{smcl} {right:version 0.0.1, 17 Nov 2020} {title:Instrumental variables regression with Conley, HAC-robust standard errors} {phang} {bf:spatial_hac_iv }{hline 2} Runs an instrumental variable regression, adjusting standard errors for spatial correlation, heteroskedasticity, and autocorrelation {title:Author} {p 4 4 2} Tim Foreman{break} RFF-CMCC European Institute on Economics and the Environment{break} timothy.foreman@eiee.org{break} {title:Syntax} {p 8 16 2} {opt spatial_hac_iv} {depvar} indepvar (endog=inst){cmd:,} lon({it:varname}) lat({it:varname}) timevar({it:varname}) panelvar({it:varname}) distcutoff({it:#}) lagcutoff({it:#}) [bartlett dropvar] {synoptset 20 tabbed}{...} {synopthdr} {synoptline} {synopt :{opt lat}} Latitude variable (degrees){p_end} {synopt :{opt lon}} Longitude variable (degrees){p_end} {synopt :{opt timevar}} Time variable{p_end} {synopt :{opt panelvar}} Panel variable{p_end} {synopt :{opt distcutoff}} Distance cutoff for spatial correlation within a time unit (kilometers){p_end} {synopt :{opt lagcutoff}} Time cutoff for autocorrelation within a panel unit{p_end} {synopt :{opt bartlett}} Use a linear bartlett window for spatial correlations, instead of a uniform kernel{p_end} {synopt :{opt dropvar}} Drops variables that Stata would drop due to collinearity{p_end} {synoptline} {title:Example} Regress y on endogenous variable x with instrument z . spatial_hac_iv y (x=z), lat(lat) lon(lon) timevar(year) panelvar(countrynum) lagcutoff(5) distcutoff(100) dropvar {title:Acknowledgements} {p}The basis for this package was written by Solomon Hsiang and Kyle Meng. It relies on ivreg2, written by Christopher F Baum, Mark E Schaffer, and Steven Stillman {browse "http://ideas.repec.org/c/boc/bocode/s425401.html":http://ideas.repec.org/c/boc/bocode/s425401.html}{p_end}