When rational choice theory meets human psychology
You chose A and D — the most common pattern, and a violation of rational choice theory.
Look at the gambles again. In both cases, you're essentially choosing between the same core bet:
The 89% is a "common consequence" — it's the same in both options of each question. According to expected utility theory's Independence Axiom, if you prefer one bet over another, adding or removing the same "common consequence" shouldn't change your preference.
If you prefer A over B, then you should also prefer "A with a common consequence C" over "B with the same common consequence C."
But most people choose A (certain $1M) in Q1, then D (risky $5M) in Q2 — reversing their preference when the common consequence changes!
The Allais Paradox exposes the certainty effect: humans dramatically overweight outcomes that are 100% certain compared to outcomes that are merely probable. The psychological difference between 99% and 100% feels massive, even though mathematically it's just 1%.
This discovery helped inspire Prospect Theory (Kahneman & Tversky, 1979), which won the Nobel Prize and became a foundation of behavioral economics.
At a conference in Paris, French economist Maurice Allais approached Leonard Savage, the legendary statistician who had helped formalize expected utility theory. Over lunch, Allais casually presented Savage with these two questions.
Savage — "as knowledgeable about rational decision-making as anybody in the world" — considered the gambles carefully... and chose A and D.
When Allais pointed out the inconsistency, Savage was deeply disturbed. He had violated his own theory! The statistician who had mathematically proven how rational agents should behave had just demonstrated that even experts fall prey to cognitive biases.
Allais published his paradox in Econometrica in 1953, but the economics establishment largely ignored it for decades. Expected utility theory was too elegant, too mathematically beautiful to abandon.
It wasn't until 1979, when Kahneman and Tversky cited the Allais Paradox as "the best known counterexample to expected utility theory" in their landmark paper on Prospect Theory, that the economics profession began to take behavioral anomalies seriously.
Maurice Allais received the Nobel Prize in Economics in 1988 — 35 years after his paradox — "for his pioneering contributions to the theory of markets." The prize vindicated what lunch with Savage had revealed: humans are not calculating machines.