1
00:00:01,725 --> 00:00:04,642
Hi. I'm here to talk about congestion,

2
00:00:04,642 --> 00:00:05,996
namely road congestion.

3
00:00:05,996 --> 00:00:08,257
Road congestion is a pervasive phenomenon.

4
00:00:08,257 --> 00:00:11,117
It exists in basically all of the cities all around the world,

5
00:00:11,117 --> 00:00:13,065
which is a little bit surprising when you think about it.

6
00:00:13,065 --> 00:00:16,028
I mean, think about how different cities are, actually.

7
00:00:16,028 --> 00:00:17,790
I mean, you have the typical European cities,

8
00:00:17,790 --> 00:00:20,782
with a dense urban core, good public transportation

9
00:00:20,782 --> 00:00:23,237
mostly, not a lot of road capacity.

10
00:00:23,237 --> 00:00:25,686
But then, on the other hand, you have the American cities.

11
00:00:25,686 --> 00:00:28,477
It's moving by itself, okay.

12
00:00:28,477 --> 00:00:31,483
Anyway, the American cities:

13
00:00:31,483 --> 00:00:34,411
lots of roads dispersed over large areas,

14
00:00:34,411 --> 00:00:35,870
almost no public transportation.

15
00:00:35,870 --> 00:00:38,080
And then you have the emerging world cities,

16
00:00:38,080 --> 00:00:39,606
with a mixed variety of vehicles,

17
00:00:39,606 --> 00:00:42,148
mixed land-use patterns, also rather dispersed

18
00:00:42,148 --> 00:00:44,747
but often with a very dense urban core.

19
00:00:44,747 --> 00:00:47,149
And traffic planners all around the world have tried

20
00:00:47,149 --> 00:00:50,803
lots of different measures: dense cities or dispersed cities,

21
00:00:50,803 --> 00:00:53,354
lots of roads or lots of public transport

22
00:00:53,354 --> 00:00:56,140
or lots of bike lanes or more information,

23
00:00:56,140 --> 00:00:59,934
or lots of different things, but nothing seems to work.

24
00:00:59,934 --> 00:01:02,937
But all of these attempts have one thing in common.

25
00:01:02,937 --> 00:01:05,288
They're basically attempts at figuring out

26
00:01:05,288 --> 00:01:08,615
what people should do instead of rush hour car driving.

27
00:01:08,615 --> 00:01:11,727
They're essentially, to a point, attempts at planning

28
00:01:11,727 --> 00:01:14,929
what other people should do, planning their life for them.

29
00:01:14,929 --> 00:01:17,381
Now, planning a complex social system

30
00:01:17,381 --> 00:01:20,456
is a very hard thing to do, and let me tell you a story.

31
00:01:20,456 --> 00:01:23,183
Back in 1989, when the Berlin Wall fell,

32
00:01:23,183 --> 00:01:25,805
an urban planner in London got a phone call

33
00:01:25,805 --> 00:01:28,355
from a colleague in Moscow saying, basically,

34
00:01:28,355 --> 00:01:30,873
"Hi, this is Vladimir. I'd like to know,

35
00:01:30,873 --> 00:01:33,895
who's in charge of London's bread supply?"

36
00:01:33,895 --> 00:01:35,548
And the urban planner in London goes,

37
00:01:35,548 --> 00:01:37,518
"What do you mean, who's in charge of London's —

38
00:01:37,518 --> 00:01:38,789
I mean, no one is in charge."

39
00:01:38,789 --> 00:01:40,671
"Oh, but surely someone must be in charge.

40
00:01:40,671 --> 00:01:44,327
I mean, it's a very complicated system. Someone must control all of this."

41
00:01:44,327 --> 00:01:47,433
"No. No. No one is in charge.

42
00:01:47,433 --> 00:01:50,034
I mean, it basically -- I haven't really thought of it.

43
00:01:50,034 --> 00:01:52,623
It basically organizes itself."

44
00:01:52,623 --> 00:01:55,200
It organizes itself.

45
00:01:55,200 --> 00:01:58,043
That's an example of a complex social system

46
00:01:58,043 --> 00:02:00,771
which has the ability of self-organizing,

47
00:02:00,771 --> 00:02:03,359
and this is a very deep insight.

48
00:02:03,359 --> 00:02:06,672
When you try to solve really complex social problems,

49
00:02:06,672 --> 00:02:08,388
the right thing to do is most of the time

50
00:02:08,388 --> 00:02:09,979
to create the incentives.

51
00:02:09,979 --> 00:02:11,923
You don't plan the details,

52
00:02:11,923 --> 00:02:14,543
and people will figure out what to do,

53
00:02:14,543 --> 00:02:16,403
how to adapt to this new framework.

54
00:02:16,403 --> 00:02:19,894
And let's now look at how we can use this insight

55
00:02:19,894 --> 00:02:21,530
to combat road congestion.

56
00:02:21,530 --> 00:02:23,762
This is a map of Stockholm, my hometown.

57
00:02:23,762 --> 00:02:27,403
Now, Stockholm is a medium-sized city, roughly two million people,

58
00:02:27,403 --> 00:02:30,192
but Stockholm also has lots of water and lots of water

59
00:02:30,192 --> 00:02:33,469
means lots of bridges -- narrow bridges, old bridges --

60
00:02:33,469 --> 00:02:36,504
which means lots of road congestion.

61
00:02:36,504 --> 00:02:40,144
And these red dots show the most congested parts,

62
00:02:40,144 --> 00:02:42,721
which are the bridges that lead into the inner city.

63
00:02:42,721 --> 00:02:45,247
And then someone came up with the idea that,

64
00:02:45,247 --> 00:02:47,084
apart from good public transport,

65
00:02:47,084 --> 00:02:49,685
apart from spending money on roads,

66
00:02:49,685 --> 00:02:53,900
let's try to charge drivers one or two euros at these bottlenecks.

67
00:02:53,900 --> 00:02:56,824
Now, one or two euros, that isn't really a lot of money,

68
00:02:56,824 --> 00:02:59,481
I mean compared to parking charges and running costs, etc.,

69
00:02:59,481 --> 00:03:02,199
so you would probably expect that car drivers

70
00:03:02,199 --> 00:03:04,637
wouldn't really react to this fairly small charge.

71
00:03:04,637 --> 00:03:05,937
You would be wrong.

72
00:03:05,937 --> 00:03:09,965
One or two euros was enough to make 20 percent of cars

73
00:03:09,965 --> 00:03:12,349
disappear from rush hours.

74
00:03:12,349 --> 00:03:15,900
Now, 20 percent, well, that's a fairly huge figure, you might think,

75
00:03:15,900 --> 00:03:18,107
but you've still got 80 percent left of the problem, right?

76
00:03:18,107 --> 00:03:19,976
Because you still have 80 percent of the traffic.

77
00:03:19,976 --> 00:03:23,243
Now, that's also wrong, because traffic happens to be

78
00:03:23,243 --> 00:03:25,935
a nonlinear phenomenon, meaning that

79
00:03:25,935 --> 00:03:29,076
once you reach above a certain capacity threshold

80
00:03:29,076 --> 00:03:32,244
then congestion starts to increase really, really rapidly.

81
00:03:32,244 --> 00:03:34,704
But fortunately, it also works the other way around.

82
00:03:34,704 --> 00:03:38,035
If you can reduce traffic even somewhat, then congestion

83
00:03:38,035 --> 00:03:41,664
will go down much faster than you might think.

84
00:03:41,664 --> 00:03:44,555
Now, congestion charges were introduced in Stockholm

85
00:03:44,555 --> 00:03:49,159
on January 3, 2006, and the first picture here is a picture

86
00:03:49,159 --> 00:03:53,142
of Stockholm, one of the typical streets, January 2.

87
00:03:53,142 --> 00:03:56,955
The first day with the congestion charges looked like this.

88
00:03:56,955 --> 00:03:59,875
This is what happens when you take away

89
00:03:59,875 --> 00:04:01,940
20 percent of the cars from the streets.

90
00:04:01,940 --> 00:04:04,650
You really reduce congestion quite substantially.

91
00:04:04,650 --> 00:04:08,308
But, well, as I said, I mean, car drivers adapt, right?

92
00:04:08,308 --> 00:04:10,810
So after a while they would all come back because they

93
00:04:10,810 --> 00:04:12,758
have sort of gotten used to charges.

94
00:04:12,758 --> 00:04:16,434
Wrong again. It's now six and a half years ago

95
00:04:16,434 --> 00:04:18,417
since the congestion charges were introduced in Stockholm,

96
00:04:18,417 --> 00:04:22,406
and we basically have the same low traffic levels still.

97
00:04:22,406 --> 00:04:24,677
But you see, there's an interesting gap here in the time series

98
00:04:24,677 --> 00:04:26,140
in 2007.

99
00:04:26,140 --> 00:04:28,529
Well, the thing is that, the congestion charges,

100
00:04:28,529 --> 00:04:31,625
they were introduced first as a trial, so they were introduced

101
00:04:31,625 --> 00:04:35,594
in January and then abolished again at the end of July,

102
00:04:35,594 --> 00:04:38,188
followed by a referendum, and then they were reintroduced

103
00:04:38,188 --> 00:04:42,388
again in 2007, which of course was a wonderful scientific opportunity.

104
00:04:42,388 --> 00:04:46,525
I mean, this was a really fun experiment to start with,

105
00:04:46,525 --> 00:04:48,306
and we actually got to do it twice.

106
00:04:48,306 --> 00:04:50,415
And personally, I would like to do this every once a year or so,

107
00:04:50,415 --> 00:04:51,903
but they won't let me do that.

108
00:04:51,903 --> 00:04:54,088
But it was fun anyway.

109
00:04:54,088 --> 00:04:57,473
So, we followed up. What happened?

110
00:04:57,473 --> 00:05:01,339
This is the last day with the congestion charges, July 31,

111
00:05:01,339 --> 00:05:03,117
and you see the same street but now it's summer,

112
00:05:03,117 --> 00:05:04,861
and summer in Stockholm is a very nice

113
00:05:04,861 --> 00:05:07,111
and light time of the year,

114
00:05:07,111 --> 00:05:09,954
and the first day without the congestion charges

115
00:05:09,954 --> 00:05:11,321
looked like this.

116
00:05:11,321 --> 00:05:13,984
All the cars were back again, and you even have to admire

117
00:05:13,984 --> 00:05:16,894
the car drivers. They adapt so extremely quickly.

118
00:05:16,894 --> 00:05:19,141
The first day they all came back.

119
00:05:19,141 --> 00:05:24,491
And this effect hanged on. So 2007 figures looked like this.

120
00:05:24,491 --> 00:05:26,822
Now these traffic figures are really exciting

121
00:05:26,822 --> 00:05:29,473
and a little bit surprising and very useful to know,

122
00:05:29,473 --> 00:05:32,232
but I would say that the most surprising slide here

123
00:05:32,232 --> 00:05:36,262
I'm going to show you today is not this one. It's this one.

124
00:05:36,262 --> 00:05:39,593
This shows public support for congestion pricing of Stockholm,

125
00:05:39,593 --> 00:05:42,262
and you see that when congestion pricing were introduced

126
00:05:42,262 --> 00:05:46,579
in the beginning of Spring 2006, people were fiercely against it.

127
00:05:46,579 --> 00:05:49,607
Seventy percent of the population didn't want this.

128
00:05:49,607 --> 00:05:51,255
But what happened when the congestion charges

129
00:05:51,255 --> 00:05:54,716
were there is not what you would expect, that people hated it more and more.

130
00:05:54,716 --> 00:05:57,869
No, on the contrary, they changed, up to a point

131
00:05:57,869 --> 00:06:01,570
where we now have 70 percent support for keeping the charges,

132
00:06:01,570 --> 00:06:03,411
meaning that -- I mean, let me repeat that:

133
00:06:03,411 --> 00:06:06,218
70 percent of the population in Stockholm

134
00:06:06,218 --> 00:06:10,211
want to keep a price for something that used to be free.

135
00:06:10,211 --> 00:06:13,837
Okay. So why can that be? Why is that?

136
00:06:13,837 --> 00:06:16,837
Well, think about it this way. Who changed?

137
00:06:16,837 --> 00:06:19,463
I mean, the 20 percent of the car drivers that disappeared,

138
00:06:19,463 --> 00:06:21,408
surely they must be discontent in a way.

139
00:06:21,408 --> 00:06:23,742
And where did they go? If we can understand this,

140
00:06:23,742 --> 00:06:27,872
then maybe we can figure out how people can be so happy with this.

141
00:06:27,872 --> 00:06:30,274
Well, so we did this huge interview survey

142
00:06:30,274 --> 00:06:32,109
with lots of travel services, and tried to figure out

143
00:06:32,109 --> 00:06:33,864
who changed, and where did they go?

144
00:06:33,864 --> 00:06:38,278
And it turned out that they don't know themselves. (Laughter)

145
00:06:38,278 --> 00:06:43,126
For some reason, the car drivers are --

146
00:06:43,126 --> 00:06:46,622
they are confident they actually drive the same way that they used to do.

147
00:06:46,622 --> 00:06:48,878
And why is that? It's because that travel patterns

148
00:06:48,878 --> 00:06:51,277
are much less stable than you might think.

149
00:06:51,277 --> 00:06:54,108
Each day, people make new decisions, and people change

150
00:06:54,108 --> 00:06:56,533
and the world changes around them, and each day

151
00:06:56,533 --> 00:06:59,786
all of these decisions are sort of nudged ever so slightly

152
00:06:59,786 --> 00:07:01,753
away from rush hour car driving

153
00:07:01,753 --> 00:07:04,337
in a way that people don't even notice.

154
00:07:04,337 --> 00:07:06,382
They're not even aware of this themselves.

155
00:07:06,382 --> 00:07:09,183
And the other question, who changed their mind?

156
00:07:09,183 --> 00:07:11,149
Who changed their opinion, and why?

157
00:07:11,149 --> 00:07:13,675
So we did another interview survey, tried to figure out

158
00:07:13,675 --> 00:07:18,051
why people changed their mind, and what type of group changed their minds?

159
00:07:18,051 --> 00:07:20,944
And after analyzing the answers, it turned out that

160
00:07:20,944 --> 00:07:24,829
more than half of them believe that they haven't changed their minds.

161
00:07:24,829 --> 00:07:27,116
They're actually confident that they have

162
00:07:27,116 --> 00:07:29,805
liked congestion pricing all along.

163
00:07:29,805 --> 00:07:31,524
Which means that we are now in a position

164
00:07:31,524 --> 00:07:34,517
where we have reduced traffic across this toll cordon

165
00:07:34,517 --> 00:07:37,728
with 20 percent, and reduced congestion by enormous numbers,

166
00:07:37,728 --> 00:07:41,275
and people aren't even aware that they have changed,

167
00:07:41,275 --> 00:07:45,186
and they honestly believe that they have liked this all along.

168
00:07:45,186 --> 00:07:48,660
This is the power of nudges when trying to solve

169
00:07:48,660 --> 00:07:51,292
complex social problems, and when you do that,

170
00:07:51,292 --> 00:07:54,922
you shouldn't try to tell people how to adapt.

171
00:07:54,922 --> 00:07:57,109
You should just nudge them in the right direction.

172
00:07:57,109 --> 00:07:58,831
And if you do it right,

173
00:07:58,831 --> 00:08:00,883
people will actually embrace the change,

174
00:08:00,883 --> 00:08:03,945
and if you do it right, people will actually even like it.

175
00:08:03,945 --> 00:08:07,270
Thank you. (Applause)