EH237 Lecture Topic 19 Difference-in-Difference 24-25 Flashcards

1
Q

What is Difference-in-Difference (DiD)?

A

A quasi-experimental design that simulates a randomized control trial to establish causality by comparing pre- and post-treatment outcomes between treated and control groups.

DiD is often used in economics to analyze the effect of a treatment or intervention.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the main goal of using Difference-in-Difference?

A

To establish causality between an independent variable and an outcome by comparing changes over time between treated and control groups.

This method aims to control for unobserved factors that may influence the outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What are the two main strategies to deal with the problem of identification in economics?

A
  • Instrumental Variables
  • Difference-in-Difference

These strategies help to ensure that the independent variable of choice is exogenous.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the role of Instrumental Variables in econometrics?

A

To clean endogenous variation from the independent variable of choice by using a predicted value from a first-stage regression.

This approach helps to isolate the causal effect of the independent variable.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are Randomized Control Trials (RCTs)?

A

The gold standard of scientific methodology that involves randomly selecting individuals for treatment and control groups to measure treatment effects.

RCTs ensure that any observed differences are due to the treatment itself.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What are Natural Experiments in the context of Difference-in-Difference?

A

Situations where individuals or administrative units are assigned to treatment by chance, allowing researchers to analyze effects without traditional RCTs.

Examples include natural disasters or policy changes.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the significance of parallel trends in Difference-in-Difference analysis?

A

It assumes that treatment and control groups behave similarly before treatment, allowing the control group to predict the counterfactual for the treatment group.

This assumption is crucial for valid causal inference.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What are the mechanics of Diff and Diff

A

Two dummy variables to capture if in treatment group and when exposed treatment (post)

Interact treatment and post to get the differential effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What does the treatment variable equal in a Difference-in-Difference model?

A

Treatment equals 1 if an individual is in the treatment group and 0 if in the control group.

This binary variable helps to differentiate between groups in the analysis.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is the post variable in a Difference-in-Difference model?

A

Post equals 1 if the time period is after the treatment came into effect and 0 if before.

It helps to identify the timing of the treatment effect.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is the effect of the 1906 San Francisco Earthquake used as an example of?

A

A natural experiment that can be analyzed using Difference-in-Difference methodology to assess impacts on affected populations.

The earthquake serves as a case study for evaluating treatment effects.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What are the critiques of Diff-in Diff

A

The 2 critical assumptions

Parallel Trends:
* Treatment and control behave similary before treatment (even if diff levels)
–> allows control to predict counterfactual of treatment

No spillovers from treatment to contrtol:
* Control group cannot be influenced by treatment

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What was the focus of Bleakley (2010) regarding malaria in Colombia?

A

To determine whether malaria exposure as a child influenced human capital formation and school enrollment rates.

The study examines both direct and indirect effects of malaria on education.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What were the direct and indirect effects of malaria eradication found by Bleakley?

A
  • Direct effect: healthier children learning better
  • Indirect effect: parents invest more in children when they are not sick

These findings illustrate the broader impacts of health interventions on economic outcomes.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What problem can arise from the use of natural experiments in research?

A

The lamppost problem, where only certain types of questions can be studied effectively using natural experiments.

This limitation may restrict the applicability of findings to broader contexts.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is the importance of panel data in Difference-in-Difference analysis?

A

Panel data allows for comparison over time across different groups, enhancing the robustness of the analysis.

It helps to control for individual heterogeneity.

17
Q

True or False: The control group in a Difference-in-Difference study can be influenced by the treatment.

A

False

The assumption is that the control group remains unaffected to accurately estimate the treatment effect.

18
Q

What is the main conclusion drawn from the study of malaria eradication in Colombia?

A

The eradication of malaria had strong positive effects on people’s income and human capital in later life.

This emphasizes the long-term benefits of health interventions.