Mental health costs of economic change are measured in a cross-level analysis that combines aggregate economic conditions (unemployment rate or absolute change in employment levels) with individual self-reports of undesirable job and financial events, psychosomatic symptoms, and help-seeking for emotional problems. The data were collected in Los Angeles County during 1978-80 in 12 quarterly surveys. Data for 4,144 Anglo respondents are analyzed in seven-way contingency tables to test two competing classes of explanations of the reported association between economic change and the use of mental health facilities. One class assumes that economic contraction affects utilization by increasing the incidence of disorder. The other class assumes that economic change affects utilization without increasing the incidence of disorder, but by influencing the decision to seek help for chronic or anticipated disorder. The results support versions of both classes of explanations. The support, however, is qualified by several factors that suggest that the widely-cited association between the economy and mental health is more complex than usually portrayed.