Background: Efforts to document different types of non-parents often distinguished those who are voluntarily childless, involuntarily childless, and temporarily childless. However, an expanded approach is needed to incorporate the role of non-biological children, and to classify individuals who do not want children despite infecundity or who are undecided about wanting children.

Objective: This study operationalizes the Attitudes, Behavior, and Circumstances (ABC) framework for application to the US National Survey of Family Growth (NSFG) data, and examines demographics and trends in types of non-parents that differ in their attitudes and circumstances surrounding having children.

Methods: Using data collected in each wave of the NSFG since 2002, this study classifies non-parents into six types: childfree, biologically childless, socially childless, not-yet-parents, ambivalent, and undecided. It then presents weighted estimates of the prevalence and demographic composition of each type.

Results: The NSFG contains sufficient information to classify 99.38\% of all respondents, with not-yet-parents consistently ranked as the most common type and childfree individuals consistently ranked as the second most common type. From 2002 to 2022--2023, there is evidence of a decline in the prevalence of not-yet-parents and an increase in the prevalence of childfree, undecided, and ambivalent individuals.

Conclusion: Using the ABC framework to operationalize types of non-parents in the NSFG can complement existing approaches to classifying non-parents.

The Research and Development Survey (RANDS) is a series of cross-sectional surveys using probability-sampled commercial survey panels. The sixth round of RANDS (RANDS 6) was administered by NORC at the University of Chicago using the AmeriSpeak Panel from August 10, 2022 to August 29, 2022. RANDS 6 contained the embedded probe questions and experiments as in previous rounds of RANDS. It also explored different administrations of questions asked on the National Survey of Family Growth (NSFG) through split-sample experiments. RANDS 6 has a known survey sampling design and can be used to produce nationally and sub-nationally representative estimates.