Product Rules for Mutually Independent Events

Why Multiple-Event Products Need a Strong Assumption

For two independent events, the probability that both occur is the product of their probabilities. With three or more events, the same calculation is valid only when the events satisfy a stronger condition: they must be mutually independent. Knowing that every pair is independent does not necessarily establish that the whole collection is mutually independent.

This distinction matters whenever a problem describes several events and asks for the probability that all of them occur. Multiplication may look natural, but it is justified by the probability model, not by the number of events or by a general rule that all intersections factor. In an actuarial setting, the independence assumption might represent separate processes with no relevant common source of dependence. It must be part of the stated or justified model.

Definition: Events $A_1,\ldots,A_n$ are mutually independent if, for every subset of at least two distinct events, the probability of their intersection equals the product of their individual probabilities. In particular, for every index set $I$ containing at least two indices, $$ P\left(\bigcap_{i\in I} A_i\right)=\prod_{i\in I}P(A_i). $$

The condition applies to every subcollection, not only to the intersection of all $n$ events. For three events, mutual independence requires the product rule for each of the three pairs and for the triple intersection. For two events, the definition reduces to ordinary independence.

The Product Rule for Mutually Independent Events

If $A_1,\ldots,A_n$ are mutually independent, their full intersection has probability

P(A_1\cap A_2\cap\cdots\cap A_n) =\prod_{i=1}^{n}P(A_i).

This formula calculates the probability that all the specified events occur. It does not calculate the probability that at least one occurs, or that exactly a specified number occurs; those are different events requiring different calculations.

A justification from the chain rule

The chain rule for intersections expresses a joint probability using successive conditional probabilities. When each conditioning event has positive probability, it gives

P(A_1\cap\cdots\cap A_n) =P(A_1)P(A_2\mid A_1)\cdots P(A_n\mid A_1\cap\cdots\cap A_{n-1}).

For mutually independent events, each conditional factor equals the corresponding marginal probability, provided the conditioning intersection has positive probability. For example, if $P(A_1\cap\cdots\cap A_{k-1})>0$, mutual independence gives

P(A_k\mid A_1\cap\cdots\cap A_{k-1}) =\frac{P(A_1\cap\cdots\cap A_k)} {P(A_1\cap\cdots\cap A_{k-1})} =\frac{\prod_{j=1}^{k}P(A_j)} {\prod_{j=1}^{k-1}P(A_j)} =P(A_k).

Substituting these factors into the chain rule produces the product rule. If an event in the collection has probability zero, the full intersection has probability zero because it is contained in that event; the product is also zero. Thus the product rule still holds in that case, without conditioning on a probability-zero event.

Worked Example: Three Independent Operational Alerts

Hypothetical model. A monitoring system has three alert events: $A$, a temperature alert; $B$, a pressure alert; and $C$, a vibration alert. Assume these events are mutually independent, with $P(A)=0.20$, $P(B)=0.10$, and $P(C)=0.05$. Find (1) the probability all three alerts occur and (2) the probability that $A$ and $C$ occur but $B$ does not.

Method. Apply the product rule to the specified intersection. Since the events are mutually independent, the three-event intersection factors into three marginal probabilities. For the second part, use $P(B^c)=1-P(B)$.

(1) All three alerts. The event is $A\cap B\cap C$, so

P(A\cap B\cap C) =P(A)P(B)P(C) =(0.20)(0.10)(0.05) =0.001.

To four decimal places, this is $0.0010$.

(2) Alerts $A$ and $C$, but not $B$. First calculate the complement probability:

$$ P(B^c)=1-P(B)=1-0.10=0.90. $$

Mutual independence also permits the event-complement combination to be factored. Therefore,

P(A\cap B^c\cap C) =P(A)P(B^c)P(C) =(0.20)(0.90)(0.05) =0.009.

To four decimal places, this is $0.0090$. The values are consistent with the model: the second intersection is more likely than the first because it requires $B$ not to occur instead of requiring $B$ to occur.

Common mistake avoided: For the second part, multiplying $P(A)P(B)P(C)$ would calculate the probability that all three alerts occur. The word “not” changes the event $B$ to $B^c$, whose probability is $0.90$, not $0.10$.

Mutual Independence Versus Pairwise Independence

Events are pairwise independent when every pair is independent. This requires $P(A_i\cap A_j)=P(A_i)P(A_j)$ for each distinct pair $i,j$. Mutual independence includes those pair conditions, but also requires the product rule for intersections of three or more events. Thus, mutual independence implies pairwise independence, while pairwise independence alone may not imply mutual independence.

The following model makes the distinction concrete. It uses a sample space with four equally likely outcomes, so each outcome has probability $1/4$. Equal likelihood is an explicit assumption of this hypothetical model.

Worked Example: Pairwise Independent Switch Events That Are Not Mutually Independent

Hypothetical model. Two switches are independently set to either on or off, and each of their four joint settings is equally likely. Let $A$ be the event that the first switch is on, $B$ the event that the second switch is on, and $C$ the event that both switches have the same setting. Determine whether $A,B,C$ are mutually independent.

Method. List the equally likely outcomes, calculate the marginal probabilities and pairwise intersections, and then test the triple intersection. This checks the definition rather than assuming that pairwise independence is sufficient.

The sample space is

\Omega=\{(0,0),(0,1),(1,0),(1,1)\},

where $1$ represents on and $0$ represents off. Each outcome has probability $1/4$. The events are

A=\{(1,0),(1,1)\},\qquad B=\{(0,1),(1,1)\},\qquad C=\{(0,0),(1,1)\}.

Each event contains two of the four outcomes, so

P(A)=P(B)=P(C)=\frac{2}{4}=\frac12.

Each pairwise intersection contains only $(1,1)$. Hence

P(A\cap B)=P(A\cap C)=P(B\cap C)=\frac14

and, for each pair, the product of the marginal probabilities is

\frac12\cdot\frac12=\frac14.

So the three events are pairwise independent. However, the triple intersection also contains $(1,1)$, giving

P(A\cap B\cap C)=\frac14, \qquad P(A)P(B)P(C)=\frac12\cdot\frac12\cdot\frac12=\frac18.

Since $1/4\ne 1/8$, the triple intersection does not factor. The events are not mutually independent, and the product rule for all three cannot be used.

Common mistake avoided: Checking all three pairs is enough to establish pairwise independence, but not mutual independence. The triple-intersection condition is a separate requirement.

Using Complements in Product Calculations

When mutually independent events are replaced by their complements in any chosen positions, the resulting events remain mutually independent. Consequently, a specified pattern such as “$A_1$ occurs, $A_2$ does not occur, and $A_3$ occurs” can be calculated by multiplying the corresponding probabilities. For each complemented event, use $P(A_i^c)=1-P(A_i)$.

The important distinction is between changing an event and changing the independence assumption. Replacing $A_i$ with $A_i^c$ changes the probability factor; it does not remove the need for mutual independence. This product calculation is appropriate only when the original events are mutually independent.

Worked Example: A Specified Pattern Across Four Independent Audits

Hypothetical model. Four separate audit controls are assessed. Let $A_i$ be the event that control $i$ passes. Assume the four pass events are mutually independent, with probabilities $P(A_1)=0.98$, $P(A_2)=0.95$, $P(A_3)=0.90$, and $P(A_4)=0.80$. Find the probability that all four pass, and the probability that controls 1, 2, and 4 pass while control 3 fails.

Method. Use the product rule for each specified intersection. For the second calculation, first replace the pass probability for control 3 with its failure probability.

All four pass.

\begin{aligned} P(A_1\cap A_2\cap A_3\cap A_4) &=(0.98)(0.95)(0.90)(0.80)\\ &=(0.931)(0.90)(0.80)\\ &=(0.8379)(0.80)\\ &=0.67032. \end{aligned}

Rounded to four decimal places, the probability is $0.6703$.

Controls 1, 2, and 4 pass; control 3 fails. The failure probability for control 3 is

$$ P(A_3^c)=1-0.90=0.10. $$

The required probability is

\begin{aligned} P(A_1\cap A_2\cap A_3^c\cap A_4) &=(0.98)(0.95)(0.10)(0.80)\\ &=(0.931)(0.10)(0.80)\\ &=(0.0931)(0.80)\\ &=0.07448. \end{aligned}

Rounded to four decimal places, this is $0.0745$. The first probability is larger, as expected: it requires control 3 to pass, an event with probability $0.90$, rather than to fail, an event with probability $0.10$.

Common mistake avoided: Using $0.90$ for control 3 in the second calculation would describe a pass, not a failure. The correct factor is $1-0.90=0.10$.

Exam Traps and Model Checks

The product rule is short, but a correct calculation depends on identifying the right event and verifying the right independence condition. Before multiplying, check the following:

Count the events in the intersection. A formula for two events does not by itself establish a product rule for a larger collection.
Check the stated condition. “Every pair is independent” gives pairwise independence, not necessarily mutual independence when there are three or more events.
Translate words into event notation. “Does not occur” means the complement, with probability $1-P(A)$.
Keep the requested event precise. The product of all marginal probabilities applies to the intersection of all specified events, not automatically to a union or to an event such as “exactly one.”
Do not infer independence from separate labels. Events described as different, separate, or occurring in different places are not automatically independent. The model must establish or assume independence.

A useful final check is to compare the result with the marginal probabilities. An intersection must have probability no greater than any event it contains. If a computed intersection exceeds one of its marginal probabilities, there is an arithmetic or event-translation error.

Key takeaway: For mutually independent events, every intersection factors into the product of its marginal probabilities. Mutual independence requires the product rule for every subcollection; pairwise independence checks only two-event intersections and may not justify a product for three or more events.

Check Your Understanding

For each problem, identify the specified intersection before applying a product rule. Treat an independence statement as a model assumption only where the problem explicitly gives it.

A hypothetical sensor network has three mutually independent event probabilities: $P(A)=0.40$, $P(B)=0.25$, and $P(C)=0.10$. Find the probability that all three events occur. (Answer: $0.0100$)
In a hypothetical review process, three mutually independent flags have probabilities $0.60$, $0.30$, and $0.80$. Find the probability that the first flag occurs, the second does not occur, and the third occurs. (Answer: $0.3360$)
Four mutually independent equipment checks pass with probabilities $0.90$, $0.80$, $0.75$, and $0.40$. Find the probability that all four pass. (Answer: $0.2160$)
Two fair switches have four equally likely joint settings. Let $A$ mean the first is on, $B$ mean the second is on, and $C$ mean their settings agree. State $P(A\cap B\cap C)$, compare it with $P(A)P(B)P(C)$, and decide whether the events are mutually independent. (Answer: $P(A\cap B\cap C)=0.25$, while $P(A)P(B)P(C)=0.125$; they are not mutually independent.)
Three events $A,B,C$ are mutually independent, with $P(A)=0.70$ and $P(B)=0.80$. If $P(A\cap B\cap C^c)=0.168$, find $P(C)$. (Answer: $P(C)=0.70$)

What You'll Learn