A series of recent research papers, including a major study by Prof. Drew Winters of Texas Tech University and his coauthors, explores how these decisions affect minority borrowers and entrepreneurs. Taken together, the papers paint a layered picture of modern lending. They show that the structure of the banking system matters deeply for minority owned businesses. They also reveal that the statistical models researchers rely on often explain surprisingly little of the ultimate lending decision, leaving room for human judgment, institutional culture, and possibly bias.

To understand the stakes, it helps to begin with the role of credit in economic life. For households, a mortgage is often the gateway to homeownership and long term wealth building. For entrepreneurs, a business loan can determine whether an idea remains a dream or becomes a source of jobs and income. In theory, lending decisions are based on risk and return. In practice, they are embedded in institutions that differ in size, incentives, and information.

One of the studies conducted by Winters and colleagues examines mortgage lending using data from the Home Mortgage Disclosure Act, commonly known as the HMDA. HMDA data provide detailed information about mortgage applications, including the race and ethnicity of applicants, income, loan amount, and whether the loan was approved or denied. For decades, researchers have used this data to investigate potential discrimination in mortgage markets.

The paper finds a striking pattern. Even after controlling for many observable characteristics, including income and loan size, statistical models explain only a small portion of the variation in loan approval decisions. In other words, the measurable factors in the data account for surprisingly little of why one applicant is approved and another is denied. The residual differences across racial groups remain substantial.

This does not automatically prove intentional discrimination in every case. HMDA data do not include all relevant risk variables, such as detailed credit scores. Yet the limited explanatory power of the models raises important questions. If observable financial metrics explain only a modest share of outcomes, then unobserved factors, institutional practices, or subjective judgments may be playing a significant role. The mortgage market, often described as highly standardized and data driven, turns out to contain more discretion and opacity than many assume.

This insight provides a broader backdrop for understanding small business lending. If even mortgage lending, with its relatively uniform products and extensive data reporting, cannot be fully explained by observable risk factors, then business lending, which is more complex and heterogeneous, may involve even greater room for discretion.

Another study turns directly to small business credit and compares lending outcomes for minority owned firms at community banks and at large national banks. Community banks are typically smaller institutions that focus on a limited geographic area. Their loan officers often know local business owners personally and gather what economists call soft information through repeated interactions. Large banks operate across states or the entire country and rely more heavily on standardized underwriting models and centralized decision making.

The study finds that minority owned businesses tend to have better access to loans at community banks than at large banks. After controlling for firm characteristics and local conditions, minority entrepreneurs are more likely to receive credit from community banks. The results suggest that relationship lending and local knowledge may help reduce barriers that minority business owners face in more impersonal, model driven environments.

These findings set the stage for a third study from Prof. Winters and his coauthors. If community banks are relatively more supportive of minority entrepreneurs, what happens when those banks disappear through mergers and acquisitions? Over the past several decades, the United States has experienced significant consolidation in the banking sector. Thousands of community banks have merged into larger institutions. While consolidation can improve efficiency and diversify risk, it also changes how lending decisions are made.

The main study assembles detailed data on the process for implementing section 1071 of the Dodd/Frank Act. Section 1071 is intended to create a small business loan database for examining compliance with Fair Lending laws, as HMDA did for mortgage loans. Interestingly, section 1071 suffers from the same data problems as HMDA in that it does not require the collection of credit variables. The study also provides a summary of the CFPB process for developing section 1071 and comments from the bankers during the process.

The purpose of this analysis is to determine if the proposed implementation of section 1071 will provide for reasonable testing of Fair Lending compliance in small business loans. The test results using only the variables to be collected under section 1071 show that minority-owned small businesses have less access to credit. However, this result is not correct because Winters and colleagues used the same data as the second study we have just described, that finds that minority-owned businesses get better access to credit at community banks than at large banks.

The researchers explored several mechanisms that could explain this pattern. One possibility is the shift from relationship based lending to more standardized credit scoring. Community banks often rely on qualitative information about a borrower’s character, local reputation, and business prospects. These factors are difficult to quantify but can be crucial for small firms with limited financial histories. Large banks, by contrast, may prioritize uniform metrics that can be applied across thousands of branches. If minority entrepreneurs are more likely to have shorter credit histories or less collateral, they may be disproportionately affected by this shift.

Another mechanism involves organizational incentives and authority. In a small bank, a local loan officer may have significant discretion to approve a loan based on firsthand knowledge. In a large bank, decisions may require multiple layers of approval and adherence to centralized guidelines. The change in structure can dampen the influence of local information and personal relationships.

The researchers are careful to address alternative explanations. They control for local economic conditions, firm characteristics, and bank level variables. They compare changes over time in affected and unaffected areas. The results remain robust across multiple specifications. The evidence suggests that the decline in minority lending is not simply due to worsening local economies or changes in borrower demand. It is closely linked to the shift in bank ownership and structure.

The consequences extend beyond individual loan approvals. Small businesses are key drivers of job creation and neighborhood vitality. When access to credit tightens for minority entrepreneurs, the effects can ripple outward. The paper presents evidence consistent with reduced small business activity in affected communities. Fewer loans can mean fewer expansions, fewer hires, and slower growth.

When viewed alongside the mortgage discrimination study, a broader theme emerges. In both mortgage and business lending, formal models explain only part of what is happening. Institutional context and organizational design matter. In the mortgage study, even detailed HMDA data leave much of the approval decision unexplained. In the small business context, the type of bank and its internal structure shape outcomes for minority borrowers.

These findings complicate simple narratives about technology and objectivity in finance. It is tempting to believe that as lending becomes more data driven, disparities will shrink. Yet the evidence suggests that models are only as powerful as the variables they include and the institutions that deploy them. If key aspects of borrower quality are difficult to measure, then discretion remains. That discretion can either help or hinder underserved groups, depending on the setting.

Community banks, according to the research, may use discretion in ways that benefit minority entrepreneurs by incorporating local knowledge. Large banks, in contrast, may use standardized processes that unintentionally disadvantage borrowers who do not fit typical profiles. Neither model is inherently virtuous or flawed. Each has tradeoffs. Community banks may face higher costs and greater exposure to local shocks. Large banks may achieve efficiencies and stability through scale. The question is how these differences affect access to opportunity.

Prof. Drew Winters has emphasized through this work that diversity in banking institutions can play an important role in promoting inclusive growth. A financial system composed only of large, centralized banks may not serve all communities equally well. At the same time, a system composed solely of small banks might lack resilience and innovation. The balance between consolidation and local autonomy has real consequences.

For policymakers, the research carries important implications. Bank mergers are reviewed by regulators who consider competition and safety and soundness. The findings suggest that regulators may also need to weigh distributional effects on minority borrowers and entrepreneurs. Aggregate lending totals can mask shifts within subgroups. A merger might leave overall loan volume unchanged while reducing credit to minority owned businesses.

The mortgage study also points to the value of transparency and data. HMDA reporting has enabled decades of research into racial disparities in home lending. Yet even with this data, much of the decision process remains opaque. Expanding data collection and improving model transparency could help researchers and regulators better understand where disparities arise and how to address them.

For the general public, these studies highlight that financial systems are not neutral backdrops. They are designed and governed by institutions with particular structures and incentives. A decision made in a corporate boardroom about a bank acquisition can influence whether a local entrepreneur secures funding. A tweak to a credit scoring model can affect who buys a home.

At a time when discussions about racial wealth gaps and economic mobility are prominent, evidence-based research is essential. The work of Prof. Drew Winters and his collaborators does not claim that every disparity is the result of overt discrimination. Nor does it argue that consolidation is always harmful. Instead, it demonstrates that institutional details matter. The size of a bank, the information it collects, and the authority it grants to local officers all shape economic outcomes.

Ultimately, the story is about access and opportunity. A minority entrepreneur seeking a loan is not just a data point. They are someone trying to build a livelihood, support a family, and contribute to their community. A prospective homeowner navigating the mortgage process is not just an application number. They are making one of the most significant financial decisions of their life.

By bringing rigorous data analysis to these questions, the research invites a more informed conversation about how to design a financial system that is both efficient and fair. It reminds us that behind every approval and denial lies a set of institutional choices. As the banking landscape continues to evolve, understanding those choices will be crucial for ensuring that opportunity is broadly shared.