The answer to this question, along with a number of other intriguing insights about urban planning, emerge from an article by Roderick M. Hills Jr. and David Schleicher, "Can `Planning' Deregulate Land Use?" which appears in Regulation magazine (Fall 2015, pp. 36-41). Here's their answer:
Dean Lueck (University of Arizona) and Gary Libecap (University of California, Santa Barbara) have confirmed empirically that the existence of easily ascertained boundaries can substantially increase property values. Comparing “metes and bounds” lots (the boundaries of which are defined by irregular and individualized lines often following natural boundaries like hills, trees, and streams) with “rectangles and squares” (a system that originally allocated properties in standardized rectangular and square lots), Lueck and Libecap determined that “rectangles and squares” demarcation results in property values that are roughly 30 percent higher than “metes and bounds”—an effect persisting 200 years after the original demarcation. Lots defined by rectangles and squares attracted more population, urbanized more quickly, and—despite the property owners’ power to customize their lots after the initial demarcation—retained their geometrical boundaries long after they were initially demarcated.
Manhattan’s street grid suggests a similar effect of simple, uniform, and rectangular lot lines. In 1811, when New York was a city of just under 100,000 people residing almost entirely south of Houston Street, the state legislature authorized three street commissioners to create a uniform street grid covering almost all of Manhattan Island. With few deviations, the result was a uniform grid of numbered streets and avenues that, according to the commissioners, would lower the cost of construction because “strait-sided and right-angled houses” were cheaper and easier to build. The grid had another advantage: rectangular lots made property rights easy to ascertain, reducing property disputes between neighbors and making it easier for outsiders to buy, sell, and improve real estate. Trevor O’Grady, a post-doctoral fellow at Harvard University, has confirmed that properties on gridded blocks are both more valuable and more densely developed than properties with irregular boundaries in non-gridded areas.
This finding may seem unsurprising, but it has a provocative lesson for current urban planning. In many urban areas, exactly what you are allowed to build on a plot of land is not altogether clear, and it can often be more about political negotiations than about concrete and drywall. Want to build an apartment building or a condo complex? Prepare for a negotiation over height, parking spaces, the design of street-level entrances, and the inclusion of units designated as "affordable." Want to build an office building? City Hall will probably only give you a building permit after a negotiation over many features of the building, potentially including height, whether mass transit or public space is included in the design, parking spaces, the mix of retail and office space, and a number of other features.
In short, the two-dimensional square of land owned by a property developer may be well-defined by right angles. But the property rights concerning the three-dimensional box of what can be built on that plot of land are often not well-defined. Instead, a system of metaphorical "metes and bounds" includes many requirements that are negotiated anew during any given project.
Like everything else in economics, there are competing theories and a tradeoff at work here. On one side, there's an argument that bargaining over each individual project--typically in process that pits the would-be developer against those already in the neighborhood--helps to assure that new projects are blended into the existing neighborhood and fit with a broader public vision. Hills and Schleicher admit some truth in this argument, but also offer two strong counterarguments. In the context of thinking about requirements for affordable housing, they write:
In two important respects, this bargaining process defeats the very goal of land-use deregulation necessary for a lasting solution to the housing affordability crisis. First, individualized bargains increase the chance that NIMBY neighbors will hijack local land to exclude housing because the individualized bargaining process provides no way for politicians representing different parts of the city to strike bargains allocating locally unwanted land uses across neighborhoods. ... Local legislatures tend to be too disorganized to enforce complex deals for allocating locally undesirable land uses (LULUs) across legislators’ districts. The result is excessive zoning restrictions created by a universal coalition against LULUs: each member votes for every other member’s proposals to keep LULUs out in expectation that every other member will do the same....
Second, individualized bargaining raises the costs of knowing what development rights one buys with the purchase of a lot. Those costs drive away real estate investors, depriving the city of capital sorely needed to remedy a desperate housing shortage. ... Comprehensive plans also promote transparency and marketability for use rights in much the same way that a simple grid system promotes a market in possession rights. By making it simple for outsiders to see what they are buying when they purchase a parcel, such plans attract more buyers, thereby increasing investment in housing. In these two respects, centralized and rigid planning can actually be a libertarian reform, while ostensibly more flexible bargaining can lead to the strangulation of a city’s housing supply.
This insight is what leads to their seemingly paradoxical title: "Can `Planning' Deregulate Land Use?" General plans that specify what you can do with your land "as-of-right"--that is, without anyone being able to block you--may be a useful step toward avoiding regulation-by-negotiation.
Their lesson is broader than just land use planning. Many government policies can be implemented through general rules, or though a series of negotiations and exceptions and discretionary decisions. In specific situations, one can often make the case that the general rule doesn't quite work, at least not this time, and so tweaking and bending and amending the rule makes sense just this time. But if every decision comes down to a negotiation, then the general rule itself lacks force. Moreover, the outcome of real-world negotiations won't be decided by an All-Knowing and Benevolent Social Planner. Actual negotiations will be a messy business of who can hire the most high-powered lawyers, spend the most money on publicity, turn out the neighborhood crowds, and cut deals with local politicians and well-placed decision-makers for their support. As Hills and Schleicher conclude:
Paradoxically, sometimes rigidity and centralization of a general framework for buying and selling land is more market-friendly than the bargaining free-for-all that keeps everyone guessing—and paying lobbyists to improve the odds of their guesses.