2005-2007 Release: These are updates and corrections related to the CHAS data released in December 2009, which used 2005-2007 ACS 3-year data.

• HUD has discovered that part of the methodology for the 2005-2007 CHAS was inconsistent with similar affordability data (including the 2000 CHAS). For jurisdictions with incomes above the US median, the 80% AMI figure used in the 2005-2007 CHAS may be substantially higher than the figure used in the 2000 CHAS, independent of actual income growth. The full list of impacted communities is available here: Analysis of uncapped 80% AMI (*xls). Further explanation of the implications of this error, and how HUD will adjust going forward, is available here: Implications of uncapped 80% AMI (.doc).
• Renter households that paid no rent and had income greater than zero were treated inconsistently in different tables. In Table 8, these households were considered to have a cost burden of zero, and therefore counted among those with no cost burden. In all other tables involving cost burden (Tables 1, 2, 3, 4, 5, 6, 7, 9, 11, and 12), such households were mistakenly counted among those for whom cost burden could not be computed. As a result, Tables 1, 2, 3, 4, 5, 6, 7, 9, 11, and 12 will overestimate the number of renter households for whom cost burden could not be calculated and underestimate the number of renter households with no cost burden. Counts of cost burdened households are unaffected by this error. This also applies to the housing problems variable, which includes cost burden (Tables 1, 2, 4, 5, 6, and 11). In the future, the Table 8 specification will be used. “Cost burden could not be computed” will only refer to households with zero or negative income; for an explanation of why these households are not assumed to be cost burdened, see the 2007 Worst Case Needs Report pages 84-85 and 89.