Organizations routinely generate gaps between groups even when no one intends to. You cannot close a gap without first understanding how it was generated. This course is about that diagnosis, and about designing interventions that actually close gaps rather than appearing to.
Every session puts you in the decision seat with real data. You will not be told the conclusion. You will find it, defend it, and then discover what your own recommendation did to people you did not consider.
A 30,598-person manufacturer has to cut labor costs. Seven criteria are on the table, each facially neutral: seniority, absenteeism, AI exposure, performance rating, geography. Choose one and defend it. Then the demographic data arrives.
A recidivism model is calibrated and produces unequal error rates by group. Work through why competing definitions of fairness cannot all be satisfied at once, and decide which one you are willing to sacrifice.
A uniform global carbon price treats every country identically. Build the case that it is fair. Then build the case that it is not, using the same numbers. Learn what selective framing looks like from the inside.
Every disparity you examine gets located on one grid. One axis asks what process produced it: allocations, meaning who is matched to which opportunity, or valuations, meaning how contributions are scored and rewarded. The other asks what behaviour drove it: differential treatment, where identity changes the decision, or disparate impact, where a neutral rule meets inputs already correlated with identity.
The cell you land in determines which intervention can work. Misplace the mechanism and the fix fails, however well intentioned.
People are sorted into different opportunities based on identity. Different decisions produce different outcomes in who gets which role, project, or resource.
Session 4 · Pay equityNeutral assignment rules produce unequal access, because the inputs they rely on are correlated with identity through prior structural processes.
Session 5 · Axis BankThe same performance is evaluated, rated, or rewarded differently depending on who produced it.
Session 2 · Meritocracy paradoxNeutral evaluation criteria produce unequal rewards, because the criteria themselves encode structural disadvantage.
Session 3 · COMPASEach session is tagged by where it sits on the course's organizing framework: whether the disparity arises from differential treatment or disparate impact, and whether it operates through allocations or valuations.
You work the data in the room, commit to a recommendation, and then see what your recommendation did. The tools are built for this course and run once, in class. They are not posted afterward, because the exercise only works the first time.
One is public, from a case on degree requirements in job postings. It takes about four minutes and it is a fair sample of how the rest feel.
A four-question walkthrough on what a degree requirement does to an applicant pool. Then the full simulator, if you want it.
degreescreen.netlify.app →Four sessions this term are taught alongside practitioners who have made these decisions with consequences attached. Session 12’s guest is still being finalized between two strong candidates. Guests join the discussion rather than lecture at it, and students ask most of the questions.
Judge, Hamilton County Municipal Court
Session 6 · November 11
Community benefits agreements from the bench: what evidence a decision-maker actually needs before authorizing an intervention.
McKinsey & Company · Michigan graduate
Session 8 · November 18
Co-author of the AutoNow case. On what the analysis looks like in the rooms where restructuring decisions are actually made, and what is usually missing from them.
Professor of Medicine · University of Michigan Health
Co-Medical Director, Kidney & Pancreas Transplantation
Session 11 · December 7
Medical Director of the Transplant ACU. Joins the session on Combing the Records to discuss the kidney-function formula from the clinical and institutional side: how the race coefficient entered practice, and what it takes to correct a waitlist after the fact.
Session 12 · December 9 · one or both confirming
Mehri co-wrote the report that produced the NFL’s Rooney Rule and is the addressee of the capstone memo assignment. Duru is the leading legal scholar on the Rule and author of Advancing the Ball. Either closes the course on turning analysis into policy change; both would be better.
Co-Founder and CEO, Working IDEAL · formerly Senior Advisor, U.S. Department of Labor
Works on exactly what this course teaches: pay-equity assessment, hiring and promotion program design, and workplace data analysis for companies, cities, unions, and nonprofits. At the Labor Department she led the development of investigative guidelines for assessing federal contractor pay systems. She has since led the racial equity assessment of BlackRock. JD Harvard, PhD Berkeley.
Alfred P. Sloan Professor of Management, MIT Sloan · Associate Dean for Community Engagement
Studies how demographic characteristics shape which relationships form, how diversity affects what a team can learn, and how organizational climate affects who stays and who performs. An Academy of Management Fellow. The network mechanisms he works on are the ones operating underneath several of the disparities students diagnose in this course.
Vice Dean & Senior Associate Dean for Professional Degree Programs, UVA Darden School of Business · formerly Head of Global Diversity and Belonging, Airbnb
Twenty-five years studying what unleashes and amplifies the contributions of individuals, particularly women and underrepresented groups, in teams and negotiations. Built and ran Airbnb’s global diversity, inclusion, equity, and belonging strategy before returning to academia. Also Professor of Public Policy at Darden’s Batten School.
Group Head – Human Resources, Axis Bank · Chairperson, FICCI HR Committee
Over 27 years across leading financial services organizations. Champions the Bank’s distinctive people practices, including GIG-A Opportunities, the workforce-flexibility program at the center of Session 5. Joined an earlier offering of this course to discuss it from the inside.
Head of Product Equity, Instagram (Meta) · currently Associate Creative Director, Experience Strategy, Apple
Built and led Instagram’s product equity function, after leading youth product strategy for Meta’s metaverse work and AI products for WhatsApp. Product leadership roles at Savage X Fenty and One Medical, and freelance product strategy for Google, Target, Reebok, and Discover.
Every case below was written for this course rather than borrowed for it. Several are published through WDI Publishing and now taught at other schools. You will be working material that did not exist until this course needed it.
A 30,598-employee workforce file, eight downsizing criteria, and a general counsel who wants an adverse-impact analysis. Ships with a browser-based simulation. Co-authored with a Michigan graduate now at McKinsey, who joins the session.
The equity framework applied to global climate policy, and an exercise in arguing competing positions from a single dataset.
An enterprise pay audit where divisional aggregates conceal what the disaggregated data shows. Third place, 2023 DEI Global Case Writing Competition.
Three decades of hand-coded coaching data. Counterfactual modelling of a policy intervention, and the difference between symbolic and substantive compliance. The capstone.
A recidivism algorithm that excludes race from its inputs and still produces unequal error rates by group. The session where students learn why competing fairness criteria cannot all hold at once.
A work-from-anywhere programme in India, designed to widen access and audited for what it actually did across hiring, project allocation, evaluation, advancement, and attrition.
Black attorney representation in Cincinnati, and the arithmetic of what closing a representation gap would actually require in hires and pipeline investment. A sitting municipal court judge joins the session.
A race coefficient inside a clinical formula, the transplant waitlist it reshaped, and what a medical director owes the patients it already passed over. Traverses all four cells of the framework on one outcome.
Mondays and Wednesdays, 8:00 to 10:20, for seven weeks. It is early, it is twice a week, and I am not going to pretend the hour is a feature of the design. It is the slot the course has.
What I will say for it: nobody drifts into an 8 a.m. elective. Everyone in that room chose to be there, which is why the discussion is better than the schedule deserves.
And the block is long enough to run a full analysis, argue about it, and change your mind before the session ends. That does not happen in seventy-five minutes. Nine of the twelve sessions put you in the data during class, and those exercises need the room.
A student from an earlier cohort put it better than I can. I cannot make this up:
Professor Rider is a great lecturer who made 8 a.m. two-hour lectures fly by.
Students also asked for one change I could actually make: move the course out of the end of the school year, when everything else is due at once. That is why it now runs in the fall.
On workload, since it is the other thing worth knowing before committing to an early start: across every offering, students have rated it typical for a course of this credit. Demanding, not punishing. The reading is bounded and it is mostly cases.
Anonymous course evaluations, across three offerings. Selected, and therefore not a random sample — a point the course itself would insist on.
Chris does a great job balancing a rigorous approach to statistics while making it friendly to those who don't have a deep technical background. He is an excellent lecturer and educator, who inspires his class to make a meaningful difference in their future workplaces.2026
I really enjoyed how we covered a breadth of case studies in areas that seem so far apart from each other — higher education, professional sports, climate change. Operating from the lens of the fundamental 2x2 framework made it easier to compare apples to apples even across different industries.2026
Even if I don't remember all the statistical analyses, I will certainly remember the three-step process for equity analytics in the future.2023
This course challenged me to look at equity in a new, empirical way. Chris is a very effective facilitator of difficult topics in a room with very diverse voices. His guest speakers are incredible and add significant value to the conversations.2023
Prof Rider presents what can often feel like a very amorphous topic in a very concrete way, rooted in data.2026
The 2x2 framework gave me a foundation to think about equity in the work environment. I'm super glad I took this course. Although I wasn't sure what to expect from it, I highly recommend it.2024
I wanted to take the course since last year, and it was one of my best classes to date.2023
I wish there was a Part B or semester-long version of this course. There is so much content to learn and practice that seven weeks doesn't feel like enough.2023
Open to first-year and second-year MBA students, Master of Management students, and other Ross graduate programs.
Enroll through the standard registration system. Search class number 37250 or ES 616-001.
Students from Rackham, the Ford School, the School of Information, Public Health, and Law are welcome, and several have taken the course.
Enrollment runs through a request submitted to the Ross Registrar's Office. They handle eligibility, seat availability, and the request itself.
rossregistrarsoffice@umich.edu(734) 647-4933
Reference ES 616-001, class number 37250, Fall B 2026.
The catalog lists TO 502, STRT 502, and MO 503, or concurrent enrollment. Treat those as guidance rather than a gate. What the course actually requires is elementary statistics and basic spreadsheet work. There is no programming.
If you think you are prepared without them, request a prerequisite waiver from the Ross Registrar’s Office. It gets routed to me, and I approve waivers for students with sufficient statistics background, relevant professional experience, or both. First-year MBA students are eligible.
rossregistrarsoffice@umich.edu · (734) 647-4933
Elementary statistics and comfort with a spreadsheet. Nobody is asked to write code. The catalog lists three prerequisite courses; if your background or professional experience prepares you without them, request a waiver from the Ross Registrar’s Office and it comes to me for approval. Students who arrive with statistics training go deeper into the technical material; the analytical reasoning is the point rather than the software.
Seven weeks, twelve sessions, two case analyses, and a final project you choose. The reading is substantial but bounded, and it is mostly cases rather than journal articles. The work is in the thinking rather than the volume.
No. It is a course about inference. The central question is whether an observed gap was produced by the process you think produced it, and whether a proposed fix targets that mechanism or a different one. Students reach different conclusions about what organizations should do, and the course does not adjudicate that.
A repeatable workflow for taking a disparity from observation to defensible inference, and the ability to tell whether a recommendation you have been handed will survive scrutiny. Students also leave with a written reference guide covering the full analytical sequence.
Consulting, people analytics, strategy, product, and law. Anywhere you will be asked whether a decision is defensible, not only whether it is profitable.
Four formats, including an original analysis of a disparity in a setting you care about, and a drafted teaching case. Strong cases receive detailed feedback and are encouraged toward publication; several have been adopted into later versions of the course.
If you are weighing whether the course fits, write to me. I answer every message myself, and I would rather talk it through than have you guess from a page.
Fit with your background, workload, what the sessions are actually like, whether it makes sense alongside your other electives.
Eligibility, seat availability, and enrollment requests, including for graduate students outside Ross. Reference ES 616-001, class 37250. Or call (734) 647-4933.