Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/63977 
Year of Publication: 
2003
Series/Report no.: 
ETLA Discussion Papers No. 861
Publisher: 
The Research Institute of the Finnish Economy (ETLA), Helsinki
Abstract: 
Linked employer-employee data from the Finnish business sector is used in an analysis of worker turnover. The data is an unbalanced panel with over 219 000 observations in the years 1991-97. The churning (excess worker turnover), worker inflow (hiring), and worker outflow (separation) rates are explained by various plant and employee characteristics in type 2 Tobit models where the explanatory variables can have a different effect on the probability of the flow rates to be non-zero and on the magnitude of the flow rate when it is positive. Most of the characteristics are defined as 5-group categorical variables defined for each industry separately in each year. We compare the Tobit results to OLS estimates, and also use weighting by plant employment. It turns out that weighted OLS results are fairly close to Tobit results. The probabilities of observing non-zero churning, inflow, and outflow rates increase with plant size. The magnitudes of the non-zero churning and inflow rates depend positively on size, but the magnitude of outflow rate negatively. High-wage plants have low turnover, whereas plants with large within-plant variation in wages have high turnover. Average tenure of employees has a negative impact on turnover. High plant employment growth increases churning and separation but reduces hiring in the next year. We also control various other plant and average employee characteristics like average age and education, shares of women and homeowners, foreign ownership, ownership changes, and regional unemployment.
Subjects: 
worker turnover
churning
employer-employee data
JEL: 
J23
J41
J63
Document Type: 
Working Paper

Files in This Item:
File
Size
329.32 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.