# RCM **Repository Path**: econometric/rcm ## Basic Information - **Project Name**: RCM - **Description**: Stata回归合成控制法 - **Primary Language**: Unknown - **License**: Not specified - **Default Branch**: master - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 3 - **Forks**: 7 - **Created**: 2021-08-23 - **Last Updated**: 2023-09-17 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README ``` *======================================== * 高级计量经济学 *======================================== * 计量经济学服务中心 *------------------------------------------------------------------------------- * 参考资料： * 《初级计量经济学及Stata应用：Stata从入门到进阶》 * 《高级计量经济学及Stata应用：Stata回归分析与应用》 * 《量化社会科学方法》 * 《社会科学因果推断》 * 《面板数据计量分析方法》 * 《时间序列计量分析方法》 * 《高级计量经济学及Eviews应用》 * 《R、Python、Mtalab初高级教程》 * 《空间计量入门：空间计量在Geoda、GeodaSpace中的应用》 * 《零基础|轻松搞定空间计量：空间计量及GeoDa、Stata应用》 * 《空间计量第二部：空间计量及Matlab应用课程》 * 《空间计量第三部：空间计量及Stata应用课程》 * 《空间计量第四部：《空间计量及ArcGis应用课程》 * 《空间计量第五部：空间计量经济学》 * 《空间计量第六部：《空间计量及Python应用》 * 《空间计量第七部：《空间计量及R应用》 * 《空间计量第八部：《高级空间计量经济学》 *------------------------------------------------------------------------------- ``` # 1、简介 RCM有效地实现了回归控制方法(rcm)，即一种用于项目评估的面板数据方法(Hsiao et al., J. Ap. Met. 2012)，该方法利用横截面相关性，通过线性回归(OLS)、套索或套索后OLS构建单个治疗单元的反事实结果。协变量(indepvars)被允许进一步改进by Hsiao and Zhou(J. Ap. Met., 2019)。RCM也支持使用反事实和/或时间来进行统计推断的安慰剂试验。关于rcm使用的详细指南，请参阅Yan和Chen(2021)。 # 2、命令下载安装 ``` rcm -- efficient implementation of regression control method (RCM), aka panel data approach (PDA) for program evaluation (Hsiao et al. 2012) ``` 在Stata中实现“回归控制方法(RCM)” 安装方法命令为: ``` ssc install rcm, all replace ``` 要求：Stata版本>= 16 安装文件包括: rcm.ado and rcm.sthlp: Stata ado and help file 其他数据文件包括growth.dta 和repgermany.dta这两个数据 # 3、语法格式为： ``` rcm depvar [indepvars] [if] [in] , trunit(#) trperiod(#) [options] ``` 注意事项： 必须使用Xtset panelvar timevar来以通常的长形式声明面板数据集;参考xtset。 RCM自动重塑面板数据集从长到宽的形式，适合于实现回归控制方法。 depvar和indepvars必须是数值变量，不允许使用缩写。 **必选项包括** Trunit(#)在xtset panelvar中指定的面板变量中给定的被处理单元(即受干预影响的单元)的单元号。注意，只能指定单个受影响的单元。 Trperoid(#)干预发生的时间周期。时间周期指的是xtset timevar中指定的时间变量，并且必须是一个整数(参见下面的示例)。注意，只有一个可以指定时间段。 **可选项包括如下** ``` Model ctrlunit(numlist) 表示控制单元用作供体池 preperiod(numlist) 表示受干预前 postperiod(numlist) 表示受干预前 Optimization criterion(sel_criterion) 用于选择最优模型的方法 estimate(est_method) method for estimating post-selection coefficients Placebo Test 使用假治疗单位和/或时间的安慰剂试验 Reporting frame(framename) 创建一个Stata数据框，以宽格式存储数据集和生成的变量 nofigure 不要显示图。默认是显示所有的图。 ``` # 4、Stata操作案例 复制Hsiao et al.(2012) 考虑1997年7月1日回归对香港实际GDP增长率的影响 处理变量时间点:1997Q3 预处理期:1993Q1-1997Q2 后处理期:1997Q3-2004Q1 处理单位:香港 对照组:10个与香港经济有密切联系的国家/地区 ### 4.1案例1 estimating the impact of political integration of Hong Kong with mainland China in 1997q3 (Hsiao et al., 2012) ``` . use growth, clear . xtset region time ``` 结果为 ``` . xtset region time Panel variable: region (strongly balanced) Time variable: time, 1993q1 to 2008q1 Delta: 1 quarter . end of do-file . do "C:\Users\Metrics\AppData\Local\Temp\STD43e8_000000.tmp" . . label list region: 1 Australia 2 Austria 3 Canada 4 China 5 Denmark 6 Finland 7 France 8 Germany 9 HongKong 10 Indonesia 11 Italy 12 Japan 13 Korea 14 Malaysia 15 Mexico 16 Netherlands 17 NewZealand 18 Norway 19 Philippines 20 Singapore 21 Switzerland 22 Taiwan 23 Thailand 24 UnitedKingdom 25 UnitedStates ``` 接下来进行RCM分析 ``` . rcm gdp, trunit(9) trperiod(150) ctrlunit(4 10 12 13 14 19 20 22 23 25) postperiod(150/175) ``` 结果为： ``` . rcm gdp, trunit(9) trperiod(150) ctrlunit(4 10 12 13 14 19 20 22 23 25) postperiod(150/175) Step 1: Select the suboptimal models (method best specified) Note: If this takes too long, you may wish to try method(lasso)(recommended), method(forward) or method(backward). Alternatively, you may restrict indepvars, and/or the donor pool by the option ctrlunit(). Selecting the suboptimal model with number of predictors 1-10... Step 2: Select the optimal model from the suboptimal models (criterion aicc specified) Comparing the suboptimal models containing different set of predictors: ----------------------------------------------------- K | AICc AIC BIC R-squared ----+------------------------------------------------ 1 | -144.7514 -146.4657 -143.7946 0.4034 2 | -160.5063 -163.5832 -160.0217 0.7937 3 | -170.6492 -175.6492 -171.1973 0.9056 4 | -171.7725 -179.4088 -174.0666 0.9314 5 | -169.7878 -180.9878 -174.7552 0.9438 6 | -164.2937 -180.2937 -173.1707 0.9477 7 | -156.6834 -179.1834 -171.1701 0.9503 8 | -146.2921 -177.7207 -168.8169 0.9517 9 | -131.7464 -175.7464 -165.9523 0.9518 10 | -111.3603 -173.7603 -163.0758 0.9518 ----------------------------------------------------- Among models with 1-10 predictors, the optimal model contains 4 predictors with AICc = -171.7725. Fitting results in the pre-treatment periods using OLS: ----------------------------------------------------------------------------------- Mean Absolute Error = 0.00611 Number of Observations = 18 Mean Squared Error = 0.00003 Number of Predictors = 4 Root Mean Squared Error = 0.00578 R-squared = 0.93144 ----------------------------------------------------------------------------------- ----------------------------------------------------------------------------------- gdp路HongKong | Coefficient Std. err. t P>|t| [95% conf. interval] ------------------+---------------------------------------------------------------- gdp路Korea | -0.4323 0.0634 -6.82 0.000 -0.5692 -0.2954 gdp路Japan | -0.6760 0.1117 -6.05 0.000 -0.9172 -0.4347 gdp路Taiwan | 0.7926 0.3099 2.56 0.024 0.1231 1.4621 gdp路UnitedStates | 0.4860 0.2195 2.21 0.045 0.0118 0.9603 _cons | 0.0263 0.0170 1.54 0.147 -0.0105 0.0631 ----------------------------------------------------------------------------------- Prediction results in the post-treatment periods using OLS: --------------------------------------------- Time | Treated Predicted Tr. Effect --------+------------------------------------ 1997q3 | 0.0610 0.0798 -0.0188 1997q4 | 0.0140 0.0810 -0.0670 1998q1 | -0.0320 0.1294 -0.1614 1998q2 | -0.0610 0.1433 -0.2043 1998q3 | -0.0810 0.1319 -0.2129 1998q4 | -0.0650 0.1390 -0.2040 1999q1 | -0.0290 0.0876 -0.1166 1999q2 | 0.0050 0.0670 -0.0620 1999q3 | 0.0390 0.0400 -0.0010 1999q4 | 0.0830 0.0445 0.0385 2000q1 | 0.1070 0.0434 0.0636 2000q2 | 0.0750 0.0398 0.0352 2000q3 | 0.0760 0.0524 0.0236 2000q4 | 0.0630 0.0318 0.0312 2001q1 | 0.0270 0.0118 0.0152 2001q2 | 0.0150 -0.0177 0.0327 2001q3 | -0.0010 -0.0177 0.0167 2001q4 | -0.0170 0.0184 -0.0354 2002q1 | -0.0100 0.0314 -0.0414 2002q2 | 0.0050 0.0500 -0.0450 2002q3 | 0.0280 0.0577 -0.0297 2002q4 | 0.0480 0.0346 0.0134 2003q1 | 0.0410 0.0538 -0.0128 2003q2 | -0.0090 0.0251 -0.0341 2003q3 | 0.0380 0.0628 -0.0248 2003q4 | 0.0470 0.0761 -0.0291 --------+------------------------------------ Mean | 0.0180 0.0576 -0.0396 --------------------------------------------- Note: The average treatment effect over the post-treatment periods is -0.0396. Finished. . end of do-file . ``` ### 4.2Use post-lasso OLS with LOOCV and all control units, and create a Stata frame "growth_wide" storing dataset with generated variables in wide form 代码为 ``` rcm gdp, trunit(9) trperiod(150) postperiod(150/175) method(lasso) criterion(cv) frame(growth_wide) ``` 结果为 ``` . rcm gdp, trunit(9) trperiod(150) postperiod(150/175) method(lasso) criterion(cv) frame(growth_wide) Step 1: Select the suboptimal models (method lasso specified) Selecting the suboptimal model... Step 2: Select the optimal model from the suboptimal models (criterion cv specified for leave-one-out cross-validation) Comparing the suboptimal models containing different set of predictors: ------------------------------------------------------------------------------------------------------------- K | AICc AIC BIC CV MSE R-squared lambda | Operation ----+------------------------------------------------------------------------+------------------------------- 1 | -136.4018 -138.1161 -135.4450 0.0004 0.0513 0.0136 | add gdp路Mexico 2 | -137.2575 -140.3344 -136.7729 0.0003 0.2495 0.0108 | add gdp路Japan 3 | -135.5407 -140.5407 -136.0888 0.0003 0.3361 0.0098 | add gdp路Finland 4 | -137.3309 -144.9673 -139.6250 0.0003 0.5354 0.0074 | add gdp路Korea 6 | -129.9034 -145.9034 -138.7804 0.0002 0.6468 0.0062 | add gdp路Australia gdp路Italy 5 | -138.0419 -149.2419 -143.0093 0.0002 0.6722 0.0059 | drop gdp路Finland 6 | -143.4155 -159.4155 -152.2925 0.0002 0.8333 0.0032 | add gdp路Taiwan 5 | -153.4700 -164.6700 -158.4374 0.0002 0.8609 0.0028 | drop gdp路Italy 6 | -147.6414 -163.6414 -156.5184 0.0001 0.8682 0.0027 | add gdp路Malaysia 7 | -153.9967 -176.4967 -168.4834 0.0001 0.9422 0.0011 | add gdp路UnitedStates 8 | -145.5446 -176.9732 -168.0695 0.0001 0.9497 0.0008 | add gdp路Canada 7 | -157.4199 -179.9199 -171.9065 0.0001 0.9522 0.0007 | drop gdp路UnitedStates 8 | -146.7505 -178.1790 -169.2753 0.0001 0.9529 0.0007 | add gdp路Philippines 9 | -133.3717 -177.3717 -167.5776 0.0001 0.9559 0.0006 | add gdp路UnitedStates 11 | -84.8554 -175.8554 -164.2805 0.0001 0.9616 0.0004 | add gdp路France gdp路Germany 12 | -35.5339 -175.5339 -163.0687 0.0001 0.9650 0.0004 | add gdp路Switzerland 11 | -92.2461 -183.2461 -171.6713 0.0001 0.9745 0.0002 | drop gdp路UnitedStates 13 | 59.2679 -180.7321 -167.3766 0.0001 0.9766 0.0002 | add gdp路China gdp路UnitedStates 12 | -44.0388 -184.0388 -171.5736 0.0001 0.9782 0.0002 | drop gdp路Canada 13 | 56.8234 -183.1766 -169.8211 0.0001 0.9795 0.0002 | add gdp路UnitedKingdom 12 | -47.6918 -187.6918 -175.2266 0.0001 0.9822 0.0002 | drop gdp路UnitedStates 12 | -54.1876 -194.1876 -181.7224 0.0001 0.9876 0.0001 | . ------------------------------------------------------------------------------------------------------------- Among models with 1-24 predictors, the optimal model contains 12 predictors with CV MSE = 0.0001. Fitting results in the pre-treatment periods using post-lasso OLS: ------------------------------------------------------------------------------------ Mean Absolute Error = 0.00404 Number of Observations = 18 Mean Squared Error = 0.00001 Number of Predictors = 12 Root Mean Squared Error = 0.00256 R-squared = 0.99481 ------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------ gdp路HongKong | Coefficient Std. err. t P>|t| [95% conf. interval] -------------------+---------------------------------------------------------------- gdp路Australia | 0.0293 0.0988 0.30 0.779 -0.2247 0.2833 gdp路China | 0.3318 0.1115 2.98 0.031 0.0452 0.6184 gdp路France | 0.4306 0.1858 2.32 0.068 -0.0469 0.9081 gdp路Germany | 0.5107 0.1917 2.66 0.045 0.0180 1.0033 gdp路Japan | -0.8833 0.1007 -8.77 0.000 -1.1421 -0.6244 gdp路Korea | -0.6836 0.0753 -9.07 0.000 -0.8773 -0.4900 gdp路Malaysia | 0.0400 0.0481 0.83 0.443 -0.0836 0.1636 gdp路Mexico | 0.0667 0.0489 1.36 0.231 -0.0591 0.1925 gdp路Philippines | -0.6231 0.1339 -4.65 0.006 -0.9674 -0.2789 gdp路Switzerland | 0.1001 0.1098 0.91 0.404 -0.1822 0.3824 gdp路Taiwan | -0.4112 0.4313 -0.95 0.384 -1.5198 0.6974 gdp路UnitedKingdom | 0.8364 0.2854 2.93 0.033 0.1027 1.5701 _cons | 0.0881 0.0220 4.01 0.010 0.0317 0.1446 ------------------------------------------------------------------------------------ Prediction results in the post-treatment periods using post-lasso OLS: --------------------------------------------- Time | Treated Predicted Tr. Effect --------+------------------------------------ 1997q3 | 0.0610 0.0896 -0.0286 1997q4 | 0.0140 0.0929 -0.0789 1998q1 | -0.0320 0.2110 -0.2430 1998q2 | -0.0610 0.2397 -0.3007 1998q3 | -0.0810 0.2472 -0.3282 1998q4 | -0.0650 0.2572 -0.3222 1999q1 | -0.0290 0.1389 -0.1679 1999q2 | 0.0050 0.0670 -0.0620 1999q3 | 0.0390 0.0515 -0.0125 1999q4 | 0.0830 0.0395 0.0435 2000q1 | 0.1070 0.0063 0.1007 2000q2 | 0.0750 0.0448 0.0302 2000q3 | 0.0760 0.0377 0.0383 2000q4 | 0.0630 0.0571 0.0059 2001q1 | 0.0270 0.1183 -0.0913 2001q2 | 0.0150 0.1144 -0.0994 2001q3 | -0.0010 0.1302 -0.1312 2001q4 | -0.0170 0.1325 -0.1495 2002q1 | -0.0100 0.0814 -0.0914 2002q2 | 0.0050 0.0722 -0.0672 2002q3 | 0.0280 0.0872 -0.0592 2002q4 | 0.0480 0.0271 0.0209 2003q1 | 0.0410 0.1020 -0.0610 2003q2 | -0.0090 0.1192 -0.1282 2003q3 | 0.0380 0.0829 -0.0449 2003q4 | 0.0470 0.0950 -0.0480 --------+------------------------------------ Mean | 0.0180 0.1055 -0.0875 --------------------------------------------- Note: The average treatment effect over the post-treatment periods is -0.0875. Finished. . ``` ### 4.3 Implement a placebo test using all fake treatment units in the donor pool ``` . rcm gdp, trunit(9) trperiod(150) postperiod(150/175) method(lasso) criterion(cv) placebo(unit) ``` ``` . rcm gdp, trunit(9) trperiod(150) postperiod(150/175) method(lasso) criterion(cv) placebo(unit) Step 1: Select the suboptimal models (method lasso specified) Selecting the suboptimal model... Step 2: Select the optimal model from the suboptimal models (criterion cv specified for leave-one-out cross-validation) Comparing the suboptimal models containing different set of predictors: ------------------------------------------------------------------------------------------------------------- K | AICc AIC BIC CV MSE R-squared lambda | Operation ----+------------------------------------------------------------------------+------------------------------- 1 | -136.4018 -138.1161 -135.4450 0.0004 0.0513 0.0136 | add gdp路Mexico 2 | -137.2575 -140.3344 -136.7729 0.0003 0.2495 0.0108 | add gdp路Japan 3 | -135.5407 -140.5407 -136.0888 0.0003 0.3361 0.0098 | add gdp路Finland 4 | -137.3309 -144.9673 -139.6250 0.0003 0.5354 0.0074 | add gdp路Korea 6 | -129.9034 -145.9034 -138.7804 0.0002 0.6468 0.0062 | add gdp路Australia gdp路Italy 5 | -138.0419 -149.2419 -143.0093 0.0002 0.6722 0.0059 | drop gdp路Finland 6 | -143.4155 -159.4155 -152.2925 0.0002 0.8333 0.0032 | add gdp路Taiwan 5 | -153.4700 -164.6700 -158.4374 0.0002 0.8609 0.0028 | drop gdp路Italy 6 | -147.6414 -163.6414 -156.5184 0.0001 0.8682 0.0027 | add gdp路Malaysia 7 | -153.9967 -176.4967 -168.4834 0.0001 0.9422 0.0011 | add gdp路UnitedStates 8 | -145.5446 -176.9732 -168.0695 0.0001 0.9497 0.0008 | add gdp路Canada 7 | -157.4199 -179.9199 -171.9065 0.0001 0.9522 0.0007 | drop gdp路UnitedStates 8 | -146.7505 -178.1790 -169.2753 0.0001 0.9529 0.0007 | add gdp路Philippines 9 | -133.3717 -177.3717 -167.5776 0.0001 0.9559 0.0006 | add gdp路UnitedStates 11 | -84.8554 -175.8554 -164.2805 0.0001 0.9616 0.0004 | add gdp路France gdp路Germany 12 | -35.5339 -175.5339 -163.0687 0.0001 0.9650 0.0004 | add gdp路Switzerland 11 | -92.2461 -183.2461 -171.6713 0.0001 0.9745 0.0002 | drop gdp路UnitedStates 13 | 59.2679 -180.7321 -167.3766 0.0001 0.9766 0.0002 | add gdp路China gdp路UnitedStates 12 | -44.0388 -184.0388 -171.5736 0.0001 0.9782 0.0002 | drop gdp路Canada 13 | 56.8234 -183.1766 -169.8211 0.0001 0.9795 0.0002 | add gdp路UnitedKingdom 12 | -47.6918 -187.6918 -175.2266 0.0001 0.9822 0.0002 | drop gdp路UnitedStates 12 | -54.1876 -194.1876 -181.7224 0.0001 0.9876 0.0001 | . ------------------------------------------------------------------------------------------------------------- Among models with 1-24 predictors, the optimal model contains 12 predictors with CV MSE = 0.0001. Fitting results in the pre-treatment periods using post-lasso OLS: ------------------------------------------------------------------------------------ Mean Absolute Error = 0.00404 Number of Observations = 18 Mean Squared Error = 0.00001 Number of Predictors = 12 Root Mean Squared Error = 0.00256 R-squared = 0.99481 ------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------ gdp路HongKong | Coefficient Std. err. t P>|t| [95% conf. interval] -------------------+---------------------------------------------------------------- gdp路Australia | 0.0293 0.0988 0.30 0.779 -0.2247 0.2833 gdp路China | 0.3318 0.1115 2.98 0.031 0.0452 0.6184 gdp路France | 0.4306 0.1858 2.32 0.068 -0.0469 0.9081 gdp路Germany | 0.5107 0.1917 2.66 0.045 0.0180 1.0033 gdp路Japan | -0.8833 0.1007 -8.77 0.000 -1.1421 -0.6244 gdp路Korea | -0.6836 0.0753 -9.07 0.000 -0.8773 -0.4900 gdp路Malaysia | 0.0400 0.0481 0.83 0.443 -0.0836 0.1636 gdp路Mexico | 0.0667 0.0489 1.36 0.231 -0.0591 0.1925 gdp路Philippines | -0.6231 0.1339 -4.65 0.006 -0.9674 -0.2789 gdp路Switzerland | 0.1001 0.1098 0.91 0.404 -0.1822 0.3824 gdp路Taiwan | -0.4112 0.4313 -0.95 0.384 -1.5198 0.6974 gdp路UnitedKingdom | 0.8364 0.2854 2.93 0.033 0.1027 1.5701 _cons | 0.0881 0.0220 4.01 0.010 0.0317 0.1446 ------------------------------------------------------------------------------------ Prediction results in the post-treatment periods using post-lasso OLS: --------------------------------------------- Time | Treated Predicted Tr. Effect --------+------------------------------------ 1997q3 | 0.0610 0.0896 -0.0286 1997q4 | 0.0140 0.0929 -0.0789 1998q1 | -0.0320 0.2110 -0.2430 1998q2 | -0.0610 0.2397 -0.3007 1998q3 | -0.0810 0.2472 -0.3282 1998q4 | -0.0650 0.2572 -0.3222 1999q1 | -0.0290 0.1389 -0.1679 1999q2 | 0.0050 0.0670 -0.0620 1999q3 | 0.0390 0.0515 -0.0125 1999q4 | 0.0830 0.0395 0.0435 2000q1 | 0.1070 0.0063 0.1007 2000q2 | 0.0750 0.0448 0.0302 2000q3 | 0.0760 0.0377 0.0383 2000q4 | 0.0630 0.0571 0.0059 2001q1 | 0.0270 0.1183 -0.0913 2001q2 | 0.0150 0.1144 -0.0994 2001q3 | -0.0010 0.1302 -0.1312 2001q4 | -0.0170 0.1325 -0.1495 2002q1 | -0.0100 0.0814 -0.0914 2002q2 | 0.0050 0.0722 -0.0672 2002q3 | 0.0280 0.0872 -0.0592 2002q4 | 0.0480 0.0271 0.0209 2003q1 | 0.0410 0.1020 -0.0610 2003q2 | -0.0090 0.1192 -0.1282 2003q3 | 0.0380 0.0829 -0.0449 2003q4 | 0.0470 0.0950 -0.0480 --------+------------------------------------ Mean | 0.0180 0.1055 -0.0875 --------------------------------------------- Note: The average treatment effect over the post-treatment periods is -0.0875. Implementing placebo effects using fake treatment unit Australia...Austria...Canada...China...Denmark...Finland...France...Ge > rmany...Indonesia...Italy...Japan...Korea...Malaysia...Mexico...Netherlands...NewZealand...Norway...Philippines...Singapore > ...Switzerland...Taiwan...Thailand...UnitedKingdom...UnitedStates... Placebo test results using fake treatment units: -------------------------------------------------------------------------------------------------- Unit | Pre MSPE Post MSPE Post/Pre MSPE Pre MSPE of Unit/Treated Unit ---------------+---------------------------------------------------------------------------------- HongKong | 0.0000 0.0198 3009.7824 1.0000 Australia | 0.0000 0.0008 18.5015 6.4202 Austria | 0.0000 0.0004 28.6729 2.0841 Canada | 0.0000 0.0020 1039.6208 0.2995 China | 0.0000 0.0016 82.3533 2.8889 Denmark | 0.0000 0.0012 43.5217 4.3274 Finland | 0.0000 0.0039 207.2381 2.8796 France | 0.0000 0.0006 73.2262 1.1454 Germany | 0.0000 0.0003 16.0078 2.7504 Indonesia | 0.0000 0.0092 284.5697 4.9035 Italy | 0.0000 0.0014 228.5263 0.9062 Japan | 0.0000 0.0054 183.8924 4.4759 Korea | 0.0000 0.0068 3721.7026 0.2798 Malaysia | 0.0007 0.0060 9.0531 100.4546 Mexico | 0.0001 0.0032 30.5062 15.8926 Netherlands | 0.0000 0.0001 4.8597 4.3291 NewZealand | 0.0001 0.0032 26.6424 18.1849 Norway | 0.0001 0.0078 55.8595 21.2922 Philippines | 0.0000 0.0198 3757.4421 0.8045 Singapore | 0.0001 0.0072 137.8805 7.9428 Switzerland | 0.0000 0.0013 58.8820 3.3993 Taiwan | 0.0000 0.0011 2852.1634 0.0595 Thailand | 0.0002 0.0060 39.4678 23.0439 UnitedKingdom | 0.0000 0.0015 161.0820 1.4022 UnitedStates | 0.0000 0.0001 285.1185 0.0517 -------------------------------------------------------------------------------------------------- Note: The probability of obtaining a post/pre-treatment MSPE ratio as large as HongKong's is 0.1200. Placebo test results using fake treatment units (continued): --------------------------------- Time | Tr. Eff. P-value --------+------------------------ 1997q3 | -0.0286 0.1667 1997q4 | -0.0789 0.0000 1998q1 | -0.2430 0.0000 1998q2 | -0.3007 0.0000 1998q3 | -0.3282 0.0000 1998q4 | -0.3222 0.0000 1999q1 | -0.1679 0.0417 1999q2 | -0.0620 0.2083 1999q3 | -0.0125 0.8333 1999q4 | 0.0435 0.1250 2000q1 | 0.1007 0.0000 2000q2 | 0.0302 0.2917 2000q3 | 0.0383 0.1667 2000q4 | 0.0059 0.8333 2001q1 | -0.0913 0.1250 2001q2 | -0.0994 0.1667 2001q3 | -0.1312 0.0833 2001q4 | -0.1495 0.1250 2002q1 | -0.0914 0.2083 2002q2 | -0.0672 0.1667 2002q3 | -0.0592 0.1250 2002q4 | 0.0209 0.3750 2003q1 | -0.0610 0.1250 2003q2 | -0.1282 0.0417 2003q3 | -0.0449 0.1250 2003q4 | -0.0480 0.0833 --------------------------------- Note: The p-value of the treatment effect for a particular period is definded as the frequency that the absolute values of the placebo effects are greater or equal to the absolute value of treatment effect. Finished. . ```