Nudging doesn't give poor people retirement savings, it just makes them poorer


Nudging — the idea that a well-designed "choice architecture" can help people make free choices that are better than the ones they would make without the nudge — has a few well-publicized success stories: the cafeteria where frontloading veggies and other healthful options gets kids to choose carrots over pizza; and the employer-side deduction for retirement savings that gets employees to put aside a little more to retire on (this insight rates a Nobel-adjacent prize*!).


It's true that people suck at saving for retirement, and there are lots of reasons for that, and some of them are undoubtedly psychological.


But we live in the age of wage stagnation, where all the economic gains have gone to a tiny minority of financial engineers and their pals, who have, not coincidentally, bought up all the necessities of lives (housing, health care, education) and turned them into money-sucking tornadoes that take an ever-larger share of the ever-small paychecks most of us get to take home.

So what happens when you "nudge" people to save more for retirement? Turns out that a lot of people aren't saving enough for retirement because they're not taking home enough pay to cover their car, student, and home loans (or rent), food, and other necessities. So when you take $3,237 out of someone's paycheck per year toward their retirement, they amass $1,563 more in consumer and auto debt, and owe $4,131 more on their homes. And since investments generally return less than debts cost, you can see where this ends up. All this from The Effect of Providing Peer Information on Retirement Savings Decisions (Sci-Hub mirror), a 2015 paper in The Journal of Finance by a group of elite business-school professors).

The late stages of capitalism bear a striking and increasing resemblance to the standard stages of grief, with denial and bargaining featuring heavily, as in:

"People are going to end up either starving or being a burden on their working children or the state because they don't earn enough to save for retirement: we're going to need to bring back defined-benefits pensions, peg Social Security to inflation, and increase the minimum wage."


"No! That way lies socialism. People should just stop being poor!" (Denial)

"We tried it. They're still poor."

"Agh! These poors have no self-control!" (Anger) "We're nudge them into saving more." (Bargaining)


"Still poor, dude. Poorer, actually."

"Fuck, I guess we're voting for Donald Fucking Trump then." (Depression)


"Guys, they're building guillotines on the Capital Mall!"

"Elizabeth Warren in 2020!" (Acceptance)

* Not an actual Nobel prize


[The] study found that four years after hire, the employees who were auto-enrolled amassed an average of $3,237 more in 401(k) contributions than those who were left to sign up on their own. (That number includes both employee and employer contributions, but not market growth.)

But the auto-enrolled employees also had an average of $1,563 more in consumer and auto debt than those who were hired before auto-enrollment. When mortgage debt is factored in, the picture becomes more complicated. The auto-enrolled employees owed $4,131 more, on average, on their homes than their colleagues who were hired before auto-enrollment.

This debt more than offsets the extra $3,237 the auto-enrolled employees contributed to the plan, including the employer match.

Oh Damn, 401(k)s Aren't Magic [Hamilton Nolan/Splinter News]


(Thanks, Fipi Lele!)

The Effect of Providing Peer Information on Retirement Savings Decisions (Sci-Hub mirror) [John Beshears, James J. Choi, David Laibson, Brigitte C. Madrian and Katherine L. Milkman/The Journal of Finance]