April 2, 2014 Brian Kelsey

Occupancy Limits

The Austin City Council voted last month to lower occupancy limits, the maximum number of unrelated people who can share a home, from six people to four people. The new four-person limit is temporary (two years) and will be applied only to newly permitted single-family and duplex properties in greater central Austin, which is roughly bounded by Highway 183, MoPac, and William Cannon (i.e. McMansion Overlay), while the city works on updating the land development code. The ordinance change was largely a response to central neighborhoods unhappy about the proliferation of stealth dorms, which refer to high-occupancy, single-family zoned properties designed for housing, presumably, more than four unrelated people.

Austin is struggling to come to grips with declining affordability as population growth, increasing wealth, and demand for central-city living put upward pressure on home prices. Austin’s priciest neighborhoods are centrally located, and still a bargain for people of means relocating from places like San Francisco, Los Angeles, and New York–or from Dallas or Houston or other places in Texas, where the majority of transplants to Austin originate. Given no apparent signs of a slowdown yet, the proposed change in occupancy limits sparked a healthy debate about housing affordability–i.e. to what extent would lower occupancy limits aimed at stealth dorms result in the proverbial unintended consequences that could make Austin even less affordable, especially for students and low-income workers.

Council Member Bill Spelman lamented the lack of data that could be consulted on the relationship between occupancy limits and housing affordability during the public hearing. A few central neighborhoods in proximity to UT-Austin contributed inventories of perceived stealth dorms, but that work was limited to only a couple of zip codes, with no independent verification of accuracy. Moreover, no such work was done for other neighborhoods located within the McMansion Overlay that would be impacted by this change. Far be it from me to suggest that central neighborhoods ever drive citywide policy in Austin, Texas, but that was among the concerns from critics of the change. On the other hand, no serious person could drive around the most impacted neighborhoods, such as parts of Hyde Park and the North Loop area, and not recognize that some high-occupancy properties were creating problems that needed to be addressed.

The question, therefore, was how to do it–a targeted policy limited to central neighborhoods reporting the greatest impact, or something more far-reaching that, according to supporters of the change, would “protect” neighborhoods across Austin from ever having to deal with the problem of stealth dorms in the first place.

The Austin Board of REALTORS convened a small group of us to see what we could do in terms of bringing a more data-driven perspective to the debate. We got about four weeks to complete a “study” of the relationship between occupancy limits and housing affordability in Austin. Since four weeks is nowhere close to enough time to finish anything that could remotely qualify as a “study” of such a complex issue, we collected and analyzed whatever data we could before City Council was scheduled to vote on the new occupancy limits on March 20. We waited until Friday, March 14, for the last of the data to come in and then had approximately 72 hours to complete a report in time to be reviewed by City Council in preparation for the March 20 meeting. You can download a copy here:

Affordability Impacts of Proposed Changes to Occupancy Limits in Austin

The report fell short in several areas, for lack of sufficient time, no peer review, and no housing economist to guide the work, for that matter. But I hope the report can serve as a starting point for continuing the effort under CodeNEXT. And in the interest of teeing up that future work, here are a few points that we didn’t have time to fully articulate in the report. They all warrant much more public discussion than they received on March 20.

1. The stealth dorm apocalypse appears to be a ways off, even in the most severely impacted neighborhoods. We identified 1,796 possible single-family zoned high-occupancy properties with five or more unrelated occupants in Austin. These properties represent approximately 0.5% of all households in the city. The most severely impacted area is 78751 (Hyde Park/North Loop), where high-occupancy properties make up 1.6% of total housing units. In other words, a significant change, albeit on a temporary basis (for now), to Austin’s housing policy was just passed based largely on the stated preferences of a few central neighborhoods where high-occupancy properties make up less than one out of every fifty homes. One out of fifty. We’re not exactly entering Walking Dead territory here.

2. Many of the neighborhoods with the greatest number of high-occupancy properties don’t seem to be at the table for this conversation. We identified 100 or more possible single-family zoned high-occupancy properties in 78702, 78751, 78745, and 78723 (78748 is not far behind with 99, but outside the McMansion Overlay).

High-Occ Map Zoomed_Revised2

More time for a real study would have provided useful perspective on how high-occupancy properties are impacting communities in 78723, 78702, and 78745, where, presumably, UT students are not the primary drivers of demand for stealth dorms. At the very least, more time would have allowed for targeted outreach to invite new stakeholders to participate, now that we know where high-occupancy properties are most prevalent.

3. And what about those well-heeled UT students who can afford the reported $1,000 per bedroom rents charged in the stealth dorms taking over neighborhoods near campus? Well it turns out that, with the exception of 78705, very few stealth dorms are inhabited by UT students. According to student address records provided by UT for the study, nearly every high-occupancy property we identified in 78705 is occupied by UT students. In fact, our estimate in the report of 64 possible high-occupancy properties in 78705 appears to be low, after reviewing the student address file that arrived too late to make it into the report.

By contrast, only three high-occupancy properties in 78751 were found in the student addresses, which means that non-UT students occupy 98% of high-occupancy properties, or stealth dorms if you prefer, in the portions of Hyde Park, Northfield, and other neighborhoods in 78751. Further, no high-occupancy properties with five or more UT students living there could be found in the three other 100+ zip codes, 78702, 78745, and 78723.

Now, a major qualifier here is that approximately 50% of students don’t have a local address on file with UT, according to the data we received. In addition, future researchers should request student address records from St. Edward’s, Huston-Tillotson, and other colleges in Austin to complete this picture. However, our preliminary analysis indicates that, outside of 78705, the “stealth dorm issue” is more a question of housing availability for workers and other non-student residents than it is about student housing.

4. Finally, while the grandfathering clause was indeed a necessary provision–the new four-person limit will apply only to new construction–it does nothing to settle many unanswered questions that need to be addressed. We are no closer to understanding the root cause of why the market is responding with high-occupancy properties in the first place. Stealth dorm opponents argue that profit-driven developers are tearing down “affordable” homes that would otherwise be available to families with children. With most central neighborhoods now well north of $200 per sq ft and therefore well out of reach for most families, this argument seems a bit disingenuous when made by serious people. On the other side of the debate, the report doesn’t offer much ammunition for people advocating greater population density in the form of higher occupancy limits in central neighborhoods, either.

As the report indicates, all we can say with a reasonable degree of certainty is that high-occupancy properties appear to be more prevalent in lower-income areas of Austin experiencing rising average rental rates, and that relationship holds even when excluding neighborhoods with the greatest number of students.

And what of the new construction that will fall under the temporary, four-person limit during the next two years? In 2013, 2,573 new single-family homes and 101 new duplexes were permitted in Austin, according to data submitted to the Census Bureau. A large portion of those new housing units are likely to be in the McMansion Overlay and subject to the four-person occupancy limit–that’s more than 5,000 housing units that may not be available to students and low-income workers who need more than three roommates to afford rent, especially in central neighborhoods with convenient access to public transportation.

Seems like that would have generated more discussion, or at least six weeks for something more resembling a proper study, in a city that professes to be so concerned with declining affordability.

Tagged: ,

Comments (11)

  1. John Boardman

    Brian, thank you for your careful analysis of the data. One issue that I don’t see addressed is the rate of growth and a projection into, say, 2024. For example in the last few years I’ve had a block of houses in my neighborhood (30th Street in Heritage) flip to be almost entirely stealth dorms, and two more houses in that section got bought by stealth dorm developers in the last month. Even more dramatically Northfield has seen block after block turn into stealth dorms the last few years.

    The issue isn’t just the current state, it’s looking at the YOY growth and projecting out 10 years from now. That’s why the occupancy reduction was over a broad area: there’s no reason to think it will be constrained to 78705 in 10 years.

  2. John, that’s a great point. We didn’t have time (or the expertise, frankly) to get into forecasting in the report, but you are right on that it should be addressed in future analysis of this issue. You could make some good progress just using the City’s permit data, but at some point I hope a housing economist can lend some time to this effort.

  3. Pamela Bell

    I am surprised that the issue of property taxes was not addressed in determining affordability. TCAD has a different way of valuing structures of rental properties/”investment” properties. To use the words of their staff, comparing “investment” property values with those of properties owned by people who live in and are going to take care of their properties for a long time (aren’t these people investors too?) is like comparing apples and oranges. TCAD gives SIGNIFICANT discounts to “investment” properties including stealth dorms, while homesteaders (owners who live in their homes) pay SIGNIFICANTLY higher property taxes. Why is that? The discounts occur in the values assigned to the structures. TCAD will not reveal the method for determining structural values. These savings on property taxes are not being factored into the affordability equation; I doubt such savings are being passed on to renters.

  4. Brian:

    Thanks for the work on study. As obvious to all, it was a very preliminary look into the numbers on this issue.

    You’ve taken the liberty to arrive at a few conclusions, based on a certain point of view.

    I’d like to take the opportunity to show how the same information you’ve highlighted, but from a different viewpoint.

    1. The percentages presented are based on total number of units. However, the ordinance only dealt with single family zoned areas and this is where the problem resides. Based on the numbers compiled by the Northfield neighborhood, we are experiencing a 7% rate of high occupancy use. This is 14 times the city average. Even using your number of 1.6% for 78751, it shows that 78751 is more than three times higher than the city average. When only single family units are taken into account, this number is even higher as there are very few high occupancy units in multi-family properties.

    The 7% rate of high occupancy units in Northfield is also more than double the threshold of 3% that is allowed by the short term rental ordinance. This is important and relevant benchmark because high occupancy units create many of the same issues that are caused by short term rentals among others.

    Taken together, the high occupancy and STR uses comprise about 10% of our neighborhood.

    So while I agree that we are not at “walking dead” territory. However, I think that if you went into any of the neighborhoods not yet experiencing this issue and told them that we are going to set aside 10% of your single family zoned lots and assign to them non-single family use, that you would get a very swift and significant outcry against that policy.

    Or looking at it from a RECA or ABOR view, if you told them that 10% of all commercially zoned property was going to be rezoned to single family use, you would get a similar or even more outraged response. The difference here is that if even a very small amount (let’s say 0.5%) of commercial property was rezoned in this fashion it would not only almost immediately be met with not only significant outcry at the CIty, County, and State level, but in all likelihood there would be almost instantaneous litigation that would accompany that lobby.

    2. Many proponents of the ordinance supported city-wide application of it, but the McMansion boundary was the compromise that was made at council.

    This issue has always been about high occupancy units, not solely students. The same issue would be occurring if we lived near a major agricultural area, maquiladora, or any other major source of low income households.

    Students living in lower occupancy units have always lived in harmony in our neighborhoods. This only became a problem when high occupancy units proliferated, especially starting in the early 2000s.

    That point aside, I would think the neighborhoods would be more likely to be right that students do live in the majority of central neighborhoods high occupancy units as they interact with those properties on a daily basis.

    3.Given that UT has about 50,000 students, the half of them with non local addresses is about 25,000. Compared that to 139 high occupancy units that your data reports in 78751 (or 834 occupants at six each), it means that just over 3% would have to be in high occupancy units for every single one of the reported units to be student occupants.

    So I’d have to disagree about your 98% conclusion –

  5. 4. For the number of 5,000 units to be cited, how many would would be built in the McMansion area, and how many of those would be built in a fashion to support high occupancy? Then how many of those units would lie within a 1/4 mile of public transit?

    The net effect on the housing market would then be tempered by the fact that any truly affected units would still be allowed an occupancy of four, which is still above the national average of occupancy and significantly above the average in the State of Texas.

    When you take this “truly affected” number and compare it to the 330,000 units that have been grandfathered, the net effect on the total Austin housing market is small.

    Additionally, it is more than plausible that the existing 330,000 units would be able to fulfill any new high occupancy needs for quite some time.

    However, since the high occupancy properties are concentrated in a few neighborhoods, the negative effects of high occupancy units on those neighborhoods becomes much more significant.

    As far as out of reach for families, many of the homes that are being torn down are much smaller than the suburban counterparts, so a 800-1200 sq. ft. home at $200 per sq.ft. is much closer to the Austin average home price and at least has a possibility of being “affordable.” Based on the data of your study, the high occupancy units being brought to market much larger and have higher per sq.ft., so are definitely unaffordable.

    Last of all, regarding “this argument seems a bit disingenuous when made by serious people.” – I am making this argument along with many of the occupancy reduction proponents. While I’m typically not that serious of a guy (please take this last comment in a somewhat comedic view), I do feel that I’m pretty genuine in my concern about this issue. There’s probably very little data in this study or any other that would support whether people are “disingenuous” or “serious.” Good luck in writing a reasonable methodology to determine the flavor of those adjectives. In fact, it MIGHT be disingenuous for a serious person to make that assertion.

    Again, I very much appreciate your work on this issue. It took a lot of courage to volunteer to work on a study on a controversial issue that had not much chance of ending in anything except further questions. I look forward to any follow-on work that arise.

  6. What the study says is that only 2% of HOUs in 78751 are occupied by UT students. Other colleges aren’t included but are far less likely to have residents living in 78751 for obvious reasons that it’s not close to their campuses. And unless there’s some logical reason why UT students living in HOUs in Northfield are substantially LESS likely to register their local address than are other UT students, the maximum reasonable conclusion you could make from this data is that around 4% of HOUs in 78751 are likely occupied by UT students.

  7. Brennan Griffin

    In response to Mike above:

    1) We need more density at the core. I think upzoning a lot more of the central city would be an excellent idea. Renters are people, too, as are students! If you’re concerned about noise and trash, lets beef up code enforcement – maybe a pro-active regime in areas with high concentrations of these types of housing, rather than reactive. Neighborhood associations reflexively attacking all attempts at density have in part caused more and more people to pack into single family homes in core areas.

    2) This sounds like an argument about all multi-family all the time, and is pretty ugly towards low-income people in general. It sounds like you’re saying poor people shouldn’t be allowed to live in multi-family, high density housing anywhere. Or maybe you’re just saying that high density housing for poor people should only be around poor people, aka lets just recreate Cabrini Green! Or are you saying it all should be low density everywhere, in which case I guess we’re just going to sprawl out to Houston?

    3) Where do you get the idea that 3% of the 25,000 students without addresses would have to be in stealth dorms in 78751? Why would you assume that? That’s just weird, when of the 25,000 addresses he does have, a very rough estimate would give you that 1.6% live in high occupancy housing throughout the city, mostly in 78705…You have to both double the percentage living in this kind of housing, and then assume that they’re all in 78751.

    4) Agreed that net effect of the final compromise may be small on the housing market, but it is an effect, and its an effect in the wrong direction – pushing sprawl, taking away affordable options. If its negatively affecting specific neighborhoods with trash and noise, then lets come up with a targeted solution for those neighborhoods addressing trash and noise complaints – pro-active code compliance for neighborhoods with high densities of this kind of housing, perhaps.

    In general, I’d rather just upzone more central city neighborhoods, or at least the major corridors like 45th, 38th, Duval, etc, reduce parking minimums, allow garage apartments and accessory apartments, and relieve some of the pressure this way. But in the meantime, we have to preserve the affordable options we have, or we’re just recreating the San Francisco problem – only rich people can afford the central city, and everyone else commutes an hour or two each way.

  8. paul

    “As the report indicates, all we can say with a reasonable degree of certainty is that high-occupancy properties appear to be more prevalent in lower-income areas of Austin experiencing rising average rental rates…” Common sense dicates that you have to connect the occupancy limit to SQUARE FOOTAGE…a one-thousand square foot duplex unit [regardless of year built] simply should not have six people living in it (related or un-related, illegals or citizens), period.

  9. Monique

    Interesting article. We are dealing with occupancy issues in Boulder, Colorado, also a university town with rising cost of housing. The link provided for this report no longer works. “Affordability Impacts of Proposed Changes to Occupancy Limits in Austin” — is there somewhere else to download it?

Leave a Reply

Your email address will not be published. Required fields are marked *