A hidden gift to manufacturing: cities should stop zoning land exclusively for industrial use.

Hills, Roderick M., Jr. ; Schleicher, David

We both now live or have lived in Brooklyn, in a neighborhood that real estate brokers call "Cobble Hill," just a few blocks from the East River. The location allows you simultaneously to get some exercise and study land use regulation by taking a run along the South Brooklyn waterfront. If you run south along Columbia Heights and Van Brunt, you enjoy one of the city's best views of the Upper Bay and Manhattan. Just by turning your head, you can see a stretch of river from the Brooklyn Bridge to the Verrazano Bridge, encompassing the towers of the Financial District, the Statue of Liberty, Governor's Island, and Staten Island. It is not far from the popular restaurants and cafes of Cobble Hill and Carroll Gardens, and the area received some press when MTV's The Real World filmed a season from a building nearby. The land seems like a prime spot for a condo with a view.

But the buildings enjoying this magnificent vista are often abandoned or underused, frequently consisting of apparently empty warehouses, some usually idle container cranes, and crumbling concrete lots surrounded by chain-link fence. The waterfront still has a couple of significant industrial employers. On Piers 7 through 11, for instance, American Stevedoring leases land from the Brooklyn Port Authority, although the rent is apparently paid from funds supplied by the state, and the cocoa beans that the company unloads are barged over to Port Newark, NJ. The Golten Marine Company operates a maritime repair facility next door to American Stevedoring. But heavy maritime industry has mostly left the area: the two most conspicuous businesses are the Fairway Grocery and the new IKEA, both commercial retailers in Red Hook that cater to city residents. These retailers are surrounded by the skeletons of crumbling warehouses and defunct cranes, the relics of Brooklyn's manufacturing past.

[ILLUSTRATION OMITTED]

The absence of housing along New York's waterfront is, in large part, the result of zoning. Since 1961, the city's zoning resolution has barred residential uses from manufacturing zones, and 30 percent of the city's shoreline is presently zoned for industrial use. Such "noncumulative" manufacturing zones (that is, zones that do not allow the "cumulation" of uses less noxious than industry in manufacturing zones) bar housing in a considerable part of New York City. According to a 2005 Manhattan Institute study by Regina Armstrong, the city zoned 22,500 acres for industrial development to the exclusion of residential uses in 2005. Cities around the country have adopted similar policies, fencing off valuable real estate in tight markets from residential and often from commercial development.

Why not allow residential uses in these manufacturing zones? Advocates of noncumulative zoning typically offer two explanations for how residential users could threaten industrial uses. First, residential users are said to burden industry with complaints and nuisance lawsuits. Keeping residences out of industrial zones is a regulatory analogue to a "coming to the nuisance" defense in tort law, preserving industrial investments from encroachments by sensitive users. Second, residential users outbid industrial users for land, driving industry out of the city in search of cheaper real estate. By excluding residential uses, noncumulative zoning protects urban industry from the threat of escalating real estate prices.

Neither of these arguments provides a compelling reason to exclude residential or commercial uses from manufacturing zones. Industry could be protected from nuisance litigation by giving industrial users a defense of regulatory compliance. Further, noncumulative zoning is an indiscriminate, inflexible, and politically invisible form of subsidy for manufacturers. It cannot be targeted at firms that provide agglomerative spillover benefits for the city, it provides far less benefits to firms than would cash subsidies of equal value, and it generates costs that are not included in a city's public budget and hence are hidden from ordinary group competition for scarce public resources. Noncumulative zoning is an idea whose time has passed.

BRIEF HISTORY OF NONCUMULATIVE ZONING

Prior to World War II, most zoning ordinances followed the model of the Village of Euclid v Ambler Realty (1926) by providing for "cumulative" or "inclusive" zoning districts defined by a hierarchy of uses. The highest use was the single-family home, and land zoned for such homes excluded most other uses, including multifamily residential uses. Land zoned for uses deemed to be more noxious than the single-family use would permit single-family homes as well as other less noxious uses. Thus, commercial, multifamily, and single-family uses would all be permitted in industrial zones, and landowners could freely convert industrial uses into other uses within those zones as the market dictated.

After World War II, municipalities began experimenting with noncumulative zones that excluded residential uses from industrial zones. During the 1950s, state courts were typically skeptical of such ordinances, regarding them as inconsistent with the anti-nuisance rationale for zoning that had been proffered in Village of Euclid as zoning's primary justification.

This focus on protecting residences from industrial uses began to change in New York City when Robert Wagner began a campaign to revise New York's zoning resolution between 1947 and 1961. Unlike the state court opinions that justified such exclusion largely in terms of protecting residential uses from the fumes and noise of industry, the studies commissioned by Wagner and his appointees argued that industrial uses were threatened by housing and that the city needed to safeguard its industrial future by reserving land exclusively for the former. After a decade of protracted political struggle, Wagner got his wish when the City Council enacted the 1961 Zoning Resolution that provided, for the first time, noncumulative manufacturing zones in New York City.

What had changed by the early 1960s that would make more compelling these calls to protect industry from residential uses? Between 1955 and 1965, New York's industry had been devastated by a revolution in transportation that deprived cities of their comparative advantage in attracting and retaining manufacturing. At the center of this revolution was the creation of an interstate highway system and the "container revolution" integrating the shipment of goods in a single metal box across rail, trucking, and shipping. Radically reducing shipping costs, this transportation revolution eliminated most of the advantage of locating factories in immediate proximity to Brooklyn's piers. By 1965, New York City had lost much of its maritime shipping business to New Jersey and factories had deserted Bay Ridge and Sunset Park in droves. New York's waterfront had been reduced to derelict shambles and the city's industrial job base had suffered staggering losses.

Responding to pressure from manufacturers and unions, politicians in New York and other big cities pressed for incentives to keep manufacturing enterprises in the city. New York's 1961 move to noncumulative zoning has been matched by other large cities' efforts to attract or retain industry. The names and details of these zoning schemes differ (Chicago calls them "planned manufacturing districts," San Francisco calls them "industrial protection zones," and Los Angeles, "industrial business zones"), but the basic concept remains the same: the city seals off some area from residential and often commercial development in order to provide benefits to urban manufacturers.

Two arguments dominate this defense of noncumulative industrial zoning. First, both planners and industrial users of urban land complain that intruding residential uses threaten industrial uses with complaints about noise, smell, or traffic. As Robert J. Hughes, the owner of Erie Basin Bargeport, the city's largest barge operator, stated in explaining his opposition to luxury waterfront condos in Red Hook, Brooklyn, "The first thing luxury condo owners will do is sue us." This anti-lawsuit justification is the exclusive reason for non-cumulative zoning offered by the New York City Planning Department's website.

Second, manufacturers and unions worry that residential users will bid up the price of land, causing landowners to hike rents on industrial users, who will respond by fleeing to the suburbs. The transportation revolution increased the elasticity of demand for industrial land so much that cities can no longer hope to retain manufacturing enterprises simply by offering proximity to customers or suppliers. By excluding residential (and, less frequently, even office and commercial uses), the central city can provide an in-kind subsidy of cheaper land to manufacturers as a bribe to get them to locate or expand in the city.

Such exclusively industrial zoning has potentially large costs. Land zoned for nonresidential uses could be an important source of residential housing. In 2005, the City of New York zoned 22,500 acres of land exclusively for manufacturing uses; developed at even a small fraction of Brooklyn's average density of 55 dwelling units per acre, this land could provide thousands of units of housing in a city with a notorious shortage of residential units. Of course, there is no way to easily calculate how much of this land would actually be used for residential uses if zoning were cumulative, but there is anecdotal evidence of high demand for industrially zoned land among residential users. In the 1930s, fully half of all New Yorkers lived in non-residential zones, and, even under the post-1961 regime of noncumulative zoning, manufacturing zones accommodate a large number of residential units--so many, in fact, that the city's crackdown on illegal conversions of manufacturing units to residential use was impeded by the threat of leaving hundreds of tenants stranded.

THE CASE AGAINST THE CASE FOR NON-CUMULATIVE ZONING

It is not self-evident that keeping manufacturers within cities is a good idea. If transportation and real estate costs are lower in less densely populated areas, then the de-industrialization of cities might be a boon rather than a bane. But there may be some justification for these policies. In what follows, we adopt what economist Timothy Bartik has called the "market failure" perspective on the problem of urban industrial uses: we try to define the circumstances under which markets in land might fail to reflect the socially optimal amount of industry in urban areas. We argue that noncumulative zoning is likely to be too inflexible and indiscriminate to identify specific industrial and commercial uses that will create increasing returns, and we suggest that a "regulatory compliance" defense and tax or grant subsidies would be better mechanisms for industrial retention if, in fact, such retention is advisable.

Stopping Nuisance Complaints The least controversial "market failure" for which exclusive industrial zones might be a plausible remedy is the problem of nuisance. On this view, residential users will inefficiently drive out industrial neighbors by complaining about nuisance costs such as noise, fumes, or traffic. Iran industrial landowner could somehow purchase, lease, or buy easements for all of the land within earshot of their industrial facility, they could protect themselves from nuisance lawsuits. But transaction costs predictably foil the effort to buy out a multitude of ill-identified potential plaintiffs in a densely populated urban area.

The problem of lawsuits, however, does not require total exclusion of housing. Why not allow residential users to occupy land in manufacturing zones, but give industrial users a defense of regulatory compliance against any nuisance or analogous lawsuits that might otherwise be applicable? If residential users prefer the grit and noise of industrial neighbors to other residential alternatives, then excluding those users from manufacturing zones entirely would seem to be excessive paternalism. Why not let the urban pioneers make their own choice on the tradeoff between low price and low clamor?

There is plenty of precedent for such a regulatory compliance defense to protect active users from quiet enjoyers. In particular, many states have passed "right-to-farm" laws that give existing farms a right to continue defined farming activities against nuisance complaints about the obnoxious side effects. "Right-to-stink" laws could serve an analogous function for industry.

One difficulty with right-to-farm legislation is that such statutes typically protect only existing uses from nuisance complaints. It is a common lament of industrial users that the presence of residences prevents their expansion as well as continuation. To the extent that one wanted to protect prospective industrial uses from existing residential users' complaints, then one would need a full defense of regulatory compliance, under which current or future uses consistent with the manufacturing zone's use schedule would be exempt from liability even if they arose after the plaintiff or complainant purchased within the zone.

Such a strong defense of regulatory compliance, in which a lot that is vacant or residential could be converted into a smoke-spewing smelter, places a heavy burden on potential residential users to research and insure against changes in the use of nearby land, either through on-site precautions (for example, triple-pane windows) or simply taking a short-term lease rather than a fee simple interest. Manufacturing zones that form the basis for a regulatory compliance defense, therefore, might need more detailed specifications of permissible uses, including emissions and decibel levels to serve as notification for housing consumers.

Subsidizing Manufacturing The major justification for exclusion of residential uses from manufacturing zones is less concerned with prevention of nuisance litigation than with stabilization of land prices. In purpose and effect, noncumulative zoning is a subsidy to draw manufacturing enterprises to the city. Zoning reduces the cost of manufacturing land in the city and, thereby, is a subsidy to new manufacturing entrants.

Does it ever make sense to influence the location of industry with such subsidies? Maybe--but the subsidies have to be targeted to the precise uses that generate spillover benefits that are not reflected in the market value of land. Noncumulative zones do not seem well suited for delivering such subsidies when compared to outright grants or tax subsidies. In particular, noncumulative zones are so indiscriminate, inflexible, and politically invisible that the deadweight costs imposed on residential users are likely to outweigh the benefits to the city's economy.

How might a subsidy rationally support the development of a city? In their efforts to explain why cities develop, economists have produced a voluminous literature usually called "agglomeration economics" or the "New Economic Geography." This literature argues that individuals and firms locate near others because of the external benefits of physical proximity. Specifically, they have identified three major reasons why firms and individuals cluster in cities:

* to reduce shipping costs for goods;

* to access deep markets (particularly labor markets, but also consumption and social markets), which provide more specialized services and employers, insurance against firm- or industry-specific risk, and quick matching; and

* for information spillovers between firms and individuals, either inside an industry or between industries, that promote both increased production and the development of human capital over time.

The market for land theoretically might lead to too little clustering of firms in cities because individual firms will fail to take into account the positive effect they have on other firms and individuals. The core case for industrial subsidies follows ordinary Pigouvian principles (see p. 2) for subsidizing beneficial spillovers: cities that derive special spillover benefits from particular firms or industries should subsidize those firms or industries until private costs are reduced to the point where they make the socially optimal decision. As long as cities differ in their ability to capture these externalities, competition among them should lead to optimal location decisions, as the city best placed to capture the externalities will bid the most to capture the firm.

However, to justify a subsidy, a firm must provide a greater externality than the various deadweight costs imposed by taxation. Taxation at the local governmental level produces two distinct sorts of deadweight losses. Aside from the ordinary excess burden caused by taxation's distorting effect on consumption or production, local taxes also influence residential choices. Unless the tax is effectively a benefits charge paid exclusively by those who receive the external benefit of the subsidy, increases in taxation may (on the margin) cause some residents to exit to other locations. And individual residents can create agglomeration spillovers just as readily as industry. If artists, actors, financial wizards, writers, or other creative types are driven out of the city because of high tax burdens, then the agglomeration economies that they generate by hanging out at cafes and exchanging ideas will be lost.

Thus, for a subsidy to be justified for a given city, the external benefits provided by the subsidized firm must be greater than (a) the sum of the cost to existing residents of local taxation (in, say, changing their consumption patterns or labor market participation in inefficient ways), and (b) the loss of external benefits generated by those who are priced out of the city by taxation. This formula implies that industrial subsidies are justified from the city's perspective only if the industrial tenant thus gained or retained will be a far better generator of beneficial spillovers than whomever they displace.

Such a formula suggests that the mechanism for providing industrial subsidies should not be indiscriminate or inflexible. Policies that provide subsidies to crudely defined categories of industry are unlikely to distinguish between industries that generate net benefits after the cost of the subsidies is taken into account. Take, for example, intellectual spillovers, which are commonly cited as the most prominent agglomeration benefit of dense industrial concentration. Cities are said to foster a "creative class" precisely because their density allows persons who benefit from intellectual spillovers to interact frequently and informally. It is, however, difficult to determine which mix of firms--whether diverse or homogenous--or which type of firms generates intellectual spillovers. There is no reason to think that it will be beneficial to subsidize manufacturing in general rather than industries with particular characteristics, such as firms with high levels of human capital and/or firms that are known to provide and rely on a high degree of intellectual ferment (e.g., filmmaking, education, medicine, software design).

The same argument can be made for increasing the labor market size. There are spillover benefits from labor market depth, but manufacturing industries do not indiscriminately generate such benefits more than other types of industries. Deep labor markets provide gains from specialization and insurance. Deep labor markets also provide labor with insurance against firm- or industry-specific risk. If a single firm or field in a big city does badly, an individual who works there can get another job without relocating. Further, deep labor markets have lower search costs for both firms and individuals, as it is easier to find the proper labor (or firm) if there are many choices. This has dynamic effects as well. Increased localization creates incentives for labor to invest in human capital, as they can be sure these investments will not be wasted.

To the extent that labor is not fungible across industries, labor market depth might provide an argument for subsidizing industries that are already large in any given urban area. By having more firms in these industries, there will be more labor demanded, more specialization, and greater insurance against firm-specific risk (although less against industry-specific risk). But this argument does not support indiscriminate manufacturing subsidies. Instead, the argument for labor-market depth provides an argument for subsidizing certain classes of industry in which cities are already strong, thereby retaining labor market depth and preserving the market's quality and specialization. Cities certainly engage in this type of subsidy policy. For instance, New York City granted Broadway theater companies a subsidy by allowing them to sell the air rights above their theaters to developers. Spraying dollars indiscriminately at the manufacturing sector, however, seems intuitively an implausible way to preserve labor market depth. One might as well deepen the labor market by simply attracting all sorts of businesses through generally lower taxes.

Finally, the traditional argument for concentrating industry in urban areas--reduction in transportation costs seems increasingly implausible as a justification for manufacturing subsidies. The importance of this as a force for agglomeration has declined as intercity transport costs have fallen dramatically in the last 50 years. It barely costs anything to ship, say, lug nuts, and so the value to other firms of having locally sourced lug nuts is now very low. As a result, it is hard to imagine a justification for urban industrial subsidies on the basis of reducing shipping costs.

NON-CUMULATIVE ZONING'S THREE FATAL FLAWS

Judged by the standards set forth above, noncumulative zoning is a poor way of subsidizing retention of industry in urban areas. Such in-kind land subsidies tend to be too indiscriminate, inflexible, and invisible to be reliably worth the deadweight costs that they are likely to impose.

Indiscriminant Subsidy The most obvious flaw with noncumulative zones is that they provide an indiscriminate subsidy to every business falling within the zoning district's schedule of uses. These uses tend to be broadly defined, often including (to the consternation of lobbyists for industrial land) not only industry but also commercial retailing. Even if retailers were excluded, however, manufacturing districts do not make fine distinctions between manufacturing uses based on the quantity and quality of the jobs they produce. New York City, for instance, has only three manufacturing use districts (light, medium, and heavy) defined by the intensity of noise or pollution produced by the permitted uses. Although these three categories are subdivided into use groups that variously permit specific commercial uses, the resolution--like other cities' zoning ordinances provides no way to distinguish (for instance) between a high-wage manufacturer using a highly skilled workforce and a low-wage manufacturer employing very few workers or unskilled nonresidents who commute from the suburbs.

Zoning categories are far too crude to discriminate between businesses worth subsidizing and businesses that produce no net gain for local residents. The result is that manufacturing districts contain those underused warehouses, parking lots, and even abandoned buildings on New York's waterfront. Even when such zones produce economically viable uses, there is no guarantee that they will produce the high-wage jobs that their boosters urge as their justification. One study of Chicago's planned manufacturing districts found that they did not do much to preserve high-paying manufacturing jobs even when they were occupied by successful commercial businesses.

The indiscriminate nature of manufacturing zones could be solved by narrowing the number of permissible uses. This, indeed, is the solution urged by New York's Industrial Retention Network, which has urged the creation of industrial zones from which profitable retail uses are excluded. But it is unlikely that the city could confidently select specific types of enterprises likely to produce, say, human capital spillovers and promote those enterprises by excluding all other uses from a parcel. Unlike a cash subsidy, the zoning subsidy risks leaving a lot vacant for an extended period of time while the city awaits a buyer who would actually put the land to the zoned use.

Inflexible Subsidy If landowners could easily obtain map amendments or use variances whenever the existing manufacturing use is, in the judgment of the city, less valuable than a proposed residential use, then the indiscriminate character of zoning districts would not inflict any deadweight cost on housing consumers. Landowners with manufacturing tenants that produce few spillover benefits would simply seek to have their parcel rezoned for residential uses, citing the low-value nature of its current use.

But land use changes, whether administrative or legislative, are costly to obtain. Unlike outright grants of revenue that are regularly reviewed through the budget process, zoning districts remain in place until the planning commission, city council, or private parties propose an amendment or variance. Landowners seeking to change zoning designations face opposition from neighbors who typically oppose any rezoning that increases density or bulk of the existing use, as well as from the tenants of the landowner and their employees and potential rival purchasers of the land. To say that it is difficult to get a variance understates the case significantly.

Of course, well connected and experienced developers can and will pay the freight to alter the zoning where the gap between the value of the existing use and the proposed residential use is extraordinarily high. By greasing the skids with community benefit agreements or providing parks and playgrounds, developers can buy allies on the city council and buy off opposition groups. But this is inefficient, unfair to those with fewer political connections, and time consuming.

The inflexibility of noncumulative zoning comes with an added cost. Because the only way to use noncumulative zoning as a subsidy is by procuring land, it generates inefficient substitution among manufacturers toward the use of more land rather than, say, more efficient machines. Not only does this provide less benefit to manufacturers than would a direct subsidy, it increases the cost to the city economy. The cost of the noncumulative zoning "tax" will be borne by other potential users of the property--that is, commercial users and residents. As these residents generate external returns too, the costs of noncumulative zoning are magnified.

Invisible Subsidy Noncumulative zoning has political as well as practical drawbacks. In comparison with the obvious alternatives of direct subsidies or tax breaks for firms, it is much less visible even to an attentive public. The reason for this invisibility is that the baseline of "neutral" treatment is much harder to perceive in zoning than in taxation. When a manufacturer or developer receives tax abatement, then it is obvious to minimally informed observers that they are receiving an unusual benefit for which they should be held accountable, because the layperson's baseline of expectations is that one normally pays taxes. Thus, New York City's 421-a tax abatement program, which provides tax relief to developers and owners of newly built condominium apartments that have certain characteristics, generated enormous controversy in the popular press because of the perception that people like Calvin Klein and Derek Jeter ought not to receive "special" tax relief. By contrast, when a manufacturer gets a cheap lot because competing bidders have been zoned out of existence, there is no intuitive baseline of expectations by which to identify or measure the benefit. No one can tell why residential users have never bid on a lot (which might not have been developed as residential housing even if rezoned), let alone the magnitude of the price reduction that the manufacturer received as a result of the zoning restriction.

Subsidy programs or tax abatements have to compete with other possible uses of local resources from schools to roads. Industrial retention does not have to go through this political crucible. As long as noncumulative zoning makes the costs invisible to the public, there is little political check to ensure that the costs of industrial retention do not massively outweigh the benefits.

CONCLUSION

The simplest solution to the problem of noncumulative zoning is to make such zones cumulative by permitting within them all uses less noxious than industry. Land markets are hardly perfect mechanisms for allocating parcels among competing uses for land. Noncumulative zones, however, are highly unlikely to outperform even imperfect markets. As methods for abating nuisance lawsuits, such zones go far beyond what is necessary to preserve the rational expectations of industrial users. As mechanisms for subsidizing the industrial users' costs of acquiring land, such zoning imposes extraordinary deadweight losses--primarily elimination of housing opportunities--while making no effort to target the cost reduction to those very specific industries that are likely to generate spillover benefits for the city or region.

Rather than rely on this indiscriminate and inflexible device for subsidizing industry, we suggest that the city rely on subsidies that actually are earmarked for businesses that produce the touted benefits. Ideally, businesses would apply for grants based on their capacity to generate intellectual or labor market spillovers, ensuring a program that is maximally discriminating and subject to regular legislative review.

One might respond that such a subsidy-based system generates deadweight losses of its own in the form of higher taxation needed to generate the necessary revenue. Ideally, the city's system of taxation would be able to tap the extra value created by cumulative zones by using the increase in property assessments generated by the looser zoning restrictions to generate more revenue for (among other things) carefully targeted industrial subsidies. In such an ideal tax system, landowners could convert their land as of right to whatever use generated the highest returns, and cities would tax the land based on the market value of a vacant parcel without respect to its actual use. Such a system would impose little deadweight loss because the landowners' actions would not change their tax liability. Unfortunately, most states' systems of property taxation deter such a rational system of revenue by limiting the ability of a local government to tax certain types of property, either by underassessing residential uses or by subjecting nonresidential uses to additional taxes from which residential users are exempt.

The best justification for noncumulative zones, in short, might be that they provide a second-best solution to an artificially constrained system of local government finance. By imposing conditions on the rezoning of noncumulatively zoned land, a city can generate various in-kind benefits (e.g., the parks that cities get from developers who are greasing the wheels in an effort to avoid the strictures of the zoning regime) that would be denied to the city if it allowed land to convert as of right. Put differently, the only way to justify noncumulative zoning--and even this would kindly be described as a stretch--is as a response to extreme pathologies of the laws governing local taxation. The energy invested in its defense by urban planners and city politicians would be more wisely devoted to improving the municipal system of taxation so that cities would have revenue to do precisely what noncumulative zones do so crudely--retain industry that actually generates benefits worth subsidizing.

BY RODERICK M. HILLS JR.

New York" University Law School

AND DAVID SCHLEICHER

George Mason University Law School

Roderick M. Hills Jr. is the William T. Comfort III Professor of Law at New York University Law School.

David Schleicher is assistant professor of law at George Mason University Law School.

The authors gratefully acknowledge the research of Jonathan Herczeg, NYU Law School Class of 2008, whose unpublished paper on the history of New York City's 1961 zoning resolution provided valuable background on the politics of the city's zoning. This paper is adapted from the authors' paper "The Steep Costs of Using Non-Cumulative Zoning to Preserve Land for Urban Manufacturing," which will be published in Vol. 77 of the University of Chicago Law Review.