What makes a process a birth-death process?

It is a continuous-time Markov chain on the integers whose transitions move only to the nearest neighbours, increasing or decreasing the count by one according to birth and death rates.

Why are birth-death processes always reversible?

Because the state space is linear and transitions are only between adjacent states, the flow between any two neighbours balances in equilibrium, so the detailed balance equations hold and give the stationary distribution directly.

Birth-Death Processes

A birth-death process is a continuous-time Markov chain on the integers that increases or decreases by one at a time, with birth and death rates that depend on the current population size.

Definition

A birth-death process is a continuous-time Markov chain on the non-negative integers whose only transitions are upward by one at a state-dependent birth rate and downward by one at a state-dependent death rate, so its sample paths change by unit steps.

Scope

This topic covers the structure of nearest-neighbour transition rates, the special cases of pure birth and pure death processes, transient solutions and extinction probabilities, the stationary distribution obtained from detailed balance, and applications to populations and queues including the M/M/1 and M/M/c systems.

Core questions

How do state-dependent birth and death rates determine the dynamics?
When does a birth-death process have a stationary distribution, and what form does it take?
How are extinction and explosion probabilities computed?
How do queueing systems arise as birth-death processes?

Key theories

Detailed balance and the stationary distribution: Because transitions are to neighbouring states, a birth-death process is reversible and its stationary distribution is obtained explicitly from the detailed balance equations as a product of successive birth-to-death rate ratios.
Extinction and absorption analysis: First-step and generating-function arguments give extinction probabilities and expected times to absorption when zero is an absorbing state, characterising whether a population dies out and how quickly.

Clinical relevance

Birth-death processes model biological populations, the spread and clearance of infections, the number of customers in a queue, and the occupancy of communication channels; the M/M/1 queue, a birth-death process with constant arrival and service rates, is the canonical example linking this topic to queueing theory.

History

The pure birth process was introduced by Yule in 1925 to model the growth of biological genera, Feller analysed general birth-death processes in the 1930s and 1940s, and the framework became central to queueing theory through the work of Erlang and his successors on telephone traffic.

Key figures

William Feller
George Udny Yule
Alfred Lotka

Seminal works

karlinTaylor1975

Frequently asked questions

What makes a process a birth-death process?: It is a continuous-time Markov chain on the integers whose transitions move only to the nearest neighbours, increasing or decreasing the count by one according to birth and death rates.
Why are birth-death processes always reversible?: Because the state space is linear and transitions are only between adjacent states, the flow between any two neighbours balances in equilibrium, so the detailed balance equations hold and give the stationary distribution directly.