(25 percent). The coefficient on BUYEROUT is also significant in Fort Lauderdale and

Los Angeles, which have 8 and 4 percent of out-of-state buyers, respectively. These

findings add to the evidence that out-of-state buyers pay price premiums.

 The coefficient estimate on the variable representing sale-leaseback transactions,

SALELEASEBACK, is generally positive and significant in three markets: Phoenix,

Riverside / San Bernardino and Washington, DC. In one market, San Diego,

SALELEASEBACK is significant but negative. The negative coefficient of the variable in

Seattle is contrary to our expectation. One possible explanation is that the result may be

influenced by an outlier, since it is based on only 30 observations. Generally, the results

present evidence supporting our hypothesis that sale-leasebacks are associated with

higher transaction prices.

 The estimated coefficient on PORTSALE is positive and significant in three

markets (Las Vegas, Los Angeles, and Sacramento) and negatively significant in Seattle.

However, none of these markets have a large number of portfolio sales from which to

draw any meaningful conclusions about observed price increases (decreases) associated

with portfolio sales. In addition, in the two markets where portfolio sales are numerous,

Washington, DC and Chicago, no significant results are observed. Therefore, there is

only weak evidence that portfolio sales of office properties are associated with price

premiums.

 The estimated year dummies present a mixed picture that is quite different than

with apartment properties. Sales in years 2000 and 2001 tend to have lower prices than

sales in 1999. The dummy coefficients show that in 2000, prices in 9 markets were lower

than prices, all else equal, in 1999. In 2001, dummy variables in 5 markets are negative,