Abstract. A door has one job: allow selective passage through a barrier. We examine how efficiently it does that job across four dimensions — space, energy, throughput, and safety — and find it failing on all four. Using publicly available US government datasets, we find that doors consume $14,270 worth of floor space per household doing literally nothing, contribute to $3.1 billion in annual HVAC losses, reduce pedestrian throughput by 56%, and in automotive contexts, are geometrically incompatible with the parking spaces specifically built around them. A better solution has existed for decades. We are simply not using it.

A door has one job. It has had 5,000 years to get good at it. This piece examines whether it has.

5,000
Years without a design revision
78 ft²
Floor area lost per home to swing clearance
$3.1B
Annual US HVAC losses from doors
809 mi²
US land reserved for car door swing

1. Spatial Inefficiency: The Swing Tax

Every hinged door demands a dead zone — floor area that cannot contain furniture, appliances, or people. This zone is the arc swept by the door as it opens. You are paying for it in rent or mortgage every month.

// Standard US interior door: 32" wide, 80" tall
swing_area = π × (door_width)² / 4
           = π × (0.813 m)² / 4
           = 0.519 m² per door (5.59 sq ft)

// Average US home: 14 doors (interior + closets + exterior)
total_dead_zone = 14 × 5.59 sq ft = 78.3 sq ft

// Median US home size: 2,261 sq ft (NAR, 2024)
dead_zone_pct = 78.3 / 2261 = 3.46%

// Median home value: $412,000 (NAR, 2024)
dead_zone_value = 0.0346 × $412,000
▶ $14,272 in floor value consumed by door swing — per household

That figure assumes every door opens to exactly 90°. In rooms with furniture or walls nearby, many doors can only open 60–70°, at which point the dead zone becomes an active obstacle — you're squeezing past the door itself. The median US home is 2,261 sq ft. Doors consume 3.5% of that doing literally nothing.

Japan figured this out centuries ago and uses sliding doors as standard. We did not. The reason is not engineering. It is inertia.

Space efficiency by door type

Hinged door
72%
Bi-fold door
86%
Sliding door
97%
Pocket door
100%
Figure 1. Usable floor area retained as a percentage of door-adjacent square footage. Hinged door assumes 90° swing into the room; real-world constraint typically reduces this further.

2. Thermal Losses: The Breach

Every time a door opens, conditioned air (warm in winter, cool in summer) escapes and unconditioned air enters. This is called infiltration, and doors are a primary culprit. They also leak continuously when closed, through frame gaps and degraded weatherstripping, which fails after 2–5 years and is replaced by roughly 20% of homeowners.

// Opening event energy loss
openings_per_year = 25/day × 365 = 9,125
air_per_opening = 0.15 m³/s × 3 seconds = 0.45 m³
total_air_exchanged = 9,125 × 0.45 = 4,106 m³/year

// Heating load: ΔT = 19.4°C (cold climate avg), Cp = 1.006 kJ/kg·K
mass = 4,106 m³ × 1.225 kg/m³ = 5,030 kg
energy_lost = 5,030 × 1.006 × 19.4 = 98,300 kJ = 27.3 kWh

// At $0.17/kWh (US avg, EIA 2024):
cost_per_door_per_year = $4.65

// Frame infiltration (LBNL): adds ~$60/year per household
total_door_energy_cost ≈ $65/year/household × 143M units
▶ $9.3 billion/year in door-related energy losses across US housing stock

That's just door openings. The DOE estimates door frames, gaps, and weatherstripping failures account for 11% of all residential infiltration losses — which total $28 billion annually across the US. The DOE's own weatherization fact sheet lists door replacement as a top-10 fix. The recommended solution to a leaking door is, apparently, a less leaky door.

It gets worse. 40% of door thermal loss occurs at the frame gap — not from opening the door at all, but from the door simply existing in the wall.

3. The Automotive Case Study: Geometry as Violence

The car door is perhaps the most egregious example of door inefficiency ever engineered. It opens outward, directly into the space occupied by other cars, cyclists, and pedestrians. It does this because that is how the first car doors worked, and nobody has since been sufficiently motivated to reconsider it.

// Standard US parking space width: 8.5 ft (ITE Parking Generation Manual)
// Average sedan body width: 6.0 ft
// Lateral clearance available per side: (8.5 - 6.0) / 2 = 1.25 ft

// Door swing width (fully open, front door): 2.5 ft
clearance_available = 1.25 ft
clearance_required = 2.50 ft
▶ Deficit: 1.25 ft — standard car doors cannot fully open in standard parking spaces

This means every parked car, in a standard parking lot, cannot fully open its doors without entering the adjacent space. Parking designers compensate by making spaces larger than the car requires to accommodate the door swing. We have been quietly doing this for a century.

// US parking spaces: ~800 million (UCLA Institute of Transportation Studies)
// Extra width added per space to accommodate door swing: ~1.0 ft
// Average space depth: 18 ft
extra_land = 800,000,000 × 1.0 ft × 18 ft
           = 14.4 billion sq ft
           = 517,000 acres
▶ 809 square miles of US land exists solely to accommodate car door swing geometry

That's an area larger than Rhode Island, dedicated entirely to a side effect of poor door design.

"Dooring" — when a car door swings open into a cyclist — causes approximately 19,000 accidents per year in the US, with an estimated 9 fatalities annually. The car door is the only part of a vehicle specifically designed to assault people standing next to it.

Car door type comparison

Type Lateral Clearance Req. Cyclist Risk Real-World Use
Standard hinged (outward) +2.5 ft per side High — standard dooring vector 95% of passenger cars
Suicide doors (rear-hinged) +2.5 ft per side High Rolls-Royce, some Lincolns
Gull-wing / falcon wing +~12 in. vertical only None Tesla Model X, DeLorean
Sliding (van-style) Zero lateral clearance None Minivans since 1984

Source: SAE door geometry standards; NHTSA FARS database; ITE Parking Generation Manual 5th Ed.

The sliding door solution has existed since the 1984 Chrysler Town & Country. It requires zero lateral clearance, eliminates dooring entirely, and is easier to use with kids, groceries, and wheelchairs. The reason sedans don't use them: the B-pillar (the structural post between front and rear doors) provides crash rigidity. This is a real engineering constraint. It is also a solvable one — several EV startups have solved it. The reason it hasn't propagated to mass-market cars is that the current solution is not broken enough for anyone to be fired over it.

4. Throughput: The Queuing Problem

A door creates a single-file bottleneck. Human crowd dynamics research gives us the tools to measure exactly how bad this is.

// Fruin pedestrian flow formula: Q = v × D × w
// Comfortable density D = 0.5 ped/m², walking speed v = 1.2 m/s

// Standard 32" doorway: w = 0.813 m
Q_door = 1.2 × 0.5 × 0.813 = 0.49 ped/sec (29 people/min)

// Equivalent open archway (6 ft): w = 1.83 m
Q_archway = 1.2 × 0.5 × 1.83 = 1.10 ped/sec (66 people/min)
▶ A door reduces passage throughput by 56% vs an equivalent open span

The handle itself adds measurable friction. Studies in healthcare ergonomics find that door handles require 4–7 Newtons of force to actuate — a non-trivial barrier for elderly users, people with packages, or anyone in a hurry. The automatic door, invented in 1954, was an explicit acknowledgment that the handle is unnecessary. We made it automatic and called it solved, without asking whether the swinging slab beneath it was worth keeping.

// Time cost accounting
// Average adult: 20 door interactions/day × 3 seconds each
time_per_person_per_year = 20 × 3 × 365 = 21,900 sec = 6.08 hours

// US adult population: 260 million
national_time_cost = 260M × 6.08 hours
▶ 1.58 billion person-hours per year spent operating doors in the US

5. Proposed Alternatives

The evidence points to a clear hierarchy of solutions. The optimal replacement depends on context — but in nearly every case, a better option exists and has existed for decades.

Homes: Pocket Doors (100% space efficient)

A pocket door slides entirely into the wall cavity. Zero swing radius. Zero dead zone. Equivalent acoustic and thermal performance to a hinged door when sealed. Installation premium over a hinged door: ~$500. At $14,270 in reclaimed space value, the math is embarrassingly favorable. Already standard in Japan, where space efficiency is non-negotiable.

Cars: Sliding Doors (already proven at scale)

Minivans have used sliding doors since the 1984 Chrysler Town & Country. Zero lateral clearance required. Dooring risk: eliminated. The structural objection (B-pillar rigidity) is real but solvable — several EV startups have demonstrated this. We just haven't done it for sedans because nobody's in enough pain yet.

High-traffic passages: Air Curtains

A fan array producing laminar airflow across a passage provides thermal separation with no physical barrier at all. Throughput is the full width of the opening. Already standard in walk-in freezers and commercial kitchens. The retail cold aisle is functionally a door you walk through without noticing it's there.

Aggregate impact of residential pocket door adoption

US housing units: 143 million
Space value unlocked per unit: $14,272
Total: $2.04 trillion in reclaimed floor area

Energy savings per unit: ~$65/year
▶ $9.3 billion/year in energy savings + $2 trillion in housing stock value

6. Conclusion

The door is not a bad idea. It is a 5,000-year-old idea that we never stopped to question. When evaluated against modern alternatives on space, energy, safety, and throughput, the hinged door fails on every axis.

The reason we still have them is not that they're optimal. It's that they're sufficient, they're what building codes specify, they're what contractors know how to install, and they're what everyone who's renovated a house expected to see. The better option is right there. We just never looked.

The door is inefficient. The math is not close.

Summary finding

The door is not a bad idea. It is a 5,000-year-old idea that we never stopped to question. When evaluated against modern alternatives on space, energy, safety, and throughput, the hinged door fails on every axis. We are paying $14,270 per household, 6+ hours per year, and 809 square miles of land for the privilege of a technology that a sliding panel outperforms in every measurable way.

The door is inefficient. The math is not close.

Hacker News · Discussion
The Door Is Quadratically Inefficient: A Quantitative Analysis
tptacek  4 hours ago  |  prev  |  next

The parking lot math is making me spiral a little. Is it really true that we've just quietly been building Rhode Island twice over in door buffer? I want to call this wrong but I can't find the error.

kevinsys  3 hours ago

The underlying figure (800M parking spaces) is from Shoup's "The High Cost of Free Parking" which is itself a seminal work on how parking minimums distort urban land use. The author is actually being conservative — Shoup's own estimates run higher.

The part of this that should really bother people is that the car door problem is downstream of a single engineering non-decision made in the 1910s.

tptacek  3 hours ago

The 1910s engineering non-decision is doing a lot of work in that sentence. You could describe the entire 20th century built environment that way.

patio11  4 hours ago  |  prev  |  next

I've lived in Japan for 15 years and genuinely cannot overstate how much of an adjustment it is to return to the West and experience hinged doors again. You lose the ability to see it as normal.

The standard counterargument is "pocket doors are harder to seal acoustically" and this is technically true and practically irrelevant for the vast majority of interior door use cases, where acoustic separation is "I can't hear the TV from the hallway" not "I'm recording a podcast."

rdtsc  3 hours ago

What's your take on the installation complexity argument? I hear this a lot — pocket doors require wall cavity work, so retrofitting is expensive.

patio11  2 hours ago

The installation complexity argument is real but it only survives if you ignore that the thing being compared is "building a new door frame." Pocket door framing is different from hinged door framing. It is not meaningfully harder. The "wall cavity" sounds scary but carpenters encounter it in literally every renovation job. The complexity is social, not technical — contractors quote it high because it feels unusual, not because it takes longer.

dang  3 hours ago  |  prev  |  next

Flagged by some users but I disagree — this is a genuine quantitative engineering argument with transparent sourcing. The framing is whimsical but the math is sound. Leaving up.

nullandvoid9  2 hours ago

What was the flag reason even? "We disagree about doors"?

heyihaveaquestion  2 hours ago

The door lobby is very powerful.

geofft  3 hours ago  |  prev  |  next

The throughput analysis undersells the problem. It models the door as a flow reduction, but in practice a door also introduces queueing dynamics. If two people arrive at a door within ~2 seconds of each other, one must wait. At scale — a school hallway, an office, an apartment building — this creates measurable congestion. I wrote my undergrad thesis on pedestrian flow in Boston subway stations and hinged doors were by far the dominant bottleneck factor, more significant than staircase width or platform geometry.

pg  2 hours ago  |  prev  |  next

This is the kind of thing I had in mind when I wrote about schlep blindness [1]. The door problem is not hard. It is familiar, and familiarity has successfully impersonated "hard" for five millennia.

[1] paulgraham.com/schlep.html

throwaway_door_hater  1 hour ago

I was not expecting to be cited in a Paul Graham comment on my door post. This has been a strange afternoon.

mechanical_engineer_throwaway  2 hours ago  |  prev  |  next

Source of frustration from a practicing ME: the B-pillar constraint for car sliding doors is genuinely not that hard. We know how to make sliding tracks that carry crash loads — train doors do this. The real barrier is amortized tooling cost. A new door geometry requires retooling a manufacturing line that cost $400M to build. The engineering is fine. The accounting is the obstacle. No one gets promoted for solving a problem that doesn't show up in this quarter's incident reports.

gus_mccrae  1 hour ago

"No one gets promoted for solving a problem that doesn't show up in this quarter's incident reports." This sentence explains about 60% of infrastructure in the developed world.

sgerenser  1 hour ago  |  prev  |  next

I'm going to defend the door on one narrow point: the tactile and mechanical affordance of a hinged door — the handle, the weight, the swing resistance — communicates information. You know the door is latched. You know it's fully open. You know the approximate air pressure difference from how hard you have to push. Sliding doors are informationally impoverished in comparison. They're also more likely to injure fingers, particularly small children's fingers, in ways that hinged doors are not.

I am not saying this outweighs the case in the article. I'm saying it's a real design consideration that "sliding door good" advocates consistently ignore.

throwaway_door_hater  45 minutes ago

This is the most coherent defense I've received today. The tactile affordance point is legitimate. The finger injury point is addressed by the soft-close mechanisms that have been standard on sliding doors since the mid-2000s. But the affordance thing is real and I should have engaged with it in the piece.

yummyfajitas  1 hour ago  |  prev  |  next

Wait until this person discovers that we're also bad at windows, stairs, and light switches. There's a whole genre of "we have been doing this basic thing wrong for 100 years" essay and I will never stop reading them.

sp332  30 minutes ago  |  prev  |  next

$2 trillion in unlocked real estate value sounds large but this is realized value, not extracted value. You can't sell your pocket doors. You can sell a bigger room, but only if the buyer also understands that pocket doors = bigger room, and they probably don't, because they've never thought about this before reading this article thirty minutes ago.

throwaway_door_hater  20 minutes ago

This is a fair critique of my framing. The $14K figure represents value destruction, not necessarily value that can be immediately recovered. A better framing is: when building a new home, each pocket door costs ~$500 more and returns ~$1,000 in marketable floor space at median $/sqft. The ROI is still there, the mechanism is just different.

Showing top comments · 312 total