Hundreds of thousands of American homeowners may be in desperate straits, facing crippling rate resets or worse, foreclosure, but others continued to purchase vacation homes and investment properties at a healthy rate.

According to a study released late last month by the National Association of Realtors� (NAR) such second-home sales declined with the overall market in 2007, but second home sales still accounted for 33 percent of all new and existing home sales. The combination of these two sales accounted for 36 percent of the total in 2006.

21 percent of all homes purchased in 2007 were for investment purposes compared to 22 percent the year before. An additional 12 percent of the home purchases were vacation homes, down from 14 percent in 2006.

In terms of numbers, the Investment and Vacation Home Buyers Survey showed that the total number of primary sales declined 10 percent to 4.34 million in 2007 from 4.82 million the year before. Vacation home sales dropped 30.6 to 740,000 from what had been a record-setting 1.07 million homes in 2006. Investment property sales accounted for 1.35 million transactions compared to 1.65 million in 2006, a drop of 18.1 percent.

The median price spent on a vacation home in 2007 was $195,000, 2.5 percent lower than the median of $200,000 in 2006. Investment property was purchased at a median price of $150,000, unchanged from 2006.

Vacation homes were most popular in the South where 41 percent of them were purchased; 25 percent were bought in the West, 19 percent in the Northeast, and 16 percent in the Midwest. 30 percent of the homes purchased were in rural areas, 20 percent each in resort areas or suburbs, and 14 percent in a city neighborhood.

Investment properties were also disproportionally located in the South (38 percent) with 23 percent in the Northeast, 21 percent in the West, and 19 percent in the Midwest; 39 percent were purchased in a suburb, 21 percent in a small town, and 20 percent in an urban area.

The typical vacation-home buyer in 2007 was 46 years old, had a median household income of $99,100, and purchased a property that was a median of 287 miles from his or her primary residence; while the typical investor was 42, earned an income of $92,900, and bought a home that was relatively close to his or her primary residence � a median distance of 27 miles. Sixty-five percent of vacation home buyers and 71 percent of investment home buyers purchased existing homes, while the remainder purchased new homes.

Lawrence Yun, NAR chief economist, said the findings suggest different cycles for each of the sectors over the past two years. "Investment-home sales declined sharply in 2006 as speculators disappeared, leaving the market to serious buyers, with the pattern continuing in 2007," he said. "Vacation-home sales rose to a new record in 2006 because there was a pent-up demand from buyers who couldn't find a property as a result of tight supplies in preceding years."

The overall sales decline in 2007 resulted from a combination of factors. "Certainly, second homes are discretionary purchases and there is a natural tendency to pull back from big-ticket items in periods of uncertainty," Yun said. "The other factor is the disruption in the mortgage market, with a significant tightening of credit during the second half of 2007. Some buyers simply adopted a wait-and-see attitude."

Yun said lifestyle factors and strong demographics remain positive for the vacation home market. "Investment considerations are secondary for vacation-home buyers, so there is some dormant underlying demand," he said. "A peak of population is moving through the prime years for buying recreational property. It is welcoming to see investment sales returning to pre-boom sales activity."

NAR's 2007 Investment and Vacation Home Buyers Survey, conducted in March 2008, includes answers from 1,965 usable responses. The survey controlled for age and income, based on information from the larger 2007 National Association of Realtors� Profile of Home Buyers and Sellers, to limit any biases in the characteristics of respondents.