---
title: "Difference-in-Differences | Intro to Public Policy"
description: "Difference-in-differences compares outcome changes in treated and untreated groups before and after a policy, helping Intro to Public Policy students estimate impact."
canonical: "https://fiveable.me/introduction-to-public-policy/key-terms/difference-in-differences"
type: "key-term"
subject: "Intro to Public Policy"
unit: "Unit 2"
---

# Difference-in-Differences | Intro to Public Policy

## Definition

Difference-in-differences is a policy evaluation method that compares how outcomes change over time in a group affected by a policy and a group that is not. In Intro to Public Policy, it is used to estimate whether a policy caused the change.

## What It Is

Difference-in-differences is a way to estimate a policy’s effect by comparing two changes: the change in the group that got the policy and the change in a similar group that did not. In Intro to Public Policy, you use it when you want to ask not just “did outcomes go up or down?” but “did the policy likely cause that shift?”

The basic setup needs data from before and after the policy, plus a treatment group and a comparison group. For example, if one state raises the minimum wage and a nearby state does not, you can compare employment or wage outcomes in both places before and after the law changes. The idea is that the untreated group shows what might have happened without the policy.

What makes this method different from a simple before-and-after comparison is the second group. If only the treated state changes, that does not automatically mean the policy caused it, because the economy might also be shifting at the same time. Difference-in-differences subtracts out the change seen in the control group, which helps account for outside trends that affect both groups.

The method depends on a big assumption called parallel trends. That means that, without the policy, the two groups would have followed similar paths over time. You do not need them to be identical, but you do need them to move in a comparable way before the intervention. If that assumption is shaky, the estimate can be misleading.

In public policy classes, this technique shows up in policy evaluation questions where random assignment is impossible. Governments do not usually assign laws randomly across states, cities, or schools, so difference-in-differences becomes a practical way to study real policies, like education reforms, health programs, or local regulations. It is one of the main tools for turning messy real-world changes into evidence about cause and effect.

## Why It Matters

Difference-in-differences matters because policy classes are full of cases where a law or program was rolled out in the real world, not in a lab. If a city adopts paid sick leave, a state expands Medicaid, or a district changes school funding, you need a way to separate the policy effect from ordinary trends. This method gives you a cleaner causal story than just comparing outcomes before and after a change.

It also connects directly to policy evaluation and feedback. Policymakers do not just want to know whether a problem exists, they want to know whether a solution worked. Difference-in-differences is one of the main tools researchers use when they evaluate whether a policy deserves to stay, expand, or be revised.

You will also see why the comparison group matters. A weak control group can make the results look better or worse than they really are. That is why public policy students are often asked to think about whether a comparison state, city, or district is actually a fair stand-in for what would have happened without the intervention.

The method is useful beyond numbers, too. It trains you to think carefully about causation, timing, and alternative explanations, which is the core skill behind policy analysis. When you see a claim that a policy “worked,” difference-in-differences helps you ask what evidence supports that claim and what else could explain the change.

## Connections

### Causal Inference

Difference-in-differences is one tool for causal inference, which means making a careful argument that one thing actually caused another. In public policy, that is the difference between saying a policy happened and saying the policy changed outcomes. The method is useful because it gets closer to causation when random assignment is not possible.

### [Quasi-Experimental Designs](/introduction-to-public-policy/key-terms/quasi-experimental-designs)

Difference-in-differences is a quasi-experimental design because it studies a real policy change without randomly assigning people or places to treatment. That makes it more realistic for public policy evaluation, where lawmakers cannot usually run a true experiment. The tradeoff is that you have to think hard about comparison groups and assumptions.

### [Randomized Controlled Trials](/introduction-to-public-policy/key-terms/randomized-controlled-trials)

Randomized controlled trials create treatment and control groups by chance, which is stronger for causal claims when it is possible. Difference-in-differences is often the next-best option when policies are assigned by geography, time, or politics instead of randomization. Comparing the two helps you see why policy research often relies on observational methods.

### [Evidence-based policymaking](/introduction-to-public-policy/key-terms/evidence-based-policymaking)

Difference-in-differences supplies evidence that policymakers can use when deciding whether to continue or revise a program. It fits the evidence-based policymaking idea because it looks for measurable outcomes instead of relying only on intuition or political pressure. Strong policy analysis often asks whether the evidence comes from a credible comparison.

## On the AP Exam

A quiz question or short essay might give you a policy case, like a state expanding Medicaid or a city changing a housing rule, and ask how researchers could estimate the effect. Your job is to identify the treated group, the comparison group, and the before-and-after change. Then explain that difference-in-differences compares those two changes to isolate the policy’s likely impact.

If you see a chart or table, look for four cells: treated before, treated after, control before, and control after. A strong answer usually mentions the parallel trends assumption and why a poor comparison group can weaken the result. If the prompt asks for limitations, point out that outside events affecting the groups differently can bias the estimate.

## difference-in-differences vs randomized controlled trials

Difference-in-differences is often confused with randomized controlled trials, but they are not the same thing. RCTs assign treatment randomly, while diff-in-diff compares outcome changes in existing groups before and after a policy. RCTs are usually stronger for causation, but diff-in-diff is much more realistic for state and local policy changes.

## Key Takeaways

- Difference-in-differences estimates a policy’s effect by comparing changes over time in a treated group and a comparison group.
- The method works best when the untreated group gives a believable picture of what would have happened without the policy.
- Parallel trends is the big assumption behind the method, so the groups should have moved similarly before the policy change.
- Public policy classes use diff-in-diff to evaluate real-world laws and programs when random assignment is not available.
- A bad control group or an outside event that hits the groups differently can make the result misleading.

## FAQs

### What is difference-in-differences in Intro to Public Policy?

Difference-in-differences is a policy evaluation method that compares how outcomes change over time in a group exposed to a policy and a similar group that is not. It helps you estimate whether the policy caused the change instead of just happening around the same time. In public policy, it is common for state, city, or school-level reforms.

### How does difference-in-differences work?

You measure the outcome before and after the policy in both the treated and untreated groups. Then you compare the size of the change in each group and subtract one from the other. That extra comparison helps remove background trends that would have affected both groups.

### What is the parallel trends assumption?

Parallel trends means the two groups would have followed similar paths over time if the policy had never happened. You do not need the groups to be identical, but you do need their pre-policy trends to look comparable. If that assumption does not hold, the estimate can be biased.

### Is difference-in-differences the same as a randomized experiment?

No. A randomized experiment assigns treatment by chance, while difference-in-differences studies an existing policy change and compares trends across groups. Diff-in-diff is useful when randomization is not possible, but it depends more heavily on assumptions about the control group.

## Related Study Guides

- [2.5 Policy Evaluation and Feedback](/introduction-to-public-policy/unit-2/policy-evaluation-feedback/study-guide/OzBNUKK77iQjTpPU)

## About This Document

Canonical Fiveable pages are available as Markdown at the same path plus `.md`.

- [llms.txt](https://fiveable.me/llms.txt): index of Fiveable's sections and URL patterns
- [llms-full.txt](https://fiveable.me/llms-full.txt): complete subject and unit listing
- [MCP server](https://fiveable.me/mcp): call Fiveable as tools instead of fetching pages (`https://fiveable.me/api/mcp`)
- [MCP server for AP teachers](https://fiveable.me/mcp/teachers): a teacher's classes, assignments and AP-rubric grading (`https://fiveable.me/api/mcp/teacher`)

## Structured Data

```json
{"@context":"https://schema.org","@graph":[{"@type":"LearningResource","@id":"https://fiveable.me/introduction-to-public-policy/key-terms/difference-in-differences#resource","name":"Difference-in-Differences | Intro to Public Policy","url":"https://fiveable.me/introduction-to-public-policy/key-terms/difference-in-differences","learningResourceType":"Concept explainer","educationalLevel":"AP® / High School","about":{"@id":"https://fiveable.me/introduction-to-public-policy/key-terms/difference-in-differences#term"},"audience":{"@type":"EducationalAudience","educationalRole":"student"},"dateModified":"2026-07-03T02:23:30.430Z","isPartOf":{"@type":"Collection","name":"Intro to Public Policy Key Terms","url":"https://fiveable.me/introduction-to-public-policy/key-terms"},"publisher":{"@type":"Organization","name":"Fiveable","url":"https://fiveable.me"}},{"@type":"DefinedTerm","@id":"https://fiveable.me/introduction-to-public-policy/key-terms/difference-in-differences#term","name":"difference-in-differences","description":"Difference-in-differences is a policy evaluation method that compares how outcomes change over time in a group affected by a policy and a group that is not. In Intro to Public Policy, it is used to estimate whether a policy caused the change.","url":"https://fiveable.me/introduction-to-public-policy/key-terms/difference-in-differences","inDefinedTermSet":{"@type":"DefinedTermSet","name":"Intro to Public Policy Key Terms","url":"https://fiveable.me/introduction-to-public-policy/key-terms"}},{"@type":"FAQPage","mainEntity":[{"@type":"Question","name":"What is difference-in-differences in Intro to Public Policy?","acceptedAnswer":{"@type":"Answer","text":"Difference-in-differences is a policy evaluation method that compares how outcomes change over time in a group exposed to a policy and a similar group that is not. It helps you estimate whether the policy caused the change instead of just happening around the same time. In public policy, it is common for state, city, or school-level reforms."}},{"@type":"Question","name":"How does difference-in-differences work?","acceptedAnswer":{"@type":"Answer","text":"You measure the outcome before and after the policy in both the treated and untreated groups. Then you compare the size of the change in each group and subtract one from the other. That extra comparison helps remove background trends that would have affected both groups."}},{"@type":"Question","name":"What is the parallel trends assumption?","acceptedAnswer":{"@type":"Answer","text":"Parallel trends means the two groups would have followed similar paths over time if the policy had never happened. You do not need the groups to be identical, but you do need their pre-policy trends to look comparable. If that assumption does not hold, the estimate can be biased."}},{"@type":"Question","name":"Is difference-in-differences the same as a randomized experiment?","acceptedAnswer":{"@type":"Answer","text":"No. A randomized experiment assigns treatment by chance, while difference-in-differences studies an existing policy change and compares trends across groups. Diff-in-diff is useful when randomization is not possible, but it depends more heavily on assumptions about the control group."}}]},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Intro to Public Policy","item":"https://fiveable.me/introduction-to-public-policy"},{"@type":"ListItem","position":2,"name":"Key Terms","item":"https://fiveable.me/introduction-to-public-policy/key-terms"},{"@type":"ListItem","position":3,"name":"Unit 2","item":"https://fiveable.me/introduction-to-public-policy/unit-2"},{"@type":"ListItem","position":4,"name":"difference-in-differences"}]}]}
```