补充下，是stata的命令

联立一起估计，是没有gmm命令的
但可以考虑单方程一条一条估计，用ivregress


Title

    [R] ivregress -- Single-equation instrumental-variables regression


Syntax

        ivregress estimator depvar [varlist1] (varlist2 = varlist_iv) [if] [in] [weight] [,
              options]


    estimator               Description
    ---------------------------------------------------------------------------------------
    2sls                    two-stage least squares (2SLS)
    liml                    limited-information maximum likelihood (LIML)
    gmm                     generalized method of moments (GMM)
    ---------------------------------------------------------------------------------------

OK，谢谢     

xtabond2可以吧

Description

    xtabond2 can fit two closely related dynamic panel data models.  The first
    is the Arellano-Bond (1991) estimator, which is also available with
    xtabond, though without the two-step standard error correction described
    below.  It is sometimes called "difference GMM." The second is an
    augmented version outlined by Arellano and Bover (1995) and fully
    developed by Blundell and Bond (1998).  It is known as "system GMM."
    Roodman (2009) provides a pedagogic introduction to linear GMM, these
    estimators, and xtabond2.  The estimators are designed for dynamic
    "small-T, large-N" panels that may contain fixed effects and--separate
    from those fixed effects--idiosyncratic errors that are heteroskedastic
    and correlated within but not across individuals.  Consider the model:

    y_it = x_it * b_1 + w_it * b_2 + u_it      i=1,...,N;     t=1,...,T
    u_it = v_i + e_it,
应该可以的