1 00:00:00,420 --> 00:00:05,220 And to come, but it's also great to be able to join online. 2 00:00:05,220 --> 00:00:11,580 So without further ado, I share my presentation with you. 3 00:00:11,580 --> 00:00:15,240 And just a word in advance. 4 00:00:15,240 --> 00:00:22,680 So behavioural analysis of humanitarian negotiators, we turn to them learning from cognition that there is a little bit from the title. 5 00:00:22,680 --> 00:00:26,850 And the reason is that it's really still work on progress. 6 00:00:26,850 --> 00:00:34,050 So everything I'm presenting today is preliminary, and we haven't totally exhausted the data yet. 7 00:00:34,050 --> 00:00:41,460 So there might be questions arising, which I cannot answer because we haven't gone there yet to analyse the data. 8 00:00:41,460 --> 00:00:48,570 So that's a joint research project by myself, as well as somebody from the Hebrew University, 9 00:00:48,570 --> 00:00:52,980 Christopher A. from the Max Planck Institute for Research on Collective Goods and Catarina. 10 00:00:52,980 --> 00:01:02,310 Look now to from the University of Hamburg. And the reason being that we have such a big research group because it's a big project. 11 00:01:02,310 --> 00:01:09,630 And I'm giving you a little bit of a sense later. So I have to put this. 12 00:01:09,630 --> 00:01:14,430 OK, so what's our research motivation? 13 00:01:14,430 --> 00:01:21,270 And you might have seen actually, he started this research before the U.N. guidance on behavioural science came out. 14 00:01:21,270 --> 00:01:28,800 But here's a quote for by the U.N. secretary general. Pro-terrorist behavioural science is a critical tool for the U.N. to progress on. 15 00:01:28,800 --> 00:01:35,400 Its members can contribute to combating poverty, improving public health and safety, preventing and managing crises, 16 00:01:35,400 --> 00:01:42,660 promoting gender and economic equality, tackling corruption, strengthening peacebuilding and all of the hostages. 17 00:01:42,660 --> 00:01:48,840 And it should also make the public sector more efficient and including the international organisation. 18 00:01:48,840 --> 00:01:55,410 So what you see here is that the U.N. by now really has been trying behavioural. 19 00:01:55,410 --> 00:02:00,270 It's following the World Bank on that which had an earlier report in 2019. 20 00:02:00,270 --> 00:02:05,910 The World Development Report was devoted to behavioural sciences. 21 00:02:05,910 --> 00:02:15,360 But what you see is that they really believe that behavioural sciences can help international organisations in their policies and, 22 00:02:15,360 --> 00:02:18,570 for that matter, the vote. 23 00:02:18,570 --> 00:02:30,420 Not hitherto international law, as well as the policies, rely to a large extent and still does on the rational choice paradigm for its act. 24 00:02:30,420 --> 00:02:40,300 So when we draught treaties, when we talk about compliance, it's all basically based on rational choice assumptions. 25 00:02:40,300 --> 00:02:45,240 These, you know, if you look at our theory, certainly by the institutionalist, 26 00:02:45,240 --> 00:02:52,110 certainly by the realists constructivist have been different, but without using cognitive sciences. 27 00:02:52,110 --> 00:02:58,950 Now, psychological and experimental research challenges this rational choice paradigm. 28 00:02:58,950 --> 00:03:06,120 So people are only boldly rational using biases and de-risk and make of cognitive mistakes. 29 00:03:06,120 --> 00:03:18,240 Now, this research comes mainly from experiments with students in the lab or laypersons, and I'm going to come back to that. 30 00:03:18,240 --> 00:03:28,800 So we used to readily assume rational choice or rational actors, and now maybe too readily assume bound, oddly rational actors. 31 00:03:28,800 --> 00:03:35,130 I'm saying this, and some of you may wonder because I think I can say that come up. 32 00:03:35,130 --> 00:03:44,580 Rudy and I have been have been since years pushing for international law, taking into account behavioural sciences. 33 00:03:44,580 --> 00:03:48,360 And now suddenly we say, oops, maybe we should be cautious. 34 00:03:48,360 --> 00:03:55,110 But you know, it should all be evidence based, and what I'm presented to you today is sort of evidence we got. 35 00:03:55,110 --> 00:04:06,470 So the question is whether we can readily use behavioural sciences without regard to specific actors and specific context. 36 00:04:06,470 --> 00:04:06,760 Never, 37 00:04:06,760 --> 00:04:18,110 Burton in an angel unbound contribution on the limitations or limits of the behavioural approach to international law says that the biggest problem of 38 00:04:18,110 --> 00:04:22,670 using behavioural insights for international law is that it's extremely difficult 39 00:04:22,670 --> 00:04:27,980 to recruit actual decision makers in a way that allows for direct study. 40 00:04:27,980 --> 00:04:35,330 And that is correct. So if you think about, for example. Studies on judges. 41 00:04:35,330 --> 00:04:37,640 You don't get international judges, 42 00:04:37,640 --> 00:04:50,120 and they're also not sufficient to prove them to have to be able to really conduct a sufficient statistical analysis we have done. 43 00:04:50,120 --> 00:04:52,830 I was others with Susan, Frank and others we have done. 44 00:04:52,830 --> 00:05:00,750 A study was international arbitrators where we do find indeed that they find show bias is on juries like everybody else. 45 00:05:00,750 --> 00:05:02,660 They are a little bit better. 46 00:05:02,660 --> 00:05:13,150 Now I got the access to the competence centre of of humanitarian negotiators and negotiations, but I need to tell you a little bit of that. 47 00:05:13,150 --> 00:05:18,860 So that is a project. And now it's installed in Geneva. 48 00:05:18,860 --> 00:05:33,680 It's adjunct to the ICRC, and it's basically driven by international humanitarian organisations such as the UNHCR, UNICEF, but also NGOs like me. 49 00:05:33,680 --> 00:05:39,220 It's also from Pierre, is a is a frequent participant, and they also finance that centre. 50 00:05:39,220 --> 00:05:46,960 And and they said, well, it would be super interesting to understand how humanitarian negotiators. 51 00:05:46,960 --> 00:05:56,220 Make decisions. So basically, we got that opportunity and the access to those professionals to better understand 52 00:05:56,220 --> 00:06:02,310 frontline humanitarian negotiators as agents of international law or more specifically, 53 00:06:02,310 --> 00:06:06,990 Joe, and to contribute to better practises through negotiation training. 54 00:06:06,990 --> 00:06:12,300 Of course, from the situation side, the the they want better, better informed training. 55 00:06:12,300 --> 00:06:23,880 And when they train now, they basically rely on rational choice negotiation theory, which may be insufficient. 56 00:06:23,880 --> 00:06:30,750 So what about front line humanitarian dossier? Does that basically all those negotiating for humanitarian needs on the ground? 57 00:06:30,750 --> 00:06:34,470 So there is not a profession of humanitarian negotiators, 58 00:06:34,470 --> 00:06:42,030 but usually all frontline humanitarians are negotiators because they constantly negotiate on something. 59 00:06:42,030 --> 00:06:49,320 Maybe it was government, it was rebel groups, whatever or communities. 60 00:06:49,320 --> 00:06:56,970 So, so they're constantly negotiating. And the question is then do humanitarian negotiators think differently than laypersons? 61 00:06:56,970 --> 00:07:00,990 And if so, when and in which situations? And why is that important? 62 00:07:00,990 --> 00:07:09,630 Because we could have gone back and I could have written an easy manual negotiation manual for the humanitarian negotiators, 63 00:07:09,630 --> 00:07:17,580 just basically using all the insides, which we have from experiments, but again, mainly the students and tell them, well, 64 00:07:17,580 --> 00:07:27,590 think about that, think about that and try to to to show them where they might be making mistakes. 65 00:07:27,590 --> 00:07:31,820 And we from the very beginning thought that, well, maybe they are different. 66 00:07:31,820 --> 00:07:38,120 They are very maybe some of you do know them. They're very impressive people. 67 00:07:38,120 --> 00:07:42,260 I have to say, but very normal people as well. 68 00:07:42,260 --> 00:07:48,740 And the question is and they negotiate sometimes in really, really difficult situations. 69 00:07:48,740 --> 00:07:54,590 So sometimes is a gun at the head. So, so how does that impact their decision making? 70 00:07:54,590 --> 00:08:02,990 And what does that imply if they are different for the external validity of behavioural experiments used by the UN more generally, 71 00:08:02,990 --> 00:08:15,350 and just more specifically, because the U.N. hitherto largely relies on if you look at who's behind, there is a condiment or something. 72 00:08:15,350 --> 00:08:21,080 They used the experiments of everyday actors or everyday subjects, so students. 73 00:08:21,080 --> 00:08:28,310 So what did we do? We originally wanted to go to the workshops, and this decision holds workshops all over the world, 74 00:08:28,310 --> 00:08:36,830 usually usually in the crisis situations, credit crisis spots. 75 00:08:36,830 --> 00:08:41,240 But due to COVID, we couldn't do it and they didn't do their workshops anymore. 76 00:08:41,240 --> 00:08:48,650 They did them online. So we did a vignette study using Qualtrics. 77 00:08:48,650 --> 00:08:55,220 We had 119 humanitarian negotiators with an average of 4.7 years of experience. 78 00:08:55,220 --> 00:09:03,830 We also tested 57 students of Hebrew and Hamburg universities was illegal international relations background, 79 00:09:03,830 --> 00:09:09,230 and we tested 150 for laypersons by April. 80 00:09:09,230 --> 00:09:15,950 Now we put a series of questions and some of them are actually framed. 81 00:09:15,950 --> 00:09:23,900 We call them friends, so they have vignettes in the humanitarian context of a typical situation they would be facing. 82 00:09:23,900 --> 00:09:27,710 And then we asked the same question in an abstract way. 83 00:09:27,710 --> 00:09:34,970 So A and B and C so untrained and we put all those questions to those subjects. 84 00:09:34,970 --> 00:09:42,560 So what our research questions. So first, are humanitarian negotiators equally subject to cognitive biases as laypeople? 85 00:09:42,560 --> 00:09:46,670 So the question of expertise, does expertise make a difference? 86 00:09:46,670 --> 00:09:52,580 Or and research question, too, was whether that bias is depend on the context of decision making, 87 00:09:52,580 --> 00:09:58,490 so whether they are deciding differently in the neutral or in the humanitarian context. 88 00:09:58,490 --> 00:10:13,400 So those were the frame versus pantry experiments when we tested for biases which are well proven, often used in all the policy. 89 00:10:13,400 --> 00:10:17,840 Advice or used by national nudge units. 90 00:10:17,840 --> 00:10:22,520 So they're very subtle bias. The first one is the hindsight bias, 91 00:10:22,520 --> 00:10:28,160 so humanitarian negotiators perceived post-event as having been more predictable than they 92 00:10:28,160 --> 00:10:34,520 actually were their policies to when faced with a concrete humanitarian negotiation. 93 00:10:34,520 --> 00:10:40,670 Humanitarian negotiators are influenced by irrelevant numeric anchors when taking decisions. 94 00:10:40,670 --> 00:10:52,850 So anchoring bias they processed three humanitarian negotiators resolve disputes based on representative views rather than deliberative reason. 95 00:10:52,850 --> 00:10:54,200 The so-called base rate, 96 00:10:54,200 --> 00:11:06,470 error and hypothesis for humanitarians are prone to overoptimism when raising the prospect of the mission's so-called overoptimism bias. 97 00:11:06,470 --> 00:11:08,210 So we tested all of this. 98 00:11:08,210 --> 00:11:17,270 So in other words, we expect humanitarian negotiators to be so susceptible to cognitive biases, but to respond differently to laypeople, 99 00:11:17,270 --> 00:11:30,860 sometimes at least, and we expect them to respond differently in humanitarian contexts or in frame situations in contrast to more abstract vignettes. 100 00:11:30,860 --> 00:11:35,870 And here are our results. So first, the hidden side bias, so the hidden side bias. 101 00:11:35,870 --> 00:11:42,750 Basically, we should focus on the. So. 102 00:11:42,750 --> 00:11:54,180 The hindsight bias basically tells you it's like it's a little bit like I knew it all along, so what it what it does, 103 00:11:54,180 --> 00:12:01,980 what it says is basically that people perceive past events as more predictable than they actually were. 104 00:12:01,980 --> 00:12:11,520 That matters and has been tested, for example, with judges when they adjudicate on tort law or in the context of police violence. 105 00:12:11,520 --> 00:12:20,910 Stuff like that. OK, now here's the frame when yet we have imagined humanitarian negotiator H, 106 00:12:20,910 --> 00:12:26,490 whose team is coordinating a medical aid in an area where kidnapping is a frequent occurrence. 107 00:12:26,490 --> 00:12:32,700 Most intense instances of kidnapping are financially motivated and thus involves demands for ransom. 108 00:12:32,700 --> 00:12:42,600 Kidnapping groups have also targeted humanitarian workers. We get daily needs medical supplies that the team is able and willing to deliver. 109 00:12:42,600 --> 00:12:46,200 We have delivered such aid to villages in the region 10 times in the past. 110 00:12:46,200 --> 00:12:52,370 In one case, the team member was kidnapped but released on ransom a few days later. 111 00:12:52,370 --> 00:13:01,520 Should age send a team member eight to the village age decides to send a team member, so that's a neutral question. 112 00:13:01,520 --> 00:13:04,990 And then we vary between the groups, right? 113 00:13:04,990 --> 00:13:10,910 So one group got each decides to send a to team member aid to the village of Ein Counters 114 00:13:10,910 --> 00:13:15,800 kidnappers on the way after the organisation has paid a substantial amount of money. 115 00:13:15,800 --> 00:13:21,170 A week later, he returns to the team. Do you think each decision was appropriate? 116 00:13:21,170 --> 00:13:29,590 And then the second variation is each decides to send a team member to the village every chance without encountering problems. 117 00:13:29,590 --> 00:13:39,280 So that's a positive scenario. Scenario. Do you think each decision was appropriate and we always have put? 118 00:13:39,280 --> 00:13:42,130 H instead of how would you decide? 119 00:13:42,130 --> 00:13:56,300 Because we want to create personal distance to the question, because when in 2019 I we conducted on the annual conference call. 120 00:13:56,300 --> 00:14:01,650 Before COVID, we conducted a workshop and we tested was few people. 121 00:14:01,650 --> 00:14:09,100 It's very, very, very well known Asian disease problem issue. 122 00:14:09,100 --> 00:14:14,140 And we totally didn't get the results which we were expecting. 123 00:14:14,140 --> 00:14:23,500 And when we afterwards talked to them, it seems that they wouldn't do something if they feel it contradicts contradicts. 124 00:14:23,500 --> 00:14:30,850 So that's actually the second study we are conducting. So we thought, but we know that they do. 125 00:14:30,850 --> 00:14:37,930 We know that they do not adhere to the principles of age all the time. 126 00:14:37,930 --> 00:14:44,530 They just wouldn't talk about it and certainly not the ICRC people, because the ICRC keeps everything secret. 127 00:14:44,530 --> 00:14:50,710 OK, so, so so we tried to create this sort of personal distance. 128 00:14:50,710 --> 00:14:59,590 Now, this is the results, negotiators are more cautious under both scenarios, but also show him side bias. 129 00:14:59,590 --> 00:15:03,910 OK, so what you see here, those are the student, this the student population. 130 00:15:03,910 --> 00:15:10,690 I left out the lay people now, this student population here and this is the negotiators. 131 00:15:10,690 --> 00:15:15,280 And what you see that this was the negative scenario. They go even less so. 132 00:15:15,280 --> 00:15:21,920 They do also show the hand side bias, but generally they're much more cautious. 133 00:15:21,920 --> 00:15:25,520 Then they people, which is in a way, a good sign. 134 00:15:25,520 --> 00:15:38,240 Now then, if this is basically the same scenario, which just abstract, OK, in 10 past ten past decisions D has decided to do a one case was a failure, 135 00:15:38,240 --> 00:15:42,300 so decided in favour of a and then you have again the positive and negative. 136 00:15:42,300 --> 00:15:51,230 OK. And this is what we find here. So here the difference between humanitarians and students disappears. 137 00:15:51,230 --> 00:15:56,870 So that basically indicates to us that the hindsight bias is context dependent. 138 00:15:56,870 --> 00:16:05,400 So humanitarian negotiators, OK? Let me turn to anchor. 139 00:16:05,400 --> 00:16:15,990 Anchoring is also a very well proven bias where you have and you can have legal anchoring, 140 00:16:15,990 --> 00:16:24,530 but you can also hear being focussed on the more better tools a numerical quantitative anchoring which is well-established in untrained situations. 141 00:16:24,530 --> 00:16:29,870 So, for example, which was the first study done by Kahneman first, 142 00:16:29,870 --> 00:16:35,670 Kahneman got the Nobel prise, the Nobel prise for all of the studies, the wheel of Fortune. 143 00:16:35,670 --> 00:16:42,210 So what they did is they manipulated the wheel of fortune and had to stop at 10 and at sixty five. 144 00:16:42,210 --> 00:16:53,340 And then after that, they asked participants to to to name the percentage of African countries in the United Nations. 145 00:16:53,340 --> 00:17:01,830 And you know, the wheel of fortune obviously doesn't have anything to do with the number of African countries in the United Nations. 146 00:17:01,830 --> 00:17:06,150 That's a study from 1974. 147 00:17:06,150 --> 00:17:16,150 And still, what what they found is that the answers were very close to the anchor, so those who got a 10 would be saying something like around 10. 148 00:17:16,150 --> 00:17:19,830 You know, it could be 12 15. There is those who got the sixty five. 149 00:17:19,830 --> 00:17:33,470 I had much higher numbers. Another one is the age of Gandhi's death, where they were asked whether. 150 00:17:33,470 --> 00:17:41,990 They had anchors of nine and 140 and again here countries that actually was at 78. 151 00:17:41,990 --> 00:17:56,630 But here again, they found an influence on on the on the edge, which was stated by the respondents, which should be unconnected and so irrelevant. 152 00:17:56,630 --> 00:18:01,920 So, so so they asked actually whether he had died at nine or at one hundred thirty four days. 153 00:18:01,920 --> 00:18:07,190 So both are irrelevant anchors in the sense of it must be clear that Gandhi has not died either. 154 00:18:07,190 --> 00:18:15,470 Mine was 140, but still it matters mattered on on on what age they then institute. 155 00:18:15,470 --> 00:18:23,720 So what, what, what? What was the vignette? You're about to go on the mission delivering medical supplies to a camp for 156 00:18:23,720 --> 00:18:29,480 displaced person people that is experiencing an outbreak of a new Ebola like virus? 157 00:18:29,480 --> 00:18:32,180 It is unclear, and we did that before COVID. 158 00:18:32,180 --> 00:18:39,620 It is unclear how many of the about 10000 people in the camp are infected, but health must be administered immediately. 159 00:18:39,620 --> 00:18:45,410 Thus, you can only work with the supplies that had you have access to a sizeable amount of a drug that has been 160 00:18:45,410 --> 00:18:52,550 used to treat patients infected by other variants of the Ebola virus and then over dinner with the team. 161 00:18:52,550 --> 00:18:58,940 Your colleague talks about a very difficult humanitarian mission she participated in on a different continent. 162 00:18:58,940 --> 00:19:03,650 So totally different situation, right? Rare. 163 00:19:03,650 --> 00:19:05,690 And then again, you had the treatment. 164 00:19:05,690 --> 00:19:14,210 Either they got a thousand people were saved or 10000 people were saved because food and clean drinking water was delivered on time. 165 00:19:14,210 --> 00:19:20,570 And then you ask how many of the people in the camp can be saved? So in absolute numbers, and here are the results. 166 00:19:20,570 --> 00:19:23,750 I hope you can see that on your screen. 167 00:19:23,750 --> 00:19:32,660 So what we see here is the negotiators and they, the high anchor indeed made a difference, but it's not statistically significant. 168 00:19:32,660 --> 00:19:37,580 The students, we didn't really find any effect. That was just laypersons oncolytic. 169 00:19:37,580 --> 00:19:40,780 We did find a nice, differentiated effect. 170 00:19:40,780 --> 00:19:49,970 OK, so it seems that the humanitarian context seems to deep biased anchoring, but still for anchoring from the presence. 171 00:19:49,970 --> 00:19:58,700 It's still visible. So the students are something in between, right, because they have a little higher background. 172 00:19:58,700 --> 00:20:02,390 But again, you know, the numbers are not so high of the students. 173 00:20:02,390 --> 00:20:07,880 So, so, so take. This was a note of caution. When does it matter? 174 00:20:07,880 --> 00:20:16,940 So we had here only the print vignette, no unframed comparison, which in a way is a pity, but, you know, exposed to all this matter. 175 00:20:16,940 --> 00:20:23,450 So we didn't find any significant anchoring effect in humanitarian negotiators or students, but in layperson's. 176 00:20:23,450 --> 00:20:28,940 So it could be that the humanitarian context might matter and needs further research because they're anchoring effect. 177 00:20:28,940 --> 00:20:38,570 Really, if you if you if you talk to experimentalists, if there's one effect which is always there anchoring, it always works. 178 00:20:38,570 --> 00:20:45,980 But here at the end, we are not sure why we could be really so be attributed to the context. 179 00:20:45,980 --> 00:20:59,410 Next one base rate neglect. So that's another very well known bias and actually probably the basis of all sorts of prejudice. 180 00:20:59,410 --> 00:21:05,440 So the base rate fallacy is basically that when you when you would be totally rational, 181 00:21:05,440 --> 00:21:13,540 when calculating probabilities to use base rates as an obvious approach for estimations where no other information is provided. 182 00:21:13,540 --> 00:21:23,030 So the classical entrance tests, for example, to universities or to, I don't know, McKinsey likes to do something like that. 183 00:21:23,030 --> 00:21:28,540 They would ask you, so how many you know, McDonald's are in the world, for example, 184 00:21:28,540 --> 00:21:35,950 so you don't have you need to have some assumption about base rates and then you try to do to go from there. 185 00:21:35,950 --> 00:21:43,550 But mostly, people make evaluations based on surface similarities rather than base rates or statistical realities. 186 00:21:43,550 --> 00:21:49,180 And the tendency to ignore base rates and prefer and so so they have the tendency to ignore base 187 00:21:49,180 --> 00:21:54,520 rates and prefer in the of information over general information when the former is available. 188 00:21:54,520 --> 00:22:06,340 And what I what you have here is like the statute. So so you have Steve who's shy and then you ask people, is a salesperson or is Steve the librarian? 189 00:22:06,340 --> 00:22:16,320 And then you immediately would say, Well, it must be, you know, you associate shyness more with a librarian than it was a salesperson. 190 00:22:16,320 --> 00:22:23,590 Right? It's like your prisoners impression, which is probably true because maybe like the librarians, you know, almost. 191 00:22:23,590 --> 00:22:32,410 But if you look at the base rates of the number of people who are actually they say it's people and who are librarians, 192 00:22:32,410 --> 00:22:39,190 then you should say, Well, it's it. That's the salesperson, OK? 193 00:22:39,190 --> 00:22:46,420 And I always did this with my students and I had I was a bit mean. 194 00:22:46,420 --> 00:22:56,920 So I put a pick of the person with glasses and dispute and and said, Well, Mark likes to listen to Mozart. 195 00:22:56,920 --> 00:23:01,720 And then what is more probable is that the literature professor or a truck driver. 196 00:23:01,720 --> 00:23:07,600 And, you know, usually they would always go for the literature professor, even though, of course, 197 00:23:07,600 --> 00:23:14,000 that much fewer literature professors than truck drivers, even though that might be changing. 198 00:23:14,000 --> 00:23:23,290 OK, now if truck driving gets more gets more attractive because they must go up with the salary. 199 00:23:23,290 --> 00:23:32,830 That was a joke now. So, so the classical based rate test is is the following. 200 00:23:32,830 --> 00:23:39,100 This has been tested was more or less the same numbers on other issues and charts because judges. 201 00:23:39,100 --> 00:23:46,210 So it's all about the base rate, and we kind of adapted it to the humanitarian context. 202 00:23:46,210 --> 00:23:49,510 So your organisation has teamed up with the producer of the track that helps 203 00:23:49,510 --> 00:23:53,260 individuals in the early stages of the virus infection to lead to a longer, 204 00:23:53,260 --> 00:23:59,680 meaningful life. The producer has given you permission to use the drug on recipients in least developed countries and 205 00:23:59,680 --> 00:24:05,260 just made it your responsibility to guarantee that the drug is not sold in developed countries. 206 00:24:05,260 --> 00:24:11,350 The organisation has found ways to offer surplus guarantee, but must now decide which communities to serve. 207 00:24:11,350 --> 00:24:16,850 The total amount of the drug is only sufficient to serve a small fraction of the population. 208 00:24:16,850 --> 00:24:21,880 We want to find the community by the administration of the drug would be most effective. 209 00:24:21,880 --> 00:24:28,630 Assume you have the following information of a population of twenty thousand ten are expected to be infected. 210 00:24:28,630 --> 00:24:36,130 You have been able to conduct tests on the entire community. The test shows positive in 90 percent of infected people. 211 00:24:36,130 --> 00:24:43,210 The test also shows positive one of a million, while also one in thousands of uninfected people. 212 00:24:43,210 --> 00:24:48,350 Your team leader asks you whether the drug should go to this community as an input for her decision. 213 00:24:48,350 --> 00:24:55,180 She wants you to quantify the probability that the individual who a positive test actually is infected. 214 00:24:55,180 --> 00:25:04,410 What's your estimate? So that was their task, and the correct answer is thirty one point one. 215 00:25:04,410 --> 00:25:11,610 OK, so this are those are the results. So the base rate neglect here, 216 00:25:11,610 --> 00:25:20,150 so refined like so so we have the students and the negotiators in the first two one and this is for the layperson strike neutral and trained. 217 00:25:20,150 --> 00:25:26,760 So so we need to redo the graphs a little bit, but didn't have time for that now. 218 00:25:26,760 --> 00:25:36,150 But you you can see the results, so you can see that the the humanitarian negotiators were better in the base, 219 00:25:36,150 --> 00:25:42,450 right neglect, but they were not better. There were actually worse in the neutral frame. 220 00:25:42,450 --> 00:25:49,230 So we did the base rate issue not only in the frame, but also in unframed on neutral context. 221 00:25:49,230 --> 00:25:55,200 And that is that is interesting because it seems again that context matters. 222 00:25:55,200 --> 00:26:04,920 And then or maybe they're just thinking more slowly when they feel they in their own context, in the Italian context. 223 00:26:04,920 --> 00:26:12,510 And here you see the layperson's came. So for all three groups, the humanitarian context matters. 224 00:26:12,510 --> 00:26:17,090 They were all better in the humanitarian context than in the neutral context. 225 00:26:17,090 --> 00:26:25,140 And humanitarians are much more likely to avoid base rate neglect in the humanitarian context, but they have worse. 226 00:26:25,140 --> 00:26:30,780 Interestingly, in comparison to students and delayed persons in the non-French neutral experience. 227 00:26:30,780 --> 00:26:43,130 So again, that shows that the context really matters. And overoptimism, that's an interesting one, because, again, overoptimism. 228 00:26:43,130 --> 00:26:48,920 Everybody is overoptimistic, and it's often explained as an evolutionary trait, because otherwise, 229 00:26:48,920 --> 00:26:55,550 why would you get married or if you ask people when they get married, whether they would get divorced and they would say no. 230 00:26:55,550 --> 00:27:03,500 But of course, like depending on the country, you have 30 percent of the population getting divorced or if you have four start-ups and start-ups, 231 00:27:03,500 --> 00:27:09,970 they all think their business will survive, even though we know that. I don't know 70 percent that goes bust after a year or two. 232 00:27:09,970 --> 00:27:16,790 Right. So overoptimism might be a very good, evolutionarily speaking, a very good trait. 233 00:27:16,790 --> 00:27:21,440 So you compare, but we also know that, you know, 234 00:27:21,440 --> 00:27:28,110 eighty five percent of the population think that they are over average drivers, which is a logical impossibility. 235 00:27:28,110 --> 00:27:33,240 And we found the same thing, for example, with judges and arbitrators. 236 00:27:33,240 --> 00:27:41,690 Say you would like the arbitrators more a bit more modest than the U.S. American judges, but more or less, 237 00:27:41,690 --> 00:27:48,450 you know, it's like something between 80 and 90 percent of them think that they're better than the average judge. 238 00:27:48,450 --> 00:27:53,700 So it's important most people think, you know, you see it every day. 239 00:27:53,700 --> 00:28:03,270 I won't catch COVID or I will not die of cancer, or even though we know that half of the population at one point in their lives has cancer. 240 00:28:03,270 --> 00:28:09,620 So, so and so there are a lot of of those issues where, you know, what's our it? 241 00:28:09,620 --> 00:28:16,460 And we wanted to see whether they are overoptimistic considering themselves because of this phenomenon and all others are but not me. 242 00:28:16,460 --> 00:28:23,720 OK, so you're working for an organisation and in the state undergoing civil war was a mounting humanitarian crisis organisation and 243 00:28:23,720 --> 00:28:30,980 as part of the condition of 10 governmental and non-governmental aid agencies active in humanitarian activity in the region, 244 00:28:30,980 --> 00:28:34,910 the governor of the province, where aid is to be delivered to the civilian population, 245 00:28:34,910 --> 00:28:39,380 has confirmed access to the coalition aid groups to the affected groups. 246 00:28:39,380 --> 00:28:42,500 You are preparing delivery of humanitarian aid. 247 00:28:42,500 --> 00:28:52,220 When you are informed that the central government has revoked the visas for all humanitarian agencies was mandates to all regions of the state. 248 00:28:52,220 --> 00:28:56,810 The central government has issued a press release according to which man is some civil 249 00:28:56,810 --> 00:29:03,200 unrest in parts of the country and therefore foreign humanitarian aid is not necessary. 250 00:29:03,200 --> 00:29:08,440 You have a country that was before Ethiopia. We did not like that. 251 00:29:08,440 --> 00:29:14,720 This is so so all that typical situations, you will find them all the time. 252 00:29:14,720 --> 00:29:20,690 You have a contact to the Ministry of Interior and at the private meeting request clarifications, 253 00:29:20,690 --> 00:29:26,120 the ministry assures you that there has been an error and the results will not be resolved. 254 00:29:26,120 --> 00:29:29,150 No formal assurance has been provided, however. 255 00:29:29,150 --> 00:29:38,240 What is the likelihood for your team to retain their visas and continue the humanitarian missions, and we ask them to give them too much information? 256 00:29:38,240 --> 00:29:43,190 And then we also ask them what is the likelihood for humanitarian workers or other 257 00:29:43,190 --> 00:29:48,790 agencies in the coalition to attend Evanson and again give the longevity for some? 258 00:29:48,790 --> 00:29:55,890 And this is what we find interestingly. 259 00:29:55,890 --> 00:30:07,360 You know, they all more or less the range through at 50, which is a classical focal point here. 260 00:30:07,360 --> 00:30:14,730 You find in the dotted line, you have the students, so they're a little bit more optimistic. 261 00:30:14,730 --> 00:30:20,790 So they exhibited the optimism bias first and you share this really don't OK? 262 00:30:20,790 --> 00:30:29,160 The layperson also don't. So so so this is interesting because as I told you before, 263 00:30:29,160 --> 00:30:38,250 overoptimism is so well proven and the only people who are not overoptimistic are those clinically depressed. 264 00:30:38,250 --> 00:30:47,670 And it seems that at least the humanitarians seem not to have overall optimism if you talk to them, 265 00:30:47,670 --> 00:30:55,650 if you go and I've been to some of those places and those workshops before COVID, they are under a lot of strain for sure. 266 00:30:55,650 --> 00:31:01,510 I'm not saying they're depressed, but they're under a lot of strength. 267 00:31:01,510 --> 00:31:09,350 So, but also since we also tested the students and the layperson, somehow context seems to matter. 268 00:31:09,350 --> 00:31:16,520 To conclude, so you really are in negotiations, shares similar cognitive biases. 269 00:31:16,520 --> 00:31:19,370 Lay people, but we cannot equate it. 270 00:31:19,370 --> 00:31:30,290 So with regard to some biases and in the frontline context, they're more cautious vis-a-vis anchoring and essentially overoptimism. 271 00:31:30,290 --> 00:31:37,550 Insights on grounded rationality we find can thus not easily be transposed 272 00:31:37,550 --> 00:31:43,850 from laypeople and context free Williams to professionals and their context. 273 00:31:43,850 --> 00:31:50,750 So our takeaway from that is that the UN needs to proceed with caution in applying behavioural economics 274 00:31:50,750 --> 00:31:58,260 and cognitive psychological insights because the experiments from the lab cannot be so easily transposed. 275 00:31:58,260 --> 00:32:08,720 I mean, external validity has always been an issue discussed in experimental economics, but there are few comparisons as laypeople. 276 00:32:08,720 --> 00:32:13,370 And then there are studies to show professionals showing as well. 277 00:32:13,370 --> 00:32:20,240 Others say that professionals show with less, but we need much more research on them. 278 00:32:20,240 --> 00:32:22,160 But. 279 00:32:22,160 --> 00:32:30,830 This sort, even though it's like nitty gritty, we are just talking about frontline humanitarian negotiators, we are just talking about for biases. 280 00:32:30,830 --> 00:32:37,550 This is the sort of research I believe we need if you want to give evidence based policy advice, 281 00:32:37,550 --> 00:32:46,210 which helps us because this research helps us to understand how actors decide when applying international law generally that we need more research. 282 00:32:46,210 --> 00:32:56,300 Sure. And and I ajello more specifically now, humanitarians are not the only person HL, but there is, of course, more to do. 283 00:32:56,300 --> 00:33:04,550 We hope to do a study was the UN Office of Counterterrorism. So they, they they are also trying to. 284 00:33:04,550 --> 00:33:06,920 But of course, you, for example, would say it's not. 285 00:33:06,920 --> 00:33:14,450 It's basically impossible to do experiments with terrorists or doing experiments in counterterrorism. 286 00:33:14,450 --> 00:33:24,790 And we are doing a second follow up study, which is which I'm not reporting here, also because the humanitarians on application of international law. 287 00:33:24,790 --> 00:33:32,250 So the interpretation of of HL, but I'm not reporting because it's still ongoing. 288 00:33:32,250 --> 00:33:43,350 OK, so I stop here. That was it from my side, and I'm much looking forward to any questions. 289 00:33:43,350 --> 00:33:56,220 Thank you, Anna, for this really informative and also intriguing presentation, and please join me to give a true round of applause to Anna Baker. 290 00:33:56,220 --> 00:34:01,900 Yes. And now we move to the discussion part. 291 00:34:01,900 --> 00:34:06,580 And. And thank you, Natasha Co-convenor. 292 00:34:06,580 --> 00:34:16,750 To remind you that you can just use a restaurant function at the bottom of the tap interaction tap or just tap. 293 00:34:16,750 --> 00:34:34,440 I have a question in the chat. Let me check whether I made some of the comments or questions. 294 00:34:34,440 --> 00:34:39,290 OK. Natasha, please. Thanks. 295 00:34:39,290 --> 00:34:42,800 Thanks, Jason. Thank you so much. And this is absolutely fascinating. 296 00:34:42,800 --> 00:34:47,760 I'm afraid this is very different territory for what I'm used to, so I apologise in advance. 297 00:34:47,760 --> 00:34:53,150 If my question is, is this is a silly one because behavioural economics are completely new to me. 298 00:34:53,150 --> 00:34:57,830 But I totally understand why you do it, and it sounds very important and fascinating. 299 00:34:57,830 --> 00:35:07,670 But one thing I wondered is kind of the the aim of this or that or the the outcome in terms of and is that 300 00:35:07,670 --> 00:35:14,480 the point of or is part of the points of the behavioural analysis to inform the shaping of legal norms? 301 00:35:14,480 --> 00:35:23,990 Or is it more about policy advice and how to how to train people on the ground or a bit of a mix of or something that I have completely overlooked? 302 00:35:23,990 --> 00:35:31,370 I'm just curious about that. Thank you. Thanks, Natasha. 303 00:35:31,370 --> 00:35:40,160 So if you allow me to share another PowerPoint read, I have a nice graph where you see how we can. 304 00:35:40,160 --> 00:35:44,290 Is that OK for you? Thank you. 305 00:35:44,290 --> 00:35:53,790 OK. Sea and open pretty. 306 00:35:53,790 --> 00:36:06,280 So that was a presentation. It's called a roadmap and some stop signs. 307 00:36:06,280 --> 00:36:19,790 But using behavioural insights in international law, so it can be many things like like, you can use it in as Queensland to. 308 00:36:19,790 --> 00:36:34,660 So what you see here is like. So so what you see here is basically so you have behavioural science interventions, and this is the targeted actor. 309 00:36:34,660 --> 00:36:45,040 So you can from an internationally, you can target the national level so you can say you give advice or give guidelines that can 310 00:36:45,040 --> 00:36:51,400 be different normative forms of law or whatever in order to nudge consumers or business. 311 00:36:51,400 --> 00:36:56,860 So this was a presentation in the context of sustainable development, and this is, of course, something. 312 00:36:56,860 --> 00:37:05,740 But, you know, look at the W.H.O., right? So you would, for example, drop the age old guidelines on plain packaging, tobacco plain packaging. 313 00:37:05,740 --> 00:37:07,930 They would be ultimately targeting consumers. 314 00:37:07,930 --> 00:37:16,360 But of course, the first step they target the states, which then transform on the national international level. 315 00:37:16,360 --> 00:37:27,640 You can use it in in in the process of negotiations like, for example, we do have some studies on whether it matters, 316 00:37:27,640 --> 00:37:34,750 whether you have positive or negative lists when you negotiate, for example, trade or the service agreement. 317 00:37:34,750 --> 00:37:43,810 We also have insights on and that's on the substantive level of the treaty on the the that default rules matter. 318 00:37:43,810 --> 00:37:49,930 So there has been this fantastic study by Gene Galbraith that whether you reduce the 319 00:37:49,930 --> 00:37:57,970 jurisdiction of the ICJ or any other international court as an opt in or opt out. 320 00:37:57,970 --> 00:38:03,730 So if it's in the treaty and you need to make a reservation, it's about how many more states stay in. 321 00:38:03,730 --> 00:38:09,550 Whereas if you have an additional protocol like, for example, in the human rights treaties, then much less states get it. 322 00:38:09,550 --> 00:38:14,500 And so, so it shouldn't matter from a rational perspective because you would assume 323 00:38:14,500 --> 00:38:19,150 while the state knows whether it wants to have to accept jurisdiction or not, 324 00:38:19,150 --> 00:38:29,590 but it matters. And, you know, reservations are very reservations and objections to reservations are super interesting topic on that, right? 325 00:38:29,590 --> 00:38:36,730 So for example, I would say, especially in human rights studies, we would need to design it differently. 326 00:38:36,730 --> 00:38:41,890 So, so so it can be used either in the process of negotiation. 327 00:38:41,890 --> 00:38:48,730 And again, here also, diplomats are trained in rational choice negotiation theory, and we would say, well, 328 00:38:48,730 --> 00:38:59,800 this is probably insufficient and there has been so many writings on on impediments to negotiation success. 329 00:38:59,800 --> 00:39:05,560 If you totally stay in a in a rational choice framework, 330 00:39:05,560 --> 00:39:12,460 so so it might matter for diplomats and negotiators to use behavioural insights, but also for the treaty text. 331 00:39:12,460 --> 00:39:16,450 And then, of course, it can be used in international organisations. 332 00:39:16,450 --> 00:39:19,240 That's the last sentence of the quote of Anthony Pettis, 333 00:39:19,240 --> 00:39:25,570 which I had put in order to ameliorate the internal workings of the international organisations. 334 00:39:25,570 --> 00:39:37,630 But also, and that's more the CC Agenda Project help their own people to deep bias in their book Pay or help them to be able. 335 00:39:37,630 --> 00:39:45,760 And that's also the aim of the situation with this sort of training is that to see behavioural biases and duress in their 336 00:39:45,760 --> 00:39:55,790 counter practise in order to be able to use them and have a better negotiation result from a humanitarian perspective. 337 00:39:55,790 --> 00:40:02,290 OK, and then international organisations. And that's what I had said when they used there. 338 00:40:02,290 --> 00:40:05,690 Yeah, and they can use it in between themselves as well. 339 00:40:05,690 --> 00:40:11,260 So there are many, many, many possibilities. 340 00:40:11,260 --> 00:40:18,500 I hope that answers your question. It does this, you know, that most of that you can all do this conceptually. 341 00:40:18,500 --> 00:40:29,440 You can all use all those biases and jurists, which we have, which we know about and apply them like this, like, you know, I can do this as my best. 342 00:40:29,440 --> 00:40:31,750 And we could have gone to seise the agenda, as I said, 343 00:40:31,750 --> 00:40:36,460 let's give them a training handbook and say, well, you know, those are all the biases they exist. 344 00:40:36,460 --> 00:40:41,590 Those are likely to come up in humanitarian negotiations. Train your people on them. 345 00:40:41,590 --> 00:40:42,970 Or we can say, well, 346 00:40:42,970 --> 00:40:52,300 maybe that is because there is we know there is a problem with external validity of those experiments because they are mainly conducted 347 00:40:52,300 --> 00:41:01,890 for students because we can so easily conduct experiments with students and be it's so difficult to get the real persons to do. 348 00:41:01,890 --> 00:41:09,010 Does that answer it? It does. Thank you so much, and I'm glad to see you in a way that it has so many potential applications because 349 00:41:09,010 --> 00:41:15,600 it sounds like it's something will be really useful to a lot of different contexts. I had just one more question if you don't mind, Chris. 350 00:41:15,600 --> 00:41:18,600 Thank you. And you don't mind to add. Thank you very much. 351 00:41:18,600 --> 00:41:28,020 It was about the process for how you come up with the vignette because they seem very carefully constructed as a kind of, 352 00:41:28,020 --> 00:41:31,770 you know, the good test test cases test example. 353 00:41:31,770 --> 00:41:37,590 I just wondered if you are drawing on experiences, particular experiences, 354 00:41:37,590 --> 00:41:43,710 how you how you pick, how you, how you frame those, those vignettes in those questions. 355 00:41:43,710 --> 00:41:48,600 I know I know some of them are deliberately abstractly framed and deliberately framed using humanitarian issues. 356 00:41:48,600 --> 00:41:54,180 But sometimes I think it can be quite tricky to know how to also come up with the wording. 357 00:41:54,180 --> 00:41:57,010 They see that they must be have to be very carefully monitored. 358 00:41:57,010 --> 00:42:01,740 So I just wondered about I was curious about the process that you go through when you're writing those. 359 00:42:01,740 --> 00:42:05,820 Yeah, and that's actually the most time intensive. 360 00:42:05,820 --> 00:42:15,320 So, so, you know, a study like this, we started in 2018, right, and finished this year, and that's only the first study. 361 00:42:15,320 --> 00:42:19,140 Well, some of it was also due to COVID and stuff like that. 362 00:42:19,140 --> 00:42:25,800 But yes, designing the vignette is the most difficult part. 363 00:42:25,800 --> 00:42:33,480 And you know, it starts with it's something on vaccination expiry or something like that. 364 00:42:33,480 --> 00:42:37,280 And then we talk to medical doctors and then all this is medically wrong. And so like a.