Why Compact-Only Spaces Failed: A Post-Mortem

Walk any garage built between 1978 and 1995 and you will find them: rows of stalls striped a foot narrower than the rest, stenciled COMPACT ONLY, with a Chevy Tahoe parked across the line. The compact space is one of parking design's clearest engineering failures. Not because the idea was foolish, but because the conditions that made it work quietly disappeared, and the stalls stayed. Understanding why it failed is worth a designer's time, because the same failure is available to anyone who designs for a vehicle mix instead of a vehicle trend.

A design that worked, briefly

The compact-only stall was born honest. In the years after the mid-1970s oil crisis, the American fleet split into two distinct populations: the big domestic sedans and wagons already on the road, and a wave of genuinely small imports and downsized domestics. A bimodal fleet invites a bimodal layout, and designers obliged. In one common arrangement, angled stalls for large vehicles ran along one side of the aisle and tight 90-degree compact stalls along the other. The design enforced itself. A driver in a full-size sedan who tried a compact stall met a brutal turning movement and doors that wouldn't open. The geometry was the policeman. Efficiency gains were real: more stalls per floor plate, smaller footprint, lower cost per space.

The failure was not in the design. It was in the assumption underneath it: that the fleet would stay bimodal.

The middle ate the fleet

It did not. Starting in the mid-1980s, aerodynamics and engine efficiency let manufacturers meet federal fuel-economy standards without keeping cars small, so small cars grew. Large cars, meanwhile, had already been downsized. The two populations converged toward the middle of the size range, and the middle is exactly where a compact stall stops being self-enforcing. Today nearly half the vehicles sold in the US sit within a foot of length and a few inches of width of the traditional small/large boundary. A "large" car can park in a compact stall now. Uncomfortably, but it can. The geometry stopped policing.

The small car itself was also vanishing as a market category: roughly half of US vehicle sales in the 1980s, about 20 percent by the 2000s, under 10 percent through the 2010s and holding there through 2022. The compact stall became a reserved space for a customer who had largely stopped existing.

The two failure modes

Once self-enforcement died, compact stalls failed in both available directions, depending on where the designer put them.

Put them in convenient locations, near the elevator or the entrance, and drivers of intermediate and large vehicles take them anyway. An oversized vehicle in a compact stall encroaches on its neighbor, which pushes the next parker off-center, which propagates down the row until a stall somewhere becomes unusable. One poached space erases the efficiency gain of several compact stalls. The ripple is the whole point of the post-mortem.

Put them in inconvenient locations instead, and small-car drivers rationally take the standard stalls near the elevator, leaving late-arriving large vehicles a choice between the back of the roof deck and a compact stall they don't fit. They choose the compact stall. Same ripple, different cause.

There is no third placement. Compact stalls work only where a facility actively polices vehicle-to-stall matching, and almost no operator will fund enforcement of stall sizing. The revenue case isn't there and the customer-relations case is worse. A design that requires a permanent enforcement budget to deliver its designed efficiency is not an efficient design.

The residue

The compact stall survives in two places: older facilities that have never restriped, and municipal ordinances that still permit or encourage compact percentages as a density concession. The first is a restriping project. The second is a trap. An owner who accepts a compact allowance to hit a required space count on paper is building stalls that will underperform their count in practice, and the shortfall lands on the operator at exactly the peak hours the count was supposed to cover.