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.
808 questions
0 votes
0 answers
33 views
Regression Discontinuity with longitudinal data and possibility of multiple treatments
I'm wondering how to approach this project idea: I found that the Department of Housing and Urban Development (HUD) requires that areas it allocates funding towards have a population of at least 50,...
- 21
2 votes
1 answer
124 views
Staggered Difference-in-Differences: Setting Reference Period to $t = -3$
This is more of an econometrics question so I haven't included any data. I'm running a staggered difference-in-differences (DiD) using the did_multiplegt_dyn ...
1 vote
0 answers
87 views
Multicolllinearity with Sun and Abraham time fixed-effects in cross-sectional (non-panel) event studies (birth cohort in my case ,*not* calendar time)
I want to preface this by noting that I am an undergraduate, so my apologies if I seem to be making glaring mistakes. I'm conducting an event study where I have cross-sectional data, and not panel ...
2 votes
1 answer
57 views
Difference-in-differences (DiD) with policy change during post-period
I’m working with pooled cross-sectional data from the Current Population Survey on California’s Paid Family Leave (PFL) program. I need guidance on modeling a difference-in-differences (DiD) setup ...
0 votes
0 answers
38 views
DiD with incomplete data
I want to test if the onset of Policy A encouraged out-migration in counties with poor internet connectivity (which is a binary variable). Policy A was implemented in the year 2006. I have internet ...
- 183
0 votes
0 answers
36 views
DiD with two pre-treatment waves
I am trying to set up a DiD model with two pre-treatment models. I want to assess if a treatment encourages individuals to migrate across counties. The treatment is implemented between waves 2 and 3. ...
- 183
0 votes
0 answers
34 views
Should I include year fixed-effects in a differences-in-differences model with multiple years, one treated group and only one treatment time?
I'm looking at the effect of a law on a continuous outcome. The law was implemented in a specific year and only affected the treated group. I don’t have more detailed data - just a column for year (I ...
- 43
0 votes
0 answers
48 views
Two models for Identifying Average Treatment Effect (ATT)
to evaluate treatment effect under two period and $N$ individuals, there are two forms of models to derive ATT: \begin{align} Y_{it}&=\alpha_0+\beta D^1_i+\gamma T_{t}+\delta\left(D^1_i\times T_{t}...
1 vote
1 answer
97 views
Interpretation of difference-in-differences (DiD) coefficient when treatment lasts for more than one period
In introductory courses, difference-in-differences (DiD) might be taught by framing the research design into a "pre-treatment" period and a "post-treatment" period. That is, there ...
- 3,403
0 votes
0 answers
58 views
Controlling for pre-treatment covariates in Difference-in-Difference
For Two-Way Fixed Effects designs, a common advice is to include pre-treatment covariates to improve parallel trends. Let's say the data is a balanced panel and I want to assess the effect of policy D....
- 1
2 votes
0 answers
99 views
Event Studies vs. Differences-in-Differences: Can adding a control group p-hack your results?
If we are comparing event studies with differences-in-differences (DiD), I would say that, in general, the structure of DiD would provide a cleaner identification given the control. What I'm wondering ...
- 922
2 votes
1 answer
127 views
Non-linear difference-in-differences (DiD) model and negative p-values
I have a Poisson diff-in-diff model. Following Puhani (2012), I'm trying to estimate the ATT by differencing the marginal effects, like this: ...
- 3,403
0 votes
0 answers
20 views
Why is the LP-DID estimator numerically identical to the static TWFE-DID estimator in the two-groups-and-multiple time-periods setting?
Dube et al. propose the "local projection (LP) based difference-in-differences approach" (LP-DID). In the setting with just two groups and T periods, where all units are treated at the same ...
1 vote
0 answers
65 views
difference-in-differences with qualitative treatment intensity
I have a staggered diff-in-diff regression with multiple time periods where each unit i is treated at different points in time t. Ignoring more advanced methods such as Callaway and Sant'Anna (2021), ...
- 383
2 votes
1 answer
90 views
Staggered diff-in-diff with different fixed effects
I have a staggered diff-in-diff regression with multiple time periods where each unit i is treated at different points in time t. The units are firms in this case, and they belong to different ...
- 383