humanitarian.info

I posted on UN Dispatch today about the risks of failing to learn from what’s gone before. Needless to say, documentation of previous humanitarian information work is thin on the ground, but that’s exactly why we need to create mechanisms for better institutional memory. The idea that Haiti was Year Zero for humanitarian information isn’t always explicit, but the general tone of a lot of online commentary definitely labours under its influence, particularly from those who don’t necessarily understand the work.

Haiti has exposed both the benefits and risks of the web-enabled response to disaster. The positive side has been written about extensively – wider awareness, greater levels of engagement, access to potentially valuable resources (whether satellite imagery or volunteer time). I’ll go on record here as believing that these are all to the good, and possibly the starting point for greater good in the future; although if we’re going to get to that future good, everybody’s going to have to check their egos at the door.

Unfortunately the negative side is almost completely ignored. Wider awareness that isn’t informed by grounded understanding is useless, and occasionally dangerous; greater levels of engagement need to harnessed in the right direction if they’re going to have any impact; and access to resources is unevenly distributed and comes loaded with assumptions that are often hidden from the people who provide those resources. The UN Dispatch blog post series is a good start but it’s no substitute for hard discussion.

Haiti also amplified another trend – the involvement of the public in debates about humanitarian principles and practice. This is something that we should welcome, but I’m a big fan of people knowing the limitations of their own experiences. I don’t feel the need to pontificate on the finer points of hospital management, because I’m not competent to discuss them; yet for some reason people feel that they’re entitled to pass judgement on the humanitarian sector based on what they just read in Wired or Newsweek.

While the recommendations in the Disaster 2.0 itself are likely to have positive impacts, they’re generally about how the humanitarian community can better utilize technology, rather than how “volunteer and technology communities” (and my word, what a frightful phrase that is) can perform better. Likewise this report on Volunteer Technology Communities does nothing more than describe what they do, rather than the impact they can have.

If we put all of this together, we can identify the problem at the heart of this debate; and possibly why some on the “V&TC” side has become increasingly defensive. In all the discussion about how social media might affect the humanitarian community, as far as I know, nobody has yet articulated a clear theory of change. That was what I was hoping for last year, when I challenged the crowdsourcing community to clearly articulate how the inputs they deliver will lead to the outputs they claim will result.

If we’re going to work actively to ensure that our use of technology is going to help anybody in post-disaster/post-conflict responses (whether 2.0 or just plain 1.0), all of us need to be much clearer about exactly how it’s going to help. I emphasise that this is a management problem, not a technology problem; and to their credit the Disaster 2.0 report writers very clearly targeted their recommendations on addressing the management issues, at least in some small part. Let’s see where this takes us.

When aid workers meet during an emergency response – in a bar, in a workshop, in a meeting – you can guarantee that within five minutes, one of them will have resorted to the single most important question in the industry. To the relative newbie, the question seems innocent enough, but beware! It’s actually a finely calibrated conversational gambit that is designed to test and establish credibility:

So, where were you working before this?

The first stage in this process is simple: where you’ve worked sends a clear signal as to how hardcore you are. If you reply “I spent two years in Ghana with the Peace Corps”, nobody will take you seriously. If you’re able to casually respond “Oh, I was in DRC last month; had a 2-month stint in Afghanistan before that; and was covering for some colleagues in Somalia until we had to be evacuated”, then you’re safe. The important factors here are: location, duration and role. If you arrived in Haiti a year after the earthquake, you are nowhere near as hardcore as somebody who arrived four days after the quake and stayed for 2 weeks.

Once initial credibility is established, this enables your interrogator to follow up: who were you working with? Again, be careful. If you say “Homeopaths without Borders”, nobody will speak to you again. This is a minefield, and I’m never quite sure how to negotiate it. If I say that I worked on a PRT in Afghanistan, it’s +3 hardcore points but -2 “humanitarian principles” points. Sometimes I think it would be easier if everybody had a D&D-style character sheet that they could just hand out when they arrive. It would be more useful than a CV. Almost everybody brings their own prejudices to the table, so you’ll have to get through this one on your own.

The final phase is, “Oh, you must know [insert name of friend X].” If you can establish a mutual friend in the sector that both of you trust, the network effect takes hold and a certain amount of trust is automatic. A mutual friend is a handy shortcut that enables you to avoid the tedious business of actually finding out whether somebody is any good at their job and should be taken seriously. However, if you name somebody who your interlocutor thinks is an incompetent buffoon, you’re going to have to do a lot of work to gain their trust – and their cooperation, if you should need it.

That first question is doing a lot of work in the aid industry. I don’t like it, but I do it myself. I sometimes catch myself worrying whether my own answers are sufficiently credible (especially after taking a reasonably long break from work), and thinking that the whole thing is ridiculous. The problem is, I can’t come up with a better alternative…

There’s no point debating about whether the George Clooney-fronted Sentinel Satellite Project will have an impact. Since I can’t think of a metric for impact that isn’t insane (surprise!) that’s not an argument that we can even start, let alone win. People working in the aid industry get the sense that there’s something wrong with the whole thing, but find it difficult to articulate the reason for that sense.

Some argue that celebrities don’t know much about the issues they pontificate on. but that clearly isn’t true. Everybody agrees that Bono knows his beans, and this article on Clooney makes clear that (with a little help from John Prendergast) he’s reasonably well-informed. It’s also true that there isn’t an exam you have to pass before you’re allowed to care about something.

Some argue that sending celebrities into disaster zones, while it draws the cameras, doesn’t really raise awareness of the issues or raise funding by the barrel. This is a stronger line of attack, but still not that strong. Considering how much it costs, and how widespread the practice is, it seems likely that the return on investment is reasonably good, as far as it can be measured.

The furthest that most people are prepared to go is to worry that the presence of Clooney – or any celebrity – distracts from the issue rather than drawing attention to it. But everybody knows that you’re not sending Shakira to Bangladesh because it makes Bangladeshi kids happy – they haven’t got a clue who she is. Shakira’s in Dhaka because you get UNICEF footage on the 6 o’clock news.¹

What makes us uncomfortable is that when you put a celebrity into the middle of (for instance) South Sudan, it throws our own culture into sharp relief. Sending a celebrity to draw attention to an issue makes us realise that a culture that needs celebrity in that way has gone wrong somewhere. We can’t articulate the real problem because we’ve been trained not to notice it, although locals do.

We think that sending in a celebrity will draw attention to the issues we care about because we’ve been trained that this is the nature of celebrity – a bright light shining in a dull world. Yet the presence of somebody deemed more important than others for nothing more than their visibility spits in the face of our humanitarian principles, which tell us that everybody is equally important.

Celebrities gain their status at the expense of any decent conception of what humans are worth. Most of the time we don’t notice it because nearly everybody in our societies is infected, and because many of us want the coin of celebrity ourselves. We imagine that earning that coin will improve our lives, but that’s a narcissistic fantasy that will later be used to sell us things.

It’s not the fault of celebrities themselves, since they’re oblivious to the nature of their status. I don’t have anything against George Clooney – I think he’s a fine actor, and I enjoy his movies – or any other celebrity. They can choose to spend their celebrity however they want, and I applaud anybody who decides to spend it drawing attention to issues that I also believe are important.

Spending that coin costs the celebrity nothing, however; in fact it buys them more celebrity at the expense of the cause that they’re promoting. Our vain hope is that the celebrity will rub off on the cause, but spending the coin of celebrity makes all of us a little poorer. Celebrity deforms the entire humanitarian enterprise; we should want no part of it.

1. I am perfectly happy to single UNICEF out here. [↩]

For me, the value of the 4636 system was never been just about the actionability of the data — I saw the reports were profoundly flawed as soon as they began to arrive. There were amazing sparks of actionability — glimpses of something we can all fumble toward somehow — but it was always clear that these were no replacement for well-tested logistics plan.

I know that you and others working in the crowdsourcing ecosystem in Haiti must have recognised this. Yet it didn’t seem like the public statements at the time reflected this; it doesn’t seem as if any of the interviews or conference presentations or media articles reflected this; and most importantly of all, it didn’t seem as if your message to the people of Haiti reflected this. It’s interesting that the rhetoric now – after the funding is in place, after the conference invitations are in, after the media attention has been secured – is much more circumspect than it was previously.

And what exactly was the value of the 4636 system from your perspective? This is what I’ve been really interested to hear.

But, while I was quite aware of how utterly limited the data were, this never made me question the basic value of an open communications system during this crisis.

I guess this is where my old-school sensibilities are a handicap, because my view is that if you’ve set up an information system and the data it’s producing is not that useful, there’s a problem with that system. If we agree for a moment that there’s nothing wrong with your technology, that means there’s a problem with the inputs – because as we all know, the basic rule is garbage in, garbage out – yet the inputs are exactly what you’re claiming are special about your system.

When I am in a crisis, I want strong systems in place that let me text anything I want.

The problem here is that your needs are not necessarily a useful guide to the needs of other people. I’ll say that again, because it’s a fallacy that affects nearly everybody starting out in relief or development: your needs are not necessarily a useful guide to the needs of other people. It’s nothing personal: my needs aren’t a useful guide to the needs of other people either. Hell, my needs probably aren’t a useful guide to your needs, or vice versa.

However I agree that open communications (along several tracks) are valuable, and that accessible communications systems can be as important as accessible food distribution networks. Different criteria apply in a disaster, though; the food distribution networks run by Wal-Mart aren’t as effective in South Sudan. This doesn’t mean that we can’t learn something from Wal-Mart – just that we need to be careful about what we learn.

The same holds for technology; context is vital and we can’t just transplant things that work really well in one place to another. And besides, there’s already a strong system in place that lets you text anything you want – it’s called the mobile phone network.

There may be a limited amount of resources on a particular day in a particular crisis — but in the long term I believe that there are millions of people who could be engaged in productive and powerful work through diverse and open platforms.

I also believe that there are millions of people who could be engaged, it’s just that I don’t believe that millions will be engaged. There’s already ways in which those people could be involved if they wanted to, and the vast majority of them simply don’t want to. I long ago ceased to be naïve enough to believe that everybody shared my value set, so I’m not condemning anybody here; I just fail to see why anybody thinks that technology will magically make people more altruistic.¹

The only vaguely convincing argument I’ve seen says that the internet – and in this case crowdsourcing – has lowered the barrier to entry to get involved in humanitarian activities. Unfortunately that doesn’t mean that more people will get involved, because the barrier to entry was also lowered on a huge range of other, more interesting activities that people will choose instead of the difficult and depressing work of humanitarian activities. A lower entry barrier is also a lower exit barrier, as Ushahidi found out somewhere between Haiti and Pakistan.

The bitter truth is that you may have to face the possibility that the crowdsourced response to Haiti was possible mainly because Haiti is within flying distance of the US, has a long history with the US (including a large diaspora community), and was a particularly media-friendly disaster. I’m not saying that this was definitely the reason why that particular crowdsourced response ever got off the ground, but I’m interested to hear what this discussion looks like from the other side, rather than just get fed another inane marketing video.

Particularly I see a great promise in networks which help people in crisis see their own neighborhood, to help each other, rather than always relying on the aid worker. So, information which seems predictable and “not novel enough” to an aid worker might be quite powerful to someone who lives down the street.

I agree completely. We are already seeing this (something I’m going to talk about soon in a follow-up article) but this isn’t the model of crowdsourcing that was developed in Haiti. I remain unconvinced that the Haiti model is a useful model for anything else – in fact I remain unconvinced that the Haiti model was a particularly useful model in Haiti. All I’ve been asking for is some evidence that it warrants the massive amount of hyperbole that surrounded it – and the opportunity costs of implementing it.

The other point to note here is that the fact that different types of information have different value to different people is trivially true. That being the case, then who gets the value in the type of information that Ushahidi et al gathered in Haiti – the aid worker or the person who lives down the street? Because it seems to me that the information gathered in Haiti might well be useful to somebody who lives down the street, but that’s not where the project was driving (because it was subject to exactly the same constraints as everybody else in reaching the street).

In this sense, I have a hard time understanding how you could question whether it has “any substantive value.” To me the value is self-evident, if complex, and problematic for traditional patterns of response.

The value may be self-evident to you, and others working on crowdsourcing projects like 4636 or Ushahidi, but they’re not self-evident to others. You can’t accept people to accept your claims just because you really, really believe; you need to present a convincing narrative. The narrative around Ushahidi is really convincing to technologists because it tells them that technologists can really really make a difference.

That narrative is a lot less convincing to people who actually do the work. All I wanted was for proponents of crowdsourcing to step up to the plate with something more than the usual vague gushing about the revolutionary power of the crowd, with some more awareness of the embedded technocratic ideology that comes with the package, and to be backed up by slightly more evidence than second-hand anecdotes from a US marine. I freely admit that I could be wrong – that was why I wrote the article – but I need to be shown to be wrong.

All tools are suitable for some tasks, and not for others. It strikes me as strange that Ushahidi is proposed as a mapping solution for citizen journalism; for election monitoring; for human rights monitoring; for emergency response; for clearing snow from American cities; and for pretty much any activity you care to mention. It strikes me as singularly unlikely that it’s a solution for all of them.

Generally speaking, the rule of tools is this: the wider the range of tasks that a tool can perform, the less effective it is for any given task within that range. There is no shame in anybody anywhere standing up and saying, “it turns out that our tool isn’t suitable for this particular taskâ€

1. I have to refer back to Shirky again. The opening chapter of Cognitive Surplus contrasts Ushahidi and Lolcats as two ways in which that surplus can be engaged; he judges one to be more constructive and one less constructive, but both are deemed more constructive than watching the television. This is bullshit. You learn more from watching Unreported World than you do from watching X Factor, but a lot more people watch the latter. Ushahidi maps to Uncovered Stories, Lolcats to X Factor, and you can do the math about how many people are likely to contribute to any given Ushahidi instance. [↩]

I’m glad that Bear vs Shark helped Chris (and others, based on emails I received at the time) to think through how we can balance “data entry” with “story telling”. The reason why this is important embodies one of my concerns about the web in general: that it has a mediating effect, distancing us from the world of the actual. You may not share that concern, but my own experience suggests that this mediation undermines the sense of solidarity that lies behind the humanitarian impulse – or indeed civic responsibility of any kind.

Another of my concerns is the ideology that is smuggled in with social media, or at least with a lot of the rhetoric about social media. I don’t use Facebook for anything other than maintaining a register of my extended network and opening up lines of communication, because (as many people have started to notice) Facebook is not neutral. Embedded in its design is a particular view of the world – this is true of all social media – and it’s not a world that I necessarily want to live in.¹

Like fish in the sea, most people are blind to this, and so discussion focuses entirely on the cool things these tools do rather than the social values that they embody. This discourse is highly valued by companies that wish to distract you from the fact that what they really want to do is sell you something, and Clay Shirky is the undisputed heavyweight master of this theatre of misdirection. While he pretends to insight into how social media is changing the world, what he’s really doing is telling the over-mediated bourgeoisie that we’ll become the revolutionary vanguard when we buy an iPhone app.

Obviously, we won’t.

For me, crowdsourcing brings together both of these concerns in a single package wrapped up in that rhetoric. That doesn’t mean that I think crowdsourcing heralds the end of civilisation as we know it, or even that I oppose crowdsourcing as a useful tool for all kinds of tasks. The reason I wrote about crowdsourcing in the last year is simply because I think the humanitarian enterprise is in dire straits, and I don’t think technology will fix that. In fact, technology can be a distraction from the real problems if it isn’t handled well.

If you read what I’ve written about crowdsourcing in the past, most of my frustrations have not been with the concept of crowdsourcing, but with the rhetoric around it:

1. The Crisis Category Error: the pretence that “crisis” is a meaningful category when it’s used to bring together complex social phenomena that are completely unrelated.²
2. The If All You Have Is A Hammer Fallacy: the claim that because crowdsourcing as a tool has a perceived utility in event type A, that utility is generalizable to event type B.
3. The Year Zero Expert Syndrome: the pronouncement that crowdsourcing has / is / will revolutionize a sector, made by people with little to no actual experience of that sector.
4. The Unstructured Use Case Error: the obfuscation of crowdsourcing as a way for communities to help themselves, crowdsourcing as a way to improve efficiencies in the humanitarian sector, and crowdsourcing as a way for outsiders to feel like they’re helping communities.

Now while you might disagree with my frustrations – and you’re welcome to do so – none of them are about the concept of crowdsourcing per se.³ That crowdsourcing exists, and can be useful, and is worth working on, I have no doubt at all; that it has implications for business as usual in relief and development, I have serious doubts but I’m ready to be persuaded. My problem is the gap between the rhetoric and the reality, and the way in which that gap is exploited to sell crowdsourcing to organisations and individuals for whom it has marginal value.

That MobileActive article did two things: presented a professional opinion (although not according to Robert Munro, who for some mysterious reason believes that experience of actually doing humanitarian coordination somehow disqualifies one from having an opinion on humanitarian coordination) and asked a lay question. So the really interesting issue for me is: Why are attempts to open up serious discussion about crowdsourcing usually painted as hostile opposition; as a futile rearguard action by entrenched vested interests who SIMPLY DON’T UNDERSTAND THE FUTURE?

Luckily, I have the answer to that (it’s one of the few answers I do have), but we need to flash back on Clay Shirky. The over-mediated bourgeoisie (myself included) cry out for a sense of purpose; there are lots of people paid to pander to our quest for purpose (such as Shirky); technology is one of the things that we feel might give us purpose, especially if the marketing material tells us it’s all about “making connections”; and the rhetoric of technocratic utopianism seems like it gives us revolutionary credentials. And if we believe that you’re the revolutionary vanguard, then obviously we need an establishment to set ourselve against. No Tsar, no Bolsheviks, after all.

The reason that I wrote that MobileActive article, and the reason why I took a more aggressive tone in it, was to balance out the sort of rhetoric that I mentioned above. My tone will differ depending on what I’m writing about, where I’m writing and who I’m writing for; I took it for granted that I was addressing an audience that was mature enough to accept and engage with constructive criticism, which is why Chris’ reply is so welcome (as were nearly all of the other comments on the original article). Not least because it has given me an opportunity to write about this much wider issue around technology in the humanitarian sector.