In response to recent reports, especially by CoreLogic that the National Association of Realtors' (NAR) Existing Home Sales data was overstated by as much as 15% to 20%, NAR released a new Frequently Asked Questions (FAQ) section to their website explaining the benchmarking process and discussing revisions to the process.
Without acknowledging potential errors in the data, NAR explained that the benchmarks used to calculate changes in home sales use the U.S. Census data compiled once every ten years. They noted that following the 2000 Census, they identified an upward drift in their data estimated at 13 percent that led to revising their data from the 1990s.
The data from the 2000 Census became the benchmark for the last decade. NAR has noted that due to the fact that the Census no longer uses a long-form questionnaire that includes questions about moving and purchasing home, in addition to the infrequency of the reporting, that a new methodology for benchmarking is required. They anticipate a new system to be launched sometime this summer and will provide the ability to re-benchmark on a more frequent basis.
It is expected that their new benchmarking system will also lead to revisions to the data from 2000 to 2010, especially from the tail end of the decade, 2009 and 2010.