An Agent-Based Explanation for the Housing Plight of America’s Mentally Ill

An Agent-Based Explanation for the Housing Plight of America’s Mentally Ill

The largest public mental health facility in the United States is not a hospital; it is the Los Angeles County Jail.

A Severely and Persistently Mentally Ill (SPMI) patient generally defines someone with a diagnosis of Schizophrenia, Bipolar Disorder, or Major Depressive Disorder, and this group constitutes about 1.7% of the US population. To better understand the condition and possible improvements, IBM Global Research and Otsuka Pharmaceuticals used an agent-based approach to model the SPMI living situations over the second half of the 20th century.

Kyle Johnson, Managing Consultant with IBM Global Business Services presented this project at the 2014 AnyLogic Conference. He works within the Advanced Analytics and Optimization branch of IBM BAO. Kyle is an expert in Data Mining, Agent-Based Modeling Simulations, and Machine Learning and an Adjunct Professor at Carnegie Mellon University.

The presentation describes an agent-based approach to explaining why prisons and jails house so many of America's most severely mentally ill. It traces this fact back to legislation passed in the 1960's, which allocated public funding away from state mental hospitals. The model contains five main components, each based upon a feature of the SPMI population's real world housing environment.

  1. The housing cycle
  2. The seven housing types
  3. The SPMI Patient Agent
  4. Different responses to identical crisis
  5. U.S. Legislation placed a ceiling on the percentage of patients who can live in long-term hospitals

This simple agent-based model improves understanding of Severely and Persistently Mentally Ill housing dynamics in multiple important ways. First, the model shows that even a weak length of stay alteration due to SPMI crises produces a unyielding effect. Second, increasing the patient time to the crisis has a significant positive impact on the population's housing makeup. Finally, the model raises ethical concerns for future mental health policy in the United States.

An effort to improve SPMI patient health in the information technology realm is currently underway with a partnership between Otsuka Pharmaceuticals and IBM. This project seeks to use care coordination information technology to help a geographic area's many health care providers work together to treat efficiently SPMI patients. The effort is quickly gaining traction and has even received the attention of the United States House of Representatives.

Learn more about this project in the case studies section of, by watch Kyle's presentation or through review of his published paper.