Friday, February 21, 2020

Portland Metro Supportive Housing Needs: Some Alternative Calculations and Comparison to Current Voucher Counts and Funding.

One of the key inputs into Portland metro consideration of a homeless tax is the need for supportive housing.

A 2019 PSU study on homelessness HERE projected a need for 4,936 supportive housing units. 

I went back and re-calculated that need using the PSU calculation formula but excluding doubled up individuals/students and using the HUD point in time count for 2019. (The PSU report used 2017 PIT data, supplemented with other homeless data).

The PDF file HERE embedded below shows those calculations for three scenarios and includes hyperlinks to the data sources. Note that all scenarios:

  • Provide supportive housing for 90% of the chronically homeless,
  • Provide supportive housing for 10% of all homeless, and 
  • Provide additional units for those those who are returning to supportive housing after previously leaving it. 
Since the count of chronic homeless is readily available from the PIT count the biggest variable in calculating supportive housing need is the total homeless count. 

The lowest estimate for supportive housing need rely on the PIT count as a base while higher need estimates add to the HUD PIT count and expand the universe of who needs rental assistance to include doubled up individuals/students.

(Pages 68-71 of the PSU report summarizes their method of need calculation).

Needed Supportive Unit Count and Percentage Differences with Different Inputs
Using updated 2019 point in time data and excluding just doubled up individuals/students yields a need that is 35% (1,730 units) to 54% (2,642 units) BELOW the 4,936 units found in the PSU report:
  1. Using HUD published 2019 point in time data yielded the smallest need, 2,293 units, 54% below the PSU base case estimate.
  2. Using a Metro February 18, 2020 memo reported 2019 PIT count point table yields a need for 2,535 units, 49% below the PSU base case estimate.
  3. And finally using PSU‘s data but excluding 19,840 doubled up individuals yields a need for 3,206 units, 35% below the PSU base case estimate.
Reduced Projected Need Costs Using Different Inputs
I applied the unit percentage difference from the base PSU case to arrive at the cost reductions for the three different scenarios. This resulted in costs that were $41.7 M to $63.7M less than the PSU base case estimate: 
  1. Using HUD's published 2019 point in time data yields a need for $55.2 M. That's $63.7M below the PSU base case need.
  2. Using a Metro February 18, 2020 memo reported 2019 PIT count yields a need  for $61.1 M, That's $57.8M below the PSU base case need.
  3.  And finally using PSU‘s data , but excluding 19,840 doubled up individuals, yields a need for $77.3M. That's $41.7M below the PSU base case need.

A CSH analysis projected supportive housing needs for 3,123 units. That's still more than projected need in these three alternative scenarios. However if the CSH data was updated to 2019 PIT HUD data I think it is likely that it would be closer to the 2,393 units I projected in my lowest need scenario. 

Funding Remaining for 5,600 to 11,900 Additional Rent Assisted Units if Tax Collection Goal is $175 Million
According to the Metro analysis 5,000 non SH rental assistance units would cost $50M ($10,000 per unit). 

If the tax collection goal is $175 million there would remaining additional amounts available for rental assistance from all of these scenarios. Adding those units to the supportive housing units for these scenarios yields these results:
  • $56M additional, using the PSU base case. (5,601 rent assisted units, total of 10,536 units).
  • $119M additional using the updated 2019 HUD point in time data. (11,971 rent assisted units, a total of 14,264 units).
  • $113M additional using the Metro Feb 18,2020 reported PIT count.(11,387 rent assisted units, a total of 13,923 units).
  • $97M using the PSU base case, less doubled up individuals.(9,771 rent assisted units, a total of 12,977 units).
Comparison to Housing Vouchers in 3 County Area
My estimate is that there are 13,500 voucher units in these three counties as of September 2019. This does NOT include additional VASH vouchers--a key supportive housing tool- that are already authorized but not in use.

Adding 5,600 to 11,900 additional rent assisted units and a total of 10,536 to 14,264 assisted units (adding the supportive housing units) would be a significant boost. 

In FY 2018 HUD obligations for the housing voucher program in these three counties was $137 million so an additional $175 million in annual Metro funding for housing would also be a significant boost.

Originally created and posted on the Oregon Housing Blog.

Wednesday, February 19, 2020

My Projections: 55,000 Metro Filers (6% of All Filers) Would be Subject to the Proposed Metro Homeless Tax; The Majority Would be Outside Multnomah County.

Two of the many unanswered questions about the proposed metro housing ballot measure are
  1. How many filers would have income sufficient to make them subject to the tax?
  2. How would these filers be distributed by county?
I constructed a table and graphs with my projections. I have pasted those below and posted the table and graph as a PDF file HERE.

My projections:

  • 55,000 filers would have incomes above the reported AGI thresholds subject to the tax. (That's 6% of all 884,000 2017 federal filers).
  • 45% of those filers would be in Multnomah county, 22% in Clackamas county, and 33% in Washington county. (So 55% of filers subject to the tax I project would be outside Multnomah county).
How I Calculated
I used 2017 IRS SOI county data that counts the number of actual federal filers by filing status and by several income groupings.

Because income groupings do not exactly match the reported AGI income tax thresholds for the proposed Metro tax I had to make projections of the percentage of filers in each income grouping who fell above the Metro AGI tax thresholds.

So for single filers for the SOI income group $100,000 to $200,000 AGI I assumed 75% of the actual filers would be subject to the tax (since the reported single filer threshold was $125,000). For joint filers with incomes above $200,000 I assumed 50% would be subject to the tax (with incomes above the $250,000 AGI threshold). And for head of household filers I use the same projected 75% distribution as single filers.

Adjusting the projected percentages that I used would of course change the total number of filers subject to the proposed tax. I suspect (but have not tested to see) that this however may not significantly change the county distribution of those filers subject to the tax.

Caveats/ Limitations
While this exercise provides a reasonable projection of the count of filers and their county distribution:

There may be better data sources that use more recent or more refined income groupings.

The actual distribution within income brackets may not be linear. So for example within the $100,000 to $200,000 AGI bracket there may be more or less than 75% of filers with AGI of more than $125,000. Above $200,000 AGI there may be more or less than 50% who have incomes at or above $250,000 AGI. 

Also the distribution of filers within income brackets in individual counties may vary from the uniform percentages I have used.

It does NOT attempt to show how much tax might be collected. Collection goals,tax rates and income exemptions will influence tax collections .  

The distribution of filers subject to the  proposed tax does not match the distribution of the homeless. This raises the prospect that filers in one county may be asked to provide resources for use to fund homeless programs in another county. (This isn’t really different than paying Oregon income tax and having it disproportionately distributed outside the county where it was collected).

Originally created and posted on the Oregon Housing Blog.

Monday, February 17, 2020

Women Benefit More From Homeless Programs that Target Homeless Families VS Programs that Target Single Homeless Persons or the Chronic Homeless.

The Portland metro homeless and supportive housing ballot measure proposal comes before Metro Council this week; the draft resolution is HERE.  (As of Sunday evening exhibits A and B referenced in this resolution were not on line. The agenda HERE may be subsequently updated to include the exhibits).

For now, I'm not clear on the specific populations targeted to benefit, how "homeless"  and  ''success" will be defined, any set asides for specific populations and areas, limits on administrative costs, and the specifics of who will be taxed and at what rate. Governance, audits, and transparency are also issues for me.

In the interim I thought it would be useful to take a look at one aspect of homelessness that has received little attention that I can remember--female homelessness.

I downloaded and organized several gender specific population counts from the 2019 PIT count for all Oregon Continuum of Care organizations. I also grouped the data to focus on the three Portland metro CoC organizations.

The graph and the table pasted below summarize my initial observations.
1. Women were a minority of ALL homeless persons. For the 3 county Portland metro area women were only 35% of all homeless persons.
2. Women are also a minority of all homeless who lived as individuals, For the 3 county Portland metro area women were only 31% of all homeless persons who lived as individuals.
3. In contrast women were a MAJORITY of persons who lived in families.For the 3 county Portland metro area women were 62% of all homeless persons in families.

Chronic Homeless Gender Information More Difficult to Find, But Women Clearly the Minority.
The HUD 2019 CoC downloadable data does not include gender for the chronic homeless. However on the 2019 PIT Dashboard for the Home for Everyone/Portland CoC there were 1,729 chronic homeless people counted including 572 women. So, women were 32% of all chronically homeless people.
The Clackamas county CoC reported that 32% of its chronic homeless population were women. 
I was not able to locate similar data for Washington County.

Bottom Line: 
Programs that target homeless individuals or the chronic homeless will primarily benefit men. Programs that target homeless families will provide more benefit to women.

Originally created and posted on the Oregon Housing Blog,

Friday, February 14, 2020

FY2018 Federal Homeless Grant Obligations, 3 County Portland and Oregon.

It is more difficult than it should be to track federal homeless grant funding on a local basis. Given recent attention being paid to homeless housing and supportive housing needs I thought it might be instruction  to START that process. 

So from USA Spending I extracted FY 2018 data on Oregon federal obligations for these 9 grant programs, including programs from HUD, VA, HHS, DOL, and the Education Department:

NOT Yet Included:
Note that this does NOT include 
  1. Any HUD CDBG, HOME,Housing Trust fund, public housing, or voucher spending for homeless programs, including the 2,000+ unit HUD VASH program targeted specifically for homeless veterans. 
  2. Any direct obligations to or on behalf of homeless persons. (SSI, VA disability, etc.)
  3. Any homeless contract obligations using other federal funding sources. 
  4. State and local spending that does not use a federal source. 
The Excel workbook  I constructed HERE, and embedded below shows for ONE fiscal year (2018).
  • Total 3 county Portland metro area homeless grant obligations of $76 million.
  • Total Oregon homeless grant obligations of $111 million.
The workbook includes pivot tables and data tables that include the names of grantees; the default (opening) view is of the 3 county Portland metro area pivot table.

Originally created and posted on the Oregon Housing Blog.

Tuesday, February 11, 2020

4th Qtr 2019: City of Portland Had 75% of Portland Metro Multifamily Units In Development Pipeline.

Using the highly recommended Berkadia apartment construction pipeline reports as a source I was able to take a closer look at the pipeline of Portland metro multifamily development for CY 2019.

The first graph I constructed below shows by quarter that the total count of MF units in development remained largely consistent during the year, with total units close to the 27,000 to 28,000 unit mark, and a range of 7,200 to 8,400 units either in construction or in lease up. 

The second table focuses in on the city location of MF Portland metro development for the 4th quarter of CY 2019. It shows that the City of Portland was capturing between 70% to 84% of MF development in various phases with an overall capture rate of 75%. 

Finally, the last table shows, for the 4th quarter, the neighborhood location of the City of Portland 20,934 multifamily units in development by phase. 

The Kerns/Buckman neighborhood (3,312 units total) had the largest share of City of Portland units under construction (22%/902 units) or in lease up (33%/548 units). Southwest Hills (4,133 units total) had the largest share of prospective (25%/1,622 units) or planned (29%/2,511 units) MF development.

Originally created and posted on the Oregon Housing Blog.