个人试过没有问题（我用的是联通宽带），你再试看看，先看看ie是否能登陆http://www.stata-journal.com/software/sj7-3/。

IE能登陆，就是STATA不能更新。不知道为什么？

我和你的问题一样啊，着急死了，高手赶紧进来看下啊，万分感激啊!

Title

    xtscc -- Regression with Driscoll-Kraay standard errors


Syntax

        xtscc depvar [indepvars] [if] [in] [weight] [, options]


    options         description
    -------------------------------------------------------------------------------------------------
    lag(#)          set maximum lag order of autocorrelation; default is m(T)=floor[4(T/100)^(2/9)]
    fe              perform fixed effects (within) regression
    pooled          perform pooled OLS/WLS regression; default
    level(#)        set confidence level; default is level(95)
    -------------------------------------------------------------------------------------------------
    You must tsset your data before using xtscc; see [TS] tsset.
    by, statsby, and xi may be used with xtscc; see prefix.
    aweights are allowed unless option fe is specified; see weight.
    See xtscc postestimation for features available after estimation.


Description

    xtscc produces Driscoll and Kraay (1998) standard errors for coefficients estimated by pooled
    OLS/WLS or fixed-effects (within) regression.  depvar is the dependent variable and varlist is an
    optional list of explanatory variables.

    The error structure is assumed to be heteroskedastic, autocorrelated up to some lag and possibly
    correlated between the groups (panels). These standard errors are robust to general forms of
    cross-sectional (spatial) and temporal dependence when the time dimension becomes large. Because
    this nonparametric technique of estimating standard errors places no restrictions on the limiting
    behavior of the number of panels, the size of the cross-sectional dimension in finite samples
    does not constitute a constraint on feasibility -- even if the number of panels is much larger
    than T.  Nevertheless, because the estimator is based on an asymptotic theory, one should be
    somewhat cautious with applying this estimator to panels that contain a large cross-section but
    only a short time dimension.

    The xtscc command is suitable for use with both balanced and unbalanced panels. Furthermore, it
    can handle missing values.


Options

    lag(#) specifies the maximum lag to be considered in the autocorrelation structure.  By default,
        a lag length of m(T)=floor[4(T/100)^(2/9)] is assumed.

    fe performs fixed-effects (within) regression with Driscoll and Kraay standard errors.  These
        standard errors are robust to general forms of cross-sectional ("spatial") and temporal
        dependence (provided that T is sufficiently large). See above.  If the residuals are assumed
        to be heteroskedastic only, use xtreg, fe robust.

    pooled performs pooled OLS/WLS regression with Driscoll-Kraay standard errors.  These standard
        errors are heteroskedasticity consistent and robust to general forms of cross-sectional
        (spatial) and temporal dependence when the time dimension becomes large.  If the residuals
        are assumed to be heteroskedastic only, use xtreg, fe robust.  When the standard errors
        should be heteroskedasticity- and autocorrelation consistent, use either regress, cluster()
        or newey, lag(#) force.  Analytic weights are allowed for use with option pooled; see weight.

    level(#); see estimation options.



Examples

        . webuse grunfeld

    Pooled OLS estimation

        . reg invest mvalue kstock, robust cluster(company)

        . est store robust

        . newey invest mvalue kstock, lag(4) force

        . est store newey

        . xtscc invest mvalue kstock, lag(4)

        . est store dris_kraay

        . est table *, b se t

    Fixed-effects (within) regression

        . est clear

        . xtreg invest mvalue kstock, fe robust

        . est store fe_robust

        . xtscc invest mvalue kstock, fe lag(4)

        . est store fe_dris_kraay

        . est table *, b se t


Reference

    Driscoll, J. C., and A. C. Kraay. 1998. Consistent covariance matrix estimation with spatially
        dependent panel data.  Review of Economics and Statistics 80: 549-560.


Notes

    - The main procedure of xtscc is implemented in Mata and largely follows Driscoll and Kraay's
      GAUSS program, which is available from http://www.johncdriscoll.net/.
    - The xtscc uses functions from Ben Jann's moremata package.


Acknowledgments

    I thank David M. Drukker and William Gould from StataCorp for their useful comments and
    suggestions.


Author

    Daniel Hoechle, University of Basel, daniel.hoechle@unibas.ch


Also see

    Manual:  [R] regress, [TS] newey, [XT] xtreg

    Online:  xtscc postestimation;
             [TS] tsset, [R] regress, [TS] newey, [XT] xtreg, [P] _robust
(remote file ends)

有时候网络慢也出现这样的事情