We analyze the relationship between housing and the business cycle in a set of 51 U.S. cities. Most surprisingly, we find that declines in house prices are often not followed by declines in employment. We also find that national permits are a better leading indicator for a city’s employment than a city’s own permits.