DID中diff相关问题

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

求助大佬，为什么此时控制组在政策实施后的个体为0？是因为在此之前我所设定的treat和post变量有误吗？检查了好几遍也没有发现错误，想了一个多小时了

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Diff DID DIF IFF treat

你只给一个DID的结果，没人能帮你检查出问题。你得把具体的研究问题描述出来，数据给出来。帮你查命令是不是有问题，你得把你的命令写出来啊。

除了命令，可能还要看看你的数据（至少一小部分代表性数据），还有变量treat和post你是怎么定义的

很有可能是treat post设定有问题

1. cd "D:\stata data\Course paper"
   
2. 
   
3. import excel "Annual data of 30 provinces (excluding Inner Mongolia)" ,clear sheet(Sheet1) firstrow
   
4. destring , replace
   
5. save 30 , replace //将第一张Excel转换为dta数据格式
   
6. 
   
7. import excel "30个省成为自贸区时间" ,clear sheet(Sheet1) firstrow
   
8. destring , replace
   
9. merge 1:m region using "30.dta"
   
10. sort region year
    
11. keep if _merge == 3
    
12. count if _merge == 3
    
13. drop _merge
    
14. save 30+21 , replace //合并两张Excel表并转换为dta数据格式
    
15. 
    
16. cd "D:\stata data\Course paper"
    
17. use "30+21.dta" , clear //调用数据
    
18. des2 //查看数据类型等信息
    
19. 
    
20. generate treat = 0
    
21. replace treat = 1 if time != 0
    
22. 
    
23. generate post = 0
    
24. replace post = 1 if (year >= time) * (time != 0)
    
25. 
    
26. generate did = treat * post
    
27. 
    
28. encode region, gen(Region)
    
29. 
    
30. order Region region year time treat post did
    
31. 
    
32. fsum, s(mean sd p50 min max)  //描述性统计
    
33. 
    
34. local vars "GDP first_in second_in third_in permanent_population enterprise_number total_import_export  road consume"
    
35. foreach v of varlist `vars'{
    
36.         generate ln_`v' = ln(`v')
    
37. }
    
38. 
    
39. fsum ln_* , s(mean sd p50 min max)  //描述性统计
    
40. 
    
41. global c1 "first_in enterprise_number total_import_export road employment_urban "
    
42. 
    
43. global c2 "ln_first_in ln_second_in ln_third_in ln_permanent_population ln_enterprise_number ln_total_import_export ln_road ln_consume employment_urban fiscal_expenditure rail fiscal_revenue CPI PPI"
    
44. 
    
45. *描述性统计表格
    
46. help sum2docx
    
47. sum GDP  first_in second_in third_in permanent_population employment_urban fiscal_expenditure enterprise_number total_import_export road rail fiscal_revenue consume CPI PPI
    
48. sum2docx GDP did first_in second_in third_in permanent_population employment_urban fiscal_expenditure enterprise_number total_import_export road rail fiscal_revenue consume CPI PPI using sum.docx, replace stats(N mean(%9.0f) sd(%9.0g) min(%9.0g) median(%9.0g) max(%9.0g)) title("Descriptive statistics")
    
49. 
    
50. *相关性分析表格
    
51. help corr2docx
    
52. corr2docx GDP  first_in second_in third_in permanent_population employment_urban fiscal_expenditure enterprise_number total_import_export road rail fiscal_revenue consume CPI PPI  using corr.docx, replace star   spearman(ignore)
    
53. 
    
54. *回归表格
    
55. help reg2docx
    
56. reg GDP did
    
57. est        store r1
    
58. reg GDP did first_in
    
59. est        store r2
    
60. reg GDP did first_in enterprise_number
    
61. est        store r3
    
62. reg GDP did first_in enterprise_number total_import_export
    
63. est        store r4
    
64. reg GDP did first_in enterprise_number total_import_export road
    
65. est        store r5
    
66. reg GDP did first_in enterprise_number total_import_export road employment_urban
    
67. est        store r6
    
68. *产业结构、规模企业、对外贸易、交通运输、人力资源
    
69. reg2docx r1 r2 r3 r4 r5 r6  using reg1.docx, replace scalars(N r2(%9.3f) r2_a(%9.2f))  b(%9.2f) t(%7.2f) title(table1: OLS regression results)  noconstant
    
70. 
    
71. *两期DID分析
    
72. reg GDP did treat post , r
    
73. reg GDP did treat post $c1, r
    
74. xtset Region year
    
75. xtreg GDP did treat post $c1 , fe r
    
76. reghdfe GDP did treat post $c1 , absorb(i.year Region)
    
77. diff GDP , t(treat) p(post) cov($c1)
    
78. 
    
79. *多期DID分析
    
80. help outreg2
    
81. 
    
82. reg GDP did $c1 i.Region i.year  ,vce(cluster Region)
    
83. outreg2  using reg2,word replace stats(coef tstat)   adjr2 addtext("Year", "YES", "Region","YES" , "Cluster","YES" )  keep(did $c1)
    
84. *在Region和year层面进行固定效应分析，在Region层面进行聚类稳健分析
    
85. xtset Region year
    
86. xtreg GDP did  $c1 i.year , fe vce(cluster Region)
    
87. outreg2  using reg2,word append  stats(coef tstat) addstat("F-value", e(F) ) adjr2 addtext("Year", "YES", "Region","NO" , "Cluster","YES") keep(did $c1)
    
88. *xtreg相比reg而言，该回归命令可以估计出更多参数
    
89. reghdfe GDP did $c1 , absorb( Region year ) vce(cluster Region)
    
90. outreg2  using reg2,word append  stats(coef tstat) addstat("F-value", e(F) ) adjr2 addtext("Year", "YES", "Region","YES" , "Cluster","YES") keep(did $c1)

复制代码

代码如上，不知道如何展现数据，部分数据为