A prototype System Model of Pierce County Homelessness

• Gerrit Nyland – Catholic Community Services – gerritn@ccsww.org – 253-304-5105
• I developed a tool that models the homeless system in Pierce County.  The objective of this tool is to spur conversations around our funding options to address homelessness (his hidden objective is to use the model to manipulate you into funding the homeless system how he thinks, in his arrogance, it should be built out. Be careful who you trust… –ed)
• The tool is still a prototype, and is mostly accurate.  The model simplifies a very complex system in order to create an interactive tool.  Most of the model uses data tracked in our community.  The model attempts to show the effect of increasing or reducing funding to different programs.  Do keep in mind it is only a model.  The list of disclaimers is like what you’d expect from a magazine drug advertisement.  A big pretty picture with 10 words in a large font, with a ridiculous amount of small type explaining why that pretty picture may not happen.  Except I’m not providing the fine text, just the pretty picture.
• Again, this is still a prototype.  Neither the County nor the City have bought off on this as either an accurate model nor a preferred way to engage in a discussion about responding to homelessness.  Mind you, neither has said it isn’t accurate.  One of the concerns from County is that it presents some rather daunting costs to end homelessness that may impede our ability to get the funding we need.  That is a legitimate concern.  Tacoma would like a City specific model, but I think that the siting of so many of the County’s services for folks experiencing homelessness within the City of Tacoma makes a model of the entire County the only realistic option. 
• I’m being a bit naughty sharing this prototype, but I’m eager to get additional feedback on this tool before I go live with it.  Since I started working in homelessland 2 years back, I’ve wanted to know the answer to the question of “what resources do we need to end homelessness”.  This is my attempt to answer that question. 
• So, pop over to https://exchange.iseesystems.com/public/gerritnyland/pchomeless/index.html#page1 and take a look. 
• This model starts with our current homeless population (a hair over 5,000 people) and predicts what the homeless population will look like in 5 years.  That is the graph in the lower left.  The red line represents all folks experiencing homelessness.  The Blue line represents a subset of the population that can only be assisted with Permanent Support Housing (a housing subsidy and case management for the rest of your life).  Click the green button to run the model. 
• There are 6 controls that increase or decrease the homeless population.  Adjust these dials and sliders, then click the “Run Model” button to see how they affect the homeless population in 5 years.
  • Annual Diversion Funding (center green dial) – this controls the amount of funding that goes to the one-time assistance program called Diversion.  Everyone experiencing homelessness is offered this one-time assistance, which can help with the deposit for an apartment, relocation costs to move to a family member who will guarantee permanent housing, or other costs that will provide stable housing.  It is an in depth conversation that helps a client identify their strengths and assets to end homelessness on their own, often with no financial assistance.  And if financial assistance is provided, it is only a one time deal.  Diversion funding only helps so much.  It is not a deep enough intervention to assist many families, but it is so cost effective, it is a no-brainer to fund it.  Remove it, and the homeless population skyrockets.  But the optimal amount is about $1.4 million.  More funding than that doesn’t decrease the homeless population, because once all the households that can be helped with it have been helped, it can’t help any additional households. 
  • Annual RRH Funding – this control adjust the amount of funding that goes to Rapid Rehousing.  Rapid Rehousing is essentially 6 months of a rent subsidy for households renting from landlords out in the community.  The idea with Rapid Rehousing is that clients will be housed and then have a chance to work on employment or mental health or increasing income or substance use.  Whatever the client wants to work on.  The housing provides space to be successful.  This program currently helps about 70% of the clients that enroll.  This is the work horse of the homeless system in Pierce County and throughout the US.  It can help most households experiencing homelessness get and keep housing.  It takes  a lot of additional resources added to Rapid Rehousing to get the red line near zero.  However, the red line doesn’t hit zero, because the blue line, representing the clients that can only be helped by Permanent Supportive Housing are still out there. 
  • Annual PSH Unit Funding – this control adjust the annual funding spent on Permanent Supportive Housing.  Again, Permanent Supportive Housing is case management and a rental subsidy (where a household pays 1/3 of their income to rent, and the subsidy covers the remaining rent and utility costs) that lasts forever.  The model is a bit unsophisticated in its display of the effect of adding funding.  The goal is to get the blue line to zero for the entire 5 years, which you can do for around an additional $4 million annually in funding.  And depending on how you provide that housing (using community landlords or building your own building) will affect the capital costs associated with these additional units (a slider lets you determine how much is project based where a provider owns the facility vs. tenant based, where the landlords in the community are paid rent for the units). 
  • Annual Homeless Prevention Funding – this control adjusts how much funding goes into homeless prevention.  We do almost none of this in our community.  People often criticize the homeless response system for funding so little homeless prevention.  The idea that preventing homelessness is far cheaper than addressing homelessness is very attractive.  Unfortunately, the reality is that it is extremely expensive to use homeless prevention to address homelessness.  The reason is that most folks facing homelessness don’t become homeless.  Looking at data, only about 5% of households facing a 3 day pay or vacate actually become homeless.  Homeless Prevention programs can cut that in half to just 2.5%, but at huge costs.  You end up assisting 100 families to prevent 2 or 3 from becoming homeless.  Those families all needed that financial support, but if you are trying to address homelessness, there are more efficient ways to do this.  If you assist 100 families with Rapid Rehousing, you are helping 70 families exit homelessness.  With homeless prevention, that same money would only stop 2 or 3 households from becoming homeless.  That data is from a couple different national studies, but since we do so little homeless prevention locally, I don’t have local data available for the calculation. So, an efficient system should have little to no homeless prevention while households are still literally homeless. 
  • The last control is the Rental Vacancy Rate slider.  It models the relationship between the rental vacancy rate and homelessness.  Looking at 75 metropolitan areas that the rental vacancy rate is tracked for, and then comparing it to the per capita homeless rate in those communities, I’ve identified a relationship between the rental vacancy rate and the per capita homeless rate.  I am arguing that it is causal.  When there isn’t enough housing stock, housing costs rise dramatically, causing homelessness.  If we could increase housing stock and increase the vacancy rate, we could slow the flow of households into homelessness.  Homelessness is very low in communities with over 7% rental vacancy rates.  Economists identify 6% or 7% vacancy rates as the level that sees only increases in rent related to inflation.  When you have rates lower than that, raw economics of limited supply force prices up.  That causes homelessness more than any other factor I’ve been able to associate with homelessness.  Domestic Violence rates, rates of disability, rates of mental illness, rates of substance use – they are all the same in communities with high homelessness as in communities with low homelessness.  It is the rental vacancy rates that correlate with homelessness.  In the model, if you increase the housing vacancy rates, you need far fewer resources to address homelessness.
• Some financial numbers on the right side show the cost of the existing system, and the cost of the system you’ve modeled.  It also shows non-housing related costs associated with homelessness.  These are real number from other communities that have quantified the costs of incarceration, crisis services, emergency rooms, healthcare and the like caused by homelessness.  Again, these aren’t local numbers, but they are none-the-less very real.  I was careful to use number from research conducted in communities most similar to Tacoma.  While it may never be possible to pay for an increased homeless response with funds we will save in the court system, it is none-the-less real savings to the community. 
• Lots of folks asked questions, and instead of taking notes of the questions, I was answering them.  I thought I was recording the discussion on my phone, but I screwed that up, so I don’t have anything other than my sieve like memory to go on for these questions.  So I interlaced what I remembered of the questions into my narrative. 
• One interesting question was, “what do we do with this?”   I think the answer is a bit complicated, but it is certainly to continue our push to increase housing stock.  That is what increases the vacancy rate, which will drive rental prices down and reduce homelessness.  And, while I don’t think everyone in the room was convinced, it really is only necessary to add housing stock to the market.  It doesn’t need to necessarily be low income housing to have a positive effect on folks with low income.  The data shows it is really a simple lack of housing stock, not at a specific pricing level, that will improve rental rates for everyone.  That said, I’ve not done adequate research on this topic, so I may have more insight to offer in the future.  We also need to push for adequate funding for Rapid Rehousing and Permanent Supportive Housing.  These are the industry standard, best practices that best address homelessness. 
• Please provide any feedback directly to me (gerritn@ccsww.org).  I’d love to hear what is confusing, doesn’t seem right or could be improved.  (He already think it’s amazing, so you don’t need to tell him that. –ed)
• Some questions I’ve received have to do with how this model takes into account race and discriminatory housing policies.  It doesn’t.  All that discrimination is real – both past and present (people really suck, sometimes) – but this model is not nearly complex enough to account for that level of detail.  And it is complex enough as it is.  Trust me.