Home Statistical Dictionary Birth-Death Process

Birth-Death Process

A continuous-time Markov chain on the non-negative integers where every transition moves exactly one step up (a 'birth') or one step down (a 'death'), never skipping states.

Stochastic ProcessesIntermediateFree

In Plain English

A birth-death process tracks a population, queue length, or inventory count that changes by exactly one unit at a time, up when something is 'born' (a new arrival, a new individual), down when something 'dies' (a departure, a death). Because jumps are restricted to nearest neighbors, the process's entire behavior is described by just two sequences of rates, birth rates $λ_{i}$ and death rates $μ_{i}$ , one pair for each state $i$ , making it one of the most tractable and widely used building blocks in queueing theory and population dynamics.

Definition

A birth-death process is a continuous-time Markov chain on state space ${0, 1, 2, \dots}$ in which, from state $i$ , the only possible transitions are to $i + 1$ (a birth, at rate $λ_{i}$ ) or to $i - 1$ (a death, at rate $μ_{i}$ , with $μ_{0} = 0$ since the state cannot go negative); all other transition rates are zero. The generator matrix is therefore tridiagonal. When a stationary distribution exists, it has the explicit product form $π_{i} = π_{0} \prod_{k = 0}^{i - 1} \frac{λ _{k}}{μ _{k + 1}}$ , obtained directly from the detailed balance condition, which always holds for birth-death processes because probability flow between any two non-adjacent states is impossible. Special cases include the Poisson process (a pure birth process with $μ_{i} = 0$ and constant $λ_{i} = λ$ ) and the M/M/1 queue (constant arrival rate $λ$ , constant service rate $μ$ ).

Formula

π_{i} = π_{0} k = 0 \prod i - 1 \frac{λ _{k}}{μ _{k + 1}}

Notation

λ_{i}

—Birth rate: rate of transitioning from state $i$ to state $i+1$

μ_{i}

—Death rate: rate of transitioning from state $i$ to state $i-1$

Properties

Because transitions are restricted to nearest neighbors, detailed balance holds automatically for any birth-death process with a stationary distribution, this is what gives the explicit, closed-form product formula for $π_{i}$ without needing to solve a general system of equations.
Queueing theory builds almost its entire elementary toolkit (M/M/1, M/M/c, M/M/1/K and further variants) as birth-death processes with different choices of $λ_{i}$ and $μ_{i}$ , e.g. constant arrival rate but state-dependent (multi-server) service rate.
Population ecology and epidemiology also lean on birth-death processes directly: simple population growth models (with birth and death rates possibly depending on current population size, capturing carrying-capacity effects) are birth-death processes by construction.

At a Glance

State spacenon-negative integers

{0, 1, 2, \dots}

Allowed transitionsonly to nearest neighbors:

i \to i + 1

(birth) or

i \to i - 1

(death)

Always reversibledetailed balance holds automatically (no flow between non-adjacent states)

Classic examplesPoisson process (pure birth), M/M/1 queue (constant birth & death rates)

Last updated September 8, 2026← Back to the Dictionary