Skip to main content

Types of Data

Studies presented in the Data Hub use two primary types of quantitative data on the non-traditional workforce: self-reported and administrative data. Self-reported data compiles survey or interview responses. Administrative data consists of records originally recorded for non-research purposes. Either type of data may be collected by public agencies, such as the Bureau of Labor Statistics, non-profit organizations, academic researchers, or private companies. In addition, there are some qualitative examinations of non-traditional work, drawing on interviews and observations to inform understandings.

Each type of data has benefits and shortcomings, and is suited to particular types of questions. Administrative data tends to be thorough in its sampling, but limited in the scope of information it can provide. Self-reported data tends to be more targeted in the information it provides, but requires careful sampling to ensure that it is generalizable, and relies on respondents’ interpretation of questions, which may or may not match researchers’ interpretations.

  Self-reported Administrative
  • Tax forms issued (1099s versus W2s issued)
  • Taxes filed (Schedules C and SE)
  • Nonemployer statistics
  • Bank account records (JPMorgan Chase Institute)
  • Private administrative data (like records of Uber drivers analyzed by Hall and Krueger)

The Data Hub brings together all of these types of data to establish as comprehensive an understanding of the non-traditional workforce as possible. Some of the research presented contains original data collected by authors, while other research offers analysis of publicly available data sets. For more information on what research is included and prioritized in the Data Hub, see our Criteria for Study Inclusion.