Questions tagged [difference-in-difference]

Often abbreviated DID or DD, this is a technique for inferring causality from observational data. It involves comparing measurements before and after a treatment occurs (hence, the growth rate) in both a group that received the treatment, and an otherwise comparable group that did not.

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

Why people normally exclude the first lead in a two-way fixed effect model?

Borusyak (2021) writes: The first lead,..., is often excluded as a normalization. In a dynamic two-way fixed effect model, when we included some leads and lags, we normally exclude the first lead. ...
0
votes
0answers
4 views

What should we control for when using DDD in country level? Should DDD avoid parallel trend testing?

A very common assumption of DiD is parallel pre-trend satisfaction. And when we subsample, we need to make sure the parallel trend assumption of the subsample also need to be statistically satisfied. ...
0
votes
0answers
7 views

Why homogeneous treatment assumption leads to "spurious identification of long-run treatment" in staggered DiD?

Borusyak, 2021 has a sentence Third, in dynamic specifications, implicit assumptions about treatment effect homogeneity lead to the spurious identification of long-run treatment effects for which no ...
0
votes
0answers
12 views

How to understand strict exogeneity?

Apart from paralell trend assumption, another assumption in parallel trend is strict exogeneity. I have read this answer from Thomas and this paper of Miller. The strict exogeneity in Miller is ...
0
votes
0answers
8 views

What does DiD “differencing-out” group- or time- invariant confounders mean?

From a paper by Miller (2021), he said The DID approach compares changes in outcomes for units that experience an intervention at particular points in time with units that do not, while “differencing-...
0
votes
0answers
15 views

Is there any reason to test Difference-in-Differences if paralell trend is weakly violated?

Normally when testing the parallel trend assumption in the Difference-in-Differences setting, we normally conduct the joint null test for the coefficients of pre-treatment periods. Normally, while the ...
0
votes
0answers
5 views

What do the coefficients of control variables in a Different-in-Differences mean?

In a setting of Difference-in-Differences (in specific, generalized Difference-in-Differences - where the laws is staggered implemented) For example: In a generalized Difference-in-Difference, ...
0
votes
1answer
12 views

Interpretation of event study difference-in-difference coefficient

I understand mathematically what a difference-in-differences model is estimating, but I want to confirm that I am 'translating' it into English/words properly. Let's say I am running the following ...
0
votes
1answer
15 views

Is it a must to test for parallel trends using different sub-groups in difference-in-differences?

I am testing the movement of a whole sample after an event by using generalized difference-in-differences (DiD) with staggered event dates. The whole sample satisfies the parallel trend assumption. I ...
0
votes
0answers
6 views

Whether we need to show all coefficients of all variables after testing DiD?

I am curious about whether it is a need to show all coefficients of all variables after testing DiD. $Y_{i,t} = \alpha_i + \beta_t + \gamma D_{i,t}$ while $i$ and $t$ are unit and time fixed effects. $...
-1
votes
0answers
13 views

Statistical significance for panel linear model for difference in differences analysis

I carried out a difference in differences analysis with fixed effects for units and times: $$ y_{it} = \eta_i + \lambda_t + \delta D_{it} + u_{it}, $$ I have a panel with several Italian villages (my ...
0
votes
0answers
15 views

Difference in difference with binary outcome -- test parallel trend assumption

I am performing this two-period difference-in-difference model, $y_{i,t} = \beta_1 T_{i} + \beta_2 \text{ after}_{i, t} + \gamma T_{i} \text{ after}_{i, t} + \mathbf{x}_{i}^T \boldsymbol{\alpha}  + \...
1
vote
1answer
26 views

What fixed effects should be controlled for in a DDD specification?

I am following work by Borusyak, 2021 which uses a triple-difference design: In triple-difference designs, the data have three dimensions, such as regions $i$, demographic groups $g$, and periods $t$....
0
votes
1answer
58 views

Clarifying the definition of DDD?

I am examining the triple diff (diff-in-diff-in-diff)(DDD) in a staggered setting. Normally, when it comes to DDD, I understood that we examine the differential movement between two sub groups based ...
0
votes
0answers
4 views

How to conclude the result of paralell-trend test following Borusyak's DiD approach?

I am doing a parallel trend test suggested by Borusyak by using event_plot My code and result is as below. I am wondering how can I conclude about the pre-trend ...
2
votes
0answers
10 views

How to read the result of the generalized Difference-in-Differences following Callaway and Sant'Anna (2020)?

I am using the generalized DiD following Callaway and Sant'Anna (2020) by applying the package csdid in STATA. My code is ...
0
votes
1answer
15 views

How many leads do we include for parallel trends testing in daily data (Difference-in-Differences)?

Regarding using generalized difference-in-differences, we normally use the joint null test for 2 or 3 years before the event year. For example, in this answer, @Thomas Bilach mentioned we may perform ...
0
votes
0answers
8 views

DiD with Several Interactions and Staggered Treatment Adoption

I have panel data that have control and treatment groups and I am running a regression that looks like this: ...
0
votes
0answers
8 views

Paralell testing for aggregate and individual variables in generalized Difference-in-Differences?

I am using a joint null test to test a parallel trend for satisfying the key assumption of DiD approach following Miller, 2021. In general, if the p-value of joint null test is insignificant (p-value &...
0
votes
1answer
6 views

Is there any reference for stopping the sample after some years after the event date (staggered Difference-in-Differences)?

I am using a generalized difference-in-differences (DiD) equation for a staggered treatment. The law (i.e., treatment) is implemented at different times. From this discussion of @Thomas Bilach here, ...
1
vote
1answer
18 views

Including pre-treatment covariates in difference-in-differences: what if my covariates determine treatment but are outcomes of treatment?

I have the following question: What should I do with covariates that affect treatment in the pre-treatment period but are affected by treatment once the treatment is in place in a difference-in-...
1
vote
0answers
18 views

Estimate the impact of in-store product placement [duplicate]

I have six variables: sales (weekly), product category, customer segment, store location, week and product placement (aisle, entrance, ...). For each category, segment and location, I observe sales ...
3
votes
1answer
55 views

Estimate the impact of in-store product placement

I have six variables: sales (weekly), product category, customer segment, store location, week and product placement (aisle, entrance, ...). For each category, segment and location, I observe sales ...
1
vote
1answer
34 views

Can I use a differences-in-differences (DID) model to identify when treatment starts?

I am analyzing the impact of a policy, but I have found that if I set the time of treatment to the time the policy was introduced, then a negative DiD coefficient does not return a significant result. ...
0
votes
0answers
15 views

Treatment group size vs. Control group size for difference-in-difference estimation

I am reading an unpublished paper that employs diff-in-diff to firm data containing a total of 2598 observations. Of these observations, 2474 are in the treatment group and 124 in the control group. I ...
0
votes
0answers
13 views

What are the difference-in-differences dealing with the treatment being turned off?

The generalized Difference-in-Differences being used to deal with staggered implementation of laws in different countries (multiple groups and multiple time periods). And due to the heterogeneous ...
0
votes
0answers
14 views

Generalized DiD - Should I drop States which are treated before my sample data begins?

I am estimating the effect of a law change on certain policies for the US states. The time period for my dataset ranges from 1980-2017. My treatment variable is a binary variable that indicates the ...
1
vote
1answer
56 views

What might be the identification challenges with a generalized DiD model where the treatment variable experiences reversals (switches on/off)?

I have a setting where my treatment variable experiences reversals across the panel units in a staggered adoption setting. To estimate the average treatment effects on the treated in a setting that ...
1
vote
1answer
77 views

Classical difference-in-differences: Coding the time (post) variable when treatment starts at different times

I have panel data with 40 treated cases and 40 control cases. I thought about the application of the 'classical' difference-in-differences (DiD) equation with the following linear regression model: $y ...
0
votes
1answer
29 views

Difference-in-differences and control variables

Im currently processing a bigger assignment, and I'm trying to reproduce the following table, which are difference-in-differences (DD) estimates of income change on voter turnout. However, as far as I'...
1
vote
1answer
31 views

Difference-in-differences with two policies

Can I use the general DiD approach with two policy variables (i.e., two treatments) as below? xtreg outcome i.policy_1 i.policy_2 i.year, fe vce(cluster id)
0
votes
1answer
149 views

How to read the result by running did_multiplegt code in Stata

One way of treating heterogeneous impact of staggered laws is to use did_multiplegt method developed by Clément de Chaisemartin. When running the code suggested, I have the result as below, I do not ...
0
votes
1answer
50 views

difference in difference specification

I am trying to see how a policy, say the treatment that happens in period 0, affects an outcome. I am interested in the time-varying effect of the policy. I have a panel data in long format, for ...
0
votes
0answers
38 views

Clustering standard errors for difference in difference

I am running a difference in difference to examine the effect of a merger on petrol prices. I am looking to see whether the prices of company A have increased due to a merger with company B. Local ...
2
votes
0answers
30 views

Confusing about the pre-trend test suggested by Goodman-Bacon, 2019's note?

Goodman, 2019 is a very famous paper decomposing the coefficients of staggered DiD. Here I do not talk about the paper itself, I talk about the note accompanying with this paper. Pre-trend testing ...
1
vote
1answer
80 views

Pre-trend testing of DiD following Pischke (2005)?

This post simplifies how to conduct a pre-trend analysis using "coefficient plotting", which follows from lecture notes by Pischke 2005. I simplified copying the description here: A formal ...
0
votes
1answer
102 views

Monthly difference-in-differences with three-way interactions

For my master thesis I am running a set of difference-in-differences (DD) estimators and lately I am a bit confused about the correct specification. I am investigating the differential impact of the ...
0
votes
0answers
16 views

How to calculate relative lift in Difference-in-Difference

In difference-in-difference finding the absolute difference is pretty straightforward difference = (Test{Post} - Test{Pre}) - (Control{Post} - Control{Pre}) But if ...
0
votes
0answers
22 views

Difference-in-differences on compliers

I'm trying to estimate a generalized DID (two periods three groups)'s local average treatment effect on Stata. The DID code used was: ...
0
votes
0answers
15 views

What is the exact meaning of the difference in differences design?

In economics, many empirical papers say that they utilized a difference-in-differences design. I have already know the canonical DID method (two groups and two time periods) But, the setup in the ...
1
vote
0answers
15 views

What does "sharp design" mean in Chaisemartin (2020)'s paper compared to Goodman-Bacon (2017)'s paper?

When studying the paper of Chaisemartin (2020) comparing to Goodman-Bacon, 2017's paper regarding heterogeneous treatment effects, a commentator said that Chaisermartin's papers have more of a sharp ...
0
votes
1answer
18 views

Diff-in-diff intention to treat

When running an RCT where different individuals are invited to participate and then randomized, what would the interpretation of diff-in-diff method be? To my understanding, diff-in-diff identifies ...
1
vote
0answers
28 views

How much is modest in this case for the usefulness of two-way fixed effects?

Regarding the usefulness of two-way fixed effect (TWFE), Guido Imbens mentioned that TWFE can be good starting point for DiD if the N and T are modest, I am wondering is there any criteria for "...
1
vote
1answer
84 views

What is the reasonable approach to deal with DiD for multiple time periods and groups? (staggered DiD)

Two-way fixed effect (TWFE) has been used for 2 decades for examining the change of some particular objectives after an event, or "generalized Difference-in-Differences (DiD)". However, ...
0
votes
0answers
37 views

Whether we need to cluster by country with staggered laws implementation?

When trying to find a way of avoiding using clustering, I saw that Abadie, 2017 have a great paper mentioned when we should cluster, summarized by McKenzie here. I used the paper of Dasgupta,2019 to ...
2
votes
1answer
156 views

Asking about clustering condition following Abadie, Wooldridge 2017

Abadie, 2017 have a paper about when we should cluster. And this paper has been summarized by McKenzie here. I used the paper of Dasgupta,2019 to link to the summarized work of McKenzie. So, in ...
1
vote
0answers
19 views

How to extract treatment effect and statistical significance in a Matched pair + diff in diff experiment

I have a unique experiment setup where for each user i look at the before and after of a metric and also, i have treatment and control group users. So basically my data looks like: ...
0
votes
1answer
38 views

How to explain the coefficient plotting as pre-trend stability testing for generalized Difference-in-Differences?

Proving the pre-trend parallel is hard in generalized DiD. From this post, Thomas Bilach suggests me a very excellent way to do so which is called "coefficient plotting" In short, it can be ...
2
votes
0answers
32 views

Difference-in-differences model with a matched control group

I need to run a difference-in-differences (DiD) model, but I'm not sure how to construct a formula for this. The problem is that the timing of events affecting the treatment group is not uniform, like ...
1
vote
0answers
35 views

How to check the effect of a treatment when you do not have control group?

I have a facebook dataset ranging from January-2010 to May-2017 in which the focus of my study are the columns including categories (Group A & Group B), Timestamp, Engagement (Sum of likes, ...

1
2 3 4 5
12