Moving Beyond the VI-SPDAT: Integrating Your System’s Values into Prioritization

Written by Alliance Research Council member Marybeth Shinn (Organizational Development Professor, Department of Human and Organizational Development, Vanderbilt University) and Molly Richard (Doctoral Student, Department of Human and Organizational Development, Vanderbilt University)

Until there’s more funding to end homelessness, communities must figure out how to prioritize and allocate limited resources, especially access to housing.

The most widely used tool has been the VI-SPDAT (Vulnerability Index – Service Prioritization Decision Assistance Tool). But in the last year, many are reconsidering the VI-SPDAT after research showed it may not be reliable and may perpetuate racial inequities.

How do we replace the VI-SPDAT and continue to determine who should be prioritized in our coordinated entry systems? This is a huge question in the field. Recently, we wrote an essay to help guide and inform these planning efforts. It summarizes what’s been going on in the field, and pulls ideas from other fields, like medicine, tasked with allocating scarce interventions. The essay doesn’t describe a silver bullet or the perfect new tool for the field to adopt. Instead, it invites communities to take this opportunity to reconsider their values and goals, and to create an assessment system that works for them. Here are some of its key takeaways:

  1. The U.S. Department of Housing and Urban Development (HUD) requires communities to use an assessment tool and prioritize people according to vulnerability, or risk. When we consider what counts as risk, we should be explicit about the values driving the conversation. Vulnerability can mean risk of death on the streets, emergency services use, physical harm, mental health risk, incarceration, risk of continued homelessness, and more. HUD describes types of risk communities can consider in assessments, but local systems must decide the outcomes they consider most important.
  1. Communities can also use weighting factors based on service-user characteristics (e.g. race/ethnicity, military service, early intervention for young children, and preventive intervention for people leaving prison). Veterans, for example, have been prioritized for housing through a value proposition (people who serve their country should not be homeless), not one based on risk or maximizing program effectiveness. But one could argue the ethics of prioritizing other populations, such as people of color (especially Black/African American or Indigenous people), because of past systemic harms and ongoing discrimination that puts them at heightened risk of homelessness. When we can’t use characteristics directly, like race/ethnicity, research from other fields encourages use of proxies like zip code of last residence to reduce disparities.

Values matter, and the ready availability of a tool, however flawed, lets people avoid the values discussion. Communities shouldn’t simply adopt a new tool because it’s popular, but should think carefully about what they are trying to prevent or foster in their approach to homelessness response.

Your Continuum of Care (CoC) can reevaluate its values by taking the following steps:


  • Bring together stakeholders such as service providers, people experiencing homelessness, and other members of your CoC governing board.
  • Review relevant aspects of your system and its data, including the different groups your CoC serves, disparities between people entering the system and exiting into housing, and the varied resources available to you.
  • Based on your system review, work with stakeholders to reach a consensus on your system’s values. For example, these values could include countering systemic racism, keeping young children out of shelter, or helping people returning from incarceration or youth leaving foster care to get a fresh start.
  • Those values should lead to the consideration of certain outcomes of concern or weighted factors. You may choose different risk/vulnerability factors for different populations (as is explored more in depth in the essay).
  • Your value considerations may lead directly to a point system, or possibly different point systems for different groups.
  • If you want to determine which consumers are most at risk of the outcomes that are of concern to your CoC, it may be useful to work with academics/researchers to develop a risk model. However, academics can’t tell you which outcomes should matter most.
  • Not all communities have the same data, staff, or resources available for developing their own tool. You can also test surveys developed in other communities with similar context and values.
  • Ideally, you should monitor how your new system is working, including a) evaluating whether it’s accurately prioritizing people based on your community’s chosen criteria and b) assessing for racial and gender inequities in who is getting housed and who is returning to the system.

There is a lot of great work going on like this around the country. In the future, we might get to a point where there is a database of available tools based on different values, and/or tools that can be adjusted to meet an individual community’s context (e.g., demographics of people experiencing homelessness and their key barriers). We aren’t quite there yet, but we can get started.

We want to highlight that we are having these conversations because we lack resources to serve everyone. Ultimately, systems will work most effectively and equitably when there are more to go around. But until then, we must continue to advocate for more resources, and ensure that we are using available resources effectively.