The overarching goal of the Workshop on Mathematics and Racial Justice is to explore the role that mathematics plays in today’s movement for racial justice. For the purposes of this workshop, racial justice is the result of intentional, active and sustained anti-racist practices that identify and dismantle racist structures and policies that operate to oppress, disenfranchise, harm, and devalue Black people. This workshop will bring together mathematicians, statisticians, computer scientists, and STEM educators as well as members of the general public interested in using the tools of these disciplines to critically examine and eradicate racial disparities in society. Researchers with expertise or interest in problems at the intersection of mathematics, statistics and racial justice are encouraged to participate. This workshop will take place over two weeks and will include sessions on Bias in Algorithms and Technology; Fair Division, Allocation, and Representation; Public Health Disparities; and Racial Inequities in Mathematics Education.
Mathematics is often viewed as one of the main tools responsible for scientific progress, and developments in mathematics are behind some of society’s most significant technological advancements. While mathematics has been used to push society forward, there are also well documented instances of mathematics being used as a tool of racial oppression. With increasing frequency, the inequities faced by the Black community are becoming more difficult to ignore, and mathematicians have increasingly been answering the call to engage with issues of social justice within their research, their teaching, within their institutions, and in the broader mathematical sciences community. This workshop is a part of that movement and makes the distinct contribution of centering issues of mathematics and racial justice. Further, we intend to approach the topic of social justice from the lens of justice for the Black community.
The workshop organizers have provided this list of Suggested Resources. MSRI is working to provide access to many of the resources listed and will update the links as they become available. If there are e-resources that you cannot get, please do not hesitate to email our librarian.
Workshop Organizers: Caleb Ashley (Boston College), Ron Buckmire (Occidental College), Duane Cooper (Morehouse College), Monica Jackson (American University), Omayra Ortega* (Sonoma State University), Robin Wilson* (California State Polytechnic University, Pomona)
* denotes lead organizers
June 9, 2021
Keynote: Seeking racial equity and social justice in mathematics teaching and learning
Robert Berry (University of Virginia)
This session will engage participants in unpacking mathematics teaching and learning focused on racial equity and social justice. Specifically, the session will explore the intersection of mathematics teaching and learning with racial equity and social justice across four critical reasons: 1) building an informed society; 2) connecting mathematics to cultural and community histories as valuable resources; 3) confronting and solve real-world mathematics as a tool to confront inequitable and unjust contexts, and 4) use mathematics as a tool for democracy and creating a more just society.
June 9, 2021
Keynote: Roles for Computing in Social Change
Rediet Abebe (University of California, Berkeley)
Recent scholarship warns that computing work has treated problematic features of the status quo as fixed, failing to address and often even exacerbating deep patterns of injustice and inequality. This begs the question: what roles, if any, can computing play to support and advance fundamental social change? Through an analysis informed by critical scholarship, we articulate four such roles -- computing as a diagnostic, formalizer, rebuttal, and synecdoche. We then illustrate how mathematical and computational tools can aid in understanding and tackling poverty and social inequities through the role of computing as a formalizer. We close with a discussion on the Mechanism Design for Social Good (MD4SG) research community, which works to bridge research and practice to ensure that such insights can be leveraged to advance social change.
Panel (moderated by Lou Matthews)
If you would like to watch this recording, please email email@example.com.
Panel: What to expect over the next 5 days
June 10, 2021
Bias in Algorithms and Technology
Algorithms undergird and dominate many aspects of modern society. Although central, the presence of algorithms in our daily lives is far from innocuous. Part of this danger is inherent, it is bound up in the fact that the operating parameters and the explicit roles of algorithms are suppressed or hidden. Bias in design or implementation further compound the potential threat of algorithms to be harmful social forces.
Determining if an algorithm acts as a mechanism by which structural inequalities are perpetuated in society is fundamentally important on moral, ethical, and legal grounds. The existence of types of “automated social bias” has been established in a variety of algorithms, from those governing housing and lending practices, to those employed in policing and the criminal justice system. We will explore many facets in the creation/lives of algorithms which adversely affect social justice during this session. Our emphasis will primarily center on bias in algorithms which threaten racial justice. We ask how mathematics can be wielded to detect, even dismantle, such bias in algorithms.Algorithms undergird and dominate many aspects of modern society. Although central, the presence of algorithms in our daily lives is far from innocuous. Part of this danger is inherent, it is bound up in the fact that the operating parameters and the explicit roles of algorithms are suppressed or hidden. Bias in design or implementation further compound the potential threat of algorithms to be harmful social forces.
Determining if an algorithm acts as a mechanism by which structural inequalities are perpetuated in society is fundamentally important on moral, ethical, and legal grounds. The existence of types of “automated social bias” has been established in a variety of algorithms, from those governing housing and lending practices, to those employed in policing and the criminal justice system. We will explore many facets in the creation/lives of algorithms which adversely affect social justice during this session. Our emphasis will primarily center on bias in algorithms which threaten racial justice. We ask how mathematics can be wielded to detect, even dismantle, such bias in algorithms.
Welcome: Bias in Algorithms and Technology
Plenary Talk: Designing for Equity
Sharad Goel (Stanford University)
Machine learning algorithms are now used to automate routine tasks and to guide high-stakes decisions. In the first part of this talk, I'll describe an evaluation of automated speech recognition (ASR) tools, which convert spoken language to text, and have become increasingly widespread, powering popular virtual assistants, facilitating automated closed captioning, and enabling digital dictation platforms for health care. Over the last several years, the quality of these systems has dramatically improved, due both to advances in deep learning and to the collection of large-scale datasets used to train the systems. There is concern, however, that these tools do not work equally well for all subgroups of the population. Indeed, I'll show that five state-of-the-art ASR systems — developed by Amazon, Apple, Google, IBM, and Microsoft — exhibited substantial racial disparities, making twice as many errors for Black speakers compared to white speakers. In the second part of the talk, I'll describe a general, consequentialist paradigm for designing equitable decision-making algorithms. Stakeholders first specify preferences over the possible outcomes of an algorithmically informed decision-making process. For example, lenders may prefer extending credit to those most likely to repay a loan, while also preferring similar lending rates across neighborhoods. One then searches the space of decision policies to maximize the specified utility. I'll describe a method for efficiently learning these optimal policies from data for a large family of expressive utility functions, facilitating a more holistic approach to equitable decision making.
Plenary Talk: Sources and consequences of algorithmic bias
Maria De-Arteaga (The University of Texas at Austin)
In this talk I will first provide a taxonomy of different sources of bias in machine learning algorithms. I will then present novel results on the effect of differential victim crime reporting on predictive policing systems (FAccT’21). Previous research on fairness in predictive policing has concentrated on the feedback loops which occur when models are trained on discovered crime data, but has limited implications for models trained on victim crime reporting data. We demonstrate how differential victim crime reporting rates across geographical areas can lead to outcome disparities in common crime hot spot prediction models, which may lead to misallocations both in the form of over-policing and under-policing. I will conclude the talk by discussing paths forward for research on algorithmic fairness, arguing that reliable assessment and design require us to center AI-assisted decisions, rather than AI predictions, as the locus of evaluation.
June 11, 2021
Disparities in Public Health
Everyone deserves the right to good health, yet unnecessary and unjust inequities persist in our healthcare system. These inequities are the result of policies and practices that create an unequal distribution of money, power and resources among communities based on race, class, gender, place and other factors. Healthy People 2020 states that, “Racism has been linked to low birth weight, high blood pressure, and poor health status. Further, the 2015 National Healthcare Disparities Report indicated that white patients receive better quality of care than 36.7% of Hispanic patients, 41.1% of black patients, 32.4% of American Indian/Alaska Native patients, and 20.3% of Asian and Pacific Islander patients.” A racial justice approach to health equity requires that we address how issues of racism, disproportionate distribution of wealth opportunities, and privilege within society produce negative outcomes for the Black community. In this session we will discuss the racial disparities that exist within healthcare, how mathematics has been used to perpetuate these problems, and how mathematics can be used to identify and mitigate these disparities.
Welcome: Public Health Disparities
Plenary Talk: The Pandemic within The Pandemic
Darius McDaniel (Leidos)
The SARS-CoV-2 pandemic has brought about a new level of fear and many more constraints and challenges. This pandemic has clearly amplified the reality of unequal health outcomes for people of color. Racial disparities are pervasive in our health care system. The intersection of numerous diseases is leading to greater morbidity and mortality in existing at-risk groups.
Researchers within many areas of Public Health and Medicine are finding that the roots of discrimination run deep and the dis-ease beneath the diseases are killing from all angles. Mounting epidemiological studies are demonstrating disparities amongst race within both communicable and non-communicable diseases.
From data on maternal-child health to Alzheimer’s and now Coronavirus, the statistics are revealing the unbalance of care and treatment. The data is clear no matter the phase of life the pandemic of racism effects a person from birth to late stage of life.
Within this talk we will aim to discuss many of the underlining roots of disparities throughout Public Health research areas. We will also provide a glimpse at research fighting for equality of health. Lastly, through personal and research experiences in Alzheimer's and COVID-19 we will aim to look at how we can use the mathematical sciences for social and systemic change.
Plenary Talk: Race and causality in health disparities research: time for a necessary paradigm shift
Emma Benn (Icahn School of Medicine at Mount Sinai)
In this talk, I intend to challenge the way in which we are traditionally taught to operationalize race in biomedical research. By introducing a causal perspective, my hope is that we can move away from a descriptive approach and towards an inferential approach that moves us closer to finding effective interventions to reduce racial disparities in health.
Panel (moderated by Julie Ivy)
Emma Benn (Icahn School of Medicine at Mount Sinai), Darius McDaniel (Leidos)
June 16, 2021
Racial Inequities in Mathematics Education
There are many different ways in which mathematics education and racial justice intersect. Mathematics as a gatekeeper and a catalyst to academic success has long been a feature of higher education in the United States. In the classroom these experiences play out at the individual level from teacher-to-student and student-to-student. At the systemic level, there are multiple overlapping and interconnected inequities based on race that affect students of color in general and Black students in particular. —Some examples are a focus on standardized testing over rich problem-solving opportunities and equitable teaching practices, funding formulas for public education that are directly connected to long-lasting racially segregated housing patterns, a deficit view of Black children in mathematics education research, overlooking Black students for advanced coursework, and the dehumanized, impersonal view that mathematics identities do not intersect with racial identities. In this session of the workshop, we will explore the barriers towards dismantling racial inequities in mathematics education that still persist despite numerous, longstanding movements to eradicate or at least ameliorate them. Further, we intend to use this space to envision what racial justice in mathematics education can look like for Black students.
Welcome: Racial Inequities in Mathematics Education
Plenary Talk: Teaching to Transgress: Mathematics as a tool for social justice
Brittany Mosby (Tennessee Higher Education Commission)
Plenary Talk: Rethinking Equity and Inclusion as Racial Justice Models in Mathematics (Education)
Danny Martin (University of Illinois at Chicago)
I present a critical perspective on equity and inclusion as racial justice models in mathematics (education). I situate my critique in the context of efforts designed to increase Black representation. These efforts include recent reforms in K-12 mathematics education. Despite these reform efforts, many Black learners continue to experience dehumanizing and violent forms of mathematics education. I suggest that these forms of mathematics education are rooted in white supremacy and antiblackness, which have always functioned as self-correcting, multi-level projects of Black exclusion. Although equity and inclusion initiatives align with progressive sensibilities, these initiatives are often accommodated in ways that do not threaten the overall functioning of white supremacy and antiblackness. Framing mathematics education as a political-racial project aligned with other political-racial projects helps to explain why inclusion at certain levels of these projects has not diminished the intractability of Black exclusion at other levels.
Panel (moderated by Jalil Cooper)
Danny Martin (University of Illinois at Chicago), Brittany Mosby (Tennessee Higher Education Commission)
June 17, 2021
Fair Division, Allocation, and Representation
The question of how finite resources can be shared is an important and interesting problem with mathematical underpinnings, historical significance, and increasing relevance to modern life. The principle of fairness has particular resonance to questions of racial justice in response to historically unfair systems and treatment. Mathematics can (and should) be involved in any discussion, analysis, or decision-making process about fair division, allocation, and representation. In this session, presenters will address some aspects of the mathematics of fairness, both in theory and application. Theoretical considerations include presentation of definitions, axioms, and theorems. Applications include algorithms for fair division of entities among multiple parties and fair allocation of goods that are not divisible. Fair representation has applications that include the apportionment to states of seats in the U.S. House of Representatives and the contemporary disputes about how the states and other regions are divided into districts for election of representatives.
Welcome: Fair Division, Allocation, and Representation
Plenary Talk: Elections and Representation
Michael Jones (Mathematical Reviews)
In this talk, I will introduce the mathematics and applications of election procedures and apportionment methods.
Elections are easy when there are only two candidates: vote by majority rule. For three or more candidates, Kenneth Arrow’s Impossibility Theorem shows that no three-candidate election procedure satisfies a set of reasonable axioms, implying that there is no “best” election procedure. After discussing the axiomatic approach in voting theory, we will review commonly used election procedures, including the use of ranked choice voting in East Pointe, Michigan as part of the resolution of a Voting Rights Act lawsuit.
In the context of the US House of Representatives, the apportionment problem is to determine the number of representatives each state receives in the House. We will review the history and mathematics of apportioning the House, including its relationship to the Electoral College. We will conclude with the recent use of apportionment methods to allocate delegates among candidates in the Democratic and Republican presidential primaries.
Plenary Talk: Once in a Decade Opportunity to Address Gerrymandering
Stephanie Somersille (Somersille Math Education Services)
Once a decade issues of fair representation in government come to the forefront and that time is now. The census numbers were just released in April and the process of allocation and redrawing district lines has begun. In the first part of this talk, we will give a brief overview of gerrymandering and the voting rights act and introduce the new Geometry and Election Outcome (GEO) metric. This metric is unique in that it uses both election data and map data to quantify gerrymandering.
For the second part of the talk we will discuss alternate voting processes which may better approach the ideal of “one person one vote”.
A recording of this panel is not available at this time.
Panel (moderated by Ron Buckmire)
Ron Buckmire (Occidental College), Michael Jones (Mathematical Reviews), Stephanie Somersille (Somersille Math Education Services)
A recording of this panel is not available at this time.
2021 Workshop Sponsors
The 2021 Workshop on Mathematics and Racial Justice is sponsored by the National Science Foundation, the American Mathematical Society (AMS), the Center for Minorities in the Mathematical Sciences (CMMS), the National Association of Mathematicians (NAM), and the Society for Industrial and Applied Mathematics (SIAM).