The most important graph of 2012? We asked around for some suggestions last week. We got 34 responses. This was mine. It answers the question: What share of each recovery since 1970 came from selling houses and cars?

Houses_and_Cars_Growth2.PNG

The message is that home and car sales power recoveries. This recovery is different from all others because we just. Aren't. Selling. Enough. Houses.

That might be changing. With home prices rising, construction hours-worked recovering, and multi-family homes making a sustained comeback, 2013 could be the year our economy breaks out of "new normal" growth and gets back to "normal normal" growth.

Behind my optimism is a trend that doesn't get a lot of play in some corners of the financial press. It's household formation. Household means is a group of people living together. It can be six roommates, a four-person nuclear family plus a grandmother in the guest room, or a a young couple of two. Formation means one more of those categories. More formations is good news. It suggests more people getting jobs, getting apartments, getting married, having kids, and (in all likelihood) spending more money to furnish their new households and express their independence.

This recovery, however, has been a story of few jobs, crowded apartments, low marriage-rates, and low birthrates. It all comes down to households. In 2007, household formation (in RED) went horizontal, clearly diverging from our two-decade growth trend (graphs below via Credit Suisse):

Screen Shot 2012-12-21 at 12.11.04 PM.png

Unemployment among Millennials is about twice the national average, and real wages for young people have declined outright since 2007. As a result, one in three older teens and twentysomethings reported moving back in with their parents. That means they weren't starting new households. They weren't paying rent, taking out mortgages, buying furniture, paying separate utility bills -- all of which fall under the Housing Category, which accounts for nearly a fifth of GDP.

Consider Florida, our fourth-largest state economy and perhaps the worst-hit by the housing crash. The graph below shows the percentage of 25- to 34-year-olds who head a household (renting or owning). "In 2006, half of Floridian young adults had their own place," Credit Suisse reports.  Five years later, 20 percent of that group had moved in with their parents or somebody else. That's an astounding demographic shock to a real-estate-centric economy.

Screen Shot 2012-12-21 at 12.22.44 PM.png

Okay, but here's the good news. Household formation is miserable now, but it's projected to pick up for a simple reason: an improving economy is bound to encourage young people to get out, buy apartments, and get married, eventually. How fast they start gobbling up apartments and houses is unclear. But Credit Suisse makes three projections: No recovery (unlikely), strong recovery (possible), and consensus recovery (plausible). Here's the impact of each recovery speed on US household formation over the next year.

Screen Shot 2012-12-21 at 12.22.28 PM.png

And here is how that would translate into more spending on houses (or "residential investment"). Basically, a strong household recovery would coincide with a residential investment boom that took us to 2005 highs.

Screen Shot 2012-12-21 at 12.20.50 PM.png

Housing probably isn't going to snap back to its pre-bubble peak in the next year. But even normal growth in residential investment would be huge. If residential investment simply returns to its long-term average (going back to the 1990s), "it would add 1.7 percentage points to overall growth in the coming year," Neil Irwin reported for the Washington Post, which would put overall growth in the coming year at about 3.2% -- almost twice as strong as economic growth in 2011, the year that supplies most of these graphs' data.

Housing is the key. And it all starts with formation.

>

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.