We use Meta (encompassing Facebook and Instagram) for our survey recruitment due to its extensive user base and comprehensive reach, making it the most expansive nonprobability web-based platform available. Meta also provides robust identity verification for respondents. Furthermore, Meta’s detailed audience targeting capabilities allows us to reach difficult-to-sample populations, such as employees from specific companies. The platform’s ability to gather employer–employee linked data is particularly valuable for studies on labor practices and the impacts of labor laws, making Meta an effective tool for survey recruitment.
Documentation
Documentation
Using a novel method of online sampling and recruitment, we field surveys to workers at over 150 of the largest companies in retail, food service, grocery, delivery and fulfillment, and hospitality. We provide unique measurement of the reality of working conditions, including work scheduling, access to paid leave, workplace surveillance, and labor standards violations. We pair this in-depth measurement of job quality with detailed measures of financial security and of the health and wellbeing of workers and their families.
Methodology
We recruit participants through targeted Meta ads on Facebook and Instagram, focusing on employees of major U.S. companies. Those who click on the ad are taken to a survey landing page and invited to consent to participate and complete the survey via Qualtrics. As an incentive, participants who complete the national survey and provide contact details are entered into a $500 gift card lottery.
In addition to cross-sectional data, we collect panel and oversample data. Panel data collection occurs through reinterview surveys, where we recontact cross-sectional participants via email or text for follow-ups, allowing us to track changes in individual worker’s conditions and wellbeing over time. Reinterview respondents are offered a gift card incentive to participate. For oversample data, we use targeted Meta ads to reach subgroups of service sector workers in specific cities or states. This approach ensures larger sample sizes within these groups, allowing us to assess local labor policy changes. For detailed information on The Shift Project’s data collection, methodology, and data validation, refer to Schneider and Harknett (2022).
Data Structure
The chart below tracks
National , Panel and Oversample data across waves.
Wave 1 (Fall 2016)
Wave 2 (Spring 2017)
Wave 3 (Fall 2017)
Wave 4(Spring 2018)
Wave 5(Fall 2018)
Wave 6 (Spring2019)
Wave 7 (Fall 2019)
Wave 8 (Spring 2020)
Wave 8.5 (Summer 2020)
Wave 9 (Fall 2020)
Wave 10 (Spring 2021)
Wave 11 (Fall 2021)
Wave 12 (Spring 2022)
Wave 13 (Fall 2022)
Wave 14 (Spring 2023)
Wave 15 (Fall 2023)
Wave 16 (Spring 2024)
Wave 17 (Fall 2024)
National Sample
Oversample
SEA
NYC / OR
SEA
SEA / NYC / OR
WA / OR
WA
NJ
Panel 1
Baseline
1st Follow-up
2nd Follow-up
Panel 2
Baseline
1st Follow-up
Panel 3
Baseline
Baseline
1st Follow-up
1st Follow-up
Panel 4
Baseline
Baseline
Baseline
1st Follow-up
2nd Follow-up
3rd Follow-up
Panel 5
Baseline
Baseline
1st Follow-up
2nd Follow-up
3rd Follow-up
Panel 6
Baseline
Baseline
1st Follow-up
2nd Follow-up
3rd Follow-up
Panel 7
Baseline
Baseline
1st Follow-up
Panel 8
Baseline
Content Map
The charts below track survey content across different waves of national data collection and outline the survey flow as it has changed over time. Hover over any survey module to see its place in the survey flow and its presence across different waves.
Average Survey Flow
Survey Modules Across Waves
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Hours & Work Scheduling
Work-Life Conflict
Demographics
Basic Job Characteristics
Household Finances & Hardships
Fringe Benefits
Parenting
Child Wellbeing
Health & Wellbeing
Place of Work
Employment Status
Paid Sick Leave
Automation & Surveillance
Unemployment Insurance
Unemployment
Job Insecurity
Social Support
Retirement
Workplace Demographics
COVID-19
Knowledge Of Law
Bullying/Respect at Work
Work Breaks
Paid Family Medical Leave
Labor Violations
Managerial Practices
Union Membership
Worker Voice
Core Data Documentation
The interactive codebook details the modules and variables included in The Shift Project’s surveys, with question wording, labels, display logic, and survey wave frequency for each variable.
Explore our comprehensive survey questionnaires designed for national and panel data collection. Each instrument offers valuable insights into the questions used and the breadth of our research.
Access our de-identified survey data through Dataverse, featuring the main cross-sectional dataset of hourly service sector workers from The Shift Project’s surveys.
More Data Resources
Qualitative Interviews
This study involved in-depth qualitative interviews with 102 retail and food service workers, yielding over 130 hours of detailed conversations about their daily work experiences. Participants were chosen from previous Shift Project survey takers. Findings from these interviews have informed survey content and led to academic presentations and publications.
Labor Policy Database
This labor policy database includes policies affecting labor conditions and requirements across U.S. states and counties from 2013 to 2022. It provides valuable insights into the evolving policy landscape and serves as a resource for labor and policy researchers interested in how these environments influence working conditions.
FAQ
-
The Shift Project sample is constructed by delivering targeted advertisements to Facebook/Instagram users who are employed at large companies. These companies are drawn from the National Retail Federation’s list of top 100 retailers and the Restaurant Business’ list of top 100 restaurant chains. While not all companies can be targeted on Meta platforms, we successfully sampled 44 of the top 50 retailers and 48 of the top 50 restaurants. The Shift data also sample respondents at 135 other large companies that are a purposive mix of retailers in the 50-100 rank, restaurants in the 50-100 rank, as well as large companies in hospitality and logistics that are not represented on the NRF or Restaurant Business lists. A subset of these companies are surveyed at every wave, with other companies rotating in and out of sample purposively depending on the wave.
-
-
We make de-identified survey data publicly available via Dataverse. This dataset includes the main cross-sectional data from The Shift Project’s surveys of hourly service sector workers.
-
The Shift survey recruitment approach yields a non-probability sample of workers, which may differ from the broader population of service-sector workers. To mitigate potential bias, we have applied survey weights that adjust our sample to reflect the universe of service-sector workers in the U.S. These weights are constructed in two stages.
-
To a greater extent than probability sample surveys conducted online or in other modes, online surveys using non-probability sampling methods face threats to data quality. We take several steps to both guard against the collection of low-quality or ineligible responses and identify and exclude them after each round of data collection closes.