1
00:00:00,000 --> 00:00:02,550
♪ [音乐] ♪

2
00:00:03,800 --> 00:00:05,800
- [旁白] 欢迎收看“诺奖得主畅谈系列”

3
00:00:07,040 --> 00:00:08,100
在这一集中

4
00:00:08,100 --> 00:00:11,570
Josh Angrist 和 Guido Imbens
与 Isaiah Andrews

5
00:00:11,570 --> 00:00:14,600
将讨论计量经济学领域是如何发展的

6
00:00:16,100 --> 00:00:18,750
- [Isaiah] 那么，Guido 和 Josh

7
00:00:18,750 --> 00:00:21,500
你们都是开发经济学实证研究工具的先驱

8
00:00:21,500 --> 00:00:23,174
所以我想了解你对这个领域的发展方向的想法

9
00:00:23,174 --> 00:00:25,300


10
00:00:25,709 --> 00:00:28,079
经济学、计量经济学，整个领域

11
00:00:28,510 --> 00:00:31,302
首先，我很想听听你是否觉得

12
00:00:32,171 --> 00:00:35,200
局部平均处理效应框架的方式

13
00:00:35,200 --> 00:00:38,510


14
00:00:38,800 --> 00:00:42,187
对经济学中的新经验法如何发展和传播

15
00:00:42,187 --> 00:00:44,300
或它们应如何发展和传播有任何榜样

16
00:00:44,560 --> 00:00:45,960
- [Josh] 这是个好问题

17
00:00:46,610 --> 00:00:47,790
你先说吧

18
00:00:47,790 --> 00:00:49,240
[笑声]

19
00:00:49,700 --> 00:00:52,940
- [Guido] 是的，所以我认为重要的是

20
00:00:52,940 --> 00:00:58,550
要提出令人信服的案例

21
00:00:58,550 --> 00:01:02,207
其中问题清晰

22
00:01:02,400 --> 00:01:05,720
且方法普遍适用

23
00:01:06,253 --> 00:01:07,560
有一件事我...

24
00:01:08,192 --> 00:01:12,000
当我回顾较新的文献时

25
00:01:12,200 --> 00:01:16,700
我真的很喜欢回归不连续性文献

26
00:01:16,700 --> 00:01:19,670
其中显然有一堆非常有说服力的例子

27
00:01:19,670 --> 00:01:23,378
这让人们可以更清晰地思考

28
00:01:23,378 --> 00:01:27,200
更仔细地研究方法问题

29
00:01:27,400 --> 00:01:28,800
有着一些清晰的应用方法

30
00:01:28,800 --> 00:01:30,600
然后允许你思考：

31
00:01:30,600 --> 00:01:33,600
“哇，这些类型的假设在这里看起来合理吗？

32
00:01:33,600 --> 00:01:38,000
我们不喜欢早期论文中的哪些方面？

33
00:01:38,500 --> 00:01:39,802
我们如何改进这些方面？”

34
00:01:39,802 --> 00:01:44,210
因此，我认为有着明确的应用方法

35
00:01:44,210 --> 00:01:46,400
来激发这些文献是非常有帮助的

36
00:01:46,800 --> 00:01:48,050
- Guido，我很高兴
你提到了回归不连续性

37
00:01:48,050 --> 00:01:49,382


38
00:01:49,382 --> 00:01:53,300
我认为 IV 和 RD、

39
00:01:54,700 --> 00:01:57,060
工具变量和回归不连续性之间有很多互补性

40
00:02:00,506 --> 00:02:03,260
回归不连续性的许多计量经济学应用
曾经被称为“模糊” RD

41
00:02:03,260 --> 00:02:04,520


42
00:02:04,520 --> 00:02:07,230


43
00:02:07,230 --> 00:02:11,620
它在截止时不是离散的或确定的

44
00:02:11,620 --> 00:02:14,900
而只在速率或强度上有着变化

45
00:02:14,900 --> 00:02:17,737
LATE 框架帮助我们理解这些应用方法

46
00:02:17,737 --> 00:02:18,740


47
00:02:18,740 --> 00:02:21,140
并为我们提供了一个清晰的解释

48
00:02:21,140 --> 00:02:25,000
比如在我与 Victor Lavy 的论文中

49
00:02:25,000 --> 00:02:28,100
我们使用了 Maimonides 规则
班级规模截断

50
00:02:28,430 --> 00:02:30,030
那么你这里了解到了什么？

51
00:02:30,290 --> 00:02:31,820
当然，你可以用线性常数效应模型
来回答这个问题

52
00:02:31,820 --> 00:02:33,900


53
00:02:34,200 --> 00:02:36,310
但事实证明我们并不局限于此

54
00:02:36,310 --> 00:02:39,889
RD 仍然非常强大和有启发性

55
00:02:40,630 --> 00:02:43,092
即使在这种情况和类型规模下

56
00:02:43,092 --> 00:02:45,866
截止和感兴趣的变量之间的相关性是局部的

57
00:02:45,866 --> 00:02:49,133


58
00:02:49,133 --> 00:02:51,000
甚至可能不是那么强

59
00:02:52,000 --> 00:02:54,999
所以肯定有着平行发展

60
00:02:54,999 --> 00:02:56,400
这也很有趣...

61
00:02:57,253 --> 00:02:59,780
当我们在读研究生时，没有人谈论回归不连续性设计

62
00:02:59,780 --> 00:03:01,220


63
00:03:01,220 --> 00:03:02,843
这是其他社会科学家感兴趣的东西

64
00:03:02,843 --> 00:03:05,300


65
00:03:05,800 --> 00:03:09,507
它与 LATE 框架一起成长

66
00:03:09,507 --> 00:03:11,927
我们都曾基于这两种应用方式和方法进行工作

67
00:03:11,927 --> 00:03:14,565


68
00:03:14,565 --> 00:03:18,377
而且能看到它的发展并变得如此重要

69
00:03:18,377 --> 00:03:19,800
我感到非常兴奋

70
00:03:20,000 --> 00:03:21,767
我认为，这是朝着可靠的识别策略

71
00:03:21,767 --> 00:03:26,086
因果效应的普遍演变的一部分

72
00:03:26,086 --> 00:03:27,441


73
00:03:29,393 --> 00:03:30,642
使计量经济学更多地关注
因果问题而不是模型

74
00:03:30,642 --> 00:03:33,300


75
00:03:33,640 --> 00:03:34,650
就未来而言

76
00:03:34,650 --> 00:03:37,660
我认为 LATE 帮助促进的一件事是

77
00:03:37,660 --> 00:03:42,008
朝着更具创造性的随机试验迈进

78
00:03:42,008 --> 00:03:44,400
其中有着一些有趣的东西

79
00:03:45,500 --> 00:03:48,460
它不可能或直接能被简单地关闭或开启

80
00:03:48,460 --> 00:03:50,700


81
00:03:51,000 --> 00:03:54,584
但你可以鼓励或阻止它

82
00:03:54,584 --> 00:03:58,200
因此，例如，你能通过经济援助
来补贴学校教育

83
00:03:59,000 --> 00:04:02,080
所以现在我们有了
一个完整的框架来解释这一点

84
00:04:03,600 --> 00:04:07,113
它打开了对以前似乎不可能的事情
进行随机试验的大门

85
00:04:07,113 --> 00:04:09,265


86
00:04:10,300 --> 00:04:12,471


87
00:04:14,500 --> 00:04:17,864
我们在麻省理工学院
Blueprint实验室中

88
00:04:17,864 --> 00:04:21,160
大量使用了这一点

89
00:04:22,360 --> 00:04:26,600
我认为，我们正在以
非常有创意的方式利用随机分配

90
00:04:28,100 --> 00:04:31,395
- [Isaiah] 与此相关，你是否看到

91
00:04:31,395 --> 00:04:34,445
有助于计量经济学研究的特定因素？

92
00:04:34,445 --> 00:04:38,443
你已经提到它与
实际出现的问题有明确的联系

93
00:04:38,443 --> 00:04:40,300


94
00:04:40,300 --> 00:04:42,862
并且经验实践通常是一个好主意

95
00:04:43,290 --> 00:04:45,000
- 这不是一个好主意吗？

96
00:04:45,700 --> 00:04:50,112
我经常发现自己坐在
一个计量经济学理论研讨会上

97
00:04:50,700 --> 00:04:52,500
比如哈佛麻省理工学院的研讨会

98
00:04:53,400 --> 00:04:56,350
我会想，“这个人在解决什么问题？

99
00:04:56,350 --> 00:04:57,960
谁面对这个问题？”

100
00:04:57,960 --> 00:04:59,800
而且，你知道…

101
00:05:01,600 --> 00:05:04,700
如果我问的话
有时会出现令人尴尬的沉默

102
00:05:04,900 --> 00:05:08,300
或者可能会有一个相当牵强的场景

103
00:05:08,800 --> 00:05:11,600
我想看看这个工具在哪里有用

104
00:05:12,500 --> 00:05:14,765
有一些是纯粹的基础工具

105
00:05:14,765 --> 00:05:16,250
我确实明白这一点

106
00:05:16,250 --> 00:05:21,735
有些人正在研究的概念基础

107
00:05:22,600 --> 00:05:25,300
它变得更像数理统计

108
00:05:25,800 --> 00:05:27,653
我的意思是，我记得一个较早的例子

109
00:05:27,653 --> 00:05:29,920
那就是我很难理解的随机等连续性的概念

110
00:05:29,920 --> 00:05:32,500


111
00:05:32,500 --> 00:05:35,070
我的论文顾问之一，Whitney Newey

112
00:05:35,070 --> 00:05:36,479
使用它取得了很好的效果

113
00:05:36,479 --> 00:05:38,821
我当时试图理解这一点

114
00:05:40,600 --> 00:05:42,034
这真的很基础

115
00:05:42,034 --> 00:05:45,200
驱动它的不是一个应用方法

116
00:05:45,890 --> 00:05:47,300
至少不是立即的

117
00:05:48,600 --> 00:05:53,200
但是大多数事情不是这样的
所以应该有着问题

118
00:05:53,800 --> 00:05:59,247
我认为这取决于这类事情的卖家

119
00:06:00,480 --> 00:06:02,250
因为有着机会成本、时间和注意力

120
00:06:02,250 --> 00:06:05,295
以及理解事情的努力

121
00:06:05,980 --> 00:06:07,200
卖家要说：“嘿，我正在解决这个问题，这里有一组结果表明它很有用，这是我得到的一些见解。”

122
00:06:07,400 --> 00:06:08,900


123
00:06:09,400 --> 00:06:12,900


124
00:06:12,900 --> 00:06:15,200


125
00:06:16,097 --> 00:06:18,280
- [以赛亚]正如你所说，乔希，已经朝着更多地考虑经济学中的因果关系和经济学中的实证工作的方向发展。

126
00:06:18,280 --> 00:06:20,700


127
00:06:20,700 --> 00:06:22,900


128
00:06:22,900 --> 00:06:26,570
这种观点的传播有什么让你感到惊讶的后果，或者你认为是经验经济学发展方式的不利因素吗？

129
00:06:26,570 --> 00:06:28,347


130
00:06:28,705 --> 00:06:31,500


131
00:06:31,500 --> 00:06:34,322
- 有时我看到有人做静脉注射，他们得到的结果似乎难以置信地大——这是通常的情况。

132
00:06:34,322 --> 00:06:38,304


133
00:06:38,800 --> 00:06:40,200


134
00:06:42,500 --> 00:06:45,220
因此，这可能是一些相对较小的干预造成的非常大的因果影响，这些干预是随机的，或者您可以证明存在一个好的设计。

135
00:06:45,220 --> 00:06:48,600


136
00:06:49,100 --> 00:06:52,260


137
00:06:52,260 --> 00:06:54,490


138
00:06:54,900 --> 00:06:57,205
然后当我看到这一点时，我认为我很难相信这种相对较小的干预会产生如此大的影响。

139
00:06:57,944 --> 00:07:00,101


140
00:07:00,101 --> 00:07:02,030


141
00:07:02,030 --> 00:07:03,720


142
00:07:04,100 --> 00:07:06,277
作者有时会诉诸局部平均治疗效果定理并说，“嗯，这些编译器——他们在某些方面很特别，他们只是从这种干预中受益匪浅。”

143
00:07:06,277 --> 00:07:08,690


144
00:07:08,690 --> 00:07:11,066


145
00:07:11,066 --> 00:07:12,700


146
00:07:13,300 --> 00:07:15,800


147
00:07:15,800 --> 00:07:17,600


148
00:07:18,100 --> 00:07:21,175
我不愿意从表面上看。

149
00:07:21,175 --> 00:07:23,750
我经常认为，当影响太大时，这是因为排除限制失败了，所以你真的没有正确的内生变量来衡量这个结果。

150
00:07:24,300 --> 00:07:26,780


151
00:07:26,780 --> 00:07:29,456


152
00:07:29,456 --> 00:07:31,380


153
00:07:32,000 --> 00:07:35,700
因此，我不太高兴看到一个通用的异质性论点被用来为我认为可能是一个更深层次的问题的事情辩解。

154
00:07:36,937 --> 00:07:39,692


155
00:07:40,022 --> 00:07:41,760


156
00:07:41,760 --> 00:07:43,800


157
00:07:45,190 --> 00:07:47,358
- [Guido] 我认为它在简化形式和结构方法之间的讨论中扮演了一个不幸的角色，我觉得这不太正确。

158
00:07:47,358 --> 00:07:50,083


159
00:07:50,083 --> 00:07:51,700


160
00:07:51,700 --> 00:07:55,510


161
00:07:56,090 --> 00:07:58,810
工具变量假设是关于行为的核心结构假设——它们来自经济……考虑代理人的经济行为，不知何故它被推向了一个我认为不是很有帮助的方向。

162
00:07:58,810 --> 00:08:01,510


163
00:08:01,510 --> 00:08:03,483


164
00:08:03,483 --> 00:08:05,200


165
00:08:07,100 --> 00:08:09,900


166
00:08:10,300 --> 00:08:15,100


167
00:08:15,100 --> 00:08:17,600


168
00:08:20,426 --> 00:08:21,663
我认为，最初，我们把事情写下来，描述正在发生的事情。

169
00:08:22,800 --> 00:08:24,067


170
00:08:24,067 --> 00:08:26,480


171
00:08:26,480 --> 00:08:29,783
人们使用了一套方法。

172
00:08:29,783 --> 00:08:32,111
我们以一种我认为包含相当多洞察力的方式阐明了这些方法的作用。

173
00:08:32,811 --> 00:08:38,361


174
00:08:39,100 --> 00:08:42,050
但它不知何故被推到了一个我认为不一定很有帮助的角落。

175
00:08:42,050 --> 00:08:45,379


176
00:08:45,379 --> 00:08:48,604
- 在简化形式与结构的语言中，我觉得有点有趣，因为局部平均治疗效果模型，潜在结果模型是非参数结构模型，如果你想像你建议的那样考虑它，Guido .

177
00:08:48,604 --> 00:08:50,306


178
00:08:50,306 --> 00:08:52,985


179
00:08:52,985 --> 00:08:54,154


180
00:08:54,154 --> 00:08:56,110


181
00:08:56,110 --> 00:08:58,600


182
00:08:58,600 --> 00:09:01,129
所以当——

183
00:09:01,129 --> 00:09:03,505


184
00:09:03,505 --> 00:09:05,116
- [圭多] 是的。
- [Josh] 当然，这种语言来自我们继承的联立方程框架。

185
00:09:05,116 --> 00:09:08,371


186
00:09:08,371 --> 00:09:09,641


187
00:09:10,400 --> 00:09:11,440
它的优点是人们在你使用它时似乎知道你的意思，但这可能是不同的人听到不同的东西。

188
00:09:11,440 --> 00:09:14,085


189
00:09:14,085 --> 00:09:15,164


190
00:09:16,559 --> 00:09:18,300


191
00:09:18,300 --> 00:09:20,480
- [圭多] 是的。我认为简化形式已被用于贬义...

192
00:09:20,480 --> 00:09:22,200


193
00:09:22,200 --> 00:09:23,540
- [乔希] 有时。 ...这并不是它最初的目的。

194
00:09:25,104 --> 00:09:28,250


195
00:09:30,100 --> 00:09:33,090
- [以赛亚] 我想在思考局部平均治疗效果框架的影响时，还有一点让我印象深刻的是，人们通常会在远远超出任何形式的正式结果实际存在的环境中诉诸局部平均治疗效果直觉已确立的。

196
00:09:33,090 --> 00:09:35,645


197
00:09:35,645 --> 00:09:37,676


198
00:09:37,676 --> 00:09:40,000


199
00:09:40,000 --> 00:09:42,358


200
00:09:42,358 --> 00:09:44,963


201
00:09:45,440 --> 00:09:49,180
我很好奇，鉴于你们为在不同环境中建立 LATE 结果所做的所有工作，我很好奇，对此有什么想法吗？

202
00:09:49,180 --> 00:09:52,390


203
00:09:52,390 --> 00:09:54,415


204
00:09:55,360 --> 00:09:57,306
- 我认为在很多情况下直觉确实会让你保持距离，但它会受到一些限制，建立正式的结果可能会有点棘手，然后可能只在特殊情况下才有效，然后你就结束了有很多可能无法完全捕捉直觉的形式。

205
00:09:57,306 --> 00:10:02,200


206
00:10:02,800 --> 00:10:04,989


207
00:10:04,989 --> 00:10:07,600


208
00:10:08,400 --> 00:10:09,490


209
00:10:09,490 --> 00:10:12,700


210
00:10:14,600 --> 00:10:16,540


211
00:10:16,540 --> 00:10:19,500


212
00:10:19,900 --> 00:10:21,550
有时我对它们有些不安，它们不一定是我想写的论文，但我确实认为直觉通常确实抓住了问题的一部分。

213
00:10:21,550 --> 00:10:24,438


214
00:10:25,148 --> 00:10:27,819


215
00:10:27,819 --> 00:10:30,912


216
00:10:33,100 --> 00:10:36,300
我认为，从某种意义上说，我们很幸运，LATE 论文在期刊上得到了处理，所以，实际上，编辑把它缩短了很多，这让我们能够专注于非常清晰、清晰的结果。

217
00:10:36,900 --> 00:10:39,250


218
00:10:39,250 --> 00:10:41,766


219
00:10:42,100 --> 00:10:46,300


220
00:10:49,924 --> 00:10:51,770
计量经济学文献中有些不幸的趋势是论文越来越长。

221
00:10:51,770 --> 00:10:53,118


222
00:10:53,118 --> 00:10:54,870


223
00:10:54,870 --> 00:10:56,550
- 嗯，你应该能够解决这个问题，伙计。

224
00:10:56,550 --> 00:10:58,915
- 我正在努力解决这个问题。 [笑声]

225
00:10:58,915 --> 00:11:01,510
但我认为这是一个例子，很明显，它很短实际上是——

226
00:11:01,510 --> 00:11:03,097


227
00:11:03,097 --> 00:11:04,750
- 你应该强制规定没有任何论文可以比 LATE 论文更长。

228
00:11:04,750 --> 00:11:06,802


229
00:11:07,269 --> 00:11:09,617
- 那……哇！那可能很棒。

230
00:11:09,617 --> 00:11:11,685
- 至少没有理论，没有理论论文。

231
00:11:11,892 --> 00:11:13,339
- 是的，我认为......

232
00:11:14,500 --> 00:11:16,800
我正在努力让论文更短，我认为今天有很多价值，因为它通常是论文的第二部分实际上并没有让你得到太多进一步理解事物，它确实使事情变得更难阅读。

233
00:11:16,800 --> 00:11:19,514


234
00:11:19,514 --> 00:11:21,573


235
00:11:21,573 --> 00:11:25,049


236
00:11:25,049 --> 00:11:26,395


237
00:11:27,000 --> 00:11:29,870


238
00:11:32,426 --> 00:11:36,179
这可以追溯到我认为应该如何进行计量经济学。

239
00:11:36,179 --> 00:11:38,070


240
00:11:38,700 --> 00:11:41,300


241
00:11:41,500 --> 00:11:43,900


242
00:11:44,800 --> 00:11:48,900


243
00:11:48,900 --> 00:11:50,010


244
00:11:51,100 --> 00:11:54,670


245
00:11:56,260 --> 00:11:57,750


246
00:11:57,750 --> 00:12:00,230


247
00:12:00,230 --> 00:12:03,765


248
00:12:03,765 --> 00:12:04,994


249
00:12:04,994 --> 00:12:07,563


250
00:12:07,563 --> 00:12:09,840


251
00:12:09,840 --> 00:12:11,600


252
00:12:11,600 --> 00:12:13,950


253
00:12:14,600 --> 00:12:16,770


254
00:12:16,770 --> 00:12:21,000


255
00:12:21,800 --> 00:12:27,530


256
00:12:28,000 --> 00:12:29,300


257
00:12:29,300 --> 00:12:32,030


258
00:12:32,600 --> 00:12:35,500


259
00:12:35,500 --> 00:12:37,000


260
00:12:37,400 --> 00:12:42,362


261
00:12:43,228 --> 00:12:46,700


262
00:12:46,700 --> 00:12:49,250


263
00:12:49,250 --> 00:12:50,800


264
00:12:52,000 --> 00:12:54,250


265
00:12:54,250 --> 00:12:56,100


266
00:12:56,100 --> 00:12:59,736


267
00:12:59,736 --> 00:13:02,143


268
00:13:02,143 --> 00:13:04,290


269
00:13:04,800 --> 00:13:07,000


270
00:13:07,600 --> 00:13:09,501


271
00:13:09,920 --> 00:13:12,805


272
00:13:13,100 --> 00:13:15,400


273
00:13:15,900 --> 00:13:17,730


274
00:13:17,730 --> 00:13:19,231


275
00:13:19,231 --> 00:13:22,616


276
00:13:22,616 --> 00:13:25,100


277
00:13:25,230 --> 00:13:27,534


278
00:13:27,534 --> 00:13:30,900


279
00:13:31,700 --> 00:13:36,068


280
00:13:37,900 --> 00:13:41,190


281
00:13:41,190 --> 00:13:42,900


282
00:13:43,200 --> 00:13:47,280


283
00:13:47,280 --> 00:13:51,180


284
00:13:51,180 --> 00:13:53,272


285
00:13:54,906 --> 00:13:56,477


286
00:13:56,477 --> 00:13:59,300


287
00:13:59,643 --> 00:14:01,992


288
00:14:01,992 --> 00:14:05,657


289
00:14:05,657 --> 00:14:07,910


290
00:14:07,910 --> 00:14:09,557


291
00:14:10,100 --> 00:14:13,389


292
00:14:13,389 --> 00:14:14,688


293
00:14:16,300 --> 00:14:17,414


294
00:14:19,000 --> 00:14:21,700


295
00:14:22,120 --> 00:14:23,370


296
00:14:23,370 --> 00:14:26,780


297
00:14:27,200 --> 00:14:28,780


298
00:14:28,780 --> 00:14:32,836


299
00:14:33,370 --> 00:14:36,325


300
00:14:36,325 --> 00:14:37,807


301
00:14:37,807 --> 00:14:39,289


302
00:14:39,289 --> 00:14:42,100


303
00:14:44,846 --> 00:14:47,500


304
00:14:47,500 --> 00:14:49,467


305
00:14:49,467 --> 00:14:54,212


306
00:14:54,212 --> 00:14:56,920


307
00:14:56,920 --> 00:14:58,950


308
00:14:58,950 --> 00:15:01,140


309
00:15:01,140 --> 00:15:02,723


310
00:15:03,900 --> 00:15:07,630


311
00:15:07,630 --> 00:15:09,430


312
00:15:09,430 --> 00:15:12,730


313
00:15:12,730 --> 00:15:15,480


314
00:15:15,480 --> 00:15:16,800


315
00:15:17,100 --> 00:15:18,550


316
00:15:18,550 --> 00:15:20,900


317
00:15:22,104 --> 00:15:23,732


318
00:15:23,732 --> 00:15:26,783


319
00:15:26,783 --> 00:15:32,772


320
00:15:32,772 --> 00:15:35,595


321
00:15:35,595 --> 00:15:38,290


322
00:15:38,290 --> 00:15:40,300


323
00:15:40,300 --> 00:15:42,300


324
00:15:42,300 --> 00:15:44,400


325
00:15:44,400 --> 00:15:45,890


326
00:15:45,890 --> 00:15:48,200


327
00:15:49,100 --> 00:15:52,200


328
00:15:52,200 --> 00:15:53,675


329
00:15:53,675 --> 00:15:56,390


330
00:15:56,913 --> 00:15:59,400


331
00:15:59,980 --> 00:16:02,840


332
00:16:02,840 --> 00:16:05,112


333
00:16:05,112 --> 00:16:07,100


334
00:16:07,380 --> 00:16:08,900


335
00:16:10,200 --> 00:16:13,929


336
00:16:13,929 --> 00:16:15,950


337
00:16:15,950 --> 00:16:17,330


338
00:16:18,200 --> 00:16:21,241


339
00:16:21,241 --> 00:16:27,430


340
00:16:27,430 --> 00:16:31,600


341
00:16:31,600 --> 00:16:33,274


342
00:16:35,500 --> 00:16:38,600


343
00:16:38,600 --> 00:16:42,893


344
00:16:42,893 --> 00:16:46,330


345
00:16:46,330 --> 00:16:48,300


346
00:16:48,600 --> 00:16:51,500


347
00:16:52,718 --> 00:16:53,868


348
00:16:53,868 --> 00:16:56,920


349
00:16:56,920 --> 00:17:00,533


350
00:17:00,800 --> 00:17:03,240


351
00:17:03,240 --> 00:17:05,420


352
00:17:05,420 --> 00:17:07,600


353
00:17:07,600 --> 00:17:09,700


354
00:17:09,700 --> 00:17:13,370


355
00:17:15,110 --> 00:17:18,230


356
00:17:18,500 --> 00:17:22,900


357
00:17:23,400 --> 00:17:24,933


358
00:17:24,933 --> 00:17:27,570


359
00:17:27,570 --> 00:17:30,300


360
00:17:31,700 --> 00:17:35,204


361
00:17:35,204 --> 00:17:38,516


362
00:17:38,516 --> 00:17:43,400


363
00:17:44,600 --> 00:17:47,020


364
00:17:47,020 --> 00:17:48,200


365
00:17:48,200 --> 00:17:50,500


366
00:17:50,500 --> 00:17:53,100


367
00:17:53,700 --> 00:17:56,720


368
00:17:56,720 --> 00:17:58,300


369
00:17:59,036 --> 00:18:01,495