I'm using Stata 14 for Mac and I'm exploring a data set. It was going well but I've hit a wall.

The question is phrased as such:

"Estimate the hours equation as a standard Tobit model (Tobit I) on the full sample. For this purpose estimate a simple (reduced-form) wage equation first (including a constant term and the standard explanatory variables, like age, age squared, education, …etc) and use this estimate to predict wages for all observations in your sample (do not account for selectivity in the wage equation at this stage). Take care of the identification issue here, since the expected (predicted) wage is to be used as explanatory variable in the hours equation to be estimated by the Tobit model. Write down the likelihood function for this model. Under what assumptions are estimated coefficients consistent? Derive wage and income elasticities of hours supplied (as defined above) for your estimates evaluated at sample means. To what extent do they differ from those derived under a)"

I'd like to ask for general advice on how to approach this question e.g.:

1) How to estimate a reduced-from equation in Stata?
2) How does that actually translate to estimation using the Tobit model?
3) What's the identification issue?

along any other relevant help you can offer.

I'm sure this is far more rudimentary compared to what you're used to but I wouldn't ask I wasn't totally confused. Any help would be much appreciated!


Thanks in advance