Decision Theory
What Is Decision Theory?
Decision theory is the interdisciplinary study of agent choices. It combines economics, statistics, and psychology to model how individuals make decisions, particularly under conditions of risk and uncertainty.
Suppose you have $100. You can keep it, or flip a coin. Heads you get $250, Tails you lose the $100. What should you do? Decision theory provides the math to answer this. It calculates the "Expected Value" (EV) of the gamble: (50% x $250) + (50% x -$100) = $125 - $50 = $75 gain. Since the expected gain ($75) is positive, a purely rational robot would take the bet. However, a human might not, because losing $100 hurts more than gaining $75 feels good. Decision theory explores this gap between mathematical optimality and human reality.
Key Takeaways
- Decision theory provides mathematical frameworks for making optimal choices.
- It distinguishes between decisions under certainty, risk, and uncertainty.
- Normative theory describes how people *should* decide (Rationality).
- Descriptive theory describes how people *actually* decide (Psychology).
- Expected Utility Theory is the cornerstone of rational choice models.
- Game Theory extends decision theory to interactions with other agents.
Normative vs. Descriptive
**Normative Analysis:** The ideal. It assumes "Homo Economicus"—a perfectly rational being who always maximizes utility. It uses tools like probability theory and utility functions to prescribe the best action. **Descriptive Analysis:** The reality. It studies actual behavior, incorporating biases, emotions, and cognitive limitations. This branch evolved into Behavioral Economics.
The Components of a Decision
1. **Acts:** The options available (e.g., Buy Stock A, Buy Bond B, Hold Cash). 2. **States:** The possible external conditions (e.g., Bull Market, Bear Market, Recession). 3. **Outcomes:** The result of an Act in a specific State (e.g., Buying Stock A in a Bull Market yields +20%). 4. **Payoffs:** The value (utility) assigned to each outcome.
Real-World Example: The Maximin Rule
A risk-averse investor chooses between two portfolios during uncertain times.
FAQs
It is a weighted average of the utility (satisfaction) of each possible outcome. Unlike Expected Value (money), Expected Utility accounts for "diminishing marginal returns" (the first million is worth more than the second million).
In decision theory: **Risk** is when you know the probabilities (like rolling dice). **Uncertainty** (or Ambiguity) is when you don't even know the probabilities (like predicting a war). Decision making is much harder under uncertainty.
Decision theory is usually about one person making a choice against "Nature" (randomness). Game Theory is about making a choice against "Other Players" who are also making choices that affect you (strategic interaction).
Yes. Option pricing models (Black-Scholes) are essentially applied decision theory. Traders constantly calculate Expected Value: "If I risk $1 to make $3, and I win 40% of the time, is this a good system?" (Yes: EV = 0.4*3 - 0.6*1 = +0.6).
A Nobel-winning descriptive theory by Kahneman and Tversky. It shows that people are loss-averse (pain of loss > joy of gain) and overweight small probabilities (lottery tickets). It corrected the flaws in classical Expected Utility theory.
The Bottom Line
Decision theory is the engine room of economics. It strips away the noise of daily life to reveal the mathematical skeleton of choice. Whether you are an algorithm designer programming a bot or an investor deciding on asset allocation, understanding the mechanics of probability, utility, and risk is foundational to making better choices.
Related Terms
More in Microeconomics
At a Glance
Key Takeaways
- Decision theory provides mathematical frameworks for making optimal choices.
- It distinguishes between decisions under certainty, risk, and uncertainty.
- Normative theory describes how people *should* decide (Rationality).
- Descriptive theory describes how people *actually* decide (Psychology).