A nationwide statistical study released in May adds further weight to existing evidence about the correlation between certain house components and the price of the home.

Written by Florida State University professors G. Stacy Sirmans and David Macpherson, the report follows up on 2003 research that examined the Philadelphia housing market. The earlier study looked at more than 140 house characteristics — “I like to say we examined everything, including the kitchen sink,” Sirmans jokes — but this new data looks at only a handful.

The study, sponsored by the National Association of Realtors, examined results from dozens of smaller, regional studies, and certain characteristics were left out because the sample size was not big enough to be statistically relevant.

Square Footage The researchers found that, on average, for every 1,000 square feet added to a home, the sale price increased by over a third. Fireplaces, air conditioning, garages, and pools all affected the value of the home by 8% to 10% (see chart, right). Kitchen and Bath Adding a bathroom increased the sale price by 8.7% on average, more than double the rate for adding a bedroom. That is in concert with what Sirmans found in his 2003 study, and what he says he's heard anecdotally from people in tune with the industry. “When somebody walks into a house, they may look at other things, but if they like the bathroom, and the kitchen, they'll buy the house.”

While the report provides useful information about the correlation between home improvements and house prices, the real reason Sirmans undertook the study was to determine if the conclusions drawn in Philadelphia in 2003 were applicable to other parts of the United States. He says that although the earlier study received widespread attention, doubts remained about its broad relevance.

“People said, ‘Hey, that's Philadelphia. What happens in Atlanta may be completely different,'” Sirmans says.

Vindication But, according to Sirmans, the national picture seems to be generally consistent with the results of the first model. “There wasn't a whole lot of difference from one region to another,” he says. “There were a few small variations, but in general it seems to indicate that maybe market preferences are more universal than what people always thought.”

Indeed, of the nine characteristics studied, the six that could be considered house “components” showed geographical variation in only one region each. Characteristics such as square footage and lot size were sensitive to location, no doubt due to the vast differences in home and land prices across the country. —Ted Cushman is a journalist in Great Barrington, Mass.