This paper measures the heterogeneity of establishment-level employment changes in the U.S. manufacturing sector from 1972 to 1986. It examines gross job creation and destruction, as well as the rate at which jobs are reallocated across plants. Using a rich dataset with over 860,000 observations on 160,000 manufacturing establishments, the study quantifies the connection between job and worker reallocation, evaluates theories of plant-level employment dynamics, and identifies new results related to labor market cyclical behavior. The study finds that manufacturing's average annual gross job creation and destruction rates are 9.2% and 11.3%, respectively. These rates reflect high job creation and destruction within narrow economic sectors. The job reallocation rate, defined as the rate at which employment positions are reallocated across establishments, is calculated by summing the rates of job creation and destruction. The high rates of job reallocation indicate that the reshuffling of employment opportunities across plants is a major reason workers change employers or move between employment and joblessness. Combining data from the Longitudinal Research Datafile (LRD) and the Current Population Survey, the study calculates that 35-56% of worker reallocation between employers or between employment and joblessness is due to job reallocation. The study also finds that most annual job creation and destruction reflects persistent establishment-level employment changes. For example, 73% of jobs created between 1974 and 1975 still existed in 1976, and 72% of jobs lost in 1974-75 were still lost in 1976. The average one-year persistence rates for job creation and destruction are 68% and 81%, respectively. This persistence indicates that most job creation and destruction cannot be implemented by temporary layoffs and recalls. Job destruction is highly concentrated, with only 23% accounted for by establishments that shrink by less than 20% over a year. This finding indicates that most job destruction cannot be accommodated by normal worker attrition rates. Together, these results suggest that job reallocation is typically associated with long-term joblessness and/or worker reallocation across employers. The study also investigates the sources of heterogeneity in establishment-level employment changes. It finds strong relationships between the intensity of job reallocation and observable plant characteristics like age, size, and ownership type. The study also draws on several theories of plant-level heterogeneity and dynamics to explain simultaneous job creation and destruction within narrowly defined sectors. The study concludes that any successful explanation for the magnitude of job reallocation must also explain why high rates of job creation and destruction occur among mature plants in narrowly defined sectors. The study also finds that the job reallocation rate exhibits significant countercyclic time variation, with rates ranging from 17% in 1980 to 23%This paper measures the heterogeneity of establishment-level employment changes in the U.S. manufacturing sector from 1972 to 1986. It examines gross job creation and destruction, as well as the rate at which jobs are reallocated across plants. Using a rich dataset with over 860,000 observations on 160,000 manufacturing establishments, the study quantifies the connection between job and worker reallocation, evaluates theories of plant-level employment dynamics, and identifies new results related to labor market cyclical behavior. The study finds that manufacturing's average annual gross job creation and destruction rates are 9.2% and 11.3%, respectively. These rates reflect high job creation and destruction within narrow economic sectors. The job reallocation rate, defined as the rate at which employment positions are reallocated across establishments, is calculated by summing the rates of job creation and destruction. The high rates of job reallocation indicate that the reshuffling of employment opportunities across plants is a major reason workers change employers or move between employment and joblessness. Combining data from the Longitudinal Research Datafile (LRD) and the Current Population Survey, the study calculates that 35-56% of worker reallocation between employers or between employment and joblessness is due to job reallocation. The study also finds that most annual job creation and destruction reflects persistent establishment-level employment changes. For example, 73% of jobs created between 1974 and 1975 still existed in 1976, and 72% of jobs lost in 1974-75 were still lost in 1976. The average one-year persistence rates for job creation and destruction are 68% and 81%, respectively. This persistence indicates that most job creation and destruction cannot be implemented by temporary layoffs and recalls. Job destruction is highly concentrated, with only 23% accounted for by establishments that shrink by less than 20% over a year. This finding indicates that most job destruction cannot be accommodated by normal worker attrition rates. Together, these results suggest that job reallocation is typically associated with long-term joblessness and/or worker reallocation across employers. The study also investigates the sources of heterogeneity in establishment-level employment changes. It finds strong relationships between the intensity of job reallocation and observable plant characteristics like age, size, and ownership type. The study also draws on several theories of plant-level heterogeneity and dynamics to explain simultaneous job creation and destruction within narrowly defined sectors. The study concludes that any successful explanation for the magnitude of job reallocation must also explain why high rates of job creation and destruction occur among mature plants in narrowly defined sectors. The study also finds that the job reallocation rate exhibits significant countercyclic time variation, with rates ranging from 17% in 1980 to 23%