Basic Difference-in-Differences Method (DID) | Estimation Methods | Stata Tutorials Topic 42

Stata Tutorials Topic 42: Basic Difference-in-Differences Method (DID) | Regression Analysis and Estimation Methods Using Stata
Hi, I am Bob. Today, we will discuss a popular identification strategy in evaluating policy effects. It is the difference-in-differences method. We use this technique to investigate whether a policy or a treatment has a significant causal effect on the outcome variable.
Please download the DID.dta dataset:
drive.google.com/file/d/1uKZF...
#Stata #DID #tutorial #DifferenceinDifferences #tabulate #regress #interactionterm #RegressionAnalysisandEstimationMethods #OLS #treatment #controlgroup #treatmentgroup #tutorial
The Stata commands in this video:
*DID using "tabulate"
tabulate treatment after, summarize(divorced) nofreq nost
display .16188525- .1064257
display .10526316- .06951027
display .05545955- .03575289
*DID using "regress"
regress divorced i.treatment##i.after
*DID with additional controls
regress divorced i.treatment##i.after fam_inc num_chi
【Some Free Courses on My KZbin Channel】
Economics in Real Life:
• Economics in Real Life
Solutions to Introductory Econometrics A Modern Approach 7th Edition:
• Solutions to Introduct...
Solutions to Microeconomics Theory and Applications with Calculus:
• Solutions to Microecon...
Introductory Stata:
• Introductory Stata (2022)
Introductory Microeconomics:
• Introductory Microecon...
Five Minute Econometrics:
• Five Minute Econometri...
Five Minute Stata:
• Stata Tutorials
【國語Mandarin】現實生活中的經濟學:
• 【Mandarin國語】現實生活中的經濟學
【粵語Cantonese】現實生活中的經濟學:
• 【Cantonese粵語】現實生活中的經濟學
【粵語Cantonese】微觀經濟學基礎:
• 粵語Cantonese微觀經濟學基礎2022
【國語Mandarin】五分鐘計量經濟學:
• 【Mandarin國語】五分鐘計量經濟學(計...
【粵語Cantonese】五分鐘計量經濟學:
• 【Cantonese粵語】五分鐘計量經濟學(...
On the Road:
• On the Road
【Become a Supporter of the channel ($2.99) to get PDF transcripts for】
1. Solutions to Microeconomics Theory and Applications with Calculus 5th Edition,
2. Solutions to Introductory Econometrics A Modern Approach 7th Edition,
3. Introductory Stata (2022),
4. Introductory Microeconomics (2022), and
5. Introductory Stata: Summary Statistics and Data Management
/ @bobwenecon

[@yangyijane]

thank you. look forward to the staggered one tutorial.

[@bihuyen3225]

Thank you so much for your video Dr. Bob. About the DID regression result you mentioned on the video, I have a small question is that since Treatment#After's P>|t| = 0.454 and it's considered not significant, so this variable can no longer be used to explain, is that correct? I would appreciate it if you could help me clarify it because I'm also working on DID regress model for my master's thesis and I'm so confused about these P>|t| value

[@BobWenEcon]

Yes. You are right. The law has no significant effect if the assumption that the two states are similar is correct.

[@user-yh3mt1dm8m]

Thank you for this explanation and insights into the DID. Please, are you able to share with me your Do File?

[@BobWenEcon]

Please see the description. For the regression, it is: regress divorced i.treatment##i.after fam_inc num_chi

[@ngbaotu249]

Thank you. May I ask what the regression model of this data is? And where is this data sourced from?

[@BobWenEcon]

Hi, I don't remember whether I collected the data from CPS or PSID many years ago.

[@yabelolake3790]

Do you think the regress command is appropriate for a categorical outcome variable like divorce before and after treatment?

[@BobWenEcon]

We can use a logit or probit model. I use the LPM here to show the DID method.

[@yabelolake3790]

@@BobWenEcon Thank you for your quick response. I am not sure, is there didlogit command? Could you do one KZbin video focusing on Logit or probit model for the DID? Sorry do you mean linear probability model (LPM)?