Regressing data from two different years

[Paul Mannschreck]

Hey all,

I want to regress ln(income) on some variables such as education, computer skills and other skill indicators. At first, I used a simple OLS regression for observations from the same year 2008. Now I am concerned about reversed causality, such that people with higher income could invest into courses to improve their skills. I want to reduce reverse causality concerns by using income data for 2014. When I run the regression, stata tells me that there are no observations which makes sense after browsing the data. Is there any "short" command where I can tell stata to use data on income for 2014 and the rest for 2008.

Thank you in advance!
Paul

[William Lisowski]

Welcome to Statalist, Paul.

There are a lot of ways your data could be organized, and a corresponding number of ways your regression command could be structured, so what you need to change our dependent variable in the way you described depends very much on the organization of your data and the form of your command.

Please review the Statalist FAQ linked to from the top of the page, as well as from the Advice on Posting link on the page you used to create your post. Note especially sections 9-12 on how to best pose your question.

The more you help others understand your problem, the more likely others are to be able to help you solve your problem.

[Phil Bromiley]

To directly address your question, if you have xtset your data (so Stata knows what a year is), you can refer to previous years with the lag L. or L2. etc. prefix
reg income L2.education

[Carlo Lazzaro]

Paul:
there's another endogeneity source in your model, as individual ability can affect both education (better educational attainments) and income (higher income).