Collection of parental leave data resumed in the 2019 SIPP for the first time since the 2008 panel. The question sequence in 2019 differed from prior panels, and both men and women were asked the questions for the first time in 2019. In response to the working paper evaluating the reintroduced parental leave content in the 2019 SIPP, enhancements were made to the 2022 data collection process. As a result, SIPP parental leave estimates from the 2022 and later are not comparable to the 2021 and prior SIPP panels due to changes in data collection. These updates included the following:

1. Questions regarding whether the respondent stopped working prior to the birth of the first child and, if so, when that occurred are now combined into one question that asks directly about duration of time not working (in weeks) prior to the birth of the child (released variable: TPREBIRTHINT). The universe for this question is now all those who worked at any time during the pregnancy (EPREGWORK=1).

2. Questions regarding the age of the first child when the parent started working are now comprised of two questions: TPSTBRTHINT contains the age of the first child when the parent started working, and EPSTBRTH_TP contains the unit of time used for reporting that age (weeks, months, or years). The universe for both questions is all those who worked for pay after the birth of the child (EBIRTHWORK=1).

3. The universe for the question regarding whether the respondent worked for pay after the birth of the child (EBIRTHWORK) is now all those who worked at any time during the pregnancy (EPREGWORK=1).

4. The universe for the questions regarding the types of leave taken before the child was born (EPREGRSN1-10) is now all those who stopped working prior to the birth of the child (TPREBIRTHINT>0).

5. The universe for the questions regarding the types of leave taken after the child was born (EBIRTHRSN1-10) is now all those who worked during pregnancy (EPREGWORK=1).

Data users should keep these changes in mind when doing analyses of Parental Leave data and should avoid making comparisons to estimates of prior years if the universe or wording of the item has changed.