Despite many great examples of expanding and replicating good policy programs, it’s difficult to apply universal rules to scaling. This is because the success of many policy projects depends on conditions and circumstances that are often particular to their first runs.
Program officers at the Bernard van Leer Foundation (BvLF) thought about this challenge as they scaled their Criança Feliz program, which provides coaching for parents in vital areas of early childhood development, working with some of Brazil’s lowest-income families.
According to BvLF’s Cecilia Vaca Jones, there are five key pre-conditions that she now looks for when determining whether it’s possible to scale a project.
Early achievements: the success of Criança Feliz
Brazil has 207 million people spread across 27 states and more than 5000 municipalities. So when the Ministry of Social Development, with support from BvLF, set out to reach three million children living under the poverty line in the first 1,000 days of their lives, they had their work cut out for them.
The ministry first had to identify the most vulnerable families in the country; those who would most benefit from specialised coaching as their babies grew. To do this, it teamed up with the Bolsa Família program, a social service that provides cash transfers to the poorest families in Brazil.
Because Bolsa Família already works with low-income families, the Ministry had instant access to a list of vulnerable households with children under three. For Vaca Jones, bundling a new service on top of a cash transfer was one of the most innovative aspects of the project.
The Ministry then began designing its service: home visits from health workers for pregnant women and the youngest children.
The next step was to train a fleet of health workers to visit these families regularly and to promote the best practices for parents to care for their new babies. These include responsive feeding, eye contact, playing, singing, and verbal and non-verbal cues, all of which have been shown to significantly impact a child’s mental and emotional development.
Coordinating this huge training effort required effective organisation at each level of government: federal, state and municipal. At the federal level, training strategies are formulated and dispersed to states. States are then responsible for making sure the strategies are applied across their municipalities, while the municipalities themselves train the home visitors, plan the family visits, and monitor and assess the program as it is implemented.
This decentralised mode of governance allowed the program to scale quickly and effectively. By the end of 2017, home visitors had visited 167,454 children and 22,946 pregnant women. While the program is still relatively new, early evaluations are promising.
The secret ingredients of scaling
In reflecting on the success of Criança Feliz, Vaca Jones identified five key ingredients that made its scaling possible.
While it’s impossible to apply a one-size-fits-all approach to expanding different kinds of programs, the areas listed below may be important indicators for scaling a program in countries with a similar governance model to that of Brazil, or programs that are also looking to “bundle” a new service with an existing one.
1. Political will
“If there is the political will to support a program at scale… all the rest is easier,” said Vaca Jones.
This was one reason for the success of Criança Feliz. The Minister for Social Development, Osmar Terra, was a huge champion of early childhood development as an under-recognised area where the Ministry could reduce poverty and improve quality of life. So committed was Terra to scaling the home visiting program quickly, that he fought to change the structure of the Ministry to hire a full-time, permanent leader to run it.
Vaca Jones explained that political will of powerful ministers is reflected in all kinds of concrete actions, from committing specific budget and personnel to a program, to making sure that its built sustainably and can continue to run.
Even if you’ve got key ministers passionate about a program, they may be hamstrung by institutional governance issues and bureaucracy.
Criança Feliz worked because Brazil has what Vaca Jones called “a minimum structure” (almost like a “minimum viable product”) that allowed its leaders to coordinate quickly across sectors and levels of government.
“BvLF has seen many programs work very well at the national level, but when those programs decentralize or deconcentrate their actions, they become weaker,” Vaca Jones explained. In designing a program for scale, it’s important to consider whether key institutions can implement it beyond its initial scope.
Effective scaling also requires a workforce that is ready and trained. Training goes beyond basic implementation: frontline workers also need the ability to, as Vaca Jones puts it, “scale and adapt, scale and adapt.”
This could entail frontline staff providing detailed feedback about whether the scaling process is working and how it can be improved, helping to ensure that quality remains high as it is replicated in different areas.
Brazil invested significant resources in home visitor training:they trained all the supervisors and frontline workers to a high level. But another challenge comes in the form of how to compensate these workers.
For Criança Feliz, this depends on state and municipality. Some visitors are volunteers, while others are recent university graduates with stipends.
But, as Vaca Jones said, “One of the challenges in any service that’s related to childhood development is how we recognise a fair stipend and fair salary for the people who are dealing with the most vulnerable human beings in the world.”
Cost-effectiveness can get overlooked when scaling: collecting evidence at large scale can be challenging and expensive.
However, Vaca Jones insisted that it’s vitally important to have enough information to cost services, particularly data on how scaling can make the process more cost effective.
In Brazil, the Ministry for Social Development, under budgetary pressure and keen to ensure long-term sustainability, was careful to identify the real administrative and investment costs required to keep improving and growing the program.
5. Data and Monitoring
Integrated data collection is key to scaling a program. Vaca Jones explained that often what’s most effective for an early childhood development program is a specific ID code that allows monitoring of individual kids.
Access to such personal information can be challenging in places like Europe with strict data policies and regulations.
On the other hand, in developing countries, the challenge of obtaining and managing good quality data in the first place still looms large. This creates difficulty for scaling, and for providing a good service in the first place.
While Brazil has strong informational monitoring systems, the challenge has been integrating the systems across different states. Collecting data from all the different municipalities and ensuring interfaces among systems work – and then collating that with the national system, is a big project.
The final word
Vaca Jones maintains that one of the secrets of the Criança Feliz program, and one key to its successful scaling, is the way that it bundled itself with the Bolsa Família cash program.
Often, early childhood programs are supported through the health departments, which seems natural and obvious as an entry point to pregnant women and their babies.
But there are many reasons why thinking outside the box and identifying other services that can be leveraged – like the cash transfer program – might be a productive way to scale. Doing so opens up opportunities for meeting some of the criteria for scaling: if you can’t find political will or good data in the health department, you can try another team who might have compatible aims.
Vaca Jones says that because the theory and methodologies in the early childhood development field can be so complex, program leaders often try to put in as much information as possible into each program design. “Then, it becomes impossible to deal with everything,” she says.
“So we need to come up with simple, cost effective programs that can easily be bundled anywhere. That’s still a challenge.”
– Megan Dent
(Picture credit: Flickr)