The following post was written by Oz Erickson and has been edited slightly for clarity. Oz is the Chairman of Emerald Fund, one of SFHAC’s founding members, and was SFHAC’s 2013 Housing Hero. The post explains the implications of the June 2016 Prop C ballot measure that raised San Francisco’s inclusionary housing requirements, 25% for the on-site option. For a rental project, this means 15% at 55% Area Median Income (AMI) and 10% at 100% AMI. Using a scenario project, he explains how nearly doubling the inclusionary housing requirements impacts rents and the financial viability of building housing. Emerald Fund does not have any projects in the pipeline that would be impacted by the legislation. Please refer to this spreadsheet throughout the post.
The current 25% affordable requirement doesn’t work. Absent increases in density/height, it is impossible to build life-safety rental projects in the city with the current 25% affordable requirement. Page 1 shows the current costs for building a 120’ tall, 162- unit building in SF. All the costs, rents and expenses are actual numbers with the exception of the construction costs which are budgeted at $420,000 a door, very real costs for a life-safety project built today with union workers.
If you look at the first spread sheet entitled “Base Case”, you will note that the yield on equity (53% of total cost) on this investment is 1.11%. There is not a chance of obtaining this equity anywhere absent very, very unusual circumstances (e.g. owning 18,000 other housing sites in San Francisco). Who would be willing to invest their money in an illiquid asset and earn 1%? Any pension fund administrator prudently investing pension funds under this scenario would be fired.
The second spread sheet shows the same project with identical costs except the land is free. The amount of equity goes down to 41% of total costs, but the return on that equity still stays at an anemic 1.79%. Maybe there are investors somewhere who would sign up for this deal, but if there are, they are few and far between.
Finally, the third spread sheet shows a financeable project. The return on costs is 5.5%, a total project return that is virtually a minimum requirement for most lenders and investors today. One could build such a project. The only problem with this project is that it requires a housing subsidy from the City of $226,000 per unit, and by per unit, we are talking about all 162 units, not just the affordable units.
The current 25% affordable requirement, once it fully kicks in, will stop the production of market-rate housing. In stopping this production with all the labor distress that will cause, it will also have two additional major negative effects. First, the most obvious one, is that no longer will the historically 12% affordable units delivered with each market-rate project be delivered. 25% of nothing is nothing; 12% of something is something.
Much more insidious is the effect that the elimination of market-rate housing will have on rent-controlled units. Over the last 4 years the very well-designed Better Neighborhoods Plans kicked in. In 2013, 2,000 net units were added; 2014-3,500; 2015- 3,000 and 2016 somewhere around 4,500 units. This increased supply has reduced rents. Over the last 14 months, rents have gone down around 6%. You ask, “But how on earth does this decline effect rent-control units?” Let’s use a simple model. Let’s say a two-bedroom unit rented in 2015 for $5,000. In 2016, let’s say that a similar unit rents for $4,700, a 6% decline. How does this decline effect a rent-control unit? Imagine a two-bedroom, two-bath rent-control, vacancy decontrol unit coming to market in 2015. It is older and not as nice as the brand-new unit, so let’s say the maximum value for this unit is $4,000, 80% of the new market-rate unit. In 2016, a similar rent-controlled unit comes on the market. At 80% of $4,700, it rents for $3,760. In 2017, let’s say that because of the delivery of 4,000 new market rate units, the rent-control unit that would have rented for $4,000 in 2015 is now down to $3,534 (compared to $4,418 for a market-rate unit).
Not a significant change, you say. However, let’s work through the numbers. The minute the 25% affordable requirement hits the market place (say 2019), construction of market-rate units stops cold. With no supply, the existing market-rate units go back to steady state 6.5% compounded rental growth, the historical growth rate in San Francisco. In 10 years, 2029, the market rate unit is at $6,633 while the $3,534 rent-control unit, limited to no more than 2% annual increase but starting in 2017, is at $4,482. If there had been no market-rate development and rents had never gone down due to the extra supply of market-rate units, that rent-control unit that rented for $4,000 in 2015 would be renting for $5,278 per month, $796 more, almost 18% higher.
With 172,000 rent-control units in San Francisco, probably around 15,000 units roll every year. When market-rate unit rents go down, the rents on these vacancy decontrol units go down as well. When market-rent unit rents go up, the rent on these vacancy decontrol units go up, but they can only go up by a maximum of 2%. At $796 savings per month on 15,000 rent-control units, the savings to the tenants is an astounding $143 million per year.
In conclusion, the current 25% affordable requirement is impossible to build to. If this requirement is not changed, not only will the well-paid, blue-collar union construction workers be thrown out of work (in 2010 the unemployment in the construction was over 30%), but the 12% of BMR market-rate units will not be delivered, and finally the lack of supply will dramatically raise rents throughout the city, even on rent-controlled units.
The City Controller will present their final recommendations for the new inclusionary levels to the Board of Supervisors on Tuesday, March 7. Download the report here.
Image: Sharon Mollerus