Work Hours as a Source of Income and Employment Instability

 

Background and Purpose Recent analyses of national data suggest that one source of income instability is within-job fluctuations in earnings (Gottschalk & Moffitt 2009). In hourly jobs, fluctuations in the number of hours employees work is a likely source of earnings instability and a likely contributor to job loss (Kalleberg, 2009; Lambert, 2008). Unfortunately, survey items included in most national datasets do not capture variations in work hours because they ask about ‘usual' work hours. In contrast, the data used for this paper are well equipped to reveal variation in weekly work hours because they come from a retail firm's payroll system which records actual hours worked on a daily basis.

The current study addresses the following research questions:

Question 1. How much do employees' work hours fluctuate week to week?

Question 2: How are hour fluctuations related to employee age, race, tenure, job status (full-time/part-time) and occupation (assistant manager/sales associate)? 

Question 3: Do fluctuating hours interfere with workers' sustained employment? 

Methods The sample includes all hourly employees (N=233) in 19 retail stores in the Chicago area between May 2008 and January 2009. The workforce is exclusively female; average age is 43. Fully 78% of employees are white, 12% are African American, and 9% are Hispanic. ‘Hour fluctuations' is measured as the standard deviation in weekly work hours across the 37-week study period. ‘Sustained employment' is the proportion of months employed between the beginning of the study period through January 2010 (calculated from corporate records).  Descriptive and bivariate statistics of work hours are used to address Questions 1 and 2. Multivariate regression that adjusts for the clustering of employees by store is employed to estimate the relationship between work hours (both average number and hour fluctuations) and sustained employment, controlling for employees' age, race, seniority, and job type (Question 3).  

Results. Question 1: Across the study period, employees worked an average of 20 hours per week (range=4 to 45 hours). On average, employees' hours varied 5.7 hours week to week, but some employees experienced greater variation than others (range = .4 hours to 15 hours). Question 2: Employee age, race and tenure are not significantly related to work hours, although tenure is positively associated with average number of weekly hours. Full-time assistant managers worked significantly more hours and experienced greater hour fluctuations than part-time sales associates. Question 3: Both hour fluctuations and average total work hours are significantly related to sustained employment, after controlling for tenure, age, race and even occupation and job status. The more hours employees worked and the less hours fluctuated, the longer they remained employed at the firm.

Conclusion: In March 2010, the Obama administration held a summit on workplace flexibility. As efforts move forward, it will be important for social workers to bring the voices of workers in hourly jobs into the public discourse. The results of this study caution that variability in work hours does not equate with flexibility and may undermine rather than bolster employment security in low-level, hourly jobs.