Expansive versus expensive cities

The WSJ points to a just-released study by Issi Romem of US cities and housing affordability which finds that expansive cities are less expensive. In other words, cities which increase their footprint or expand outwards, are more likely to keep housing affordable. The standard narratives on the growth of cities have tended to present an either-or relationship between the emergence of haphazard suburban sprawls as cities expand outwards and vertical development by liberalizing zoning regulations. This study questions that conventional wisdom.   

It captures the steady expansion of US cities from the forties onwards. Cities like New York, which expanded aggressively earlier have almost stopped growing.

In contrast, others like Atlanta continue to grow outwards even today.

The difference between these two types of representative cities is captured in the graphic below on housing prices. It clearly shows that housing prices have risen less in cities which have expanded outwards, even when their populations have increased faster.

Romem explains the relationship, 

Expensive cities gained population as well, but they did so despite the constraints on housing supply, and in the absence of such constraints their population would have grown much more. Legacy cities’ populations grew only slightly or even decreased, as indicated by the absence of a circle. The chart shows that, with the exception of legacy cities, housing price growth is inversely related to cities’ outward expansion. 

At least three things are driving the relationship. One, massive amounts of housing were built on rural land in expansive cities and helped keep housing prices there in check, whereas the restricted outward expansion of expensive cities limited their supply of housing and contributing to housing price growth. Even though correlation alone does not imply causation, there should be no doubt that cities’ degree of outward expansion affected their housing prices directly. Two, land use policy impeding densification – as opposed to expansion – is likely to be stricter in the same cities whose outward growth is curbed, and such impediments to densification contributed to housing price growth as well. Three, recall that housing price growth sets in motion a sorting process that yields a more affluent population, which is prone to tightening land use regulation. This process means that housing price growth can indirectly cause cities to expand more modestly, which once again contributes to the relationship in the chart.

For developing country cities, where urbanization is only just gathering pace, the choice is not between deregulation or geographical expansion. They need both. They are like New York of 1960s or earlier. They are also like the Atlanta of today, but with far more regulated antecedent zoning regulations. 

In the years ahead, these cities will face massive immigration pressures. The more successful cities will be especially vulnerable. Business-as-usual will leave them with unmanageable sprawls and their economic vibrancy will be crushed. These cities need aggressive deregulation by way of much larger Floor Area Ratios (FARs), calibrated releases of massive extents of vacant lands currently occupied by public agencies, and gradual expansion of the city boundaries. Unfortunately, none of these are happening in any satisfactory manner.