this is really a two-part question.
First, if you're an individual trying
to decide how to manage your finances, what's the best approach?
And, second, if you fancy yourself a
policy expert, should you promote a mandatory-savings program like Australia's,
give people options, or step back entirely?
The very short answer to the first
question is this: pay
off high-interest-rate debt first.
begs the question: when does the interest rate count as
"high"? Most of the time, that's when the rate is higher than
the investment return you can reasonably expect from your IRA or 401(k).
If you're paying 15% interest on a credit card (at the low end according to the
and you expect to earn 6% in your IRA, this is obvious: pay off the
credit card debt first. If you have a home mortgage with an interest rate
of 5%, don't be in a rush to pay it off early at the expense of retirement
savings. There is a grey area, though: if you have an employer
match on your contributions, don't think of that as "free money."
Think of it as boosting the
interest rate that you earn over time -- for example, a pure
dollar-for-dollar match might be similar to the effect of doubling the
investment return* -- so that retirement savings can come out as the
"winner" in your comparison. At the same time, if you expect to
gain from the tax advantages of an IRA or 401(k), that's worth some extra
interest in the comparison calculation.
(*This isn't literally true, and the
math is a lot more complicated but the point is that it might not always be a
sure thing to go for the match at a cost to continued interest-rate compounding
on your debt.)
This seems obvious, no? But
here's a claim from the National
Employment Law Project that caught my attention over the
weekend: in light of a paper voicing skepticism of state-run auto-IRA
plans because they "could hurt low-income participants because they will
be investing in lower-return retirement plans instead of paying off
high-interest shorter-term revolving debt, such as credit card debt," the
author, Michele Evermore, responded,
these findings do not make a
persuasive case for limiting access to savings vehicles for low-income workers.
That’s because the multiplier effects of starting to save for retirement while
young are hugely beneficial for the majority of workers.
The error Evermore makes, of course,
is that of thinking of retirement savings and debt as two entirely separate
buckets, and promoting the beneficial effect of compound interest in the former
while ignoring its pernicious effect in the latter case.
Here's an example:
Imagine you find yourself at age 25
with a desire to save for retirement but also with $10,000 in credit card
debt. You've done your budgeting and have $3,000 per year that you can
apply to paying off that debt or saving. If you apply all of that to
savings and earn 6% in investment earnings over time, then you can end up with
$17,000 five years later, and that will continue to increase over time.
But if you never make payments on your debt, and have a 15% interest rate, you
will owe $20,000, and have a -$3,000 net worth. If you pay off your debt
instead, you will have a $0 net worth, which isn't great but is in fact more than -$3,000.
What if you have an employer match
which doubles your contribution, so that a $3,000 contribution to savings is
worth $6,000? In the short-term, that boost gives you a positive net
worth very quickly, but the debt's interest rate will compound enough to outpace
the investment returns, particularly if that savings rate doesn't grow over
time. So, yes, contribute
at least enough to receive the full match, but not at the expense of paying
down your debt at least to the degree to keep the interest rate compounding at
The bottom line: compounding
investment returns are great. Compounding debt interest can be ruinous.
So what does this have to do with
auto-IRAs or mandatory savings? Experts worry that forcing people to
save, like Australia
does, or creating strong nudges that have a similar effect, will
just send more low-income workers into debt, and economics researchers have
been trying to find out what happens in the real world.
There does appear to be some good
news, in the form of a study, "Borrowing to
Save? The Impact of Automatic Enrollment on Debt," by a team of
Harvard researchers (thanks to Scott Graves for the link via twitter
@SHGraves29), which examines the outcomes of a "natural experiment"
when the U.S. Army began auto-enrolling its civilian new hires into the Thrift
Savings Program with a rate of 3%, and found that among those enrollees,
"bad debt" did not increase but car loan and home mortgage balances
did. Is this good news, because car and home loans have low interest rates
in any case? Or, to the contrary, does this indicate that new enrollees
previously had been able to fund car purchases, and make greater down payments,
out of nonretirement savings which no longer exists? -- But, on the other hand,
if the autoenrollment program had been harming people by taking away money that
was otherwise going to emergency funds, then presumably researchers would have
found higher levels of credit card debt as participants resorted to paying
repair bills with credit card debt.
It's all a muddle, and relies heavily
on Americans being able to navigate through combining retirement savings with
their other financial needs. That's one of the reasons that in at least
some countries with mandatory or auto-savings, the lowest tranche of income is
excluded entirely (see "Should
Poor People Save For Retirement?") with their income
needs in retirement taken care of with a flat anti-poverty-focused Social
Security benefit. In the United States, on the other hand, we've got a
Social Security formula that is weighted towards low income but leaves people
likely to struggle with determining how it fits in with their actual
circumstances and savings needs. In a perfect world, auto-IRA programs
such as OregonSaves and equivalent programs coming online in Illinois and
California would provide meaningful counsel to participants to help them
identify whether and how much they should save, given their income level and
other circumstances, but they fall
short, and, realistically, given the Social Security formula, it's
difficult to prescribe a simple rule of thumb.
And here's one final wrinkle to
consider: in an extreme case of a person with high levels of personal,
unsecured debt, and a high IRA account balance, that balances is protected against
creditors in the case of a bankruptcy. Had that individual chosen, over
the course of a lifetime, to pay off debt but never been able to save, she
would have no recourse to go back to creditors and say, "please give me my
money back so I can afford to retire."
for the original article.