Executive Summary
Over the past several decades, the national savings rate has plunged, even as the system for retirement preparedness has shifted increasingly towards defined contribution plans that rely on workers choosing to save in order to succeed. The end result is a generation of baby boomers that have found themselves woefully behind on retirement, notwithstanding all the "save more and spend less" advice that has been laid upon them for years.
Yet recent research suggests that perhaps the key to resolving this is to stop telling people to cut their spending now and save more, and instead to simply encourage them to save more tomorrow, instead. While this doesn't necessarily solve the challenge of the baby boomer who is already on the eve of retirement, the research suggests that for those in Generation X and Generation Y who still have years or decades until retirement, it may be a far more effective approach. The concept is rather straightforward - just commit to saving most of next year's raise, instead of cutting your spending now - yet the simple elegance is backed by a number of important behavioral finance concepts, including aversion to a loss of current lifestyle and taking advantage of our tendency towards hyperbolic discounting to make (future) saving less painful.
While there are some real world challenges to implementing a Save More Tomorrow approach - in part because planners lack some of the tools necessary to fully automate the process the way it's been done in the early research - it nonetheless raises the question of whether our "traditional" approaches to retirement advice, like "save more and spend less" or "save X% of your income every year" are due for a radical rethinking. By focusing on saving more tomorrow - and by not spending more tomorrow - perhaps we can actually find a better path to guide today's future retirees towards success.
Save More Tomorrow (SMarT) Program
The "Save More Tomorrow" program is the brainchild of behavioral finance researchers Shlomo Benartzi and Richard Thaler (the latter is the author of the book "Nudge" which has been previously reviewed on this blog). The basic idea is relatively straightforward: instead of asking people to save more now, ask them to save more in the future instead.
For instance, retirement plan participants were asked to commit to increasing their savings rate next year by 3% (assuming their annual raise would be at least a little more than 3%), and in their initial study Benartzi and Thaler found that while only 28% of employees accepted and implemented advice from an advisor to immediately increase their too-low savings amount, a whopping 78% of those who refused to increase their savings now were willing to increase their savings by 3% per year in the future. After making the commitment, only 2% of the participants dropped out after a year, and only about 20% dropped out after four years. Yet the end result was astonishing: after four years, those who accepted the recommendation to increase savings had managed to boost their rate from 4.4% to 8.8%, but those who participated in the SMarT program lifted from 3.5% (a lower base as they appear to have been less inclined towards saving in the first place) to a whopping 13.6% savings rate. In some subsequent scenarios with other firms, the researchers found that the effects were often even more dramatic for those who were not already saving at all (i.e., starting from a baseline of 0%) than they were for those who had been saving already but were trying to increase further.
Notably, these tests were done with real plan participants; thus, the SMarT program is not merely a theory, but has survived some real world environments. Although the programs were not implemented in a true experimental fashion - e.g., participants self-selected into the program and whether to save more now or later, rather than being randomly assigned - the authors take pains to share in their research why the results are probably quite robust and should sustain for many/most retirees.
Using Financial Irrationality For Good
The reason why the SMarT program seems to work so well is that it takes a lot of irrational behaviors we tend to have and frames them for good, instead of allowing them to inflict self-harm.
For instance, we all would prefer to have $1 in hand today than $1 in the future; we discount the value of things we don't receive until later (which is why we demand a rate of return for parting with our money). However, as behavioral finance research has found, we don't apply this very consistently; it turns out that we discount a lot for things we have to delay at all, and then only modestly discount further from there. For example, I might demand $2 next year to give up $1 today, but might give up that same $1 to get "just" $3 five years from now. Technically, that means the initial discount for waiting one year was far greater than the subsequent discount for later years (100% discount rate for 1 year, but only about a 10% discount rate for the subsequent 4 years); this phenomenon has been labeled as hyperbolic discounting in the research. While this behavioral "quirk" makes it difficult for us to commit to save now - because we so overweight the value of the present - it also makes it easier for us to make big commitments in the future (because we deeply discount their future value!).
Similarly, another big challenge to saving is the simple fact that we are loss averse; we don't like giving things up, and choosing to save more now inherently means we must give up - i.e., "lose" - something now from our spending to afford the saving. By contrast, when we commit to save more in the future, we never actually give up anything we currently enjoy. In fact, the SMarT research generally only committed people to saving some or most of their future raises, not all of them. Thus, over time, the individual's net income and therefore spending actually continues to rise over the years... it just rises by far less than the total amount of the raise!
The net result, as shown from some of the research above, is that after a few years of compounding the effect, spending may rise slightly but savings rises significantly. For instance, an individual who earns $50,000/year (spending all of it) and gets a raise of 4% per year - while also committing to increase savings by 3% per year or 75% of the raise - ends out after 5 years with a salary of $60,833, with spending of about $52,500 and saving of about $7,500/year. Thus, the client ends out with a whopping 12.5% savings rate, even while still spending more money every year - and following a path far easier than trying to figure out how to carve out 12.5% (or 10% or 5% or any amount of savings) to save from current income that's already being spent.
In other words, the client achieves a 12.5% savings rate by not giving up anything from his/her lifestyle... except a commitment to increase that lifestyle by less than the amount of each raise in the future! Or viewed another way, self-control around spending turns out to be a lot easier when we're redirecting future higher spending, rather than today's lifestyle spending.
Save More Tomorrow Strategies For Financial Planners
Notwithstanding the efficacy of a Save More Tomorrow approach from both the behavioral research theory, and the real world results, in practice it seems planners are often reluctant to implement plans that defer savings for the future. Typically, this seems to be built around the basic assumption that if a client is struggling to save now, it's unlikely that behavior will change in the future. Ironically, perhaps this means planners themselves are succumbing to hyperbolic discounting themselves, irrationally discounting the likelihood and relevance of saving in the future and overweighting the importance of saving immediately!
On the other hand, one unfortunate challenge of the approach from the planner's perspective is that, unlike the testing environment of Benartzi and Thaler, we cannot fully automate and "force" clients to save in the future to follow-thru on their commitment. By contrast, in the aforementioned research, the saving was truly automated once the participant committed; if/when/as the raise finally came, the savings contribution was immediately and automatically adjusted. The net result: with no further action necessary on the client's part, every raise just turned out to be more of a savings raise and only slightly a spending/paycheck raise. Given the systems available now to planners, it would be difficult to fully automate this in the same manner outside the qualified plan environment... and to the extent that implementation and follow-through requires additional steps later, there is an increased risk that clients really won't follow through when the time comes (although planners in an ongoing monitoring process could at least partially mitigate this problem by intervening quickly when raises occur to help ensure clients follow-through on their future-saving commitment).
Nonetheless, this still raises the question of whether a SMarT style approach wouldn't be a better baseline for teaching fiscal responsibility and encouraging saving. After all, another way to frame "save more tomorrow" is to simply say "don't spend [much] more tomorrow" - in other words, it's not even a commitment to cut spending and lifestyle, but simply a commitment to increase it by less than the amount of an individual's raises over time. Arguably, that's a much better way to help clients manage their spending and savings behaviors than simply insisting and repeating to spend less and save more, even as the falling savings rate for decades has made it clear that a huge number of people simply cannot implement that advice. At what point do we acknowledge that the in-the-present focused "spend less and save more" approach just isn't working very well; or alternatively, given how low the savings rate, how much is there to lose with a new and potentially better approach!?
Another benefit of a more focused approach on controlling spending over time - rather than just telling people to save more and spend less - is that it avoids the fundamental flaw of the "save-a-percentage-of-income" approach: that committing to saving 10% or 20% of your income each year also equates to permission to spend 80% to 90% of your raise every year and continuously raise lifestyle costs! Sustained over time, this leads many workers to lifestyles that require enormous sums of money to afford in retirement, and leads to a scenario where savings in the early years turn out to be woefully inadequate because spending and lifestyle rises to much later. Viewed from this perspective, the SMarT approach encourages people to save the majority of their future raises, while save-a-percentage-of-your-income encourages people to spend a majority of their future raises; compounded over a period of years or decades, this results in a drastic difference in both spending/lifestyle, savings rate, and accumulated savings, where those who merely save X% of their income lag further and further behind over time.
Unfortunately, the reality is that a save-more-tomorrow approach really only works when there are some tomorrows left for saving; in other words, this is less about solving the retirement crisis for baby boomers who are closer to retirement and have fewer tomorrows (years) available to save, and more about how to maximize the savings and success of today's Gen X and Gen Y workers. On the other hand, this also represents a significant opportunity, as most Gen X/Y workers have not yet reached their peak income/earning years, and therefore still have time to lock in their spending and lifestyle today and ensure it doesn't rise too fast in the future! So the next time you sit down with a younger client who asks about retirement advice, consider whether you're going to tell them they need to start saving more today for retirement success, or if it would be better to tell them to start saving tomorrow, instead?
In the meantime, for those who are interested in further information, see below for Shlomo Benartzi's TED talk about the "Save More Tomorrow" approach.
Very interesting. I like the approach a lot from a behavioral perspective, although I don’t think it has to completely replace the “save now” approach. Even if you can commit to saving a tiny percentage of your current income, I think it’s beneficial just to get in the habit of saving and develop the tools to do so. But for big, long-term savings increases, this approaches seems like it might have a lot of merit.
My one big hang-up with it, besides the obstacles to true commitment that you described, is that it relies on external forces to increase your saving. In other words, you’re counting on regular raises at a large enough rate to significantly increase your savings level. These raises may or may not come, they may or may not be regular, and they may or may not be at the level you expected. On the other hand, you have direct control over your current spending and savings levels. You don’t have to hope for a future raise in order to save more now. That, to me, is a big advantage of the old-fashioned approach.
I think the real lesson here is that there probably needs to be a balance between the two. At some point, a person needs to learn how to save money. Whether that point is today or next year, it has to happen. But we also have to be realistic about the psychology involved, and the approach you describe here can be a great way of using it to your advantage. Some combination of the two approaches seems likely to lead to the best results.
Michael,
Excellent article! I’ve read about this topic via your previous post a few months ago and had watched the TED talk. This is something that my wife and I actually do in regards to our own planning and as such, I’ll share below.
As we are both two years out of undergrad (and I’m in school for my master’s), a good portion of our income goes toward student loan payments, housing and other “living” expenses. While we use a budget and still pay ourselves first, we just cannot feasibly save 15 or 10% of our income just this moment. Nonetheless, we are paying down student loans quicker because being free of student loan debt is a priority for us, as a couple – regardless of whether or not an investment rate of return happens to be more for any given year.
By doing this, though, we have mutually agreed to save 50% of all salary raises (regardless of the percentage), bonuses and/or any extra income with 25% being used to pay additional principal on student loans and the other 25% to be used as “spending money”. Eventually, as our student loans disappear, we will be able to go to a 75/25 savings/spending split or some other combo between 50/50 and 75/25.
In two years we have already seen a noticeable improvement. We went from saving $2,575 or 3.99% of our income in 2012 (July ’11 – June ’12) to an estimated (but should be accurate) $3,543 or 5.33% of our income for 2013 (July ’12 – June ’13). That’s $968 more in savings in just one year while also giving us an equal amount to spend. IF we are able to maintain this, we will hit a savings rate 10.313% (assuming 3% annual salary increases) by the end of year 6 (June 2017).
I do realize, though, that the last two raises were the typical 3% but I know mine will vary going forward. So, we are really only increasing spending by 1.5% while inflation (if above 1.5% will eat away at our budget. So, as Matt said above, this requires us to continually budget and make sure that we continue to decrease expenses here and there and pay ourselves first. However, as our debt primarily lies in the form of student loans, and we pay extra, we will soon have much more of our budget to help offset any differences that inflation has caused.
As with anything, I think it requires periodic reviewing. I do think that this strategy could be of value to planners and ultimately to clients.
Wow talk about counterintuitive. I didn’t think I was strange in that I am willing to temper my current spending in order to save/invest for security in the future, but these studies have me thinking otherwise. My philosophy on personal finance is that it’s personal, so whatever gets people toward a stronger financial future is ok in my book. The save more in the future and not today strategy wouldn’t make me comfortable though, so I won’t pursue it. Do whatever works for you, but do SOMETHING.
In Australia we have a compulsory savings system (called superannuation) where the employer must contribute 9.5% of the employees annual wage as savings into the system and every employee is free to “top-up” the employer contribution to savings (salary sacrifice). I read also that in Australia the average graduate salary is $50k and that graduates typically experience wage growth of between 30% and 40% in the first 4-years following graduation. It got me thinking – we could offer a discounted engagement program (financial plan) for 1st and 2nd year workers aged under 30 on the basis that they committed to future savings based on their wage increases (in keeping with the article and findings etc). So in our proprietary software we modeled a starting salary of $50k at age 22, employer contributions at the 9.5% rate and then factored in an additional contribution of 15% of any wage increase in the years that follow. We increased wages each year by between 3-4%, topping out at age 47 (peak earnings estimate) and then also aggressively reduced income by -10% each year from age 60 to 65 (during which time the 15% top-up would naturally cease). Our system suggested the Estimated Retirement Income (the amount the saver could withdraw each year of their retirement until death) increased from $70,500 to $79,500. Now clearly – and just as Thaler et al would need to do – to make such a conclusion we need to make assumptions about asset class returns and here’s what we noticed. The above ERI conclusion assumes (over the long-term) an inflation rate of 2.5% and a cash rate of return of 5%, however if we reduced the cash rate to 2% (more realistic in Australia but still unrealistic in the U.S), the ERI plummeted to $31k. If inflation adjusts upwards to 4%, the ERI falls further to $20k. It is easy to imagine the dashed retirement dreams if one’s income falls from $80k to $20k.
(NOTE: We believe it is not prudent for planners to assume historical rates of return for traditional asset classes and instead, they should simply assume risky assets will return a “premium” above cash. The outcome described above assumes premiums will persist at the same rate they have historically, but that too may be unrealistic especially for equities (Hussman et al), in which case the ERI outcomes I have described would deteriorate even further).
My point is this (as our software clearly shows) adviser efforts to help clients adjust for their behavioral biases (and increase savings) are admirable, but stand no chance in the tsunami of wealth destruction that is coming as a result of misguided government policy to actively manipulate rates to unheard of low levels. Our system reveals it would take a truly heroic increase in savings rates to correct for the the massive transfer of wealth from savers to debt-accumulators that is occurring right now as a result of these irresponsible policies. If historical risk premiums evaporate too, may god help us all.