In 1975, Ann Landers famously reported that 70% of parents had buyer’s remorse: If they had their lives to live over against, they wouldn’t have kids.  Landers’ study subsequently made it into some statistics lectures as an illustration of the danger of a self-selected sample.  In 1976, Newsday re-ran Landers’ survey on a random sample of Americans, and found that 91% of parents did not have buyer’s remorse.

When I looked into the Newsday survey, I learned two surprising facts.  First, as far as I can tell, this question was never again put to a random American sample.  (The closest thing I could find was a 2003 Gallup poll that asked non-parents over the age of 40 about non-buyer’s remorse).  Second, the Newsday survey was almost impossible to track down.  The stats profs only knew about it by word of mouth, so they didn’t have a citation.  My RA couldn’t find it online, and Newsday‘s only advice was, “Try microfiche.”  Microfiche!

The upshot was that my current and past RA went on an archaeological expedition to the Library of Congress – and saw first-hand what research was like back in the bad old days.  In the end, though, they found the original article.  It’s up on my personal webpage if you’re curious.

What does the Newsday survey say?  First and foremost, the hearsay about the rarity of regret is accurate.  In fact, since some people didn’t answer this question, fully 93% of the actual responses were positive.  Other interesting results:

  • Women had more regret than men: 9% of women had buyer’s remorse, versus just 5% of men.  While many will say this result is obvious, remember that there is virtually no gender gap on “desired family size.”
  • Young (under 25) and old (65+) had the most regret: 15% and 13% respectively.
  • Blacks had much more regret: 19%, versus 6% for whites.
  • Regret sharply falls as income rises.  13% with income under $5000 (in
    1976 dollars) had buyer’s remorse, versus only 4% with incomes of $25k+.
  • Regret sharply falls as education rises.  12% of drop-outs admitted regret, versus 3% of college grads.

Unfortunately, I still don’t have the original data, so I can’t use regression to sort out the effects of gender, age, race, income, and education.  Now that we have the date (6/13/76), it’s conceivable that Newsday will be able to give me more info, but that seems like a long-shot.

Other interesting results: The survey also asked people how many children they would have if they had a “do-over.”  If you read the table, it looks like there is a moderate tendency to want more: Respondents have 2.66 but want 2.84.  You might think that this result happens solely because the survey includes young parents who have not yet reached their target; but it also holds for respondents 45 and older.

If you read the accompanying text, however, you learn that these results are only for the sub-sample of “people who had children and also said they would like to have children again.”  If you factor in zeroes for all the parents with buyer’s remorse, there might even be a slight tendency for parents to want fewer kids than they’ve got.  Keep in mind, though, that almost all parents clearly want all the kids they’ve got; the results are driven by a small minority of the disappointed.

OK, so what’s the take-away? 

First of all, even though child-free advocates continue to cite the famous Ann Landers survey, it was discredited over thirty years ago.  Almost no one regrets having kids. 

Second, you might dismiss the Newsday results as mere status quo bias – “Everyone thinks that whatever they did was for the best.”  But you probably shouldn’t.  The 2003 Gallup study
finds that about two-thirds of childless people over 40 wish they had
kids.  Buyer’s remorse is rare; non-buyer’s remorse is common.

Third, almost everyone wants as many kids as they’ve got.  You can object that it’s partly an endowment effect, but it’s still real.

Fourth, the small amount of regret that exists is heavily concentrated in the left tail of the SES distribution – once again confirming the hypothesis Scott Beaulier and I advance that “the poor deviate more” from neoclassical rationality and self-control assumptions.