I’ve written multiple times about the concern that if people are prodded or even compelled to save for retirement who can ill afford it, it may have the effect of making them worse off, if rather than the proverbial forgoing of lattes at Starbucks they end up in with greater debt levels and in particular, debt with greater interest rates than the investment earnings they can reasonably expect to earn in their accounts, and especially for the lowest earners for whom Social Security (and various means-tested social assistance benefits) provides a solid standard of living compared to their pre-retirement income so that arguably high levels of retirement savings are not as important.
As long as individuals are able to decide for themselves the best way to manage their money, they may be making exactly the right decision. But this is a risk that needs more analysis as more states launch into auto-enrollment IRA programs (see my “Should Poor People Save For Retirement?” and “Three Worries About State Auto-IRA Programs”) and when individuals receive the faulty advice that retirement savings is more important than debt pay-off (no, it’s not; the mistake they make is praising the benefits of compound interest for retirement savings but neglecting its pernicious effects for debt).
And the Melbourne Mercer Global Pension Index, which I discussed in general terms last week, sheds some light on this question in an international context. This index measures 37 countries’ pensions/retirement systems by looking at a whole host of factors beyond the generosity of benefit formulas. Among these are
- The replacement rate for a range of incomes;
- The minimum pension benefits for low earners;
- The household savings rate and the degree of household debt (calculated separately then combined into a single index factor in the report);
- The amount of real funding in the system, through private and employer savings and state systems with true advance funding; and
- The homeownership rate.
And one oddity of the data is that number two ranked Denmark lands at 37th, or dead last, in the measure of savings rate/household debt (aggregated so that a better ranking means a combination of better savings and lower debt). Similarly, number three ranked Australia is 34th in this ranking, and number two ranked Netherlands is 31st.
The report itself likewise notices this and discusses it as “an interesting relationship between pension assets and household debt.” In fact, this isn’t just a fluke of the top-ranked pension system countries – the correlation is very strong, with a factor of 74.4 percent. (See here for a quick explainer on correlation.)
Why would countries with the best retirement systems simultaneously have the highest levels of debt?
Here’s the report’s explanation:
“There are likely to be several causes of this strong relationship but the well-known wealth effect is probably a major factor in many economies. That is, consumers feel more ﬁnancially secure and conﬁdent as the wealth of their homes, investment portfolios or accrued pension beneﬁts rise. In short, if your wealth increases, you are more willing to spend and/or enter into debt. . . .
“Notwithstanding these developments, the growth in assets held by pension funds is also likely to be a major contributor. That is, households feel more ﬁnancially secure in the knowledge that increasing funds have been set aside for the future thereby enabling them to borrow additional funds prior to retirement. Such an outcome is not a bad thing. The assurance of future income from existing pension fund assets enables households to improve both their current and future living standards. This situation stands in contrast to those who are relying on pay-as-you-go social security beneﬁts which can be adjusted by governments thereby reducing long term conﬁdence in the system.”
But is this something to be shrugged off?
Looking at the report’s own data on subindexes, there is not an especially strong correlation between debt/savings levels and homeownership (r = .21). In other words, this high debt level may result in homeowners owning larger homes than otherwise, but it is not increasing homeownership itself. (Again, it’s not possible to test the correlation for debt levels alone as only the consolidated index factor, with higher = less savings and more debt, is included in the final report.)
What instead does the debt/savings subindex correlate to?
It is strongly inversely correlated with the “real funding” subindex which measures public and private assets held to fund pensions, with a correlation factor of -.61. In fact, Denmark and the Netherlands tie for first on this measure. In other words, the more retirement savings in a country, the greater the household debt and the weaker the personal savings rate. This makes a certain amount of sense, and maybe it’s just fine for people to save less if they know they have generous pensions, but I’m far less comfortable with an increased willingness to actually go into debt.
It is similarly negatively correlated (-.60) with the measurement for provision of minimum benefits, but it has no particular connection to levels of replacement-rate benefits (-.04). (The lack of correlation may in part be due to the report’s method of calculating the latter index, giving “full credit” to any country with replacement ratios of between 70% to 100% of pay.) Does this mean that it’s the certainty of minimum benefits, rather than the prospect of benefits at a given level relative to pay, that contributes to high debt/low savings? And, within any individual such country, how is this debt/forgone savings distributed among the rich, the poor, and the in-between?
Or does this all mean, in the end, that there just isn’t enough known about the issue?
As always, you’re invited to share your comments at JaneTheActuary.com!