Friday, July 20, 2012

Hyperbolic discounting and Nashville's growing pension problem

The city of Nashville uses discounting to decide how much to save for its future pension obligations. For a pension that pays out $100,000 in 20 years, Nashville must save $20,485=$100,000/(1.0825)^20 today, using an 8.25% discount rate.   If the city invests the $20,485, and earns 8.25%, the savings will compound and be worth $100,000 in twenty years.  If however, the investments return less than 8.25% (in fact they have done much worse),then the city will not have saved enough when the future finally gets here.  Of course, a more realistic discount rate, say 6.5%, would mean much higher current savings, 28,380=$100,000/(1.065)^20 to fund the same future pension.  But higher savings means less current spending, and spending is politically popular.

If voters were perfectly rational, they would recognize that their cities are not saving enough to fund their future pension obligations. 

That they don’t seem to care has long been recognized by psychologists, and even has a name, “hyperbolic discounting.”  It means that most people make decisions using discount rates that are too big.  In other words, they place too much weight on the present, and not enough weight on the future.  Businesses, like politicians, take advantage of this irrationality by, for example, offering a low “teaser” price which goes up in the future, or by offering a low price on a consumer durable, like a pod-coffee maker, and then charging a high price on the consumables,like the pod.  When deciding whether to purchase the pod-coffee “system,” consumers place too much weight on the “current” low price of the machine, and discount too heavily the “future” high price of the pods.  By shifting most of the system costs to the future, the coffee company makes the system appear cheaper, which increases demand.


  1. 16. I am really wondering if this is beneficial at all…

  2. What should the discount rate be then?
    How do you properly price a discount rate for government, or is this impossible (ala the socialist calculation debate?)

  3. I saw a paper that argued for a discount rate equal to the thirty year govt rate plus some risk premium, e.g., one or two percent.