Questions tagged [difference-in-difference]

It is about the difference-in-difference method which is quite popular for quasi-experimental studies

Filter by
Sorted by
Tagged with
0 votes
1 answer
27 views

Difference in Difference - how to choose the control group (difference with no clear trend, parallel series or cointegrated?)

I'd like to use difference in difference methodology to test the impact of a treatment. Being the formula of the regression: ...
GabyLP's user avatar
  • 101
0 votes
0 answers
22 views

Assessing multiple Treatment Effects in Overlapping Interventions for difference-in-difference

I am currently working on evaluating the impact of two sequential treatments, Treatment 1 and Treatment 2, on two distinct groups in our country: Group A and Group B. Treatment 1 is administered at ...
Marble's user avatar
  • 111
0 votes
1 answer
32 views

Pooled difference-in-differences with limited panel

I wish to estimate a pooled difference-in-differences specification in the following form: $y_{i,t}=\alpha_i+\theta_t + \beta \times \mathbf{1}\left(t > \bar{t}\right) \times Treatment_i + \...
Serb's user avatar
  • 13
2 votes
0 answers
32 views

Individual level covariates used in this difference-in-difference paper: a mistake?

This paper uses a differece-in-differences (DID) approach to assess the effect of adding quizzes and assignments on exam grades. The data come from 3 different sections that one of the authors taught, ...
sams's user avatar
  • 21
0 votes
0 answers
10 views

DiD when pre-start and post-start period durations differ by individual?

Can a generalization of Difference in Differences (DiD) be done when the durations of pre-start and post-start periods vary by individual? I believe that the mean pre-start-duration and mean post-...
jbuddy_13's user avatar
  • 103
1 vote
0 answers
30 views

How to add interaction term into staggered DID (differentiate the effect between large company and small company)

I have a staggered DID model as follows: Here I is an indicator of whether the distance of the calendar year from the event year equals to q ( where q takes -3 to +3 and is my time window ). \delta ...
Eileen's user avatar
  • 11
0 votes
0 answers
26 views

Event study with individual fixed effects

Suppose we have a treatment that varies at the group level before and after some t, where g is rather small with a couple of categories. Would we estimate an event study with group or individual fixed ...
Papayapap's user avatar
  • 1,781
2 votes
0 answers
122 views

How do I set up a staggered difference-in-differences design in R?

I have data which looks like this: ...
econdid's user avatar
  • 21
0 votes
0 answers
28 views

How to analyze the effect of a policy change?

I am currently analyzing the impact of a public policy change, specifically one that was introduced in 2018 but later changed in 2020, changed the eligible rules. While I have used DiD and Event-study ...
Claudia_ECON's user avatar
2 votes
2 answers
159 views

How to correct selection bias in an econometric analysis?

We know that selection bias occurs when the treatment and control groups are not comparable, leading to differences in the outcome that are not solely due to the treatment. How does one address the ...
funcard's user avatar
  • 43
0 votes
0 answers
49 views

How to estimate staggered difference in difference with continuous treatment?

I want to estimate a staggered difference in difference with continuous treatment. The data looks something like this: $$ \begin{array}{ccc} Individual & year & CT_{i,t} \\ \hline 1 & ...
TeTs's user avatar
  • 181
1 vote
1 answer
38 views

Is matching combined with Diff-in-Diff a bad idea?

I remember seeing such a comment on Twitter long time ago but I see so many prominent papers doing it that I doubt it can really bad. Does somebody know more (e.g. papers) on this?
Papayapap's user avatar
  • 1,781
2 votes
1 answer
57 views

Is Difference-in-Differences still valid when treat or control is determined by subjects?

I have a question about the difference-in-differences method. I am considering using it to evaluate a policy, the UK's Help-to-Buy (H2B) equity loan scheme. The housebuilders can choose to join the ...
Yunlong Huang's user avatar
2 votes
1 answer
49 views

Diff in Diff: Why do we control for unit and time fixed effects instead of just treatment status and pre/post indicator?

In a multi-period DiD analysis, the usual specification is: $$y_{it} = \alpha_i + \theta_t + \beta D_{it} + \epsilon_{it} $$ where $D_{it}$ is the interaction between ...
Rainroad's user avatar
  • 243
0 votes
0 answers
12 views

Argument the coefficient of diff-in-diff after matching

I am trying to use the propensity matching score to match the control and treatment groups based on their characteristics (can be done by gmatch, psmatch2, tseffect match...) However, after matching, ...
Phil Nguyen's user avatar
  • 1,090
2 votes
1 answer
63 views

Which covariates can I include in my fixed-effects regression?

I am doing a difference-in-difference analysis of an event that affected several states in the US. I am interested to understand the effects of this event on state-level unemployment rates. I have ...
funcard's user avatar
  • 43
1 vote
2 answers
126 views

Diff-in-diff with two control groups: compare parallel trends

Suppose in a diff-in-diff problem, I have two control groups, but just one can be used as comparison, so I want to check which of them provides a better comparison. My idea is to compare the parallel ...
Oalvinegro's user avatar
1 vote
0 answers
26 views

Short vs. long horizon event studies?

I am studying the impact of announcements by a Central Bank on a stock's returns. To measure these impacts, I employ an event study (where my event window is 10 days and my estimation window is 60 ...
angelavtc's user avatar
2 votes
0 answers
28 views

Two-way fixed effects with two treatments and interaction between both treatments

I am interested in estimating the effect of two different treatments (T1 and T2) and the interaction of the two (T1 x T2) in a two-way fixed effects event-study design. Both T1 and T2 have a "...
aeiz's user avatar
  • 21
1 vote
0 answers
30 views

Heterogeneous treatment effects with two post-treatment time and without never-treated groups

I am trying to evaluate a policy with staggered policy time (the policy was implemented in two different years), and there are no never-treated groups. The data (four-time period, highly unbalanced - ...
Chang's user avatar
  • 11
0 votes
0 answers
14 views

Difference in Differences with panel data/repeated cross section with dummy

I have to do a public policy evaluation (on R) based on an article where the authors use a quarterly survey to study the effect of a parental leave reform in 2015(reducing the length of leave) on the ...
R S's user avatar
  • 1
2 votes
0 answers
54 views

Meaning of non-anticipation assumption in panel context

I am moving this discussion from this comment section to a full question. Consider a binary policy $D_{it} \in \{0,1\}$ that is not randomly assigned to some units in period $t=\tau$. We have ...
Papayapap's user avatar
  • 1,781
1 vote
3 answers
75 views

Discussing Difference-in-Difference Assumption of the treatment assigment

while reviewing some of the literature on DiD and Event Study designs, I come up with a general question, since DiD is by definition a quasi-experimental approach to determining causal effects when ...
John M. Riveros's user avatar
1 vote
0 answers
32 views

Positively correlated within subject residuals in diff-in-diff shows a deflated false positive error?

My question is regarding in the difference in difference model, I am trying to see the impact of serial correlation within subjects on the treatment effect inference. In my simulation, I have two ...
Yang Song's user avatar
1 vote
1 answer
29 views

Heterogeneity of time invariant characteristic in event study model

I am trying to estimate the impact that the onset of a medical disease has on a number of outcomes, call them $O$. To do this, I am using an event study model with individual and time fixed effects... ...
salve_regina's user avatar
1 vote
2 answers
269 views

Why do we need Exogeneity of Treatment in a Diff in Diff

One of the assumptions of DiD I learnt is that the treatment must be random and unrelated to the outcome. I'm a little bit confused about why we need this assumption. If I look at the typical picture ...
Rainroad's user avatar
  • 243
4 votes
1 answer
64 views

Differences in Differences with Small Violations of Parallel Trends

I am interested (for no particular reason) in estimating a hypothetical Differences-in-Differences model with one period of treatment. However, we observe small non-linear violations of the parallel ...
wflewis's user avatar
  • 45
3 votes
0 answers
33 views

Intuition Behind Difference-in-Differences with Continuous Treatment?

It's very clear to me how a difference-in-difference model works when there is a binary treatment. If the model is $y_{i,t}=\beta_{0}+\beta_{1}P_{t}+\beta_{2}T_{i}+\beta_{3}(P_{t}*T_{t})+u_{i,t}$ ...
econogist's user avatar
5 votes
1 answer
74 views

Is program evaluation (DiD, RD) a structural estimation?

Consider program evaluation methods such as IV, Diff-in-Diff, and RD. According to Haile (2021): "Typically program evaluation requires more than descriptive analysis: one must counterfactually ...
Frank Swanton's user avatar
1 vote
1 answer
80 views

Diff-in-diff parallel trends with a positive outcome

I am trying to test for parallel trends when using diff-in-diffs, where my outcome variable is nonnegative (basically a percentage -- the share of subjects with some property). I run into a simple-...
Serb's user avatar
  • 13
1 vote
1 answer
86 views

How to define treatment & control groups properly?

I’m working on a project examining the effect of a 2016 cash transfer on fertility. Who is eligible for the cash? All families with: 2+ children, or 1 low-income or disabled child. The data doesn’t ...
Anna Bokun's user avatar
2 votes
1 answer
73 views

Difference-in-difference robust to heterogeneous treatment effect - Gendron-Carrier et al. specification

I am trying to extend the results of Gendron-Carrier et al. (2022) article published in the American Economic Journal : Applied Economics which is about the effect of subway opening on pollution. I ...
Galactus's user avatar
0 votes
0 answers
13 views

Cross sectional survey - difference in difference campaign evaluation

I would like to use two cross sectional surveys to evaluate the impact of a media campaign on public opinions. The problem is how to identify the "untreated" at baseline? Technically ...
macster's user avatar
1 vote
1 answer
76 views

Handling diff in diff with multiple treatments, but each treatment is applied once to differents groups

Say each year I observe the effect of a movie-level treatment on the reviews of 10 movies (1 treated movie and 9 non-treated movies) over two periods (pre and post treatment). Each year, the treatment ...
Nicolas L's user avatar
2 votes
1 answer
96 views

Difference in differences with one post and two treatment variables

I want to test the following hypotheses with a difference in differences (DiD) OLS model: 1. The post grant effect on the dependent variable is stronger for patents with at least one rejection than ...
Nils_Denter's user avatar
0 votes
3 answers
311 views

Do anticipation effects almost always bias the DiD ATE or ATT towards a null result and increase standard errors?

I am trying to implement a difference in difference model. Suppose that some treated units anticipate the treatment. Then, that would mean that they respond to the treatment prior to the treatment ...
Daycent's user avatar
  • 250
2 votes
0 answers
485 views

Difference-in differences survival models

I am in a setting where I'd like to estimate the average treatment effect on the treated (ATT) using a cox-proportional hazards model in a DID style setting. I'll adopt the setting of Mastrobuoni and ...
Agar's user avatar
  • 21
0 votes
0 answers
32 views

DiD with 2 groups

I am thinking about a DID approach. Typical DID can be written as follows: $$Y_i=\beta_0+\beta_1D_i+\beta_2T_i+\beta_3 T_i D_i+\epsilon_i$$ $Y_i$ is an outcome, $D_i$ is a binary treatment, and $T_i$ ...
hrkshr's user avatar
  • 195
0 votes
0 answers
25 views

Testing method for the performance of R&D support program

Suppose that Government aims to initiate R&D support program to firms The objective of the support program is to rise R&D expenditures and productivity. I have the following data variables: Q: ...
studentp's user avatar
  • 120
1 vote
0 answers
12 views

Difference in difference papers that look at multiple cohorts at a point in time?

I'm trying to see if a CCT targeted at school age children can mitigate adverse impact of weather shocks faced in childhood on their wages in adulthood (at age 30 say) using a setting similar to Duflo,...
Jan3's user avatar
  • 185
1 vote
0 answers
18 views

Can I use difference in difference if some households in treatment group become part of control group?

I have data on households which can be eligible for a conditional cash transfer (CCT) if household income is below a threshold. The CCT is introduced in year t for eg. Now some households which have ...
Jan3's user avatar
  • 185
0 votes
1 answer
72 views

What should we do if we find the anticipation effect?

In the Difference-in-Differences estimator, one key assumption is the anticipation effect, meaning that the real event date maybe a couple of years before the real event year (yearly data). I am ...
Phil Nguyen's user avatar
  • 1,090
1 vote
0 answers
126 views

DAG for Difference in Difference Studies

I have been looking for DAGs to represent diff-in-diff research designs but did not come across any. How do you represent state/time fixed effects and time-varying effects on a DAG?
abby's user avatar
  • 11
1 vote
0 answers
24 views

What is the purpose of testing "additional effect" in "strict exogeneity" in Diff-in-Diff?

The strict exogeneity are different in using OLS and Difference-in-Differences (DiD) as pointed here. From a recently published paper of Miller, 2021, in the Reviewing DID Assumption part (sorry I did ...
Nguyen Lis's user avatar
1 vote
0 answers
36 views

What does no anticipation following Athey, 2021?

From a paper of Athey, 2021, Journal of Econometrics, I saw the assumption of no anticipation effect is Assumption 2: For all units i, all time periods t , and for all adoption dates $\alpha$ , such ...
Nguyen Lis's user avatar
1 vote
0 answers
107 views

How to test anticipation effect in Diff-in-Diff following Borusyak, 2021?

In general, two most important assumptions that all Difference-in-Differences must satisfy are "no anticipation effects" and "parallel trend". The parallel trend is tested by ...
Nguyen Lis's user avatar
2 votes
1 answer
195 views

Should we test for an anticipation effect in difference-in-differences when a natural event occurs?

Normally, we test for anticipation effects as an assumption when using difference-in-differences. However, it seems that it is necessary for examining the impact of laws on firms' behaviour. For ...
Nguyen Lis's user avatar
0 votes
1 answer
32 views

What does "inter alia" in "difference-indifference" testing?

today my paper get rejected due to this reason when I used the Difference-in-Difference as my methodology In addition, it would need to address concerns about causal estimates as there are a number ...
Nguyen Lis's user avatar
0 votes
0 answers
20 views

How many leads we use for parallel trends testing in daily data (difference-in-differences)?

In a difference-in-differences setting, if we deal with yearly data, we normally test the parallel trends assumption for 3 or 4 years before the event dates. However, when dealing with daily data, how ...
Phil Nguyen's user avatar
  • 1,090
2 votes
1 answer
91 views

Are there any papers which used the event study methodology NOT in a financial market context?

I was advised to implement an event study in the context of air pollution levels after an event had happened. Now my problem is, that I can not really find any introductory literature (or papers), ...
Ballaballala's user avatar