Contrasting alternative pathways to scale
Written with Florencia Guerzovich
This is the third post in a post series about how to shape and navigate pathways to scale in social accountability. In post 1, we explained that scale is a complex change process. We questioned and reorganized how practitioners and the literature view scale — a to-do recommended in a recently published synthesis paper from the Action for Empowerment and Accountability (A4EA) programme. In post 2, we introduced a “resonance” pathway to scale, which has been largely overlooked in theory, but resonates with many practitioners we talk to.
In this post, we’ll discuss the theoretical underpinnings of two pathways that are more familiar in the literature: best practice and resistance. We’ll also provide examples to illustrate them.
Pathways have a main thrust — anchored in social learning (resonance), pressure (resistance), or technical expertise and “rigorous” evidence (best practice) — even if they are not “pure types” in the real world. The discussion should help to explain why identifying potentially different pathways matters.
This pathway assumes that change happens chiefly in response to technical expertise and rigorous evidence. Partly as a result, it is less attentive to socio-political factors, such as the challenge of perceived resistance by status quo forces or the potential for social learning among different actors.
For years, many stakeholders’ approach to social accountability scale up, much like in traditional development approaches, largely focused on designing best-practice solutions. This approach to scale up assumes that technical experts who produce knowledge can determine the unique form of social accountability mechanisms that work and then use their authority and knowledge to promote cross-contextual convergence (i.e., scale up) towards those arrangements (Rodrik 2008). Forms of “rigorous” evidence used for this purpose can include transnational measurement and benchmarking and comparisons in different settings to incentivize a race to the top and Randomized Control Trials (RCTs) — the “gold standard (see Duflo and Kremer, 2015; Kremer et al. 2019 for what this looks like).”
The Community Score Card © in Malawi
One example of this comes from CARE Malawi which is the story of government uptake of a community scorecard model in Ntcheu district. Between 2011 and 2015, CARE Malawi implemented the Maternal Health Alliance Project (MHAP). It conducted a cluster-randomized control evaluation which demonstrated impressive results in various health outcomes such as post natal and home visits and an increase in the use of modern family planning. In July 2017, the National Community Health Strategy (2017–2022) was approved and recommended the roll out of enhanced social accountability mechanisms at community level, with explicit mention of scorecards. Yet, the line item of the budget for these mechanisms totalled $0 (p.66 if you want to check). Indeed, the strategy did not reference “The Community Score Card ©,” as such. So, staff one of us spoke to questioned how much influence they may have had in influencing the policy (especially as other organizations also use scorecards in Malawi). Ultimately, the RCT didn’t suffice to convince policy makers to adopt the CARE tool nationally. But, was this a reasonable expectation in the first place?
If one looks at this case searching with a full replication or bust mindset, the story gets cut short unless there is wholesale success. As it turns out, Ntcheu district did partially take up a lighter version of the scorecard model at a reduced scale in 5 (rather than 10) facilities between November 2016 and February 2020. Uptake entailed significant adaptations. These included being run by the District Health Management Team rather than by Community Based Organizations (CBOs) or CARE itself. CARE only providing targeted technical assistance, there were fewer facilities, indicators were streamlined, and there was a requirement to share with zonal and national government stakeholders (so the district team could lobby them for funding, because they didn’t have much). This pivot therefore suggests that perceived opposition from government was low, but appetite for social learning between CARE and the district was reasonably high (even if the district didn’t want CARE’s ideal model).
If you check Ntcheu District’s Development Plan (2017–2022), you’ll find a Transparency and Accountability Project ‘promoting citizen engagement by using scorecards and promoting service delivery by using Public Expenditure Tracking [PETS] and developing Service Charters in the budget.’ This suggests that the district had a wider interest and saw some value in accountability generally. While other NGOs (e.g., World Vision) and funders are mentioned, strangely, CARE is not. So, it’s reasonable to suggest that what further adaptations might be made are likely to be informed by more than just a “best practice” application of a trademarked tool (indeed, despite historical unwillingness to experiment with new social accountability tools, CARE Malawi later promoted PETS themselves).
This form of adaptation might be a bittersweet pill to swallow for those expecting full replication and compliance with plans rather than lots of adaptation, as explained by Callie Simon from Save the Children in a recent webinar. Yet, if one takes a “shades of grey” mindset and considers the fit of resonance with the context, new possibilities potentially arise. In fact, on 11 November 2021, we heard in a webinar that a World Bank operation could be opening a different window of opportunity to scale up an adaptation of certain components of social accountability interventions in Malawi. So, the story of scale up may not be over and exploration could be continued.
However, what has and may go on will be invisible to the eye for those expecting best practice replication and clear-cut results, instead of grappling with tangible, but imperfect, ones. In fact, in a recent Twitter thread, Innovations for Poverty Action — IPA (pre-eminent evangelists for replication) acknowledged the following:
The second tweet is effectively the point of departure for the resonance pathway, rather than an afterthought, and requires as IPA acknowledges, someone to think of a completely different model for understanding what will be the most promising investments over time.
Lant Pritchett’s recent blog on what he learned from Rukmini Banerji also questions the long-held assumption that the (supposed) quality of evidence is always the key, as partners:
“Did not find that having rigorous evidence of an effective action that raised learning in and of itself created a heavily trodden path to their door. In fact, they generally found that the reaction from the mainstream education establishment, both in government and in civil society, was indifference, if not outright hostility… [it] did not fit with the reigning agendas of the time of enrollment expansion and upgrading of the physical school infrastructure (e.g., boundary walls) and quantifiable inputs.”
To appropriate Jane Austen, it is not a truth universally acknowledged, that a policymaker in possession of a good fortune, must be in want of an RCT.
The conceptualization and operationalization of this pathway typically starts with the assumption of an adversarial relationship of resistance between ‘us’ and ‘them’: civil society actors and status quo interests entrenched in the sector and the state (see Fox, 2016). Associated with this is the belief that change happens through productive conflict (resistance, tension, pressure) between both sides, and typically that the primary motor for scale up is civil society agenda setting in spaces created by coalitions of relatively powerless or excluded groups and forcing (or squeezing) status quo interests to change through pressure.
It is this pressure which is deemed to give civil society demands sufficient clout to deter poor behavior and incentivize stakeholder compliance locally and at scale (see Joshi, 2013; Fox, 2016; Anderson, Fox and Gaventa, 2020). While it is not always acknowledged, this is an adversarial relationship, and there are “goodies” and “baddies.” Adversarial “countervailing power” by civil society often through multi-pronged campaigns tend to include some form of “naming and shaming” from campaigners and potentially also including protest and litigation at different administrative levels is increasingly seen as key to success (see Gaventa and McGee, 2010; Joshi, 2013; Fox, 2016; Joshi, 2017).
A resistance approach has several different versions — with various nuances that move certain specifications to the middle rows of the box above — more than we can cover in a single blog post (we will come back to this in future posts). Some versions of a resistance approach pair civic pressure from the outside with government disruptors on the inside. Some exert pressure from outside (e.g., international institutions or foreign countries) to hold national institutions to account, others do not. But, what unites these is the view that pressure and productive conflict is what triggers change. There is some research in motion which we anticipate will clarify these distinctions further.
So, we look forward to that to see more clearly what variations of resistance look like. We’re also hopeful that such research will help to clarify the conditions under which these models may be a good bet, how they unfold, for what purposes, and for how long. The main point here is that while this bet may potentially consider the value of social learning or evidence to a limited degree, this is not the main thrust (as ARC’s diagram shows, pressure rather than persuasion is still the primary force).
Another characteristic of many forms of a resistance approach is that, much like those looking to scale up through a replication of best practice lens, at least until recently, unless there is a clear-cut, headline-grabbing, “unqualified successes,” results aren’t considered to be worth celebrating. Yet, there is much celebration of evocative campaign “wins,” even if these aren’t sustained.
We should remember that Rome wasn’t built in a day. In a complex sector, it is unreasonable to expect homogeneous, generalisable, and rapid success across a wide range of contexts. We have started to see increasing recognition of “incremental,” “transitory” and imperfect progress. This is welcome. The latest A4EA synthesis piece seems to conclude closer to the starting point of the resonance pathway than earlier iterations of the work, albeit with consistent normative caveats:
“Starting from localised analysis and interventions in multiple locations, we can identify commonalities and trends across these localities. This will allow us to focus on issues or solutions that are common to multiple localities and can be addressed at higher levels through issue-based projects. This would also enable issue-based projects working at national or provincial levels to be informed by different ways of working at a local level.”
As much scholarship has focused on inspiring examples of giant leaps, we were surprised (and pleased) to hear so clearly that “small steps matter.” Nonetheless, it’s worth considering some of the trade offs of a big bang approach to scale up.
The tradeoffs of sabotage in Uganda
Pressure being the primary force is well captured in the Human Resources for Health (HRH) Campaign in Uganda (2011–2012), which we’ve discussed before. It’s one of the examples widely used for inspiration. The case demonstrates how civil society actors pivoted from what was labelled a “gentleman’s approach” to advocacy and oversight (more akin to a resonance pathway) to one which the Ugandan Ministry of Health considered an attempt at “sabotage.” According to Jillian Larson’s study, the Ministry of Health ‘felt that it was under attack by CSOs [and that they had] used their position in working groups and committees to steal insider information which was then brought before parliament.’ Following peaceful demonstrations outside the Supreme Court, and “naming and shaming” of government officials for having received VIP healthcare abroad, citizens sent SMS to parliamentarians saying: “We are watching you: Refuse to pass the budget unless it includes the increase you promised.” As mentioned in a previous blog, the confrontational pivot was effective in contributing to the government agreeing to a higher staffing budget.
A window of opportunity opened, but then quickly closed. The HRH campaign’s mobilization and headline-grabbing results have been interpreted by many as a great success. Many large-scale mobilizations raise expectations and, sometimes, create headlines. There are some that survive and flourish, but also many others breakthroughs that have stagnated or disappeared. We pay less attention to the latter.
The outcome is more complex once we move from a black and white world to one of shades of grey. First, as Larson’s interviews revealed, some CSOs saw this success as a ‘pyrrhic victory’ because their role ‘declined as a result of changes carried out by the executive immediately following the campaign’s budget victory,’ with reduced transparency from Cabinet as a result.
Second, the campaign didn’t have sustainable results. GOAL’s Accountability Can Transform Health (ACT Health) program (2016–2018) adopted a relatively similar (but, we believe, softer) campaigning approach to address exactly the same staffing issue in 18 districts and to some degree at national level. This shows that commitments didn’t stick.
Indeed, the HRH campaign created a significant backlash, which appeared to have a long tail for subsequent outcomes, including the allocation and implementation of resources to deliver health results.
In a background study on social accountability and sanctions we found this kind of blowback is by no means unique to Uganda’s HRH Campaign. Nor indeed is this restricted to authoritarian regimes. We found blowback in nearly half of 35 cases we reviewed. Relevant effects included discrediting advocates, threatened loss of funding, reprisals for advocates and whistle-blowers, including court cases and staff being relocated for speaking up, the denial of access to services, an increased concentration of poorly performing staff in some areas, and even threats of violence to MPs and cabinet reshuffles for those who worked with campaigners.
Given the widespread blowback of confrontation, we have some doubts regarding the conditions under which and how much official oversight might credibly advance civil society resistance at scale. It’s interesting to note partly overlapping insights from the SOAS-ACE anti-corruption research program. Mushtaq Khan recently argued that corruption is lower in sector projects when individuals with above average income and the social capital that comes with it have a self-interest and capacities to engage in monitoring people of the same standing — shared networks play a big role in the argument. Khan calls this horizontal or peer to peer monitoring. He underscores its potential over traditional “vertical monitoring” by state accountability institutions and over transparency and voice interventions. He also warns about doing no harm through calls for others to resist, including when their lives may be at risk.
Florencia works in Santa Catarina, Brazil, with multi-stakeholder groups that co-produce accountability. These groups include colleagues from local, state, and national Supreme Audit Institutions (SAIs). Learning by doing and action research informs this work. Yet, we are cautious about generalizing the insights — empirically, there are variations in the possibilities, goals, functions and effectiveness of partnerships between civil society and oversight institutions within Brazil and within organizations, too.
Pathways in context
In this blog, we made the case that it’s important to distinguish pathways to scale. In the next blog post in our series, we’ll discuss four additional concrete contextual factors to move beyond saying that “context matters” to building knowledge on how context shapes decisions in practice.
These two building blocks — multiple pathways and their conditioning factors — are crucial to ensuring that we can make more productive use of funding available for Transparency Participation and Accountability’s (TPA) evidence architecture at field, portfolio, and program levels that bridge the gap between the decisions that stakeholders are supporting, doing, and evaluating TPA. Along with a different way of evidencing scale, multiple pathways and conditioning factors are key to help the field move beyond doom loop narratives to a more mature effort to grow and sustain the results, the work, and the diversity of people and organizations who do it around the world.