Mission-oriented banks in Texas and underserved businesses: lessons from the Paycheck Protection Program

Emergency Capital Investment Program Picks Up Where PPP Left Off

During the final weeks of 2021, the largest government program ever targeted to community financial institutions announced deployment of nearly $9 billion across the country. Named the Emergency Capital Investment Program (ECIP), this unprecedented amount of money directed explicitly to Community Development Financial Institutions (CDFIs) and Minority Depository Institutions (MDIs) hopes to expand on where the Paycheck Protection Program (PPP) left off—by improving access to capital for under-resourced communities and small businesses.

The ECIP is one of several programs, including the CDFI Rapid Response Program and the State Small Business Credit Initiative, aimed at mitigating lending disparities and spurring small business health in a post-pandemic world. These are worthy goals, given that most small employers report ongoing negative impacts of COVID-19—impacts that are particularly pronounced for very small businesses and minority-owned firms.

Although it is too soon to have outcome data on the ECIP, we can still explore how useful expanding MDI and CDFI lending might be to underserved small firms by using data on loan access and forgiveness from its predecessor, the PPP.

In doing so, we can ask the following: What effect did these mission-oriented banks have in helping the most vulnerable small businesses gain access to the PPP? Were the policy changes that aimed to foster more-inclusive lending efficacious? And what role did mission-oriented banks play in PPP loan forgiveness? This brief explores these questions in hopes of gleaning lessons learned from this quickly implemented $800 billion program, which can be helpful for small business policy and/or future economic crises.

MDIs and CDFIs

MDIs and CDFIs are mission-oriented banking institutions. Together we refer to these entities as Community Financial Institutions (CFIs). MDIs are federally insured banks or credit unions where at least 51 percent of the voting stock is owned by persons of color, or where a majority of the board of directors is of color, and the institution primarily serves a predominantly minority community. A recent Dallas Fed study on MDIs shows that across the country, MDIs hold combined assets of $320 billion. There are 24 MDIs located in Texas, including nine Asian MDIs, 13 Hispanic, one Black and one Native American MDI.

CDFIs explicitly lend to low- and moderate-income (LMI) communities and historically underserved consumers. Nationwide, they manage more than $222 billion aimed to improve access to capital for low-income borrowers, borrowers of color and women. There are 45 CDFIs located in Texas.[1]

PPP Timing and the Increasing Role of CFIs

Anecdotally, we know that MDIs and CDFIs played an important role in allowing businesses of color, microfirms and those operating in LMI communities access to the PPP.[2] After the initial concerns about reaching vulnerable firms in the early days of PPP, regulations shifted to foster greater involvement of community lenders more explicitly.

An overview of key PPP changes is below:

As shown above, the initial $349 billion of funding ran out in 13 days, and reports indicated a large disparity in firms accessing funding. For example, the National Small Business Association found that while a majority of applicant firms with more than 20 employees were approved, just 18 percent of those with 10 or fewer employees were. When the PPP was reauthorized on April 27, the additional $310 billion came with a few changes to remedy that imbalance. Critical among these changes was the $30 billion set aside for CFIs.

This early 2020 change was not the only one to use MDIs and CDFIs to improve the distribution and reach of the program. After the Consolidated Appropriations Act was signed into law in December 2020 the PPP was reauthorized with an additional $284 billion—open initially only to CFIs. Finally, on May 4, 2021, the PPP was closed to all lenders except CFIs.

So did these changes result in a corresponding jump in PPP loans made by MDIs and CDFIs? The answer seems to be yes.

The total number of loans made by CFIs ranged from 0 in mid-July 2020 to a high of more than 10,600 in May 2021. Similar patterns exist for the share of loans made by CFIs: Rates approached 100 percent in May 2021 but loans made by CFIs accounted for 0 percent of total PPP loans in the early months of 2020.

The influence of CFIs does appear to increase as time goes on, suggesting that the more-inclusive regulatory changes did in fact have some impact. As Chart 1 illustrates, about 80 percent of loans through MDIs or CDFIs were made in 2021, compared with just over 50 percent for non-CFIs.

MDIs, CDFIs and Vulnerable Small Businesses

We are also interested in if the greater inclusion of community institutions, as was hypothesized, really led to better access for underserved small firms.[3]

For the purposes of this analysis, we define vulnerable small businesses by owner demographic, geographic location and firm size:

• Firms located in LMI neighborhoods.
• Microfirms (five or fewer employees including the business owner).
• Firms owned by people of color.
• Firms owned by women.
• Firms owned by veterans.

These firms were more likely to be in poorer financial condition prior to the pandemic and more likely to struggle accessing capital.[4]

Microfirms

Most PPP loans went to microfirms (firms with five or fewer employees including the business owner), and this share significantly increased as time progressed. Overall, about 75 percent of PPP loans went to microfirms. However, this figure is disproportionately small compared with microfirms’ overall share of the total small business population, which is closer to 94 percent.[5] Looking at the role that MDIs and CDFIs played, microfirms were about twice as likely to receive their loan through an MDI or CDFI than larger firms (12 percent versus 6 percent).

LMI Neighborhoods

As Chart 2 suggests, the share of loans made in LMI neighborhoods is somewhat positively correlated with the date the loan was approved.[6]

The dates during which more than 50 percent of the total loans approved were in LMI communities occurred during those windows open exclusively to community lenders. There is in fact a statistically significant relationship between a recipient’s location in an LMI neighborhood and the likelihood that their PPP loan servicer is a CFI. While 28.7 percent of all PPP loans went to LMI neighborhoods in Texas, 36.4 percent of MDI and CDFI loans did.

Firm-Owner Demographics

To properly address the rest of the vulnerable firm question, accurate data on firm-owner demographics would be ideal. However, as mentioned in a previous article, self-reported firm-owner demographic data are sparse.

Without consistent self-reported demographics in the SBA data, analysis of women-owned and veteran-owned firms is unreliable. Regarding minority-owned firms, however, we can use the share of people of color in a ZIP code as a proxy for share of firms of color. We can therefore explore if MDIs and CDFIs were indeed more likely to lend in majority-minority communities.

A strong pattern emerges for ZIP codes in which residents of color are the majority (Chart 3).

Here, the association between the date and the share of loans made in majority-minority communities is strongly positive.

Looking at individual loan-level data, then, it is not surprising that CFIs appear more likely to lend in neighborhoods with high shares of people of color. For instance, while 55.6 percent of all PPP loans were approved in ZIP codes in which the majority were residents of color, 76.3 percent of CFI-serviced PPP loans were in neighborhoods of color. These associations are statistically significant at the 99 percent confidence level.

Overall, it does appear to be the case that MDIs and CDFIs were critical in improving access to the most vulnerable small firms (Table 1).

Table 1: CFIs Directed Larger Shares of PPP Loans to Underserved Small Business
Share of loans going to:	CFI share (%) (N=99,069)	Non-CFI share (%) (N=840,664)
Microfirms	76.3	53.4
LMI neighborhoods	36.4	27.8
Majority-minority ZIPs	85.9	73.8
NOTE: All reported differences are statistically significant at the 99 percent confidence level. SOURCES: Small Business Administration PPP data; author’s calculations.

MDIs, CDFIs and Hard-Hit Industries

Understanding the role of CFIs in reaching vulnerable small firms, we can now look at the question of industry concentration.

According to analysis from the Texas Comptroller’s office, the two sectors experiencing the greatest employment losses in 2020 were 1) arts, entertainment and recreation, and 2) accommodation and food services. Both of these are considered part of the leisure and hospitality “supersector” and both experienced employment drops far larger than the Texas average. This is not surprising since they rely heavily on social interactions, which became less prevalent in the COVID era. In fact, by November 2020, leisure and hospitality accounted for more than 38 percent of total lost nonfarm jobs year over year in Texas.

Looking at the distribution of PPP by industry, about 18 percent of PPP loans went to the retail sector, totaling more than $7 billion (Table 2). Professional and business services small firms represented about 17 percent of total recipients and received more than $10.7 billion in the aggregate.

Table 2: Top Texas Sectors by PPP Loan Approval
	Share of total PPP loans in Texas (%)	Total amount approved (billions of dollars)	Average per PPP loan (dollars)	Share of total lost jobs in Texas (%)
Retail	18	7.01	41,553	3
Professional and business services	17	10.76	67,621	0
Education and health services	10	8.05	85,105	17
Leisure and hospitality	9	7.46	86,506	38
Construction	8	7.64	96,858	7

Compared with its share of overall job losses in 2020, the leisure and hospitality sector appears underrepresented in the share of total PPP loans received. Looking at the share of reported jobs also suggests underrepresentation: leisure and hospitality accounts for less than 18 percent of all jobs impacted by the PPP.

Here, we once again find that CFIs played an outsized role in assisting firms in the hardest-hit industry to secure a PPP loan. While representing 9 percent of total loans, leisure and hospitality represented 11.5 percent of all loans made through CFIs.[7] The majority appear to be driven by MDIs, with 16 percent of their loans going to this industry in Texas. Digging in further, while MDIs of all categories disproportionately lent to the leisure and hospitality sector, Asian banks stand out (Table 3).

Table 3. Leisure and Hospitality Represented Nearly a Quarter of Texas PPP Lending at Asian MDIs
	Total PPP loans in Texas	Share of loans made to leisure and hospitality (%)
Asian MDIs	15,763	22***
Black MDIs	773	12**
Hispanic MDIs	23,258	11***
Native American MDIs	161	10
** Denotes statistical significance at the 95 percent confidence level. *** Denotes statistical significance at the 99 percent confidence level. SOURCES: Small Business Administration PPP data; author’s calculations.

Did MDIs and CDFIs help the PPP reach the industries most impacted by the economic crisis? At least in the case of MDIs, the answer appears to be yes. While CDFIs were more likely to lend to the arts and entertainment sector, they were significantly less likely to lend to restaurants and hotels.

Putting It All Together

Now that we have an idea of the relationships between CFIs and various business demographics, it would be good to understand if those relationships hold while controlling for a variety of other factors.[8]

Overall, it appears that when controlling for other factors, a PPP loan was more likely to be originated through a CFI if the borrower was in the leisure and hospitality industry, the loan size was less than $150,000, or the firm was located in ZIP codes with higher shares of residents of color. While MDIs were less likely to originate loans to microfirms and start-ups (defined as businesses open two years or less), CDFIs were more likely to originate these loans. CDFIs were also more likely to make loans later on in the PPP timeline. MDIs, however, were more likely to lend to firms in rural neighborhoods, Contrary to our earlier finding, CFIs now appear less likely than non-CFIs to reach LMI neighborhoods. This change is driven by the inclusion of the variable representing the share of residents of color within a given ZIP code.

MDIs, CDFIs and PPP Forgiveness

Finally, we took a look at the role MDIs and CDFIs played in PPP loan forgiveness. As originally conceived, recipients of PPP funds would see those loans fully forgiven, provided they met certain requirements. So far, we have discussed how CFIs were more inclusive in extending PPP loans to vulnerable firms. But were they also more likely to forgive these loans? We restrict our observations to loans approved in 2020 because of data limitations for later dates of forgiveness.

Overall, the vast majority of PPP applicants have had their 2020 loans forgiven. Of the 409,889 loans approved in 2020, 87 percent were fully forgiven (Table 4).

Table 4. Most PPP Loans Approved in 2020 Fully Forgiven
	Forgiveness Share Category
	100% forgiven	50% to 99.99% forgiven	Less than 50% forgiven
Number of loans	356,908	28,233	24,748
Median loan amount ($)	23,193	32,300	17,500
Median unforgiven amount ($)	0	4,673	16,600

We are interested in what factors increase the likelihood that a firm’s PPP loan is less than 50 percent forgiven, as well as 0 percent forgiven. This could have negative implications for the business’ financial condition moving forward, particularly if that firm expected the loan to be forgiven.

We estimated the impact of variables on the likelihood that a minority share of a loan is forgiven and on the likelihood that the loan is not forgiven at all. Overall, it appears that microfirms, start-ups, and those in the leisure and hospitality sector have a higher likelihood that only a small portion of their loan will be forgiven. Firms that originally indicated a higher portion of the loan would be devoted to payroll are also far more likely to struggle with forgiveness approval. This may be due to misaligned expectations about the amount of dollars needed to support payroll, or the firm may have had challenges retaining its workers despite the PPP.[9]

Location in an LMI neighborhood is not significantly related to the likelihood that a loan will be forgiven. However, the share of residents of color in a business’ ZIP, our proxy for firms of color, does increase the likelihood that a business owner will find full forgiveness more challenging. Importantly, holding all else constant, origination of a PPP loan at an MDI or CDFI strongly reduces the likelihood that only a small portion, or no portion, of the loan will be forgiven. Finally, as expected, loans under $150,000 are significantly more likely to be forgiven.[10]

Policy Implications

Based on this analysis, it appears that MDIs and CDFIs played a significant role in not just helping underserved firms access the PPP program in Texas but also in having that loan be forgiven.

A few important implications surface: Initiatives like the ECIP that boost the capacity of MDIs and CDFIs in Texas may indeed help mitigate some of the disparities in small business financial health. Keeping PPP outcomes in mind, we find the ECIP promising in its directing of dollars specifically to mission-oriented banks. But the program isn’t perfect. The demand outweighs supply, and regulations were a disadvantage for those smaller CFIs that are structured as an S-Corporation, restricting their access to funds. Furthermore, long-term struggles with undercapitalization will likely take more than ECIP to overcome.

Beyond the capital infusions needed, technical assistance and technology improvements may also help MDIs and CDFIs expand their reach. The private sector can also play a role to partner with CFIs, provide liquidity, and support operational efficiencies through technology and technical assistance. In fact, non-CFI banks that invest in CFIs may receive credit for their Community Reinvestment Act assessments.[11] Another suggestion from the Milken Institute is co-location of CDFI and MDI services within other local services including community colleges and department stores. Large banks can also use existing ATM infrastructure to provide no-fee services for MDI or CDFI customers.

The State Small Business Credit Initiative (SSBCI), reauthorized with the passage of the American Rescue Plan, offers a total of $10 billion across the country to support small businesses emerging from the pandemic. Incentivizing greater participation of MDIs and CDFIs in this program could greatly improve the impact, particularly because the SSBCI contains a set-aside for businesses owned by socially and economically disadvantaged people and very small firms, which are more likely to be reached by CFIs.

Finally, the lessons learned from the early implementation of the PPP should not be forgotten in case of a future financial crisis. Earlier involvement of CFIs, through promotion of opportunities for MDI and CDFI lenders and set-asides for them, could help mitigate the disparities in accessing emergency aid. It could also help improve understanding of the terms of forgiveness.

Closing the gaps in access to funding and technical assistance and the broadening of networks for vulnerable firms will require more than improving capacity and technical assistance for MDIs and CDFIs. But as this article suggests, it could be a step in the right direction.

Notes

1. Because a small business owner could submit an application for PPP through any approved financial institution—including beyond state lines—we include MDIs and CDFIs not just in Texas but across the country.
2. See “CDFIs and MDIs Continue to Cater to Small Businesses Despite PPP Barriers,” the National Community Reinvestment Coalition, July 31, 2020.
3. We define underserved firms as those owned by people of color, small firms (under five employees) and those located in low- and moderate-income areas. We recognize that women-owned and veteran-owned firms are also vulnerable, but the paucity of data in these cases precludes deeper analysis.
4. See “2016 Small Business Credit Survey: Report on Women-Owned Firms,” Fed Small Business, 2016, and “2019 Small Business Credit Survey: Report on Minority-Owned Firms,” Fed Small Business, December 2019.
5. Author’s calculations using data from the U.S. Small Business Administration Office of Advocacy “Frequently Asked Questions,” October 2020.
6. The definition of LMI community is a census tract in which the majority of residents earn less than 80 percent of the area median income. For more information, see “Defining ‘Low- and Moderate-Income’ and ‘Assessment Area,’ ” Federal Reserve Bank of Minneapolis, March 8, 2018.
7. This difference in lenders is statistically significant at the 99 percent confidence level.
8. See appendix.
9. See appendix.
10. For loans under $150,000, the SBA offered a streamlined forgiveness application to expedite forgiveness processing for these groups.
11. As part of its Advance Notice of Proposed Rulemaking, the Federal Reserve Board of Governors is considering further incentivizing investments in CDFIs and MDIs through changes such as the elimination of geographic restriction on investments.