Non-classical Measurement Error

[Panos Markou]

[Cross-posted here as well: https://stats.stackexchange.com/questions/330619/non-classical-measurement-error]

I have a question similar to this thread. Although there was one answer, it did not capture the essence of the poster's question.

In essence, I am trying to estimate y = a + bx* + e. Unfortunately, I do not observe x*, but rather a proxy x = x* + v. I know that in such cases, the model suffers from endogeneity (classical error-in-variables.

In my case, however, v ≥ Ø and so E[v] ≠ Ø.

The question is can I still use traditional instrumental variable techniques, or why not? If not, what options do I have? How can I implement this in Stata?

Thanks,
Panos

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

I suspect you've created a problem you don't need. If you think of the proxy as x = a + x* + v where a is a constant, then a could be seen making the mean of the error zero. I think this suggests standard instrumental variables estimation would be OK.