1
00:00:01,440 --> 00:00:04,079
Now we're recording. Let's hope this works.

2
00:00:04,080 --> 00:00:15,120
And maybe we reduce the so that first of all, can you just say your name and your position and affiliation here in Oxford?

3
00:00:15,360 --> 00:00:25,320
Yeah. My name is Max Rosen, and I am the director of the Oxford Martin Program on Global Development.

4
00:00:25,950 --> 00:00:30,989
Thanks very much. And first of all, can you just tell me a little bit about yourself and how you got you're an economist,

5
00:00:30,990 --> 00:00:35,280
how you got interested in that subject in in economics?

6
00:00:35,310 --> 00:00:42,060
Yes, I suppose that was like it took me some time to actually come up with this idea of studying economics.

7
00:00:42,210 --> 00:00:45,090
Now, in retrospect, it makes a lot of sense, but at the time I had no idea.

8
00:00:45,510 --> 00:00:50,969
I come from Germany and somehow, weirdly, economics doesn't have a high status in Germany.

9
00:00:50,970 --> 00:00:56,070
It's not a very it's like it's a much lower status than back then here in the U.K., in the U.S.,

10
00:00:56,670 --> 00:01:03,270
where we have very few economists, like major economists, say the last one was Marx or something.

11
00:01:05,460 --> 00:01:13,950
And so when I started studying at the university, which was in Austria, I studied philosophy and geoscience.

12
00:01:15,000 --> 00:01:19,020
And the idea at the time, which doesn't make that much sense in retrospect,

13
00:01:19,020 --> 00:01:26,760
was that I had these interests, both in philosophy and ethics, but also in numbers and maths and science.

14
00:01:27,390 --> 00:01:32,190
And I thought with geoscience and philosophy, I can somehow combine them into one.

15
00:01:32,910 --> 00:01:35,340
But I think if I had known that economics was a thing,

16
00:01:35,340 --> 00:01:43,590
then this would have been the right overlap between ethical questions and quantitative research or along.

17
00:01:43,590 --> 00:01:52,470
But it took me like I did the two undergrads and I found out about economics, got fascinated by by that research.

18
00:01:52,960 --> 00:01:57,900
I remember whether reading Amartya Sen, an Indian economist that impressed me very much,

19
00:01:57,900 --> 00:02:05,370
it still impressed me very much, then ended up studying economics and later did my PhD in Austria.

20
00:02:06,030 --> 00:02:11,490
And the research topic that I was focusing on at the time was inequality of incomes.

21
00:02:13,620 --> 00:02:19,080
And at the time the big inequality researcher was here at Oxford.

22
00:02:19,410 --> 00:02:26,520
His name is Tony Atkinson, and I randomly sent him an email one afternoon,

23
00:02:26,910 --> 00:02:35,070
sending him my poor research that I was doing there at the time, and he was kind enough to write back.

24
00:02:35,790 --> 00:02:44,969
And a couple of months later he invited me to come here to Oxford and I visited him for four months and thought I'd go home after four months.

25
00:02:44,970 --> 00:02:48,020
But now it's ten years later and I'm still here. And you?

26
00:02:48,030 --> 00:02:51,180
I read that you did a lot of travelling during your studies.

27
00:02:51,990 --> 00:02:59,490
Yeah, I wanted. To what extent that influenced your focus on inequality in a global sense?

28
00:03:01,350 --> 00:03:07,979
It definitely influenced me a lot. Like it's true. Like at the time when I was in Australia I was it was great there.

29
00:03:07,980 --> 00:03:16,350
We had these three month summer breaks and overall undergrads there are a bit more to than over here in Oxford.

30
00:03:16,350 --> 00:03:20,230
So people are just not thinking of doing internships or anything like that.

31
00:03:20,250 --> 00:03:27,000
Everyone just went away. So I like maximised travel over time and actually I had my money that way.

32
00:03:27,000 --> 00:03:33,059
So in the spring I would always go and do bicycle tour guiding across Europe like in

33
00:03:33,060 --> 00:03:39,900
Portugal and Poland and Sweden and France mostly really got some money together.

34
00:03:39,900 --> 00:03:45,720
And then in the summer I would go and travel like three months to Africa,

35
00:03:45,900 --> 00:03:51,600
three months to South America, three months through Asia, like always on the go.

36
00:03:52,600 --> 00:04:00,270
Mm. I don't recall like this moment where it was like it would somehow push me into this research.

37
00:04:00,270 --> 00:04:03,959
But obviously just seeing this vast inequality in the world,

38
00:04:03,960 --> 00:04:14,520
the poverty in most countries in the world really obviously makes you ask the question of what's going on, like how is that possible?

39
00:04:14,520 --> 00:04:22,259
Like why are some people so rich and other people are so very poor, which essentially is the central question of economics.

40
00:04:22,260 --> 00:04:27,600
So I don't recall like a clear path, but somehow it must have been in the back of my mind.

41
00:04:27,780 --> 00:04:34,530
Mm hmm. So you got you arrived here in Oxford in 2012, 2012, 13th February 2012.

42
00:04:34,770 --> 00:04:40,590
And I thought I saw something that said, oh, we're going data started in 2011 which it so had it already.

43
00:04:40,650 --> 00:04:43,440
Yeah. So our world in data existed before you were right.

44
00:04:43,440 --> 00:04:50,970
I mean, like, I started thinking about this project like it didn't exist, so it wasn't actually at the beginning so much tied to Oxford.

45
00:04:51,390 --> 00:04:54,480
Oh, I see. Well, sorry, I should. I jumped ahead a little bit.

46
00:04:54,630 --> 00:04:59,460
So you're you're currently currently mainly known.

47
00:05:00,150 --> 00:05:03,959
For developing this incredible resource world in data.

48
00:05:03,960 --> 00:05:08,850
So what? What was the origin story? I'm building data lake there.

49
00:05:08,850 --> 00:05:21,630
The lake. The origin was really the year before I came to Oxford when I was in Brazil and I was working on my Ph.D. and overall I thought.

50
00:05:23,680 --> 00:05:28,270
It's just not well known enough. Like, I was like when I was like,

51
00:05:29,110 --> 00:05:34,179
like connecting that back to the earlier question about the times and look like I was just not

52
00:05:34,180 --> 00:05:39,340
understanding even the most basic trends of like how living conditions in the world have changed.

53
00:05:39,760 --> 00:05:44,739
I remember sitting in the lecture of Kansas Crespo Quaresma,

54
00:05:44,740 --> 00:05:50,620
who later became a close collaborator and was supervising me during the time of the Ph.D.,

55
00:05:51,580 --> 00:05:55,080
and he was showing these long run trends in declining poverty.

56
00:05:55,090 --> 00:05:58,219
And I just didn't believe them. I thought, that's just made up.

57
00:05:58,220 --> 00:06:05,590
Like it was like. Now it's strange to even think that because so much of the last years I spent researching these questions.

58
00:06:05,830 --> 00:06:07,300
But at the time, I just really didn't know.

59
00:06:07,810 --> 00:06:13,780
And so I thought, like, I am interested in all of these things, like how you travel to all of these places.

60
00:06:13,780 --> 00:06:18,700
I read the newspaper and I don't know even the most basic developments in the world.

61
00:06:19,780 --> 00:06:27,219
And when I was in Brazil, Brazil at the time was was in a much better place.

62
00:06:27,220 --> 00:06:32,650
Like there was a lot of optimism, a lot of like like enthusiasm about the future.

63
00:06:34,090 --> 00:06:37,710
This was good pre bonds. Bolsonaro. Exactly. Yeah. It was 2011.

64
00:06:37,750 --> 00:06:42,280
There was like a period of high economic growth. There was lots of development.

65
00:06:43,870 --> 00:06:51,909
And and also, like, what happens often in poorer places is when they develop, they can develop extremely fast, much,

66
00:06:51,910 --> 00:06:59,410
much faster than the countries where I grew up in, where the development is, is stretched out over decades or centuries.

67
00:06:59,890 --> 00:07:03,250
And so there was like you could just accede with your eyes that people knew,

68
00:07:03,250 --> 00:07:07,660
like how things are changing, how health is improving, how poverty is declining.

69
00:07:08,110 --> 00:07:08,770
And then I thought,

70
00:07:09,040 --> 00:07:16,990
I need to convey that information to others that are not aware of these basic developments and thought I would write a book about it.

71
00:07:17,800 --> 00:07:26,080
And for the book, I started collecting data on all of these relevant aspects, and that got out of hand.

72
00:07:26,080 --> 00:07:31,420
And I still haven't written the book, but I had collected more data than I probably need for a book.

73
00:07:31,930 --> 00:07:34,330
And that turned into the website. Mm hmm.

74
00:07:35,050 --> 00:07:42,970
So, I mean, the website must have involved thinking about how the data needed to be organised and how the data needed to be accessed.

75
00:07:44,080 --> 00:07:47,110
Yeah, exactly. Like that came also later.

76
00:07:47,120 --> 00:07:50,769
Like at the beginning, it was just like there wasn't any plan for building a website.

77
00:07:50,770 --> 00:07:55,780
I was just like collecting that on my, on my computer. I just had these like spreadsheets full of data.

78
00:07:56,500 --> 00:08:01,110
And to a good extent it was actually Tony who I mentioned earlier, the inequality researcher.

79
00:08:01,570 --> 00:08:06,610
I can remember this conversation with him where he published a book on earnings inequality,

80
00:08:06,610 --> 00:08:12,070
and there was 80 countries and he was so frustrated that there was a typo somewhere in a table there.

81
00:08:12,160 --> 00:08:20,110
Like it was a book. It was a book mostly of presenting data in a book format, which obviously is not a great idea.

82
00:08:20,110 --> 00:08:24,520
But the publisher, I think it might have been Oxford University Press thought it was a good idea.

83
00:08:25,000 --> 00:08:34,270
And so they so he had this like 200 page book with tables off the tables at a table and one table on page 70 had a typo,

84
00:08:34,270 --> 00:08:41,259
and he was frustrated that he couldn't fix it and was talking about how it should all be online and really, you know, it's that's not the way to go.

85
00:08:41,260 --> 00:08:45,000
And he was like present quite a bit for publishing data online.

86
00:08:45,010 --> 00:08:48,840
So that was definitely part of the thinking.

87
00:08:48,850 --> 00:08:57,830
And then I started building the websites in maybe 2012, but that was just a project in the evenings and weekends.

88
00:08:57,850 --> 00:09:09,530
My main job was to finish my Ph.D. and to do research on income inequality, and then I kept on doing that for the coming years.

89
00:09:10,630 --> 00:09:16,720
I didn't publish it, I don't think, until 2014, um, maybe 2015.

90
00:09:17,590 --> 00:09:25,450
Um, and just built this website on my computer like I had like a local server running on my computer was building the website there.

91
00:09:25,810 --> 00:09:32,080
I think that the mistake that many founders are doing where you're too shy to actually release

92
00:09:32,080 --> 00:09:40,090
the work to the world and thought it needs to be all like perfect before I put it out.

93
00:09:40,900 --> 00:09:46,210
And that's obviously just a bad idea because you want to actually get the work out,

94
00:09:46,570 --> 00:09:53,890
especially if you don't have any like big reputation or any like risks in this in this regard to learn as quickly as possible from your readers.

95
00:09:54,970 --> 00:10:01,300
But I didn't do that. And in 2014 or 15 it was then online.

96
00:10:02,480 --> 00:10:12,280
I look that up quickly. I wrote this history of our and data for the website,

97
00:10:13,120 --> 00:10:22,600
which goes all the way except like until 2019 and since the pandemic, everything has been so crazy that it is not in the.

98
00:10:23,320 --> 00:10:30,460
May 2014. I exactly. In May 2014, I launched it over the following six months, 20,000 visitors every month.

99
00:10:30,820 --> 00:10:34,180
And at the time, I just couldn't believe it that people would actually visit it.

100
00:10:34,510 --> 00:10:38,560
Like I had like Google Analytics open and it was just like mega excited when,

101
00:10:39,160 --> 00:10:43,960
when someone was like on the website and I saw someone popping up in India or something.

102
00:10:44,140 --> 00:10:48,640
So how many different kinds of data were you collecting and where did you get it from?

103
00:10:49,660 --> 00:10:56,590
Um, I tried like it's I think one reason why I waited so long was that I tried to be very broad.

104
00:10:56,620 --> 00:11:05,230
I tried to cover health and food and hunger and inequality, obviously poverty, economic growth, violence.

105
00:11:05,500 --> 00:11:13,660
So I tried to make it very broad early on because I thought that's kind of the unique angle of of this publication.

106
00:11:15,250 --> 00:11:21,520
And I got it from a rate like it's very it was very similar to how it is now.

107
00:11:21,520 --> 00:11:26,839
It's a range of sources from like things that are very straightforward to to get like the

108
00:11:26,840 --> 00:11:34,059
U.N. data on population figures to things that are somewhere in the middle of a spectrum,

109
00:11:34,060 --> 00:11:43,780
like the appendix of a particular paper where there's like some PDF with some table that you have to convert into,

110
00:11:43,870 --> 00:11:50,079
like an actual spreadsheet that you can use the data. So a bit like middle of the spectrum.

111
00:11:50,080 --> 00:11:58,540
And then sometimes I would produce datasets based on bringing different sources together,

112
00:11:58,540 --> 00:12:03,189
very like what became a very central part of the COVID work much later,

113
00:12:03,190 --> 00:12:11,530
where our job was in many ways to produce datasets and to collate the data from many different sources.

114
00:12:11,650 --> 00:12:18,940
But it wasn't like that at the beginning as well. Mm hmm. So you're aggregating data collected by other lots of different organisations.

115
00:12:19,120 --> 00:12:25,990
And then how were you representing it at the time when you actually presented the to the website publicly?

116
00:12:26,290 --> 00:12:32,919
Yeah, that was also similar to how it is now. Like it was always like it was always a strong focus on data visualisation.

117
00:12:32,920 --> 00:12:39,940
I always thought that there at this small extra step from turning a spreadsheet into an actual visualisation,

118
00:12:40,180 --> 00:12:43,830
a graph or a diagram of some kind of world map. Mm hmm.

119
00:12:44,020 --> 00:12:55,270
Like, actually just provides a lot of value. And I still find it odd how little emphasis there is in the presentation or the publication of data,

120
00:12:55,720 --> 00:13:00,250
where most publications take it to the spreadsheet level,

121
00:13:00,250 --> 00:13:07,600
but then neglect a bit the tools that actually bring this data to life and allow people to interact with it.

122
00:13:07,780 --> 00:13:11,260
And I think that's what made it so appealing to so many people when it first came out.

123
00:13:11,680 --> 00:13:15,430
Exactly. Yeah, exactly. I mean, it wasn't really all that much to it,

124
00:13:15,430 --> 00:13:25,240
like it was just to tell a story that's like a story with my my partner and my girlfriend for many years.

125
00:13:25,480 --> 00:13:34,000
She, like, knew about hours and days before and her dad also knew about it before and he was like excited at seeing our data.

126
00:13:34,000 --> 00:13:39,280
Wrote her an email that I saw that even and she and he is excited that I our in

127
00:13:39,310 --> 00:13:43,930
data access and she was like replying I well obviously I know and data they're

128
00:13:43,940 --> 00:13:56,590
just taking the data from the UN and republishing it in charts which is true like in much of the work that we do is pretty simple in that in that way.

129
00:13:59,280 --> 00:14:02,859
And I mean who were your main users early on?

130
00:14:02,860 --> 00:14:05,920
Do you think it I mean, was it media, was it other researchers?

131
00:14:05,930 --> 00:14:14,589
Was it the general public meeting with her? It was like it was I was definitely surprised that there were many more than I had thought.

132
00:14:14,590 --> 00:14:21,280
That was like just a lot more enthusiasm from non-experts than I had thought.

133
00:14:21,370 --> 00:14:28,629
Like just people who, like, I, I still really like social media and sharing research on social media.

134
00:14:28,630 --> 00:14:33,520
And I was starting that early on in this in those days and it really went well,

135
00:14:33,520 --> 00:14:42,940
like just putting out a nice Web map that shows you when I go to work and I was like, I remember like a very early one was two months.

136
00:14:42,940 --> 00:14:49,000
One was child mortality levels in 1970 and one was tech, which has 11, I guess, in 2011 or something.

137
00:14:49,450 --> 00:14:55,750
And just that contrast of how in all countries in the world, child mortality had declined.

138
00:14:56,680 --> 00:14:59,710
That was interesting to lots of people and lots of people were shared.

139
00:14:59,830 --> 00:15:07,470
So there was always a large share of the population who are just, like, generally interested people.

140
00:15:08,050 --> 00:15:14,680
Mm hmm. And so, I mean, were you able to get funding eventually to get some more people working on it?

141
00:15:15,130 --> 00:15:22,900
Yeah. Yeah, that's true. That was then basically the big next step where I didn't have the ambition at all to.

142
00:15:23,320 --> 00:15:29,620
Search for funding. I didn't think that this would ever become a university project, and it was Tony who pushed for that idea.

143
00:15:30,040 --> 00:15:37,449
He had the idea that really there should be an institution that like carries this work forward,

144
00:15:37,450 --> 00:15:40,480
there should be people like there should be a team of people working on it.

145
00:15:41,470 --> 00:15:48,490
And he had also the idea to approach the Nuffield Foundation in London to ask them for funding.

146
00:15:49,960 --> 00:16:05,770
And then we I wrote the first of many grant applications over £75,000 and they rejected it and sent some referee reports.

147
00:16:05,770 --> 00:16:11,950
And I had to rewrite everything. And ten months later or something they finally accepted it and I got the 75,000

148
00:16:12,640 --> 00:16:16,330
and I started working with the first three colleagues over the summer back then.

149
00:16:18,250 --> 00:16:21,450
And that was I think in 2015. Yeah.

150
00:16:21,550 --> 00:16:33,310
Yeah. And was the work just to build on and expand and make the database better and better?

151
00:16:33,310 --> 00:16:38,830
Or were you within your unit also carrying out your own research on the data that you'd collected?

152
00:16:40,330 --> 00:16:47,950
Um, there's always, I think in the big picture,

153
00:16:48,190 --> 00:16:54,790
I think it's a bit of an unusual project where like maybe the way to think about it is

154
00:16:54,790 --> 00:17:02,170
that there isn't the status of data in especially social science research isn't very high,

155
00:17:02,500 --> 00:17:03,130
oddly enough.

156
00:17:03,910 --> 00:17:12,040
So I think of the natural sciences, it's often a bit better where you can produce a data set, and that's a major publication and people cite it.

157
00:17:12,040 --> 00:17:17,619
And that's that's the way that you have an academic career in the social sciences.

158
00:17:17,620 --> 00:17:18,790
That's not the case.

159
00:17:19,690 --> 00:17:27,370
Like broadly, they're the models is that you produce a data set and then you analyse the dataset and you answer a particular question.

160
00:17:27,370 --> 00:17:32,440
And that's the thing that that gets you an academic career.

161
00:17:32,800 --> 00:17:41,860
But producing a data set as such isn't regarded as, as a, as a valuable contribution, which I think is, is a big mistake.

162
00:17:41,980 --> 00:17:48,280
There's a lot of emphasis on on methods of analysing that data on statistical methods, econometrics in economics.

163
00:17:50,260 --> 00:17:56,200
But if you have crappy data, then all of your fancy statistical methods don't help you that much further either.

164
00:17:56,200 --> 00:18:00,910
So I think there should be a bit of a rebalancing where the status of data is is higher.

165
00:18:02,410 --> 00:18:07,360
And so in some respects, like in that respect, this project has always been unusual.

166
00:18:07,450 --> 00:18:13,750
Where I often left it at, here's the best data that we can produce.

167
00:18:14,260 --> 00:18:18,669
Like it's not for others to make use of that data and, and take it forward.

168
00:18:18,670 --> 00:18:28,120
And that's has been very much the case for COVID, especially where our work was not to analyse what we can learn from the data about,

169
00:18:28,240 --> 00:18:31,719
I don't know, lockdowns or vaccinations or any travel restrictions,

170
00:18:31,720 --> 00:18:35,140
but all of these many questions that people might ask based on this data,

171
00:18:35,470 --> 00:18:43,240
our job was always sort like the way that I thought of our work was we bring we bring the global

172
00:18:43,240 --> 00:18:49,510
data together in one place so that others have the basis on which they can do it in their research.

173
00:18:49,960 --> 00:18:53,020
And that has always been has also been true before.

174
00:18:53,020 --> 00:18:57,310
So we always do some research, but the balance is different,

175
00:18:57,460 --> 00:19:07,360
I think from most academics where our focus is more on the data production and data presentation and not so much on the research.

176
00:19:08,350 --> 00:19:17,380
And moving on from just academics, you would think that this kind of data is what policymakers should be basing their decisions on.

177
00:19:17,830 --> 00:19:25,810
Do you think data has been neglected by policymakers? I mean, obviously the question is always the answer is always, yes, you do it like this.

178
00:19:26,560 --> 00:19:34,180
But I mean, maybe I'm not so down on them like that's so pessimistic on it.

179
00:19:34,180 --> 00:19:42,489
I think I think COVID has been like this kind of intense statistical data literacy course for the entire world.

180
00:19:42,490 --> 00:19:47,139
And I think overall it has gone pretty well like there was there was obviously there

181
00:19:47,140 --> 00:19:52,000
was like misuse of data and people like twisting the data and in unhelpful ways.

182
00:19:52,000 --> 00:20:06,760
But by and large, I think I was like happy to see how, how the, how central the data was in the arguments of policymakers talking about with them.

183
00:20:07,180 --> 00:20:18,340
Yeah. Yes, we even go it go. Oh I see. No, I'm I'm really thinking about when you started the project, you think data with less was less of a thing.

184
00:20:18,880 --> 00:20:22,360
Yeah. A policymaker. So yeah I think overall the last ten years where.

185
00:20:23,220 --> 00:20:31,080
Ten years in which the status of data and the use of data has increased just in all spheres from journalism to policymaking.

186
00:20:33,450 --> 00:20:35,760
So I think this is really going in a good direction.

187
00:20:36,090 --> 00:20:43,260
Obviously, anyone, everyone who's working in that wishes we would be much further and is annoyed by all of the problems.

188
00:20:43,260 --> 00:20:47,790
But I think it's it's been it's now less neglected than it was back then.

189
00:20:47,820 --> 00:20:52,500
Mm hmm. And just to sort of summarise what again, before you got to it.

190
00:20:52,860 --> 00:21:03,719
The the the content of the data was or is geared towards essentially how how good a life people have.

191
00:21:03,720 --> 00:21:09,150
Is that. If you could narrow it down to to what it's relevant to.

192
00:21:09,150 --> 00:21:12,480
It's relevant to how well people are living in whichever country there.

193
00:21:12,930 --> 00:21:15,480
Yeah, I think that's that's like fair.

194
00:21:15,840 --> 00:21:27,270
Like now we usually frame it around the world's largest problems, which obviously includes a lot of progress and shortcomings around how people live.

195
00:21:27,720 --> 00:21:37,350
But it's also a bit broader than that. It also includes especially non non-human animals and the biosphere around us.

196
00:21:37,430 --> 00:21:44,310
Like there's been a lot of work by my colleague Hannah Ritchie on on those topics.

197
00:21:44,790 --> 00:21:49,560
So it's not just about living conditions. It was also, to a large extent around the world more broadly.

198
00:21:49,920 --> 00:21:59,340
But overall, I, I very much want it to be a publication of data that's relevant to how people to people's living

199
00:21:59,340 --> 00:22:04,950
conditions and what it means for people to live at that particular time in that particular place.

200
00:22:05,790 --> 00:22:10,260
So lots of data that's available in the world doesn't fit that.

201
00:22:10,410 --> 00:22:17,069
We're not presenting data on sports where maybe sports is also relevant to how

202
00:22:17,070 --> 00:22:24,650
people think about it or financial data or like all of these aspects don't.

203
00:22:25,260 --> 00:22:27,809
I'm not I'm not so much part of the focus.

204
00:22:27,810 --> 00:22:38,430
It's it's really problems that people have and the progress that we might achieve against these problems and the world's biggest problems that.

205
00:22:38,850 --> 00:22:43,319
How many are there? Are there just a large number of them.

206
00:22:43,320 --> 00:22:49,860
But I mean, what are the top five? The top five, I would think.

207
00:22:52,230 --> 00:22:56,310
I mean, like the highest level, it's kind of like global health and early death.

208
00:22:56,880 --> 00:23:01,080
It's hunger and poor nutrition.

209
00:23:01,950 --> 00:23:08,130
It's all of the economic aspects from centrally about poverty.

210
00:23:09,690 --> 00:23:17,690
Um, it is all of the, a range of environmental problems from climate change to biodiversity losses.

211
00:23:18,720 --> 00:23:27,120
Um, it's existential risk or catastrophic risks of very large scale risks, including pandemics,

212
00:23:27,690 --> 00:23:35,260
but also, for example, the worst case wars or nuclear, the use of nuclear bombs.

213
00:23:37,130 --> 00:23:43,020
Um, access to education and the quality of education is a big topic.

214
00:23:44,010 --> 00:23:57,330
Um, the functioning of the public sector and the provision of public goods from global health to education to social care, that range of, of aspects.

215
00:23:58,920 --> 00:24:06,870
It's a long list, but yes, yeah, those are some of the. And you are capturing or letting in a few spreadsheets and a lot of spreadsheets.

216
00:24:07,230 --> 00:24:10,920
That's right. Yeah. So let's okay, let's arrive at COVID now.

217
00:24:10,920 --> 00:24:16,979
So can you remember where you were when you first heard about the the what was going on in China

218
00:24:16,980 --> 00:24:22,350
and how it might look that it looked as if it might affect beginning it become a global problem,

219
00:24:22,350 --> 00:24:34,110
right? Yes, I do. I remember very well I had lunch maybe on the early January, like maybe 7th of January or something in George Street.

220
00:24:34,470 --> 00:24:40,320
And I had a burger and I was reading a message from a friend of mine who was here at the university as well,

221
00:24:40,320 --> 00:24:45,959
Moritz Kramer, who is an expert in infectious diseases.

222
00:24:45,960 --> 00:24:56,760
And he was sending me some data from China and was telling me about how his sick worried and getting interested in what's what's happening there.

223
00:24:56,760 --> 00:25:02,960
He was like looking at travel out of China to understand whether that disease might possibly spread.

224
00:25:03,390 --> 00:25:10,520
I mean, it was very early days where it was still a bit of a debate of whether human to human transmission was possible.

225
00:25:10,530 --> 00:25:14,490
Like it's like, yeah, feels very far removed now.

226
00:25:15,120 --> 00:25:18,930
Yeah. Early January was the first time I came across it.

227
00:25:19,080 --> 00:25:22,410
Mm hmm. And how soon was it? How long was it before you?

228
00:25:22,450 --> 00:25:29,440
Thought that that that our the data should be collecting data on the pandemic itself.

229
00:25:30,130 --> 00:25:36,880
Like the week after I went to South Africa and I was in South Africa for the rest of January

230
00:25:36,880 --> 00:25:42,910
and then in Tanzania for early February because my partner is doing research in Africa.

231
00:25:43,630 --> 00:25:50,100
And so we were further away from from like The Daily News and what's going on there.

232
00:25:50,110 --> 00:25:51,820
I was also further away just from work.

233
00:25:52,480 --> 00:26:01,959
Um, I was working from there, but less so like we were like in some remote places there and that I think like slowed things down.

234
00:26:01,960 --> 00:26:06,040
Like, I wish I obviously it was great to be in South Africa and Tanzania,

235
00:26:06,040 --> 00:26:14,620
but I think it slowed things down a bit and it was really only when I came back from there that we started working on it.

236
00:26:15,170 --> 00:26:23,829
And I was I was definitely worried when I was in Africa. I remember being worried about travels and I thought like, this guy has gone crazy.

237
00:26:23,830 --> 00:26:29,460
Like, what's he talking about? Why is he worried about travel suddenly?

238
00:26:29,560 --> 00:26:37,650
Um, and then I had actually told this story yesterday because we had like a similar conversation last night.

239
00:26:37,660 --> 00:26:45,040
Um, I think the real, like a real big moment was Marc Lipsitch, you know, the epidemiologist at Harvard.

240
00:26:45,340 --> 00:26:54,040
Um, he, he was at some point early February, he was saying he suspects now I don't want misquote him,

241
00:26:54,040 --> 00:27:01,810
but I think 70% of the world would eventually have COVID or some like majority of the world would have COVID at some point.

242
00:27:02,320 --> 00:27:06,790
And I like I was in bed reading this and I remember thinking like, that's just insane.

243
00:27:06,790 --> 00:27:12,550
Like, what's he talking about? But then it was I was thinking like, I wouldn't quite know why he would be wrong.

244
00:27:13,360 --> 00:27:22,659
So that, that like realisation that there's really nothing that stops this from going from the couple of hundred or a thousand cases

245
00:27:22,660 --> 00:27:35,230
currently in China to just billions of cases was when I like got worried but I wasn't still that wasn't quite the time when um,

246
00:27:35,680 --> 00:27:42,130
when we worked on it because we, we initially like, we thought it's not our job.

247
00:27:43,410 --> 00:27:49,000
I, I mean, we're talking about, I'm, uh, my research focus was poverty.

248
00:27:49,000 --> 00:27:53,559
It's not infectious diseases or pandemics. We had the bad luck.

249
00:27:53,560 --> 00:28:04,930
That's on the 1st of January 31st of December, I guess a colleague who is an expert on infectious diseases on influenza just left us.

250
00:28:05,740 --> 00:28:13,330
She took on a new job in Seattle. So we were also lacking someone in the team with the expertise.

251
00:28:13,330 --> 00:28:22,840
So we were thinking we shouldn't really like we should stay in our lane and should focus on those things that we understand and shouldn't.

252
00:28:23,110 --> 00:28:26,439
Yeah, like we shouldn't create any misinformation.

253
00:28:26,440 --> 00:28:34,150
We were just worried that we would do do something wrong. So we stayed like we stayed away from it until late in February.

254
00:28:35,920 --> 00:28:46,629
And then late in February there was, there was a weekend where I, I basically say then I was like, I didn't sleep much.

255
00:28:46,630 --> 00:28:53,590
Like I was just constantly working on the data, looking at like trying to understand what the researchers are saying.

256
00:28:53,590 --> 00:28:58,010
And I was I was thinking, this is going to be very, very, very, very bad.

257
00:28:58,210 --> 00:29:01,240
I, I, I got very worried at the end of February.

258
00:29:02,230 --> 00:29:06,790
And then we also and then I think and and then like,

259
00:29:07,310 --> 00:29:17,799
I think one crucial aspect was early on was that there was lots of news reporting on case counts here in the UK.

260
00:29:17,800 --> 00:29:31,150
There were these headlines of, um, there are four cases found in Bristol, there are three cases found in Bath or like these very low case counts.

261
00:29:31,150 --> 00:29:40,570
And I think the, the messaging from the media and from many commentators online was, don't worry, these numbers are tiny.

262
00:29:42,220 --> 00:29:48,310
But the crucial thing that was missing there was that there were also no tests.

263
00:29:48,880 --> 00:29:52,270
And if there are no tests, you also don't find cases.

264
00:29:52,270 --> 00:29:58,270
So nobody knew about asymptomatic transmission then I thought, yeah, exactly.

265
00:29:58,270 --> 00:30:08,020
Like and maybe the, the epidemiology is, was for others to comment on.

266
00:30:08,020 --> 00:30:14,170
But I thought like and other colleagues, Esteban Ortiz Ospina, a close colleague in the team,

267
00:30:15,730 --> 00:30:22,300
Joe Haskell, Hannah Ritchie and others, they were pointing out that around the around the testing.

268
00:30:22,400 --> 00:30:24,740
There's actually some data work to be done.

269
00:30:25,640 --> 00:30:33,020
Countries were starting to say something like they were like there was no spreadsheet with with with the number of tests,

270
00:30:33,020 --> 00:30:38,750
but they were like press statements or something from the health minister or something about like how many tests have been done.

271
00:30:40,610 --> 00:30:44,299
And so we thought, we need to communicate that. You won't find many.

272
00:30:44,300 --> 00:30:47,840
You won't find more cases than there are tests done and there are not many tests done.

273
00:30:47,840 --> 00:30:55,010
So I think that's the crucial information for you to know if you want to make sense of these low case counts.

274
00:30:55,580 --> 00:31:01,370
That was the one thing. And the other thing was just the power of of exponential growth.

275
00:31:01,640 --> 00:31:09,800
Exactly. Yes. I guess like there's this quote of the biggest flaw, the human mind, is that it's not able to understand exponential growth.

276
00:31:10,430 --> 00:31:17,389
And I think in the early days of oh, maybe throughout the pandemic, but especially in the early days of the pandemic, it was like on full display,

277
00:31:17,390 --> 00:31:28,190
just like how how that that knack of that understanding was was in the way of people understanding where this might be, might be going.

278
00:31:28,670 --> 00:31:37,460
And that that weekend was like the turning point. And then we shifted into, like, intense mode for the weeks after.

279
00:31:37,580 --> 00:31:41,090
Mm hmm. So you began by collecting data on tests.

280
00:31:41,090 --> 00:31:43,670
What what other data did you move on to collect?

281
00:31:44,000 --> 00:31:49,310
Yeah, the test was an early one, but I had like, that's another aspect that was, was like triggering this.

282
00:31:51,050 --> 00:32:01,490
The data that was published at the time was from the show was that it's the first source that we turn to and the and they and

283
00:32:01,490 --> 00:32:09,020
then we thought like one thing that we could usefully do is to turn like the way that they presented that data was in PDFs.

284
00:32:10,070 --> 00:32:13,400
So they had every day a status report or what they called it.

285
00:32:14,300 --> 00:32:19,850
And there was a table in that PDF just listing the number of cases in the various countries.

286
00:32:19,850 --> 00:32:28,310
And so we thought, well, like one thing that would make this a whole lot better is if someone copies these numbers from this spreadsheet,

287
00:32:28,400 --> 00:32:31,280
from these PDFs and puts them into a spreadsheet.

288
00:32:31,910 --> 00:32:38,750
And so Hannah and I were on the calls every morning, like Hannah Ritchie, my, my colleague, for many years.

289
00:32:39,200 --> 00:32:43,160
And we would just type up the the PDFs from the W.H.O.

290
00:32:43,580 --> 00:32:46,570
So she would be like, I don't know where she was.

291
00:32:46,580 --> 00:32:57,350
Maybe she was here or I know like they must have had a spreadsheet to get the PDF and it stayed like this for ages,

292
00:32:57,350 --> 00:33:02,299
like it was, it was crazy, but they just didn't have it in a spreadsheet online.

293
00:33:02,300 --> 00:33:07,129
And then she would be on the call with me and then she would be like. Philippines 12.

294
00:33:07,130 --> 00:33:11,540
Thailand four. Uh, China 3412.

295
00:33:11,540 --> 00:33:13,910
And I would just like typing these numbers every morning.

296
00:33:14,360 --> 00:33:20,030
And then like after we did this for like three days, I was maybe on the second day we realised that there were just like many flaws,

297
00:33:20,030 --> 00:33:29,059
like the number of cases, like the number of cumulative cases up to yesterday was higher than the number today or something like that.

298
00:33:29,060 --> 00:33:32,410
So just like obvious errors in the data.

299
00:33:32,420 --> 00:33:38,540
So then we got even more worries. And did you communicate with W.H.O.?

300
00:33:38,550 --> 00:33:46,640
Yeah, yeah, yeah, exactly. We like we wrote to them like like we stayed throughout the street and we stayed in close contact with the people there.

301
00:33:48,020 --> 00:33:51,889
Um, and then we also got in touch with the European CDC.

302
00:33:51,890 --> 00:33:55,580
They would, I think they were doing really excellent statistical work.

303
00:33:55,580 --> 00:33:56,870
The Centres for Disease Control.

304
00:33:57,080 --> 00:34:07,639
Yeah, because they are in Stockholm and they also reported case counts, um, death counts then globally or just the European,

305
00:34:07,640 --> 00:34:11,930
you know, globally that they had some of the data they only did for Europe.

306
00:34:11,930 --> 00:34:16,040
But the case in death comes, they did for a long time and they did it globally.

307
00:34:16,040 --> 00:34:18,500
And it was quite the heroic effort by those guys.

308
00:34:19,070 --> 00:34:28,010
They were up, I think in the in March we were having calls with them and they were always up like at 4:00 in Stockholm or something.

309
00:34:28,700 --> 00:34:35,179
And like that remained true for a very long time that the data publication in

310
00:34:35,180 --> 00:34:40,420
many countries was like much worse than you could imagine or you would hope.

311
00:34:41,390 --> 00:34:50,060
And so they would have to go to various different websites, to social media accounts, typing up the case codes and death counts.

312
00:34:50,060 --> 00:34:53,090
And the European CDC did a really good job there early on.

313
00:34:53,390 --> 00:35:00,890
And then we thought, okay, like now there's another opportunity for us to actually do some useful work where we take the data

314
00:35:00,890 --> 00:35:09,350
from the CDC and make it available in these tools that we have the web maps and line charts.

315
00:35:09,350 --> 00:35:11,090
And all of this, like what we talked about earlier,

316
00:35:11,440 --> 00:35:15,500
is doing this kind of extra step from the spreadsheet to do something where you can actually use it.

317
00:35:17,120 --> 00:35:21,230
But that was also not that easy. That was like maybe early March then.

318
00:35:22,530 --> 00:35:26,609
Like in February. I was still doing this with some other tools.

319
00:35:26,610 --> 00:35:29,700
I was using data recorder and Google spreadsheets,

320
00:35:29,700 --> 00:35:40,410
and I was doing this one off visualisations because our main tool that we intended to develop wasn't set up to report daily data.

321
00:35:41,220 --> 00:35:45,959
So a tour was set up to show changes over centuries or decades.

322
00:35:45,960 --> 00:35:55,500
And so this the shortest interval that the data a model could could handle was years, but not days.

323
00:35:56,250 --> 00:35:59,640
And we were a tiny team. We were six at the time.

324
00:36:01,170 --> 00:36:09,870
And the the engineers, the data at and the web developers in the team who were two at the beginning, they were saying, oh my God,

325
00:36:09,870 --> 00:36:14,909
this is going to be just going to take us a long time to turn this into a tool that

326
00:36:14,910 --> 00:36:19,290
can actually handle daily data and that which would work with with this demand.

327
00:36:19,830 --> 00:36:24,540
But then we had a new colleague just joining on the 1st of March. We hired him weeks before.

328
00:36:24,540 --> 00:36:33,370
So that was his first day of work. Brett Units and Brett came in and worked through the night or two nights like

329
00:36:33,390 --> 00:36:37,140
incredibly fast and turned this whole thing around and like actually made it possible,

330
00:36:37,140 --> 00:36:41,600
like, very hacky. But he made it possible to actually present daily data.

331
00:36:41,610 --> 00:36:49,650
There's still like there was like in the back end of our tour, there was like a checkbox that says something like a year is day.

332
00:36:50,010 --> 00:36:54,440
So and then the days are stored as like as years.

333
00:36:54,800 --> 00:36:57,810
So it was a very hacky solution, but it works.

334
00:36:57,810 --> 00:37:02,130
And then in early March, we, we did those things we like.

335
00:37:03,180 --> 00:37:08,399
We showed the testing data soon. I'm not quite sure when that happened, but early and in March,

336
00:37:08,400 --> 00:37:18,660
I guess we showed the European CDC data soon and in much of the communication we tried to emphasise exponential growth.

337
00:37:19,950 --> 00:37:23,069
There was another tool that Daniel got rid of, another engineer,

338
00:37:23,070 --> 00:37:38,820
another developing the team developed where we thought it's just hard to understand exponential growth as like in percentage increase per day.

339
00:37:39,240 --> 00:37:43,049
So we would focus on the doubling time. That was the idea at the time.

340
00:37:43,050 --> 00:37:44,640
That's what I always imagine.

341
00:37:44,640 --> 00:37:52,710
Lily Somebody talks about lily ponds and it's how you your lily pond gets to have COVID and then the next day it's fully covered.

342
00:37:53,130 --> 00:37:57,150
And so that's how I cope with exponential growth.

343
00:37:58,230 --> 00:38:02,940
Yeah, it's just so much more intuitive that you go for 50%, 100% in one go and right?

344
00:38:02,940 --> 00:38:08,309
Yeah. Yeah, exactly. So then he built this tool that would reading the data every day.

345
00:38:08,310 --> 00:38:11,280
And then for all the countries in the world, it would show the current doubling time.

346
00:38:11,280 --> 00:38:16,709
And we would see like it's five days here, seven days here, and you would see these double to date.

347
00:38:16,710 --> 00:38:20,440
And times are short. If. If. Hmm.

348
00:38:20,760 --> 00:38:25,270
Mm hmm. And did you have to write?

349
00:38:25,350 --> 00:38:29,430
Were you able to raise more funding or did you get more funding to increase the size of your team?

350
00:38:29,520 --> 00:38:34,389
Yeah. So we were like funding was quite crucial, I guess.

351
00:38:34,390 --> 00:38:41,300
So sort of like thinking about it in this, uh, um, like university discourse that we're in,

352
00:38:41,850 --> 00:38:47,160
um, because it was really crucial was that we didn't have restricted funding.

353
00:38:48,000 --> 00:38:57,329
So we were always keen to not allow funders to interfere with what we want to do with our data.

354
00:38:57,330 --> 00:39:02,210
And that was really crucial at the beginning of the pandemic because it allowed us to shift from one day to the other.

355
00:39:02,650 --> 00:39:03,200
Yeah, yeah.

356
00:39:03,270 --> 00:39:12,839
So we didn't have any like deliverables or something that were like lined up where we would have to finish some paper or a report or something.

357
00:39:12,840 --> 00:39:22,560
So there was like we were just free to say, okay, we're not doing climate change for next month at least we're now doing, we're not doing COVID.

358
00:39:23,540 --> 00:39:27,200
Um, so that was, that was good.

359
00:39:27,600 --> 00:39:28,739
Like, the funders were just great.

360
00:39:28,740 --> 00:39:40,110
They're, um, which now helps me to make this case to future funders and that, that we would get unrestricted funding.

361
00:39:41,460 --> 00:39:48,870
Um, and then some, um, some more funding came in like during the following months.

362
00:39:49,110 --> 00:39:52,020
But it took a bit of obviously like the main,

363
00:39:53,040 --> 00:40:00,600
the main focus was just to work on the data and make sense of the data, clean the data, find the best sources,

364
00:40:00,810 --> 00:40:11,280
compare different sources against each other, try to speak with epidemiologists, learn from them whether what we are doing is sensible.

365
00:40:11,280 --> 00:40:19,220
Like what I said earlier, that we were worried that we don't have an expert. So we were like lots of epidemiologists were just really helpful.

366
00:40:19,230 --> 00:40:23,410
Natalie Dean I remember she was she was great. And. Giving us feedback.

367
00:40:23,980 --> 00:40:27,340
She was in Miami at the time. I think she's now at a different university.

368
00:40:27,640 --> 00:40:30,820
Eric Topol was was great early on helping us.

369
00:40:30,820 --> 00:40:35,290
Adam Kucharski from London School of Hygiene and Tropical Medicine.

370
00:40:36,070 --> 00:40:37,360
Jason Hendry.

371
00:40:37,360 --> 00:40:46,490
Moritz Kramer Like lots of people who are experts in the field, Benedetta is herself Benedetta that a native who left us just a few months earlier.

372
00:40:46,900 --> 00:40:53,170
All of them were really great in giving us feedback where we were trying to make sure that we are not spreading any misinformation,

373
00:40:53,170 --> 00:40:56,739
that what we are doing is correct. And it was that typical of the way you worked?

374
00:40:56,740 --> 00:41:02,290
I have. Another thing I'm interested in is the extent to which work that was undertaken as a result

375
00:41:02,290 --> 00:41:07,690
of the pandemic was more or less collaborative than the way you'd worked previously.

376
00:41:09,400 --> 00:41:12,430
So have you always been in constant touch with experts?

377
00:41:12,640 --> 00:41:15,790
Yeah, I think that's true. Like, I mean, we were always a small thing.

378
00:41:15,790 --> 00:41:20,660
We're still a small team, and so we can't be experts in all these different subjects and the like.

379
00:41:20,680 --> 00:41:23,709
That's, I think, the big advantage from being orthodox of Martin School,

380
00:41:23,710 --> 00:41:29,200
where it's this interdisciplinary department that brings people together from many different disciplines.

381
00:41:29,920 --> 00:41:38,530
It's also the advantage of being at Oxford, where you just have experts on pretty much anything that you ever would need an expert on.

382
00:41:38,530 --> 00:41:42,129
So we are we always ask people you just mentioned we're not in Oxford.

383
00:41:42,130 --> 00:41:45,280
Those were all elsewhere. Some of them were they? Moros were us.

384
00:41:45,580 --> 00:41:48,760
Jason was serious. Some, some some were here.

385
00:41:51,580 --> 00:41:54,730
And I'm sure, like, I forget really important people who helped us.

386
00:41:58,800 --> 00:42:03,900
Do that. I think you answered the question that that that collaborative way of working was typical of the way you worked anyway.

387
00:42:03,900 --> 00:42:06,510
So it wasn't as if it was a new thing for you. No.

388
00:42:06,750 --> 00:42:15,100
Yeah, but we were definitely like it was just that everything was just faster and way more intense than ever before.

389
00:42:15,120 --> 00:42:20,339
Like in Mojave. Like I. But often slept in the office.

390
00:42:20,340 --> 00:42:26,250
Like, because, like, I would wake up in the middle of the night and think of something new.

391
00:42:26,370 --> 00:42:28,919
I would try to minimise the hours of sleep that I had.

392
00:42:28,920 --> 00:42:39,150
So I was for like I would put a mattress in next to my desk and just like crash there for a couple of hours when, uh, when I was too tired.

393
00:42:39,870 --> 00:42:43,680
But I would do the building not close. I mean, were you able to come into the.

394
00:42:43,800 --> 00:42:47,220
No, no, I was at home. Oh, yeah. I have an officer home. Yeah.

395
00:42:47,430 --> 00:42:51,000
I mean, yes. Yeah. Like there's a like a guest room kind of.

396
00:42:51,030 --> 00:42:54,720
Right. And I put the the desk there and that and put a mattress there.

397
00:42:55,380 --> 00:43:01,860
Uh, actually, I had enough from. From the racking up in the middle of the night to do our not fighting sleep like.

398
00:43:01,860 --> 00:43:05,550
Yeah. So it would just, like, physically live in that, in that room.

399
00:43:05,730 --> 00:43:11,850
Mm hmm. And, yeah, spending a lot of time on calls, presumably on.

400
00:43:12,240 --> 00:43:13,920
Yeah. Yeah, exactly. There was, like, lots of.

401
00:43:13,920 --> 00:43:20,489
Cause, like, there was also a, like a a difficult thing was that we just hired we were six before the pandemic.

402
00:43:20,490 --> 00:43:24,870
We just hired four colleagues or so. So we would from 6 to 10.

403
00:43:24,870 --> 00:43:29,790
That's I mean, that's almost doubling. So there was also a lot of growth in the team.

404
00:43:29,790 --> 00:43:35,819
We had to onboard the new colleagues. They were entering the team at this moment of like absolute madness.

405
00:43:35,820 --> 00:43:40,680
So that that came on top of it. Some of them were in the US.

406
00:43:40,680 --> 00:43:45,690
So basically my day was always in like two shifts where I would do the mornings.

407
00:43:45,900 --> 00:43:49,650
Hannah wakes up at like four or five, she works up, wakes up really early.

408
00:43:49,650 --> 00:43:57,540
So I would work with her in the mornings. Then after lunch I would sleep for 2 hours and then I would do a night shift

409
00:43:57,540 --> 00:44:03,899
with the people who are in Americas and many calls trying to coordinate that,

410
00:44:03,900 --> 00:44:07,170
that everything's moving forward. And. And this could work.

411
00:44:07,320 --> 00:44:22,380
Mm hmm. Mm hmm. And yes, I mean, did you go beyond the testing and the and the the cases and deaths data?

412
00:44:22,390 --> 00:44:26,260
What other kinds of kinds of data did you built into your system?

413
00:44:26,320 --> 00:44:30,520
Yeah, so that grew over time.

414
00:44:32,470 --> 00:44:40,120
I think in another early Oxford project where we worked together here with many colleagues from the university was on policy measures.

415
00:44:40,840 --> 00:44:46,270
That's a team that's based at the Blavatnik School of Government, although the Global Governance Tracker.

416
00:44:46,300 --> 00:44:49,450
Yeah, yes, yes. I've interviewed under Patrick. Oh, yeah.

417
00:44:49,480 --> 00:44:54,520
Yes. And I had a question specifically about that. But now you've you've pre-empted me now.

418
00:44:54,520 --> 00:44:56,950
No, I mean, yes, I was interested in the extent to which.

419
00:44:57,430 --> 00:45:05,500
Apart from that one, because there are other people who collect and present data like Johns Hopkins, for example.

420
00:45:06,010 --> 00:45:10,180
What how did you fit into this kind of ecosystem of data aggregators?

421
00:45:10,880 --> 00:45:14,500
Yeah. So Johns Hopkins.

422
00:45:14,500 --> 00:45:22,090
Johns Hopkins, they were also early. We didn't use them early because we thought the quality, I guess, of European CDC was just higher.

423
00:45:22,930 --> 00:45:26,620
Um, Johns Hopkins was always done by volunteers. Oh, really?

424
00:45:26,620 --> 00:45:34,870
I didn't realise that. Yeah, like, it's. I mean, they have a team there that was coordinating the work at like at the university.

425
00:45:34,870 --> 00:45:43,010
But many of the contributions and the updates of the data, the changes of the data, they were all coordinated through GitHub and there was, um,

426
00:45:44,020 --> 00:45:47,469
and I think overall they did an amazing job,

427
00:45:47,470 --> 00:45:54,010
but I think early on we were sometimes a bit worried that they weren't as responsive as we would have wanted.

428
00:45:54,010 --> 00:45:58,600
And we thought that the European CDC was more on top of things and thought they did a better job.

429
00:45:59,740 --> 00:46:08,530
Um, we later then switched to the Johns Hopkins data when the European CDC stopped doing daily updates.

430
00:46:08,720 --> 00:46:14,140
Right. But that was more than a year later.

431
00:46:14,170 --> 00:46:20,130
Yeah. Now we're we're data from that coordinated by the Johns Hopkins colleagues.

432
00:46:21,220 --> 00:46:31,540
And early on, what like we definitely did a lot of data then on like variations of the same.

433
00:46:31,540 --> 00:46:35,079
Like, I mean sometimes these things are so trivial,

434
00:46:35,080 --> 00:46:41,800
like the case counts and the death counts that are published by the European CDC were daily figures,

435
00:46:42,520 --> 00:46:49,809
but there was there was obviously part, obviously, but there was less reporting on weekends because people had fewer tests.

436
00:46:49,810 --> 00:46:54,490
So you would have this these weekly patterns in the data.

437
00:46:54,490 --> 00:47:01,660
So you would want to smooth this data and look at a seven day moving average or a 14 day moving average.

438
00:47:02,290 --> 00:47:10,540
And overall, it's like a bit odd maybe for people who are familiar with working with data and statistics,

439
00:47:10,540 --> 00:47:17,950
but for especially for journalists, that's very helpful if someone actually turns the daily figures into a seven day moving average.

440
00:47:18,460 --> 00:47:24,310
So even these small things of presenting a seven day moving average of the same metric

441
00:47:24,310 --> 00:47:29,440
is is helpful for many journalists because they can they can then directly use it.

442
00:47:30,640 --> 00:47:36,100
Another crucial thing there was to always publish the data in the same place,

443
00:47:36,610 --> 00:47:43,120
because lots of the data sources would constantly shift and like the UK I think eventually did.

444
00:47:43,120 --> 00:47:47,919
Amazing in presenting that data, what it like was one of the best countries in presenting the data,

445
00:47:47,920 --> 00:47:52,660
but early on it was shifting around quite a lot and it was difficult for journalists or

446
00:47:53,230 --> 00:47:58,840
epidemiologists also to access the data because suddenly it has shifted to another place.

447
00:47:58,840 --> 00:48:04,780
The format has changed. The the column labels in the in the spreadsheet have changed.

448
00:48:04,780 --> 00:48:12,939
And so if you have the column labelled hardcoded in your in your analysis, then it just messes up with your analysis.

449
00:48:12,940 --> 00:48:17,560
So like standardising all of that data, that was that was very crucial early on.

450
00:48:18,430 --> 00:48:25,899
But you asking about other data that we saw like exactly like we did the tests and then a thing that

451
00:48:25,900 --> 00:48:32,380
we spend a lot of energy on is trying to get people to understand the importance of the positive rate.

452
00:48:33,010 --> 00:48:40,360
So the share of tests that come back positive as an indicator for whether a country is testing enough or not,

453
00:48:41,560 --> 00:48:52,000
where the basic idea is, if a lot of your tests come back positive, then you might miss a ton of other tests.

454
00:48:52,540 --> 00:49:01,180
And um, when on the other hand the positive rate is very low, then it means your testing is, is at a large scale and relative to the, to the outbreak.

455
00:49:01,630 --> 00:49:09,610
And we try to spend a lot of energy explaining the value of that metric, which took a really long time.

456
00:49:10,600 --> 00:49:20,390
Um. And then these days we're publishing many, many more metrics, but those were added later.

457
00:49:20,400 --> 00:49:29,300
I think the next big change was at the end of March, maybe when the first excess mortality data became available.

458
00:49:30,240 --> 00:49:37,760
Whether that was from NASA, was it here for the UK like I think at the beginning of the people who have especially interested in Italy?

459
00:49:38,180 --> 00:49:44,050
Oh, yes. Just because that was two weeks or so happening earlier than than elsewhere.

460
00:49:44,060 --> 00:49:50,780
So then there were like the first reports coming out of Lombardy and the regions there in the north of Italy,

461
00:49:51,380 --> 00:49:54,320
suggesting that excess deaths were just very high.

462
00:49:55,490 --> 00:50:01,700
And so the number of confirmed deaths of confirmed cases was a huge undercount of what was actually going on.

463
00:50:03,290 --> 00:50:10,159
That's gotten a lot of attention. And then that data became available from more and more countries.

464
00:50:10,160 --> 00:50:14,510
And we worked with colleagues here at the university analysing this data.

465
00:50:15,050 --> 00:50:24,430
John Mutambara and Janine Aron Um, we were then coordinating also with journalists.

466
00:50:24,440 --> 00:50:35,059
I mean that's also a very interesting aspect to think about that a lot of this early data was done by journalists and which was I mean,

467
00:50:35,060 --> 00:50:39,709
which is really frustrating. It's, it should really be the W.H.O. who's on top of it.

468
00:50:39,710 --> 00:50:45,710
But it was in many cases journalists who did amazing work.

469
00:50:46,310 --> 00:50:57,049
I just yeah. Like and at that, in addition to communicating after work, they just collated a lot of data and and did a really amazing work.

470
00:50:57,050 --> 00:51:05,960
Amazing job. And it was people like The Economist I thought was, was really great because they were publishing everything, open access immediately.

471
00:51:06,970 --> 00:51:12,480
Um, so they would also allow others to actually scrutinise the quality of the data.

472
00:51:13,130 --> 00:51:16,010
They made it available to others to access the data.

473
00:51:16,550 --> 00:51:23,600
The New York Times, that amazing in the US bringing together the data like even early in February, I think not quite sure about that,

474
00:51:23,600 --> 00:51:33,500
but they started very early bringing together the data because the US was lacking an agency that would bring together, um, countrywide statistics.

475
00:51:34,370 --> 00:51:37,849
Um, and that was also then true for the access.

476
00:51:37,850 --> 00:51:43,999
That's where there were many calls with the people from The Economist trying to make sense of that data.

477
00:51:44,000 --> 00:51:47,770
And that was then a big theme in going into April.

478
00:51:49,760 --> 00:51:55,340
And then the next really big change was vaccinations.

479
00:51:56,570 --> 00:52:06,560
So that that was I think that was. So then the summer came, we like things were quiet where we're becoming a bit more quiet.

480
00:52:06,990 --> 00:52:12,950
Um, I was able to go home and see my parents family in Germany.

481
00:52:12,980 --> 00:52:21,920
Um, I had a bit of a break. Um, but we were very sure that the next winter was going to, was going to be tough.

482
00:52:21,920 --> 00:52:28,370
So we were like spending the summer preparing our tours and like getting everything cleaned up so that we would be ready for the next winter.

483
00:52:29,720 --> 00:52:32,930
And then in the next winter things were pretty intense again.

484
00:52:32,930 --> 00:52:37,040
So we were like very tired, everyone in the team.

485
00:52:37,040 --> 00:52:44,150
And then on just before Christmas, like I wasn't like just reporting deaths and disease all day long.

486
00:52:45,990 --> 00:52:56,149
It was like just not nice. So just before Christmas, then finally the vaccines were approved and we were just excited that this was happening.

487
00:52:56,150 --> 00:53:04,160
So we thought, Oh, amazing, let's do this nice map where we can actually use the colour green instead of red.

488
00:53:04,160 --> 00:53:09,920
And we showed that like now the first vaccine there was this elderly woman here in the UK.

489
00:53:09,950 --> 00:53:14,779
Yeah, I remember like December 2020 as a, we only had like one data point.

490
00:53:14,780 --> 00:53:23,179
It would like this, we all just got the vaccine, but we were not thinking that this would become a big theme then.

491
00:53:23,180 --> 00:53:27,139
Edwards And what Matua is, is now the head of the data team.

492
00:53:27,140 --> 00:53:31,040
He joined us early in 2020. He was a friend of mine.

493
00:53:31,040 --> 00:53:40,489
He's a friend of mine, but he was a friend before. And he like he was like he has had quite the extraordinary 2020 that he got in touch with me.

494
00:53:40,490 --> 00:53:44,600
And in March 2020, he was saying, Hey, I'm seeing that you're doing a lot of this COVID work.

495
00:53:44,990 --> 00:53:49,040
I'm locked. I'm in lockdown in in Paris.

496
00:53:50,030 --> 00:53:53,509
I have time. I know how to work with data.

497
00:53:53,510 --> 00:53:57,050
If you have anything that you need my help with, just let me know.

498
00:53:58,250 --> 00:54:05,540
And then two weeks later, actually, I think on the testing data, we were like, yeah, we actually hopefully if we would have some, some help on that.

499
00:54:05,540 --> 00:54:12,210
So anyway, I started working in maybe March or April to.

500
00:54:12,600 --> 00:54:17,429
Early 20:20 a.m. And he was amazing.

501
00:54:17,430 --> 00:54:21,870
He was like from the start, we did an amazing job coordinating the work of others,

502
00:54:22,320 --> 00:54:30,809
like just doing very high quality work and communicating it also very well to journalists,

503
00:54:30,810 --> 00:54:34,620
to social media, to epidemiologists that were relying on us.

504
00:54:35,340 --> 00:54:38,969
And so he did more and more, and he ended up doing two full time jobs.

505
00:54:38,970 --> 00:54:46,290
So he had like this job from 8 to 4 for his like regular job in in France and then from four to midnight.

506
00:54:46,290 --> 00:54:51,239
Or is that. Yeah, I guess that's another 8 hours.

507
00:54:51,240 --> 00:54:55,650
He would work with us. So we had like full two full time jobs and he was just working constantly.

508
00:54:56,070 --> 00:55:03,690
So he was also retired from Libya. At some point. He obviously dropped out that first job, but he still kept on working a lot.

509
00:55:04,080 --> 00:55:06,149
And then at night, end of the year, when I was saying that,

510
00:55:06,150 --> 00:55:10,530
we were like happy that we could finally report some positive developments and we had these vaccines.

511
00:55:10,860 --> 00:55:16,740
He was like, Ah, we should really maybe think about bringing the vaccinations data together.

512
00:55:16,770 --> 00:55:25,470
Like obviously it will be soon. The World Health Organisation that that will release this data, they do the COVAX initiative and so on.

513
00:55:25,980 --> 00:55:34,170
So it'll be then, but we could just start and do a weekly update so it's not going to be too intense.

514
00:55:34,170 --> 00:55:40,379
We just do it weekly and I have for some stupid reason, thought that this was like a reasonable approach.

515
00:55:40,380 --> 00:55:51,300
So I said, Yeah, let's do that. Sounds fun. And then it became probably one of the most intense periods like early 20, 21 was,

516
00:55:51,480 --> 00:55:59,310
was very intense because this vaccine rollout happened much more categorically than people had hoped.

517
00:56:00,000 --> 00:56:04,050
There was many, many rich countries did very poorly.

518
00:56:05,190 --> 00:56:09,390
My home country, Germany, I was very frustrated by how slowly everything happened.

519
00:56:10,650 --> 00:56:15,120
There was this story that Israel did, this incredibly fast rollout.

520
00:56:15,120 --> 00:56:18,090
So there was this huge discrepancy which was like constantly in the headlines.

521
00:56:18,600 --> 00:56:24,330
And we had started putting this data together and suddenly we were responsible for that job.

522
00:56:24,750 --> 00:56:27,750
That was like one thing like during the pandemic that we learned.

523
00:56:27,750 --> 00:56:29,729
Like whenever you start doing something,

524
00:56:29,730 --> 00:56:37,920
then the the first minute people are thankful that you do it and the second minute they demand that you do it and in the perfect quality all the time,

525
00:56:37,920 --> 00:56:52,680
every day without a break. And if the data of some small country isn't updated Sunday at 10:00, then you get in like, yeah, you get emails like often.

526
00:56:52,690 --> 00:56:58,709
Also like I think we all sort of struggled because these emails were often not nice where people are

527
00:56:58,710 --> 00:57:06,360
just like everything was so political that people would see some political motive in whatever you do.

528
00:57:06,360 --> 00:57:12,599
Like if you've, if you've updated, if you've not updated the data for that country in the last 4 hours,

529
00:57:12,600 --> 00:57:21,299
then it's because you have this vendetta against this particular country and you don't like their right wing or left wing government.

530
00:57:21,300 --> 00:57:24,150
And like I have no idea who was even in power in these places.

531
00:57:24,810 --> 00:57:29,670
It's really far from the truth, but they would to see some kind of motive in whatever is happening.

532
00:57:29,670 --> 00:57:32,729
And if you would update the data, then the other people would be like, Oh,

533
00:57:32,730 --> 00:57:37,170
you're just updating it because you want to let them shine in the best light.

534
00:57:37,170 --> 00:57:44,670
Or it was like so much of this kind of emails is coming our way and social media of just people being angry for no good reason.

535
00:57:46,410 --> 00:57:48,960
And that happened a lot. And this vaccine work.

536
00:57:49,980 --> 00:57:55,920
And then weirdly, the WTO just never got a lot, got around to actually publishing this data on the vaccine.

537
00:57:55,920 --> 00:58:07,530
So for many, many weeks, it was Edward sitting in his in his kitchen in Paris building this data that everyone relied on,

538
00:58:07,980 --> 00:58:12,000
that every newspaper in the world was relying on his data.

539
00:58:12,750 --> 00:58:19,230
And it was he was getting it from individual health services, health executive departments.

540
00:58:19,410 --> 00:58:28,469
And if so, he would get it like from like he would spend the morning or the day just going through all of the different websites,

541
00:58:28,470 --> 00:58:35,129
which you would hope is there some place where you can actually find the data in the spreadsheet and you could just read it in?

542
00:58:35,130 --> 00:58:39,930
And some countries obviously did that. And then he built this great place that would take the data out of the spreadsheet

543
00:58:39,930 --> 00:58:43,140
and put it in a spreadsheet and job was done and you can automate that.

544
00:58:43,590 --> 00:58:53,340
But many countries didn't do it like that. So then it was there were places where they just released the vaccine counts in press conferences.

545
00:58:53,760 --> 00:59:01,260
So then he's sitting there listening to a press conference in Hungarian of the assistant to the Health Minister.

546
00:59:01,260 --> 00:59:07,470
And you have to figure out what the word for second dose in Hungarian is so that

547
00:59:08,160 --> 00:59:12,150
you can hopefully find something like a slide or something in that because.

548
00:59:12,220 --> 00:59:19,120
Taken to type up the number, or it'd be countries where they would only release the numbers in a photo,

549
00:59:19,120 --> 00:59:23,910
like in a kind of like a screenshot of, of a table in social media.

550
00:59:23,920 --> 00:59:29,200
So you would need to know the Facebook page of the health minister of that country,

551
00:59:29,200 --> 00:59:34,059
because that's where that country is, that the only place where that country would release the numbers.

552
00:59:34,060 --> 00:59:37,210
So he would like piece that together from all these various places.

553
00:59:37,660 --> 00:59:49,630
And that's how he'd remained for the first quarter of 2021, which was the time when the the infection rates were going up a lot as well.

554
00:59:50,140 --> 00:59:59,710
Yeah, yes. I'd say that there was this big wave here in the UK just after New Years in particular, I think very early January 2021.

555
01:00:00,550 --> 01:00:05,830
And we have the data here. Yes, I'm sure you do. You can look it up how this works.

556
01:00:05,830 --> 01:00:12,610
But yeah. And then it was this one.

557
01:00:13,060 --> 01:00:17,500
Yes. Yeah. So like it peaked on January 10th, so.

558
01:00:19,060 --> 01:00:22,780
Yeah. And here it's like during Christmas was this underreporting for sure right away.

559
01:00:22,780 --> 01:00:31,510
Like just few tests were done. But there was this. Yeah, first massive wave and second massive wave.

560
01:00:33,010 --> 01:00:40,120
And then after that there was a lot of work with the vacc, the various variants that became a big theme.

561
01:00:41,200 --> 01:00:59,499
Um, where we were like one variant was following the next and it was this covariance dot org team that made this data like the,

562
01:00:59,500 --> 01:01:06,970
presented this data based on the G as a IDE, uh, sets on these, on these variants.

563
01:01:07,630 --> 01:01:14,380
And then access steps became like obviously also a big state, a big one in excess deaths.

564
01:01:16,270 --> 01:01:19,419
I think one of the, like I was praising The Economist earlier,

565
01:01:19,420 --> 01:01:24,430
but like I think the most useful thing that they did was to present estimates

566
01:01:24,430 --> 01:01:33,070
of excess deaths for countries around the world and call it they're quotes.

567
01:01:33,190 --> 01:01:43,749
Sandra in uh at the economist did this this project where they tried to estimate how many how many

568
01:01:43,750 --> 01:01:50,889
people and you try to estimate a baseline of how many deaths you would expect in normal years.

569
01:01:50,890 --> 01:01:59,800
And you try to then, um, calculate the difference between what you observing and what you are, what you, what you would expect.

570
01:02:00,710 --> 01:02:03,910
And those excess deaths are available for,

571
01:02:04,090 --> 01:02:10,219
for many rich countries because in rich countries there are good records of the number of deaths from past years.

572
01:02:10,220 --> 01:02:20,050
So you can use that as a baseline to compare it with. But most countries in the world don't actually have these despite those statistics,

573
01:02:21,010 --> 01:02:28,059
and often the majority of deaths aren't actually recorded at the births on record as the number of people are unreported.

574
01:02:28,060 --> 01:02:32,070
So like all of that data is lacking. So you can't calculate excess deaths.

575
01:02:32,270 --> 01:02:34,030
It's not trivial to estimate that,

576
01:02:34,450 --> 01:02:41,020
but I think it was very crucial to to do that because the world was focusing very much on the number of confirmed deaths.

577
01:02:41,650 --> 01:02:50,799
And in rich countries, that's a reasonable number to know, like how many people have actually died due to the pandemic.

578
01:02:50,800 --> 01:02:55,270
But in many poor countries, it's a huge undercount, surely,

579
01:02:55,270 --> 01:03:01,870
of the number of people that have actually died because the testing is so low, because the vital statistics on there.

580
01:03:02,290 --> 01:03:05,739
And so these estimates of excess mortality, I think were very crucial.

581
01:03:05,740 --> 01:03:13,060
I think it was like one of the maybe yeah, I think it was one of the most important projects throughout the pandemic that

582
01:03:13,450 --> 01:03:18,520
he and his colleagues did this to these estimates of excess deaths throughout.

583
01:03:19,450 --> 01:03:28,719
And it's still frustrating. Like I have now seen several emails to the people at The Guardian because if you go to the Guardian website,

584
01:03:28,720 --> 01:03:32,170
they have a tracker of how the pandemic is going.

585
01:03:32,710 --> 01:03:46,090
Let's make sure that I'm not spreading misinformation, but I think it's still like this COVID tracker, uh, corona virus here.

586
01:03:48,070 --> 01:03:52,040
Which countries have the latest global COVID data?

587
01:03:52,240 --> 01:03:56,469
Total deaths, 6.5 million. And that's just the light. Like they're just spreading lies.

588
01:03:56,470 --> 01:04:04,540
And I told them many times that this is not true. Like there were obviously many more deaths than the 6.5 million, with 6.5 million confirmed deaths.

589
01:04:05,230 --> 01:04:10,390
That's the case. But confirmed deaths are a fraction of the number of total deaths in the world.

590
01:04:10,900 --> 01:04:16,710
Because in the most poor. Countries. A few of the deaths are only recorded.

591
01:04:17,070 --> 01:04:23,100
And so I think throughout the pandemic, because of these kind of practices from from journalists in particular,

592
01:04:24,990 --> 01:04:30,120
the world has been just misinformed of the severity of of what was actually happening.

593
01:04:30,600 --> 01:04:36,749
If you compare it with these estimates on excess deaths that are now available from three major sources,

594
01:04:36,750 --> 01:04:39,330
the economist that does these weekly updates of it,

595
01:04:39,660 --> 01:04:50,130
the W.H.O. and the NIH and here in Washington, all of their estimates are around 23 million deaths or so.

596
01:04:50,580 --> 01:04:59,920
So like news outlets like The Guardian, they are under understate the severity of of what was actually happening by.

597
01:05:00,600 --> 01:05:06,550
Yeah. Like it's it's a it's a quarter of what they're actually reporting as the number of deaths.

598
01:05:07,930 --> 01:05:13,979
What I mean the next map they on this website is just as much of a nonsense like which countries currently

599
01:05:13,980 --> 01:05:18,890
have the highest case rates you can't know this from this is no testing exactly as again there's no testing.

600
01:05:18,930 --> 01:05:23,120
Like it's it's crazy. Like how which countries currently have the highest death rates?

601
01:05:23,130 --> 01:05:29,459
It's just not true that no one is dying in Africa. No. And and I think because of this because of this information,

602
01:05:29,460 --> 01:05:33,360
there's also not enough initiative to actually share the vaccines with with

603
01:05:33,360 --> 01:05:37,710
the poorer countries that actually the new data is coming from Johns Hopkins.

604
01:05:38,130 --> 01:05:41,940
Yeah, but it's it's just the number of confirmed deaths. So they need to explain that.

605
01:05:41,960 --> 01:05:47,100
Yes. Mercifully, that is that these numbers are just a fraction of what's going on.

606
01:05:48,930 --> 01:05:53,100
So these excess deaths, I thought, were like very valuable projects.

607
01:05:54,960 --> 01:05:57,960
And that's that's included now in your in your data. Yeah.

608
01:05:58,110 --> 01:06:01,200
Yeah. Which which is great that they make it publicly available.

609
01:06:01,620 --> 01:06:05,639
Mm hmm. So you don't I interrupted very rudely when you started talking about your

610
01:06:05,640 --> 01:06:08,880
collaboration with the people at Blavatnik and the Government Response Tracker.

611
01:06:09,210 --> 01:06:19,050
Oh, yeah. They're, um, yeah, they did this very helpful data bringing together all of the policy interventions that were happening at the time,

612
01:06:19,050 --> 01:06:28,740
which countries closed their schools, which countries restricted large gatherings and they made it all available for on their website.

613
01:06:28,740 --> 01:06:34,020
They benefited a lot from the, uh, alumni community of the Blavatnik School.

614
01:06:34,020 --> 01:06:40,110
They have people in all corners of the world, so they would report from their home countries and what's happening.

615
01:06:40,940 --> 01:06:49,380
Um, so it'd be possible for somebody to look at what they collected about what governments did with what you collected about death rates,

616
01:06:49,680 --> 01:06:55,410
you know, case rates and make draw some conclusions about the effectiveness of policy making.

617
01:06:55,770 --> 01:07:02,549
Yes, exactly. Like and that happened a lot, uh, during the during the pandemic.

618
01:07:02,550 --> 01:07:05,820
There was, I was part of this group I have to look up.

619
01:07:06,600 --> 01:07:12,840
They had some complicated acronym. Um, I was it of.

620
01:07:15,890 --> 01:07:22,500
First. Oh, is it called?

621
01:07:26,560 --> 01:07:35,500
Like we had this working group with, with a number of epidemiologists coordinated by the UK government.

622
01:07:36,310 --> 01:07:43,330
That was, that was very much trying to, to do those analysis of what can we actually learn from,

623
01:07:43,420 --> 01:07:47,050
from how other countries are reacting to the pandemic.

624
01:07:48,610 --> 01:08:01,660
And they would very much rely on the data from from their colleagues at the School of Government and don't quite know where the name is.

625
01:08:03,170 --> 01:08:13,570
Yeah. So, I mean, how would you assess how well policymakers around the world used the data that was available?

626
01:08:14,830 --> 01:08:23,380
Because there were as I said, there were a number of groups who worked extremely hard to aggregate data in in different ways.

627
01:08:23,450 --> 01:08:27,580
So there was in a way, there was no shortage of no deficit of information.

628
01:08:29,260 --> 01:08:32,829
But policymaking was was very different.

629
01:08:32,830 --> 01:08:36,370
And policy decisions were very different in different countries. That's right.

630
01:08:36,910 --> 01:08:42,550
Yeah. Well, so I was kind of trying to make the point that this was was going well in many ways.

631
01:08:43,270 --> 01:08:50,560
I don't know. It was also obviously frustrating. Like I was very frustrated with Germany, which I thought did quite well early on in the pandemic,

632
01:08:51,730 --> 01:08:58,600
especially compared with other countries were especially that took testing very seriously scaled up testing early on.

633
01:08:59,290 --> 01:09:03,159
Um developed some of the tests in, in Germany.

634
01:09:03,160 --> 01:09:09,550
Just used them early on. Um, and one of the vaccines was developed in a German lab.

635
01:09:09,640 --> 01:09:13,000
Oh yeah. That's like local patriotism. Like, that's just done this.

636
01:09:13,210 --> 01:09:17,740
That's very close from where I grew up. Their own minds and.

637
01:09:17,950 --> 01:09:21,819
And the place where I grew up is 20 minutes from there. Exactly.

638
01:09:21,820 --> 01:09:33,880
Like not a big. The company is called Biontech, founded by the daughter and son of Turkish immigrants who came to to the area where I'm from,

639
01:09:34,630 --> 01:09:43,780
uh, a couple of decades ago and it was developed by then, but they needed a partner that could produce them at large scale.

640
01:09:43,780 --> 01:09:52,809
So they worked with Pfizer like it's mostly known as the Pfizer vaccine and trying to keep the patriotism going.

641
01:09:52,810 --> 01:09:57,520
And and they keep on referring to it that the biontech vaccine.

642
01:09:57,520 --> 01:10:03,100
But he added they did amazingly well then build one of these first MRA vaccines.

643
01:10:04,090 --> 01:10:08,559
But it was the actual approval of the vaccination rollout that Germany was slow on that.

644
01:10:08,560 --> 01:10:16,690
Right. Um, the, the approval I think was, was with the other European countries and that was like the Europeans.

645
01:10:17,170 --> 01:10:22,180
It was an EU. Yeah. The European Medical Agency is what it's called.

646
01:10:23,350 --> 01:10:26,409
Um, uh, no.

647
01:10:26,410 --> 01:10:34,770
But they was like, I think what was frustrating on, on the vaccine rollout was that they hadn't purchased many vaccines before,

648
01:10:35,110 --> 01:10:40,330
before the vaccine was approved and they were too hesitant, like also like in coordination with the European Union.

649
01:10:42,970 --> 01:10:54,010
But that was definitely frustrating. I think there were so many good ideas on what's called an advance market commitment, uh, to,

650
01:10:55,210 --> 01:11:03,850
to finance such and such goods that don't exist yet that you, that you, that, that, that there's an advanced market for such goods.

651
01:11:04,360 --> 01:11:09,400
There were so many good ideas around it had to have it had worked for other vaccines in the past.

652
01:11:10,030 --> 01:11:13,570
Michael Cramer and Drexler Glenister um,

653
01:11:14,200 --> 01:11:20,529
well mostly famous as development economists and develop these ideas many often for use these ideas for

654
01:11:20,530 --> 01:11:30,519
development um policies they put them all in place just the decade before it was all ready to go like it.

655
01:11:30,520 --> 01:11:35,259
I think the world could have moved much faster if that was more widely adopted.

656
01:11:35,260 --> 01:11:38,379
The US did it to a good extent with this.

657
01:11:38,380 --> 01:11:42,550
Um, what was it called? The Project Warp Speed.

658
01:11:42,940 --> 01:11:52,089
Oh, yes, yes, yes, yeah. Yeah. And the COVAX, AMC that the W.H.O. and others coordinated was also using that.

659
01:11:52,090 --> 01:11:58,450
But I think on the European level, from what I understand, this could have been used more.

660
01:11:58,990 --> 01:12:05,770
But what I meant, like why I was frustrated, was just how like the winter 20, 20, 20, 21,

661
01:12:05,770 --> 01:12:15,729
I think Germany's very pretty where it was clear where things were heading and the it was election times during this very

662
01:12:15,730 --> 01:12:24,660
federal country where the individual states have a lot to say and the coordination between the states took very long until.

663
01:12:25,680 --> 01:12:31,440
Actions were actually taken. And I think then the restrictions had to be had to have been had to be much more

664
01:12:31,440 --> 01:12:38,130
severe than than what would have been possible and maybe otherwise testing.

665
01:12:39,450 --> 01:12:45,750
We could have been like way further at that point in time and wasn't used as much.

666
01:12:47,850 --> 01:12:54,150
And um, I never thought that slipped my mind.

667
01:12:55,320 --> 01:13:00,120
But you said when we were talking earlier, before, before we really got into talking about COVID,

668
01:13:00,120 --> 01:13:07,680
you did say that you felt that the experience of the pandemic had raised the status of of data generally.

669
01:13:07,690 --> 01:13:16,739
Right. And do you think that's dissipating now, or do I mean these that there's so much madness going on politically at the moment?

670
01:13:16,740 --> 01:13:22,170
We're talking in in the beginning of September 2022.

671
01:13:22,620 --> 01:13:29,940
Right. And that a lot of the political debate that's going on at the moment, if you can even call it that,

672
01:13:31,590 --> 01:13:37,410
seems not to be focussed on what the real problems are, the big problems, the really big problems that you've been talking about.

673
01:13:38,010 --> 01:13:41,760
From a UK perspective, I mean it probably happens a bit elsewhere as well.

674
01:13:42,540 --> 01:13:44,400
I mean, that's for sure. That's right.

675
01:13:45,450 --> 01:13:55,020
Like, I wish obviously that was what would be better, but I think lots of people learned to interpret data in ways that that it was just,

676
01:13:55,290 --> 01:14:02,489
I think, impressive to see the quality of the conversation even on social media that's often given such a bad reputation.

677
01:14:02,490 --> 01:14:13,840
I thought it was often great to see how people discuss growth rates and doubling times and oh numbers and ah and our know it's,

678
01:14:15,390 --> 01:14:21,930
I think it's encouraging to see to see that and and I would not think that this goes away.

679
01:14:21,930 --> 01:14:33,030
I think the world is moving into a direction where, uh, an understanding of numbers is just much more at the centre of, of debate between people.

680
01:14:33,210 --> 01:14:36,510
Yes. Like there's a lot of screaming on the big political platforms,

681
01:14:36,510 --> 01:14:45,270
but I think lots of people are are getting a bit of a more nuanced understanding based on on data and evidence.

682
01:14:45,570 --> 01:14:52,350
Mm hmm. And do you think as far as the general public is concerned, I mean, you know, I come from a background in in science communication,

683
01:14:52,350 --> 01:15:00,660
and there's been a a push for 30 or 40 years now to raise the level of understanding of science in in the general public.

684
01:15:01,260 --> 01:15:04,110
Do you think social science has got a bit further to go on that?

685
01:15:04,320 --> 01:15:13,560
I mean, is there a sense that communicating with the public is part of the responsibility of academics and social side for sure?

686
01:15:13,590 --> 01:15:19,260
Like, I think it's the like we had we touched on a point of that a little bit earlier on when,

687
01:15:19,260 --> 01:15:24,750
when we talked about the value of producing data, publishing data and social science.

688
01:15:25,290 --> 01:15:29,369
And I think overall, it's just about it's about incentives.

689
01:15:29,370 --> 01:15:39,920
There's just no big incentive for social scientists to communicate with with a wider audience that the one thing that counts,

690
01:15:39,930 --> 01:15:47,250
like the central thing that counts is academic publications, academic publications and highly regarded high impact journals.

691
01:15:48,540 --> 01:15:53,699
And so if you don't value if people speak with an audience with a with a wider public

692
01:15:53,700 --> 01:15:59,190
audience and make an effort of communicating that if that isn't rewarded for your career,

693
01:15:59,700 --> 01:16:06,719
then it's understandable that people who are in universities and have lots of other warriors to to take care of,

694
01:16:06,720 --> 01:16:15,270
from teaching to supervision to grant applications, don't find the time and the energy to to do that in addition to everything.

695
01:16:15,270 --> 01:16:20,729
So I think the universities need to change and value the engagement with the

696
01:16:20,730 --> 01:16:25,650
public in addition to the academic research that obviously should also come.

697
01:16:25,830 --> 01:16:30,360
Mm hmm. I look this up is called the International Best Practice Advisory Group.

698
01:16:30,360 --> 01:16:34,620
That right. There was a meeting that we had at some point,

699
01:16:34,700 --> 01:16:40,290
like every week and where often these questions of what can we learn from the

700
01:16:41,370 --> 01:16:49,140
government's policy tracker and and the COVID data was very central to the conversations.

701
01:16:49,410 --> 01:16:54,060
And what was the output of that group? I mean, did that feed back into policy making?

702
01:16:54,540 --> 01:17:03,089
I hope so. Like that was that's beyond what I understood, that I took part in these conversations also not very often,

703
01:17:03,090 --> 01:17:08,740
but I took sometimes part in it because I wanted I wanted to learn what data.

704
01:17:08,760 --> 01:17:12,750
There were lots of epidemiologists there. There were people from government there.

705
01:17:13,140 --> 01:17:15,540
And I wanted to understand what data they would require,

706
01:17:16,530 --> 01:17:23,340
what what kind of things we could focus our energy on so that it would be useful for them overall.

707
01:17:23,520 --> 01:17:33,190
I very much often try to. Very much tried to speak with the users of the data so that I hear from them what's what's needed.

708
01:17:33,820 --> 01:17:45,219
It's like this. I think that's maybe another criticism of of of academia, where I think too much of academia is supply side driven,

709
01:17:45,220 --> 01:17:52,150
where it's some academic thinking, I want to do this thing, I want to spend my energy researching that thing.

710
01:17:53,770 --> 01:18:00,610
And they don't spend enough time thinking about Does anyone need the answer on that question from me?

711
01:18:01,260 --> 01:18:06,550
Like what? I actually open questions here. Like, am I answering a question that others.

712
01:18:07,360 --> 01:18:14,200
I had a much better position to answer. Has maybe someone else really answered this in a way that that I wouldn't even be able to achieve?

713
01:18:14,200 --> 01:18:20,500
I think we're thinking too little about the demand side for the academic work.

714
01:18:20,830 --> 01:18:28,180
And it's also like often that frustrating because then academics are often, I think, frustrated that their work isn't being picked up.

715
01:18:28,870 --> 01:18:36,519
And partly that's what you said, that maybe the political discourse isn't relying on evidence enough and isn't like demanding it enough.

716
01:18:36,520 --> 01:18:41,770
But I think it's partly also the blame of academics that don't think hard enough about which

717
01:18:41,770 --> 01:18:46,590
questions are actually worth answering and which evidence is actually needed by the world.

718
01:18:46,800 --> 01:18:51,670
Is the world actually waiting for me to finish this paper or why not?

719
01:18:52,900 --> 01:19:00,670
And what do you what are you doing now? And are you is Kogan still high up in your priorities or have you been able to

720
01:19:00,670 --> 01:19:05,350
go back to some of the other these other huge problems in global problems?

721
01:19:05,410 --> 01:19:08,500
I mean, we worked on other problems throughout the pandemic.

722
01:19:08,500 --> 01:19:12,760
There was always other work going on and on, various other topics.

723
01:19:14,140 --> 01:19:27,310
But a lot of our energy, like our definitely the the focus of our energy during the in all of 2020 and 2021 was very much on on COVID.

724
01:19:28,120 --> 01:19:33,609
That has changed now a bit. Not so much because we don't do this work anymore.

725
01:19:33,610 --> 01:19:41,319
Like all of the data that we do is still the daily updates you find, the latest vaccination case and deaths and excess mortality.

726
01:19:41,320 --> 01:19:47,770
All of the many metrics that we spoke about, you find them all always up to date, but the updating has become much easier.

727
01:19:48,370 --> 01:19:52,149
So the governments are doing have have learned a lot in the last two years are doing a

728
01:19:52,150 --> 01:19:58,030
much better job in providing this data and making it easy to automate the data updates.

729
01:19:58,600 --> 01:20:02,799
It's also that other international organisations, including the W.H.O.,

730
01:20:02,800 --> 01:20:08,980
are now doing a much better job in producing the data so that we can rely on on their

731
01:20:08,980 --> 01:20:14,950
data also because countries are reporting directly to the W.H.O. in many cases.

732
01:20:15,160 --> 01:20:21,610
So all of that has made our work much easier really since the beginning of this year, 2022.

733
01:20:22,770 --> 01:20:32,410
Um, and, and so code work is still ongoing, but it has become easier.

734
01:20:33,340 --> 01:20:40,540
And then on, in addition to this, we, we grew the team a lot like we are six at the beginning of the pandemic.

735
01:20:40,540 --> 01:20:42,999
I said, and I think we're now 28 or 29.

736
01:20:43,000 --> 01:20:53,920
Well yeah as has grown a lot so so it's like a big focus in the last two years was also hiring finding and was that on the back of the COVID work.

737
01:20:53,920 --> 01:21:00,640
I mean I'm just is it going to is your funding going to fall off a cliff now that you haven't got COVID drive?

738
01:21:00,730 --> 01:21:05,080
I hope not that no. I think we are we are good with with funding.

739
01:21:05,080 --> 01:21:09,370
We and like, funding is good.

740
01:21:09,370 --> 01:21:16,870
We we have the support for the next two years more for sure.

741
01:21:17,950 --> 01:21:22,780
And we just, we are just received a good, a bigger grant.

742
01:21:23,530 --> 01:21:35,170
Um, and we are also not increasing further. So the growth of the team and hiring was, was a big theme, but it has come to an end.

743
01:21:36,160 --> 01:21:48,340
Natasha ahuja has just joined us yesterday and there might be a few hires that we have to do in the next year,

744
01:21:48,340 --> 01:21:52,750
but it's not going to be it's not going to be as crazy as the last two years.

745
01:21:54,130 --> 01:22:04,240
So we are finally actually able to lift our heads and just look a bit further ahead and and plan a bit like the next months or years.

746
01:22:04,420 --> 01:22:10,659
During COVID, it was it was always whatever happened with the pandemic was determining what was happening.

747
01:22:10,660 --> 01:22:17,649
And that week it was very immediate. And that's not the case, fortunately anymore.

748
01:22:17,650 --> 01:22:24,220
So we we do a range of other topics, like one that I'm excited about now is is it the longest standing project?

749
01:22:25,670 --> 01:22:35,150
Of our world in data, a big database on the history of war where obviously is a big question of is the world becoming more peaceful?

750
01:22:35,160 --> 01:22:41,510
Is the world becoming more violent? Can we learn anything about what makes what regions more peaceful?

751
01:22:41,510 --> 01:22:50,990
And so on? There's also a ton of research on this, but the big central database for the long run is of very poor quality.

752
01:22:52,040 --> 01:22:54,859
There's great data from especially from the Second World War onwards,

753
01:22:54,860 --> 01:23:00,890
but over the last centuries the data is poorer also because there was just no institution that had good funding to do this.

754
01:23:01,250 --> 01:23:10,910
And so for the last six, seven years or something, we've been working on this big project of piecing together a global database on war casualties.

755
01:23:11,000 --> 01:23:15,020
But going back historically, yeah. Over centuries, yeah.

756
01:23:15,020 --> 01:23:24,380
To like trying to go back to 1300. Of course the data is very patchy, but the earliest records that are included in our database are from 1300.

757
01:23:26,000 --> 01:23:31,370
And now we're building a big tool to make this data available to the world.

758
01:23:31,670 --> 01:23:35,330
And the idea there's also like this data is not going to be perfect either.

759
01:23:36,140 --> 01:23:40,330
But in contrast to the earlier work, it's documented well.

760
01:23:40,370 --> 01:23:46,669
So you can actually know from before each war and each data points for the number of deaths in each.

761
01:23:46,670 --> 01:23:49,190
What? You can know where where we have that data from.

762
01:23:49,640 --> 01:23:58,850
And so the idea is that it's becoming a living database, like a constant work in progress where other people can contribute to the database.

763
01:23:58,850 --> 01:24:04,940
And if you are an expert in the history of, say, Japan and you know,

764
01:24:04,940 --> 01:24:11,570
of some wars that happened in Japan that aren't in our database, you would be able to to add them to the database.

765
01:24:11,600 --> 01:24:17,060
Or if you find that the data source that we rely on for some war in Japan is actually a very bad one.

766
01:24:17,900 --> 01:24:21,680
You can you can change that and add another data.

767
01:24:21,890 --> 01:24:27,770
So it's like it should improve in quality over time. That's one project that we just got back to.

768
01:24:28,460 --> 01:24:33,410
And are you including civil wars as well as wars between countries?

769
01:24:33,570 --> 01:24:38,899
Yeah, it would include both. Yeah, I think that's one of the difficulties in this.

770
01:24:38,900 --> 01:24:42,650
Like it's very hard to draw the borders between. Yeah, what is a war.

771
01:24:42,710 --> 01:24:45,710
Exactly. Exactly where does one war started? Does it end?

772
01:24:45,710 --> 01:24:51,080
Like is the Second World War? Is that one massive war or is that a series of many different wars?

773
01:24:51,080 --> 01:24:57,160
And should you aggregate the numbers into one massive number for all of the people that have died in the Second World War?

774
01:24:57,170 --> 01:25:02,270
Or should you break it down by what happened in regions of the Pacific and regions in Europe?

775
01:25:02,270 --> 01:25:10,400
And like, I mean, it's, it's a question of, uh, where you like, how you structure this information.

776
01:25:10,640 --> 01:25:17,280
And then another big one is who actually dies in a, in a war like what is actually a war death.

777
01:25:17,300 --> 01:25:24,410
Like you can have a very narrow definition of that, like deaths on the battlefield between combatants.

778
01:25:25,010 --> 01:25:27,020
Or you can have a very wide definition of that,

779
01:25:27,380 --> 01:25:38,930
where war deaths also those people who died in famines or the outbreaks of infectious diseases that were caused by the war, maybe.

780
01:25:39,890 --> 01:25:48,080
And that makes sense to some extent, because often, especially famines and the restrictions of food supply is actually used as a weapon of war.

781
01:25:48,090 --> 01:25:54,979
So it makes some sense to include them. But it's I think what's crucial is that you have some structure to this and you

782
01:25:54,980 --> 01:26:00,050
actually provide the information of what your definition of of a death is.

783
01:26:00,950 --> 01:26:02,360
So that's one one project.

784
01:26:03,440 --> 01:26:19,190
Another I have to ask, do you have an insight into what the the answer is as to whether where whether deaths are, um, proportionately less over time.

785
01:26:19,580 --> 01:26:30,469
Right. I would think it's, it's definitely not easy to answer because, um, I mean, what's,

786
01:26:30,470 --> 01:26:38,180
what's for sure true is that many world regions live through incredibly peaceful periods in history.

787
01:26:38,570 --> 01:26:39,680
Like my own hometown,

788
01:26:39,680 --> 01:26:48,200
there's like a historian that was looking through the southwest of Germany and tracking that the wars that were ongoing in the south west of Germany.

789
01:26:48,620 --> 01:26:52,489
And I think there has been some way since I read this,

790
01:26:52,490 --> 01:26:59,059
but I think they don't find them a period that's anywhere as long as has been has been peaceful for anywhere

791
01:26:59,060 --> 01:27:09,530
as long as the period now from 45 to today in the last 2400 years or something like and then like 400 B.C.,

792
01:27:10,640 --> 01:27:17,750
the data obviously gets very poor. So who knows what exactly was going on there, but it was just basically war following after war.

793
01:27:17,900 --> 01:27:22,940
I mean, as any anyone who ever looked into Europe's history would know.

794
01:27:24,410 --> 01:27:30,210
And then I think it's also. Not a question that you can just answer with the data on board, as you would also.

795
01:27:31,230 --> 01:27:35,190
I think those people who make the argument that the world has become more peaceful say

796
01:27:35,190 --> 01:27:40,049
that there were structural changes in the world that made the world more peaceful.

797
01:27:40,050 --> 01:27:46,680
And we have to interpret the data on war deaths in light of these structural changes.

798
01:27:46,690 --> 01:27:55,470
So it would be wrong just to look at these outcome variable because in the end especially big was fortunately don't happen very frequent.

799
01:27:55,770 --> 01:28:04,530
Lee And so you might be actually mistaking this period of the accident of war for a signal that you actually can't,

800
01:28:04,600 --> 01:28:09,389
uh, can't see in the data yet just because they are so infrequent.

801
01:28:09,390 --> 01:28:14,430
And you would be fooled by randomness, as they say, and it's just the randomness of data.

802
01:28:14,520 --> 01:28:20,070
So the argument that looks at these structural changes I think is convincing to some extent.

803
01:28:20,110 --> 01:28:28,139
So I think there are four big ones. One is that democracies are much less likely to go to war with each other,

804
01:28:28,140 --> 01:28:34,230
and the world surely has become much more democratic over the last 200 years.

805
01:28:35,610 --> 01:28:40,200
Uh, so more democracies should make the world more peaceful.

806
01:28:40,200 --> 01:28:45,989
We have the evidence that for this democratic peace theory backing this up to, to a good extent,

807
01:28:45,990 --> 01:28:52,260
a few examples, if any of that would count as democracy use fighting peace, fighting wars against each other.

808
01:28:52,260 --> 01:29:02,969
The idea there is that as an authoritarian leader, you have a lot to win from a war and little to to to lose.

809
01:29:02,970 --> 01:29:10,049
So you might like extend the glory of your reign of that country a lot and you might just,

810
01:29:10,050 --> 01:29:18,120
like increase your territory and your wealth, but you might not be the first one to actually run into the battles yourself.

811
01:29:18,780 --> 01:29:25,110
While democracies, on the other hand, individuals have often little to win from a war and much to lose.

812
01:29:25,530 --> 01:29:33,299
So once citizens have some say in whether the country is fighting a war or not,

813
01:29:33,300 --> 01:29:39,090
they are much less likely to actually volunteer for for running into the next war.

814
01:29:39,420 --> 01:29:41,280
So that's this change in democracy.

815
01:29:41,280 --> 01:29:51,150
Another one is increase in trade, that it's just cheaper in many ways to trade than to raid another country and get the resources that way.

816
01:29:52,140 --> 01:29:58,770
Trade has increased a lot. A third one is diplomacy, that diplomacy has increased a lot between nations around the world,

817
01:29:58,770 --> 01:30:04,739
and that for some conflicts, at least, diplomatic solutions are found.

818
01:30:04,740 --> 01:30:08,670
So that increases or decreases the likelihood of war.

819
01:30:08,940 --> 01:30:13,440
And the fourth one, I think is also a big one is about just cultural change.

820
01:30:13,440 --> 01:30:18,210
Like if you look at the like even at the high intellectuals from 100 years ago,

821
01:30:18,840 --> 01:30:27,329
it was very common to praise war as this rite of passage for young men, especially, that like to become a man.

822
01:30:27,330 --> 01:30:31,260
You actually have to fight a war. And it'd be wrong to not fight a war.

823
01:30:31,260 --> 01:30:34,499
And like, there's like glory in war. And for you, a fatherland.

824
01:30:34,500 --> 01:30:38,830
And and all of this rhetoric was extremely common 100 years ago.

825
01:30:38,850 --> 01:30:42,360
Like throughout many periods in history, it's much,

826
01:30:42,360 --> 01:30:53,280
much rarer for for young guys to get excited about going into a war these days and just looking at time with about 2 minutes left.

827
01:30:53,850 --> 01:30:57,540
So I'm going to skip all those questions, so I'll just get to the last one.

828
01:30:59,670 --> 01:31:02,819
So, uh, yes.

829
01:31:02,820 --> 01:31:14,520
So has your work on this changed in any way, your approach to your work or and how would you like things to develop in the future?

830
01:31:15,480 --> 01:31:17,190
Um, yeah, for sure.

831
01:31:17,490 --> 01:31:23,640
Lots of changes came together, like obviously this much bigger team, really excellent colleagues that joined us in the last two years.

832
01:31:24,720 --> 01:31:28,470
That's a bit hard to disentangle from the changes due to the pandemic,

833
01:31:29,100 --> 01:31:38,399
but things have become a whole lot more like we're still not like us, still feels very much like a start up in the early stages,

834
01:31:38,400 --> 01:31:45,510
but compared with two years ago, like those that have processes in place, like how the data is updated, all of these things have improved a lot.

835
01:31:45,510 --> 01:31:49,440
Just because colleagues are really taking care of these things much more.

836
01:31:51,390 --> 01:32:01,320
I think how like I think it definitely also changed my understanding of like existential risks.

837
01:32:01,500 --> 01:32:13,920
Like we touched upon it a bit, but I think seeing how bad in many ways this pandemic has also gone definitely increased

838
01:32:15,150 --> 01:32:20,820
my perception of how how likely or how yet how high the risks are of of future pandemics.

839
01:32:21,870 --> 01:32:25,320
Um, especially now in technologies like in.

840
01:32:25,750 --> 01:32:31,240
Much more are available to change pathogens in many ways.

841
01:32:32,830 --> 01:32:43,030
And I think one of the things that I was thinking about a lot this last days is just a frustration of

842
01:32:43,030 --> 01:32:48,610
how little the world is doing to prepare for future pandemics and reduce the risks for pandemics.

843
01:32:48,660 --> 01:32:55,149
So I think I would definitely, based on what I was reading before, I was expecting a pandemic.

844
01:32:55,150 --> 01:33:03,920
Like, I mean, any epidemiologist that you ever speak to is expecting pandemics or is definitely expecting outbreaks.

845
01:33:04,240 --> 01:33:11,950
And what do they say? Like outbreaks are outbreaks are certain, pandemics are optional.

846
01:33:12,520 --> 01:33:20,530
And so, like everyone always knows that pathogens jump from from animals, animal hosts through to humans.

847
01:33:20,530 --> 01:33:27,399
And that happens. And there's always a risk of a pandemic. So that maybe didn't surprise me that much, but I'm definitely surprised.

848
01:33:27,400 --> 01:33:36,219
Now, two years of two and a half years into the pandemic, that relatively little is happening in preventing future pandemics.

849
01:33:36,220 --> 01:33:45,250
So I think I would like in also in the team, we very much want to make the case for the risk of future pandemics.

850
01:33:45,250 --> 01:33:52,149
And when I focus on that, I think that's one lesson that we take away from the last two and a half years,

851
01:33:52,150 --> 01:34:02,530
that it was bad enough let's not suffer through more lockdowns and not suffer through more millions of deaths from from these preventable tragedies.

852
01:34:02,530 --> 01:34:05,739
And is there better data to collect on pandemic preparedness?

853
01:34:05,740 --> 01:34:12,309
Because there was a date I mean, there was a ranking of pandemic preparedness for wasn't there?

854
01:34:12,310 --> 01:34:17,840
And yet the countries that were at the top of that ranking didn't necessarily do all that well know.

855
01:34:17,860 --> 01:34:24,849
Like it was one of these indexes that I messed up, I don't know, some data and was I think was nonsense to start with.

856
01:34:24,850 --> 01:34:28,570
And then it was just proven that it was wasn't all that informative.

857
01:34:29,110 --> 01:34:36,250
No, I think this is not all that much. I mean, once you have the data point about a terrible pandemic, it's kind of too late.

858
01:34:36,850 --> 01:34:41,709
But but like, there are things that we I think we can do, for example,

859
01:34:41,710 --> 01:34:47,860
and there isn't actually a great dataset on the history of the pandemic throughout history.

860
01:34:47,920 --> 01:34:54,069
Mm hmm. So that's a very obvious thing to actually do, like lots of academic papers on the risk from pandemics.

861
01:34:54,070 --> 01:35:02,380
They start with some kind of literature review and speak about the Justinian's plague and the Black Death and maybe the stars.

862
01:35:02,680 --> 01:35:09,790
So they pull together a couple of data points, but it would be just helpful for the community to have like this one data's data set on pandemics.

863
01:35:09,790 --> 01:35:16,390
So that's a very obvious thing that we can do. And it would just show hopefully to people like these things happen all the time.

864
01:35:16,570 --> 01:35:19,810
Let's just make sure that COVID was the last one.

865
01:35:21,880 --> 01:35:29,500
That's a very obvious data project. Another one is, I think, a big risk for future pandemics from influenza.

866
01:35:29,680 --> 01:35:33,940
Many past pandemics were from from influenza viruses.

867
01:35:34,300 --> 01:35:37,810
So we want to just work more on influenza.

868
01:35:37,840 --> 01:35:40,600
We currently bring together the data historically,

869
01:35:40,600 --> 01:35:48,340
and the W.H.O. does a good job in bringing together also real time data on influenza so that we make that data available.

870
01:35:49,900 --> 01:35:53,650
Saloni Da Tani, a colleague who is focusing on infectious diseases,

871
01:35:53,650 --> 01:36:04,910
she's working currently there on communicating and researching why some flu winters, like it's a very seasonal disease wave some flu.

872
01:36:04,930 --> 01:36:10,870
Winters are also particularly bad. Why does that happen and why do we have these big differences?

873
01:36:10,900 --> 01:36:14,920
What can we learn from that? And how can we maybe reduce the risk from from influenza?

874
01:36:16,300 --> 01:36:22,210
Because I hope that this is maybe useful. I think it's I mean, the sad thing,

875
01:36:22,220 --> 01:36:25,480
like maybe like yesterday I was speaking with a colleague who was who was more

876
01:36:25,480 --> 01:36:31,600
upbeat than I about the the chances of preparing better for future pandemics.

877
01:36:31,600 --> 01:36:39,429
He was making the point that this is only now starting this moment when when governments can possibly actually get involved in that,

878
01:36:39,430 --> 01:36:45,579
because now we're coming to the end of actually like at least for for many rich countries, we're coming,

879
01:36:45,580 --> 01:36:46,149
fortunately,

880
01:36:46,150 --> 01:36:55,629
to the end or at least into a better phase where like a large share of the population is vaccinated and things are calmer for those people,

881
01:36:55,630 --> 01:36:58,800
for those experts in governments that are focussed on infectious diseases.

882
01:36:58,810 --> 01:37:03,129
So that he was making the point that from now onwards those people are actually

883
01:37:03,130 --> 01:37:09,060
freed up to actually make the case about getting prepared for future pandemics.

884
01:37:09,310 --> 01:37:14,230
Hopefully he's right and I'm too too early in my pessimism.

885
01:37:17,170 --> 01:37:21,219
But for example, I think this like index on pandemic preparedness was a bit nonsense.

886
01:37:21,220 --> 01:37:24,250
But I think some other things are just obvious.

887
01:37:24,250 --> 01:37:32,210
Like I think one thing that. I would make the case for that's useful and many of the owners just make the case for is much better.

888
01:37:32,360 --> 01:37:43,730
B So that especially in in even worse pandemics with, with, with more terrible viruses, you actually have a chance to,

889
01:37:44,520 --> 01:37:55,099
to have to have people that have to do the most important roles in keeping people fed and electricity on and hospitals running,

890
01:37:55,100 --> 01:38:02,299
that those people actually have excellent PPE that's produced before the pandemic actually happens, that's actually available somewhere.

891
01:38:02,300 --> 01:38:08,210
And and you can count how many how many countries prepare for for that aspect.

892
01:38:08,420 --> 01:38:15,410
That would be like one other aspect that you but I think is a sensible metric that actually tells you something about the preparedness of a country.

893
01:38:15,750 --> 01:38:22,700
Mm hmm. Good. I think we have to stop because we've got to deadline, then turn off the record.

894
01:38:23,340 --> 01:38:24,020
Let's do that.