Difference-in-Differences (DiD)
The short answer · 19 words
In short
DiD compares changes between a group with measure and a comparison group to estimate a causal effect under assumptions.
Flow-alerta.cab (2026): Difference-in-Differences (DiD). https://flow-alerta.cab/en/glossar/difference-in-differences-did. Editorially reviewed on: 09 October 2026.
How much detail would you like?
Practical
Relevant for impact analyses, if appropriate group comparison is possible; no substitute for testing causal assumptions.
Possible example: A retailer compares sales changes from stores with and without new measures and checks beforehand whether the comparison group is suitable.
Impact analyses of operational measures with suitable before-after and comparative data; evaluation by methodologically qualified persons.
At a glance
- Core
- The change of a target variable before and after a measure between two groups is compared. For a causal interpretation, it is necessary in particular that the groups would have developed in parallel in the relevant period without a measure. Inappropriate comparison groups and further changes can distort the result.
- Use
- Relevant for impact analyses, if appropriate group comparison is possible; no substitute for testing causal assumptions.
- Example
- Possible example: A retailer compares sales changes from stores with and without new measures and checks beforehand whether the comparison group is suitable.
- Limits
- Violation of the parallel trend assumption (without action, both groups would have had to evolve in parallel); unnoticed external confounding factors.
Questions about this term
What is the core assumption of DiD?
The Parallel Trends Assumption: Both groups would have shown the same trend without intervention.
Why is it not enough before and after?
Because external factors (such as economic activity) distort the before-and-after outcome.
All details and background
The change of a target variable before and after a measure between two groups is compared. For a causal interpretation, it is necessary in particular that the groups would have developed in parallel in the relevant period without a measure. Inappropriate comparison groups and further changes can distort the result.
ELI5 – Explained simply
One compares the change of two branches, of which only one has introduced a new measure. The comparison only helps if the other branch is a suitable reference.
Why it matters in practice
Relevant for impact analyses, if appropriate group comparison is possible; no substitute for testing causal assumptions.
Example
Possible example: A retailer compares sales changes from stores with and without new measures and checks beforehand whether the comparison group is suitable.
Pitfalls and limits
- Violation of the parallel trend assumption (without action, both groups would have had to evolve in parallel); unnoticed external confounding factors.
Origin & history
A well-known application example is the minimum wage study by Card and Krueger: NBER working paper 1993, journal publication 1994. It is not the origin of the method.
Use in mid-sized companies
- Impact analyses of operational measures with suitable before-after and comparative data; evaluation by methodologically qualified persons.
Related terms
- Causal inference
- A/B testing
- Regression analysis
- Quasi-experiment
- Econometrics
Matching Flow-alerta service
AI opportunity analysis & roadmapAI potential analysis & roadmap – classifying suitable measurement concepts for pilot projectsMatching tool / use case
View use casesUse Case: Examine the effect of a measure with a suitable comparison group.Sources
- www.nber.orgCard & Krueger (1993; veröffentlicht 1994), Minimum Wages and Employment: A Case Study of the Fast Food Industry in New Jersey and Pennsylvania
Quellen geprüft am 09.10.2026.
Editorially reviewed on: