A study examined the earnings losses suffered by a group of experienced workers who separated from their firms in the early and mid-1980s. The quarterly earnings histories of a large number of Pennsylvania workers covering the period 1974 through 1986 merged with employment information about their firms served as the data set. Workers' earnings losses were estimated by comparing their postdisplacement earnings to their expected earnings had they not been displaced. The relationship between earnings losses and various worker and employer characteristics was examined. Experienced workers were found to have incurred substantial immediate and persistent earnings losses after they separated from their firms. Even 5 years after separation, many displaced workers' earnings were still down by more than 25% of their predisplacement earnings. Moreover, displaced workers' earning began diverging from their expected levels 2-3 years before they left their firms. Existing government programs were found to be incapable of compensating for more than a small portion of displaced workers' losses because most losses accumulate after workers are reemployed. It was recommended that, instead of bolstering existing programs, policymakers assist displaced workers by introducing income or wage subsidies. (Contains 10 tables; 29 figures; 73 references.) (MN)