The Current Population Survey gives detailed information at the individual level on labor market participation, either in employment or nonemployment, and the weekly hours worked by detailed reason. The definition of full-time and part­time employed by reason given by the Bureau of Labor Statistics (BLS) follows from questions asked of employed workers about working hours and availability. The BLS definition and terminology is “At work” by full- or part-time and reason. A worker is referred to in this paper as “Full-time” if they are classified by the BLS as “At work 35 or more hours.”[1] Workers who report working a total of 1-34 hours at all jobs during the reference week are classified as “At work part-time.” Part-time workers are also further classified by reason for working part-time. Workers who are classified as “At work part-time for economic reasons” (P T E ) are individuals working part-time hours who report wanting and being available for full-time work and provide the reason for being at work part-time as either “Slack Work/Business Conditions,” “Failure to Find Full-time Work,” or “Seasonal Work/Changes in Demand.” Those who report being “At work part-time for non-economic reasons” (P T N ) are individuals “At work part-time” who report reasons for working part-time that are not related to employer demand, or report that they are unwilling or unable to work full-time. In this paper, I refer to the BLS classifications of “At work full/part-time/part-time by reason” simply as “Part-time employed (for econ/non-econ. reasons)” and “Full-time employed.”

This is different from the typical definition of “employment” in that absent workers are not counted in this population. This definition relies only on actual hours worked, and does not necessarily coincide with being in the “usual” full-time labor force, which consists of those in the labor force who report usually working 35 or more hours per week (regardless of current employment or actual hours worked in the survey week). Similarly, the part-time labor force consists of employed and unemployed individuals who usually work less than 35 hours per week.

The sample time-frame considered for the gross flows analysis of Part-time employ­ment is 1994-2014, after the CPS redesign. The sample time-frame initially studied for the gross flows analysis was chosen due to the fact that there are substantial changes in the 1994 redesign affecting the classification of workers into full-time and part-time employed, especially by reason. Polivka and Miller (1998) document that prior to the redesign, the CPS survey structure only asked those who reported working less than 35 hours what was their usual hours worked per week. This leads to an underestimation of the size of the part-time labor force by 9.8% relative to the redesign. The 1994 redesign also corrected an over-estimate of the part-time employed for economic reasons. The reported number of of part-time employed for economic reasons is over-stated by 20% relative to the unrevised CPS. This difference is primarily due to the unrevised survey assuming desire for and avail­ability to work full-time based on the reported reason given for being part-time employed, rather than asking this question directly. For details on the CPS variables and responses used to classify workers, see Appendix A.3.1