Garages in new American homes: fifty years of change

What we found

We joined two Census Survey of Construction files year by year, from 1973 to 2025. One tracks garages (parking_cust.xls). The other tracks house size (squarefeet_cust.xls).

The two move together almost in step: a correlation of 0.947 across 53 years. The median new single-family home grew from 1,525 to 2,142 square feet, up 40 percent. The share of new homes built with a garage rose from 65.2 to 89.3 percent, up 24.1 points.

The two times the link broke are the interesting part. In the financial crisis, median house size dipped in 2008 and 2009, then climbed past its pre-crisis level by 2012. Garage share fell from 89.8 percent in 2007 to 85.5 percent in 2010, and was only 87.4 by 2012.

The second break is happening now, and it looks different. From 2021 to 2025, for the first time in 53 years, both lines fall together. Median size fell from 2,273 to 2,142 square feet. Garage share fell from 92.2 to 89.3 percent.

Why compare garages with house size?

Because a garage is part of the same square-footage decision. The Census Survey of Construction publishes a set of files on new single-family homes, all from the same survey of builders. They cover parking, square footage, stories, foundation type, bedrooms, bathrooms, wall material and more.

Any two can be joined by year, because they describe the same homes. Neither series is derived from the other.

We chose square footage because it ties most directly to what a garage is. A garage is floor area under the same roof, built for a car instead of for living. When a builder decides how big to make a house, the garage bay is part of that same decision.

Stories and foundation type are real variables for a future page. But they describe the house’s shape and footing, not the square-footage choice a garage is part of.

HyreGarage analysis: that direct link is why the correlation should not surprise anyone, and why it is still worth publishing. A figure of 0.947 confirms the garage did not become standard on its own.

It became standard as part of the house getting bigger, through the same building decisions, in the same years and mostly the same places. The interesting part is the two times the link stopped holding.

In the financial crisis, houses kept growing but garages did not

Look at 2007 to 2012 in the chart and the two lines pull apart. Median house size dipped for two years, from 2,277 square feet in 2007 to 2,135 in 2009. Then it climbed through the rest of the crisis, reaching 2,306 by 2012, bigger than before the recession.

Garage share did not bounce back the same way. It was 89.8 percent in 2007 and fell to 85.5 percent by 2010, with one small uptick in 2009. It had only recovered to 87.4 by 2012, while size was already at a new high.

HyreGarage analysis: this looks like a clear builder choice, not chance. In a downturn, new homes skew toward buyers who can still get a mortgage. The Census size data shows those buyers did not buy smaller houses on average.

The number of homes built collapsed (a separate fact our flagship page notes), but the homes that did get built kept growing. What builders cut instead, on this chart’s evidence, was the garage. It was cheaper to drop than living space, which the listing photos show first.

That hints the garage is a cost lever for builders, not just a side effect of house size. If a bigger house always meant a garage, the two lines would not have split in the years when construction loans got harder to get. It is a hint, not proof.

Since 2021, houses and garages are shrinking together

The second break is recent, and the numbers show a new pattern, not a repeat of 2008.

From 2021 to 2025, median new-home size fell from 2,273 to 2,142 square feet, most of the way back to its 2019 level.

Garage share fell over the same years, from 92.2 to 89.3 percent, its lowest since 2010.

HyreGarage analysis: unlike 2008–2012, the two series are moving the same way, together, for the first time in the 53 years we can measure.

Housing-market commentary most often explains the size drop this way. After mortgage rates jumped in 2022, many buyers were priced out of the large, garage-heavy homes of the 2010s. Builders returned to smaller, lower-cost “starter” homes.

If that is right, it explains why garage share is falling with size this time. A small starter home is more likely to sit on a small lot, where a garage takes a bigger share of the space you can build on. The garage is not just a cheap line item to cut.

To be precise about the limits: this is the most common explanation in general housing commentary, not a cause stated by the Census. We have not checked a lot-size or starter-home series ourselves.

What the Census data shows on its own is narrower and solid: 2021–2025 is the only five-year period in 53 years when median size and garage share both fell at once.

What we did not try, and what would extend this page

A regional version of this correlation. The flagship page shows the regions caught up at very different speeds. Is the size-and-garage link equally strong in each region, or driven mostly by the South, which had the most room to change on both? The same files can answer that. We have not run it.

The stories and foundation comparisons. Both are outlined above as possible future pages rather than run here, partly because foundation type would likely just repeat the regional story.

A lot-size series. Our explanation for the 2021–2025 joint drop relies on smaller starter homes on smaller lots. That is a common account in housing commentary, but we have not checked it against a lot-size series of our own. We call it the most likely explanation, not a confirmed one.

Any claim about cause. This page shows a strong link and two periods where it breaks. It does not show, and does not claim, that either variable caused the other to move, in general or in either period.