UCLA Housing Voice

Ep 79: Who Pays For Inclusionary Zoning with Shane Phillips

• UCLA Lewis Center for Regional Policy Studies • Season 4 • Episode 1

Share episode

Inclusionary zoning policies use the market to produce affordable housing, but nothing comes for free. So who pays? Shane takes the guest seat to discuss his analysis of IZ in Los Angeles, making the case that it’s not developers or high-income renters who bear the cost, but all renters — poor, middle income, and wealthy alike.

Show notes:

Shane Phillips  0:04   
Hello! This is the UCLA Housing Voice Podcast, and I'm your host, Shane Phillips. 

We're back for Episode 79, the first episode of our fourth season. 

Seasons aside, this episode has two other firsts. It's the first interview where I'm not hosting, and also the first where I appear as a guest. That sounds impossible after almost three and a half years, but I'm pretty sure it's true because I don't recall ever being more stressed out ahead of a recording, and that seems like the type of thing I would remember. 

We're doing a short run of episodes on inclusionary zoning policies, and this one features my own research on the subject. It includes a very cool development simulator built by our friends to the north at the Terner Center and Terner Labs, and it really homes in on the question of who pays for IZ. Inclusionary zoning mandates or incentivizes affordable units in market rate developments, and my guess is most people assume it's either developers or landowners who pay in the form of reduced profits, or it's the relatively affluent renters who live in the market rate units of those mixed income projects. Or maybe it's both. I don't think that's right. As I see it, the price is not paid just by the renters in these IZ projects, but by virtually all renters, from the upper income to the very, very low. I make that case in a report published earlier this year, and in this conversation with Mike Lens. 

I do want to thank the Turner Center and Labs for their support, including a bunch of very helpful guidance and feedback from Carolina Reid, Issi Romem, Alex Casey, David Garcia, Ben Metcalf, and Sarah Karlinsky. This project had to be rethought and reworked more than once, but with your input, I'm really happy with where it ended up. 

Only a little bit coincidentally, this episode is being released on October 16th, the same day that the Lewis Center and Terner Center are jointly accepting a Housing Heroes Award from the Housing Action Coalition. It worked out nicely that today we could also share an episode from this jointly published report. We are honored to receive the recognition, and it is absolutely a tribute to the great work done by all of our center's staff, faculty, students, and affiliates. Nice work, y'all. 

The Housing Voice podcast is a production of the UCLA Lewis Center for Regional Policy Studies with production support from Claudia Bustamante and Gavin Carlson. As always, you can reach me at shanephillips@ucla.edu. With that, let's get to my conversation with Mike Lens, or rather his conversation with me.  

Mike Lens  2:46   
This is Mike Lens, the guest host of the UCLA Housing Voice podcast today, and I have the distinct pleasure of introducing our normal host, Shane Phillips. Shane Phillips needs no introduction, of course, to our listeners. He's the host of the show, but Shane does not often provide his bio here, so I will do it. Shane is the manager of the housing initiative at the UCLA Lewis Center for Regional Policy Studies, where I work closely with him. He has taught public policy as an adjunct instructor at the University of Southern California. Shane previously worked as the director of public policy for the Central City Association, a downtown LA advocacy organization, and Shane is the author of numerous reports and papers on housing policy, one of which we will talk about today. And he is also the author of the 2020 book, The Affordable City. Shane, how did I do? Did I miss anything?  

Shane Phillips  3:45   
I guess maybe the only thing you missed was all the totally irrelevant things I did before, like working in a tuberculosis research lab as a cable installer at Comcast, as a package loader and sorter at UPS, but maybe those aren't so important in this case.  

Mike Lens  4:00   
That is an amazingly deep reservoir of experience you have, and I certainly would not consider any of those things irrelevant. Tuberculosis sounds more important than even housing, potentially. As always, Shane, we will start with a tour. You have heard something like 77 tours of the world's favorite places of our guests. Are there any left? If you were giving us a tour around your hometown or a place you've lived and want to show off, where would you take us?  

Shane Phillips  4:30   
Yeah, I didn't really know what to do here. I now understand the pressure our guests feel when they have to answer this question. I could have talked about Seattle, but I'll just talk about LA and some very local places that I really value that I feel like we haven't really heard much about. So I live in Lincoln Heights, which is just a couple miles from downtown LA up by Chinatown, and the LA River runs right by it. And from about Chinatown up north 10 miles or so is a really nice section of the river with a bike and a walking path all along it. Fairly wide, nice for riding fast, riding slow, walking, rollerblading, whatever you're into. There's coffee shops and various places to stop along the way. So that's one of my favorite places just to kind of go outside in the city and not be amongst cars. And I think a lot of people share the love of the river, especially along that section. Sort of in the opposite direction for my house is the LA State Historic Park, which opened sometime since I've lived here in LA the last 11 years, it's probably been open six or seven years or so. I think there's a whole saga of, you know, decades of actually getting that thing built. But it's become a really great resource for kind of the wider community, not just as a daily stopping ground for a lot of people, but also, it hosts a lot of events on weekends, it has shows just almost every weekend during the summer. And I think it's just one of my favorite places in LA as well. When people come to LA from other places, I almost always take them to Sonoratown. So just to pick a food place, this is a taco place in downtown LA. And so I'm just gonna plug Sonoratown and we can move on into the research.  

Mike Lens  6:10   
That’s all super cool. And the something that I don't know, if listeners know about you, is, you do not own a car.   

Shane Phillips 
No, I do now.   

Mike Lens 
You do now?   

Shane Phillips  6:21  
 Yeah, I'm ashamed to admit it. It's been several years.  

Mike Lens  6:25   
Oh, you had several years in which you lived in Los Angeles without a car?  

Shane Phillips  6:29   
No, no, no. I lived here for maybe eight years or so without a car. Right. So it's only been the last three or so two or three that I've owned a car myself. Yeah, it was partly because of the job at UCLA. Yeah, actually, when I got the job at UCLA, I still did not have a car. And I was biking on my electric bike about 18 miles each way for about six months every day. And it was not super tiring day to day, but just like that's a long time to be biking. And it was all on local streets for the most part, except for one little part along the LA River. And then COVID hit, worked at home pretty much nonstop for a couple years. And then we were going into the office here and there, you know, once or twice a week. And at that point, I had lost all comfort biking that kind of distance and in that kind of environment. And I didn't really get comfortable again, because it was so infrequent. Yeah, and tried taking the bus as I had sometimes done train and bus sort of a combo to get to UCLA, that would sometimes take two hours or more one way. And I was just done with that. So I ended up buying a car, have a nice little Nissan Leaf that I drive into work one day a week and maybe one other time, the rest of the week and the rest of the time, it just sits in the parking lot.  

Mike Lens  7:46   
It's like an electric bike, but for your ankles.  

Shane Phillips  7:50   
Much bigger and expensive too. Yes. Not as good for the environment.  

Mike Lens  7:54   
Okay. On to our research done by Shane. So our conversation today is about an analysis Shane did earlier this year. It's titled Modeling Inclusionary Zoning's Impact on Housing Production in Los Angeles, Trade-offs and Policy Implications, which was the product of a collaboration between the Lewis Center, our esteem center and our good friends up at Berkeley at the Turner Center. To summarize really quickly, Shane explored how changes to the transit oriented community program in Los Angeles could affect housing production and the possible consequences for rents and affordability in the city. TOC, the Transit Oriented Communities program, lets developers build taller and denser in exchange for making some units affordable to lower income residents. And so the question was, what happens if you ratchet the affordability requirement up or down, but leave everything else the same? He did this analysis using something called the Terner Housing Policy Simulator, which he will introduce in a minute. He found some very big trade-offs between lower income housing production and overall housing supply. But what I think is most interesting and important about this report is how he connects this to rents in the private rental market. Inclusionary zoning is often framed as a policy whose costs are paid by higher income tenants in newer market rate housing or better yet, by developers. But Shane argues pretty persuasively that the costs are also borne by lower and middle income renters. Because of that, the benefits to the small number of renters who get a spot in the affordable units produced by inclusionary zoning may be outweighed by the costs to the much larger number of renters who do not. I think this analysis helps to clarify who pays for inclusionary zoning and who benefits, and it offers a new way of thinking about how those costs and benefits measure up to each other. Shane, let's start here. You've written about LA's TOC program and density bonuses before, and in your book, The Affordable City, you offer some guidance on how to minimize unintended consequences from inclusionary zoning, but you hadn't done any empirical research on the effects of IZ. Why did you decide to dive into the subject now? And since you're not the first person to study inclusionary zoning, what does previous research tell us about its effects on housing production and affordability? 

Shane Phillips  10:13   
Yeah, so I think the genesis of this project was in some ways our conversation, which I think it was with Mike Manville as the guest host, with Emily Hamilton on inclusionary zoning. I think that's episode 31 of the podcast, so this was more than a year ago. But that conversation and her research really stuck in my brain for a long time. And specifically what interested me was how she found that inclusionary zoning in the Washington DC area was associated with faster price growth. The cost per square foot of housing went up faster in places that had inclusionary zoning, but she didn't find an effect on housing production, which you would expect because if inclusionary zoning is increasing prices, how is it doing so if not by reducing production? It doesn't really make sense that developers could just increase the price of all the new housing they're building, and people would just willingly pay that. Sort of arbitrarily increasing prices doesn't usually work. And so I got really interested in thinking through what other explanations there could be for this. And where I started this project was in a totally different place, thinking about how maybe there's something going on where renters who are able to get into these affordable units might be people who otherwise would have had to leave the city or chosen to leave the city because they just couldn't afford to stay. And so somehow that's increasing overall demand for housing in the city and pushing all the prices up. Maybe it has something to do with the fact that even if production is not falling, you are trading away some market rate units for below market rate units. And because rent increases are probably driven by the availability of market housing, for the most part, maybe that smaller supply in that segment of the housing stock is contributing to faster rent growth. All of these were, I don't know, plausible to me, but I had no real way of researching them. And so I started thinking about other ways of looking at this. And other research is also kind of mixed on inclusionary zoning. Some of it finds that production goes down, but prices aren't really changed. Others find that prices go up more quickly, but production doesn't change. Some find really no effect. And so I just felt like more could be done in this space, but because I'm not really an empiricist, I guess, primarily, it's not my main expertise, I think I'm a little better at just thinking through how this stuff works and trying to explain how the costs and benefits are allocated and just the processes by which housing markets operate. That was kind of how I approached this project. The reason I framed this, there's a couple of reasons I framed this as what happens if we increase or decrease inclusionary zoning requirements without changing anything else. One reason is just that the Turner Housing Policy Simulator, which we'll talk about, made it very easy to do that and not so easy to do other kinds of analysis. But I think more importantly, I had in mind the regional housing needs assessment or allocation policy in California and this new cycle of planning that California cities are in, and specifically in Los Angeles, where we have this goal of building 456,000 housing units over eight years, when we built about fewer than 150,000 over the previous decade. And the state, through its policy, has this, frankly, totally unrealistic goal that 40% of that 456,000 units will be affordable to lower and very low income households. And I think that is probably putting a little pressure on local officials to maybe ramp up existing inclusionary requirements to meet those goals, because usually they tend to be around 10, 15%. And even if you built the 456,000 units, which to be clear, we're not going to do, but even if you were to do that, if you only had a 15% inclusionary, then you're still going to fall very far short of your low and very low income housing goal.  

Mike Lens  14:23   
And I think that does dovetail into differentiating between some of the goals of inclusionary zoning and the related programs, particularly TOC and density bonus that are at play in your analysis in particular. So a lot of times people equate IZ with affordability requirements. LA's TOC program does work to get below market units into market rate projects, but it's voluntary and developers who provide enough affordable units also receive bonuses to help offset the cost. So should we be making a distinction between inclusionary zoning and density bonus programs, or are they operating kind of one in the same?  

Shane Phillips  15:09   
I've had this question a few times and I don't really know the answer for sure. I can see arguments on both sides. Calling the transit oriented communities program and inclusionary policy might muddle things a bit since it does a lot more than simply require deed restricted affordable units. TOC projects also get a faster approval process, which we've talked about in the past. They get up to 80% more units. They get additional height and floor area, reduced setbacks and parking requirements. The policy is also restricted to multifamily zone parcels near transit and it offers bigger bonuses near higher frequency transit service. So in that sense, it's definitely much more than just an inclusionary policy, but at the same time, I think it's actually pretty rare for a city to mandate affordability in isolation to do nothing except say from now on, all buildings need to include or all buildings over a certain size need to include X percent of affordable units. Inclusionary policies are very often accompanied by upzoning of one kind or another. And they're also often geographically constrained like the TOC program is. And since pairing inclusionary policies with upzoning can from my perspective, make the difference between something that has a net positive impact versus a net negative, I think there's good reason to put them under the same header.  

Mike Lens  16:31   
Yeah, that makes sense. And for the listeners out there that are not in the weeds of California planning or Los Angeles planning and zoning policy- Get it together. The transit oriented- Get on board. Yes. The transit oriented communities program or TOC as we've been calling it here and there is essentially Los Angeles' version or almost a ramp up of the state's density bonus program. Is that how you would call it?  

Shane Phillips  16:59   
Yeah, yeah. The policy went into place in 2017 as a result of a ballot initiative and whole long story, but it goes quite a bit further than what the state density bonus allowed for, which was generally capped out at either a 35 or 50% bonus to the number of units. The maximum bonus for TOC is 80% units and it offers other things that go beyond the state density bonus as well.  

Mike Lens  17:22   
There you go. So Shane, this is an analysis of inclusionary zoning, but it's a modeling exercise or a simulation is another way of thinking it. It's not an empirical study. So just I'm sure we'll repeat this later. These are not things that have actually happened as a result of IZ changes or affordability requirements. These are simulations based on the best models that you and the Terner Center could come up with. So can you tell us a little bit more about this housing policy simulator, what it does, how we should interpret the findings and any limitations or warnings our audience should have in their minds about misinterpreting the results?  

Shane Phillips  18:06   
So the Terner housing policy simulator was created by Mapcraft and Terner Labs. Mapcraft is a consultant that we've also worked with and is great. Terner Labs is a sister organization of the Terner Center for Housing Innovation at Berkeley. And it was created for the city of Los Angeles or alongside of them. The simulator, it was created to test how changes to policy might move housing production up or down, which we care about because we know that when cities build more housing, their rents and home prices tend to grow more slowly. And that's good for affordability. You can refer to episodes 5, 10, 16, 27, 31, 38, and so on. We've talked about this a lot. The simulator tests these changes by combining a bunch of different data sources to run development feasibility calculations called proformas on land throughout the city. It basically evaluates the most profitable use of each parcel of land, which in many cases is going to be whatever exists there right now. But in many cases, it's both legal and financially feasible to build something bigger. A 100 unit apartment building with ground floor retail where a strip mall exists today, for example. For all of the roughly 800,000 parcels in the city, the simulator identifies that highest and best use and applies a development probability over the next 10 years, since we know that not every bit of land with redevelopment potential actually gets redeveloped in near term. So back to that 100 unit project, maybe historical data shows that about 5% of sites with similar characteristics were redeveloped over the previous decade in Los Angeles. So that site has an expected value of 5 units. That does not mean it's going to be redeveloped into 5 units. It'll either be redeveloped to 100 or not redeveloped at all. But because we can't predict which of the 5% of sites like this will be redeveloped, we just count 5% of the capacity of each one and it gives us roughly the same result as if we could perfectly predict what would happen to every parcel at the end of 10 years. We put all of those expected unit counts for every parcel in the city together and the total is our expected production over a 10 year period. So that's sort of a base model for the simulator, taking into account the zoning code and zoning map, building codes, specific plans, historic overlay zones, local ordinances, state policies, market conditions like rents and interest rates, and construction costs. I don't know if all of these things are in the model, but many of them are. We get a ballpark estimate of future production. What I think is really cool about the simulator is it then lets you change a bunch of those inputs to see how they affect production. I guess the simulator wouldn't be too useful if you could only model what we have right now. So in this case I can ask what if I take the base floor area or density limit for each parcel and increase it by 50% or decrease it by 30%? What if I increase or decrease development fees or the time it takes to get a project approved by the city planning department? Or what if construction costs fall or rents rise more quickly? All of these things affect the cost and profitability of developing housing and so they also affect what gets built and how much of it gets built. I can ask questions like that for the whole city or just for specific parts of the city. The question I actually did ask is what happens if we increase or decrease the affordable unit requirement needed to use the density bonus that is provided by the TOC program? Again, part of my motivation for asking that question is this very aggressive affordable housing goal being driven by state policy. In my view that goal is completely unrealistic from a financial perspective. I'm talking specifically about the 40% of units being low or very low income, and it may also do harm if it leads us to prioritize the share of affordable housing units of the total housing that's built over the number of affordable units, which I think is the more important metric. Or if we end up prioritizing a modestly larger number of affordable units at the expense of a much worse overall housing shortage. Chris Elmendorf, Nick Marantz, and Pavo have a good paper from several years ago that touches on this and we'll put that in the show notes. As for how this should all be interpreted, I will do the researcher thing and say there are many reasons to be cautious. I shouldn't over interpret these results. It is a simulation and not an analysis of what actually happened when inclusionary zoning requirements were increased or decreased. We can't do that analysis because it didn't happen. But even if it did, I wouldn't be able to study what would have happened if the requirements hadn't changed. What can you do? But I think it's important to say I also agree with what Anthony Orlando said in our last episode and what Kevin Corinth and Jacob Faber and probably some others said in earlier episodes, which is basically that if we limited ourselves to only studying things with perfect experimental designs, then we wouldn't be able to study much at all really. When we're faced with important questions and imperfect data, we ultimately just have to do our best and try to communicate that uncertainty as well as we can. And I think if we do that, it's usually better than the alternative of essentially throwing up our hands and deciding there's no way to draw any possible conclusions about whether a policy is good or bad. So last thing on this, in the case of this analysis, I'll just say that the specific figures are, I think, much less important than the general magnitudes and directions of these numbers. If the simulation says we expect to build 300,000 units in the baseline scenario and increasing the IZ requirement reduces that to 200,000 units, I wouldn't put too much stock in those specific numbers, but I'm much more confident and willing to stand behind the assertion that the policy will reduce rather than increase production and that the drop in production is probably closer to one-third than to one-tenth or three-quarters. So I'd say readers and listeners should focus more on how the numbers relate to each other than exactly how big or how small they are.  

Mike Lens  24:18   
That's an important point. And a couple reflections on some of the things you said. First, I would say that it's, of course, possible that we're all living in a simulation, so we need to make sure that we're accepting of simulations in our lives.  

Shane Phillips  24:38   
Someone turned the IZ dial to 11% in 2017 is what you're saying.  

Mike Lens  24:42   
Exactly. There's a grand designer out there who's just spinning our dials. The other insight is I think it's really fascinating. This is researcher inside baseball that you just listed Kevin Corinth and Jacob Faber back to back as basically saying the same thing. Listeners can look up those two scholars. There are some differences. Let's jump into the first part of the analysis. You start off by simulating housing production over a 10-year period for 41 scenarios from 0% inclusionary to 40%. You're asking how many units we'd expect to be built if developers could get all those juicy TOC bonuses from setting aside 0% of their units as deed restricted affordable housing. Then you're asking the same question at 1%, 2%, 3%, et cetera, all the way up to 40%. So for each scenario, the simulator is going to pop out a number and you just multiply that number by the IZ requirement to see how many of those units are for lower income households. That's all plotted on a chart figure one in your report, which has the IZ requirement from 0% to 40% on the x-axis and the number of units on the y-axis with separate lines for market rate units below market rate units for extremely low income households and total units. So now that I've described some of the aspects of this figure, what should our listeners be seeing right now in their mind's eye?  

Shane Phillips  26:15  
Yeah, this is the best part of doing a podcast is explaining charts in verbal form. I'm going to do my best here though. So going through this chart, unsurprisingly, we get the most overall production from the simulator at 0% IZ. It's about 400,000 total units and they're all market rate since there's no inclusionary requirement. In the real world, that does not mean we wouldn't build any affordable units since subsidy programs like the low income housing tax credit produce quite a bit with or without IZ. But for the purposes of this analysis, we're just going to call it zero. At 1% IZ, we see a sharp drop in market rate production and a modest increase in below market units. We lose about 71,000 market rate units and add 3,300 below market units. The reason for such a large gap being there is really interesting and important from a policy design perspective. So we can definitely come back to that in a little bit, but just to put it aside for a minute, another interesting result is that total production is pretty stable between 1% and 5% IZ. We're basically trading market rate units for below market rate units on a one-for-one basis in that range. And that's actually related to why there's such a big drop between zero and 1%. From 1% to 40% IZ, the decline in total units is pretty steady. At 1% IZ, we've got about 330,000 units and at 40% we're all the way down to 96,000 total units. There are some small kinks in the line going from one to 40, but it's pretty consistent overall. I should also mention that the real world TOC policy imposes up to an 11% requirement if you're providing extremely low income units, which most developers do. And so that's our baseline scenario for some later analysis, but it doesn't really matter much for the purposes of describing this chart. So we have a big drop in total production from 0% to 1% IZ, then a steady and smaller decline all the way to 40. The story for market rate units is pretty similar. It just declines a little bit faster because some of the losses are offset by more below market rate units. The more interesting part of this is the line for below market production, which doesn't just follow the same trajectory from start to finish. Here what I found is a fairly quick rise as we go from 0% to about 10% IZ, and then the growth slows and slows before peaking at about 25%, at which point it starts to actually fall. So you can think of like a very wide parabolic curve with both ends of the curve facing down, just an arch, essentially a very, very wide arch. To put some numbers to that at 12% IZ, the simulation gives us 34,000 extremely low income units. And I should say extremely low income means it's rented at a price affordable to people earning 30% of area median income. In Los Angeles, this is something on the order of 30, $35,000 or so. So again, 12% IZ is 34,000 ELI units, double the IZ requirement and then some a little bit to 25%. And the number of ELI units rises to 49,000. So doubling the IZ requirement does not come anywhere close to doubling the number of affordable units. From there, it's actually all downhill. As the IZ requirement increases from 25 to 40%, the number of market rate units and the number of below market rate units both steadily fall. So at that point, not only is IZ costing us market rate units, which it does all the way from zero to 40 and is maybe bad, but perhaps acceptable to some people, it's actually costing us affordable units as well. It's really a lose-lose. And it points to what you often hear from critics of inclusionary housing policies, something along the lines of it's better to have 10% of something than 50% of nothing. If an inclusionary policy is so strong that it's deterring most new development rather than perhaps just some, then it's producing neither market rate units nor affordable units. I have a little commentary for something that did not make it into the report. Speaking of different political views, some listeners are probably familiar with the Laffer curve, which charts total tax revenues at different tax rates. It posits that there is some point along the curve between a 0% tax rate and a 100% tax rate that maximizes government revenues. I think the idea behind the Laffer curve is just self-evidently true. There has to be some point and it's not going to be 0% because that would raise no money, and it's not going to be 100% because a lot of people just wouldn't work at all if all their money was taxed away. I think something similar applies here with inclusionary zoning. In the report, I had initially described this as a sort of Laffer curve for inclusionary zoning, but we decided to drop that since in the real world, the concept is completely politicized and Arthur Laffer himself is a bit of a wacko. I also just learned recently that he helped pass Proposition 13 in California way back in 1978, so don't want to give that guy any more credit than he deserves. 

Mike Lens  31:29   
Understood. I wonder where Arthur Laffer himself these days would fit along the political spectrum curve. He might have been labeled as a wacko then, but now he might be considered relatively sober.  

Shane Phillips  31:43   
Well, I think he got a presidential medal of freedom in 2019.  

Mike Lens  31:49   
Uh-oh. He has been honored by a bedfellow that we don't like to keep. That was a very, very nice rundown, I think. If the city or any city were hanging their inclusionary zoning hat entirely on this study, which even as amazing as you are, Shane, and as great as this simulation is, nobody should ever direct policy based on one study. But if they were, I think the clear takeaway would be you can feel free to dive in to a zero to... Going from 0% inclusionary to 1%, right? There are trade-offs that you have to understand, but going from zero to one is a big step. You're going to likely compromise a lot of your market rate housing production. Going beyond 1% after that, you're not going to see that same fall-off, but then after 25%, you're likely to lose out on everything you want, right? Where you're losing not just market rate units, but you're not gaining affordable units at any rate that's likely to catch up to the market rate units. Is that a fair summary there?  

Shane Phillips  33:12   
That is what the chart shows. I do want to offer a little additional caution. First is just assuming that this chart would look the same in any other city. Every city has different market conditions, mixes of rents, construction costs, cap rates, utility requirements, fees, incomes, all the rest. And so I have had some people ask me, does this same thing apply to my city? And the answer is no. And also, does it apply equally to different affordability levels? Because in this case, we're looking at extremely low income units and the answer again is no. And so again, this is why I go back to the general shape of these things and the relationship of one scenario to another is probably more important than the specific numbers. I would also say, I think the way you framed that question sort of implied that maybe the 25% is the optimal number because it gives you the most affordable units. And I would say that is almost certainly not true. Yes. Yes. Agree. The peak is the point at which the marginal increase in affordable units is effectively zero from increasing the inclusionary requirement further. And the loss of market rate units remains very high. It just keeps going down. So going from 20% IZ to 25% IZ only adds about 3,000 affordable units, but it subtracts more than 37,000 market rate units. So between those two scenarios, we're trading away more than 12 market rate units for every one affordable unit. It's certainly true that some household will benefit from that affordable unit and that does matter, but 12 more will lose out in some way. And the city will lose out on tax revenues and some of those homes will be built somewhere else where there probably aren't as many good jobs and there's maybe a lower quality of life, at least for what many people are looking for. And most importantly, I think rents across the city will go up just that little bit faster because there are fewer available homes and competition between tenants is just that little bit stronger. I just think that losing out on 12 homes to produce one is a very steep price to pay. I think my main lesson from this is that there is no optimal IZ requirement and to the extent that one exists, it might be no IZ at all and we should come back to that. But for places that have IZ and want to get better outcomes from it, I do think this kind of analysis can probably give you a sense for whether you'd be better off moving to a slightly higher or a slightly lower requirement. If you're at 15% IZ right now and you find that increasing it to 20% loses you a whole bunch of market rate housing, but only modestly increases your affordable housing production, that's probably a bad deal. And if you find that you could drop it to 10% or 12% and gain a lot of market rate housing without losing much affordable housing, that might be worth doing.  

Mike Lens  36:02   
Okay, so I want to come back to the big drop in production that you see going from 0% IZ to 1% and some of those other, shall we call them, kinks in the line that you see at a few points. As a reminder to listeners, Shane gets an estimated 400,000 units over 10 years in the no IZ scenario and that falls all the way down to 330,000 at 1% IZ, so a drop in 70,000 units. 1% is a pretty trivial requirement, Shane, so why does it have such a big, big impact?  

Shane Phillips  36:39   
This one is interesting because in some ways it's the result of a shortcoming of the model, but it's something that actually captures a problem we see all the time in the real world. What's happening here is that housing is a lumpy good, meaning a building can't be divided into infinitely smaller pieces. The smallest unit it can be divided into is literally a unit, a single dwelling unit. If I'm building a 10 unit apartment building and the city requires me to rent 1% of my units to a low income tenant, I can't just rent out the closet or a bathroom in one of those units. I also can't usually round down, I almost never am allowed to round down and provide zero affordable units. In almost every city where inclusionary zoning exists, I have to round up. So in the case of my 10 unit apartment building, a 1% IZ requirement on paper becomes a 10% IZ requirement in practice. If I'm building a five unit project, then that 1% IZ is functionally a 20% requirement. So the reason we get a big drop in production right off the bat at 1% is because for a lot of projects it actually amounts to a 5, 10, or 15% requirement. It's also why the kinks in the line show up at 5, 10, 15, and so on. At those points, a lot of smaller projects are able to meet the IZ requirement exactly without tipping into having to provide one more unit. But when the IZ requirement increases from 5 to 6% for example, the 20 unit project goes from providing one affordable unit to providing two. The same thing happens going from 10 to 11%, from 20 to 21, and so on. This is one reason cities tend to exempt smaller projects from inclusionary requirements. It's also why I'm a big advocate for allowing developers to pay in lieu fees, at least on these fractional unit requirements. So if we have a 15% IZ requirement, a developer building a 10 unit project would be obligated to provide 1.5 affordable units, right? Developers cannot build one and a half units, but requiring them to build two is really kind of unequal taxation. It's not treating everyone the same and it may make some projects infeasible. And if that's the case, we're actually going to get zero affordable units, so not a great outcome. To solve that problem, we might just have them build one affordable unit onsite and then allow them to pay a fee equal to maybe half the cost of building another or half the subsidy portion of building another, I think that's probably more fair, and invest that money into another building either buying or purchasing and turning a unit affordable elsewhere. In practice, it's not an easy thing to figure out exactly how much a developer should pay in lieu of building half an affordable unit. I think we talked about this actually in our episode with Jake and Max a few episodes ago. Cities very often undersized their in lieu fee and end up getting fewer onsite units. But that said, I don't think it's incredibly challenging to find a reasonable amount here. And it would make for a much fairer system, you know, to say nothing of producing more housing of both market rate and affordable variety.  

Mike Lens  39:49   
Yeah, that is absolutely fascinating way to think about it because I don't think a lot of folks who look at inclusionary zoning percentages really think about the fact that in a lot of cities, most of your developments are going to be south of 100 units. And therefore, the percentage is technically going to be, or a larger, sorry, the percentage is just going to increase. If you hold the line at 5% of 50 units or whatever, then you're going to have a big share of your units that are supposed to be affordable. And then you say, okay, well, we'll exempt people below 20 units or developments below 50 units. Well-  

Shane Phillips  40:36   
You're going to get a lot of 19 or 49 unit projects.  

Mike Lens  40:39   
And 49s. There's a reason why economists hate hard cutoffs in things like this because they know that people can do a little bit of arithmetic and see which side of that hard line they want to be on, right? They're not going to provide 51 units. They're going to provide 49. They're not going to provide 21. They're going to provide 19. And then people are going to get out of this inclusionary policy more often than you'd like. As you mentioned, a way to smooth those lines out is to allow people to instead go with in lieu fees where you essentially pay some amount of money to compensate the city, compensate future low-income renters, whatever you want to think of it for the fact that you're not going to provide as much affordable housing as the letter of the law percentage suggests or mandates. And as you said, how we determine the precisely perfect fee there is challenging sometimes, but is a better system than these hard cutoffs probably. So now that we have a sense for what happens to market rate and below market rate housing supply when we raise or lower the IZ requirement according to your simulations, let's move on to what this all means for housing affordability. Shane, you tackled this question in two parts, starting with who benefits and then comparing it to who pays. First, who benefits from IZ and how much do they benefit?  

Shane Phillips  42:15   
So the main beneficiaries of the affordable IZ units are, of course, the people who get to live in them and pay, especially in the case of Los Angeles, very deeply discounted rents. In LA in 2019, which is when a lot of the inputs for the simulator were where they come from, the median rent in a new multifamily apartment was around $2,500 per month. And the restricted rent on an extremely low income unit was $503 per month. So a very, very significant discount. That means in year one, each ELI unit is being subsidized with private funding to the tune of $2,000 per month. Take that out to a decade in the future after all this housing has been built, adjust for typical rent increases over that time. And you've got the total value of private subsidies in year 10 for the various scenarios I just gave for 5% IZ, 11%, 16% and 25%. So I just picked those for reasons we don't need to get into, but there was a reason for it. The value of private subsidies in year 10 range from $551 million to $1.67 billion per year. 25% IZ produces the most private subsidy because it produces the most affordable units of all the scenarios. So there was no need to look really further beyond that. I do want to be clear that these numbers are only the benefits that can be easily quantified and there are definitely other benefits to affordable inclusionary units as well. For one thing, maybe there are benefits to having people with different incomes live in the same building. To the lower income people, but maybe for the higher income people as well. The same applies if IZ encourages some amount of racial or ethnic integration, which is a goal of policy here in California in particular. That is totally plausible. Beyond the savings on rent, the person living in that affordable unit might be at lower risk of homelessness or other problems caused by housing instability. They might just have more stable housing overall. And they'll probably place some value on the predictability of their rent being tied to the area median income rather than just whatever the market will bear. And then there's the fact that a dollar of subsidy benefiting an extremely low income household probably means much more to them than it does to whoever's most likely to be paying it. And even beyond that, not only is their marginal utility per dollar, we would call it, probably higher than the people paying for that subsidy, there are also probably many more people paying than receiving, which is maybe a nice segue to the question of who pays.  

Mike Lens  44:50  
Just to reiterate, you're getting a little bit into the distributional consequences here, right? We might, individuals with lower incomes, of course, are going to benefit from these subsidies, period. They might benefit more from these subsidies than the people paying lose out, right, from those subsidies. And then whether or not we can really measure that as a society, of course, we hopefully have some attention to the plight of people with less and we want to provide them with these subsidies just to kind of, you know, restate that. And so those benefits are fairly substantial and I think they're important and good, but in some ways the measurement of that is the easy part, right? And measuring the costs of producing those inclusionary units is a bigger lift. And I think I'm informed by your conversation in episode 31 with Emily Hamilton about how this is more complicated to figure out kind of what those costs are. And so how did you do it in this study?  

Shane Phillips  46:02   
I didn't! So as you said, measuring the effects of inclusionary zoning policies is very difficult. I would have loved to do that, but who's got the time, right? I took the easy way out for this analysis in part because it was a smaller analysis, but I think an upside to my approach is it let me focus instead on explaining how IZ is probably being paid for, which I feel like has gotten even less attention than its empirical effects. Ideally, I could have answered a question like, if we start with no IZ and then impose an 11% affordability requirement and have the TOC bonuses, the density bonuses in place in both scenarios, and that causes 10-year housing production to fall from 400,000 units to 260,000 units, what effect does that have on rents? Or if we increase the requirement from 11% to 16%, what effect does that have on rents? I just don't have the counterfactual and I am better at producing podcasts than PhD level statistical analysis, so I am not able to answer that.  

Mike Lens  47:13   
I'm kind of middling at both, so that's fine.  

Shane Phillips  47:18   
What I can do though is turn the question around a little bit. I can't say how much these different scenarios would increase rents, but because I know the ballpark value of the affordable units produced in each scenario, I can answer the question, how much would rents need to increase for the costs paid by renters on the private market to equal the benefits received by renters in deed restricted affordable housing? Taking us back to that 11% IZ scenario, in that simulation, we end up with 31,800 affordable units at the end of 10 years, and the value of private subsidies invested in those units is about $1.1 billion annually. So we've got 31,800 direct beneficiaries, say households. On the other side of the ledger, we have all the renters in the private market who don't live in public housing or deed restricted housing that already exists or is built through public subsidy programs like the low-income housing tax credit or housing supported by project-based vouchers. In the city of Los Angeles, that's about 85% of renters or 740,000 households. In 2019, the median renter household paid about $1,550 per month in rent. We assume rents for these households increase 4% per year on average as a baseline, which is the same assumption that we made for the market rate units to calculate the benefit at the end of 10 years of all of those. So now the question is how much faster do rents need to increase on those 740,000 households beyond that 4% that we've already got baked in for them to be paying at least an extra $1.1 billion in rent by year 10? And the answer maybe unfortunately is not that much. It's just a lot of people, so they don't have to pay a whole lot more for them to exceed that value. I find that if rents rise just 0.6 percentage points faster each year, and that's going to be compounding, so it's in addition to the 4%, then the cost to these many renters on the private market matches the benefit to the fewer renters who get into those affordable units. Another way of putting that is that the costs negate the benefits if by year 10 rents have gone up at least 49.9% instead of 42.3%. That's just total over a 10 year period. And as a reminder, going from 0% IZ to 11% reduces total production in the simulation by 110,000 units, more than a quarter, and it reduces market rate production by 140,000 or more than a third. The 25% IZ scenario reduces market rate production by more than 60%. And in that case, rents would have to grow an extra 0.9 percentage points faster each year to negate the subsidies going to those deed restricted, extremely low income units. The reality is none of us knows for certain how big an effect these supply shocks would have on rent growth, but 35% lower market rate production causing rents to rise 0.6 percentage points faster per year, or 60% reduction leading to 0.9 percentage point faster annual rent growth. That sounds pretty plausible to me, if not conservative. Emily's empirical research in DC found that each year IZ policies were in place was associated with a 1.1% increase in price per square foot, just as an example. So it's actually faster, even though that is looking at home prices rather than rents. But I think if those numbers sound plausible to others, they should motivate, I think, at least a bit more caution in places considering those kinds of policies.  

Mike Lens  51:01   
So I think this is slightly counterintuitive, Shane. And so I want to kind of emphasize this a little bit. So because the population of market rate renters is so large, it doesn't take a lot of rent increase for the average one of those renters for those costs to pile up pretty fast, right? But that doesn't necessarily mean that each individual renter, market rate renter, is absorbing a large rent increase necessarily. It's just saying that we have a very big population of people who are going to likely see some kind of rent increase, and those costs are then going to, in the aggregate, get quite large.  

Shane Phillips  51:56   
Yeah, yeah. I'll say more about the renters because I think who these people are is really important. So again, this is the population of renters who don't live in rent restricted housing. And it is, I think, a key message of the analysis. As with a lot of restrictive land use policies, IZ is often framed as a cost imposed on developers, but that's pretty unlikely. In the long run, cities can't really reduce developer profits. Developers aren't going to just accept, say, 10% returns in LA when they can get 12% in Long Beach or in Houston. That doesn't mean they're just not going to build in LA if they impose more restrictions. It just means that they have less money available to pay for land that they would like to build on, and the less they can pay, the fewer willing sellers they're going to find. That's ultimately the mechanism by which land use restrictions and fees and so on worsen affordability. Less land gets sold to developers, fewer homes get built, and scarcity drives rent and price growth. So renters and home buyers ultimately pay that cost. And after maybe accepting that the costs aren't being borne by developers necessarily, I think a lot of people then jump to assuming it's just the people renting and buying new homes that will pay the costs. It's the market rate tenants in these mixed income buildings. I think that is probably wrong too, because again, the effect of this policy is slower housing production, and that affects the whole market. We just re-released our interview with Evan Mast recently, if people haven't heard it and want an in-depth explanation for why that's the case. So now we're looking at the whole population of renters, and while some of them make a lot of money, most do not. This is another point I really want to emphasize. In the city of Los Angeles, the median household income for people who own their home is $130,000 as of 2023. For renters, it is $62,000. It is less than half. We are not unique in this regard. This is true of most cities. There are 1.46 million households in Los Angeles, and 300,000 of them are renter households making less than $35,000 per year. A significant number of these people do live in public housing or deed-restricted housing, but most don't. Along with most other renters, they are paying some of these costs. They're probably paying lower rents, and so faster rent growth means a smaller dollar amount for them, but this is also a population who is least capable of handling any kind of additional rent growth. Part of why I felt compelled to write this report and to emphasize this part is because I think people believe it's developers and rich renters who are paying these costs, but I don't think that's right. It is renters really up and down the income spectrum. And meanwhile, something that frustrates me is homeowners whose incomes are more than double theirs are actually growing a bit wealthier off of all of this scarcity rather than contributing to the problem in terms of paying something in. And I am one of those homeowners, for the record, listeners, and so is Mike. So this is something I feel very personally.  

Mike Lens  55:04   
Yeah. And I mean, straight up, I have owned my home for, what, 12 years. I am quite certain it's doubled at least in value in that time, and I don't remember building any inclusionary housing or having some kind of inclusionary housing bill hit my home. So Shane, is the lesson here that inclusionary zoning is just a bad policy and cities should be doing something entirely different? What alternatives do local elected and public officials have for creating affordable units if this is not the best way to do it? Building deed restricted affordable housing is expensive, very expensive. Each unit requires hundreds of thousands of dollars in subsidies, and seems like inclusionary zoning is one of the few tools we have available to produce these homes in cash-strapped cities.  

Shane Phillips  56:01   
Yeah. I know I am coming across as pretty pessimistic about IZ here.  

Mike Lens  56:06   
We've definitely lobbed some bombs on IZ today.  

Shane Phillips  56:11   
I'm sure we will hear a little bit about it from some folks, and we always welcome your comments and questions. But I just don't think IZ is going to get us where we need to go in the long run. In a totally different context, one of our previous guests, Haydn Shelby, has this quote about a slum upgrading program in Bangkok that she describes as blurring the boundaries between radical alternatives to the market and neoliberal alternatives to the state. A lot of liberal and lefty people, which I count myself among, see IZ as this radical alternative to the market, but in my view it is much more the neoliberal alternative to the state. It's what we are left with when we give up on actually funding affordable housing adequately. Affordable housing is very expensive, and I just don't know why we would think we could make developers build it for us for free without any unintended consequences. I am genuinely, genuinely sympathetic to the public officials who are polled in a lot of different directions on this. Voters say they want the kind of affordable housing that even poor people can afford, but they don't want to pay for it, and they don't want it built near them even if someone else pays for it. I'm just frustrated though by the idea that IZ is going to be a really meaningful contribution to the solution here, and I think there are reasons to think it could actually take us in the wrong direction. I just don't think anyone really believes IZ is going to be a large share of the solution, or I just don't think they should. There are very few IZ programs across the country that have produced more than a few hundred units total over their entire lifetimes, and even most of the very anti-market, or at least market skeptical people out there, I think will concede that high enough requirements will eventually choke off development. We can disagree about where that point is, but I think most people agree it is somewhere. Even the San Francisco Board of Supervisors voted 10 to 1 last year to reduce its inclusionary mandate from 22% to 15% in response to all the changes after the COVID pandemic, and if that group of people can agree that they've gone too far in restricting the market, then I think that really says something.  

Mike Lens  58:25   
Yeah, that's the soak the developers band of merry people.  

Shane Phillips  58:31   
For the most part, yeah. I don't want to go so far as to say inclusionary zoning is always a bad thing. Policymakers and advocates should be very cautious about whether they implement new IZ policies, but I'm really most concerned about cities that are considering ramping up these programs that already exist, especially if they're planning to do so without increasing their development bonuses or somehow compensating for the cost of those higher requirements. But frankly, even then, I also argue in this report that bonuses offer diminishing returns, and an even more recent episode, our one with Anthony Orlando on building heights and construction costs, goes a long way to explaining why. Part of why I don't want to be an absolutist about IZ is I've seen from LA's experience with TOC the Transit-Oriented Communities Density Bonus Program that we've had here in LA since 2017. This is a policy that has voluntary participation, and that's important, and it pairs big redevelopment bonuses with reasonable inclusionary requirements. And I think that has been, by most accounts, a win-win for the city. Carolina Reid, who helped advise on this project at Terner Center, had a great recommendation while this report was still underway to not just look at the effect of the TOC program with different IZ requirements, but to also see how production changed if we dropped both the TOC bonuses and the IZ. The result was that when we compared TOC bonuses with an 11% IZ requirement, that scenario, which is essentially the status quo policy in the city, to a scenario with no TOC bonuses and no IZ requirements, the TOC policy produced more market rate housing and more affordable housing. And as we know from another study some of us Lewis Center folks published in the Journal of the American Planning Association last year, the TOC policy also did other great things like make some important process reforms that sped up project approvals by about 30% on average.  

Mike Lens  1:00:29   
And I assume we will have those papers in the show notes as well. As always. Great. Shane, I feel like we often end with something about the political feasibility of some of the brilliant ideas that our scholars come to us with, and we've gestured at this, but the politics of inclusionary housing seem rather difficult to escape. In a lot of cities, we probably need to be investing hundreds of millions or billions of dollars in affordable housing, and that's what you've talked about in this last response. But they don't have that kind of money to spare in most cities. It's easy to attack developers and frame them as the bad guys or at least a source of easy money, and people generally believe that the developers are the ones paying the costs of inclusionary zoning. Really, anytime you have concentrated costs like this, or at least the perception of concentrated costs, and you have a villain like developers, rightly or wrongly, the politics are very ripe for passing ordinances like this and their diffusion across the country. So the question, Shane, is, is there a solution to not, of course, building some kind of political coalition that destroys all inclusionary zoning policies – I don't think that's what we would advocate for – but for producing a politics or a framing or a communication around this issue that makes it a little bit clearer what the trade-offs are?  

Shane Phillips  1:02:06   
I think there is some hope, and I'll point to a few things. I want to add, actually, though, in addition to the costs being seen as concentrated on unsympathetic group developers, there's also concentrated benefits for a much more sympathetic group, which is just low-income households. When someone like me implies that we should get rid of inclusionary zoning, and maybe that just means fewer affordable units, deed-restricted, very deeply affordable units, in exchange for everyone maybe paying $100 less on their rent, the fact that the total value of everyone's lower rents or their savings on rents exceeds the total value of the benefits to these extremely low-income households, for example, it doesn't change the fact that those extremely low-income households are probably going to be a lot worse off, and that's a very visible problem. Some of them might become homeless, and so I very much don't want to come across as dismissive of IZ or the politics here, or just the real human cost of picking one path or another. But as I said, I do think there are some ways forward. I agree with everything you said about why IZ ends up being a go-to in a lot of jurisdictions, and will again underline my sympathy for the limited tools that local officials have to solve that problem. Building housing at scale requires a whole lot of money, and it is money that's not going to be paid back, and local jurisdictions just don't have enough to do all of this on their own. But I'll say a few things about principles and share one case study that I think is promising. At a high level, my view is that land use policies are good for creating the conditions for broad housing abundance and affordability. Zoning codes, building codes, various development standards, they should be designed to ensure we have a wide variety of housing options and more than enough homes in the places people want to live. If we have that, homes will be more affordable across the board. What land use policies are much less suited to is producing deeply affordable housing, the kinds of deed restricted units that a very low or extremely low income person can afford without other assistance. For that, we need subsidies. There's not really any way around that. I think where inclusionary zoning goes wrong is it is attempting to produce affordable housing by substituting land use policy in place of broadly shared taxes and subsidies. Although that does partially achieve the goal of building affordable homes, it undermines what I think should be the primary land use goal of maintaining affordability for all across the board, market wide. The case study I want to talk about here is Portland, Oregon. I think it's rightly got quite a bit of attention over the past several months. They just recently expanded their program citywide to offset the cost to actually give a property tax rebate to developers who comply with their inclusionary zoning requirements. So essentially they say, if you provide, I don't recall the percent, let's say it's 15% of your units for low income households, then we will offer you a 10 year property tax abatement either on just the affordable units or in some cases on the entire project. I think this is a good approach because it doesn't burden developers with extra strings attached like upfront construction subsidies do where you're going to have extra labor requirements and other things that just add a lot of cost. And the abatement expires after 10 years. So to the extent that the tax break spurs development that otherwise wouldn't have occurred, you get to collect revenue on that building forever after those first 10 years have passed. I see the Portland model as a way to get the total housing production of the 0% IZ scenario, but then get the deed restricted affordable housing production of these 15 or 20% scenarios. I think it's really where we need to go. And it really aligns with my view that land use policy should be for broad affordability and we need to spend money if we want to get affordable units. And we should.  

Mike Lens  1:06:20   
Yep. I couldn't agree more. And Shane, I want to thank you for this opportunity to be the primary host. And I-  

Shane Phillips 
How's it feel?  

Mike Lens 
You know, I really had to be a lot more locked in today. I got to admit, but it was a lot of fun.  

Shane Phillips  1:06:37   
I had to prepare a lot more as a guest. I actually feel a lot more sympathy for our guests now. So it was probably good to turn the tables around once in a while.  

Mike Lens  1:06:45   
Well, in some ways you had both jobs in preparation, but thank you, Shane Phillips.  
Shane Phillips  1:06:50   
Thank you, Mike Lens.  

Mike Lens 
Cool.   

Shane Phillips 
You can read more about the Lewis Center's work on our website, lewis.ucla.edu. Show notes and a transcript of the interview are there too. The UCLA Lewis Center is on the socials, I'm on Bluesky at shanedphillips, and Mike is on Twitter at mc_lens. Thanks for listening. We'll see you next time.  

People on this episode

Contributor

Michael Lens

Host
Contributor

Michael Manville

Host
Contributor

Paavo Monkkonen

Host
Contributor

Shane Phillips

Host