{smcl} {* 22June2020}{...} {cmd:help npsynth} {hline} {title:Title} {p2colset 5 18 20 2}{...} {p2col: {hi:npsynth} {hline 1}}Nonparametric Synthetic Control Method{p2colreset}{...} {title:Syntax} {p 8 17 2} {cmd:npsynth} {it: outcome} {it:varlist}{cmd:,} {cmd:trperiod}{cmd:(}{it:#}{cmd:)} {cmd:bandw}{cmd:(}{it:#}{cmd:)} {cmd:panel_var}{cmd:(}{it:varname}{cmd:)} {cmd:time_var}{cmd:(}{it:varname}{cmd:)} {cmd:trunit}{cmd:(}{it:#}{cmd:)} {cmd:kern}{cmd:(}{it:{help npsynth##kerneltype:kerneltype}}{cmd:)} [{cmd:npscv} {cmd:n_grid}{cmd:(}{it:#1,#2}{cmd:)} {cmd:save_res}{cmd:(}{it:filename}{cmd:)} {cmd:w_median} {cmd:gr_y_name}{cmd:(}{it:name}{cmd:)} {cmd:gr_tick}{cmd:(}{it:#}{cmd:)} {cmd:gr1} {cmd:gr2} {cmd:gr3} {cmd:save_gr1}{cmd:(}{it:graphname1}{cmd:)} {cmd:save_gr2}{cmd:(}{it:graphname2}{cmd:)} {cmd:save_gr3}{cmd:(}{it:graphname3}{cmd:)}] {title:Description} {pstd}{cmd:npsynth} extends the Synthetic Control Method (SCM) for program evaluation proposed by Abadie and Gardeazabal (2003) and Abadie, Diamond, and Hainmueller (2010) to the case of a nonparametric identification of the synthetic (or counterfactual) time pattern of a treated unit. The model assumes that the treated unit - such as a country, a region, a city, etc. - underwent a specific intervention in a given year, and estimates its counterfactual time pattern, the one without intervention, as a weighted linear combination of control units based on the predictors of the outcome. The nonparametric imputation of the counterfactual is computed using weights proportional to the vector-distance between the treated unit's and the controls' predictors, using a kernel function with pre-fixed bandwidth. The routine provides a graphical representation of the results for validation purposes. {phang} According to the {cmd:npsynth} syntax: {phang} {it:outcome}: is the target variable over which measuring the impact of the treatment {phang} {it:varlist}: is the set of covariates (or observable confounding) predicting the outcome in the pre-treatment period {phang} {cmd:kern}{cmd:(}{it:kerneltype}{cmd:)} specifies the type of kernel function to use for building synthetic weights. {phang} {cmd:trunit}{cmd:(}{it:#}{cmd:)} specifies the treated unit, with value {it:#} in 'panel_var'. {phang} {cmd:trperiod}{cmd:(}{it:#}{cmd:)} specifies the time in which treatment starts. {phang} {cmd:bandw}{cmd:(}{it:#}{cmd:)} specifies the bandwidth of the kernel weighting function. {phang} {cmd:panel_var}{cmd:(}{it:varname}{cmd:)} specifies the panel variable. {phang} {cmd:time_var}{cmd:(}{it:varname}{cmd:)} specifies the time variable. {title:Options} {phang} {cmd:npscv}: allows for computing the optimal bandwidth minimizing the pre-treatment RMSPE. The default length of the grid over which finding the optimal bandwidth is 20, which means that the bandwidths' grid is [0.1, 0.2, ..., 2]. This option returns the optimal bandwidth in the e-class object {inp:e(opt_band)}. {phang} {cmd:n_grid}{cmd:(}{it:#1,#2}{cmd:)} allows to specify the length of the grid over which finding the optimal bandwidth. The default values are (1,20), which means that the bandwidths' grid is [0.1, 0.2, ..., 2]. {phang} {cmd:save_res}{cmd:(}{it:filename}{cmd:)} allows to save the treated factual and counterfactual time patterns in {it:filename.dta}. {phang} {cmd:w_median} specifies that the unique vector of synthetic weights is calculated by the yearly weights' median (the default uses the mean). {phang} {cmd:gr_y_name}{cmd:(}{it:name}{cmd:)} allows to give a convenient name to the outcome variable to appear in the graphs. {phang} {cmd:gr_tick}{cmd:(}{it:#}{cmd:)} allows to set the tick of the time in the time axis of the graphs. {phang} {cmd:gr1}: allows to plot the the pre-treatment balancing and parallel trend graph. {phang} {cmd:gr2}: allows to plot the overall treated/synthetic pattern comparison graph. {phang} {cmd:gr3}: allows to plot the overall pattern of the difference between the treated and synthetic pattern graph. {phang} {cmd:save_gr1}{cmd:(}{it:graphname1}{cmd:)} allows to save graph 1, i.e. the pre-treatment balancing and parallel trend. {phang} {cmd:save_gr2}{cmd:(}{it:graphname2}{cmd:)} allows to save graph 2, i.e. the overall treated/synthetic pattern comparison. {phang} {cmd:save_gr3}{cmd:(}{it:graphname3}{cmd:)} allows to save graph 3, i.e. the overall pattern of the difference between the treated and synthetic pattern. {marker kerneltype}{...} {synopthdr:kerneltype_options} {synoptline} {syntab:kern} {p2coldent : {opt epan}}uses a Epanechnikov kernel{p_end} {p2coldent : {opt normal}}uses a Normal kernel {p_end} {p2coldent : {opt biweight}}uses a Biweight (or Quartic) kernel{p_end} {p2coldent : {opt uniform}}uses a Uniform kernel{p_end} {p2coldent : {opt triangular}}uses a Triangular kernel{p_end} {p2coldent : {opt tricube}}uses a Tricube kernel{p_end} {synoptline} {pstd} {cmd:npsynth} returns the following objects: {pmore} {inp:e(bandh)} is the bandwidth used within the selected kernel function. {pmore} {inp:e(RMSPE)} is the Root Mean Squared Prediction Error of the estimated model. {pmore} {inp:e(W)} is the vector of (kernel) weights. {title:Remarks} {pstd} - The panel dataset must be perfectly balanced, and must not contain missing values. {pstd} - Please remember to use the {cmdab:update query} command before running this program to make sure you have an up-to-date version of Stata installed. {title:Example} . use Ita_exp_euro , clear . tsset reporter year . global xvars "ddva1 log_distw sum_rgdpna comlang contig" . npsynth ddva1 $xvars , panel_var(reporter) time_var(year) /// trperiod(2000) trunit(11) bandw(0.4) kern(triangular) gr1 gr2 gr3 /// save_gr1(gr1) save_gr2(gr2) save_gr3(gr3) /// gr_y_name("Domestic Direct Value Added Export (DDVA)") gr_tick(5) {title:References} {phang} Abadie, A., Diamond, A., and Hainmueller, J., 2010. Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California's Tobacco Control Program, {it:Journal of the American Statistical Association}, 105, 490, 493-505. {p_end} {phang} Abadie, A., and Gardeazabal, J., 2003. The Economic Costs of Conflict: A Case Study of the Basque Country, {it:American Economic Review}, 93, 1, 112-132. {p_end} {phang} Cerulli, G. 2015. {it:Econometric Evaluation of Socio-Economic Programs: Theory and Applications}, Springer. {p_end} {phang} Cerulli, G., 2019. A flexible Synthetic Control Method for modeling policy evaluation, {it:Economics Letters}, 182, 40-44. {p_end} {title:Author} {phang}Giovanni Cerulli{p_end} {phang}IRCrES-CNR{p_end} {phang}Research Institute on Sustainable Economic Growth, National Research Council of Italy{p_end} {phang}E-mail: {browse "mailto:giovanni.cerulli@ircres.cnr.it":giovanni.cerulli@ircres.cnr.it}{p_end} {title:Also see} {psee} Online: {helpb synth} {p_end}