Attachment O - 2017 Advance Notice of Schedule Question Discussion

Advance notice of work schedules

In addition to questions about workers’ access to leave, work schedules, and ability to work from home, the proposed 2017 Leave Module to the American Time Use Survey (ATUS) includes questions about advance notice of work schedules, workers’ control over their schedules, and flexible start and stop times. These questions will provide an additional dimension to analyses of workers’ job flexibility data.

The sponsor of the Leave Module, the Department of Labor’s Women’s Bureau, expressed interest in including these additional questions to measure precarious schedule arrangements. They cited questions used in the National Longitudinal Survey of Youth 1997 (NLSY97), including one on advance notice specifically.

Advance notice is an important indicator for the predictability of one’s work schedule. The less notice workers have about their work schedules, the less ability they have to rearrange personal responsibilities to meet their work requirements. Schedule unpredictability is especially problematic for parents who may have difficulty arranging child care with short notice. Employees seem to be divided between two main groups: one with very short notice and one with considerable advance notice. This “predictability gap” has not received much attention previously.

BLS included the following question on advance notice in the initial draft of the proposed Leave Module Questionnaire. This question was fielded in the NLSY97 from 2011 to 2015, and also included in the 2014 General Social Survey. The Women’s Bureau wanted to ask this question of workers who, in an earlier question, said that their employers decide their work schedules without employee input.

Advance Notice Question from NLSY97, Rounds 15-17

How far in advance do you usually know what days and hours you will need to work?

• One week or less

• Between 1 and 2 weeks

• Between 3 and 4 weeks

• 4 weeks or more

BLS developed the Leave Module questionnaire in February and March of 2016 and conducted cognitive testing in April 2016. The first test cases took well over five minutes to complete. BLS sets a limit of five minutes on ATUS modules to limit respondent burden. BLS staff began identifying questions to eliminate in order to decrease the length of the module in mid-April.

It was at this same time that a BLS researcher recommended that ATUS staff consider capturing more detail than “one week or less” when recording responses to the advance notice question. This recommendation was made because results from the 2011-12 NLSY97 show that 38 percent of early career employees know their work schedules one week or less in advance¹.

Based on this suggestion, BLS staff discussed changing the question with senior survey methodologists at BLS who were also involved with the questionnaire design and testing. BLS staff considered further refining the response options to include the following:

• Less than 1 week (with a follow-up probe, “How many days?”)

• Between 1 and 2 weeks

• Between 3 and 4  weeks

• 4 weeks or more

Respondents who said they had less than one week notice would be asked, “How many days?” However, BLS staff were concerned about adding followup questions, as doing so would increase the time required to administer the module. BLS staff also felt the Leave Module sample size would not support detailed analysis of scheduling by the number of days advance notice workers receive their schedules. As mentioned earlier, the Women’s Bureau had previously indicated that their primary interest in examining the responses to this question was for workers who said their employer decides their work schedules without employee input, which is a subset of the 5,800 workers who are expected to complete the Leave Module. Due to these factors, BLS staff were doubtful that refining the response options in this way would produce meaningful analytical value.

However, BLS staff did decide to refine the categories to separate “less than one week” from “one week.” It was thought that “one week” may be a common, modal response, and so it seemed like a logical place to divide the category. Additionally, having less than one week notice of one’s work schedule would make it challenging to plan for appointments, childcare, and other personal matters, and so it would be a meaningful analytical distinction. The additional response options were also clarified to reduce ambiguity when administering the question over the phone, as is done in the ATUS. The response options will not be read over the phone unless necessary, and so the clarification helps the interviewers correctly classify responses. The final question as it appears on the proposed Leave Module is as follows:

Advance Notice Question in proposed Leave Module

How far in advance do you usually know what days and hours you will need to work?

• Less than one week

• From 1 to 2 weeks (not including reports of 2 weeks)

• From 2 to 3 weeks (not including reports of 3 weeks)

• From 3 to 4 weeks (not including reports of 4 weeks)

• 4 weeks or more

The proposed Leave Module for the ATUS was submitted for comment in the Federal Register Notice on July 26, 2016 (81 FR 48850) and again on November 30, 2016 (81 FR 86341). One comment was received expressing support for the proposed Leave Module. BLS did not receive any comments on the proposed advance notice question, its response options, or any other item in the proposed Leave Module.

After the proposed Leave Module questionnaire was finalized and the OMB package was submitted, BLS staff proposed a change to the NLYS97 advance notice response options. The proposed changes for Round 18 of the NLYS97, to be fielded beginning in the fall of 2017, will include:

Advance Notice Question in proposed NLSY97, Round 18

How far in advance do you usually know what days and hours you will need to work?

• 3 days or less

• 4 to 7 days

• Between 1 and 2 weeks

• 3 weeks or more

• ALWAYS WORKS SAME SCHEDULE

At this time, there are no plans to field the advance notice question in the ATUS beyond the 2017 Leave Module. If the question is proposed in additional years, BLS staff will use results of the 2017 Leave Module question and results from the proposed changes to the NLYS97 question to determine if further refinement will produce meaningful and comparable results.

1 Susan J. Lambert, Peter J. Fugiel, and Julia R. Henly. “Precarious Work Schedules among Early-Career Employees in the US: A National Snapshot,” University of Chicago, 2014. https://ssascholars.uchicago.edu/sites/default/files/work-scheduling-study/files/lambert.fugiel.henly_.precarious_work_schedules.august2014_0.pdf