At a basic level, a city can be understood as two types of spaces: public spaces and private spaces. For as long as humans have built cities, we have been deeply interested in the top-down design of public spaces, from Hippodamus’ street grid to the monumental plazas of Tenochtitlan. The centralization of this element of cities makes sense: urban streets, parks, and plazas are capital- and land-hungry public goods, which might be underprovided if a central authority did not make sure land was set aside for them. At the same time, the design of private spaces has historically been a largely distributed affair, with the built landscape between blocks and parks designed in a decentralized process among thousands of smallholders, including residents, businesses, developers, and architects.
This changed in the twentieth century. In virtually every major American city, local governments regulate every detail of urban design, including maximum building heights, maximum lot coverage, minimum setbacks, and maximum floor area ratios. While the original intention of these rules varies, their cumulative effect is to completely control the minutiae of how we develop urban land, such that in most cities, the job of an architect or site planner is mostly one of rote compliance. On top of these traditional controls, cities are increasingly adding explicitly design-oriented regulations, including things like ground-floor transparency requirements, contextual setbacks, and design rules for facades.
A steelman argument for this centralized approach to urban design might read something like this: The way buildings engage with the public realm—including facades and site plans—can impose positive or negative side effects, or externalities, upon neighbors. For example, along a residential street, a building with the same facade treatment and deep setback as its neighbors can create an orderly streetscape, producing a positive externality. Along a walkable commercial corridor, on the other hand, a building with a deep setback accommodating a large surface parking lot might undermine the success of the corridor as a whole, producing a negative externality. Thus, it is incumbent on planners to control urban design, both to cultivate positive externalities and prohibit negative externalities.
There are two issues with this system: First, planners often lack the knowledge needed to know when to attempt to control urban design at all. Unlike traditional externalities, such as noise, smoke, light, or odor, “good” and “bad” design are ambiguous and subject to disagreement among reasonable people. A design that might seem to you like a disruptive and ugly oddball might seem to someone else like a welcome and exciting bit of variation.