I may be misunderstanding something and need some help

So I am looking at inheritances and want to see what kind of effect receiving one could have on the labour supply of an individual. I want to use a difference in difference approach.

In the data 2 is = 0. So the treated group are those that did not expect an inheritance but received one in time period 2, while the untreated are those that did not expect an inheritance and did not receive one.

Then I can execute the regression and I do get a result.

The result is not that important since I still need to change something with the data anyway, but I did get a result.

So I am wondering, did I do this all correctly? I have one group of individuals in time period 1 and divide them into two depending on if they received or did not receive an inheritance in time period 2. Is this consistent with diff in diff since it usually is done with 2 separate groups of individuals were one received some sort of treatment and the other didn't?

From my stand point it should be doable. I would eliminate the need for a common trend assumption and the difference between these two groups would be the actual difference. Any comments?

So I am looking at inheritances and want to see what kind of effect receiving one could have on the labour supply of an individual. I want to use a difference in difference approach.

Code:

bys id: gen treatment=0 if (expectation==2 & f.gift_received==2) | (l.expectation==2 & gift_received==2) bys id: replace treatment=1 if (expectation==2 & f.gift_received==1) | (l.expectation==2 & gift_received==1)

Code:

mi estimate, esampvaryok: reg job_hours treatment time i.time#i.treatment, robust

Code:

ultiple-imputation estimates Imputations = 5 Linear regression Number of obs = 1,114 Average RVI = 0.0278 Largest FMI = 0.0463 Complete DF = 1110 DF adjustment: Small sample DF: min = 686.75 avg = 993.23 max = 1,103.02 Model F test: Equal FMI F( 3, 1095.7) = 4.63 Within VCE type: Robust Prob > F = 0.0032 -------------------------------------------------------------------------------- job_hours | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------------+---------------------------------------------------------------- treatment | -3.349237 1.606441 -2.08 0.037 -6.501285 -.1971902 time | -1.93333 .9063042 -2.13 0.033 -3.711605 -.1550552 | time#treatment | 0 1 | -.3600156 2.466202 -0.15 0.884 -5.19907 4.479039 1 0 | 0 (omitted) 1 1 | 0 (omitted) | _cons | 36.44618 .6652147 54.79 0.000 35.14008 37.75228

The result is not that important since I still need to change something with the data anyway, but I did get a result.

So I am wondering, did I do this all correctly? I have one group of individuals in time period 1 and divide them into two depending on if they received or did not receive an inheritance in time period 2. Is this consistent with diff in diff since it usually is done with 2 separate groups of individuals were one received some sort of treatment and the other didn't?

From my stand point it should be doable. I would eliminate the need for a common trend assumption and the difference between these two groups would be the actual difference. Any comments?

## Comment