{smcl} {* 5 Feb 2010}{...} {hline} help for {hi:xtivreg2} {hline} {title:Extended IV/2SLS, GMM and AC/HAC, LIML and k-class regression for panel data models} {p 4}Full syntax, fixed effects and first differences models {p 8 14}{cmd:xtivreg2} {it:depvar} [{it:varlist1}] {cmd:(}{it:varlist2}{cmd:=}{it:varlist_iv}{cmd:)} [{it:weight}] [{cmd:if} {it:exp}] {bind:[{cmd:in} {it:range}] {cmd:, {{cmd:fe} | {cmd:fd}}} } {bind:[{cmdab:i:var(}{it:varname}{cmd:)} {cmdab:t:var(}{it:varname}{cmd:)}} {cmd:gmm} {cmd:bw(}{it:#}{cmd:)} {cmd:kernel(}{it:string}{cmd:)} {cmd:liml} {cmd:fuller(}{it:#}{cmd:)} {cmd:kclass(}{it:#}{cmd:)} {cmd:coviv} {cmd:cue} {cmd:cueinit}{cmd:(}{it:matrix}{cmd:)} {cmdab:cueopt:ions}{cmd:(}{it:string}{cmd:)} {cmdab:r:obust} {cmdab:cl:uster}{cmd:(}{it:varlist}{cmd:)} {cmd:orthog(}{it:varlist_ex}{cmd:)} {cmd:endog(}{it:varlist_en}{cmd:)} {cmdab:red:undant(}{it:varlist_ex}{cmd:)} {cmd:fwl(}{it:varlist}{cmd:)} {cmdab:sm:all} {cmdab:noc:onstant} {cmd:first} {cmd:ffirst} {cmd:savefirst} {cmdab:savefp:refix}{cmd:(}{it:prefix}{cmd:)} {cmd:rf} {cmd:saverf} {cmdab:saverfp:refix}{cmd:(}{it:prefix}{cmd:)} {cmd:nocollin} {cmd:noid} {cmdab:l:evel}{cmd:(}{it:#}{cmd:)} {cmdab:nohe:ader} {cmdab:nofo:oter} {cmdab:ef:orm}{cmd:(}{it:string}{cmd:)} {cmdab:dep:name}{cmd:(}{it:varname}{cmd:)} {bind:{cmd:plus} ]} {p 4}Replay syntax {p 8 14}{cmd:xtivreg2} {bind:[{cmd:,} {cmd:first}} {cmd:ffirst} {cmd:rf} {cmdab:l:evel}{cmd:(}{it:#}{cmd:)} {cmdab:nohe:ader} {cmdab:nofo:oter} {cmdab:ef:orm}{cmd:(}{it:string}{cmd:)} {cmdab:dep:name}{cmd:(}{it:varname}{cmd:)} {cmd:plus} ]} {p 4}Version syntax {p 8 14}{cmd:xtivreg2}, {cmd:version} {p}{cmd:xtivreg2} may be used with time-series or panel data, in which case the data can be {cmd:tsset} before using {cmd:xtivreg2}; see help {help tsset}. {p}All {it:varlists} may contain time-series operators; see help {help varlist}. {p}{cmd:aweight}s, {cmd:fweight}s, {cmd:iweight}s and {cmd:pweight}s are allowed; see help {help weights}. {p}The syntax of {help predict} following {cmd:xtivreg2} is {p 8 16}{cmd:predict} [{it:type}] {it:newvarname} [{cmd:if} {it:exp}] [{cmd:in} {it:range}] [{cmd:,} {it:statistic}] {p}For the fixed-effects estimator, {it:statistic} is {p 8 23}{cmd:e}{space 11}v_it, the idiosyncratic component of the error term{p_end} {p}and is available only for the estimation sample. {p}For the first-differences estimator, {it:statistic} is {p 8 23}{cmd:e}{space 11}v_it - v_it-1, the first-differenced idiosyncratic component{p_end} {p 8 23}{cmd:xb}{space 10}(X_it - X_it-1)*b_hat, the fitted values{p_end} {p}and are available in and out of sample. Use {cmd:e(sample)} if wanted only for the estimation sample. {title:Description} {p}{cmd:xtivreg2} implements IV/GMM estimation of the fixed-effects and first-differences panel data models with possibly endogenous regressors. It is essentially a wrapper for {cmd:ivreg2}, which must be installed for {cmd:xtivreg2} to run (version 2.1.11 or above of {cmd:ivreg2} is required for Stata 9; Stata 8.2 requires {cmd:ivreg28}). {cmd:xtivreg2} supports all the estimation and reporting options of {cmd:ivreg2}; see help {help ivreg2} for full descriptions and examples. In particular, all the statistics available with {cmd:ivreg2} (heteroskedastic, cluster- and autocorrelation-robust covariance matrix and standard errors, overidentification and orthogonality tests, first-stage and weak/underidentification statistics, etc.) are also supported by {cmd:xtivreg2} and will be reported with any degrees-of-freedom adjustments required for a panel data estimation. {p}The degrees-of-freedom adjustments depend on whether the estimation is fixed-effects or first-differences, and whether it uses the {cmd:cluster} option. {p}For fixed-effects estimation without {cmd:cluster}, the covariance matrix and regression statistics (identification and overidentification statistics, first-stage regressions and tests, etc.) are adjusted for the number of fixed effects N_g. With large-sample statistics, the covariance matrix has the adjustment (N-N_g); with small-sample statistics, the adjustment is (N-N_g-K), where K is the number of regressors. {p}For fixed-effects estimation with {cmd:cluster}, {cmd:xtivreg2} makes {it:no} degrees-of-freedom adjustment for the number of fixed effects. This follows the formulation of a cluster-robust covariance matrix for the fixed-effects model as originally proposed by Arellano (1987); see, e.g., Wooldridge (2002), p. 275. Stata's official {cmd:xtivreg}, {cmd:xtreg} and {cmd:areg} (as of version 9.1, October 2005), by contrast, use the (N-N_g-K) adjustment, which is somewhat conservative in this context. {it:However}, the approach used by {cmd:xtivreg2} requires that no panel overlaps more than one cluster. That is, either the panel variable is identical to the cluster variable, or panels are uniquely assigned to clusters. If any panel is contained in more than one cluster, the {cmd:xtivreg2} approach is invalid, and it will exit with an error. {p}First-differences estimation makes no degrees-of-freedom adjustment, irrespective of whether {cmd:cluster} is used. {p}Other features of {cmd:xtivreg2} and differences vs. official {cmd:xtivreg}: {p 3 3}{cmd:xtivreg2} supports only the fixed effects and the first-differences panel models; the option {cmd:fe} or {cmd:fd} is required. GLS random effects is not supported. {p 3 3}{cmd:xtivreg2} does not estimate or report a constant with the fixed effects model {cmd:fe}. {p 3 3}If {cmd:ivreg2} version 3.0 or later is installed, {cmd:xtivreg2} supports 2-way clustering; see help {help ivreg2} for details. {p 3 3}First-differences estimation with {cmd:xtivreg2} yields estimates identical to {cmd:ivreg2} when the latter is supplied with all variables expressed in first-differences. {p 3 3}{cmd:xtivreg2} allows use of time series operators. {p 3 3}For fixed-effects estimation, the data must either be {cmd:tsset}, the panel id variable set with {cmd:iis}, or the panel id variable supplied to {cmd:xtivreg2} with the {cmd:ivar} option. For first-differences estimation, the data must be {cmd:tsset}. {p 3 3}{cmd:xtivreg2} supports all types of weights. {p 3 3}The R-squared reported by {cmd:xtivreg2} for the fixed-effects estimation is the "within R-squared" obtained by estimating the equation in mean-deviation form. {p 3 3}{cmd:xtivreg2} supports simple fixed effects and first-differences estimation with no endogeneous variables, i.e., {cmd:(}{it:varlist2}{cmd:=}{it:varlist_iv}{cmd:)} can be omitted. {title:Examples} {p 8 12}{stata "use http://fmwww.bc.edu/ec-p/data/macro/abdata.dta" : . use http://fmwww.bc.edu/ec-p/data/macro/abdata.dta }{p_end} {p 8 12}(Layard & Nickell, Unemployment in Britain, Economica 53, 1986, from Ox dist) {p 8 12}{stata "tsset id year" :. tsset id year} {col 0}({cmd:xtivreg} vs. {cmd:xtivreg2}, fixed effects) {p 8 12}{stata "xtivreg2 ys k (n=l2.n l3.n), fe small" : . xtivreg2 ys k (n=l2.n l3.n), fe small} {p 8 12}{stata "xtivreg ys k (n=l2.n l3.n), fe small" : . xtivreg ys k (n=l2.n l3.n), fe small} {col 0}({cmd:xtivreg2} vs. {cmd:ivreg2} vs. {cmd:xtivreg}, first-differences) {p 8 12}{stata "xtivreg2 ys k (n=l.n l2.n), fd small first" : . xtivreg2 ys k (n=l.n l2.n), fd small first} {p 8 12}{stata "ivreg2 d.ys d.k (d.n=ld.n l2d.n), small first" : . ivreg2 d.ys d.k (d.n=ld.n ld2.n), small first} {p 8 12}{stata "xtivreg ys k (n=l.n l2.n), fd small" : . xtivreg ys k (n=l.n l2.n), fd small} {title:Citation of xtivreg2} {p}{cmd:xtivreg2} is not an official Stata command. It is a free contribution to the research community, like a paper. Please cite it as such: {p_end} {phang}Schaffer, M.E., 2010. xtivreg2: Stata module to perform extended IV/2SLS, GMM and AC/HAC, LIML and k-class regression for panel data models. {browse "http://ideas.repec.org/c/boc/bocode/s456501.html":http://ideas.repec.org/c/boc/bocode/s456501.html}{p_end} {title:References} {p 0 4}Arellano, M. 1987. Computing Robust Standard Errors for Within-Groups Estimators. Oxford Bulletin of Economics and Statistics, Vol. 49, pp. 431-34. {p 0 4}Baum, C.F., Schaffer, M.E., and Stillman, S. 2003. Instrumental Variables and GMM: Estimation and Testing. The Stata Journal, Vol. 3, No. 1, pp. 1-31. Unpublished working paper version: Boston College Department of Economics Working Paper No 545. http://fmwww.bc.edu/ec-p/WP545.pdf {p 0 4}Wooldridge, J.M. 2002. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press. {title:Author} Mark E Schaffer, Heriot-Watt University, UK m.e.schaffer@hw.ac.uk {title:Also see} {p 0 19}On-line: help for {help ivreg2}, {help overid}, {help ivendog}, {help ivhettest}, {help ivreset}, {help xtoverid}, {help condivreg} (if installed) {p_end}