TL;DR
Standard economic theory suggests everyone should love annuities, yet almost no one buys them—a phenomenon known as the "Annuity Puzzle." This paper demonstrates that this reluctance isn't necessarily a rational choice, but a result of how the product is framed: when viewed as a "consumption" tool for life-long spending, annuities are highly popular, but when framed as an "investment" with a risky return, they are rejected.
What Problem Does This Paper Address?
In neoclassical economics, a risk-averse individual should ideally convert a significant portion of their wealth into a life annuity to hedge against the risk of outliving their savings. However, in the real world, voluntary annuitization is rare.
Economists have traditionally tried to explain this via "revealed preference"—assuming consumers are optimizing perfectly and simply find annuities unattractive due to high fees, inflation risk, or a desire to leave bequests. This paper challenges that assumption, testing the hypothesis of imperfect optimization. The authors explore whether the "Annuity Puzzle" is actually a "Framing Puzzle," where the psychological context of the decision (investment vs. consumption) and specific reference points (like the purchase price or habitual spending) dictate consumer choice more than the underlying financial math.
Methods
The researchers conducted two large-scale internet surveys using the Zoomerang panel, targeting individuals over age 50. They employed a "forced-choice" experimental design where respondents compared the financial decisions of two fictitious people and decided who made the better choice.
The core of the methodology was the random assignment of participants to different "arms" or frames:
- Consumption Frame: Used words like "spend" and "payment," focusing on how much money is available each month until a certain age or death.
- Investment Frame: Used words like "invest," "earnings," and "return," focusing on account balances, interest rates, and the value of the estate upon death.
The authors then introduced "reference points" to test for loss aversion:
- Modified Consumption Frame: Included the $100,000 purchase price to see if "losing" that lump sum made the annuity less attractive.
- Habitual Spending Levels: Introduced a baseline monthly spending target ($1,200, $1,500, or $1,800) to see if the fear of falling below a "habit" influenced the choice.
- Principal Protection: In the investment frame, they tested annuities that guaranteed the return of the initial investment (e.g., $100k) even if the person died early.
Datasets
The study utilized two primary datasets collected via online surveys:
- Survey 1 (December 2007): 673 participants.
- Survey 2 (April 2008): 3,382 participants (approx. 483 per experimental arm).
- Demographics: The sample was 43% female, 57% male, with 54% over age 60. The authors noted that while randomization was slightly unbalanced regarding age and gender, they corrected for this using multivariate regression analysis.
Key Insights and Ideas
- Framing is the Dominant Factor: The shift from an investment mindset to a consumption mindset changes the annuity from a "risky bet on your life" to "insurance against poverty."
- Robustness of the Consumption Frame: Even when the authors explicitly mentioned the $100,000 purchase price (a potential "loss" reference point), the preference for annuities remained high. This suggests that the "investment" language itself, rather than just the mention of the cost, is what drives people away.
- The Principal Protection Paradox: In an investment frame, people hate annuities because they fear "losing" their principal if they die early. Adding a guarantee to return the principal makes the annuity twice as attractive, even though this guarantee actually reduces the monthly payout (the insurance value).
- Lack of Sharp Inflection Points: Contrary to strict Prospect Theory, there wasn't a "magic" reference point. People didn't care significantly more about protecting exactly $100,000 versus $80,000 or $110,000. They simply wanted some protection against a total loss of principal.
Key Results
- The Framing Gap: In the original Investment Frame, only 21% of respondents preferred a life annuity over a traditional savings account. In the Consumption Frame, that number jumped to 72%.
- Price Transparency: Adding the $100,000 purchase price to the consumption frame only slightly lowered the preference to 68%, a statistically insignificant change compared to the original consumption frame.
- Principal Protection Impact: In the investment frame, a standard annuity had a 21% preference rate. Adding a $100,000 principal guarantee increased the preference to 47%.
- Spending Habits: Setting a "habitual" spending level of $1,200 (which the annuity could cover) pushed preference to **80%**, but higher spending targets that the annuity couldn't meet ($1,800) saw preference drop back to 72%.
Limitations and Future Work
- Stated vs. Revealed Preference: The study measures what people say they would do in a survey, not what they do with their actual life savings. While the authors argue survey results are a valid proxy for psychological bias, real-world stakes might change behavior.
- Complexity: The study used "actuarially fair" simplified products. Real-world annuities are often complex, with varying fees and inflation riders that might confuse consumers regardless of framing.
- Policy Implications: The authors suggest that 401(k) plans should report balances as "monthly income projections" rather than "lump sums." Future research could investigate whether changing these regulatory disclosure requirements actually increases annuitization rates in real-world retirement plans.
Summary updated