« How can the price of oil move so much in one day? | Main | Risk Free No Longer »
Spend More Today
In Nudge Thaler and Sunstein motivate their Save More Tomorrow plan with the following unfortunate illustration:
Consider, for example, the case of Tony Snow, the former White House press secretary, who resigned at age fifty-two in 2007 to return to the private sector. He said his motivation for leaving was financial. "I ran out of money," he told reporters...Before serving as press secretary, Snow worked a much more lucrative gig as a Fox News Channel anchor. But he arrived at the White House not having learned Retirement 101 lessons. "Snow conceded: 'As a matter of fact, I was even too dopey to get in on a 401(k).'
Sadly, Snow's choices now look optimal. Ok, I know that may be in poor taste but let's try to rescue this observation with some theorizing. Are we more likely to commit the error of saving too much or too little?
There are people who don't save much because they have very low incomes, their behavior does not seem to be in error, especially when we take into consideration the various welfare programs that will cover people in their old age.
So let's focus on people with moderate to high incomes. Thaler and Sunstein say that we are more likely to make errors when the benefits are upfront and the costs are delayed. Eating too much chocolate being a classic example. Ok, that suggests we may save too little.
T. and S. also argue that the less frequent a decision the more likely are errors. Frequency, however, cuts both ways - we only die once - so that's a wash.
Over confidence and in particular the idea that we are special and will live a long life suggests the error is saving too much. Note that we also tend to think that our partner will be alive as well. My wife once asked me whether we were saving enough for "our" retirement. "Sure," I said, "don't forget one of us will probably die before the other and I'm not saving for your future husband." "Why," she replied with a sigh, "can't economists be more human?"
Availability bias probably also suggests we save too much - we see people who saved too little in the street but the ones who saved too much are dead and gone.
In theory, optimal saving equalizes the marginal utility of income across one's lifetime - some programs like Kotlikoff's ESPlanner attempt to calculate such an eqi-marginal utility flow and Kotlikoff's finding is that a large fraction of Americans, some 40%, are saving too much. Kotlikoff's program takes into account that we may need less wealth when we are old and retired (e.g. less transportation for work related reasons) but not that the marginal utility of wealth may be lower when we are old. (e.g. Money's not so valuable if you don't need it to or can't use it to attract a mate.) Thus over-saving may be even more common than Kotlikoff suggests.
I do not know which error is more prevalent but if we are to be neither spendthrift nor miser we need to recognize both types of error.
Posted by Alex Tabarrok on July 17, 2008 at 07:20 AM in Economics | Permalink
Comments
I don't understand why you refer to dying as "a wash". This makes absolutely no sense at all. If you die during the act of washing, then that perhaps is true, but that is such a minority exception case as to throw your phrase into the realm of the ludicrous. Dying has almost nothing to do with washing. There's no correlation whatsoever. Quite what you were actually trying to communicate is a mystery.
Posted by: Ian Tindale at Jul 17, 2008 8:02:14 AM
If we tend to expect we will live longer than average, an annuity that provided a set income stream until death would be a very popular vehicle for retirement investment, with most investors overestimating how much they will get out of it. Yet aside from SS, which is hardly a choice, this is not observed.
Posted by: Cyrus at Jul 17, 2008 8:03:27 AM
Isn't the relevant question, that given most of us will have no way to calculate the optimal amount to save for retirement, would we rather save too much or too little?
Posted by: Hei Lun Chan at Jul 17, 2008 8:15:21 AM
Is Ian being sarcastic or has he never heard of the term in that context? *grin*
Posted by: JB at Jul 17, 2008 8:48:03 AM
Hei Lun Chan pinpoints the right question to ask yourself. Or, to make the question more pointed, "Would you rather die with a year's worth of expenses in your bank account, or run out of money a year before you die?"
Posted by: Brock at Jul 17, 2008 8:54:17 AM
I agree with the points above, but also consider that if you "save too much" early on, you can just choose to retire sooner. That will make your earlier saving "efficient" because you will have more years of "lost" income to replace with your savings. Conversely, if you save too little, you will be forced to retire later. Therefore, saving "too much" early on seems a wise decision- it gives you flexibility and peace of mind.
Posted by: Cliff at Jul 17, 2008 9:14:51 AM
I think few people would agree on how much is too much. Some people want to have lots of money in the bank because it makes them feel good and safe and happy, plus they can leave it to their children; others don't actually want to ever retire unless absolutely necessary, and don't mind debt at all, and aren't even scared of bankruptcy.
Furthermore, the opportunity cost of saving or spending earlier in life differs for each person. For some saving makes them feel good along the way, for others spending makes them happy. In that case, what happens during retirement may just be the cost of the earlier decision, but still worthwhile - they would make the same choice again even if they knew their time of death.
So, this idea that there is some "right amount" based on when you die is silly, as is the idea that you can measure who saved "too much" and who "too little."
Posted by: liberty at Jul 17, 2008 9:19:03 AM
If you consider the payoff matrix involved in a 2x2:
save at high rate
save at low rate
x
die young
die old
Well, with this matrix it's pretty clear saving more is the thing to do. Being old and poor is to be avoided. Even if you saved a lot and died relatively young, you may leave spouse and kids who will appreciate your thoughtfulness in providing for them. [consider that you don't buy insurance on your life for YOUR benefit, but for THEIRS]
There may be people somewhere who are spending far too little now so they can save up too much for later, but it's hard to believe they live in the U.S.
Posted by: ZBicyclist at Jul 17, 2008 10:06:00 AM
This is the problem with most behavioral law & econ. As psychology, it is very good and useful. There are plenty of phenomena (denial, resentment, etc.) that don't fit easily within a frame of rational self-interest (at least not for me). And these phenomena, in the aggregate, seem to wash out (some are in denial; some are OCD). The endowment effect is my least favorite. Is there really a difference between the endowment effect and consumer surplus?
In order to sort through these "predicitable irrationalities" we should at least sort out how much of the fluctuations in supply and demand are due to external asymmetries rather than bounded rationality.
Posted by: Michael F. Martin at Jul 17, 2008 10:51:28 AM
I applaud your willingness to criticize a guy who just died. However, I think this is an area where standard economic models are a little off. Regardless of our income, my household can decide we "need" to spend 105% of it. If I starting making $1 million per month, at first we would feel rich but within a year I bet I would be worrying at night about paying for the maintenance on the private jet.
I am not saying mortgages are bad, etc., I'm just saying that people psychologically adjust to a given level of consumption. I wish I hadn't spent so much in grad school on my credit card, after reading "rational" critiques of debt aversion.
Posted by: Bob Murphy at Jul 17, 2008 11:06:01 AM
My thinking aligns with ZBicyclist in that being old and poor is the worst option of any scenario - far worse than being young and not having that daily mochaccino - and in that any unused assets will be passed along.
Also, Americans have a negative savings rate as a whole - I don't see oversaving as a systemic problem.
Posted by: meter at Jul 17, 2008 11:12:11 AM
Wasn't Snow concerned about what he left his wife and kids? I doubt that he saw his choices as optimal.
Posted by: jim at Jul 17, 2008 11:12:15 AM
Cyrus,
We fear inflation. Annuities are great, but if we're planning on living for decades we want something with higher returns.
Posted by: jb at Jul 17, 2008 11:12:33 AM
Perhaps the stress from realizing he didn't have the savings to retiere to the life he was accustomed contributed to his health problems.
Posted by: sort_of_knowledgable at Jul 17, 2008 11:15:54 AM
The whole discussion so far ignores the value of money given to offspring. Clearly, this has a higher value for some than for others, so it confounds the whole notion of attempting to calculate an optimal savings rate - unless, of course, you can figure out a way to impute the value that people give to providing an inheritance to their children ($1 for me = .$50 for my children?).
Posted by: Ryan at Jul 17, 2008 11:15:58 AM
This whole argument is based of the idea that saving is a sacrifice and spending is pleasurable. Is it not possible that someone could enjoying saving or that spending beyond one's means could lead to an empty unhappy life? It has been floated that people with frugal dispositions (not tightwads) are actually happier on average than spendthrifts. See: http://tierneylab.blogs.nytimes.com/2007/01/15/tightwads-and-spendthrifts/
One last thought, isn't real happiness freedom? And isn't freedom in a capitalist/consumerist system money? If it is, than I would prefer to expire with freedom in the bank.
Posted by: Rick at Jul 17, 2008 11:19:18 AM
I read that passage just a few days ago and was similarly startled. Bad taste issues aside, it shows how facile the Thaler-Sunstein arguments are. I would say this is one of the problems writing a "popular" book -- except for the fact that their scholarly articles are not much more rigorous.
Posted by: Mario Rizzo at Jul 17, 2008 11:42:01 AM
Invest in family. They can help you in old age and it is fun now.
Posted by: floccina at Jul 17, 2008 11:42:04 AM
My grandmother always told us to spend our money when we were young and can still enjoy it.
I have ignored her advice and saved a lot but this savings allows me to worry less and the worry was stressful to me. Also I have never felt completely competent at any job I have had except dish washing in a restaurant, so for a long time I made sure that I could live on about $15,000/year. Now that I have a lot of income and savings I spend much more than that without worry. BTW I still think that a family of 4 with at least one smart person can live well in the Gainesville Florida area on $20,000/year. I was still working in restaurants at 30 (washing dishes, cooking and managing) and I think that gives me perspective on things. Like:
I think that a lack of money is not the problem for poor Americans, even with low income I could save money.
Working in a restaurant doing restaurant work which is manual labor can be more enjoyable for some people (me) than doing some white collar jobs, the only thing is that the pay is generally lower.
Being a computer programmer which is what I do now, while well paid can be stressful. Bugs/ people depending on your software where cooking is and certainly dish washing and even managing was less stressful to me.
People with good character can often get a nice place to live very cheaply. We once lived in an apartment above an old widow’s place where she invited us to rent it very cheap because she did not want to take a chance on renting to strangers.
Posted by: floccina at Jul 17, 2008 11:43:47 AM
You forgot to mention that Kotlikoff's software assumes you die at a set age with probability one. That removes an enormous amount of uncertainty.
Posted by: n at Jul 17, 2008 12:24:42 PM
Obviously, everybody has to make their own estimation of the current utility of money versus the future utility of money. One piece of personal relevance that isn't mentioned here is the shape of the need for money over one's lifespan as one raises a family.
Looking at my own numbers, it was clear to me that our money needs are peaking in my 30s & 40s as I raise my 4 kids and prepare to send them to the colleges of their choice. However, my earnings profile should increase monotonically over my lifespan (hopefully exceeding inflation, but you know, academia).
I came to the same conclusion as T & S and have been effectively arbitraging my income over time, taking on debt now to improve quality of life under the assumption we can live cheaply later and take care of the accumulated debt. In theory, this raises the current quality of life of my kids in particular, which we hope has long-term value for them.
Note that risk management under this strategy requires reliance on credit being available and you probably need more insurance than if you were saving (and to be very careful about your credit rating).
Also, while over-saving means you'll leave money for your family should you die with money in the bank, you missed the opportunity to let them spend/invest it in improving their own quality of life. There's an implicit assumption that it is better to invest it in your future use than your children's current use. That may be optimal for you, but it's not totally clear you're really being optimally nice to your family.
Posted by: Paul J. Reber at Jul 17, 2008 12:25:41 PM
I am not sure Thaler and Sunstein's example was unfortunate. More like prescient.
Posted by: PJ at Jul 17, 2008 12:38:31 PM
Since 52 % of marriages end in divorce , why save for "our" retirment?
Posted by: k at Jul 17, 2008 1:28:19 PM
I think we'll be able to look back at this post in 20 to 30 years and laugh when we realize many Americans were not saving enough. Remember that the national savings rate is pretty close to zero, which means that a good portion of Americans are likely saving less than 0. Many are borrowing against their retirement accounts in an attmept to keep up their lifestyles.
In this post, there's no mention of uncertainty with regards to return on investments or cost of living over time. Not to mention that risk aversion can be rational when maximizing utility. The argument in the post seems to revolve around simply looking at some known psychological biases and extrapolating that they cause us to save too much. I am left unconvinced.
Posted by: Dave at Jul 17, 2008 1:41:27 PM
hey, if you die too early, at least you won't be around to regret your error.
more seriously, if you think about the case where you save too much, you spend your life thinking you're doing fine, and then by the time your mistake becomes evident, you are too dead to know about it. as a person prone to worrying and regret, I find this a significant bonus. sure, I'd live a more miserly life, but I'd never really realize it (and we know from research that happiness is not so tied to how much we have, beyond some point).
the only problem is that now that Alex has made me aware of this issue, I'll spend the rest of my life worrying about whether I screwed up and saved too much, until I get hit by that bus. thanks, Alex.
Posted by: Mike at Jul 17, 2008 1:49:59 PM
