out — particularly micro-businesses led by women, minorities, and immigrants. This past oversight appears to have been rectified by the recent reforms including a two-week exclusive application period for smaller businesses, changing the PPP formula for sole proprietors, expanding PPP access to immigrant entrepreneurs and those with non-fraud felonies or student loan delinquencies. All these changes have been implemented since early March 2021. As illustrated in Figure 2, micro-firms’ loan approvals steadily climbed significantly above the rest since Mid-March 2021 when all the reform measures were put in place. The number of loans approved for micro firms increased by almost 250% from 72,134 in Round 1 to 184,871 in Round 2. Likewise, the number of loans awarded to non-employer firms — sole proprietors, self-employed individuals, and independent contractors — rose by approximately 486% from 25,488 in Round 1 to 123,761 in Round 2 (see Figure 3).

Figure 3: Number of Approved Loans by Types of Small Firms

Phase two effectively channeled loans to service sector firms that were severely affected by the pandemic

Figure 4: Time Trends in Business Revenue Loss by NAICS Super Sectors

Since March 2020, across all Industry Supersectors, the City’s small businesses experienced a large decline in revenue (see Figure 4). The leisure and hospitality super-sector suffered the heaviest losses, experiencing a revenue loss of 90% during the initial lockdown period and struggling through June 2021 with daily declines of more than 70% compared to January 2020. The Trade, Transportation and Utilities (TTU) super-sector is another sector that has been experiencing steep and persistent revenue losses. TTU lost more than 50% of its revenue when Covid-19 hit. During the second and third quarters of 2020, its revenue losses seemed to level off as the city came out of New York Pause. Yet it has returned to a downward trajectory since the fourth quarter of 2020. Compared to these two super-sectors, the Professional and Business Service sector experienced smaller but still large and persistent revenue losses of around 25%.

Figure 5: Funnel Chart: Distribution of PPP Loans by Industry Sectors

Examining PPP loans by industry⁴ reveals how effectively Round 2 of the PPP reached small businesses in hard-hit sectors such as Transportation, Retail Trade Services, Accommodation and Food Services, Other Services, and Art and Entertainment. As shown in Panel A of Figure 5, there are large concentrations of these loans among the following sectors: Utility/Transportation; Other Services; Professional, Scientific, and Technical (PST) Services; Accommodation and Food Services; and Retail Trade. Small firms in hard-hit sectors received 60% of the PPP loans that flowed to the city, while representing about 40% of the City’s total firms with fewer than 500 employees.

Panel B in Figure 5 shows how the total amount of Round 2 of PPP lending was distributed across the city’s business sectors. It is tricky to evaluate the industries’ share of PPP funds because shares could simply reflect the size of firms by sector. This is by design; the loan amounts are calculated to cover payroll expenses. Notably, the accommodation and food service sector, where firms were allowed to receive funding for 3.5 times their 2019 average monthly payroll expenses, drew $2.2 billion, which is about 18% of the total PPP dollars distributed in New York City. Close behind was PST with 12.6% of the pie (about $1.6 billion). Interestingly, the utilities and transportation sector claimed 9.1% of the total pool of money deployed in the city even though this sector received the most loans. This is because many firms in this sector are self-employed workers including gig workers, which translates into a larger number of smaller loans.⁵ More importantly, the small firms in hard-hit sectors collectively received 53.8% of the pie.

Shut out of Round 1 PPP, small firms in majority-minority zip codes got much needed help in Round 2

Our previous post-mortem study of Round 1 of PPP uncovered the uneven loan distribution that favored white and affluent neighborhoods in the city. This is not surprising given that PPP funds were distributed through banks whose top lenders were large commercial banks that have a long history of reserving their services for existing customers who tend to fit the aforementioned demographic. In round 2, the program was revised with an overt intention of reaching minority and underserved small businesses and helping them to defy the odds in a time of crisis. Nevertheless, Round 2 of PPP loans were once again pushed out through the banking system.

Did the second round of PPP relief stem racial inequity and reach minority owned businesses? This is a challenging question to answer given that about 78.3% of Round 2 of PPP loans have gone to borrowers who did not report their race or ethnicity. In the absence of sufficient data on borrowers’ race and ethnicity, the present analysis relies on the racial/ethnic make-up of modified zip code tabulation areas. Analyzing borrowers’ business location where racial or ethnic minorities make up a vast majority of the population offers suggestive evidence for racial equity in Round 2 PPP.

Figure 6. Geographic Distribution of Round 2 Loans by Modified Zip Code Tabulation Areas. Additional data on geographic distribution of loan amounts and number of jobs retained in Round 2 by Modified Zip Code Tabulation Areas:

Focusing on PPP loan counts by modified zip code tabulation areas, we find that Round 2 of PPP reached most neighborhoods in the city, delivering much-needed relief to firms in majority-minority neighborhoods (see Figure 6). Unlike Round 1, the top five zip codes with the highest number of Round 2 PPP loans included three majority-minority zip codes in the outer boroughs of NYC (11377, 11208, and 10456) along with two majority-white zip codes (10001 and 11211). Midtown Manhattan’s 10001 Zip code which includes Hudson Yards received 3,927 loans (about 1.6% of the city’s total loans) and nearly 22.6% of borrowers in this zip code received loans in excess of $150,000 or higher. Close behind is Woodside (zip code 11377), a Queens middle-income working-class neighborhood where South Asians and Hispanics make up nearly 80% of the population. While there were 3,408 Round 2 PPP loans (or 1.4% of the city’s total loans) issued in Woodside, only about 3.2% of the loans were above $150K. Similarly, 3,362 businesses in Brooklyn’s zip code 11208 (East New York) claimed 1.4% of the city’s total number of loans but only about 1% of the PPP loans were above $150K. Zip code 10456 (Claremont and Morrisania), a group of low-income majority-Hispanic neighborhoods in Bronx, received 3,259 loans. A common PPP lending pattern emerged in these three majority-minority zip codes in that two-thirds of each areas’ PPP loans were made to non-employer firms. In addition, the leading sector in these zip codes is the transportation industry. More specifically, half of the total loans in Woodside and a third of the total loans in East New York and Bronx zip code 10465 went to the taxi and limousine service industry, with the majority being for-hire car drivers. Lastly, there were 3,004 loans approved in Williamsburg’s 11211 Zip code (about 1.3% of the city’s total loans). Of those 3,004 loans, 46.2% went to the Leisure & Hospitality sector businesses in the area.

Figure 7. Access to PPP Loans by Communities of Color and PPP Rounds. Additional data on access to PPP by communities of color:

Figure 7 highlights how Round 2 of PPP reached firms in majority-minority zip codes that were underserved by Round 1 of PPP. As shown, nearly 59% of Round 2 PPP loans went to zip codes with populations where non-whites are a majority of the population, while only about 36% of Round 1 PPP loans flowed to these communities. While PPP credit access improved for all majority-minority zip codes, their share of PPP loan volume did not increase uniformly across all communities of color. Majority black and majority-Black/Hispanic zip codes in particular saw their share of loans increased two-fold in Round 2.

Utilization of PPP by industry varied across racial/ethnic makeup by zip code. Firms in majority-minority zip codes received 144,286 loans in Round 2. Within communities of color, the Utility and Transportation Services (UTS) were a large beneficiary of the program, scoring 37.3% of all loans deployed in majority-minority zip codes. About 67.5% of those loans in the utility and transportation services went to firms in the subsector of taxi and limousine services. The bulk of them were independent contractors. Meanwhile, Round 2 PPP loans were widely sought out by businesses across most industry categories in majority white zip codes. But the PST sector including Management Services and the UTS received 16.6 % and 15.9% of 100,564 loans distributed in majority-white zip codes.

Non-mainstream financial Institutions partnering with fintech companies greatly expanded access to PPP relief for small businesses in underserved communities

Figure 9. Lenders by the percentage of Round 2 PPP Loans to women and communities of color

While the first round of PPP funding was mostly distributed by banks⁶, which put underbanked or unbanked borrowers at a disadvantage, in Round 2 smaller and non-bank lenders partnering with fintech companies outperformed commercial banks in terms of addressing the needs of firms in majority-minority zip codes (See Figure 9). For example, Cross River Bank — a small community bank headquartered in New Jersey partnering with Divvy and other fintech companies such as Lendio, BlueVine, Gusto, Intuit, and Veem — was the top lender in the city by loan count (25,131), and about 63.9% of their borrowers were firms in communities of color. JPMorgan Chase & Co ranked 2nd by originating 25,100 Round 2 loans for the city’s small businesses, but they were less likely to serve firms in communities of color. Harvest Small Business Finance, LLC, a non-bank SBA 7(a) lender partnering with Womply Fast Lane PPP, ranked 3rd in providing loans to a total of 22,247 small firms, more than three-quarters of which were businesses in majority-minority zip codes. Customers Bank, a lender partnering with a UK-based fintech company “OakNorth,” handed out 16,783 loans, approximately 64.2% of which went to businesses in majority-minority communities. Bentworth Capital. LLC, another authorized non-bank lender partnering with Womply Fast Lane PPP, provided 14,886 loans, roughly 80% of which went to businesses in majority-minority zip codes.

The table below provides a high-level summary of the PPP loans originated by the top 15 lenders in the city, highlighting how these top 15 lenders served the smallest firms with less than 3 employees and the hard-hit industry. Follow the link below the table for an interactive chart. Readers can sort the table by clicking on the column header, or search for a specific lender using the search bar.

Table 2. Summary of Round 2 PPP loans originated by the top 15 lenders in New York City. Url for the interactive table:

Meanwhile, the apparent success came at the cost of taxpayer’s money, putting an extra 334.8 million dollars in lenders’ pocket

One of the key lessons learned from Round 1 of PPP was that the fee structure incentivized lending toward large loans. The government enhanced the fees lenders could charge for approving PPP loans of $50,000 or less in order to create incentives to make small loans. In December 2020, an important change to the fee structure of the PPP program was authorized. Fees for loans up to $50K changed from a flat 5% of the loan amount to an amount equal to the lesser of 50% of the principal amount or $2,500. The fees for loans of more than $50,000 remained the same for Round 2. That is, for loans between $50,000 and $350,000, banks earned a 5% fee, and for loans of $350,000 and above, they earned 3%.

Of the 244,850 Round 2 PPP loans distributed in the city, 85% were for $50,000 or less, totaling $3.25 billion. Lenders originating these small loans collectively claimed about $497 million in fees (see figure 10). That is a 15% profit margin, which seems high relative to what lenders are obligated to do for loan origination or forgiveness.⁷

In Round 2, the average bank fees for small loans were $2,365 while the average loan size was $15,454. Multiplied across the $3.25 billion Round 2 loans of $50,000 or less, taxpayers had to pay an extra $334.8 million that they would not have paid in Round 1 for the same loan amounts and that could potentially have been used to stabilize thousands more small firms.

Figure 10. Estimated Bank Fees Earned on PPP Loans

We urgently need to invest in government’s digital financial infrastructure to channel sharply targeted relief to those in need at the lowest transaction costs

Unlike other countries such as the United Kingdom and the Netherlands, the United States government lacks the requisite social insurance and digital infrastructures to be able to precisely capture how much workers lost (or would lose) in wages due to the lockdown and rapidly deploy adequate relief to affected firms and their employees. Understandably, PPP was designed to leverage banking infrastructure, prioritizing speed over technical and allocative efficiency.

As highlighted above, our post-mortem study of Round 2 PPP loans across the city’s industries and neighborhoods found that the reform measures successfully channeled the second wave of PPP loans to micro-businesses, firms in majority-minority communities and the hard-hit industries that were left behind in the first iteration of the program. Yet this success came at the costs of program efficiency, putting an extra $334.8 million into lenders’ pockets that could have been channeled to additional needy recipients. The extra program cost spent on bank fees, $334.8 million, is an excessive amount for the relatively minimal role the banks played in processing PPP loans: taking e-forms of applications and supporting materials; confirming the borrower’s math and certification; submitting the information through SBA’s PPP portal⁸; and ultimately cutting a check.

In the midst of an unfolding economic crisis, speed is the key to delivery success. But now is the time to shift our attention to the longer-term solution that could improve efficiency in our government disbursement scheme. When economic relief flows through middlemen, government and taxpayers inevitably need to play by their rules. Another problem with a middleman-mediated system is that one cannot use it unless the middlemen grant access. Luckily, FinTech lenders dramatically and rapidly expanded PPP access in Round 2. But a recent study found that FinTech’s track record with potentially misrepresented PPP loans was far from perfect.

We could cut out the middlemen entirely. The SBA has already done that with the Economic Injury Disaster Loan (EIDL) program. Currently, SBA has digitized the EIDL services by dedicating their pre-existing Common IT platforms (i.e., SBA Connect) to serve the EIDL, including identity management, application process, and digital payment systems. In order for governments to provide direct and rapid relief, targeted capacity investment is required including the adoption of digital technology for government disbursement schemes.

Going forward, our goals should be delivering support rapidly and equitably, but also efficiently. The more money that is saved on fees — which rose to excessive levels in the PPP program — the more can be given to additional recipients.

¹ Under the new program rules, loans are available both as “second-draw” loans (maximum $2 million) for borrowers who received funding in the first round, and as “first-draw” loans (maximum $10 million) to first-time borrowers. Each qualifying small business will receive PPP funding based on the submitted 2019 average monthly payroll of the business applicant, multiplied by a factor of two-point-five (2.5). As such, funding for two and a half months of the small business’s payroll expenses will be provided for by the second draw loans. In addition, the new coronavirus stimulus bill includes a special calculation for businesses in the food industry and hospital industry, wherein a factor of three-point-five (3.5) is set to give restaurants and other food businesses a larger loan amount.

² Similar to the first round, 60% of PPP2 funds should be used for payroll and 40% can be used for non-payroll expenses in order to qualify for loan forgiveness. The new list of permissible costs and expenses includes:

  • Business software or cloud computing software expenses that facilitate business operations (e.g., tracking of payroll, sales or billing functions, or inventory).
  • Costs related to property damage resulting from vandalism in 2020 that were not covered by insurance.
  • Certain supplier costs that are (a) essential to the borrower’s business and (b) were incurred pursuant to agreements or purchase orders in effect prior to the borrower’s applicable covered period.
  • Capital expenditures incurred to facilitate compliance with safety standards issued by HHS, CDC, or OSHA (e.g., new ventilation systems, outdoor space expansions, and purchases of PPE).

³ The American Rescue Plan Act (ARP) of 2021, signed into law in March 2021, added $7.25 billion for a total authorization of $960.3 billion.

⁴ Using two digit NAICS codes

⁵ Of 69,832 loans that were disbursed to the utilities and transportation sector, approximately 66.1% went to firms in the taxi service industry (i.e., NAICS code 485310), the overwhelming majority of which (more than 99.6% borrowers) are self-employed workers.

⁶ In Round 1, the four large commercial banks in the city churned out 42% of area total PPP loans.

⁷The lender’s obligations do not appear to be particularly burdensome for either loan origination or forgiveness. Lenders are required to confirm 1) receipt of the borrower’s certifications contained in the Loan Forgiveness Application Form; 2) receipt of supporting documentation that supports payroll and nonpayroll costs; and 3) borrower’s calculations of specific line items in support of the request for PPP loan forgiveness — including the determination of any limitation on forgiveness of nonpayroll costs, which is capped at 25 percent. However, the Lender Responsibilities Rule ( clarifies “the lender does not need to independently verify the borrower’s reported information if the borrower submits documentation supporting its request for loan forgiveness and attests that it accurately verified the payments for eligible costs”. This means that a lender has almost no liability if the borrower’s certifications or documentation were fraudulent.

⁸Interestingly, the SBA PPP portal is designed to take lender’s PPP loan request on behalf of borrowers as well as their loan forgiveness decision to the SBA while borrowers are allowed to submit only the loan forgiveness request to their lender through the portal.


This series will continue to explore the U.S. Census Bureau’s Household Pulse data. In the current environment of persistently high unemployment, it is important to monitor timely, detailed data. The Pulse data is one important tool that can help us to assess COVID-19’s impact as the pandemic evolves.

Previous blog posts in this series have examined the pandemic’s initial labor market impact, unequal burdens of child and dependent care across gender, labor market recovery, inequity in the Paycheck Protection Program and the housing crisis. For more information on the Pulse data, including future releases, and NYC Opportunity’s use of the data see the NYC Opportunity website.

This publication of the COVID data Analysis was written by Jihyun Shin (, Deputy Director of the NYC Opportunity’s Poverty Research Team.


The Poverty Research Team at NYC Opportunity is responsible for the development of the NYCgov Poverty Measure. The alternative NYC Poverty measure, in comparison to the official U.S. measure of Poverty.
The work has received nationwide attention and contributed to the development of the Federal Supplemental Poverty Measure.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store