Arellano and Bond does not allow interactions terms nor factor-variable operators

[Alejandro Torres]

Dear statalisters,

I am writing you because I am trying to run an xtabond with variables that include interactions but I received a message saying that "interactions are not allowed", also I have an i.industry factor variable, but it says that factor -variable operators are nor allowed neither.

I really appreciate your help to know how to solve this problems.

Thank you very much,

Alejandro

[Andrew Musau]

Use Roodman's xtabond2 instead. It allows factor variables.

ssc install xtabond2

[Alejandro Torres]

Thank you very much Andrew, I am going to test it now.
Best regards !
Alejandro

[Sebastian Kripfganz]

Although it is shameless self promotion, you might want to consider xtdpdgmm as an alternative to xtabond2:

• Kripfganz, S. (2019). Generalized method of moments estimation of linear dynamic panel data models. Proceedings of the 2019 London Stata Conference.
• XTDPDGMM: new Stata command for GMM estimation of linear (dynamic) panel data models

[Alejandro Torres]

Dear statalisters, thank you for your help:

Andrew, I am trying xtabond2 but I received a message saying

Equation not identified. Regessors outnumber instruments

, what does it mean and what can I do? Thank you

sebastian, thank you for your message, I tried your command all but one variables are omitted, may I ask you how to solve it?

Thank you very much again to both.

Alejandro

[Andrew Musau]

Andrew, I am trying xtabond2 but I received a message saying
Equation not identified. Regessors outnumber instruments
, what does it mean and what can I do? Thank you

It's exactly what it states, you have more regressors than instruments and your model is not identified. Try reducing the number of independent variables and see if you get results.

[Sebastian Kripfganz]

Dear statalisters, thank you for your help:

Andrew, I am trying xtabond2 but I received a message saying , what does it mean and what can I do? Thank you

sebastian, thank you for your message, I tried your command all but one variables are omitted, may I ask you how to solve it?

Thank you very much again to both.

Alejandro

Did you specify any instruments at all with the gmm() and/or iv() options? Without seeing the actual code you typed it is difficult to provide any advice.

[Giovanni DiPietra]

Dear statalisters,

I know this is an old post, but regarding op question "Arellano and Bond does not allow interactions terms",

the fact that xtabond2 or xtdpgmm allow for interactions also indicate that, say for two predetermined x1t and x2t , a new variable x3t = x1t * x2t can also be dealt with as predetermined too, wrt the way AB methods create the instrument matrix?

Thank you so much [and sorry if this is the wrong place for such question!]

[Sebastian Kripfganz]

The interaction between two predetermined variables would normally also be treated as predetermined, yes.

[Giovanni DiPietra]

Thank you Sebastian Kripfganz !