Firms, Industries, and Unemployment Insurance: An Analysis Using Employer-employee Data from Canada - ARCHIVED

The exploration of newly available administrative data in a number of countries has led to a growing realization that a careful study of the interaction between employer and employee characteristics is needed to fully understand labour market outcomes. The objective of this paper is to develop this theme by examining the design of social policy and its interaction with the labour market. The focus is on the Canadian unemployment insurance (UI) program. This analysis uses administrative data on the universe of employees, firms, and UI recipients in Canada over an 11 year period to examine the operation of UI from the perspective of the firm, paying particular attention to longitudinal issues associated with the pattern and causes of cross-subsidies. The findings show that persistent transfers through UI are present at both industry and firm levels. These cross-subsidies are concentrated among a small fraction of firms. An analysis using firm fixed effect indicates that almost 60 percent of explained variation in persistent cross-subsidies can be attributed to firm effects. Calculations of overall efficiency loss are very sensitive to the degree to which firm level information is used. A full appreciation of how social programs like UI interact with the labour market requires recognition of the characteristics and human resource practices of firms, and might be more fruitfully explored by implicit contract models of unemployment.