GlossaryStaffing, accountability, and public reporting
Outcome measure
An outcome measure tracks a result or change that a program or organization seeks to influence, such as improved access, reduced harm, or changes in people's experience.
An outcome measure tracks a result or change that a program or organization seeks to influence, such as improved access, reduced harm, or changes in people's experience. It differs from simply counting work performed. An observed outcome change does not automatically prove that the program caused it.
Connect the measure to the goal
The Government Accountability Office's performance guidance distinguishes measuring results from evaluating why results occurred. A clear outcome measure identifies the desired condition, who or what is measured, the period, and the calculation. It should also identify data limitations.
Outcomes can be influenced by factors outside an agency's control. Reporting practices, community conditions, population changes, and other services may affect the measure. These influences matter when interpreting improvement or assigning responsibility.
Example
Hypothetical example: A referral program counts completed referrals as an output and tracks whether eligible participants obtain the intended service within 30 days as an outcome. During one period, referral counts rise while service access remains unchanged.
The program delivered more referral activity without a measured improvement in the stated access result. That does not explain why access remained unchanged; an evaluation would need to examine capacity, eligibility, follow-through, and other factors.
Common confusion
An output describes services delivered. An outcome describes a result of interest. The same program can need both measures to explain what it did and what changed.
A causal impact estimate asks what difference the program made compared with an appropriate alternative. A simple before-and-after outcome trend does not automatically supply that estimate.
Reading the evidence
Check the baseline, eligible population, follow-up period, missing observations, and comparison method. Define success before interpreting the numbers. A broad label such as “community safety” needs specific measures; one statistic cannot represent every resident's experience or every dimension of an agency's work.