The Institute for Family Studies surveyed 1,518 US adults in 2021 and found a clear asymmetry in fertility regret. 37% of all US adults wish they had more children than they currently have, including both parents who want additional children (24%) and childless adults who want to start (13%). On the other side, only 14% of parents wish they had fewer children. The desire for more consistently outweighs the desire for fewer, roughly two-to-one.
This asymmetry aligns with Gilovich and Medvec’s prediction that inaction regrets dominate over time. The gap between desired and actual fertility has been a persistent structural feature of US demography for decades, and IFS trend data shows it is widening: the share of Americans who say the ideal family size is three or more children recently hit a 50-year high. People are not adjusting their preferences downward to match their behavior — they continue to wish for more even as birth rates fall.
The comparison is imperfect in several ways. “Wishing for more” is not the same as regretting a specific past decision — it may reflect wistfulness about circumstances (finances, partner, timing) rather than decisional regret. The populations are also not cleanly symmetric: the 14% is measured among parents, while the 37% includes childless adults. And the largest confound is involuntariness: many people who want more children face fertility limitations, financial barriers, or partner disagreement, making the “inaction” label somewhat misleading. They did not choose fewer children — they ended up with fewer.
A complementary signal from the children themselves: a YouGov poll of 2,000 US adults (October 2023) found that 66% of adults who grew up as only children wish they had had at least one sibling, against 34% who were satisfied with the arrangement. The child’s sibling-wish rate (66%) runs nearly double the parent’s wish-for-more rate (37%), a gap that hints at a systematic blind spot: parents may underestimate how much their child will want a sibling, while the child feels the absence as an adult. These are different populations answering different questions, not directly comparable rates, but the directional divergence is worth registering.