The logical framework or logframe continues to reign supreme in the aid sector, not just for ‘development’ interventions, but also for humanitarian action, peace work and human rights and governance interventions.
In principle, that is a good thing: a logframe forces us to think in a disciplined manner about what we want to achieve, how we intend to go about it, and how we assess whether we make progress and achieved what we set out to. Where it goes horribly wrong is when logframes become fixed into contracts and can at best be changed with great difficulty and through a time-consuming chain of authorisations. That was not the original intent: Logframes were developed to encourage disciplined thinking, but not to do it once and then fix it for months or years.
Fixing agreements in contracts is something we do regularly in our everyday lives: we buy a new kitchen, have a garage built next to our house, bring in a caterer for a family gathering, hire an accountant to do our accounts etc. We want contracts that set clear expectations and often detailed specifications, with an agreed price. And we can do that for goods and services where the expectations are well known, and the supplier has control over most factors.
‘Determined design’ plans that then become legal obligations through contracts, are appropriate for situations that David Snowden’s Cynefin framework calls ‘complicated’. There are many elements involved but there are identifiable cause-effect links. You may need quite some expertise to make it happen, but that exists, and the contractor can do the job because s/he has control over most elements and factors that influence implementation. Most, not all: as we have seen in recent weeks, global supply chains, for example in wood, fuel, or microchips, can get disrupted, leading to temporary shortages and significant delays, often beyond original expectations and even the terms of contracts.
Determined design plans and contractually fixed logframes however do not work for situations that have a higher degree of complexity and unpredictability. The private sector talks about a world that is becoming increasingly VUCA: Volatile, Unpredictable, Complex and Ambiguous. There are many factors and many other actors that influence what actually happens, over which the intervener has no control and no or only minimal influence. There is no known pathway of intervention that will guarantee the desired results. We can make assumptions about what we think may work in a given context, but then have to test in against ‘reality’, to find what seems to work and for how long. Based on our learning-by-doing, we will discover we made assumptions that turn out not to have been correct, and will have to adjust and adapt, perhaps modestly, perhaps radically. ‘Staying the course’ as originally planned, when it turns out no longer to be relevant or not to be working, would be a serious mistake and profoundly bad practice. The private sector calls this being ‘agile’, the aid sector talks about ‘adaptive management’.
Yet, though the aid sector talks about adaptive management, in practice donor requirements, - contracts and -behaviours often continue to call for overly detailed ‘determined design’ plans, expect continued monitoring of indicators that may no longer be relevant, and commission evaluations to see whether the grant holder has delivered as per the original proposal - even if the situation has changed completely or learning during implementation signaled that the original plan had to be changed. In the worst case, the grant-holder will only be paid when the promised results are ‘delivered’. That doesn’t work for aid-sponsored interventions in unstable social, political, and security situations, where we are dealing mostly with complex social and political problems, not complicated technical ones.
Fortunately (sic), ‘thanks to’ the COVID-19 pandemic, for the past 20 months national governments around the world, including those providing Official Development Assistance (ODA), have been able to experience the VUCA world, and been forced to practice whole-of-government adaptive management.
We all started out in the ‘chaos’ sphere of the Cynefin framework, when this virus spread much faster and wider than anticipated and started causing large-scale morbidity and mortality in Italy (for the European world). Our governments, confronted with a ‘new’ phenomenon (although pandemics were definitely on the anticipatory risk maps!) did the right thing and tried to find some actions that would have some mitigating influence though not control of the pandemic i.e. you try to shift from a situation of ‘chaos’ to one of ‘complexity’. The first reaction was radical and blunt: full lockdowns. Then we started learning, slowly and by testing against reality, whether and what degree of social distancing and what type of masks could help reduce the infection spread. We struggled to get enough PPE and orders were placed for what sometimes turned out deficient equipment, or equipment that did not arrive as planned because everyone in the world was competing to get hold of it. It took long months to get some more evidence about what protective measures had what level of impact. We did not have confident indicators at the outset, nor uniform data collection and -reporting methods. Meanwhile, research was taking place globally, with heavy public investment, to find a vaccine. In the face of necessity, established protocols and requirements before officially approving a vaccine were shortened. Then our governments struggled to get enough testing and vaccination locations to go to scale. Inevitably, that took longer than everyone wished. Meanwhile, the overload in intensive care units was reduced somewhat, at least in some countries, but many people with other diseases had to wait for their care. Malaria deaths, for example, increased while all medical attention was focused on controlling COVID-19. Still, after a good year, the economically stronger countries had shifted from ‘chaos’, through ‘complexity’ a bit more into the ‘complicated’ realm, at least with regard to controlling the spread of COVID-19. We got a somewhat better grip on relevant indicators, though there remains some dispute about them, and we know there is significant underreporting. Even on the health front, we are not there yet: we introduced vaccines without knowing how long the immunity would last, what possible side effects they may have in the shorter- and the longer-term, whether they were safe for young children and pregnant women, and how well they would do against a mutating virus. Quite some complexity left.
The COVID-19 challenge was not just a public health one though: The lockdowns caused rapid economic havoc. Public expenditure policies, plans, and ceilings had to be thrown out of the window, and all sorts of measures introduced to try and mitigate these economic impacts. The richer countries generally did quite well here, although in some it exposed and exacerbated the socio-economic inequalities; in poorer countries with many people in the informal economy the rise in poverty is staggering. Governments were in unchartered waters also here, and questions about how long, strong and wide these economic safety nets would be cast remained – and remain- uncertain. Fortunately, private charitable initiatives went faster to bridge the gap before government measures could kick in, or for those who fell through the net.
It took some 15 months for some countries, most of them ODA providers, to find a managed, but still fragile, balance between controlling health and economic impacts. But we are by no means out of the ‘complexity zone’: the mental health impacts of the situation have been huge as well; public debt has often increased significantly and will need to be managed over years to come, and a significant resistance to vaccination has come up, for now mostly against COVID-19 vaccines, in future perhaps against vaccination more general. Governments cannot control the mindsets and behaviours of those opponents and struggle to find strategies that have a mix of countering misinformation, persuasion and modest (vaccine passports) or hard pressure (e.g. Italy’s hardline approach). We cannot say with full confidence what is going to work here and by when, we can only try different approaches and find out by doing.
On many occasions, our publics have been asking health ministers and prime ministers for confident statements when the situation would be under control, restrictions be lifted, and life return to normal. In other words, when would the government deliver the ‘results’ we want? When they did make too confident statements, reality repeatedly proved stronger than promise. We know more about the virus and about the vaccines developed so far and are therefore more into the realm of ‘complicated problems’ but there remain significant unknowns there, so we still sit also in the ‘complex problem’ zone.
To come back to our main point: Did our ODA-giving governments, when faced with the COVID-19 pandemic, draw up a detailed determined design plan, underpinned by a logframe, with a heavily earmarked budget (fixed, detailed budget lines), that they then executed faithfully to ‘deliver’ the desired results? No, they did not.
They struggled in the face of many uncertainties, some manifesting themselves early on, other appearing later. Much of this required dilemma management. Dilemma management is a frequent occurrence in situations of acute crisis, violent conflict, and disruption, that logframes do not recognise. These governments had to weigh many considerations against each other. They also had to make assumptions about the relevance and effectiveness of certain approaches, but then test them against reality. They often had to revisit these assumptions and adjust the approaches. Over time, parts of the challenge have moved into the ‘complicated zone’, with some (still emerging) scientific ‘evidence’ underpinning them. But dealing with current and possibly future resistance to vaccination is not a ‘complicated’ problem. Human beings can be even more volatile and unpredictable than a virus. This is a complex social and political problem, not a complicated technical problem. The same holds for aid-sponsored interventions in volatile, usually conflict-affected, environments, with a multitude of actors that no single intervener controls or can even influence.
And that is just for the situation in the richer countries. Look at the health and economic situations globally, and we remain with far more complexity and still quite some chaos.
If anything, this whole struggle to ‘manage’ the COVID-19 pandemic has given ODA-providing governments an unprecedented experience of VUCA situations, and their intrinsic complexity and chaos. There were no known ‘best’ or even ‘good’ practices: they had to learn-by-doing. For various aspects we are still trying to find one or more paths that take us in the desired direction. Had our governments tried to approach this with rigid logframes and detailed and inflexible contracts with their citizens about the results they would ‘deliver’ and by when, they would have discovered, painfully, the futility of this.
In evaluating the performance on their domestic COVID-19 challenges, should we hold ODA-providing governments to account for their initial plan and -statements, in the spring of 2020? No, that would not be reasonable: What we will look at is the quality of their adaptive management, and of their communications about these adaptations with their key stakeholders.
Based on that wider governmental experience, can aid administrations now enter into the necessary conversations about when and where detailed planning and rigid agreements are appropriate, and when and where they are not? The Global Learning on Adaptive Management (GLAM) initiative, initiated by DFID and USAID in 2018, has already given insights about what differentiates smart adaptation (‘adaptive rigour’) from clumsy adaptation, and what monitoring, evaluation and learning mean in practice, in situations that require adaptive management (MEL4AP). Let’s start from there.