L​ongevity Estimates and Married Couples

By Brad Thomason, CPA 
 
Today I want to point out something that you might not have ever thought of.  I say that simply because it’s a point I seldom if ever hear anyone else speaking or writing about.  Anyway, here it is: when you are looking at planning at the household level, it is the death of the second spouse that is most significant for income purposes. 
 
Don’t read that as “the death of the first spouse doesn’t matter.”  Of course it matters.  But the point is that the period over which income is needed will be defined by the second death, not the first. 
 
I am endlessly having conversations with guys who list out for me all of the relatives they have ever had who died before reaching the age of X.  This is provided as some sort of proof that the speaker will not live longer, either (though in about half these conversations I’ve had over the years, the person telling me the tale is already older than the genetic clock has ordained…).  I listen patiently.  Then I remind him that none of that has anything to do with his wife’s prospects for a long life.  After all, presumably (though I NEVER point this out), she has different genetic lineage.  A somewhat surprised and embarrassed look usually comes into being; they typically stop talking.  A few tell me, obviously chagrined, that I’ve raised a good point.
 
Which is the point of bringing it up here.  In most cases, one spouse lives longer.  It’s not always the wife, nor does that even matter for what we’re talking about.  What matters is that the household is going to need income as long as either one of them is alive.
 
That’s the basic message.  But now that you’ve got that, let’s move on to the master class.
 
Let’s assume that there were some reliable way to say that a person had a 20% chance of making it past such and such an age.  I’m not aware of one, but let’s pretend for purposes of this little exercise, ok?
 
OK, so here in our mental laboratory there’s a 20% chance of living past, I don’t know, let’s say…92.
 
If there are 40 households, which we define as being made up of a married couple, how many households would we expect to run out of money if all of them only planned for income through age 92?
 
Did you come up with 8 households? 40, times 20%?
 
Sorry, but that’s not the right answer.
 
The correct answer is 16 households.
 
How?  Because 40 households is 80 people.  And 20% of 80 is 16.
 
The first-to-die from each household most likely lands in the 80% who don’t make it to 92.  If you want to think about it in simplified terms, if we just snap our fingers and all the husbands are gone (sorry, guys) then we are already at 50%, 40 deaths out of our original population of 80.  But notably, all of the households still need income at this point. 
 
As the statistics play out over the next few years, the number of households goes down.  But once we reach the point of 20% survivorship, there are still 16 people; and the odds are that most if not all are living in different households.
 
It is fair to mention that if we played this out many times in real-world groups of 80 we would see some that worked out this way, and other cases where both spouses in a household lived to be 95 while their neighbors across the street never made it out of their seventies.  And so forth.  But don’t let the possibility of different permutations get in the way of understanding the core principle.  There is nothing in the assumed probabilities which would be violated if the final 16 survivors were all living in different houses.  That’s the important take-away here.
 
This is but one of those situations in the retirement puzzle where the actual outcome is different than what it appears to be on first look.  It serves as a good example of why it takes time to really dig into the details, in order to do the job as well as it can be done.  Sometimes you have to give yourself time to realize that what you thought was the right answer, really can’t be.
 
Couples need to remember that they are planning for the income needs of the household, not the individuals.  While doing that, they need to further understand that any predictions or assessments they make about the odds of running out of money have the potential to actually be twice as much as they predicted, due to the mechanical aspects of one spouse likely dying before the other.