Data & Democracy Research Initiative
Researching data science interventions for democracy
Big things are happening!
The outfit formerly known as MGGG (aka the Metric Geometry and Gerrymandering Group) is entering a new chapter as the Data and Democracy Research Initiative (DDRI) in our new home at the University of Chicago. DDRI is a big tent that supports projects on democracy, governance, and data. Our lab (now the Data and Democracy Lab) will anchor the initiative.
To help everyone stay updated, we’re excited to launch a twice-per-year DDRI newsletter where you can keep up with our research and software projects. Whether you’re interested in new research directions, software releases, upcoming events, or general news from the lab, we’ll compile all of it here.
We hope you’ll stay engaged and connected with us as we continue to evolve!
Contact UsResearch Highlights
New papers
Some recently published papers by members of the lab and collaborators include reversible recombination in SIAM Review, Models of random spanning trees in Random Structures & Algorithms, weighted voting in EC’26, and a new polarization index in Annals of Applied Statistics.
Current collaborative projects
Some of our recent oddball ideas: an “autocomplete” model of preference formation; a new segregation score that captures diffusion effects in American cities; and a new framework that looks at how badly voting rules fail various guarantees, not just how often.
Student research
Students in the lab have been pursuing cool projects on representation and spatiality, the geometry of voting rules, the “veto core” in social choice theory, the shape of the space of redistricting plans, and new topological characterizations of Arrow’s impossibility.
Outreach
Election modeling team
Our community partners often ask for help figuring out potential impacts of changing the electoral system in a particular city, county, or state. During the summer, three lab interns completed studies for a number of localities including revisiting several places we’ve studied before (Portland, OR, Santa Clara, CA, and Chicago, IL) plus a new look at San Diego, CA. We used probabilistic models to generate possible voter preference profiles and then compare representational outcomes under plurality, IRV, STV, Top-Two, Top-Four, and more.
Explore more of our past election system case studies →Partners and projects
We love to work with teams around the country who do grassroots and civil rights work in the democracy space. Some recent partner work includes modeling for the Multiracial Democracy Project at George Washington Law, building mapping modules for collaboration with the Brennan Center and for Navajo Nation, and designing an interactive tool for exploring Latino turnout for UnidosUS.
Software Updates
Districtr
v2.0
Districtr, our public-facing package for drawing districts and communities, has recently gotten a huge overhaul on the frontend and backend. New features include:
- Updated election data through 2024 (in most locations)
- Map coverage for all 50 states, DC, and Puerto Rico
- A new “Super Draw” mode that lets you break precincts down to the block level to obtain legally accurate maps
- Advanced evaluation tools available on every map – and more to come!
- A new Content Management System (CMS) that makes it easier for city officials and other partners to assemble a gallery of public feedback
The legacy version of Districtr will remain available as a read-only copy, so existing work can still be accessed.
VoteKit
v3.5
We call VoteKit our “Swiss army knife for computational social choice research” – that means you can use it to study electoral systems, from practical ones like ranked choice to far-fetched ones like OneShotMultiRandomDictator. The recent changes reflect user requests in the last few months, and more is on the way.
- More formats supported for candidate names
- More support for reproducibility
- Better forms of queries to find particular ballots of interest
- New voting rules like SimultaneousVeto and SerialVeto
- Animations for ranked choice elections (and more on the way!)
- A user-friendly front end called VoteKit Lite gives access to some functionality for the Python-phobic
GerryChain
v1.0
GerryChain was originally launched back in 2018 and has been in continuous development since then, but the releases are only just catching up! This package implements the recombination family of algorithms for making districting plans – and a whole lot more. New highlights include:
- A new graph backend abstraction, no longer tied directly to
networkx.Graph, increasing speed 2–3x - A cleaner, more composable ReCom/proposal API
- Support for multi-member districts
- Support for Python 3.11–3.14 and Pandas 3.0
- Attention to backwards compatibility
GerryTools
v2.0
GerryTools is a “sidecar” for GerryChain that supports plotting and mapping and other reporting features.
- Dramatically faster scoring of plans and ensembles (partly thanks to a Rust-backed scoring engine)
- Improved interoperability with GerryChain
- A new modular plotting API for histograms, boxplots, violin plots, seats-votes curves, maps, and more
- Improved tools for comparing plans and publication-oriented LaTeX/TikZ output
- Better tools for interacting with the Census API
RustReCom
v0.2
RustReCom is our super fast implementation of recombination in Rust, a high performance language with good memory safety guarantees.
- Unified command line tool for chains and heuristic optimization
- Better customizability, such as for edge weights
- Written for compatibility with BENDL, our bespoke highly compressed data format
Big shoutout to the team for all of their hard work on these new releases! Please feel free to send your feedback to code@data-democracy.org as we continue to improve these tools.
Talks & Events
danah boyd
Data Are Made, Not Found: A Story of Politics, Power, and the Civil Servants Who Saved the US Census
View event →Jobs
Research Director, Data & Democracy Research Initiative
The Research Director will run the research initiative at a high level and serve as the public face of DDRI’s external engagement. The role carries primary responsibility for the lab’s partnerships, fundraising, and public communications — working at the intersection of math, law, policy, and open-source software to translate academic research into policy recommendations and implementation with governmental, legal, and civic partners.
Apply →Stay connected with DDRI.
If you’re interested in getting in touch to discuss our research, outreach or events, email us to start a conversation - we’d love to hear from you. Keep an eye out for new semi-annual newsletter editions to stay up to date on our work!
Contact Us