Scaling has become a buzzword in international development, where the received wisdom is to ‘scale what works’. However, whilst this is a paradigm that suits private investment in international development and science more broadly, complex problems require nuanced solutions. Robert Mclean, John Gargani and Dena Lomofsky, argue that a new conception of scaling – scaling impact- can better serve the public good.
As researchers and evaluators, we are often met with the common wisdom: “scale what works.” It sounds simple enough. But the most pressing problems in development are wicked – we don’t know what works, what might work depends on context, and context is complex. This is why research and innovation are critical. And why the common wisdom about scaling typically falls short. Unlike in the private sector, and as attractive as it may seem to donors and social enterprise, when it comes to development outcomes, faster, bigger and more is not necessarily better. Instead, we need to shift our focus toward achieving positive impact at optimal scale.
COVID-19 and the limitations of conventional scaling
Our recent experience with COVID-19 is a painful example of the limitations of conventional models for scaling. We know a great deal about viruses, vaccines, and the spread of disease. But when confronted with a global pandemic, the common wisdom failed us:
- we lacked a ready-made solution to scale
- social distancing helped slow the disease spread, but it hurt culture and commerce, making it impossible to scale the good without the bad
- the crisis severity was affected by local context, like health system capacity, and that context is then impacted by the crisis
- there are trade-offs to be made – for example, between public health and economic return – and we often require a justified balance between the two.
With COVID-19, we cannot simply ‘scale what works’. And when research and innovation eventually produce vaccines and other solutions, it seems unlikely we will be able to ‘simply’ scale them up either. The complexity is too great. The same is true for ‘development’ most of the time. In high, middle, and low-income countries alike.
when the goal is the public good, scaling up, out or deep is only valuable if it leads to a commensurate change in positive impact
The conventional conceptions of ‘scaling’
Common conceptions of scaling are scaling up (bigger), out (more) or deep (quality).
Scaling up increases output. For example, a health education program might train more nurses, and a policy research organization might promote the implementation of a new policy with evidence.
Scaling out expands or replicates production. The health education program might open new training sites. The research organization might promote the same policy in new places.
Of course, you might pursue more than one approach at a time, scaling up and out.
While these strategies may be of great value to research and development professionals, alone they are insufficient for creating and sustaining ‘good change’.
Shifting the focus to scaling impact
Much of the scaling up, out or deep logic comes from the successful experience of the private sector, where scaling is equivalent to growth, expansion and control. And indeed, if the goal is private return, these may be worthy aims.
But when the goal is the public good, scaling up, out or deep is only valuable if it leads to a commensurate change in positive impact.
Scaling Science is a new approach to scaling that embraces the mindset and methods of science to scale impact for the public good, by focusing on impacts over actions.
Four guiding principles for scaling impact
Following a review of over 200 research projects across the Global South, we suggest four guiding principles for scaling impact (see the full argument in this open access book). The principles suggest a starting place. They require critique, testing, and improvement. Our hope is that they spark a more critical approach to scaling impact – embedded in, and supported by, science.
Justification: Scaling is not an imperative, it’s a choice justified by evidence and values. Scientific evidence helps us understand whether an innovation can scale. Our shared values help us understand whether it should scale. Justification starts with the questions, “Why scale?” and, “Who decides?”
Optimal scale: Optimal scale balances the magnitude, variety, equity, and sustainability of impacts in a way that stakeholders endorse. There are typically trade-offs among these dimensions, and stakeholders may not agree on how to balance them.
Coordination: Scaling takes place in complex systems and requires the participation of an evolving set of actors. Their efforts may be cooperative, competitive, or complementary, and their roles may change over time. Consequently, coordination demands a high level of planning, adaptation, and flexibility, along with a deep understanding of the system we place our innovations into.
Dynamic evaluation: Scaling is an intervention (we change actions to improve impacts) and it can be evaluated. Scaling produces dynamic change (it affects systems that in turn affect it) so it requires dynamic evaluation. Historically, impact evaluations have been concerned with how, why, for whom, and under what conditions actions produce impact. Dynamic evaluation goes a step further. It is concerned with how, why, for whom, and under what conditions changing actions changes impact.
more policies, programs, products and technologies will not make a better world. Research must inform these interventions and how they optimize impacts
Pathways to impact at scale
Scaling Science, understood as a justified and coordinated focus on impact at optimal scale (rather than just bigger or more), may already be a part of your work – even if you’re not calling it such. And you don’t have to be a vaccine scientist for this to be the case. In this table, we illustrate how scaling matters for a variety of purposes.
But remember, more policies, programs, products and technologies will not make a better world. Research must inform these interventions and how they optimize impacts.
| Pathway | Scaling is when research is used to… | For example, … |
| Policy | … inform a new policy for public good, or perhaps influence the replication, adaptation or extension of the policy into new jurisdictions to amplify its impact. | … a research program uses evidence from one country that successfully implemented a tax on sugary drinks, to inform policy in another country to achieve a similar impact on public health. |
| Program | … design a program, improve an existing program’s quality, change the program to fit a new context, or form partnerships with others to improve overall impact. | … evidence from a successful national centre of science excellence is used to establish similar centres in other countries within the region. |
| Behaviour, practice or skill | … design behaviour or practice change interventions – such as awareness campaigns, or evidence-based guidelines – and study the roll-out of these interventions for people and organizations. | … sharing results with communities of successful early antenatal visits encourages more pregnant women to visit a health care facility in first three months of pregnancy. |
| Product or technology | … produce new goods and services, make existing products/technologies more accessible, or optimize the value-chain underpinning a good or service like a fertilizer, software, vaccine, or internet platform. | … agricultural researchers work with farmers to develop a more nutritious variety of potato, and farming cooperatives help build the markets and supply chains to reach consumers equitably. |
| Methodology | … develop, re-orient, or otherwise optimize a way of knowing and/or doing that will generate social impact. | … users of a novel participatory research approach share the method with researchers in a neighbouring discipline, and thereby expand the benefits of stakeholder inclusion to a new field of practice. |
Advancing Scaling Science
The research supported by IDRC provides a starting point by offering guiding principles, pathways, and case studies for scaling impact.
However, for a science of scaling to develop, we need to work together. We need those with greatest knowledge across disciplines, geographies and objectives, and those most impacted to co-create in these settings.
We invite you to join in this effort, to share your ideas, experiences and lessons. We want to build on this work and unlock its full potential. We want to scale its impact.
The Scaling Science approach comes from an exploration of research and innovation across the Global South, supported by the International Development Research Centre (IDRC). It is currently being used/tested as a conceptual framework in an external evaluation of IDRC’s strategy to scale research results. Join the conversation about scaling impact on Twitter @ScalingScience.
Note: This article gives the views of the authors, and not the position of the LSE Impact Blog, nor of the London School of Economics. Please review our comments policy if you have any concerns on posting a comment below
Photo credit: JOSHUA COLEMAN on Unsplash
Great and very necessary article! Great how you articulate what a lot of people feel/sense when working on development projects. What is next?
In our own field, the political ecology of access to natural resources, scale is an essential research tool. This means research needs to look ‘up and outward’ to recognise higher-order scales and their effect on local issues. Local level soil degradation or deforestation might, conventional scientific research commonly reveals, be caused directly by the actions of land use managers or peasant farmers. But the agrarian household economy sits within a ‘chain of explanation’ [Piers Blaikie] of regional national and international forces that scientists and indeed development practitioners are much more uncomfortable scrutinising. Understand these better help to trace the cause of a resource management issue up to ultimate causes – in our case, commonly the actions of international capitalist agribusiness companies, who cajole governments into allowing them land grabs in the Global South, which disrupt local livelihood systems, dispossessing farmers of traditional land and therefore forcing them onto marginal land where erosion and fertility loss is harder to contain. This would be a classic ‘reproduction squeeze ‘ [Bernstein] or ‘accumulation by dispossession’ [Harvey] .
In the case of a peasant movement to fight the above corporation [a movement or protest is a potentially successful development intervention or response], such a protest ‘jumps scales’ when it becomes a regional, national or international movement. Its effects also have to be considered at multiple scales.
In brief, there is no utility in considering scales as part of “scaling impact” for development interventions, without first researching the political economy in its scaled dimensions. What Pawson et al call, rather poorly, “context” in realist analysis. Otherwise all efforts to “up scale” anything will be doomed to fail – imagine IDRC trying to support land reform to empower farmers to conserve soils and forests and sustain livelihoods, only to find it is blocked by regional government officials who are corrupt, because they are poorly paid by the state, which is dysfunctional because of an ethnic dispute or conflict, worsened by colonial borders and international corporate actions to obtain land. Just a notional example, but where to start? Surely at higher order scales, and with the injustices there, not locally.
This is a great article. I like the fact that it emphasizes the need to use evidence from research to make decisions on how to scale impact but with a greater emphasis on looking at context.
The second point that I found intirguing but true is the one about ‘more policies, programs, products and technologies will not make a better world’. I think this is very true particularly in global south where countries have great policies and they keep formulating more yet very little improvements in the lives of their populations.