Schedule(d) Volatility in Hourly Service Work: A New Measure and Its Policy Implications

• In Event: Determinants and Consequences of Work Schedule Quality for Child and Family Well-Being

Abstract

Background. In the US, work opportunities for those with lower levels of formal education have moved toward service employment, especially among households with children. Service work is characterized by high levels of volatility, which is primarily driven by employers, not workers. Service workers must be generally available to work whenever scheduled in order to maintain employment. 

Methods. This study uses data collected from a sample of hourly service workers with young children (aged 2-8), recruited through venue-based sampling at retail, food service, and hotel businesses in Philadelphia, PA in Fall 2019. Participants were sent daily text-message-based surveys about their work hours, including start and stop times and schedule changes, for 30 days. The analytic sample includes 750 participants who completed at least 22 daily surveys.   With these data, we create a novel measure of how much time each week hourly service workers supply for their jobs, given that workers must hold time open to be available if scheduled. To do so, we consider participants' work by day of the week (Monday, Tuesday, Wednesday, etc). We identify the earliest start time and latest stop time of shifts on that day of the week (3 to 5 observations for each day of the week), including time scheduled but not worked (ie. the start and stop times of shifts that were canceled and initially-scheduled stop times for days workers were sent home early). We then calculate the total hours ever worked or scheduled for each day of the week and sum across all days to get effective hours supplied each week. We use these data to answer the following questions:

- How much time do respondents supply for their jobs?
- How much do they actually work?
- What is the difference between supply and work–i.e., the wastage, or time respondents supply in order to maintain employment that is not ultimately demanded and paid for by their employer? 

Results. Our novel metric captures previously unmeasured volatility in service workers’ schedules, showing that workers supply, on average, 60 hours per week but only work 36. Nearly all participants (89%) supply at least full-time (40) hours per week; 18% actually work and get paid for full-time hours. Additionally, while the average number of days worked per week among our sample is 4.5, the majority (54%) of our respondents work each of the seven days of the week at least once during the month (in other words, they have no reliable day off). Finally, we find negative autocorrelation in work hours from week to week. This is a way to stabilize monthly income, so workers want to be able to make themselves available when hours are offered—otherwise monthly earnings and hours volatility would be even higher. At the same time, the more workers make themselves available, the higher their effective labor supply and potential wastage.  

Implications. These findings have implications for families' ability to maintain eligibility for the main federal government supports for families with children and to access tax credits designed to reward parents’ labor supply.

Author

• Anna Gassman-Pines, Duke University
  Presenting Author