Investigating the function and consequence of credit ratings
How do we as a society decide what is valuable or worthwhile? Economic sociologist Davon Norris, Ph.D., is working to understand how society’s tools for determining what is of value and worthwhile are steeped in patterns of inequality, especially racial inequality. Davon focuses on how this inequality empirically manifests in his work that studies credit, debt, and finance – particularly when it comes to government credit ratings and consumer credit scores.
Davon joined the University of Michigan from The Ohio State University (OSU), where he received his Ph.D. in Sociology in 2022. As a LSA Collegiate Fellow in Organization Studies, Davon will become an Assistant Professor of Organizational Studies in 2024.
From Accounting to Sociology
Davon nurtured a lifelong educational interest in math and accounting. After his uncle entered college as an accounting major, Davon solidified his high school interest in accounting and finance.
It wasn’t until he had started working in finance as a valuation analyst after obtaining his bachelor’s degree in accounting from OSU that he started to realize finance may not be the right career choice for him.
“The actual work was good but I began to question how we valued things.” He saw how financial models often animated patterns of inequality and wanted a different career path that allowed him to dive deeper into those issues. He noted, “I was exposed to academic research during my undergraduate work and I started to think a more academic-oriented space is how I wanted to go,” he said.
In deciding what type of program fit his interests, he “wondered how economists talked about finance and financial markets – and how public policy folks were talking about it?” When Davon came across a sociology article on accounting and the development of capitalism, he realized sociologists talked about the issues he cared about in a way that piqued his interest. “I came to sociology in part because sociologists were talking in a way that more strongly resonated than other disciplines…”
Moving from accounting and finance to sociology might be a strange path. However, Davon says, “For me it made perfect sense as far as career transitions go. I went from being a valuation analyst at Ernst and Young to studying valuation as a process.”
Political “double-standard” in credit ratings
As he moved into sociology, Davon observed how models and formal criteria were often thought of as a way to reduce the influence of biases, but in his work in finance he saw how the creation of models could distort such a view. “Rating criteria and formal standards about the relevant factors for an evaluation hold promise for creating transparency, eliminating biases, and generating meritocratic evaluations through standardization and uncertainty reduction,” Davon said. “Yet little is known about whether the criteria used in fact eliminate biases or introduce new complexities.”
This drove Davon to leverage his accounting background and build an original dataset that approximated the evaluative criteria used in rating the credit worthiness of city governments. Davon tested whether a rating agency’s scoring criteria were applied evenly across cities. He found evidence of a political double standard where liberal-leaning cities were evaluated according to a more strict application of rating criteria requiring these cities to have higher performance for similar ratings compared to their conservative counterparts.
“When it came to really important criteria like debt per capita and medium family income, what I was seeing is that liberal cities with lower income had lower scores relative to liberal cities with higher income. The results were pretty much what we expected.” But there was no similar correlation for conservative cities. “Regardless of whether conservative-leaning cities had lower income or higher income, they had similar and high ratings.”
Davon said, in some ways, the results made sense. “If the job of a credit rating agency is to determine whether a city government is going to repay its bonds and debts, then just a straightforward application of criteria might not make sense. For example, Detroit might issue a bond to invest in infrastructure, but maybe they won’t be as successful at paying their debts, or Detroit might face pressure from citizens to take on more debt. From the agency’s perspective, they might say somewhere like Mesa, Arizona might not succumb to those pressures and will be able to minimize pressures like these as much as possible. Even if they have high debt at the moment, they have historically done a good job paying off debt, so they will have a higher credit rating.”
Those kinds of ideas about which kinds of governments are going to behave in which ways “get implicated in the ratings itself and how they get created.” Davon explains how such cultural ideas of what it means to be a liberal-leaning or conservative city have important implications for how those cities are evaluated and ultimately how those cities access financing to build infrastructure, invest in public safety, or provide key social supports. As a result, Davon argued, “These findings highlight a mechanism producing inequality between cities and demonstrate how generating meritocratic and unbiased evaluations requires more than the existence and transparency of rating criteria.”
An image of the front and back of a 1927 Montgomery Ward credit reference letter.
What does a government’s credit rating – and goal of a good credit score mean for its residents?
The impact of credit ratings is more far-reaching than many realize, Davon said.
“Obtaining a good credit rating is an important financial goal for governments because good credit yields lower interest rates and shows fiscal responsibility,” Davon said. “The tension – and sometimes the problem for residents – is that although good credit gives governments access to cheaper borrowing, working to obtain good credit may lead governments to act in ways, which ultimately, are unfavorable to residents.”
Davon and co-author Elizabeth C. Martin used state government credit ratings from 1996 to 2012 to demonstrate that state credit rating increases were associated with high economic insecurity for their residents. “We argue that findings illustrate how good credit can often become detrimental, once we consider the potential tradeoffs.” You can read more in their December 2021 paper in the Sociological Forum.
Government credit ratings and race
In his work, Davon saw how biases were not always needed to perpetuate inequality in credit ratings or scores. “The divide between liberal cities and republican cities and their scores led me down this path. But when we think about these overt biases, they are embedded in all of these other things.”
This is especially the case in how racism and racial inequality show up in contemporary financial markets – often in insidious ways that many don’t realize or consider. “While ratings and scores give a veneer of individualized objectivity, their actual inputs reflect decades of racial disadvantage,” Davon said.
For example, Davon noted that “racial inequality in household income has remained the same from the 1970s to today (as CID faculty Robert Manduca’s work shows us).” If such a, what he terms, “racialized input” is used in creating a score or rating then, “We’re not using race, [but] we are embedding the legacy of racism into the score,” Davon said.
He discusses this point in his paper published in Social Problems. Davon demonstrates how “the use of such racialized inputs embeds historical racism in ratings allowing racial inequality to persist and escape cognition as seemingly race-neutral inputs ‘explain away’ racial disparities.”
These findings epitomized how scores and ratings “push up against the epistemological seams of how we understand and identify inequality,” he said.
“How do social scientists and policy makers understand and conceptualize racial inequality?”
Davon began to think “not only about the ways that inequalities are perpetuated, but also the ways that social scientists and policymakers understand how racial inequality presents.”
He was curious because what he argued in his piece in Social Problems was that “in the context of credit scores and ratings, statistically significant race variables that are the hallmark of identifying pernicious racial inequality…don’t exist because those variables are not being used to create the score or rating.”
Instead, “Racialized inputs allow [racial inequalities] to escape cognition” as seemingly fair characteristics like income explain away statistically significant racial inequalities. “Even if we accept the premise that [a racialized input is] acceptable to determine credit worthiness, it is significantly unequal,” Davon said.
A photo of a vintage (circa 1970) credit application from the former retailer Montgomery Ward. Note that the application states, “IF MARRIED, ALL ANSWERS MUST APPLY TO HUSBAND.”
“How did all data became credit data?”
Davon wanted to understand how and whether policymakers conceptualized those “explained away” inequalities as problematic or not because there is a larger movement in consumer credit scoring where politicians, corporations, and consumer advocates are increasingly saying “‘well, the scores are bad but they’re bad because so many people don’t have a score, so we should expand the data used to get a credit score and get more people scored, and doing so will create more racial inclusion.’”
Davon points out that many are seeing all manner of data as relevant data to be used in credit scores. Credit scores are typically concerned with “how much do you have in debt, have you missed any debts/payments? What kind of debt? How long have you had it?” Davon notes, “Increasingly, companies are adding more questions and inputs, like what are you spending money on? What is the activity in your checking/savings account? What is your job title? What did you major in college?”
Davon’s dissertation works to understand how we came to this point by analyzing the history of consumer credit scores. He asks how expanding the data used in credit scores can be understood to generate racial inclusion as his prior work and that of others discusses how fundamentally problematic scores and ratings are because of the racialized inputs that go into them.
Davon argues in his dissertation that a key problem is that while people’s ability to repay debt is a function of all manner of socially driven factors, scores treat all data as if they are individually determined. He traces how this individualized understanding emerged with the passage of the 1974 Equal Credit Opportunity Act and evolved through the early 21st century. He pays particular attention to how legal institutions and politicians grew increasingly unable to identify pernicious inequality, especially racial inequality in credit scores, and how that inability to see inequality informed the growing comfort with expanding the data used in scores.
He notes that his findings have important implications for the future of racial inequality in the face of the expansion of credit scoring. “Seemingly, there is comfort in that additional information. And most seem to think the expansion of credit scoring is going to be a good thing; my dissertation concludes this is going to be a bad thing in part because of how we understand what inequality looks like,” he said. “I’m on the team that if we want to really substantively generate a society that can be broadly considered racially, socially, and economically just, then perhaps we should not be scoring people at all.”
Since arriving at the University of Michigan this summer, Davon has expanded his focus to the broader structure of inequality in the United States and the role that debt and credit scores play in ameliorating or exacerbating that structure.
When wealth inequality researchers and concerned policymakers discuss debt, there is often a focus on the kinds of debt wealthy—and frequently White—people have. “They don’t focus on the debts that poor and minority folks have. It’s all about mortgages, student loans, and credit card debts, it’s less about past due bills, legal fees, payday loans,” he said. Because of this, “There are a lot of empirical questions left to answer that I am determining how best to approach and make sense of.”
At the top of Davon’s list to answer is whether or not it really is good to have a good credit score? Do people with good credit “find themselves in a better economic position later down the road?” he asked. In collaboration with an interdisciplinary team of scholars across several universities (including Ohio State, Dartmouth, and the University of Wisconsin-Madison), he will be leveraging unique proprietary credit report data in the coming months to begin to answer this and many related questions on credit and debt.
Davon said the community of experts here at the U-M have helped to put him in the best place to answer his next research questions. “There are so many people here that are thinking about things that I think are important, creating this ecosystem of resources to help me with my research. There’s probably no better university that exists that would be a better fit for me than the University of Michigan.”
Davon Norris’ research has been published in outlets such as Social Forces, Socio-Economic Review, Social Problems, and Sociological Forum, and has received awards from the Future of Privacy Forum and American Sociological Association. His work has been funded by the American Sociological Association. Davon received his Bachelor of Science in Accounting (2014), Master of Arts in Sociology (2018) and Ph.D. (2022) in Sociology all from The Ohio State University. Learn more about Davon at his website.