Author(s)

Lisa Daoud & Edmond Wach

The relationship between agility and the use of information technologies can be obvious, it is not systematic. The 2018 Autumn School on Humanitarian Aid brought together experts on this issue from different horizons. This article is based on the discussions that took place on that occasion and a number of messages that emerged about using information technology in aid programmes.

 

Introduction

How can we improve our decision-making in crisis contexts? How can we adapt our responses more rapidly? How can we ensure that we learn continuously by monitoring change? Very often, the answer to these three questions is the same: information technology! It seems obvious: if rapid decision-making processes depend on having information in real time, and if the continuous improvement of projects depends on having access to reliable data, information systems and IT tools would appear to be a major asset, or even the way forward. As such, it is revealing that the concept of agility comes from the field of software development to describe a method which brings software designers closer to end users, thus reducing development cycles.

As humanitarians we have seen the birth and growth of IT solutions that allow contextual and project monitoring data to be collected, processed and combined. For example, consider the changes that have been brought by RedRose1in terms of managing cash transfer programmes, or HDX2in terms of needs analysis. The organisation of international aid can now be based on data, which is often quantitative, and which is supposed to allow decision-makers in the field and at headquarters to make more enlightened decisions, at the right time, based on convincing evidence. In the business sector, community platforms – with Airbnb and BlablaCar at the forefront – also seem to show that technological advances can allow more participation and control by those who use a service: the International Rescue Committee (IRC), for example, took inspiration from this to develop Serviceinfo, a platform that allows Syrian refugees in Lebanon to rate humanitarian services based on the principle of crowdsourcing in order to be more reactive and closer to beneficiaries – in other words, more agile. But is the relationship between technological tools and the quality of aid really so obvious?

Though we have to recognise what information technology has contributed to agility, this must not prevent us from observing that it is not always as positive.

IT solutions seem necessary, at least as components of an agility plan, but what practices should be adopted to ensure that the technology really contributes to agile, high-quality aid?

Message 1: When a technology is used, this needs to be for specific uses and for a specific decision-maker. 

The ICRC’s report on the use of technology during interventions in conflict zones concludes that “The future (and the present) is digital”. Being “digitally-prepared” would allow organisations to supply more accountable and high quality services. There is nevertheless a nuance in the report: the deployment of technological solutions needs to be seen as a “means to an end, rather than an end in itself – which is still often the case”. So, what is the end? Why do we want data in a crisis context?

Answering these questions and clarifying the fact that data should serve operational and institutional decision-making is already half the battle. What decisions need to be made and where does this take place? In other words, what decision-maker is the data directed at and for what ‘purpose’?

The development of information technology allows faster access to data. The geographical hubs at headquarters in Paris, Brussels or Geneva can gain access (sometimes in real time) to their teams’ contextual and project monitoring data. As such, it is tempting to give the responsibility of making decisions to top management who are like the captain at the helm, and who feel that, in this big global village, they have as much, if not more, information than in the field. But does this not run contrary to the idea of contextual, rapid and effective decision-making? Has the centralisation of decision-making made possible by the dissemination of information somehow blocked the agility of responses in the field? As for the monitoring and evaluation of agility, which was described in a recent study as an ‘under-developed art’3, it does not appear to have found a ‘champion’ for the moment. Worse still, it is sometimes the need for agility that is given as a reason for the lack of preparation and structure of monitoring systems. In the end, though the monitoring of indicators does contribute to continuous improvement, the systems that are created often only contribute to… donor reporting.

If it is intended to increase agility, the digitisation of a process needs to be preceded by an analysis of the decisions that the process aims to support and their logic. If roles and responsibilities are not formalised and the decision-making method is not clarified, it will be difficult to choose and configure an appropriate technology that allows efficiency and quality in general to be improved: indeed, there are different needs depending on the level targeted. As such, the example of organisational indicators is revealing: top management teams often instinctively want to install ‘business intelligence’ software that allows combined indicators to be visualised. But is it really a priority for country offices to send information about ‘the number of facilities created’ every month when managing activities and measuring change represent challenges of another nature which require much more fine-tuned analysis and relevant IT solutions? The use of the technology will be different in the field, and different too between the different levels in a particular context. We therefore have to accept that what works for one decision-maker will not necessarily meet the needs of all decision-makers, at every level and in all organisations. In the absence of a universal tool, and given how difficult it is to transpose one set up (or one tool) to another context without adapting it, adopting the appropriate technology involves situation analysis and risk-taking.

What is more, using technology makes us question decision-making processes: what determines decision-makers’ choices? Is it experience, shared knowledge, or objective data? No doubt it is a little of all three, and it is interesting to note that recent studies (outside the humanitarian sector4) show that individuals tend to rationalise and/or justify their decisions based on data after the fact rather than the other way around, as is often assumed. It is also frequently interpersonal relations that produce information and knowledge (and therefore decisions), rather than self-generated information based on data of varying reliability. Studies by Groupe URD5 and ALNAP6 on decision-making should help to move the debate forward.

Message 2: If individuals and organisations do not change their tech culture beforehand, digitalisation will not have any effect, or it may even make the current situation worse.

The aid sector is not immune to the paradoxes that society in general has been faced with regarding new information and communication technologies. If they are used with certain intentions, these technologies can make the world more participatory and accountable (freedom of expression, decentralisation of transactions), but with others they become instruments that make processes more cumbersome with potential totalitarian tendencies (labelling and surveillance, and even the identification and punishment of deviant behaviour). In short, ‘tech’ is what we make of it.

As we know them today, new information and communication technologies are oriented towards quantity and output. Data do not intrinsically contain any ‘knowledge’ and the data produced by institutional information systems rarely allow real-team reflection (they need at least to be cleaned and probably to be contextualized and triangulated). What is more, this mass of data that is difficult to digest – particularly without special skills such as those of a data scientist – is not sufficiently used to analyse the relevance of the response, and sometimes just contributes to a sort of ‘data war’: as implied by Alice Obrecht from ALNAP, the more data you have, the greater your share of the ‘humanitarian market’.

What is more, information software has an annoying habit of fixing data methods, processes, calculations and categorisations, in order to ensure that their formulas and algorithms are in good health. Many of us, I am sure, when we are analysing a system with a view to digitising it, dream of hearing a user say “I always do it like this regardless of the situation or the context”. In other words, the opposite of agility… And yet, the informal functioning, changed objectives and rapid decision-making that are the enemies of IT developers, are particularly common in humanitarian action. Consequently, if algorithms are given a major role in supporting decision-making, they need to be adaptable, transparent and accompanied by ‘human’ appeals and monitoring mechanisms that guarantee that the project has the necessary flexibility (for example the Proxy Means Test used in Lebanon to select the refugees to be included in cash transfer programmes). Also, the financial dimension should not be neglected: relevant information systems require continuous investment in order to anticipate changes.

In terms of mentalities within the sector, there are two preconditions: on the one hand, we must stop being ‘tech optimists’ at all costs, believing in short-term investment and considering the harmonisation of tools to be a magic solution. As François Grünewald from Groupe URD says, we need to be ‘knowledge-driven’ rather than ‘data-driven’. Besides, we need to promote the values and approaches that we want to use, but we also need to identify the actions and decisions that we want to make more efficient so that the technologies that are developed represent and serve them. In other words, we need to respect our humanitarian culture!

Message 3: New information and communication technology will not save the world, or the humanitarian system…

… and will not make up for lack of skills (for example in management), poor contextual analysis, or the failure to listen.

Organisations, but especially the individuals within these organisations, should invest primarily in high-impact technology (rather than those with a presumed impact). Anyone who has attended a forum on aid and technology will have noticed how the humanitarian sector is not immune to ‘trends’. In the words of Sarah Selford, who currently runs the Humanitarian Data Center in The Hague, “It is not all about big data”. The challenge for operational organisations is rather to manage and give meaning to a variety of non-standardised data bases of limited size that are nothing like the big data that certain directors talk about. This is confirmed by what Simon Johnson experienced at GeONG 2018: participants at the conference felt that the impact of blockchain, cryptocurrencies and voice-assistance had been overestimated, while OpenStreetMap, HXL and IATI are little-known ‘heroes’ (who receive very little investment). Pilot tests of all kinds of technology, even the most futuristic, are useful to help the sector move forward, but sometimes detract from current challenges. We regularly see humanitarian organisations’ Innovation Units expanding and being given significant budgets to adapt the latest fashionable technologies, while their operational departments remain unable to use IT tools properly (for example, in terms of mobile data collection, case management, etc.).

Of course, the Famine Action Mechanism – the joint initiative by the ICRC, the United Nations, the World Bank, Microsoft, Amazon and Google – aims to prove the contrary. This project, which was launched in 2018, aims to “help to anticipate famine and periods of food insecurity before these events happen”, by combining artificial intelligence and ‘machine learning’. Though it will be interesting to see the concrete results of this innovative mechanism, we nevertheless know that the poor reactivity of actors to famine is caused by other factors: the lack of operational partners on the ground, access problems and the human factor, all of which cannot be tackled by technology.

 

Conclusion

The link between technology and the agility of aid is therefore not so obvious and raises questions about our practices. Do we really need real-time data at headquarters or regional levels in order to make decisions? This is not really so obvious if the decisions themselves are not made in the right place and the technology appears to only be legitimising this illusion of knowledge and control. Do information systems necessarily produce reliable data? This needs to be qualified in the light of the frequent methodological weaknesses of data collection in the field. Basing our analyses purely on data, without experience or knowledge of the context, is often useless, and the technological veneer can even sometimes legitimise inappropriate data7. Finally, are aid actors really capable of adapting their responses due to feedback from IT tools such as crowdsourcing? Numerous structural weaknesses remain and technology will not be able to solve all of these. It may even continue to mask them if the current trend continues without any serious questioning of approaches and practices.

 

Lisa DAOUD

A graduate of Science-Politiques Grenoble, Lisa Daoud worked for humanitarian NGOs for 6 years both in complex crisis contexts (Ivory Coast, Sudan and Lebanon) and at headquarters before joining Groupe URD. She helped operational staff to improve the quality of their projects (by analysing monitoring and data management systems) and accountability towards crisis-affected people.

Edmond WACH

Having worked for more than 10 years both in the field and as a technical focal point with different NGOs, including SOLIDARITES INTERNATIONAL and Terre des Hommes, Edmond WACH currently works for CartONG where he is the Information Management Project Manager. CartONG provides services and tools in mapping, mobile data collection and information management for the humanitarian and development sectors.

 

Sources:

  1. RedRose offers a unique e-card to manage distribution activities.
  2. Humanitarian Data Exchange
  3. Cf. the report by Christian Aid Ireland and ODI, Learning to make a difference Christian Aid Ireland’s adaptive programme management in governance, gender, peace building and human rights, 2018.
  4. https://hbr.org/2018/10/how-to-make-sure-youre-not-using-data-just-to-justify-decisions-youve-already-made
  5. https://www.urd.org/Evidence-based-decision-making-for,2754
  6. https://www.alnap.org/help-library/operational-humanitarian-decision-making-infosheet
  7. https://en.wikipedia.org/wiki/Garbage_in,_garbage_out

Pagination

p.13-16