Within Variation Programming

[Jim Stevenson]

I am attempting to learn how to program Stata and want to calculate within variation as individuals change jobs. I am including a sample of my data below but to summarize I have data about corporate CEOs within job 1, 2 and each job thereafter. I want to test how certain governance provisions (the governance score factor is a self-calculated total) impacts corporate outcomes, such as M&A activity. I would prefer to include all years at each job but if necessary can do averages. This is not a traditional event study as the number of years individuals are in jobs and the years covered vary based on the specific individual and position. I intend to use controls such as revt and capx. My dada includes dummies for each Executive.



Thank you for any help you can offer.

Executive	Firm	Exec_Firm	Transaction_Count	governance_score	revt	capx
3	186	1		10	837.1	53.7
3	186	1		10	903	82
3	186	1		10	1001.6	138.1
3	186	1	1	10	1247.7	93.2
3	186	1		10	1565.3	167.3
3	186	1		10	1678.6	86
3	255	2	1	3	712.9	104.052
6	189	3	1	13	15453.71	696.488
6	189	3	3	13	15658.5	846.016
6	226	4		7	32347	658
6	226	4		7	17786	373
6	226	4		7	18424	412
14	20	5	5	5	49219.4
14	20	5	8	5	47603.63
14	20	5		5	49648.5
14	20	5	2	5	43868.5
14	193	6	8	13	11587	386
14	193	6	6	13	10493	305
14	193	6	9	13	10550	271
14	193	6	11	13	11526	280
14	193	6	13	13	11924	320

[Joro Kolev]

I do not know what is "within variation as individuals change jobs."

You might want to explain better what you want to calculate.

[Jim Stevenson]

Joro, Thank you for the follow up. I want to calculate the Transaction Count score for each individual between job 1 and job 2. For example, I would like to compare the number of Transactions of Elon Musk at SpaceX, twitter and Tesla as a way to test how much the governance score influences firm decision making. So the within variation is holding the individual constant how much variation do I get in the number of transactions.