Detroit Blight Analysis

Title: Detroit Blight Analysis

January 2017 - August 2017


Property blight affects more than 20% of properties in Detroit. The City of Detroit issues tickets to owners of these blighted parcels, which incentivize residents to maintain their properties. However, the compliance rate for these tickets is under 10%, which leaves tens of millions of dollars in unpaid fines. In this paper, we seek to understand why compliance is so low and how violations could be better enforced to effectively address the city’s blight epidemic. To this end, we build a predictive model that forecasts ticket compliance, perform in-depth analysis on the groups of residents who own blight-ticketed properties, and investigate how compliance varies between these very different groups.

Leader(s): Allie Cell (slack: @acell, email: acell)

MDST Participants: Adam Rauh (@amrauh), Bhavika Reddy Jalli (@bhavikaj), Colin Tan (@colintan), Kurt Ayalp (@kayalp), Xinyu Tan (@xinyutan), Subramanian Ramanujam (@akashram), Joshua Bochu (@jbochu)

Faculty Advisor: Danai Koutra

Looking for: --

External Collaborators: Robert Goodspeed (U of M), Miriam Blanks-Smart (Detroit DAH), Joel Howrani Heeres (Detroit)

Achieved Milestone: paper accepted to Data Science for Social Good (DSSG) conference; gave oral presentations at DSSG and at the Urban Informatics Unconference; presented poster at MIDAS Symposium.

Deliverables: Paper for DSSG conference

Link to the Code:

Link to the Dataset:

Link to the Paper: