We use information from a unique survey of Dutch firms to compare the standard proxy for job creation, which is based solely on stocks of employment in individual firms at two points in time, to the correct measure that includes counts of actual jobs. Both are compared to a measure of labor turnover that counts movements of individuals into and out of jobs. We find: 1) The standard proxy for job turnover differs little from actual job turnover; 2) Most mobility is into and out of existing jobs, not to new nor from destroyed jobs; 3) A large fraction of all hires are by firms where employment is declining, and a large fraction of all fires are by firms where employment is expanding; 4) Worker turnover is roughly three times as large as job turnover; 5) Simultaneous hiring and firing exists and is mostly due to unobservable heterogeneity in the work force.