Compass Research on Zillow Tax Criticized

Compass released research this week asserting that homes featured on Zillow sold for 1.3% less than those Zillow had banned, a figure they termed the 'Zillow tax' and estimated to be $5,590 per home. However, within two days of the publication, three distinct groups of economists challenged Compass's findings. Their criticisms focused on the sample size used, the omission of crucial control variables, and the fact that all listings analyzed originated solely from Compass's own business operations. These critiques collectively questioned the validity of the 'Zillow tax' designation. While these points were acknowledged as valid by many, the core issue identified by the economists was not the sample construction but rather what Compass's research actually measured. The initial claim that homes on Zillow sold for 1.3% less was deemed a significant misstatement. Compass's research release stated that the company examined 296,966 of its listings between January 2025 and May 2026. Out of this extensive dataset, 806 listings were identified as having been banned by Zillow. According to Compass's analysis, the banned homes achieved 100% of their asking price, whereas the homes listed on Zillow achieved 98.7% of their asking price. Compass then extrapolated this 1.3% difference, multiplied it by the average price of an American home, to arrive at the $5,590 'Zillow tax' figure. The fundamental flaw identified by the economists is that this calculation does not accurately reflect the actual amount of money a seller receives. The asking price, they argue, is not an objective fact about a house but rather a figure determined by the agent. To illustrate this point, consider two identical homes, each valued at $1 million. The first home is listed at $1 million and sells for $987,000, representing 98.7% of its asking price. The second home, listed by Compass (and not on Zillow) at $950,000, sells for its asking price of $950,000, achieving 100% of its asking price. Compass's methodology would suggest the second house performed better. However, the seller of the second house actually lost $37,000 compared to the first house, despite achieving 100% of its asking price. This discrepancy arises because Compass agents set the asking prices for both sides of this comparison. Therefore, Compass is effectively scoring a test that its own agents designed, a situation likened to a student writing, taking, and grading their own final exam. This methodological issue, where the entity conducting the analysis also controls the parameters of measurement, undermines the credibility of the 'Zillow tax' claim and suggests that the reported 1.3% difference is not a reliable indicator of actual seller proceeds or the impact of Zillow's listing policies.