The URL can be used to link to this page

26 Susana Barrios From:Christine Schachter <christines@pwr.net> Sent:Monday, June 26, 2023 5:48 PM To:City Clerk; Public Comment; Ashleigh Aitken; Natalie Rubalcava; Jose Diaz; Carlos A. Leon; Norma C. Kurtz; Stephen Faessel; Natalie Meeks; Berenice Ballinas; Valeria Sandoval; Sarah Bartczak; Taylor N. Griffin; Nadia Villafana; Robert Mendoza; Nam Bartash; Joel Saldivar; Cameron Wessel; City Manager; Jim Vanderpool; Andy Nogal Cc:Phil Hawkins; Tim Shaw Subject:\[EXTERNAL\] PWR Letter re June 27, 2023 City Council Agenda Item No. 26 Attachments:PWR Letter to Anaheim City Council re Agenda Item 26_6.27.23.pdf; Below-Market Housing Mandates as Takings - Measuring their Impact.pdf; Do Affordable Housing Mandates Work? Evidence from Los Angeles County and Orange County.pdf Warning: This email originated from outside the City of Anaheim. Do not click links or open attachments unless you recognize the sender and are expecting the message. Honorable Mayor Aitken and Members of the City Council: Please find attached a letter from the Pacific West Association of REALTORS® (PWR) regarding item no. 26 on the June 27, 2023 meeting agenda. Thank you for your time and consideration. Sincerely, Christine Schachter Government Affairs Director Pacific West Association of REALTORS® 1601 E. Orangewood Ave. Anaheim, CA 92805 714-245-5500 (main) 1 714-221-8474 (direct) christines@pwr.net Serving Orange County & Los Angeles County 2 PACIFIC WEST ASSOCIATION OF REALTORS® • ANAHEIM OFFICE - MAIN LONG BEACH OFFICE 1601 East Orangewood Ave., Anaheim, CA 92805 5000 East Spring St Suite #110, Long Beach, CA 90815 (714) 245-5500 | www.pwr.net June 26, 2023 Anaheim City Council 200 S. Anaheim Blvd. Anaheim, CA 92805 Re: June 27, 2023, Council Agenda Item No. 26, Housing Affordability Ad Hoc Committee Update Honorable Mayor Aitken and Members of the City Council: The Pacific West Association of REALTORS® (PWR) is the local real estate association that represents Los Angeles and Orange Counties, including the city of Anaheim, and PWR’s 13,000 members promote homeownership, private property rights, and responsible government. We recognize the need both locally and statewide to address the housing affordability crisis, which has made housing a substantial economic burden on working families, senior citizens, and residents on fixed incomes, and we respectfully request that the City of Anaheim support housing policies that are centered around process improvements, incentives and adaptable development standards as opposed to counterproductive measures such as market controls or government mandates. The consideration of an affordable housing development fee or inclusionary housing program by the City’s Housing Ad Hoc Committee is problematic without having obtained public input, a competitive bid process for a study of other potential alternatives, and market analysis of existing incentives. A comparable case is that of Santa Ana where burdensome inclusionary housing requirements imposed by the city council effectively halted residential construction across the city. Unfortunately, inclusionary zoning does not address the factors that contribute to the high cost of housing such as land prices, lack of available sites, developer fees and exactions, cumbersome permitt ing processes, etc. For your reference, please find the following attachments on this topic. - “Below Market Housing Mandates as Takings: Measuring Their Impact”, Independent Policy Report, by Tom Means, Edward Stringham, and Edward Lopez, November 2007. - “Do Affordable Housing Mandates Work? Evidence from Los Angeles County and Orange County”, Reason Public Policy Institute , by Benjamin Powell and Edward Stringham, June 2004. We stand ready to support efforts that are proven to work but placing the burden on current residents by driving up housing prices is not the solution. For so long, Anaheim has been a leader in finding innovative ways to build a vibrant and thriving commun ity and we hope to see that continue. Sincerely, Phil Hawkins Tim Shaw Christine Schachter Chief Executive Officer Government Affairs Director Government Affairs Director 3065*/(Housing affordability has become a major issue in recent years. To address the problem, many cities have adopted a policy known as below-market housing mandates or inclu- sionary zoning. As commonly practiced in California, below-market housing mandates require developers to sell 10–20 percent of new homes at prices affordable to low-income households. Many developers, however, argue that the program is in violation of the takings clause of the U.S. Constitution because it forces devel- opers to use some of their property to advance a public goal. Nevertheless, in Home Builders Association of Northern California v. City of Napa (2001), the court ruled against the regu- latory takings argument, saying that below- market housing mandates are legal because (1) they offer compensating benefits to developers and (2) they necessarily increase the supply of affordable housing. This study investigates these claims in the following way: Section 2 discusses the his- tory of regulatory takings and discusses why below-market housing mandates may be con- sidered a taking. Section 3 investigates how much below-market housing mandates cost developers. Section 4 investigates econometri- cally whether below-market housing mandates actually make housing more affordable. Our research indicates that the deci- sion by the California Courts of Appeal is on shaky ground. Below-market housing mandates require developers to forego sub- stantial amounts of revenue and they provide little offsetting benefit. A mandate in Marin, California, for example, would require devel- opers to forfeit roughly 40 percent of revenue from a project, and builders are offered almost nothing in return. We can see how below-market housing mandates affect housing markets by using econometrics to analyze data of price and quantity for California cities in 1990 and 2000. Our regressions show that cities that impose a below-market housing mandate actually end up with 10 percent fewer homes and 20 per- cent higher prices. For developers, inclusionary zoning has an effect similar to a regulatory taking. For society in general, affordable housing man- dates decrease the supply of new housing and increase prices, which exacerbates the afford- ability problem. Below-Market Housing Mandates as Takings: Measuring their Impact Tom Means, Edward Stringham, and Edward LopezNovember 2007 Executive Summary Independent Policy Reports are published by The Independent Institute, a nonprofit, nonpartisan, scholarly research and educational organization that sponsors comprehensive studies on the political economy of critical social and economic issues. Nothing herein should be construed as necessarily reflecting the views of The Independent Institute or as an attempt to aid or hinder the passage of any bill before Congress. Copyright ©2007 by The Independent Institute All rights reserved. No part of this book may be reproduced or transmitted in any form by electronic or mechanical means now known or to be invented, including photocopying, recording, or infor- mation storage and retrieval systems, without permission in writing from the publisher, except by a reviewer who may quote brief passages in a review. The Independent Institute 100 Swan Way, Oakland, CA 94621-1428 Telephone: 510-632-1366 • Fax: 510-568-6040 Email: info@independent.org Website: www.independent.org ISBN 978-1-59813-023-2 1. Introduction High housing prices in recent years are mak- ing it increasingly difficult for many to pur- chase a home. Prices have been rising all over the United States, especially in cities on the East and West Coasts. In San Francisco, for exam- ple, the median home sells for $846,500 (Said, 2007, p.c1), which requires yearly mortgage pay- ments of roughly $63,000 (plus yearly property taxes of $8,500).1 Not only is the median home unaffordable to most, but there is a dearth of affordable homes on the low end, too. In San Francisco, a household making the median income of $86,100 can afford (using traditional lending guidelines) only 6.7 percent of existing homes (National Association of Homebuilders/ Wells Fargo, 2007). Households making less are all but precluded from the possibility of home ownership (Riches, 2004). As a proposed solution, many cities are adopting a policy often referred to as below- market housing mandates, affordable housing mandates, or inclusionary zoning (California Coalition for Rural Housing and Non-profit Housing Association of Northern California, 2003). The specifics of the policy vary by city, but inclusionary zoning as commonly prac- ticed in California mandates that developers sell 10–20 percent of new homes at prices affordable to low-income households. Below-market units typically have been interspersed among market- rate units, have a similar size and appearance as market-rate units, and retain their below-market status for a period of fifty-five years.2 The pro- gram is touted as a way to make housing more affordable, and as a way to provide housing for all income levels, not just the rich. In contrast to exclusionary zoning, a practice that uses housing laws to keep out the poor, inclusionary zoning is advocated as a way to help the poor. Because of its expressed good intentions, the program has gained tremendous popularity. First introduced in Palo Alto, California, in 1973, the program has increased in popularity in the past decade Below-Market Housing Mandates as Takings Measuring their Impact Tom Means, Edward Stringham, and Edward Lopez the independent institute2 | and is now in place in one-third of the cities in California (Non-Profit Housing Association of Northern California, 2007). And it is spread- ing nationwide, having been already adopted in parts of Maryland, New Jersey, and Virginia (Calavita, Grimes, and Mallach, 1997). But the program is not without controver- sy.3 In Home Builders Association of Northern California v. City of Napa (2001), the Home Builders Association maintained that by requir- ing developers to sell a percentage of their development for less than market price, the “ordinance violated the takings clauses of the Federal and State Constitutions.” A ruling by the Court of Appeals in California stated that affordable housing mandates are legal and not a taking because (1) they benefit developers, and (2) they necessarily increase the supply of afford- able housing. This report investigates these claims by examining the costs of the programs and reviewing econometrically how they affect the price and quantity of housing. Our report is organized as follows: Section 2 discusses the history of regulatory takings deci- sions by the courts and relates them to affordable housing mandates. It provides a brief overview of regulatory takings decisions and discusses the arguments about why affordable housing man- dates may or may not be considered a taking. When government allows certain buyers to pur- chase at below-market prices, it is making sellers sell their property at price-controlled prices. If sellers are not compensated for being forced to sell their property at a below-market price, that may be considered a taking. Section 3 investigates how much affordable housing mandates cost developers. By calculat- ing the price-controlled level and comparing it to the market price, we can observe the costs to developers each time they sell a price-controlled home. After estimating how much the program costs developers, we discuss to what extent they are being compensated. We find that the alleged benefits to developers pale in comparison to the costs. Section 4 investigates econometrically whether below-market housing mandates actu- ally make housing more affordable. Using panel data for California cities, we investigate how below-market housing mandates affect the price and quantity of housing. We find that cities that adopt below-market housing mandates actually drive housing prices up by 20 percent and end up with 10 percent fewer homes. These statistically significant findings thus bring into question the idea that mandating affordable housing neces- sarily increases the amount of affordable hous- ing. Section 5 concludes by discussing why, con- trary to Home Builders Association of Northern California v. City of Napa (2001), below-market housing mandates should be considered a tak- ing. 2. Below-market Housing Mandates and Takings What are “takings,” and should affordable hous- ing mandates be considered a taking? The most familiar form of taking is when the government acquires title to real property for public use, such as common carriage rights of way (roads, rail, or power lines). Precedent for these types of takings is evident in early U.S. jurisprudence, which institutionalized the principle that the govern- ment’s chief function is to protect private prop- erty.4 As such, the government’s takings power was limited in several key respects. Most impor- Below-Market Housing Mandates As Takings | 3 tant, the nineteenth-century Supreme Court prohibited takings that transferred property from one private owner to another and upheld the fundamental fairness doctrine that no indi- vidual property owner should bear too much of the burden in supplying public uses. But government’s takings power has expanded over time. Takings restrictions were gradually eroded beginning in the Progressive Era and accelerating during the New Deal, as the Supreme Court increasingly deferred to leg- islative bodies and an ever-expanding notion of public use. Starting in the latter half of the twentieth century, the stage was set to approve takings for “public uses” such as urban renewal (Berman v. Parker, 1954), competition in real estate (Hawaii Housing v. Midkiff, 1984), expan- sion of the tax base (Kelo v. New London, 2005), and other types of “economic development tak- ings” (Somin, 2004). By the final decade of the twentieth century, one prominent legal scholar described the public use clause as being of “nearly complete insignificance” (Rubenfeld, 1993, p.1078). Regulatory takings differ in that they are generally not subject to just compensation, because they rest on the government’s police power, not the power of eminent domain. Regulatory takings differ also in that the owner retains title to the property but suffers attenu- ated rights. For example, a government might rezone an area for environmental conservation and thereby prevent a landowner from develop- ing his property. But does an owner still own his property if he is deprived of using it according to his original intent? These were the essential characteristics of the regulation challenged in Lucas v. South Carolina Coastal Council (1992).5 In that case, David Lucas owned two plots of land that he bought for nearly $1 million and intended to develop. But the South Carolina Coastal Council later rezoned his property, stating that it would be used for conservation. The Court sided with Lucas, saying that if he was deprived of economically valuable use, he must be compensated. Under Lucas, federal law requires compensation if the regulation dimin- ishes the entire value of the property, such that an effective taking exists despite no physical removal. This so-called “total takings” test is one of several doctrines that could be used to judge regulatory takings. For example, the diminution of value test could support compensation to the extent of the harm done to the property owner. This was the Court’s tendency in the 1922 case Pennsylvania Coal v. Mahon, which found that a regulatory act can constitute a taking depending on the extent to which the value of a property is lowered.6 So the Lucas Court was not up to something new. As a matter of fact, the concept of regulatory takings was discussed by key fig- ures in the American founding era and became an important topic in nineteenth-century legal scholarship as well. 7 Following in this tradition, the Lucas Court addressed several sticking points with regulatory takings law. For example, the majority opinion cited Justice Holmes as stating the maxim that when regulation goes too far in diminishing the owner’s property rights, it becomes a taking. However, as the majority opinion pointed out, the Court does not have a well-developed stan- dard for determining when a regulation goes too far to become a taking. Finally, and most impor- tant for our purposes, the Lucas Court also stressed that the law is necessary to prevent poli- cymakers from using the expediency of police power to avoid the just compensation required under eminent domain. The Lucas Court exam- the independent institute4 | ined regulators’ incentives and voiced its dis- comfort with the “heightened risk that private property is being pressed into some form of pub- lic service under the guise of mitigating serious public harm.” Because they rezone land, requiring owners to provide a public service of making low-income housing, below-market housing mandates seem like they fit into the Lucas Court’s description of what could be considered a taking. This spe- cific issue, however, is still being debated in the courts. In 1999, the Home Builders Association of Northern California brought a case against the City of Napa for mandating that 10 percent of new units be sold at below-market rates. The Home Builders Association argued that the affordable housing mandate violated the Fifth Amendment’s takings clause stating that “pri- vate property [shall not] be taken for public use without just compensation.” The trial court dis- missed the complaint, and in 2001, the Court of Appeals decided against the Home Builders Association, arguing that “[a]lthough the ordi- nance imposed significant burdens on develop- ers, it also provided significant benefits for those who complied.” 8 In addition, the California court argued that because making housing more affordable is a legitimate state interest, then below-market housing mandates are legitimate, because they advance that goal. Judge Scott Snowden (who was affirmed by Judges J. Stevens and J. Simons) wrote, “Second, it is beyond question that City’s inclusionary zoning ordi- nance will ‘substantially advance’ the important governmental interest of providing affordable housing for low and moderate-income families. By requiring developers in City to create a mod- est amount of affordable housing (or to comply with one of the alternatives) the ordinance will necessarily increase the supply of affordable housing.” 9 The Home Builders Association’s subsequent attempts to have the case reheard or reviewed by the Supreme Court were denied. So the Court’s argument rests on two propositions that it considers beyond question: (1) affordable housing mandates provide sig- nificant benefits to builders that offset the costs, and (2) affordable housing mandates necessarily increase the supply of affordable housing. Both of these are empirical arguments that can be tested against real-world data. We investigate these propositions in the following two sections. 3. Estimating the Costs of Below- market Housing Mandates If one wants to state that “[A]lthough the ordi- nance imposed significant burdens on develop- ers, it also provided significant benefits for those who complied,” one needs to investigate the costs of below-market housing mandates in these pro- grams. Yet when this statement was issued by the Court in 2001, there had been no study of the costs.10 The first work to estimate these costs was done by Powell and Stringham (2004a). Let us here provide some sample calculations and then present some data for costs in various California cities. Once we present the costs, we can consider whether the programs have significant, offsetting benefits for developers. First let us consider a real example from Marin County’s drafted Countywide Plan.11 According to the plan, affordable housing man- dates would be designated for certain areas of the county (with privately owned property). In these areas, anyone wishing to develop their property would have to sell or lease 50–60 per- cent of their property at below-market rates.12 Below-Market Housing Mandates As Takings | 5 The plan requires the below-market-rate homes to be affordable to households earning 60–80 percent of the median income, which means price-controlled units must be sold for approxi- mately $180,000–$240,000.13 How much does such an affordable housing mandate cost devel- opers? New homes are typically sold for more than the median price of housing, but for sim- plicity let us assume that new homes would have been sold at the median price in Marin, which is $838,750. For each unit sold at $180,002, the revenue is $658,748 less due to the price control. Consider the following sample calculations for a ten-unit project in Marin that show how much revenue a developer could get with and without price controls. As these calculations show, the below-mar- ket housing mandate decreases the revenue from a ten-unit project by $3,293,740, which is roughly 40 percent of the value of a project. This is just one example, and there are many more. Powell and Stringham (2004a and 2004b) estimate the costs of below-market housing mandates in the San Francisco Bay Area, Los Angeles, and Orange counties. By estimat- ing how much units must be sold for at below- market rates and comparing this to how much homes could be sold for without price controls, one can estimate how much money below-mar- ket housing mandates make developers forgo. Even using conservative estimates (to not over- estimate costs), these policies cost developers a substantial amount. Figure 1 shows that in the median San Francisco Bay Area city with a below-market housing mandate, each price-con- trolled unit must be sold for more than $300,000 below the market price. In cities with high hous- ing prices and restrictive price controls, such as Los Altos and Portola Valley, developers must sell below-market-rate homes for more than $1 million below the market price. One can estimate the costs imposed by these programs on developers by looking at the cost per unit times the number of units built. This measure is not what economists call dead- weight costs (which attempts to measure the lost gains from trade from what is not being built), but just a measure of the lost revenue that devel- opers incur for the units actually built. In many cities, no units have been built as a result of the program, but nevertheless, the costs (in current prices) are quite high. The results for the San Francisco Bay Area are displayed in figure 2. In five cities—Mill Valley, Petaluma, Palo Alto, San Rafael, and Sunnyvale—the amount of the “giveaways” in current prices totals over $1 bil- lion. The next important question is whether developers are getting anything in return. If Mill Valley, Petaluma, Palo Alto, San Rafael, and Sunnyvale were to issue checks to develop- Sample calculations for a ten-unit, for sale development in Marin County Scenario 1: Development without price controls Revenue from a ten-unit project without price controls [(ten market-rate units) x ($838,750 per unit)] = $8,387,500 Scenario 2: Development with below-market mandate Revenue from a ten-unit project, with 50 percent of homes under price controls set for 60 percent of median-income households [(five market-rate units) x ($838,750 per unit)] + [(five price-controlled units) x ($180,002 per unit)]= $5,093,760 the independent institute6 | ers totaling $1 billion, one could say that even though there was a taking, there was also a type of compensation. But the interesting aspect about affordable housing mandates as practiced in California and most other places is that gov- ernment offers no monetary compensation at all. In fact, this is one of the reasons why advo- cates of the program and governments have been adopting it. In the words of one prominent advocate, Andrew Dieterich (1996, p. 41), “a vast inclusionary program need not spend a public dime.” In contrast to government-built housing projects, which require tax revenue to construct and manage, affordable-housing mandates impose those costs onto private citizens, namely housing developers. Here we have private parties losing billions of dollars in revenue and receiv- ing no monetary compensation in return. Monetary compensation for developers is not present, but are affordable housing man- dates accompanied by nonmonetary benefits? The Court in Home Builders Association v. Napa (2001) stated that “[D]evelopments that include affordable housing are eligible for expedited processing, fee deferrals, loans or grants, and density bonuses.” 14 According to California Government Code section 65915, government must provide a density bonus of at least 25 per- cent to developers who make 20 percent of a project affordable to low-income households. The value of these offsetting benefits will vary based on the specifics, but for full compensa- tion to take place, these benefits would have to be more than $300,000 per home in the median Bay Area city with inclusionary zoning. One could determine in two ways that the offsetting benefits were worth more than the costs.15 The first way would be if one observed the building industry actively lobbying for these programs. But in California and most other areas, the building industry is usually the most vocal opponent of these programs. In Home Builders Association of Northern California v. City of Napa the court provided no explanation of why the Home Builders Association would be suing to stop a program if it really did provide “significant benefits for those who complied.” If the programs really did benefit developers, Source: Powell and Stringham (2004a, p.15)Source: Powell and Stringham (2004a, p.15) Below-Market Housing Mandates As Takings | 7 there would be no reason why developers would oppose them. Why don’t builders want to sell units for hundreds of thousands less than market price for each unit sold? Or why don’t California builders want to forgo billions in revenue? All of the builders with whom we have spoken have stated that the offsetting “benefits” are no ben- efits at all. For example, a city might grant a density bonus, but the density bonus might be completely unusable, because density restric- tions are just one of a set of restrictions on how many units will fit on the property. Other con- straints such as setbacks, minimum require- ments for public and private open space, floor area ratios, and even tree protections make it extremely complicated to get more units on the property. Conventional wisdom suggests that building at 100 percent of allowable density will maximize profits, but in reality developers tend to build out at less than full density. The City of Mountain View recently passed a policy requiring developers to provide an explanation for projects that failed to meet 80 percent of the allowable density.16 Prior projects had averaged around 65 percent of allowable density. So giving builders the opportunity to build at 125 percent of allowable density is often worth nothing, when so many other binding regulations exist. The second and even simpler way to deter- mine whether the affordable housing mandates provide significant benefits to compensate devel- opers for their costs would be to make the inclu- sionary zoning programs voluntary. Developers could then weigh the benefits and costs of par- ticipating, and if the benefits exceeded the costs, the developers could voluntarily comply. A few cities in California tried to adopt voluntary ordi- nances, and perhaps unsurprisingly, they did not attract developers. One advocate of afford- able housing mandates argues that the problem with voluntary programs is “that most of them, because of their voluntary nature, produce very few units” (Tetreault, 2000, p.20). From these simple observations, we can infer that the significant “benefits” of these pro- grams are not as significant as the costs. In this sense, the program has the character of a regu- Source: Powell and Stringham (2004a, p.15) the independent institute8 | latory taking. In addition to observing whether builders would support or voluntarily partici- pate in these programs, we can also analyze data to observe how these programs affect the quan- tity of housing. If the Court in Home Builders Association v. Napa is correct that the benefits are significant, then we would predict that imposing an affordable housing mandate would not affect (or it would encourage) housing production in a jurisdiction. If, on the other hand, the program is not compensating for what it takes, we would predict that cities with the program will see less development than in otherwise similar cities without the program. Here the program is a tak- ing that will hinder new development. 4. Testing How Below-market Housing Mandates Affect the Price and Quantity of Housing The court in Home Builders Association v. Napa puts forth an important proposition, which we can examine statistically. The court states: “By requiring developers in City to create a mod- est amount of affordable housing (or to comply with one of the alternatives) the ordinance will necessarily increase the supply of affordable hous- ing” (emphasis added). Although the court sug- gests that it is an a priori fact that price controls will increase the supply of affordable housing, the issue may be a bit more complicated than these appellate judges maintain. Before getting to the econometrics, let us consider some simple economic theory and simple statistics about the California experience. First, if a price control is so restrictive, developers cannot make any prof- its and so the price control can easily drive out all development from an area. Cities such as Watsonville adopted overly restrictive price con- trols, and they all but prevented development until they scaled back the requirements (Powell and Stringham, 2005). Over the course of thirty years in the entire San Francisco Bay Area, below-market housing mandates have resulted in the production of only 6,836 affordable units, an average of 228 per year (Powell and Stringham, 2004a, p. 5). Controlling for the length of time each program has been in effect, the average jurisdiction has produced only 14.7 units for each year since adopting a below-market housing mandate. Since the programs have been imple- mented, dozens of cities have produced a total of zero units (Powell and Stringham, 2004a, pp. 4–5). So unless one defines zero as an increase, it might be more accurate to restate “necessarily increase” as “might increase.” Economic theory predicts that price con- trols on housing lead to a decrease in quantity produced. Because developers must sell a per- centage of units at price-controlled rates in order to get permission to build market-rate units, this policy also will affect the supply of market-rate units. Powell and Stringham (2005) discuss how the policy may be analyzed as a tax on new hous- ing. If below-market-rate housing mandates act as a tax on housing, they will reduce quantity and increase housing price. This is the exact opposite of what advocates of below-market-rate housing mandates say they prefer. So we have two competing hypotheses, that of economic theory, and that of the court in Home Builders Association v. Napa. Luckily, we can test these two hypotheses by examining data for housing production and housing prices in California. Our approach is to use panel data, which has a significant advantage over simple cross- sectional or time-series data. Suppose a city adopts the policy, there is an unrelated statewide Below-Market Housing Mandates As Takings | 9 decline in demand, and housing output falls by 10 percent. A time-series approach would still have to control for other economic factors that might have changed and reduced housing out- put. One would still need to compare the reduc- tion in output from a city that adopted the policy to a nearby similar city that did not. A cross- sectional approach can control overall economic factors at a point in time but will not control for unobserved city differences. Our approach is to set up a two-period panel data set to control for unobserved city differences and to control for changes over time. The tests, which we explain in detail below, will enable us to see how adopt- ing a below-market-rate housing mandate will affect variables such as output and prices. 4.1. Description of the Data The first set of data we utilize consists of the 1990 and 2000 census data for California cities. The 2000 census data are restricted to cities with a population greater than ten thousand, while 1990 census data are not. A decrease in popula- tion for some cities during the decade resulted in a loss of fifteen cities from the sample. We do not include the 1980 census, because there were few policies in effect during this decade (Palo Alto passed the first policy in 1972). Focusing on this decade also highlights some economic issues. From 1987 to 1989, housing prices grew very rapidly. Prices for the first half of 1989 grew around 25 percent, only to fall by this amount for the second half of the year, and continue to slide as the California economy declined. For some areas, prices did not recover to their origi- nal level until halfway through the 1990 decade. The California economy grew faster in the sec- ond half of the decade due to the dot-com boom in the technology sector. Data from the RAND California Statistics Web site provided average home sale prices for each city for the 1990 and 2000 period. The RAND data do not report 1990 home sale prices for some cities, resulting in a loss of more observations. Summary statis- tics are provided in table 1. Data on the policy adoption dates came from the California Coalition for Rural Housing and Non-profit Housing Association of Northern California. Table 2 describes the summary statistics of the policy variables that we con- structed. IZyr is a dummy variable defined to equal one if the city passed a below-market-rate housing ordinance that year or in prior years. As noted above, differences in population cutoff points and missing 1990 housing prices reduced the sample of cities that passed (or did not pass) an ordinance. Starting in 1985, our sample con- tains fifteen California cities that had passed an ordinance. The number increased to fifty-nine cities by the end of 1999. The last column reports the difference between decades. In other words, iz95delta reports the number of cities that passed an ordinance between 1985 and 1995. The differ- ence variables are fairly constant and capture a large number of cities that passed ordinances during the decade. Focusing on the 1990–2000 decade should allow us enough observations to capture the impact of the policy. 4.2. Empirical Tests Jeffrey Wooldridge (2006) provides an excellent discussion of how to test the impact of a policy using two-period panel data. Our approach is to specify a model with unobserved city effects that are assumed constant over the decade (1990– 2000) and estimate a first-difference model the independent institute10 | to eliminate the fixed effect. We also specify a semilog model so that the first difference yields the log of the ratio of the dependent variables over the decade. Estimating the models in logs also simplifies the interpretation of the policy variable coefficient as an approximate percent- age change rather than an absolute difference in averages. For the policy variable, we define IZyr as a dummy variable equal to one if the policy was in effect during the current and previous years. To see the importance of the first-differ- ence approach, consider a model specified for each decade. Level Model: lnYi,t = β0 + d0YR2000i,t + d1IZyri,t + β1 X i,t + a i + vi,t (Equation 1) i = city t = 1990, 2000 The dependent variable is either housing output or housing prices, YR2000 is a dummy variable allowing the intercept to change over the decade, IZyr is the policy dummy variable, and the X are control variables. The error term contains two terms: the unobserved fixed city component (a i) considered fixed for the decade (e.g., location, weather, political tastes); and the usual error component (vit). If the unobserved fixed effect is uncorrelated with the exogenous variables, one can estimate the model using ordinary-least-squares for each decade. The coef- ficient for IZyr measures the impact of the pol- icy for each decade.17 Unfortunately, estimating the level model may not capture the differences between cities that passed an ordinance and the ones that did not. In other words, suppose cit- ies with higher housing prices are more likely to adopt the policy. The dummy variable may cap- ture the impact of the policy along with the fact that these cities already have higher prices. The above issues can be addressed by dif- ferencing the level models to eliminate the fixed city effect, which yields the first-difference mod- el.18 First-Difference Model lnY i,2000 - lnY i,1990 = d0 + d1IZyr i,2000 - d1IZyr i,1990 + β1Xi,2000 - β1X i,1990 + vi,2000 - v i,1990 (Equation 2) i = city which can be rewritten as: ln(Yi,2000/Yi,1990 ) = d0 + d1ΔIZyri,t + β1ΔXi,t + Δvi,t (Equation 3) i = city t = 2000 Eliminating the unobserved fixed city effect, which we show below in the last two col- umns of tables 3 and 4, has an important effect on estimating the impact of the policy variable. Differencing the panel data also yields a dummy variable that represents the change in policy par- ticipation over the decade (an example of this is the iz95delta appearing in tables 2 through 6). When policy participation takes place in both periods (1990 and 2000), the interpretation of the differenced dummy is slightly different from the usual policy treatment approach. The differenced dummy variable predicts the aver- age change in the dependent variable due to an increase (or decrease) in participation. To see the advantage of the first-difference approach, we first estimated (without control variables, which we will add in tables 5 and 6) the un-differenced equations of the log of aver- Below-Market Housing Mandates As Takings | 11 age housing prices and output (lnYi,t = β0 + d1IZyri,t) over various lagged policy dummies. The first four columns in table 3 report the esti- mated coefficients (d1) for each lag year for the level models. The left two columns show the coefficient estimates for the five regressions that look at housing prices in 1990 and have iz1985, iz1986, iz1987, iz1988, or iz1989 as the policy vari- able. The third and fourth columns in table 3 show the coefficient estimates for the five regres- Variable Observations Mean Standard Deviation Minimum Maximum Population 2000 N=446 65,466 (197,087)10,007 3,694,834 Population 1990 N=431 58,468 (187,014)1,520 3,485,398 Households 2000 N=446 22,251 (68,673)1,927 1,276,609 Households 1990 N=431 20,512 (66,074)522 1,219,770 Housing Units 2000 N=446 23,278 (71,843)2,069 1,337,668 Housing Units 1990 N=431 21,745 (70,331)597 1,299,963 Density 2000 (persons/acre)N=446 7.62 (6.06)0.42 37.32 Density 1990 (persons/acre)N=431 6.87 (5.88)0.08 37.01 Median Household Income 2000 N=446 52,582 (21,873)16,151 193,157 Median Household Income 1990 N=431 38,518 (14,543)14,215 123,625 Per Capita Income 2000 N=446 23,903 (13,041)7,078 98,643 Per Capita Income 1990 N=431 16,696 (8,070)4,784 63,302 Rents/Income 2000 N=446 27.60%(3.1%)14.4%50.1% Rents/Income 1990 N=431 28.9%(2.7%)14.9%35.1% Average Home Price 2000 N=360 300,594 (235,436)49,151 2,253,218 Average Home Price 1990 N=352 206,754 (112,804)52,858 1,018,106 Table 1 Summary Statistics the independent institute12 | sions that look at housing prices in 2000 and have iz1995, iz1996, iz1997, iz1998, or iz1999 as the policy variable. For example, the 0.389 in the first row indicates that cities with inclusionary zoning in 1985 had 47.6 percent (exp(0.389) - 1) higher than average prices in 1990, and the 0.627 in the first row indicates that cities with inclu- sionary zoning in 1995 had 87.2 percent higher- than-average prices in 2000. For both decades, the impact increases slightly as the lag period is decreased, though the impact for the 2000 period is much larger than the 1990 period. Table 2 Summary Statistics – Policy Variables Variable # of cities with inclusionary zoning (in that year)Variable # of cities with inclusionary zoning (in that year)Variable Change in # of cities with inclu- sionary zoning (over 10 years) iz1985 15 iz1995 50 iz95delta (which is iz1995-iz1985)35 iz1986 19 iz1996 52 iz96delta (which is iz1996-iz1986)33 iz1987 19 iz1997 54 iz97delta (which is iz1997-iz1987)35 iz1988 22 iz1998 54 iz98delta (which is iz1998-iz1988)32 iz1989 23 iz1999 59 iz99delta (which is iz1999-iz1989)36 Dependent Variable: ln(Price) Level models for 1990 data Level models for 2000 data First-difference models (2000–1990) Policy Variable Coefficient of Policy Variable Policy variable Coefficient of Policy Variable Policy variable Coefficient of Policy Variable iz1985 .389 iz1995 .627 iz95delta .312 iz1986 .431 iz1996 .642 iz96delta .298 iz1987 .431 iz1997 .637 iz97delta .278 iz1988 .442 iz1998 .637 iz98delta .270 iz1989 .457 iz1999 .642 iz99delta .265 Table 3 Summary of Policy Coefficients from Fifteen Regressions on the Price of Housing by Model and by Lag Year Below-Market Housing Mandates As Takings | 13 The estimated coefficients (d1) for 1990 and 2000 range from 0.389 to 0.642 and indicate that cities with inclusionary zoning have 48–90 percent higher housing prices, but this does not take into consideration the possibility that cities that adopted the policy already had higher prices when they did so. To account for this potential problem, the first-difference model estimates how changes in the policy variable (adopting a below- market housing ordinance) alone affect housing prices. The last two columns of table 3 report the first-difference estimates (ln(Yi,2000/Yi,1990) = d0 + d1ΔIZyri,t). For example, the 0.312 in the last column of the first row indicates that cities with below-market housing mandates have 36.6 percent higher prices. Each of the estimated coefficients in table 3 are significant at the 1 percent level. The results in the last two columns indicate that below- market housing mandates have increased the price of the average home by 30 to 37 percent. The results for housing output (the number of units) are even more interesting. These results are presented in table 4. The estimates of d 1 for the level models for 1990 and 2000 are positive and statistically significant at the one percent level, which indicates that cities with inclusion- ary zoning have more housing production, but similar to the housing price regressions do not take into consideration the possibility that cit- ies that adopted the policy already were grow- ing when they adopted the policy. Again, we need to look at the difference in output based on cities adopting the policy. The last two col- umns in table 4 show how changes in the policy variable (adopting a below-market-rate housing ordinance) alone affect the quantity of hous- ing. Eliminating the unobserved fixed effect by differencing the data switches the sign of the policy variable from positive to negative (though most are statistically insignificant with- out control variables). This switch in sign of d1 provides strong evidence of the importance of eliminating the unobserved fixed city effect. The negative impact increases in size and statistical significance when control variables are added to the first-difference model. Dependent Variable: ln(Housing Units) Level models for 1990 data Level models for 2000 data First-difference models (2000–1990) Policy Variable Coefficient of Policy Variable Policy variable Coefficient of Policy Variable Policy variable Coefficient of Policy Variable iz1985 .777 iz1995 .665 iz95delta -.045 iz1986 .751 iz1996 .614 iz96delta -.024 iz1987 .751 iz1997 .585 iz97delta -.027 iz1988 .679 iz1998 .585 iz98delta -.038 iz1989 .653 iz1999 .618 iz99delta -.051 Table 4 Summary of Policy Coefficients from Fifteen Regressions on the Quantity of Housing by Model and by Lag Year the independent institute14 | Tables 3 and 4 indicate the importance of differencing the data and removing the unob- served fixed city effect.19 The next set of regres- sions in table 5 report first-difference estimates for housing prices for the five-year and one year lag while adding other control variables that may change over time.20 The other models (using lag periods iz96delta, iz97delta, and iz98delta) yielded similar results. Adding income, whether median household income or per capita income, increases the size of the estimated policy effect. All policy estimates of d1 are larger than 0.20, suggesting that cities that impose an affordable housing mandate drive up prices by more than 20 percent. Dropping the insignificant variables and adjusting for heteroscedasticity had little impact on the policy and income variables. The final set of results in table 6 reports the estimated effects on housing quantity for the same lag periods as the price estimates. The results are nearly identical for the other lag peri- ods (iz96delta, iz97delta, and iz98delta). Adding control variables increases the policy impact and its statistical significance. Substituting the num- ber of households for the number of units as the dependent variable does not alter the main results. Adjusting for heteroscedasticity did increase the statistical significance levels slightly for the policy variable. The negative policy coef- ficients (-0.104 and -0.097) suggest that cities that impose an affordable housing mandate reduce housing units by more than 10 percent. Table 5 Regression Results of How Below-market Housing Mandates Affect the Price of Housing: First-difference Model with Control Variables Dependent Variable: ln(average price 2000/1990) Independent Variable Coefficients and (Standard Errors) Coefficients and (Standard Errors) N=431 N=431 Constant 0.001 (0.025) -0.009 (0.025) iz95delta 0.228*** (0.038) iz99delta 0.217*** (0.037) median income 0.173*** (0.0126) 0.178*** (0.0125) density -0.007 (0.011) -0.008 (0.011) population -0.0017 (0.00661) -0.00112 (0.00662) rent %-0.002 (0.005) -0.003 (0.005) Adj. R-Squared 0.4332 0.4300 Below-Market Housing Mandates As Takings | 15 5. Conclusion Our research provides answers to two important questions: How much do below-market housing mandates cost developers, and do below-market housing mandates improve housing affordabil- ity? After showing that below-market housing mandates cost developers hundreds of thou- sands of dollars for each unit sold, we discussed how developers do not receive compensation in this amount. Next we investigated how these policies affected the supply of housing. Using panel data and first difference estimates, we found that below-market housing mandates lead to decreased construction and increased prices. Over a ten-year period, cities that imposed a below-market housing mandate on average ended up with 10 percent fewer homes and 20 percent higher prices. These results are highly significant. The assertion by the court in Home Builders Association v. Napa that “the ordinance will necessarily increase the supply of affordable housing” is simply untrue. The justification for the decision that below-market housing mandates are not a tak- ing rests on some extremely questionable eco- nomic assumptions. We are not sure about the amount of economics knowledge of Judges Scott Snowden, J. Stevens, and J. Simons. Below-market housing mandates are simply a type of price control, and nearly every econo- mist agrees that price controls on housing lead to a decrease in quantity and quality of hous- ing available (Kearl et al., 1979, p.28). Because these price controls apply to a percentage of new housing, and builders must comply with them Table 6 Regression Results of How Below-market Housing Mandates Affect the Quality of Housing: First-difference Model with Control Variables Dependent Variable: ln(units 2000–1990) Independent Variable Coefficients and (Standard Errors) Coefficients and (Standard Errors) N=431 N=431 Constant -0.056** (0.023) -0.054** (0.023) iz95delta -0.104** (0.042) iz99delta -0.097** (0.041) median income 0.0683*** (0.0132) 0.0660*** (0.0131) density 0.113* (0.011) 0.114 (0.011) population 0.0233* (0.00729) -0.0230* (0.00729) Adj. R-Squared 0.2921 0.2911 Note: *, **,*** denotes significance at the .10, .05, .01 levels, two-tailed test. the independent institute16 | if they want to build market-rate housing, price controls also will affect the supply of market-rate housing. Because price controls act as a tax on new housing, we would expect a supply shift leading to less output and higher prices for all remaining units. New names for price controls, like “inclu- sionary zoning,” make the policy sound innocu- ous or even beneficial (who can be against a policy of inclusion?), but in reality the program is a mandate that imposes significant costs on a minority of citizens. The costs of below-market housing mandates are borne by developers and other new homebuyers who receive little or no compensation. From this perspective, below- market housing mandates are a taking no dif- ferent in substance from an outright taking under eminent domain. Below-market housing mandates represent the sort of abuse the Lucas Court forewarned, and they should rightly be considered a taking. In terms of economics, below-market housing mandates only differ from an outright taking in degree—there is not a “total taking” but a partial taking and clearly a diminution of value without any compensation. The amount of harm imposed by below-market housing mandates should inform their status under the law. References Allen, Charlotte. 2005. “A Wreck of a Plan; Look at How Renewal Ruined SW,” Washington Post, July 17, 2005. Calavita, Nico, Kenneth Grimes, and Alan Mallach. 1997. “Inclusionary Housing in California and New Jersey: a Comparative Analysis.” Housing Policy Debate, 8(1): pp.109–142. California Coalition for Rural Housing and Non-profit Housing Association of Northern California. 2003. Inclusionary Housing in California: 30 years of In- novation. Sacramento: California Coalition for Rural Housing and Non-profit Housing Association of Northern California. Construction Industry Research Board. Burbank, CA: Building Permit Data 1970–2003, at www.cirbdata. com. Dieterich, Andrew G. 1996. “An Egalitarian Market: The Economics of Inclusionary Zoning Reclaimed.” Ford- ham Urban Law Journal 24: pp. 23–104. Ely, James W. Jr. 2005. “’Poor Relation’ Once More: The Supreme Court and the Vanishing Rights of Property Owners.” Cato Supreme Court Review: pp. 39–69. Fischel, William A. 2006. “Before Kelo,” Regulation (Win- ter 2005–06): pp. 32–35. Follain, Jr., James R. 1979. “The Price Elasticity of the Long-Run Supply of New Housing Construction,” Land Economics 55(2): pp. 190–199. Gelinas, Nicole. 2005. “They’re Taking Away Your Property for What? The Court’s Eminent-Domain Ruling is Use- less as Well As Unjust.” City Journal (Autumn 2005). Glaeser, Edward L., Joseph Gyourko, and Raven E. Saks. 2005. “Why Have Housing Prices Gone Up?” Ameri- can Economics Review 95(2): pp. 329–33. Glaeser, Edward L., Joseph Gyourko, and Raven E. Saks. 2005. “Why Is Manhattan So Expensive? Regulation and the Rise in Housing Prices.” The Journal of Law & Economics, 48:2: pp. 331– 369. Green, Richard K Stephen Malpezzi, and Stephen K. Mayo. 2005. “Metropolitan-Specific Estimates of the Price Elasticity of Supply of Housing, and Their Sources.” American Economic Review 95(2): pp. 334–39. Kautz, Barbara Erlich. 2002. “In Defense of Inclusionary Zoning: Successfully Creating Affordable Housing.” University of San Francisco Law Review, 36:4: pp. 971–1,032. Kearl, J. R., Clayne L. Pope, Gordon C. Whiting, and Larry T. Wimmer. 1979. “A Confusion of Econo- mists?” American Economic Review 69: pp. 28–37. Mayer, Christopher J., and C. Tsuriel Somerville, 2000. “Residential Construction: Using the Urban Growth Model to Estimate Housing Supply.” Journal of Urban Economics 48: pp. 85–109. National Association of Home Builders/Wells Fargo. 2007. “The NAHB/Wells Fargo Housing Opportunity Index.” Washington, DC: National Association of Home Builders. Non-Profit Housing Association of Northern California. 2007. Affordable by Choice: Trends in California Inclu- sionary Housing Programs. San Francisco: Non-Profit Housing Association of Northern California. Pipes, Richard. 1999. Property and Freedom. New York: Alfred A. Knopf. Below-Market Housing Mandates As Takings | 17 Powell, Benjamin, and Edward Stringham. 2004a. “Hous- ing Supply and Affordability: Do Affordable Housing Mandates Work?” Reason Policy Study 318. Powell, Benjamin, and Edward Stringham. 2004b. “Do Af- fordable Housing Mandates Work? Evidence from Los Angeles County and Orange County.” Reason Policy Study 320. Powell, Benjamin, and Edward Stringham. 2004c. “Afford- able Housing in Monterey County. Analyzing the Gen- eral Plan Update and Applied Development Econom- ics Report.” Reason Policy Study 323. Powell, Benjamin, and Edward Stringham, 2005. “The Economics of Inclusionary Zoning Reclaimed: How Effective are Price Controls?” Florida State University Law Review 33(2). Quigley, John M, and Steven Raphael. 2004. “Is Housing Unaffordable? Why Isn’t It More Affordable?” The Jour- nal of Economic Perspectives 18(1): pp. 191–214. Quigley, John M. and Steven Raphael. 2005. “Regulation and the High Cost of Housing in California.” Ameri- can Economic Review 95(2):pp. 323–28. Quigley, John M., and Larry A. Rosenthal. 2004. “The Ef- fects of Land-Use Regulation on the Price of Housing: What do We Know? What Can We Learn?” Berkeley Program on Housing and Urban Policy, Work- ing Paper 1052. At http://ideas.repec.org/p/cdl/ bphupl/1052.html. RAND California Statistics, at www.ca.rand.org/cgi-bin/ homepage.cgi. Riches, Erin. 2004. Still Locked out 2004: California’s Af- fordable Housing Crisis. Sacramento: California Budget Project. Richer, Jerrell. 1995. “Explaining the Vote of Slow Growth.” Public Choice 82: pp. 207–23. Rubenfeld, Jed. 1993. “Usings.” Yale Law Review, 102:5 (March): pp. 1,077–1,163. Said, Carolyn. 2007. “Home Prices Rise in July Even as Sales Fall to 12-Year Low.” San Francisco Chronicle, August 16, 2007. Somin, Ilya. 2004. “Overcoming Poletown: County of Wayne v. Hathcock, Economic Development Takings, and the Future of Public Use.” Michigan State Law Review 4: 1,005–40. Tetreault, Bernard. 2000. “Arguments Against Inclusionary Zoning you can Anticipate Hearing.” New Century Housing, vol. 1, no. 2 (Oct. 2000): pp. 17–20. Thorson, James A. 1997. “The Effect of Zoning on Housing Construction.” Journal of Housing Construction, 6: pp. 81–91. Westray, Laura L. 1988. “Are Landlords Being Taken by the Good Cause Eviction Requirement?” Southern Califor- nia Law Review, 62: p.321. Wooldridge, Jeffrey M. 2006. Introductory Econometrics, A Modern Approach, 3rd ed. Thomson South-Western. Cases Cited Action Apartments Association v. City of Santa Monica, 123 Cal. App. 4th 47 (2007) Berman et al. v. Parker et al., 348 U.S. 26 (1954). Hawaii Housing Authority v. Midkiff, 467 U.S. 229 (1984). Home Builders Association of Northern California v. City of Napa, 89 Cal. App. 4th 897 (modified and republished 90 Cal. App. 4th 188) (2001). Kelo et al. v. City of New London, Connecticut, 545 U.S. 469 (2005). Lucas v. South Carolina Coastal Council, 505 U.S. 1003 (1992). Poletown Neighborhood Council v. Detroit, 410 Mich. 616, 304 N.W. 2d 455 (1981). Notes 1 Assuming a 30-year fixed-interest-rate mortgage with an interest rate of 6.3 percent. 2 For details about the program, see California Coali- tion for Rural Housing and Non-Profit Housing Association of Northern California (2003) and Powell and Stringham (2004a). 3 For review of the literature, see Powell and Stringham (2005). 4 “The country that became the United States was unique in world history in that it was founded by individuals in quest of private property. . . . [T]he conviction that the protection of property was the main function of government, and its corollary that a government that did not fulfill this obligation forfeited its mandate, acquired the status of a self-evident truth in the minds of the American colonists.” Pipes (1999, p.240). 5 Lucas v. South Carolina Coastal Council, 505 U.S. 1003 (1992). 6 Pennsylvania Coal v. Mahon 260, U.S. 393 (1922). 7 As 1egal scholar James Ely writes, “In his famous 1792 essay James Madison perceptively warned people against government that ‘indirectly violates their property, in their actual possessions.’ Although Madison anticipated the regulatory takings doctrine, the modern doctrine began to take shape in the last decades of the nineteenth century. For example, in a treatise on eminent domain published in 1888, John the independent institute18 | Lewis declared that when a person was deprived of the possession, use, or disposition of property ‘he is to that extent deprived of his property, and, hence . . . his property may be taken, in the constitutional sense, though his title and possession remain undisturbed.’ Likewise, in 1891 Justice David J. Brewer pointed out that regulation of the use of property might destroy its value and constitute the practical equivalent of outright appropriation. While on the Supreme Judi- cial Court of Massachusetts, Oliver Wendell Holmes also recognized that regulations might amount to a taking of property. ‘It would be open to argument at least,’ he stated, ‘that an owner might be stripped of his rights so far as to amount to a taking without any physical interference with his land.’” (Ely, 2005, p.43, footnotes in original omitted.) 8 Home Builders Association of Northern California v. City of Napa (2001), p. 188. 9 Home Builders Association of Northern California v. City of Napa (2001), pp. 195–6. 10 The California Coalition for Rural Housing and Non-profit Housing Association of Northern Califor- nia (2003, p.3) stated, “These debates, though fierce, remain largely theoretical due to the lack of empirical research.” 11 Marin County is one of the highest-income and most costly areas in the San Francisco Bay Area. 12 http://www.co.marin.ca.us/EFiles/Docs/CD/ PlanUpdate/07_0430_IT_070430091111.pdf (accessed August 19, 2007). To simplify the specif- ics, developers have the choice of selling 60 percent of homes to low-income households or 50 percent of homes to very-low-income households, which calculates to roughly the same loss of revenue, so for simplicity we will focus on the latter scenario. 13 Median income for a household of four is $91,200, so a household earning 80 percent of median income earns $73,696, and a household earning 60 percent of the median income earns $55,272. The specific affordability price control formula will depend on cer- tain assumptions (for example, the level of the interest rate in the formula), but using some standard assump- tions we can create an estimate (assuming homes will be financed with 0 percent down, a 30-year, fixed-rate mortgage, and an interest rate of 7 percent, and as- suming that 26 percent of income will pay mortgage payments and 4 percent of income will pay for real estate taxes and other homeowner costs). This formula gives us how much a household in each income level could afford and the level of the price controls. In Marin County, a home sold to a four-person household earning 80 percent of median income could be sold for no more than $240,003, and a home sold to a four-person household earning 60 percent of the median income could be sold for no more than $180,002. The price controls may be set at stricter levels, depending on the city ordinance. For example, the City of Tiburon sets price controls for “affordabil- ity” much more strictly than the above formula. Its ordinance assumes an interest rate of 9.5 percent and assumes that 25 percent of income can be devoted to a mortgage. According to Tiburon’s ordinance, a “moderate,” price-controlled home can be sold for no more than $109,800. 14 Home Builders Association of Northern California v. City of Napa (2001), p.194. 15 Powell and Stringham (2005) discuss this issue in depth. 16 Policy on Achieving Higher Residential Densities in Multiple-Family Zones, (September 13, 2005). 17 For those readers unfamiliar with semilog models, d1 provides an interpretation of the policy variable as a percentage change. The estimate of d1 is interpreted as the approximate percentage change in Y for cities that pass an ordinance. When the estimate of d1 is large (greater than 10 percent), the more accurate estimate is %ΔY = exp(d1)–1. 18 The first difference model is the fixed-effects model when there are two time periods. 19 Controlling for the endogeneity of the policy variable will have little or no impact. The data reveal that cities that passed an ordinance also have higher housing prices on average. It may be that higher-priced cit- ies are more likely to pass an ordinance. Given our results, we have some doubts about whether this will impact our conclusion. First we lagged the policy variable from one to five years and found very little variation in the OLS estimates. A lag of five years (for a potential dependent variable) should reduce or eliminate the potential bias. Second, the first-differ- ence approach reduced the price effect and signifi- cantly changed the output effect by controlling for unobserved fixed effects. Finally, there are some limits to finding instrumental variables for a first-difference model. Clearly it would not be appropriate to use any of the 2000 data to control for policies passed in earlier years. One could use the 1990 census data, but even here there are some cities that passed the policy prior to 1990. For these reasons, we believe control- ling for endogeneity will not change the basic results. 20 The income and population variables are rescaled in units of ten thousand to simplify the coefficient presentation. Below-Market Housing Mandates As Takings | 19 About the Authors Tom Means is Research Fellow at the Independent Institute and Professor of Economics at San Jose State University and serves as Director of the Center for Economic Education. He earned his Ph.D. in economics from UCLA in 1983 and has been with San Jose State for twenty-five years. Professor Means teaches the grad- uate microeconomics and econometrics seminars. His research focuses on applied economics in the areas of public choice, labor economics, and forensic economics. He has published in a variety of journals, including Public Choice, The Southern Economics Journal, and The Journal Of Forensic Economics. Since 2004 he has served on the City Council of Mountain View and is currently Vice-Mayor. Edward P. Stringham is Research Fellow at the Independent Institute, Associate Professor of Economics at San Jose State University, President of the Association of Private Enterprise Education, editor of the Journal of Private Enterprise, editor of two books, and author of twenty articles in refereed journals including the Journal of Institutional & Theoretical Economics, Quarterly Review of Economics & Finance, and Journal of Labor Research. Stringham has been discussed on more than 100 broadcast stations including CBS, CNBC, CNN, Fox, Headline News, NPR, and MTV and in hundreds of newspapers world- wide. Stringham earned his Ph.D. from George Mason University in 2002, and has won the Templeton Culture of Enterprise Best Article Award, Paper of the Year Award from the Association of Private Enterprise, Best Article Award from the Society for the Development of Austrian Economics, Second Place in the Independent Institute Garvey Fellowship Awards, and Distinguished Young Scholar Award from the Liberalni Institut and the Prague School of Economics. Edward J. Lopez is Research Fellow at the Independent Institute and Professor of Law and Economics at San Jose State University. His main area of research is in public choice and law and economics, with emphases on empirical models of creative expression, technological innovation, political ideology, and political insti- tutions. Additional areas of research include antitrust regulation, property rights, campaign finance, term limits, and federal fiscal policy. Professor Lopez has taught courses in microeconomics, macroeconomics, law and economics, public finance, public choice, and mathematical economics. He earned a Ph.D. from George Mason University in 1997. Professor Lopez joined the faculty of San Jose State in the fall of 2005. Previously he held appointments at the University of North Texas and George Mason University, and he served as staff economist on the Joint Economic Committee of Congress. THE INDEPENDENT INSTITUTE is a non-profit, non-par- tisan, scholarly research and educational organization that sponsors comprehensive studies of the political economy of critical social and economic issues. THE IndEpEndEnT InSTITUTE 100 Swan Way, Oakland, California 94621-1428, U.S.A. Telephone: 510-632-1366 • Facsimile: 510-568-6040 E-mail: info@independent.org • Website: http://www.independent.org Founder and President David J. Theroux Research Director Alexander Tabarrok Board of Advisors, Center on Entrepreneur- ial Innovation Bruce L. Benson Florida state university George Bittlingmayer u niversity oF Kansas Peter J. Boettke, GeorGe Mason university Reuven Brenner, McGill university, c anada Enrico Colombatto, u niversity oF torino; international centre For econoMic r esearch, i taly Price V. Fishback, university oF a rizona Peter Gordon, university oF southern c aliFornia P. J. Hill, Wheaton colleGe Randall G. Holcombe Florida state university Daniel B. Klein GeorGe Mason university Peter G. Klein university oF Missouri Chandran Kukathas university oF utah; australian deFence Force acadeMy, australia Robert A. Lawson c apital university Stan Liebowitz university oF texas at dallas Stephen E. Margolis north carolina state university Roger E. Meiners u niversity oF texas, a rlinGton Michael C. Munger duK e university Robert H. Nelson u niversity oF Maryland Benjamin Powell suFFolK university William F. Shughart II u niversity oF Mississippi Randy T. Simmons u tah state university Russell S. Sobel West v irGinia university Gordon Tullock GeorGe Mason university Lawrence H. White u niversity oF Missouri at st. louis The Center on Entrepreneurial Innovation pursues research into entre- preneurship and the dynamic process of markets and technological innova- tion, without regard to prevailing popular or political biases and trends. The goal is to explore important areas that might otherwise be ignored, includ- ing questions normally considered “out-of-the-box” or controversial, but which might well be crucial to understanding and getting at real answers and lasting solutions. As a result, the Center aims to cut through the intellectual poverty, noise, and spin of special-interest-driven public policy in the U.S. and elsewhere. CENTER ON Entrepreneurial Innovation INDEPENDENT STUDIES IN POLITICAL ECONOMY For further information and a catalog of publications, please contact: THE INDEPENDENT INSTITUTE 100 Swan Way, Oakland, California 94621-1428, U.S.A. 510-632-1366 · Fax 510-568-6040 · info@independent.org · www.independent.org THE ACADEMY IN CRISIS: The Political Economy of Higher Education | Ed. by John W. Sommer AgAINST LEVIATHAN: Government Power and a Free Society | Robert Higgs A LIENATION AND THE SOVIET ECONOMY: The Collapse of the Socialist Era | Paul Craig Roberts A MERICAN HEALTH C ARE: Government, Market Processes and the Public Interest | Ed. by Roger Feldman ANARCHY AND THE LAW: The Political Economy of Choice | Ed. by Edward P. Stringham ANTITRUST AND MONOPOLY: Anatomy of a Policy Failure | D. T. Armentano ARMS, POLITICS, A ND THE ECONOMY: Historical and Contemporary Perspectives | Ed. by Robert Higgs BEYOND POLITICS: Markets, Welfare and the Failure of Bureaucracy | William Mitchell & Randy Simmons THE CAPITALIST R EVOLUTION IN L ATIN AMERICA | Paul Craig Roberts & Karen Araujo THE CHALLENgE OF LIBERTY: Classical Liberalism Today | Ed. by Robert Higgs & Carl P. Close CHANgINg THE gUARD: Private Prisons and the Control of Crime | Ed. by Alexander Tabarrok THE CHE gUEVARA MYTH AND THE FUTURE OF LIBERTY | Alvaro Vargas Llosa CUTTINg gREEN TAPE: Toxic Pollutants, Environmental Regulation and the Law | Ed. by Richard Stroup & Roger E. Meiners DEPRESSION, WAR, AND COLD WAR: Studies in Political Economy | Robert Higgs THE DIVERSITY M YTH: Multiculturalism and Political Intolerance on Campus | David O. Sacks & Peter A. Thiel DRUg WAR CRIMES: The Consequences of Prohibition | Jeffrey A. Miron ELECTRIC CHOICES: Deregulation and the Future of Electric Power | Ed. by Andrew Kleit THE EMPIRE HAS NO CLOTHES: U.S. Foreign Policy Exposed | Ivan Eland ENTREPRENEURIAL ECONOMICS: Bright Ideas from the Dismal Science | Ed. by Alexander Tabarrok FAULTY TOWERS: Tenure and the Structure of Higher Education | Ryan Amacher & Roger Meiners THE FOUNDERS' SECOND AMENDMENT | Stephen P. Halbrook FREEDOM, FEMINISM, A ND THE STATE | Ed. by Wendy McElroy H AzARDOUS TO OUR HEALTH?: FDA Regulation of Health Care Products | Ed. by Robert Higgs HOT TALk, COLD SCIENCE: Global Warming’s Unfinished Debate | S. Fred Singer JUDgE AND JURY: American Tort Law on Trial | Eric Helland & Alex Tabarrok LIBERTY FOR LATIN AMERICA: How to Undo Five Hundred Years of State Oppression | Alvaro Vargas Llosa LIBERTY FOR WOMEN: Freedom and Feminism in the Twenty-first Century | Ed. by Wendy McElroy M Ak INg POOR NATIONS RICH: Entrepreneurship and the Process of Economic Development | Ed. by Benjamin Powell MARk ET FAILURE OR SUCCESS: The New Debate | Ed. by Tyler Cowen & Eric Crampton MONEY AND THE NATION STATE: The Financial Revolution, Government, and the World Monetary System | Ed. by Kevin Dowd & Richard H. Timberlake, Jr. NEITHER LIBERTY NOR SAFETY: Fear, Ideology, and the Growth of Government | Robert Higgs & Carl P. Close OPPOSINg THE CRUSADER STATE: Alternatives To Global Interventionism | Ed. by Robert Higg & Carl P. Close OUT OF WORk: Unemployment and Government in Twentieth-Century America | Richard K. Vedder & Lowell E. Gallaway PLOWSHARES AND POR k BARRELS: The Political Economy of Agriculture | E. C. Pasour, Jr. & Randal R. Rucker A POVERTY OF R EASON: Sustainable Development and Economic Growth | Wilfred Beckerman PRIVATE R IgHTS & PUBLIC ILLUSIONS | Tibor R. Machan R ECLAIMINg THE A MERICAN R EVOLUTION: The Kentucky & Virginia Resolutions and Their Legacy | William J. Watkins, Jr. R EgULATION A ND T HE R EAgAN ERA: Politics, Bureaucracy and the Public Interest | Ed. by Roger Meiners & Bruce Yandle RESTORINg FREE SPEECH AND LIBERTY ON CAMPUS | Donald A. Downs R ESURgENCE OF THE WARFARE STATE: The Crisis Since 9/11 | Robert Higgs RE-THINk INg gREEN: Alternatives to Environmental Bureaucracy | Ed. by Robert Higgs & Carl P. Close SCHOOL CHOICES: True and False | John Merrifield STRANgE BREW: Alcohol and Government Monopoly | Douglas Glen Whitman STREET SMART: Competition, Entrepreneurship, and the Future of Roads | Ed. by Gabriel Roth TA x INg CHOICE: The Predatory Politics of Fiscal Discrimination | Ed. by William F. Shughart, II TA x INg ENERgY: Oil Severance Taxation and the Economy | Robert Deacon, Stephen DeCanio, H. E. Frech, III, & M. Bruce Johnson THAT EVERY M AN BE ARMED: The Evolution of a Constitutional Right | Stephen P. Halbrook TO SERVE AND PROTECT: Privatization and Community in Criminal Justice | Bruce L. Benson THE VOLUNTARY CITY: Choice, Community and Civil Society | Ed. by David T. Beito, Peter Gordon & Alexander Tabarrok TWILIgHT WAR: The Folly of U.S. Space Dominance | Mike Moore W INNERS, LOSERS & MICROSOFT: Competition and Antitrust in High Technology | Stan J. Liebowitz & Stephen E. Margolis W RITINg OFF IDEAS: Taxation, Foundations, & Philanthropy in America | Randall G. Holcombe Additional copies of this Independent Policy Report are available for $10.00 each. To order, visit www.independent.org or call (510) 632-1366. The Independent Institute · 100 Swan Way · Oakland, CA 94621 · info@independent.org · www.independent.org The Independent Institute is a non-profit, non-partisan, scholarly research and educational organization that sponsors comprehensive studies of the political economy of critical social and economic issues. The politicization of decision-making in society has too often confined public debate to the narrow reconsideration of existing policies. Given the prevailing influence of partisan interests, little social innovation has occurred. In order to understand both the nature of and possible solutions to major public issues, the Independent Institute’s program adheres to the highest standards of independent inquiry and is pursued regardless of prevailing political or social biases and conventions. The resulting studies are widely distributed as books and other publi- cations, and are publicly debated through numerous conference and media programs. In pursuing this uncommon independence, depth, and clarity, the Independent Institute seeks to push at the frontiers of our knowledge, redefine the debate over public issues, and foster new and effective directions for government reform. By Benjamin Powell, Ph.D and Edward Stringham, Ph.D Project Director: Adrian T. Moore, Ph.D June 2004 DO AFFORDABLE HOUSING MANDATES WORK? EVIDENCE FROM LOS ANGELES COUNTY AND ORANGE COUNTY 320 POLICY STUDY A division of the Los Angeles-based Reason Foundation, Reason Public Policy Institute is a nonpartisan public policy think tank promoting choice, competition, and a dynamic market economy as the foundation for human dignity and progress. Reason produces rigorous, peer-reviewed research and directly engages the policy process, seek- ing strategies that emphasize cooperation, flexibility, local knowledge, and results. Through practical and innovative approaches to complex problems, Reason seeks to change the way people think about issues, and promote policies that allow and encourage individuals and volun- tary institutions to flourish. Reason Foundation Reason Foundation’s mission is to advance a free society by develop- ing, applying, and promoting libertarian principles, including indi- vidual liberty, free markets, and the rule of law. We use journalism and public policy research to influence the frameworks and actions of poli- cymakers, journalists, and opinion leaders. Reason Foundation is a tax-exempt research and education organiza- tion as defined under IRS code 501(c)(3). Reason Foundation is sup- ported by voluntary contributions from individuals, foundations, and corporations. The views are those of the author, not necessarily those of Reason Foundation or its trustees. Copyright © 2004 Reason Foundation. Photos used in this publication are copyright © 1996 Photodisc, Inc. All rights reserved. Reason Public Policy Institute Policy Study No. 320 Do Affordable Housing Mandates Work? Evidence from Los Angeles County and Orange County By Benjamin Powell, Ph.D. and Edward Stringham, Ph.D Project Director: Adrian T. Moore, Ph.D Executive Summary alifornia and many urban areas nationwide face a housing affordability crisis. New housing production has chronically failed to meet housing needs, causing housing prices to escalate. Faced with demands to “do something” about the housing affordability crisis, many local governments have turned to “inclusionary zoning” ordinances in which they mandate that developers sell a certain percentage of the homes they build at below-market prices to make them affordable for people with lower incomes. The number of cities with affordable housing mandates has grown rapidly, to about 10 percent of cities over 100,000 population as of the mid-90s, and many advocacy groups predict the trend will accelerate in the next five years. California was an early leader in the adoption of inclusionary zoning, and its use there has grown rapidly. Between 1990 and 2003, the number of California communities with inclusionary zoning more than tripled—from 29 to 107 communities—meaning about 20 percent of California communities now have inclusionary zoning. Inclusionary zoning attempts to deal with high housing costs by imposing price controls on a percentage of new homes. During the past 20 years, a number of publications have debated the merits of inclusionary zoning programs. Nevertheless, as a recent report observed, “These debates, though fierce, remain largely theoretical due to the lack of empirical research.” Our recent report, Housing Supply and Affordability: Do Affordable Housing Mandates Work?, filled the empirical research void. We measured the actual performance of these ordinances in the San Francisco Bay Area. This study follows up on our previous study by examining data from communities in Los Angeles County and Orange County to evaluate the effects of inclusionary zoning and examine whether it is an C effective public policy response to high housing prices. In Los Angeles and Orange Counties, 13 cities have an affordable housing mandate. These communities vary in size and density with different income levels and demographics, so they provide a good sample to tell us how inclusionary zoning is working in Southern California. These are our findings: Inclusionary Zoning Produces Few Units Since its inception, inclusionary zoning has resulted in few affordable units. The 13 Los Angeles and Orange County cities with inclusionary zoning have produced only 6,379 affordable units, with 70 percent of those units being produced in Irvine. After passing an ordinance, the median city produces less than eight affordable units per year. Inclusionary zoning cannot meet the area’s affordable housing needs. Inclusionary Zoning Has High Costs Inclusionary zoning imposes large burdens on the housing market. For example, if a home could be sold for $500,000 dollars but must be sold for $200,000, the revenue from the sale is $300,000 less. In half the Los Angeles County and Orange County jurisdictions this cost associated with selling each inclusionary unit exceeds $575,000. In current prices the cost of inclusionary zoning in the average jurisdiction is $298 million, bringing the total cost for all inclusionary units in Los Angeles and Orange County to date to $3.9 billion. Inclusionary Zoning Makes Market-priced Homes More Expensive Who bears the costs of inclusionary zoning? The effective tax of inclusionary zoning will be borne by some combination of market-rate homebuyers, landowners, and builders. How much of the burden is borne by market-rate buyers versus landowners and builders is determined by each group’s relative responsiveness to price changes. We estimate that inclusionary zoning causes the price of new homes in the median city to increase by $33,000 to $66,000. In high market-rate cities such as San Juan Capistrano and Laguna Beach we estimate that inclusionary zoning adds more than $100,000 to the price of each new home. Inclusionary Zoning Restricts the Supply of New Homes Inclusionary zoning drives away builders, makes landowners supply less land for residential use, and leads to less housing for homebuyers—the very problem it was instituted to address. We find that new housing production drastically decreases the year after cities adopt inclusionary zoning. For all 13 cities average production of housing fell the year following the adoption of inclusionary zoning. In the eight cities with data for seven years prior and seven years following inclusionary zoning, 17,296 fewer homes were produced during the seven years after the adoption of inclusionary zoning. In those cities 770 “affordable” units were produced. One must question whether 770 units are worth the cost in terms of 17,296 fewer homes. By discouraging production of 17,296 homes in those eight cities, $11 billion worth of housing was essentially destroyed. Inclusionary Zoning Costs Government Revenue Price controls on new development lower assessed values, thereby costing state and local governments lost tax revenue each year. Because inclusionary zoning restricts resale values for a number of years, the loss in annual tax revenue can become substantial. The total present value of lost government revenue due to Los Angeles and Orange County inclusionary zoning ordinances is upwards of $752 million. Price Controls Do Not Address the Cause of the Affordability Problem Price controls fail to get to the root of the affordable housing problem. Indeed, by causing fewer homes to be built they actually make things worse. The real problem is government restrictions on supply. Supply has not kept up with demand due to these artificial restrictions. One recent study found that 90 percent of the difference between physical construction costs and the market price of new homes can be attributed to land use regulation. The solution is to allow more construction. When the supply of homes increases, existing homeowners often upgrade to the newly constructed homes. This frees up their prior homes for other families with lower income. Inclusionary zoning restricts this upgrade process by slowing or eliminating new construction. With fewer new homes available, middle- and upper-income families bid up the price of the existing stock of homes, thus making housing less affordable for everyone. Conclusion Inclusionary zoning has failed to produce a significant number of affordable homes due to the incentives created by the price controls. Even the few inclusionary zoning units produced have cost builders, homeowners, and governments greatly. By restricting the supply of new homes and driving up the price of both newly constructed market-rate homes and the existing stock of homes, inclusionary zoning makes housing less affordable. Inclusionary zoning ordinances will continue to make housing less affordable by restricting the supply of new homes. If more affordable housing is the goal, governments should pursue policies that encourage the production of new housing. Ending the price controls of inclusionary zoning would be a good start. Policy Study No. 320 Table of Contents Introduction................................................................................................................1 The Housing Market and Inclusionary Zoning in Los Angeles and Orange Counties .....3 Costs Associated with Below-Market Units ..................................................................6 A. Estimating the Effects of Price Controls by City...................................................................................9 B. Who Bears the Burden of Inclusionary Zoning?................................................................................11 C. The Effect of Price Controls on Housing Construction......................................................................15 The Fiscal Cost of Price Controls to State and Local Government...............................18 Conclusion................................................................................................................21 About the Authors.....................................................................................................22 Related Reason Foundation Studies...........................................................................23 Endnotes...................................................................................................................24 DO AFFORDABLE HOUSING MANDATES WORK? 1 Part I Introduction he number of cities with affordable housing mandates has grown rapidly, to about 10 percent of cities over 100,000 population as of the mid-90s, and many advocacy groups predict the trend will accelerate in the next five years.1 California was an early leader in the adoption of inclusionary zoning, and its use there has grown rapidly. Between 1990 and 2003, the number of California communities with inclusionary zoning more than tripled—from 29 to 107 communities—meaning about 20 percent of California communities now have inclusionary zoning.2 Thirteen cities in Los Angeles and Orange Counties have inclusionary zoning. The median price of new housing is $450,000 in Los Angeles County and $660,000 in Orange County.3 Such high prices affect all but the wealthiest families’ chances of buying a new home. Of metropolitan areas with more than one million residents, the Los Angeles-Long Beach Metropolitan Area and the Orange County Metropolitan Area respectively rank five and six as the least affordable areas in the nation. Table 1: Least Affordable Metropolitan Areas in the Nation Metro Area Least Affordable Metropolitan Areas Share of Homes Affordable for Median Incomes Family Income San Francisco, CA PMSA* 1 9.2% $86,100 San José, CA PMSA 2 20.1% $96,000 San Diego, CA MSA 3 21.6% $60,100 Oakland, CA PMSA 4 23.9% $74,500 Los Angeles-Long Beach, CA PMSA 5 34.4% $55,100 Orange County, CA PMSA 6 37.7% $75,600 Sacramento, CA PMSA 7 43.7% $57,300 Portland-Vancouver, OR-WA PMSA 8 46.6% $57,200 Boston, MA-NH PMSA 9 48.2% $74,200 Riverside-San Bernardino, CA PMSA 10 49.6% $50,300 New York, NY PMSA 11 49.9% $62,800 Miami, FL PMSA 12 58.1% $48,200 Denver, CO PMSA 13 59.6% $69,900 Bergen-Passaic, NJ PMSA 14 61.5% $78,900 Newark, NJ PMSA 15 61.1% $78,700 Source: Data are from the “Housing Opportunity Index: First Quarter 2002” (Washington, D.C.: National Association of Homebuilders), *PMSA and MSA are census designations meaning, respectively, Primary Municipal Statistical Area and Municipal Statistical Area. Faced with demands to “do something” about the region’s housing affordability crisis, many local governments have turned to inclusionary zoning ordinances. Inclusionary zoning is a name for artificially T 2 Reason Public Policy Institute lowering the price, and therefore the value, on a percentage of new homes. Builders and subsequent owners are forced to sell the homes so that they are “affordable” to specific income levels. The price controls are set using different formulas so that the “inclusionary” units will be affordable to either “Very Low,” “Low,” or “Moderate” income households, or some combination thereof. “Very Low” income is most often classified as up to 50 percent of county median income, “Low” as 50-80 percent of median, and “Moderate” as 80-120 percent of median. The percentage of units targeted as inclusionary units varies by jurisdiction, ranging from 5 to 25 percent of the new homes constructed in a project. Typically, the inclusionary units must be constructed within the project and be of the same size and quality as the market- rate units. Some jurisdictions exempt small developments while others require builders to pay an in-lieu fee for developments of 10 homes or fewer to get out from under the price controls. Still others allow in-lieu fees for projects of all sizes. Ostensibly, some jurisdictions also offer incentives for compliance. These can take the form of “density bonuses” (giving builders the option to increase the density of their developments instead of making more of the units affordable), fast-track permitting (speeding up the process of issuing permits for new development), fee waivers, or exemptions from growth controls. In a few voluntary inclusionary programs, incentives are offered in exchange for a builder committing to sell at the price- controlled rates. But most inclusionary zoning programs are mandatory, requiring all builders to participate. The proliferation of inclusionary zoning raises important public policy questions: Is it effective—does inclusionary zoning lead to a substantial increase in affordable housing production? Is it efficient—how do inclusionary zoning’s costs compare to its benefits? Is it equitable—does inclusionary zoning fairly apportion the cost of providing affordable housing? Until recently these questions had not been adequately addressed. During the past 20 years a number of publications debated the merits of inclusionary zoning programs. Nevertheless, as the 2003 report Inclusionary Housing in California: 30 Years of Innovation observed, “These debates, though fierce, remain largely theoretical due to the lack of empirical research.”4 Without knowing the economic and other real- world consequences of inclusionary zoning, policymakers have difficulty assessing the merits or faults of inclusionary zoning. Our recent report, Housing Supply and Affordability: Do Affordable Housing Mandates Work?, filled the empirical research void. 5 We measured the actual performance of these ordinances in the San Francisco Bay Area. We found that the San Francisco Bay Area inclusionary ordinances produced few “affordable” units, drove up the price of market-rate homes, and dramatically decreased the supply of new construction. Paradoxically, in the Bay Area, “affordable” housing mandates actually made most housing more expensive. The track record of affordable housing mandates in the Bay Area is consistent with the predictions of economic theory. Inclusionary zoning ordinances act like a tax on new development. Taxes decrease the supply of new housing and increase prices of the few homes that are built. This basic economic model should apply in other regions as well.6 This study follows up our prior one to see if the empirical record of inclusionary zoning is consistent with economic theory in another region of California. We use data from communities in Los Angeles County and Orange County to evaluate the effects of inclusionary zoning and examine whether it is an effective public policy in Southern California. We include in our analysis all the cities in these counties with inclusionary ordinances. Included are: Agoura Hills, Brea, Huntington Beach, Irvine, Laguna Beach, Long Beach, Monrovia, Pasadena, Rancho Palos Verdes, San Clemente, San Juan Capistrano, Santa Monica and West Hollywood. These communities have various sizes and densities with different income levels and demographics, so they provide a good sample to measure the effects of inclusionary zoning in Southern California. DO AFFORDABLE HOUSING MANDATES WORK? 3 Part 2 The Housing Market and Inclusionary Zoning in Los Angeles and Orange Counties number of studies document high housing prices and the affordability crisis in California.7 Offering a temperate climate, cultural and natural resources, and job growth, Los Angeles County and Orange County have become increasingly desirable places to live. The percentage of homes affordable to a family earning median income is only 34.4 percent for Los Angeles-Long Beach Metro and 37.7 percent for Orange Metro.8 Families earning less than median income have even fewer homes available in their price range. In response to the affordable housing crisis, 13 local governments in Los Angeles and Orange County have adopted inclusionary zoning requirements (Figure 1) and remaining cities now face loud calls from planners and advocacy groups to adopt inclusionary zoning as well. Figure 1: Number of Los Angeles County and Orange County Cities with Inclusionary Zoning 0 2 4 6 8 10 12 14 19761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001Number of Los Angeles County and Orange County Cities with Inclusionary Zoning A 4 Reason Public Policy Institute Table 2 shows the jurisdictional requirements and the number of price-controlled units produced by city. Several communities could not report how many affordable units had been produced under the program, demonstrating a simple unwillingness by city officials to keep track of how effective the policy is in spite of its costs. Our calculations of averages and costs exclude these cities. Table 2: Southern California Cities with Inclusionary Zoning City Year imposed Percent of new units under price controls Target levels VL=Very Low; L=Low; M=Moderate Number of price- controlled units produced by program Average number of price-controlled units produced per year since program inception Agoura Hills 1987 10 M 50 3.1 Brea 1993 10 VL, L, M 278 27.8 Huntington Beach 2001 10 L 313 156.5 Irvine 1977 5 VL, L, M 4,469 171.9 Laguna Beach 1985 25 VL, L, M 139 7.7 Long Beach 1992 5 M * * Monrovia 1990 20 M 280 21.5 Pasadena 1991 15 L, M 14 1.2 Rancho Palos Verdes 1997 5 VL, L * * San Clemente 1980 4 VL 627 27.3 San Juan Capistrano 1995 30 VL, L 196 24.5 Santa Monica 1985 10 VL, L * * West Hollywood 1986 20 L, M 13 0.8 Sources: California Coalition for Rural Housing and Non-Profit Housing Association of Northern California, Inclusionary Housing in California, (Sacramento, CA: California Coalition for Rural Housing), 2003;.and Calavita and Grimes, “Inclusionary Zoning in California: The Experience of Two Decades.” Journal of the American Planning Association v 64 no.2,1998, p. 152. * Inclusionary Housing in California does not report any units for these cities. Advocates of inclusionary zoning herald price controls as the solution to the affordability crisis. They point to the inclusionary units produced and declare the program to be a success. While the program has been a boon to the few families who luck out on getting the artificially reduced homes, the ripple effect distortion in the market caused by inclusion zoning is overwhelming, costing far more. Obviously, a more thorough assessment of inclusionary zoning is necessary. From an overall production perspective, how effective has inclusionary zoning been? The numbers do not look good. Compared to the region’s overall affordable housing needs for this period, inclusionary zoning clearly has not made a significant contribution to solving the region’s affordable housing crisis. For the 13 cities, the Southern California Association of Governments projects the current 7.5 year affordable housing need for very low, low, and moderate income households to be 12,460.9 But in the 27 years that inclusionary zoning has been implemented in Los Angeles and Orange Counties, inclusionary zoning has resulted in the production of only 6,379 affordable units. Of those, 4,469 were in Irvine, which built a number of the units to settle a lawsuit for not providing “affordable” housing. That averages to only 236 units per year, with 165 in Irvine and 71 in all other cities. Controlling for the length of time each program has been in effect, the average jurisdiction with inclusionary zoning produces only 34 units each year since adoption of its inclusionary zoning requirement. DO AFFORDABLE HOUSING MANDATES WORK? 5 The disparity between the regional housing need and inclusionary zoning production is shown in Figure 2. In Figure 2, the front (red) columns represent the average yearly production of affordable housing reported by cities (only for years when cities had inclusionary zoning) multiplied times 7.5, and the back (green) columns represent the 7.5 year need for affordable housing in the cities with inclusionary zoning. The number of units expected from inclusionary zoning does not meet most cities’ needs for affordable housing. Huntington Beach is the most notable exception. Laguna BeachRancho Palos VerdesAgoura HillsMonroviaWest HollywoodSan Juan CapistranoBreaLong BeachHuntington BeachPasadenaSanta MonicaSan ClementeIrvine0 1,000 2,000 3,000 4,000 5,000 6,000 Figure 2: Housing Needs Versus Expected Units Produced under Inclusionary Zoning "Affordable" units produced through inclusionary zoning. (Calculated for 7.5 years by multiplying average units per year produced under inclusionary zoning times 7.5.) 7.5 year housing needs according to the Southern California Association of Governments "Regional Housing Needs Assessment." From an overall production standpoint, inclusionary zoning has not been effective. Some advocates of inclusionary zoning respond to this poor record by calling for more vigorous and numerous restrictions. Instead, jurisdictions need to fundamentally reexamine if price controls are an effective way of producing more affordable housing. Policymakers should analyze the actual consequences of inclusionary zoning and judge whether the poor results achieved by inclusionary zoning are caused by the very nature of these laws. Looking at the number of below-market units created by programs only begins to reveal inclusionary zoning’s effect on affordability. Our findings suggest that inclusionary zoning actually leads to less housing and higher prices. 6 Reason Public Policy Institute Part 4 Costs Associated with Below-Market Units upporters often promote inclusionary zoning as a costless way of providing affordable housing. Many highlight the number of units produced under inclusionary zoning and then claim the program to be a success. But the costs of these units and programs are often missed. For example, West Hollywood has had inclusionary zoning since 1986, and the program has led to 13 affordable units. The initial reaction might be to consider the program worthwhile simply because 13 units were built. But accurately judging the efficacy of a program requires looking at its costs. What were the costs of producing each of those units? We all agree that the goal is to help low-income households, but we must recognize that some ways are better than others. If two methods cost the same amount but one helps more, we should choose the one that yields greater benefits. Or, if two methods yield the same benefits but one costs less, we should support the one with lower costs. Even though many cities have adopted inclusionary zoning, to date no one has comprehensively estimated the program costs. Without looking at the costs of inclusionary zoning, one cannot determine if better ways to provide affordable housing exist. By definition, whenever sellers must sell a unit at a government-set price, they cannot sell that unit at the market price. For example, for a home to be “affordable” to a low-income household in West Hollywood, we estimate that the home must be sold for $147,000. If a new home could be sold for $588,000 but must be sold for $147,000, the revenue from the sale is $441,000 less. When someone forgoes one opportunity to take another, economists refer to this as the “opportunity cost.” The opportunity cost of selling a “Low” priced unit for $147,000 is not selling the unit for $588,000, i.e., $441,000. Keep in mind that this does not measure production costs. Rather, it represents the lost revenue per sale of price-controlled units. First, let us consider the cost associated with each inclusionary unit by city. We calculate the cost for each unit by subtracting the regulated price from the market price.10 Most inclusionary zoning ordinances mandate that homes be affordable to some combination of very low income, low income and moderate income households. Very low income is typically defined by up to 50 percent of median, low income is defined by up to 80 percent of median, and moderate income is defined by up to 120 percent of median. The California Department of Housing and Community Development provides income levels for four-person households (Figure 3).11 S DO AFFORDABLE HOUSING MANDATES WORK? 7 Lo s A n g e le s C o u n ty O ra n g e C o u n ty $0 $10,000 $20,000 $30,000 $40,000 $50,000 $60,000 $70,000 $80,000 $90,000 $100,000 Figure 3: 2003 Income Levels for Four-Person Households Defined by California Department of Housing and Community Development Very Low Income Low Income Moderate Income Inclusionary zoning sets price controls such that homes can be “affordable” at the specified income levels. Table 3 indicates sample price controls for homes to be “affordable” to the four-person households in the respective income groups. We assume homes will be financed with 0 percent down, a 30-year fixed-rate mortgage, and an interest rate of 7 percent. We assume 26 percent of income will pay mortgage payments and 4 percent of income will pay for real estate taxes and other homeowner costs. This formula gives us how much a household in each income level could afford. We decided to use conservative assumptions so that we would not overestimate the costs of inclusionary zoning. Different jurisdictions use different formulas for calculating their price controls; actual price controls will differ accordingly. To the extent that families can afford less than our calculations assume or that jurisdictions set price controls more stringently than we assume, the costs of inclusionary zoning will be significantly higher than our estimates. Table 3: Sample Price Controls for Homes to be “Affordable” to Different Income Groups County Very Low Price Control Low Price Control Moderate Price Control Los Angeles County $91,837 $146,875 $215,265 Orange County $123,101 $184,001 $295,379 We can then compare the level of the price controls to the market price of homes. The more restrictive the price controls, the greater the cost for each unit. Figures 4, 5, and 6 compare the median price of existing homes in each county to our sample price controls. The heights of lower (red) bars represent the price controls: “very low” in Figure 4, “low” in Figure 5, and “moderate” in Figure 6. The top of the upper (green) 8 Reason Public Policy Institute bars represent the 2003 average market price of new homes by county. The difference between the market price and the price-controlled price (the height of the red bar) is the cost of providing the affordable unit. Figure 4: “Very Low” Price Controls Compared to Average Market Price by County $0 $100,000 $200,000 $300,000 $400,000 $500,000 $600,000 $700,000 Los Angeles County Orange County "Very Low" price control Cost Associated with selling "Very Low" unit Figure 5: “Low” Price Controls Compared to Average Market Price by County $0 $100,000 $200,000 $300,000 $400,000 $500,000 $600,000 $700,000 Los Angeles County Orange County "Low" price control Cost associated with selling "Low" unit DO AFFORDABLE HOUSING MANDATES WORK? 9 Figure 6: “Moderate” Price Controls Compared to Average Market Price by County $0 $100,000 $200,000 $300,000 $400,000 $500,000 $600,000 $700,000 Los Angeles County Orange County "Moderate" price control Cost associated with selling "Moderate" unit Comparing the figures, the “moderate” price controls are not as restrictive as the “low” price controls and impose less of a cost. When price controls are at the market price we would not count them as costly. In reality price controls set near the market price also cause builders to lose revenue because the price controls come with other restrictions. Inclusionary zoning ordinances almost always impose restrictions on the resale price of below-market units. The reasoning seems straightforward: the subsidized units should remain affordable for future buyers, and the initial buyers should not be able to cash out on the windfall profits of acquiring a price-controlled unit.12 These affordability controls limit appreciation to some formula based on inflation, or they simply mandate that the home be “affordable” to the equivalent income groups calculated at the time of sale. Resale price controls typically last 30 years or more and are renewed upon each sale. Because home ownership is a long- term commitment and affordability controls last a number of years, price-controlled homes are simply less valuable. A. Estimating the Effects of Price Controls by City By comparing the market price to the average level of the price controls in each city, we can estimate the average cost of each price-controlled unit and the total costs for each city. Each ordinance targets different income levels, so each city’s price controls will vary. For example, if a city in Orange County required that 15 percent of new units be “affordable” and its only target income group was “very low,” we assumed that 15 percent of units needed to be sold for $123,101 each. Or, if a city in Orange County required that 15 percent of new units be “affordable” and its only target income group was “low,” we assumed that 15 percent of units needed to be sold for $184,001. 10 Reason Public Policy Institute For cities with more than one target income group, for the sake of simplicity we took the average level of the price controls. For example, if a city in Orange County required that 15 percent of new units be “affordable” and the target income groups were “very low,” “low,” and “moderate,” we assumed that 5 percent of the units needed to be sold for $123,101 each, 5 percent for $184,001 each, and 5 percent for $295,379 each. Taking the average of those figures, we arrive at our estimate that 15 percent of units need to be sold for $200,827 each. Because many towns targeting multiple income groups do not target each income group equally, our estimates will not be 100 percent accurate. If a city targeting multiple income groups requires more “very low” units, our estimates of the costs of zoning will be on the low side. On the other hand, if a city requires more “moderate” units, our estimates will be on the high side. In addition, when a jurisdiction required 10 to 15 percent of units to be affordable, we always chose the lower bound and ignored the upper bound in order not to overestimate the costs of inclusionary zoning.13 Once we arrived at the average price control for each city, we then subtracted it from the market price for each city.14 For example, we estimate that a new home in West Hollywood could be sold for $588,530. West Hollywood requires that 10 percent of homes be priced at “low” and 10 percent at “moderate,” which we estimate at $146,875 and $215,265, an average of $181,070 per home. That means 20 percent of homes would need to be sold for $407,460 less than market price. In other words, the cost of providing each inclusionary unit in West Hollywood is $407,460. In high-priced jurisdictions these losses can be quite high. Figure 7 shows the average cost associated with selling a price-controlled unit based on the standards in those cities and the market prices. In cities with more restrictive price controls and higher land values, the cost is higher. In the median city the cost of providing each inclusionary unit is $577,726. Figure 7: Average Cost Associated with Selling Each Price-Controlled Unit $0 $200,000 $400,000 $600,000 $800,000 $1,000,000 $1,200,000 $1,400,000 $1,600,000 Long Beach Monrovia West Hollywood Pasadena Brea San Juan Capistrano Huntington Beach Irvine Agoura Hills San Clemente Santa Monica Rancho Palos Verdes Laguna Beach DO AFFORDABLE HOUSING MANDATES WORK? 11 The cost of each inclusionary unit is large. Next let us look at the cost per unit times the number produced in each city (Figure 8).15 This gives a measure of the aggregate cost of inclusionary units by city for those that report creating affordable homes under inclusionary zoning. Figure 8: Average Cost Associated with Selling Each Price-Controlled Unit Times the Number of Units West Hollywood, $5,296,977 San Clemente, $519,491,982 Huntington Beach, $180,828,261 Brea, $124,456,235 Monrovia, $79,989,814 Agoura Hills, $29,678,632 Irvine, $1,000,000,000 Pasadena, $5,851,457 Laguna Beach, $204,114,750 San Juan Capistrano, $110,050,263 According to our estimates, the costs associated with producing inclusionary units in Los Angeles County and Orange County have been $3.8 billion. Combining data from Los Angeles and Orange Counties with data from cities in the Bay Area that just adopted their programs, we can get a more accurate estimate of the costs of inclusionary zoning. We find that the median city’s cost of below-market units was more than $79 million. B. Who Bears the Burden of Inclusionary Zoning? The costs of inclusionary zoning are largely hidden. None of the costs imposed on the housing market shows up on any city’s annual budget, but they still exist. Who ends up paying for that $3.8 billion for below- market rate homes? One can debate exactly who bears the costs, but they are necessarily borne by someone. Because they are imposed on the new housing market—and not paid for by government—the costs will be borne by some combination of developers, new homebuyers, and landowners. Exactly who shoulders more of the burden depends on market conditions and supply and demand. All theory and evidence suggest that the costs of inclusionary zoning, effectively a tax, will not be borne by builders but by new homebuyers and landowners.16 Construction is a competitive industry with relatively free entry. Local market conditions will determine exactly how the burden is split. If buyers are more sensitive than sellers to changes in price, then landowners will bear most of the tax. This happens when more 12 Reason Public Policy Institute buyers have many options, such as living in similar or nearby areas. If sellers are more sensitive than buyers to changes in price, then new homebuyers will bear most of the tax. This happens when landowners have more options, such as being able to devote their land to commercial, industrial, or other endeavors. If profits are abnormally high, other builders will enter the market and undercut prices, thus bringing profits down. Conversely, if profits are abnormally low it will drive would-be builders to invest in other endeavors. When a tax in the form of inclusionary zoning is placed on builders, it decreases the number of profitable projects that they want to undertake in that jurisdiction. Builders will vote with their feet and undertake fewer projects in jurisdictions with price controls and more in neighboring jurisdictions without price controls. The quantity of housing produced will decrease where there are price controls, but increase in other places where there are not price controls, pushing some homebuyers away from their first choice of locations, and for developers profit rates at the margin will remain the same. Price controls may not stop all development, but new construction will decrease. In order for development in a price-controlled city to be profitable enough to attract builders, one of two things has to happen. Either market-rate home prices must increase, or land prices must decrease to compensate the builder for his losses due to price controls. Even with price controls on a portion of development, builders can still earn the normal rate of return if other home prices increase or land prices decrease. The likely result will be some combination of the two. Both effects lead to a decrease in the quantity of new housing as market-rate buyers will be able to afford less housing and/or landowners will supply less land for residential development due to low market prices. Raising home prices for other new homebuyers creates a paradox because the alleged goal of inclusionary zoning is to make housing more affordable, not less. Decreasing land prices also decreases the quantity of new housing because it discourages landowners from providing their land for residential projects. Instead, more land will be put to uses in which the final product is not subject to price controls. Thus, the restriction on the supply of land restricts the supply of new homes. Advocates of inclusionary zoning tend to assume that the below-market rate units are subsidized out of builder profits, but economics predicts that builders are actually least likely to bear the burden. In the very short run, if builders own the land when the ordinance was passed, they would bear part of the burden. But in the long run, builders are most able to avoid the tax because they can simply move their construction to more profitable locations. The land cannot move, and buyers are often attached to living in a particular locale. Landowners and new homebuyers will end up paying for the subsidy on the price-controlled units. Inclusionary zoning effectively acts as a tax on the production of market-rate units because developers must sell a percentage of units at a loss to gain permits to sell market-rate units. If market prices went up by the exact amount of losses on the price-controlled units, buyers would bear the full burden of the tax. If market prices did not change at all, builders and landowners would bear the full burden of the tax. In most situations buyers and sellers each bear part of the tax burden. Regardless of who bears the burden, because some units are price-controlled and others are not, the losses from price-controlled units must be spread over some combination of buyers and sellers of the remaining units. We calculate the effective tax in each city by looking at the average cost associated with each inclusionary unit and the number of market-priced units over which the cost will be spread. To do this we multiply the cost of each inclusionary unit times the percentage mandated by each city and then divide by the percentage of market-rate homes. To illustrate, for San Juan Capistrano each price-controlled unit has an associated cost DO AFFORDABLE HOUSING MANDATES WORK? 13 of $561,480 (Figure 7) and 30 percent of units must be sold at those price controls (Table 2). The calculation would be [($561,480)X(0.30)]/(0.70)= $240,634. To make it more concrete, if a project had 10 units, three must be sold at a loss of $561,480. Spreading the loss over the remaining seven units gives a tax of $240,634 per market-rate unit. Figure 9 shows the effective tax on new home purchases imposed by inclusionary zoning. Inclusionary zoning imposes sizeable taxes on each newly constructed home. The median city with inclusionary zoning is effectively imposing $65,952 of taxes on each market-rate home. Figure 9: Effective Tax Imposed on New Market-Rate Units Caused by Inclusionary Zoning $0 $100,000 $200,000 $300,000 $400,000 $500,000 $600,000 Long Beach Irvine San Clemente Brea Rancho Palos Verdes Huntington Beach Agoura Hills Monrovia Pasadena West Hollywood Santa Monica San Juan Capistrano Laguna Beach Cities with higher land values and more restrictive price controls impose the highest effective tax on new homes. In Laguna Beach the equivalent tax on a market-rate home if a developer built and sold an affordable home is approaching $500,000 per newly constructed home. After having calculated the amount of the tax, we can approximate who bears the brunt of the tax. The California Department of Housing and Community Development (HCD) takes the position that inclusionary zoning translates into higher prices for new homebuyers. HCD has consistently held this position through both Republican and Democratic Administrations: Under most inclusionary programs, which typically include an in lieu fee [whereby the builder pays a fee to opt out of the inclusionary zone requirements] option, the cost of subsidizing low-income housing units is underwritten by the purchasers of market-rate units in the form of higher housing prices. This practice of cost shifting is particularly detrimental to a home buyer who marginally qualifies for a mortgage yet earns too much to receive governmental assistance.17 We have consistently…asked local jurisdictions to analyze an inclusionary program as a potential governmental constraint. The reasoning for this is that most programs of this sort impose a fee or dedication requirement upon developers which is passed on to consumers of new market rate housing, raising the price of the market rate housing.18 14 Reason Public Policy Institute Others believe the brunt of the tax will be borne by some combination of builders and landowners. 19 Figure 10 estimates price increases on new homes under the three scenarios. If the lower bound is accurate (when buyers only pay 50 percent of the tax), the price of new homes is increased by $30,000 or more in 8 of 13 cities. If the upper bound is accurate (when buyers pay all of the tax), the price of new homes is increased by $30,000 or more in 12 of the 13 Los Angeles County and Orange County cities with inclusionary zoning. Agoura Hills is the median city. Inclusionary zoning increases new home prices there by $32,976 in scenario one, $55,400 in scenario two, or $65,952 in scenario three. Although the goal is to produce more affordable housing, inclusionary zoning is actually producing the opposite effect. Inclusionary zoning translates into significantly higher prices for market-rate homebuyers. By creating price controls on a percentage of units, it taxes other new units and leads to higher housing prices. Long BeachIrvineSan ClementeBreaRancho Palos VerdesHuntington BeachAgoura HillsMonroviaPasadenaWest HollywoodSanta MonicaSan Juan CapistranoLaguna Beach$0 $50,000 $100,000 $150,000 $200,000 $250,000 $300,000 $350,000 $400,000 $450,000 $500,000 Figure 10: Increases in Price of New Homes Caused by Inclusionary Zoning (Under Three Different Assumptions About Who Bears the Costs) Assuming 50% of tax is borne by consumers Assuming 84% of tax is borne by consumers Assuming 100% of tax is borne by consumers DO AFFORDABLE HOUSING MANDATES WORK? 15 To the extent that sellers bear more of the burden of taxation, the housing market also faces negative consequences. Because builders can move to jurisdictions without inclusionary zoning, they will not bear the burden of the inclusionary zoning tax. Thus, landowners will bear most of the sellers’ portion of the burden. Inclusionary zoning ordinances decrease the value for which landowners can sell undeveloped land to homebuilders. Because landowners receive lower prices, they will supply less land for residential development, and fewer homes will be built. Governments already give landowners incentives to supply land for commercial and industrial uses instead of residential ones. Since Proposition 13 limited increases in residential property taxes, governments began creating incentives for developing commercial real estate instead of residential because it generates more revenue. This has become known as the “fiscalization of land use.” One study described how local governments responded to limits on property taxes this way: Local municipalities employ two primary methods for revenue generation: the imposition of heavier exaction fees for new development and the promotion of retail development in order to maximize sales tax revenues. This has had a direct, deleterious impact on new housing production. Rather than adopt land-use policies that advance or incentivize new housing production, developing new retail centers — such as big box developments, entertainment complexes, and shopping destinations — emerged as the primary approach for increasing local government revenue. Consequently, residential development (and other forms of development) suffered due to a lack of incentives or outright disincentives.20 Inclusionary zoning ordinances add yet another disincentive to provide land for residential development. When part of the burden of taxation is borne by landowners, we should expect inclusionary zoning to decrease the supply of new housing. C. The Effect of Price Controls on Housing Construction In addition to increasing prices, inclusionary zoning leads to a decrease in new housing. Economics clearly predicts that the quantity of construction will be lower after the adoption of inclusionary zoning. But advocates of inclusionary zoning advance an alternate hypothesis that the quantity of construction will be the same (or higher) after the adoption of inclusionary zoning. By looking at the data of housing construction, we can get an idea of which hypothesis is correct. One test is to look at the amount of new construction in years prior and years following the adoption of an inclusionary zoning law. We examined Construction Industry Research Board yearly housing permit data for single and multifamily dwellings to compute average construction pre- and post-ordinance. For example, San Clemente adopted its ordinance in 1980 and Long Beach adopted its ordinance in 1992. We would thus compare San Clemente housing construction in 1979 and 1981, and Long Beach housing construction in 1991 and 1993. We also can compare housing production for the seven years prior and the seven years following the ordinance, so for San Clemente we would compare housing production in 1973-1979 to housing production in 1981-1987 and for Long Beach we would compare housing production in 1984-1990 to housing production in 1992-1998. Because ordinances have been adopted throughout the past 26 years (Figure 1), economy-wide phenomena such as business cycles should not be biasing the data in either direction. For example, some cities adopted their ordinances during down times while others adopted their ordinances during up times. 16 Reason Public Policy Institute The data indicate that inclusionary zoning does indeed lead to a decrease in new construction. For the one-, three-, five- and seven-year averages before and after the ordinances, the production of housing decreased after the adoption of inclusionary zoning. As price controls are in place for more time, the decrease in housing production adds up. Data on housing production seven years prior and seven years following the ordinance exist for 8 of the 13 cities. In those cities in the seven years prior to the adoption of inclusionary zoning 28,296 homes were produced, whereas in the seven years following the adoption of inclusionary zoning only 11,000 homes were produced. In those eight Los Angeles County and Orange County jurisdictions, inclusionary zoning appears to decrease housing by 17,296 units. That amounts to a decrease in housing production by 61 percent. If those 17,296 units would have been worth $650,000 per home, then the value of housing not built because of inclusionary zoning is approximately 11 billion dollars. For those jurisdictions, in only seven years the average destruction of value per city is $1.4 billion. Figure 11: Total Production of Housing 7 Years Prior and 7 Years Following the Ordinance for Eight Jurisdictions' Available Data 0 5,000 10,000 15,000 20,000 25,000 30,000 Before After Recall that over 27 years inclusionary zoning in Los Angeles County and Orange County has only led to 6,379 affordable units, which amounts to 1,653 every seven years. In those eight jurisdictions only 1,534 units have been produced. Controlling for the length of time for each ordinance, those cities in total average 110 units per year since an ordinance has been adopted. Multiplying the yearly production rate by seven gives the expected number of “affordable” units over seven years, which amounts to 770 in all of those eight cities. Although those cities together had an estimated increase in 770 “affordable” units the seven years following the adoption of inclusionary zoning, the total number of homes not built was 17,296 (Figure 12). Is a policy that creates 770 “affordable” homes at the expense of discouraging 17,296 market-rate homes worth it? This is crucial because most entry into the housing market by lower-income families is by buying older homes freed up when middle-income families move into new homes.21 Reducing the overall production of DO AFFORDABLE HOUSING MANDATES WORK? 17 housing both drives up prices and means that the people crowded out of the housing market are the lower- income would-be homeowners. Figure 12: Comparing the Increase in "Affordable" Units to the Overall Decrease in New Construction Associated with Inclusionary Zoning 770 -17,296 -25,000 -20,000 -15,000 -10,000 -5,000 0 5,000 Additional statistical work on inclusionary zoning is needed. The data indicate that the number of units pushed out of the market by inclusionary zoning is much larger than the number of “affordable” units built. Advocates of price controls must recognize that their programs lead to only a handful of below-market units coupled with a sharp decrease in market-rate homes. Because we cannot directly observe the thousands of homes never built, the costs of the program go largely unseen. Also unseen are those 17, 296 families that cannot buy homes because inclusionary zoning prevented the construction of additional homes. Is a program that destroys over $10 billion worth of housing and prevents thousands more families from getting a home than it places in an “affordable” unit worth the high costs? Decrease in overall new construction associated with inclusionary zoning (for eight cities over seven years) Inclusionary units produced (for eight cities over seven years) 18 Reason Public Policy Institute Part 5 The Fiscal Cost of Price Controls to State and Local Government ot only do price controls lead to a decrease in the quantity of housing and an increase in prices for consumers, but price controls also lead to decreased revenue for both state and local government. Inclusionary zoning ordinances are often sold to policymakers as the proverbial free lunch, with proponents claiming “A vast inclusionary program need not spend a public dime.”22 Even if market-rate buyers and landowners end up paying the price of the subsidy, so the argument goes, at least local governments need not spend revenue to create affordable housing. Proponents write, “From a local agency standpoint, inclusionary zoning provides affordable housing at no public cost” (emphasis added).23 The story, however, is not that simple. The advocates fail to take account that inclusionary zoning leads to direct losses in state and local government revenue. Inclusionary units demand and receive the same municipal services as market-rate homes. There is no evidence that providing municipal services to price-controlled homes is less costly than providing to market- priced homes. The cost of inclusionary zoning to governments comes from the fact that price-controlled homes cost the same to service but generate less revenue. Because the values of the homes are set at below- market rates, the assessed values are lower and so their property tax is lower. Thus, although governments may not spend “a public dime” to produce price-controlled homes, they take on an obligation of providing municipal services while receiving lower annual tax revenues. The cost to government from price-controlled units is the difference in the annual tax revenue that would have been generated had the same homes been assessed at market prices. If the real estate tax rate is 1 percent per year, a $700,000 dollar home generates $7,000 in government revenue, whereas a $200,000 home generates $2,000 in government revenue. To calculate the yearly tax revenue lost, we take the difference between current market price and the price-controlled price times 1 percent (for the property tax) for each unit. Multiplying times the number of units in each jurisdiction gives us a rough measure of the lost tax revenue per year. Biasing our numbers downward is the fact that we do not count the lost revenue from the homes never produced because of price controls. Biasing the numbers upward is the fact that not all market-rate homes are assessed at current prices due to Proposition 13. Also, many of the “affordable” units do not remain affordable if resale restrictions are absent. But the numbers illustrate the limit as homes are frequently resold and reassessed at current prices. They also approximate how much revenue would be gained if price-controlled units were reassessed at market rates. Figure 15 shows the yearly loss in combined state and local revenue due to price controls." to "lost state and local revenue combined become significant, as depicted in Figure 13. N DO AFFORDABLE HOUSING MANDATES WORK? 19 We do not believe that the goal is to maximize tax revenue at the expense of low-income households, and we are not advocating raising real estate taxes for low-income residents. But before considering inclusionary ordinances, governments must look at their budgets and examine whether better ways of helping low-income households exist. Figure 13: Yearly Loss in Combined State and Local Government Revenue Due to Price Controls (Assuming All Units Are Assessed at Current Prices) Irvine, $26,162,480 West Hollywood, $52,970 Laguna Beach, $2,041,148 San Juan Capistrano , $1,100,503 Huntington Beach, $1,808,283 Brea, $1,244,562 Monrovia, $799,898Agoura Hills, $296,786 Pasadena, $58,515 San Clemente, $5,194,920 Government would be well advised to consider these yearly costs before adopting inclusionary zoning. It is important to note that the lost tax revenue occurs not just in one year but every year that the price controls are in existence. The total present value of lost government revenue is upwards of $752 million (Figure 14).24 Although inclusionary zoning is often pitched to governments as a zero-cost method of creating affordable housing, the costs from lower assessed valuations are quite large. Both state and local governments bear some of the burden of lost tax revenue caused by inclusionary zoning. Property tax revenue goes to the state government, and a portion is rebated back to city and county governments. The exact amount returned to each jurisdiction varies significantly, so our above estimates measure the combined total of lost tax revenue without distinguishing the particular splits between local and state governments. Importantly this implies that although inclusionary zoning policies are usually debated and implemented at the city and county levels, state legislators should be concerned with these policies too. Each additional local inclusionary zoning ordinance adversely impacts the tax revenue not just of its own jurisdiction but also decreases the state’s tax revenue. 20 Reason Public Policy Institute Figure 14: Present Value of Yearly Loss in State and Local Government Revenue Due to Price Controls (Assuming a Discount Rate of 3 Percent and that All Units Stay Assessed at Current Prices) Irvine, $512,796,153 Pasadena, $1,566,694 Agoura Hills, $3,543,016 Monrovia, $21,416,815 San Clemente, $77,287,289 West Hollywood, $1,671,031 San Juan Capistrano, $16,372,700 Huntington Beach, $44,336,762 Laguna Beach, $48,666,401 Brea, $24,393,971 DO AFFORDABLE HOUSING MANDATES WORK? 21 Part 8 Conclusion nclusionary zoning should only be enacted if the goal is to make housing more expensive and decrease the quantity of new housing. Our findings in Los Angeles County and Orange County are consistent with the experience of the San Francisco Bay Area. Inclusionary zoning hurts homebuyers and will price out most low-income families. Despite the good intentions of those who support inclusionary zoning, economics tell us that price controls on new housing will have the unintended consequence of reducing the quantity of new homes built. Rather than helping, inclusionary zoning will actually make the affordability problem worse. We have shown that inclusionary zoning imposes significant costs on the housing sector. Those costs are passed on to landowners and buyers of market-rate homes. Higher housing prices will result. Something should be done about the affordability crisis, but price controls are not the answer and may be part of the problem. Southern California cities will never be able to rely on inclusionary zoning to meet their housing needs. In fact, inclusionary zoning has led to a decrease in housing production. Rather than continuing to impose these policies, jurisdictions would do well to eliminate them. By ending price controls on new construction, builders would have an incentive to supply more housing. The worst possible solution to the affordability crisis is to pass policies that result in restricting the supply of housing. Inclusionary zoning is one such policy. I 22 Reason Public Policy Institute About the Authors enjamin Powell is an Assistant Professor of Economics at San José State University and an Adjunct Scholar with Reason Foundation. He received his Ph.D. from George Mason University in 2003. He has numerous publications in scholarly journals, policy papers, and the popular press. Edward Stringham is an Assistant Professor of Economics at San José State University and an Adjunct Scholar with Reason Foundation. He received his Ph.D. from George Mason University in 2002. He is winner of the Paper of the Year Award from the Association of Private Enterprise, Best Article Award from the Society for the Development of Austrian Economics, and Second Prize from the Independent Institute Garvey Essay Contest. Stringham serves on the Executive Committee of the Society for the Development of Austrian Economics and on the Executive Committee of the Association of Private Enterprise Education. Powell and Stringham's other recent work on housing includes a policy study, "Estimating the Effects of Price Controls in the Redevelopment of the Fort Ord Military Base" and testimony before the Board of the Fort Ord Reuse Authority. Powell and Stringham also have the entry on "Housing" in the forthcoming Concise Encyclopedia of Economics. Note: The authors appreciate research assistance from Ilkay Pulan, Daocheng Zhu, and a research grant from the California Building Industry Association. B DO AFFORDABLE HOUSING MANDATES WORK? 23 Related Reason Foundation Studies Supply and Affordability: Do Affordable Housing Mandates Work?, by Benjamin Powell and Edward Stringham, Reason Foundation Policy Study No. 318, April 2004, http://www.rppi.org/ps318.pdf Smart Growth in Action, Part 2: Case Studies in Housing Capacity and Development from Ventura County, California, by William Fulton, Susan Weaver, Geoffrey F. Segal, and Lily Okamura, Reason Foundation Policy Study No. 311, May 2003, http://www.rppi.org/ps311.pdf San José Demonstrates the Limits of Urban Growth Boundaries and Urban Rail, by Randal O'Toole, Reason Foundation Policy Study No. 309, April 2003, http://www.rppi.org/ps309.pdf Smart Growth in Action: Housing Capacity and Development in Ventura County, by William Fulton, Chris Williamson, Kathleen Mallory, and Jeff Jones, Reason Foundation Policy Study No. 288, December 2001, http://www.rppi.org/ps288.pdf Smart Growth and Housing Affordability: Evidence from Statewide Planning Laws, by Sam Staley and Leonard C. Gilroy, Reason Foundation Policy Study No. 287, December 2001, http://www.rppi.org/ps287.pdf Urban-Growth Boundaries and Housing Affordability: Lessons from Portland, by Samuel Staley, Reason Foundation Policy Brief No. 11, October 1999, http://www.rppi.org/urban/pb11.pdf Repairing the Ladder: Toward a New Housing Paradigm, by Howard Husock, Reason Foundation Policy Study No. 207, July 1996, http://www.rppi.org/ps207.pdf 24 Reason Public Policy Institute Endnotes 1 See the statements of the housing advocacy coalition National Housing Conference at www.nhc.org 2 National Housing Conference, Inclusionary Zoning: The California Experience, Washington, D.C.: National Housing Conference, 2004, http://www.nhc.org/nhcimages/California%20IZ/CaIZ04.pdf. 3 Meyers Group “Meyers Group Summary Statistics” January-March 2004. 4 California Coalition for Rural Housing and Non-Profit Housing Association of Northern California (2003). Inclusionary Housing in California: 30 Years of Innovation, p.3. 5 Powell and Stringham, Housing Supply and Affordability, pp. i-v, pp. 1-45. 6 For an in depth explanation of the economics of inclusionary zoning see Housing Supply and Affordability: Do Affordable Housing Mandates Work? . This study also contains discussions about the long-term effects of affordability controls and the current debate on inclusionary zoning. Each of those sections is equally applicable to the Southern California market. 7 Southern California Association of Governments State of the Region 2003 , pp. 35-42 (Los Angeles: Southern California Association of Governments, 2003) 8 National Association of Homebuilders, Housing Opportunity Index: First Quarter (Washington, D.C.: National Association of Homebuilders, 2002). 9 Calculated by taking the 7.5 year Southern California Association of Governments 1998-2005 Regional Housing Need Assessment for the 13 cities and dividing by 7.5. 10 We are not measuring what economists refer to as social costs, which would include the value of the lost consumer and producer surplus associated with inclusionary zoning. We are simply estimating the monetary amount that a seller must forgo when selling at the restricted price. 11 Income categories are adjusted by household size. Compared to a four-member household, a household with five members can have an 8 percent greater income, a household with six members a 16 percent greater income, etc. Households with three members will have a 10 percent lower income, two members 20 percent lower and one member 30 percent lower according to the Department of Housing and Community Development. Because the four-member household is the baseline, we focus on four- member households throughout the paper. 12 Barbara Kautz, “In Defense of Inclusionary Zoning: Successfully Creating Affordable Housing” University of San Francisco Law Review, vol. 36, 2002, p. 1014, gives an additional legal reason: “An inclusionary ordinance that does not limit the resale prices of for-sale units (creating ‘premium pricing’ for the first buyer) may be vulnerable to attack for ‘not advancing a legitimate state interest.’” 13 For example, inclusionary zoning in Santa Monica requires 10-20 percent of units be affordable, and its target groups are “Very Low” and “Low.” In this case we would assume only 10 percent needed to affordable and the share was split 5 percent for “Very Low’ and 5 percent for “Low.” Because data for specific requirements of each city are currently unavailable, we decided to make simplifying assumptions and again err on the side of lower costs of inclusionary zoning. 14 We estimate market price by city by comparing 2004 data of the average price of new homes by county compiled by Meyers Group and 2000 Census median price of existing homes by city. Because new homes in Los Angeles County and Orange County typically sell for more than the 2000 median price of DO AFFORDABLE HOUSING MANDATES WORK? 25 existing homes, and because new home price data by city is difficult to assemble, we adjust the 2000 Census city data based on each county's price differential for new homes. In Los Angeles County the ratio of the price of new homes to the price of existing homes is 2.2 and in Orange County that ratio is 2.6. For example, the 2000 Census median price of existing homes in the city of West Hollywood was $263,400, so we multiply that by 2.2 to estimate that a new home would be sold for closer to $588,530. 15 The figure looks at costs in today’s prices. The divergence between current price controls and the price at which the units currently could sell gives us comparable numbers in today’s dollars. It does not calculate the home price for the year the price-controlled units were built or calculate the price controls in that year. Donna Jones, “Homes, Good and Cheap: Low-income Buyers Get High-Quality Homes,” Santa Cruz Sentinel, January 10, 2004, p.9. reports on one project built in 1998 that helps illustrates our assumptions. When homes were first sold, government set the prices at around $160,000. Today government sets their prices at $280,000, and “they would be worth at least $800,000 on the open market.” For the purposes of our calculations, we would subtract the restricted price of $280,000 from the market price of $800,000 to arrive at the $520,000 difference. That price control (and that difference divergence between the market price and the restricted price) is no longer imposed on the initial seller but is now imposed on the current owner. If 10 homes were in a project, the equivalent cost from the price controls would be 10 times that number. 16 The requirement of subsidized housing has the same effect as a development tax. The developer makes zero economic profit with or without inclusionary zoning, so the implicit tax is passed on to consumers (housing price increases) and landowners (the price of vacant land decreases). In other words, housing consumers and landowners pay for inclusionary zoning.” Robert Burchel and Catherine Galley, “Inclusionary Zoning: Pros and Cons,” in The California Inclusionary Housing Reader, (Sacramento: Institute for Local Government, 2003), p.29. 17 California Department of Housing and Community Development letter to the city of Fairfield, July 16, 1996. 18 California Department of Housing and Community Development letter to the city of Fairfield, April 26, 2001. 19 Laura Padilla, “Reflections on Inclusionary Housing and a Renewed Look at its Viability” Hofstra Law Review, vol. 23, 1995, p.576. 20 Joel Kotkin and Thomas Tseng, Rewarding Ambition: Latinos, Housing, and the Future of California (Malibu, California: Pepperdine University School of Public Policy, 2002). 21 Howard Hussock, Repairing The Ladder: Toward a New Housing Policy Paradigm, Reason Foundation Policy Study No. 207 (Los Angeles: Reason Foundation, July 1996). 22 Andrew G. Dieterich, "An Egalitarian Market: The Economics of Inclusionary Zoning Reclaimed" Fordham Urban Law Journal, vol. 24 (1996) p. 41. 23 Several studies emphasize the assertion that there is no direct cost. See Kautz, “In Defense of Inclusionary Zoning,” p.5; Robert Burchel and Catherine Galley, “Inclusionary Zoning: Pros and Cons,” in The California Inclusionary Housing Reader (Sacramento: Institute for Local Government, 2003), p.29; and Marc Smith, et.al., “Inclusionary Housing Programs: Issues and Outcomes,” Real Estate Law Journal, Fall 1996. 24 We take into account the length that each city intends to impose price controls (30 years for the median city) and then calculate the present value, assuming a 3 percent discount rate. Reason Public Policy Institute 3415 S. Sepulveda Blvd., Suite 400 Los Angeles, CA 90034 310/391-2245 310/391-4395 (fax) www.rppi.org